Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Development and validation of a disease specific quality of life questionnaire for potentially curable… Clifton, Joanne Caroline 2003

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2003-0354.pdf [ 15.11MB ]
Metadata
JSON: 831-1.0091183.json
JSON-LD: 831-1.0091183-ld.json
RDF/XML (Pretty): 831-1.0091183-rdf.xml
RDF/JSON: 831-1.0091183-rdf.json
Turtle: 831-1.0091183-turtle.txt
N-Triples: 831-1.0091183-rdf-ntriples.txt
Original Record: 831-1.0091183-source.json
Full Text
831-1.0091183-fulltext.txt
Citation
831-1.0091183.ris

Full Text

D E V E L O P M E N T A N D V A L I D A T I O N OF A DISEASE SPECIFIC Q U A L I T Y OF LIFE QUESTIONNAIRE FOR P O T E N T I A L L Y C U R A B L E PATIENTS WITH C A R C I N O M A OF THE ESOPHAGUS by JOANNE CAROLINE CLIFTON  A THESIS SUBMITTED IN PARTIAL F U L F I L M E N T OF THE REQUIREMENTS FOR THE D E G R E E OF M A S T E R OF SCIENCE IN THE D E P A R T M E N T OF H E A L T H C A R E A N D EPIDEMIOLOGY  We accept this thesis as conforming To the required standard  THE UNIVERSITY OF BRITISH C O L U M B I A August 2003 © Joanne Caroline Clifton, 2003  I  http://www.library.ubc.ca/spcoll/thesauth.html  UBC Rare Books and Special Collections - Thesis Authorisation Form  j  -  In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r an advanced degree at- the U n i v e r s i t y o f B r i t i s h Columbia, I agree t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r a g r e e . t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y purposes may be g r a n t e d by t h e head o f my department o'r 'by h i s o r h e r r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . 1  Department o f UlZfrOTU  C/\-£g  /  &Plh£frsnL-Q(^ / C  The U n i v e r s i t y o f B r i t i s h Columbia Vancouver, Canada Date  1 of 1  8/25/2003 7:35 P M  ABSTRACT Objectives: To develop, pre-test and validate a disease specific quality of life ( Q O L ) questionnaire ( E Q O L ) for potentially curable patients with esophageal carcinoma, for use with the E O R T C Q L Q - C 3 0 questionnaire, assessing treatment associated quality of life. Methods: Questionnaire development Patients enrolled in three centres. Literature reviews, patients, family members, and health care professionals generated 195 items symptoms (55); emotions (53); physical functioning (17); activities of daily living ( A D L ) (48); leisure/social (22). 38 patients identified items of importance and assigned importance ratings on a five point Likert scale. Impact scores were calculated as frequency times mean item importance. Item impact scores less than 20/100 were excluded. Pearson's Correlation Co-efficients compared domains with the M O S SF-20. Fifteen items remained in the E Q O L . Questionnaire validation Sixty-five patients enrolled in four centres. E O R T C Q L Q - C 3 0 , E Q O L , M O S SF-36 and a Global Rating of Change Questionnaire were completed at baseline, one week, one month, three months, and six months after treatment began. Reliability was assessed comparing mean baseline and one week scores. Responsiveness was assessed between mean scores of changed and unchanged patients, and a responsiveness index was calculated. The M O S SF-36 was used for criterion validity. Construct validity included four a priori predictions. Results: Mean scores between baseline and one week for unchanged patients were not significant (p>0.05), except in the physical function domain (p=0.025). Symptom, physical function and social domains were responsive to change at all time intervals (p<0.05). Emotional function was responsive at one and three months, A D L ' s at one and six months. Magnitude of change was significant when direction of change was stated. Between better and worse, magnitude of change was significant in all domains except at six months in symptoms, emotional and physical domains. The minimal clinically important difference was consistently around 0.5 for all domains. M i n i m a l , moderate and large effect ranges were established. Only two out of 16 time intervals had poor correlations with the M O S SF-36, establishing criterion validity. O f the four a priori predications for construct validity, only the second part of one prediction, in the emotional function domain, was not confirmed. Conclusions: Quality of life can be assessed using the E Q O L . Future studies can further test the E Q O L where validation was not confirmed.  ii  TABLE OF CONTENTS  ABSTRACT  ii  T A B L E OF CONTENTS  iii  LIST OF T A B L E S  ix  LIST OF FIGURES  xi  ACKNOWLEDGEMENTS  xiii  1. I N R O D U C T I O N  1  1.1 I N T R O D U C T I O N T O C H A P T E R  1  1.1.1 Carcinoma of the Esophagus  1  1.1.2 W h y Study Carcinoma O f The Esophagus?  2  1.1.3 Anatomy and Physiology of the Esophagus  2  1.1.4 Etiology and Pathogenesis  2  1.1.5 Symptoms and Complications  3  1.1.6 Diagnosis of Esophageal Carcinoma  3  1.1.7 Treatment of Esophageal Carcinoma  4  1.1.7.1 Surgical Resection  4  1.1.7.2 Radiotherapy  5  1.1.7.3 Chemotherapy  7  1.1.7.4 Multi-Modality Therapy  7  1.1.7.5 Treatment Summary  8  1.1.8 Effect O f Esophageal Cancer On Quality O f Life 1.2 Q U A L I T Y O F L I F E  8 9  1.2.1 History  9  1.2.2 Definition  10  1.2.3 W h y Measure Quality O f Life?  10  1.3 Q U A L I T Y O F L I F E I N S T R U M E N T S 1.3.1 Generic quality of life instruments  12 12  1.3.1.1 Overview  12  1.3.1.2 The Short Form-36  13  iii  1.3.2 Specific Quality O f L i f e Instruments 1.3.2.1 The E O R T C Q L Q - C 3 0 1.4 D I S E A S E - S P E C I F I C Q U E S T I O N N A I R E D E V E L O P M E N T  13 14 15  1.4.1 Item Generation  16  1.4.2 Item Reduction  16  1.4.3 Pretesting  18  1.5 P S Y C H O M E T R I C S  18  1.5.1 Reliability  18  1.5.1.1 Internal Consistency  19  1.5.1.2 Test-Retest Reliability  19  1.5.2 Responsiveness  19  1.5.3 Validity  20  1.5.3.1 Face Validity  20  1.5.3.2 Content Validity  21  1.5.3.3 Construct Validity  21  1.5.3.4 Criterion validity  22  1.6 C A R C I N O M A A N D Q U A L I T Y O F L I F E  22  1.7 E S O P H A G E A L C A R C I N O M A A N D Q U A L I T Y O F L I F E  23  1.8 S U M M A R Y  24  2. METHODS  26  2.1 O V E R V I E W O F S T U D Y  26  2.1.1 Ethics Approval  26  2.2 M E T H O D S - I N S T R U M E N T D E V E L O P M E N T P H A S E  26  2.2.1 Objectives  26  2.2.2 Choice of Instruments  26  2.2.2.1 E O R T C Q L Q C - 3 0  26  2.2.2.2 M O S SF-20  27  2.2.3 Patient Population  27  2.2.4 Item Generation  28  2.2.5 Item Reduction  28  2.2.5.1 Inclusion Criteria  30  iv  2.2.5.2 Patient Sample  30  2.2.5.3 Questionnaire Administration  31  2.2.5.4 Statistical Methods  32  2.2.6 Pre-Testing  ,  2.3 M E T H O D S - V A L I D A T I O N P H A S E  33 34  2.3.1 Objectives - Validation Phase  34  2.3.2 Choice of instruments  34  2.3.2.1 E O R T C Q L Q c - 3 0  34  2.3.2.2 M O S SF-36  34  2.3.2.3 Global Rating of Change Questionnaire  35  2.3.3 Time sequence  35  2.3.3.1 Permission To Use The Instruments  35  2.3.3.2 Orientation of Study Personnel  35  2.3.3.3 Preparation of the Questionnaire Study Pack  35  2.3.4 Patient Population  36  2.3.5 Patient Sample  36  2.3.6 Questionnaire Administration  37  2.3.7 Statistical Methods  38  2.3.7.1 Demographics  38  2.3.7.2 Global Rating of Change  38  2.3.7.3 E O R T C - Q L Q 3 0 Scoring  38  2.3.7.4 E Q O L Scoring  41  2.3.7.5 M O S SF-36 Scoring  42  2.3.7.6 Reliability Analysis  44  2.3.7.7 Responsiveness Analysis  45  2.3.7.8 Face and Content Validity  48  2.3.7.9 Criterion Validity  49  2.3.7.10 Construct Validity  50  3. RESULTS  53  3.1 D E V E L O P M E N T P H A S E R E S U L T S 3.1.1 Item Reduction Results  53 53  3.1.2 Pre-Testing Results  58  3.2 V A L I D A T I O N P H A S E R E S U L T S  58  3.2.1 Study Participation  58  3.2.2 Demographics  59  3.2.3 Reliability  61  3.2.4 Responsiveness Analysis  62  3.2.4.1 Part I - Change in Group B  62  3.2.4.2 Part II - Change between Group A and Group B  65  3.2.4.3 Part III - Responsiveness Index  68  3.2.5 Face and Content Validity  68  3.2.6 Criterion Validity  70  3.2.7 Construct Validity  71  3.2.7.1 Dysphagia  71  3.2.7.2 Chemoradiotherapy  74  3.2.7.3 Stage of Disease  77 '  3.2.7.4 Symptom Rating  95  3.3 S U M M A R Y  105  3.3.1 Reliability  105  3.3.2 Responsiveness  106  3.3.2.1 Part I - Change in Patients W h o State Change  106  3.3.2.2 Part II - Magnitude of Change  107  3.3.2.3 Part III - Responsiveness Index  107  3.3.3 Criterion Validity  108  3.3.4 Construct Validity  108  4. DISCUSSION A N D C O N C L U S I O N S  Ill  4.1 I N T R O D U C T I O N  Ill  4.2 P R E - T E S T I N G  Ill  4.3 R E L I A B I L I T Y  Ill  4.4 R E S P O N S I V E N E S S  112  4.4.1 Responsivenss to Change When Patients State Change  112  4.4.2 Magnitude of Change Between Changed and Stable  113  vi  4.4.3 Responsiveness Index  114  4.5 F A C E A N D C O N T E N T V A L I D I T Y  114  4.6 C R I T E R I O N V A L I D I T Y  114  4.7 C O N S T R U C T V A L I D I T Y  115  4.8 S T U D Y P O P U L A T I O N S  116  4.8.1 Gender  116  4.8.2 Employment Status  117  4.8.3 Treatment Groups  117  4.9 O T H E R M E A S U R E S O F Q U A L I T Y O F L I F E  118  4.10 Q U A L I T Y O F L I F E A N D E S O P H A G E A L C A R C I N O M A  119  4.11 U S E O F T H E E Q O L  120  4.12 L I M I T A T I O N S  121  4.12.1 Patient Numbers  121  4.12.2 Rate of Dropout  122  4.12.3 Confounding Problem  122  4.13 C O N C L U S I O N S  122  125  5. R E F E R E N C E S  APPENDIX I  Consent Form (Questionnaire Development)  131  A P P E N D I X II  Health Professionals Input Questionnaire/Interview  133  A P P E N D I X III  Family Input Questionnaire/Interview  134  A P P E N D I X IV  Initial Patient Input Questionnaire/Interview  135  APPENDLX V  Item Reduction Demographics Form  136  APPENDIX VI  Item Reduction Phase Interview  138  A P P E N D I X VII  Item Reduction Response Sheet (195 Items)  141  A P P E N D I X VIII  M O S SF-20 Short-Form Health Survey  146  A P P E N D F X DC  Consent Form (Questionnaire Validation)  149  APPENDIX X  Validation Demographics Form  151  APPENDIX XI  Self-Administered Questionnaire Interview  153  A P P E N D I X XII  E Q O L Questionnaire (15 Items)  154  vn  A P P E N D I X XIII  E O R T C Q L Q - C 3 0 Questionnaire  158  APPENDIX XIV  M O S SF-26 Health Survey  160  APPENDIX X V  Global Rating of Change Questionnaire  166  viii  LIST OF TABLES Table 1 Scoring the E O R T C Q L Q - C 3 0  41  Table 2 Scoring the E Q O L  43  Table 3 Global Change Group  45  Table 4 Demographics - Item Reduction Phase  54  Table 5 Top Five Impact Scores by Treatment Group  55  Table 6 Twenty-Two Items and Their Impact Scores  56  Table 7 Number of Patients Completing Study  59  Table 8 Demographics - Validation Phase  60  Table 9 Reliability Results  61  Table 10 Change in Symptom Domain - Group B  62  Table 11 Change in Emotional Domain - Group B  63  Table 12 Change in Physical Function Domain - Group B  64  Table 13 Change in Activities of Daily L i v i n g Domain - Group B  64  Table 14 Change in Social Domain - Group B  65  Table 15 Symptom Domain - Mean Differences Between Groups  67  Table 16 Responsiveness Index Results  69  Table 17 Criterion Validity - Correlations Over Time  70  Table 18 Number of Domain Questions in E Q O L and SF-36  71  Table 19 Mean Change in Dysphagia Score by Treatment Group  73  Table 20 Physical Function Change Baseline to One Month by Treatment Group...75 Table 21 Mean Change in Physical Function Scores by Treatment Group  75  Table 22 Emotional Function Change at One Month by Treatment Group  76  Table 23 Mean Change in Emotional Function Scores by Treatment Group  77  Table 24 Symptom Change Baseline to Three Months Early V s Late Stage  79  Table 25 Symptom Change A t Three Months Early V s Late Stage  80  Table 26 Symptom Change Baseline to Six Months Early V s Late Stage  80  Table 27 Symptom Change at Six Months Early V s Late Stage  82  Table 28 Physical Function Change Baseline to 3 Months Early V s Late Stage  83  Table 29 Physical Function Change at Three Months Early V s Late Stage  83  Table 30 Physical Function Change Baseline to Six Months Early V s Late Stage...85  ix  Table 31 Physical Function Change at Six Months Early V s Late Stage  85  Table 32 Emotional Function Change Baseline to 3 Months Early V s Late Stage ...86 Table 33 Emotional Function Change at Three Months Early V s Late Stage  87  Table 34 Emotional Function Change Baseline to 6 Months Early V s Late Stage ...88 Table 35 Emotional Function Change at Six Months Early V s Late Stage  89  Table 36 Activities of Daily L i v i n g Change Baseline to Three Months B y Stage....89 Table 37 Activities of Daily L i v i n g Change at Three Months Early V s Late Stage .90 Table 38 Activities of Daily L i v i n g Change Baseline to Six Months B y Stage  91  Table 39 Activities of Daily L i v i n g Change at Three Months Early V s Late Stage .92 Table 40 Social Function Change Baseline to Three Months B y Stage  93  Table 41 Social Function Change at Three Months Early V s Late Stage  94  Table 42 Social Function Change Baseline to Six Months B y Stage  94  Table 43 Social Function Change A t Six Months Early V s Late Stage  95  Table 44 Baseline to 3 Month Physical Function Scores B y Symptom Severity  96  Table 45 Baseline to 3 Month Emotional Function Scores B y Symptom Severity...98 Table 46 Baseline to 3 Month A D L ' s Scores B y Symptom Severity  100  Table 47 Baseline to 3 Month Social Function Scores B y Symptom Severity  102  Table 48 Baseline to 3 Month Symptom Scores B y Symptom Severity  103  LIST OF FIGURES Figure 1 Survival A l l Esophageal Cancer  6  Figure 2 Survival Esophageal Cancer B y Radiotherapy  7  Figure 3 Pre-Testing Form  33  Figure 4 Patient Sample B y Study Centre  36  Figure 5 Patient Sample by Treatment Group  37  Figure 6 Validation Phase - Patient Time Line  40  Figure 7 E Q O L Response Options  43  Figure 8 Testing the Reliability, Responsiveness and Validity of the E Q O L  46  Figure 9 Dysphagia Frequency at Baseline for 65 Patients  72  Figure 10 Dysphagia Frequency at Baseline and One Month for 47 Patients  73  Figure 11 Stem and Leaf Plot Physical Function at One Month .  74  Figure 12 Stem and Leaf Plot Emotional Function at One Month  76  Figure 13 Stage of Disease  78  Figure 14 Stem and Leaf Plot Symptoms at Three Months Early V s Late Stage  79  Figure 15 Stem and Leaf Plot Symptoms at Six Months Early V s Late Stage  81  Figure 16 Stem and Leaf Plot Physical Function at 3 Months Early V s Late Stage .82 Figure 17 Stem and Leaf Plot Physical Function at 6 Months Early V s Late Stage .84 Figure 18 Stem and Leaf Plot Emotional Function 3 Months Early V s Late Stage ..87 Figure 19 Stem and Leaf Plot Emotional Function 6 Months Early V s Late Stage ..88 Figure 20 Stem and Leaf Plot A D L ' s at Three Months Early V s Late Stage  90  Figure 21 Stem and Leaf Plot A D L ' s at Six Months Early V s Late Stage  92  Figure 22 Stem and Leaf Plot Social Domain at 3 Months Early V s Late Stage  93  Figure 23 Stem and Leaf Plot Social Function at 6 Months Early V s Late Stage  95  Figure 24 Stem and Leaf Plot Physical Function Baseline by Symptom Severity ....97 Figure 25 Stem and Leaf Plot Physical Function 3 Months by Symptom Severity ..97 Figure 26 Stem and Leaf Plot Emotional Function Baseline by Symptom Severity.98 Figure 27 Stem and Leaf Plot Emotional Function 3 Months Symptom Severity ....99 Figure 28 Stem and Leaf Plot A D L ' s Baseline Scores B y Symptom Severity  100  Figure 29 Stem and Leaf Plot A D L ' s 3 Month Scores B y Symptom Severity  101  Figure 30 Stem and Leaf Plot Social Function Baseline B y Symptom Severity  102  xi  Figure 31 Stem and Leaf Plot Social Function 3 Months B y Symptom Severity  103  Figure 32 Stem and Leaf Plot Symptom Baseline Scores B y Symptom Severity  104  Figure 33 Stem and Leaf Plot Symptom 3 Month Scores B y Symptom Severity  105  ACKNOWLEDGEMENTS First, I would like to thank my supervisory committee - Dr. Joel Singer, Dr. Chris Lovato and Dr. Richard Finley - for their invaluable help and support throughout this project. I would like to pay particular thanks to Dr. Finley who was involved from the beginning and who's encouragement, patience, support and advice, not only in this project, but in numerous others, has encouraged me to continually challenge myself and to successfully meet those challenges. M a n y thanks to Dr. Richard Finley, Dr. Gary Gelfand, Dr. Jan L i m , Dr. Richard Inculet, Dr. Richard Malthaner, Dr. Lawrence Tarn and Dr. A n d y Graham for their time and effort in recruiting patients for this study. Thanks also to Dr. Ruth Milner for her assistance with the development phase of the questionnaire. Many thanks also to all the research assistants/nurses who interviewed study patients and collected data in the various centres - Sheena Majewski in Vancouver, Wendy Valstar and Jane Gardner in Calgary, Cindy Dudar in Winnipeg, Deb Fenlon and Alexandra Martiniuk in London. I would also like to extend my thanks to all the patients who took time out of their daily lives to participate in this study. I am very grateful to the British Columbia Lung Association for their funding support of the development phase of this questionnaire and the National Cancer Institute of Canada for their funding of the validation phase of this study. Finally, I would like to thank my family and friends who encouraged me to continue and never doubted my ability to complete this project. In particular, I would like to thank my Mother and Father from whom I draw inspiration to overcome hurdles and who have taught me, by example, to be strong and determined; and to my two sons Michael and Jeffrey who constantly inspire me by their very existence, and for whom I hope to set an example of achievement and positive self-worth.  xiii  CHAPTER ONE - INTRODUCTION  1.1 I N T R O D U C T I O N T O C H A P T E R The objective of Chapter One is to lay the groundwork for why we chose to study this disease by providing a brief background to the disease itself - carcinoma of the esophagus -  including anatomy, physiology, etiology, pathogenesis and treatment.  Following that, some background on quality of life including history, definition, why we measure quality of life, the types of instruments available.  Next the phases of  questionnaire development w i l l be presented - item generation, item reduction, pretesting and validation. Psychometric properties w i l l be discussed. Finally, a brief review of carcinoma and quality of life are described and more specifically, a brief review of esophageal carcinoma and quality of life.  1.1.1  C a r c i n o m a O f The Esophagus  Esophageal carcinoma is one of the most difficult solid organ tumors to cure or palliate. The two major histological types are squamous cell carcinoma and adenocarcinoma of the esophagus or cardia. While only about 1600 new cases occur annually in Canada, a 1  study published in 1991, found the incidence of adenocarcinoma of the esophagus and gastroesophageal junction in North America increasing faster than any other cancer.  2  The same authors, in 1998, found that the incidence of adenocarcinoma of the esophagus had risen over 350% in white males since the mid 1970's . A t least 1,450 deaths due to 3  esophageal cancer occurred in Canada in 2001. The incidence of this disease in females 4  also rose, but remained much lower than among males. The five-year survival rate has been estimated between 10% and 2 0 % .  5-18  Poor survival is influenced by a number of  factors: i) advanced stage of disease upon presentation often resulting in only palliative therapy (66%)  19  ii) non-resectable tumors with distant metastases (20%)  20  ii) frequent  lymph node involvement (74%) iii) submucosal spread and extension of the tumor into 19  surrounding structures (54%)  21  There is little controversy regarding management of  palliative patients. However, those who are potentially curable present a challenge, both for the patient and physician in the management of their disease.  1  1.1.2  Why study carcinoma of the esophagus?  Patients suffering from esophageal carcinoma have poor survival and associated quality of life ( Q O L ) .  Patients have great difficulty swallowing solid foods and many have  difficulty swallowing liquids.  A s a result, patients lose weight and energy and find it  difficult to function normally in their day to day life. A s well, the esophageal tumor may penetrate surrounding organs and cause pain. Patients who undergo treatment find that the treatment itself also impacts on their quality of life. They must now deal with a fear of the cancer, a fear of its treatment and ultimately a fear of dying.  Different forms of  treatment exist for this disease - each having their own associated impact on patients quality of life. In order to understand this impact fully, we must quantify the effect of the various treatment options on the patient's quality of life.  1.1.3  Anatomy and Physiology of the Esophagus  The esophagus is a tube located behind the trachea which is approximately 10 to 16 inches long and forms part of the digestive tract, normally lined with squamous epithelium.  Its major function is to transport food from the mouth to the stomach. Food  is propelled toward the stomach during the process of swallowing in which the circular muscles of the esophagus squeeze in an orderly fashion. A s this process occurs, gland like cells produce mucus which acts as a lubricant and facilitates the transport of food through the esophagus. The valve at the junction of the esophagus and stomach stops the contents of the stomach, which are acidic, from backing up into the esophagus and damaging the lining of the esophagus.  A c i d reflux occurs when the valve isn't  performing properly and can produce damage to the esophagus resulting in columnar metaplasia, dysplasia and eventually cancer.  1.1.4  Etiology and Pathogenesis  The two most common types of primary carcinoma of the esophagus are adenocarcinoma and squamous cell carcinoma.  In addition to primary carcinoma of the esophagus,  metastases may occur in the esophagus from other primary cancers such as melanoma, breast or lung cancer.  Cells close to the outside of the body are known as squamous  2  cells, hence squamous cell carcinoma is indicative of a cancerous growth in this area of the esophagus. A cancer of the glandular cells, closer to the stomach, is adenocarcinoma.  Etiology of the disease is poorly understood, however the presence of risk factors such as smoking and heavy alcohol intake and diet appear to play a significant role in squamous cell carcinoma.  Risk factors for adenocarcinoma include: i) Barrett's esophagus in  which there exists abnormal cells in the esophagus which may be precancerous; ii) achalasia which is characterized by inefficient contractions of muscles in the esophagus; iii) gastroesophageal reflux in which frequent irritation of the esophagus can trigger cell changes which can lead to cancer; iv) cigarette smoking; and v) obesity.  There also  appears to be a correlation with age - incidence of the disease is fairly uncommon up to the age of 55, after which it increases dramatically.  1.1.5  Symptoms and complications  This disease usually presents late and the patient may already be suffering from local or distant metastases. Symptoms, due to the primary tumor, include dysphagia, weight loss, bleeding, heartburn and regurgitation.  Dysphagia, inability to swallow, is the most  common presenting symptom and is often an indicator of advanced disease in many cases,  22  hence the medical profession is often faced with the challenge of treating patients  with advanced tumors.  1.1.6  Diagnosis of esophageal carcinoma  A n initial assessment of the extent of disease is usually provided by a barium swallow, often followed by an esophagoscopy. Biopsies and brushings of all abnormal lesions are performed allowing histopathologic examination to secure the diagnosis.  Additional  investigations to rule out the presence of distant metastatic disease include further imaging studies by chest x-ray, C T scan of the chest and abdomen, and ultrasound. Finally, a surgical biopsy is performed and cell type and stage of the cancer is determined.  3  Staging illustrates the progress of the disease in an individual by assessing: i)  tumor size and location - whether the tumor is contained in the lining of the esophagus, has moved, extended into the esophageal wall, advanced through the esophageal wall or has begun to invade nearby tissues  ii)  nodal involvement - whether the cancer has or has not extended to nearby lymph nodes  iii)  distant metastases  This process allows the physician to determine the best course of treatment.  1.1.7  Treatment of esophageal carcinoma  The purpose of treating esophageal carcinoma is cure and improvement of symptoms and quality of life. radiotherapy.  Single  modality treatment includes  Multi-modality  treatments  include  surgery, chemotherapy, and  neo-adjuvant  radiotherapy  or  chemoradiotherapy followed by surgical resection, surgical resection followed by radiotherapy, and a combination of chemotherapy and radiotherapy as primary treatment. Advances in multi-modality treatment have resulted in some improved outcomes, but five-year survival remains at less than 2 0 % " . The chosen treatment modality w i l l depend largely on clinical stage, patient status and local practices.  Surgical treatment, either  alone or with adjuvant therapy, remains the preferred treatment for potentially curable esophageal carcinoma.  1.1.7.1 Surgical resection There is little controversy regarding management of palliative patients. However, those who are potentially curable present a challenge both to the patient and the physician in management of the disease. Although radiation therapy, chemotherapy, laser therapy and combination therapy have been used to treat localized esophageal cancer, surgical resection (esophagectomy) remains the standard approaches exist in performing an esophagectomy. i)  treatment.  Three major surgical  23  Trans-hiatal esophagectomy with the cervical esophagus anastomosed to a via a narrow gastric tube in the left neck.  4  ii)  The second surgical approach is through a right posterolateral thoracotomy. The stomach is mobilized through a laparotomy and anastomosed to the cervical esophagus through a left neck incision.  iii)  The  third  approach  is  a left  thoracoabdominal  esophagectomy  with  anastomosis of the distal stomach or small bowel to the esophagus in the left chest. In all approaches, either a pyloromyotomy or pyloroplasty is carried out.  A feeding  jejunostomy is constructed and inserted to provide nutrition during the perioperative period. Patients are treated with intravenous fluids and nasogastric drainage for one week after the surgery. Patients with a feeding jejunostomy are started on full-strength tube feeding on the first postoperative day.  Usually around the time of the seventh  postoperative day, a contrast radiography of the upper gastrointestinal tract is obtained to examine gastric emptying and the integrity of the esophagogastric anastomosis.  Complications of surgical resection can include hospital death, bleeding, anastomotic leak, aspiration, dysphagia, early satiety, delayed gastric emptying, pneumonia, reflux, heartburn, dumping syndrome, diarrhea, and hoarseness.  The survival rate for esophageal carcinoma after treatment remains dismal. The five-year survival following surgical resection has been estimated around 20%. In 1992, a five year survival rate of 20% or greater was achieved in a limited series of esophageal cancer 13  patients undergoing transhiatal esophagectomy.  23  Finley, et al ~, following their 1995  study, showed an overall five-year survival rate of 20% for patients undergoing either surgical resection alone (n=249) or pre-operative radiation therapy followed by surgical resection (n= 121). (Figure 1) 1.1.7.2 Radiotherapy Radiotherapy can be administered pre-operatively followed by surgical resection, postoperatively, alone, or in combination with chemotherapy regimes. regimens vary in the treatment of this disease.  Radiotherapy  Six randomized trials ' "' ' " 6  8  9  14  16  of  preoperative radiotherapy versus surgery alone, used low to moderate doses ranging from  5  20 Gy in 10 fractions to 53 Gy in 20 fractions. The six randomized studies above showed no significant survival benefit with the addition of adjuvant radiotherapy. Some studies assessing adjuvant radiotherapy have even shown five-year survival rates of only 10%. " 24  2 7  The schedule for administration of treatment is usually between one to four weeks  prior to surgery. The schedule used for esophageal carcinoma would be intraluminal brachytherapy 1,500 cGy over 30 minutes and external beam radiotherapy 4,000 cGy over four weeks.  Possible side effects include skin sensitivities.  Figure 1  Survival All Esophageal Cancer n=370  m  0  a  12  •  24  a  36  •  48  •  60  •  72  •  84  1  96  Survival Months  The associated five-year survival with radiotherapy does not differ significantly from that of surgical resection alone as seen in Figure 2 which shows the five-year survival demonstrated by Finley, et al in their previous study"'. Survival curves were generated using the Kaplan-Meier method and comparison of the curves using the log rank test for significance. Five-year survival for those receiving adjuvant radiotherapy was 25% and for those receiving surgery along 17%.  The difference in survival curves was not  significant.  6  Figure 2 Survival E s o p h a g e a l C a n c e r By Radiotherapy (n=370)  c  0.0  J 0  + 12  24  36  48  60  72  84  no XRT (n=249) censored  96  Survival Months  1.1.7.3 Chemotherapy Chemotherapy alone has been administered prior to surgical treatment.  Various drugs,  dosages and regimens for administration exist depending on whether the chemotherapy is the primary treatment for the disease or if it is being administered in combination with radiotherapy or surgical intervention.  Depending on the drug being used, various  toxicities and side effects can exist ranging from nausea and vomiting to some less common effects such as hypersensitivity reactions, severe myalgias and neuropathies, optic neuritis, elevated liver enzymes, C N S effects, and angina.  Again, survival is not significantly improved with the use of chemotherapy. 27  randomized studies assessing preoperative chemotherapy versus surgery alone" " two reported a significant survival benefit with the use of chemotherapy.  In six 32  only  The pooled  results of the studies showed no significant difference. 1.1.7.4 M u l t i - M o d a l i t y T h e r a p y A number of multi-modality therapies exist.  The most common in Canada are a  combination of chemotherapy and radiotherapy as primary treatment. Even though it is seen as primary treatment, often the patient w i l l go on to surgical resection at a later date.  7  Another common multi-modality treatment is neo-adjuvant radiotherapy followed by surgical resection or post-op radiotherapy.  1.1.7.5 Treatment summary Various treatments for esophageal cancer exist in Canada. The three most common are: i)  Surgical resection alone  ii)  Neo-adjuvant radiotherapy followed by surgical resection or post-operative radiotherapy  iii)  chemoradiotherapy as primary treatment  Strong evidence points to replacing the use of single modality chemotherapy or radiation therapy with a combination of the two. However, studies have consistently failed in showing any significant improvement in five-year survival for patients with potentially curable esophageal carcinoma. W e are confronted with a disease which has symptoms that are difficult to control and in which current treatments have modest efficacy and significant morbidity and mortality. Faced with this situation, quality of life must assume considerable importance.  1.1.8 Effect of esophageal cancer on quality of life The main symptom of this disease is progressive dysphagia. Some patients who present are only able to swallow liquids and consequently have suffered significant weight loss and malnutrition.  Complications and i l l effects of treatment have been documented.  Very little research has been conducted on the impact of this disease on a patient's quality of life, yet we know that symptoms of the disease, and the effects of treatment do affect patients during their day to day life, at work, and at rest. Given that survival does not change significantly with present treatment modalities and given the fact that for many of the patients in this population, cure is not possible - the quality of their remaining life is critical.  Choice of therapy might best be decided by assessing the  quality of life associated with each treatment modality.  8  1.2  QUALITY OF LIFE  1.2.1  History  Quality of life (qol) measurement has become, over the past decade, a useful tool in clinical trials.  Results are often important indicators of health status and patients'  perception of how their lives have been affected on a day-to-day basis,  increasingly,  quality of life measurement has a role to play in determining the best treatment options as well as in clinical practice decisions.  However, until recently the success of treatment of an illness was measured by outcomes such as death, morbidity or cure. Very little effort was given to assessing the affect of treatment on patients lives - quality of life. The first formal scale was developed in 1947 and is still in use today - the Karnofsky Performance Scale.  33  The appearance of this  scale coincided with the W o r l d Health Organization defining, in 1948, health as a state of complete physical, mental and social well being, not merely the absence of disease or infirmity.  34  One of the first actual qol measures was developed by Priestman and B a u m  35  in 1976. This measure was used to assess the effect of chemotherapy on patients with advanced breast cancer.  It was recognized that for cancer, limiting studies to survival  data was missing an important element of assessment of the disease and its treatment by not differentiating between treatment interventions.  This important development was  followed, in 1977, with the recognition of qol as a scientific concept by Index M e d i c u s ,  36  and shortly after that, by the development of a health profile questionnaire which could 37  be used to evaluate qol on varied populations - The Sickness Impact Profile.  The  38  development of utility measures based on economic policy followed.  A number of qol  measures for chronic conditions such as hypertension , rheumatic disease , 39  lung disease , and inflammatory bowel disease 41  42  40  chronic  appeared in the mid 1980's.  The  development of new questionnaires for assessment of quality of life continue to appear today.  9  1.2.2  Definition  The concept of quality of life would appear to be rather simple, yet there is no universally accepted definition of the term. A s previously stated, the W H O defined health as "a state of complete physical, mental and social well being, not merely the absence of disease or infirmity"  34  which forms the basis of any definition of quality of life, yet agreement has  not been reached. Schipper and his colleagues suggested in 1990 that "quality of life represents the functional effect of an illness and its consequent therapy upon a patient, as perceived by the patient" . 43  Spilker, in his attempt at defining the concept  44  identified  four major areas or domains that must be considered when assessing qol: physical status and functional abilities; psychological status and well-being; social interactions; and economic status. Later, in 1994, G i l l and Feinstein  45  defined qol as "a reflection of the  way patients perceive and react to their health status and to other, non medical aspects of their lives, rather than only being a description of patients health status".  Therefore, qol is a subjective concept that is influenced by the physical, emotional and social aspects of a patient's life.  It can also be influenced by factors unrelated to the  disease process or treatment, such as education, environment, income.  It is a  multifactorial concept, which means that from the patient's perspective, it represents the final common pathway of all the physiological, psychological and social influences of the entire therapeutic process.  1.2.3  W h y measure quality of life?  Quality of life measurement can be useful in a number of ways: assessing the impact of chronic disease, or assessing efficacy and effectiveness of therapeutic interventions such as new drugs or surgical operations.  Direct physiological measures have always been  important indicators for clinicians in assessing evidence of efficacy of a surgical intervention  or drug.  These  measures  include  hospitalization, and rate of disease recurrence.  reduction in  mortality,  rate  of  However, by relying solely on these  indicators, the development of an important phenomenon w i l l be missed - specifically that two patients with the same clinical criteria often have dramatically responses to treatment.  different  Physiological or biomedical outcomes do not address all  10  important aspects of an intervention. Ability to function normally, freedom from pain, and other physical, psychological and social symptoms will be missed if no quality of life assessment is done. For example, we may see longer survival with a given treatment, but is it at the expense of the patient in terms of unacceptable pain and suffering?  On the  other hand, we may see easing of symptoms but with no measurable change in the physiological functioning. W e may not be addressing the issues of most importance to the individual patient - whether it be decrease of symptoms, ability to care for oneself or ability to perform pleasurable activities, such as gardening, long walks, etc. In order to use this information as effectively as possible, we must assess the patient's quality of life.  Quality of life assessment also allows the patient to become involved in their own treatment process.  Interestingly, a study performed by Nelson in 1989  4 6  assessing  outcomes of medical services, found that patients liked completing questionnaires because they felt that they were providing important information to their doctor. Clinicians and patients may differ greatly over their perceptions of a patient's quality of life, hence the addition of this information can only work towards an improvement in the understanding and judgement of health professionals.  Additionally, quality of life measurement can provide useful information in terms of cost to the health care system by exploring use and expenditure patterns. Integration of qol measures with economic analysis has become increasingly important in defining optimal strategies for disease management.  When all costs and quality of life are taken into  consideration, the cheapest drug is not necessarily the most cost effective.  Finally, patients, clinicians, and health care administrators all have an interest in the wide aspects of effects of medical interventions. Through the information gathered by quality of life measurements we can provide the most complete assessment of the usefulness of an intervention to all those involved.  11  1.3  QUALITY OF LIFE INSTRUMENTS  In order to scientifically assess quality of life, we must have the best tools available. In this case, questionnaires are used. M a n y quality of life questionnaires exist in various formats including both health related and non-health related attributes. In clinical trials, more focus is placed on health related quality of life ( H R Q O L ) which allows for refinement and application to a clinical setting. These tools have been grouped into two major categories: generic and specific.  Generic quality of life instruments  1.3.1  1.3.1.1 Overview Generic questionnaires sample a wide range of areas and are designed to be applicable in many populations. They evaluate function, pain and discomfort, mobility, activities of daily living and physical, emotional and cognitive function. The advantage of generic questionnaires is that they can facilitate the comparison of different disease groups by comparing the effects of interventions across the different diseases. However, a major disadvantage to using generic instruments is that aspects particular to a disease may be missed along with small, but clinically important changes in the patient's quality of life. McHorney  47  presents a time line over the past 35 years of the evolution of generic health  measures. W i t h the announcement of the W H O ' s definition of health there emerged a number of clinically based, global rating scales going beyond organ function to the assessment of human function.  The Karnofsky  performance status was the first to  attempt to better understand treatment effectiveness. Single item indicators were initiated in the 1950's followed by the development of measurements of qol emerging from the social indicators movement of the 1960's. Unified indexes of mortality and morbidity also emerged during this time. A n abundance of generic instruments were developed in the 1970's including the Quality of Well-Being Scale , the Sickness Impact Profile 4 8  the McMaster Health Index Questionnaire , and the Nottingham Health Profile . 50  51  4 9  ,  All  were found to be reproducible, internally consistent and sensitive to change when compared with other measures of disability.  12  However, some of these measures are limited by their length, often taking up to 40 minutes to complete and only available in an interviewer format.  Shorter generic 52  instruments were developed including the Psychological Well-Being Index , the General Health Rating Index  53  and the M O S Short Form-3 6 . 54  1.3.1.2 The Short Form-36 The Short Form-36 (SF-36) and the Nottingham Health Profile appeared to be the most commonly used general health, quality of life, questionnaire. The SF-36 came out of the Medical Outcomes Study and is a shortened version of the original measure. Prior to the SF-36 there was a shorter version - t h e S F - 2 0 . T h e S F - 1 2 55  56  is an evener shorter version  derived from the SF-36. However, the SF-36 provides a higher level of precision with more scale response levels. The SF-36 has eight health domains: physical functioning (10 items); role limitations due to physical and emotional problems (4 items); pain (2 items); general health perceptions (5 items), vitality (4 items), social functioning (2 items), emotional well being (5 items) and change in health (1 item).  Within each  dimension the items are scored, coded, summed and transformed to a scale from 0 (poor health) to  100 (perfect health) . The SF-36 can either be interviewer or  administered.  57  self-  It is relatively brief and can be completed in approximately 5 to 10  minutes. Responses are rated on 3, 5, or 6 point scales rather than the Yes/No responses available in the Nottingham Health Profile. A statistical package is available providing ease of computing scores. The questionnaire includes both positive and negative aspects of disease, has a high response rate and good test-retest reliability.  It has been well  validated in a number of patient populations and languages including England, Australia, New Zealand, Sweden and Germany.  1.3.2 Specific quality of life instruments Specific quality of life questionnaires focus on areas of importance to specific aspects of health, whether it is a specific population, a disease, a certain function, or a problem.  A  specific instrument w i l l focus on areas that might be missed with only a generic instrument. Disease-specific instruments have been developed for many conditions and have proven to be useful in clinical trials. The advantages to using a disease-specific  13  instrument is that they increase responsiveness. The disadvantage is that, because they include only dimensions relevant to the disease being measured, they cannot be compared across disease groups, hence their generalizability is threatened.  A great number of  disease-specific questionnaires exist, yet none so far for potentially curable carcinoma of the esophagus.  1.3.2.1 The E O R T C Q L Q C - 3 0 A variety of cancer specific instruments have been developed. Some problems include reporting  bias,  low  reliability  coefficients,  length  administered, processing time and ceiling effect.  of  the  instrument,  physician  The European Organization for  Research on Treatment of Cancer ( E O R T C ) initiated a research program in 1986 to develop an integrated, modular approach for evaluating the quality of life of patients in international clinical trials in oncology . The intent was to generate a core questionnaire 58  which incorporates a range of physical, emotional and social health issues relevant to a broad range of cancer patients regardless of their specific diagnosis.  This core  questionnaire would be supplemented with diagnosis-specific and/or treatment specific questionnaire modules.  The E O R T C decided on the modular approach to ensure a  sufficient degree of generalizability for cross-study comparisons and a level of specificity adequate for addressing questions of particular relevance in a given clinical trial . 59  A first questionnaire consisting of 36 items was designed in 1987 to be cancer specific, multidimensional in structure, self-administered, and applicable across several cultural settings.  The instrument was field tested  current 30 items.  60  and the questionnaire was revised to the  The E O R T C Q L Q C - 3 0 has nine multi-item scales: five functional  scales (physical, role, cognitive, emotional and social); three symptom scales (fatigue, pain, nausea & vomiting);, and a global health and quality of life scale. Several single item symptom measures which are commonly reported by cancer patients such as dyspnea, appetite loss, sleep disturbance, constipation and diarrhea are also included. The questionnaire also assesses the perceived impact of the disease and treatment on the patient's financial status.  14  The questionnaire was validated in a multi-centered study of 305 non-resectable lung cancer patients in Western Europe, North America, Australia and Japan.  A tumor  specific module for lung cancer consisting of nine items was used to supplement the QLQ-30. The questionnaire was administered twice - once following diagnosis but prior to treatment and once during the period of treatment,  with 28 days being the average  time between administrations. Average time for completion of the questionnaire was 11 minutes.  The mean item-scale correlations across all nine scales was .53 for the pre-  treatment administration and .59 for the on-treatment administration. The authors state that 192 tests of item-discriminate validity were conducted, with success noted in 96% of the cases for both pre-treatment and on-treatment.  A l l inter-scale correlations for both  pre-treatment and on-treatment were statistically significant (p<0.01).  The strongest  correlations were observed between physical functioning, role functioning and fatigue scales. Eight of the nine multi-item scales met the minimal standards set for reliability before and/or during treatment. Evidence of item convergent validity was defined by the authors as a correlation of .40 or greater between an item and its scale. The functional and symptom scales were able to distinguish between patients differing in terms of performance status, weight loss, and treatment toxicity. Statistically significant changes in functional and symptom levels were observed for patients whose performance status had either improved or deteriorated from the pre-treatment assessment to the ontreatment assessment, showing a responsiveness to change. A limitation of the study is that it was conducted solely with lung cancer patients, however clinical validity was established. The E O R T C also conducted a study to validate the questionnaire for use in patients with breast cancer, colorectal cancer and prostate cancer with the development of disease specific modules for those patient populations.  1.4 D I S E A S E - S P E C I F I C Q U E S T I O N N A I R E D E V E L O P M E N T A disease specific quality of life questionnaire should have certain characteristics.  It  should cover the major domains of qol, be patient generated, self-administered in its final format and take less than 20 minutes for the patient to complete. It should also be valid, responsive to change, and reliable. Guyatt, et al  6 1  '  6 2  have developed a widely accepted  15  methodology for this type of questionnaire development. There are three distinct phases to questionnaire development: item generation, item reduction and validation.  1.4.1 Item generation It is important to begin with a pool of items which encompasses all the ways in which a patient's life has been affected by the disease being studied. Interviews with particularly articulate patients is key in the process since we are interested in assessing the quality of life of the patient - a concept which is very subjective and w i l l often be different, even among patients with the same clinical severity of disease.  The aim is to capture the  individual's assessment of the impact of their disease on the quality of their own life. Interviews with family members may shed some light on the situation, but it is the impression of the patients themselves that is placed in the highest regard. Interviews with health professionals, care givers and family members are useful in generating items of importance to be included in the phase of the questionnaire development. A review of the existing literature and a review of previously established generic instruments and instruments designed for closely related diseases, may also be useful in generating items. But the main emphasis should come from the patients themselves.  1.4.2 Item reduction In the second phase of questionnaire development - the item reduction phase - patients with the disease are asked whether or not an item has affected them over a specified period of time and then asked to rate the importance of that item using a Likert type scale. Once all patients have been interviewed and rated each item that affects them, an impact score is calculated for each item. The impact score is calculated by multiplying the item frequency (the number of patients selecting the item) by the mean item importance. Impact Scores = item frequency X mean item importance Another method of reducing items is factor analysis. In this method, items that have high correlations with one another are grouped together and items which are not strongly associated with one of the domains or factors, generated by the factor analysis, are excluded from the final questionnaire. analysis in detail  63  Juniper, et al has discussed the use of factor  and is reluctant to use it for item reduction of health related quality of  16  life questionnaires. A disadvantage to factor analysis is that particular items which may be excluded by the analysis may be of great importance to patients with the disease. Since the main objective is to establish the impact of the disease on the patient, one could be missing a very important aspect by deciding not to include items which are excluded by the factor analysis.  Yet, we cannot simply rely on the impact scores, either.  Clinical judgement plays an  important role in deciding which items w i l l be included in the final questionnaire. Item responsiveness, the ability to detect change if it is indeed present, is another aspect which must be considered. For example,  there may be an item which has no possibility of  changing over time due to an intervention or because of the natural history of the disease. In an evaluative instrument, the item would be excluded since it would only add to the measurement noise and not provide any insight into the qol of the patient population. Here is where clinical judgement must play a role.  The same item may be deemed  important by the patients completing the questionnaire in this phase of the study and one must consider whether it may be a future target for therapy in deciding whether to include the item.  In addition to the impact scores and clinical judgement one must also consider: excluding items which cannot demonstrate change over time with the interventions that are currently available; inclusion of items related to the specific goals of the intervention; inclusion of an adequate number of items for each domain to be measured; and exclusion of redundancies.  Once it has been decided which items to include, they should be  grouped into domains and item responses w i l l be developed. Usually, in an evaluative instrument, Likert type scales are used - either five or seven point scales. One is equal to no impairment and seven (or five) is equal to total impairment. Visual analogue scales may also be used but are more difficult to administer and not as easy to interpret as Likert scales. ' 64  6 5  It is important to be very specific in defining the time period of evaluation.  Periods of time w i l l depend on the given disease and the type of intervention taking place, but must also consider the limitations of the patient's memory. T w o weeks is a time frame which is often used and appears to have good recall. Asking patients to recall  17  how they have been feeling over a period of time greater than two weeks could cause difficulties.  1.4.3 Pretesting Once the questionnaire has been constructed and the response options developed, the instrument should be pre-tested on a small group of patients to ensure that the questions are understandable in language and in concept and that the appropriate wording is used. Patients involved in the pre-testing should represent disease severity, educational background, age and gender of the actual patients who would be involved in a clinical study. The questionnaire w i l l then be modified to eliminate ambiguities, delete offensive questions, and ensure that a full range of response options are used. Pre-testing should be repeated until no more changes are necessary.  1.5 P S Y C H O M E T R I C S The Concise Oxford dictionary defines psycho metrics as "the science of measuring mental capacities and processes". the techniques  The actual term was originated by Galton in 1883 and  and models of analysis he used form the basis of present day  psychometrics . Present day psychometric criteria for evaluating health items, indexes, 66  and scales refer primarily to assessments of validity, responsiveness and reliability.  67  1.5.1 Reliability Reliability is a measurement property which is crucial to discriminative and predictive instruments allowing measurement of differences between patients at a given point in time. A n evaluative instrument must be able to assess clinically important change within an individual patient. Reliability is the psychometric property that determines the signal 68  to noise ratio by detecting if there is small variation between replicate measures. Therefore, reliability is the first step in providing evidence of the value of an instrument by demonstrating the instrument's ability to measure in a reproducible fashion.  69  It is the  ratio of the variability between individual to the total variability and is expressed as a number between 0 (no reliability) and 1 (perfect reliability).  For example, if an  instrument, administered under similar conditions, obtained a different weight each time  18  it is administered, the coefficient would be near 0. Acceptable values of reliability are debatable but generally a reliability of 0.80 is deemed adequate for basic research.  69  1.5.1.1 Internal consistency Internal consistency is a function o f the number of items and their co-variation within a scale measuring a particular construct.  It is used primarily in constructing and  70  evaluating summary scales and examines the inconsistency or non-equivalence of different questions intended to measure the same concept. The main procedures used to assess internal consistency include corrected item-total correlation, split-half and Cronbach's alpha reliability coefficients.  67  These tests are used to estimate the extent to  which these items express the same basic attitude. Cronbach's alpha  7 1  is the coefficient  most often used today and is calculated using a one way repeated measure analysis of variance model with items functioning as the repeated measure.  The simplest way to  compute Cronbach's alpha is to use the correlation values between each item and the number of items in the domain.  1.5.1.2 Test-retest reliability Test-retest reliability is the relationship between scores obtained by the same person on two or more separate occasions.  It is generally accepted that a reliability of 0.90 is  required for comparisons between individuals, but for group comparisons, the 0.50 - 0.70 range is acceptable.  72  Problems are associated with test-retest reliability are the number  of changes that can arise between the dates of administration of the questionnaire, such as the conditions of the administration of the questionnaire (i.e. different interviewers), factors affecting patients in their daily lives, length of time between questionnaire administration (i.e. recall memory) or a change i n the patient's health status (i.e. disease progression). Therefore the questionnaires should be completed with only a brief period of time between them.  1.5.2 Responsiveness The responsiveness of an instrument is its ability to detect clinically important change when it truly exists, even i f that change is small. Therefore it,is critical to include items  19  that are sensitive to change.  In order to determine responsiveness, the questionnaire is  administered to a group of patients who mirror the characteristics of patients who would be included in any planned clinical trial with a particular intervention. The inclusion and exclusion criteria should be the same.  The questionnaire is administered prior to  treatment to an intervention, if possible, of known efficacy and at a specified interval post intervention.  It may include more than one post intervention  application.  If the  instrument is responsive, it w i l l indicate an improvement in quality of life. The larger the difference in questionnaire scores in patients where there is real change, the greater the responsiveness. Three strategies have generally been used to assess responsiveness.  A  paired t-test can be used to compare baseline and follow-up scores to determine if patients have truly changed their health status. In order to determine if the instrument is able to distinguish between those patients who have changed and those who have remained stable, an unpaired r-test can be used to determine if the magnitude of change in the instrument score differs between stable subjects and those whose health related quality of life has changed. Finally, a responsiveness index can be calculated from the minimal important difference and the pooled within-subject standard deviation of change in patients who remain stable and those who change.  Calculating the responsiveness  index is extremely useful as it can be used to calculate sample sizes needed for clinical studies.  1.5.3 V a l i d i t y The ability of an instrument to measure what it is intended to measure is known as validity. Generally, validity is classified in four categories: face validity, content validity, construct validity and criterion validity.  1.5.3.1 Face validity Face validity refers to the acceptability of the questionnaire items to both the patient completing the questionnaire and the person administering the questionnaire.  Some  considerations must be taken into account - the questionnaire must first be geared toward the correct intellectual level of the patient and at the same time be acceptable to the clinicians involved.  Feinstein  has described face validity as the application of  20  enlightened common sense, which is a mixture of ordinary common sense plus a reasonable knowledge of pathophysiology and clinical reality.  If the questionnaire is  aimed at the wrong audience or constructed in an unsatisfactory manner, it w i l l lack clinical sensibility and w i l l be inadequate regardless of the amount of statistical information obtained from its use.  1.5.3.2 Content validity Content validity is closely related to face validity.  It is a judgement of whether the  questionnaire contains all the relevant, important content or domains.  In assessing  content validity, patient population and item sufficiency should be addressed.  Ideally,  face and content validity should be based on some form of review by a panel of experts.  1.5.3.3 Construct validity Construct validity examines how a questionnaire relates to other measures of disease, consistent with the hypothetical concepts (or constructs) being measured. It tests whether an hypothesized association between the questionnaire measure and a measure of the same concept or a different concept is concerned. This is best achieved by making a prior predictions about the results of the questionnaire. For example, depending on the disease being studied, one might predict that correlations exist between certain domains in the questionnaire such as physical functioning, mobility and pain. Another example would be to predict whether an association exists between a physical test documenting stage of disease, and the symptom domain of the questionnaire. In the development of a disease specific questionnaire for asthma, the investigators made predictions about the correlation of within subject changes in quality of life scores with other indices of asthma function.  74  Convergent and discriminant validity are aspects of construct validity.  Convergent  validity refers to the extent to which measures of the same concept correlate with each other. Discriminant validity refers to the extent to which measures do not correlate with other measures that are intended to be different.  21  1.5.3.4 C r i t e r i o n validity Criterion validity is the correlation of the instrument with some other measure of the disease, a criterion or gold standard.  In most circumstances a gold standard for  measuring quality of life does not exist. In this case, another well validated quality of life instrument - usually generic - is often used.  1.6 C A R C I N O M A A N D Q U A L I T Y O F L I F E A number of specific instruments exist which are of interest to those dealing with cancer. The V i s i c k Scale  75  was designed for patients undergoing gastric surgery. Physician  administered, it is reliant on the impression of the interviewer, which may not reflect the patient's true feelings. Since it was not developed for esophageal carcinoma, it is unlikely assess those areas of most importance to patients with esophageal cancer, thus its use in this setting is quite limited. The Karnofsky Performance Index, mentioned earlier, is one of the oldest and best known cancer specific instruments  76  . This instrument is  administered by an observer and assesses symptoms, physical activity, self-care and ability to work.. Modifications have been made to the scale to make it more cancer specific. These modifications include the World Health Organization Scale and the 77  70  Eastern Cancer Oncology Group ( E C O G ) Performance Status  While this instrument is  easy to administer and score, it is not ideal. Again, it is a physician recorded scale and as such is open to bias and variability and has been found to have low reliability coefficients. The Quality of Life Index 79  80  is a widely used physician administered  instrument. It examines five areas: activities of daily living; occupation; perception of health; support and outlook on life. It is easily administered and its psychometric properties have been well established. However, it has moderate correlation with patients' ratings and fails to sample many of the important areas which are relevant to clinical investigations.  81  The Functional L i v i n g Index-Cancer  82  was intended for use in  assessing patients with a wide variety of cancer diagnoses. It is well validated and selfadministered, but it lacks specificity which limits its use in clinical trials. The Visual Analogue Scales use lines of a standard length (usually 10cm) with endpoints that represent extremes of a particular variable. For example, one end of the line would represent no pain, whereas the opposite end of the line would represent the worst possible  22  pain. Patients mark any point on the line which they feel corresponds to their condition. Linear Analogue Self-Assessment instruments utilize this same approach in order to 83  quantify patient responses and has been used in studies assessing breast cancer.  While  both of the latter two instruments are simple and easy to administer, they require more time to score and all suffer from the ceiling effect. As described earlier, the E O R T C Q L Q - 3 0 is a self-administered modular system. The base questionnaire includes questions of concern to cancer patients, regardless of the type of cancer diagnosis. The advantage of this questionnaire is its known psychometric properties and its ability to adapt to specific clinical settings. 1.7 E S O P H A G E A L C A R C I N O M A A N D Q U A L I T Y O F L I F E 84  A literature review was conducted by Gelfand and Finley  in 1994 to begin to assess the  extent to which quality of life has been assessed in studies on esophageal carcinoma. A Medline search resulted in 7569 papers identified on esophageal neoplasms. O f these, 3941 addressed therapy and only 44 considered quality of life. A n additional seven papers were identified from the references. Once the non-English publications were excluded, there remained only 37 papers addressing quality of life and esophageal cancer. The authors reviewed the remaining papers in detail using a system to classify the calibre of the papers. 1. Excellent articles were those that used well documented quality of life instruments with known psychometric properties appropriate for the patient population and with appropriate statistical testing. 2. G o o d articles used instruments with some documented psychometric properties, but clearly not state of the art, and lacked proper statistical evaluation. 3. F a i r articles were those that used quality of life instruments without know psychometric properties. 4. P o o r articles considered the use of quality of life to be important but failed to use any validated instrument to assess it. Only seven papers were deemed either excellent or good. O f the three excellent papers, none used instruments specific to esophageal cancer and the dysphagia scale used was not standardized. A l l three studies assessed only palliative patients. The four good  23  papers did use cancer questionnaires but none specific to esophageal cancer and attempts to measure quality of life were not ideal. The poor group of articles only mentioned quality of life but did not make any attempt to assess it.  It is clear from the extensive review by Gelfand that formal quality of life assessment for esophageal carcinoma has received minimal attention. There is a definite need for the development of a new instrument to assess this patient population. Given the quantity and calibre of instruments available and the effort and expense required to develop a new one from scratch, we decided that the best option was to modify  an existing  questionnaire.  1.8 S U M M A R Y • Esophageal carcinoma is a devastating disease with approximately 1600 new cases diagnosed annually in Canada. • Dysphagia is the most common symptom which often presents late and may indicate local or distant metastases. • Single and multi-modality treatments are available including surgical resection, chemotherapy, radiotherapy or a combination, for example chemoradiotherapy  or  radiotherapy followed by surgical resection. • Five year survival is not significantly different between the various treatment modalities and is estimated between 10% and 20%. • Quality of life is a subjective concept influenced by the physical, emotional and social aspects of a patient's life. It can be influenced by factors unrelated to the disease process or treatment, such as education, environment and income. • Quality of life questionnaires were first developed in the 1970's and continue today. • Generic quality of life questionnaires sample a wide range of areas and are designed to be applicable in many populations. One advantage is their ability to compare different disease groups. • Specific quality of life questionnaires focus on areas of importance to a particular population, disease, a certain function, or a particular problem.  A specific instrument  focuses on areas that might be missed with a general quality of life questionnaire.  24  Questionnaire development includes three phases: i) Item Generation in which information is gathered from patient input, health experts input and literature reviews. ii) Item Reduction in which patient interviews determine importance ratings and impact scores for items affecting quality of life. iv) Pre-testing to ensure that questions are understandable in concept and that appropriate wording has been used. Questionnaire validation includes: i)  Reliability is an instruments ability to measure differences between patients at a given point in time.  It is determined by administering a  questionnaire at determined intervals on stable population to see i f it w i l l produce the same results. ii)  Responsiveness is an instrument's ability to detect change when change exists, to measure the magnitude of change and to determine the minimal clinically important difference for a patient population.  iii)  Validity is the ability of the instrument to measure what it is intended to measure and includes face, content, criterion and construct validity.  Currently, no disease specific questionnaire exists for measuring esophageal carcinoma.  25  CHAPTER TWO - METHODS  2.1 OVERVIEW OF STUDY A disease specific quality of life questionnaire should have certain characteristics. It should cover the major domains of Q O L , be patient generated, self-administered in its final format and take less than 20 minutes for the patient to complete. It should be reliable, responsive to change and valid.  There are three distinct phases to questionnaire development: item generation, item reduction and pre-testing. This chapter w i l l discuss the development of our disease specific questionnaire - the Esophageal Cancer Quality of Life Questionnaire ( E Q O L ) .  2.1.1 Ethics Approval A n application containing the study protocol, consent form and data forms was submitted to the Clinical Ethics Committee at the University of British Columbia and to the Vancouver Hospital and Health Sciences Centre. Full approval of the protocol was obtained for both the development and the validation phases of this study.  2.2 METHODS - INSTRUMENT DEVELOPMENT PHASE  2.2.1 Objectives The objective of this phase of the study was to first develop the instrument using rigorous scientific methodology. This development phase included item generation, item reduction and pre-testing.  2.2.2 Choice of Instruments  2.2.2.1 EORTC QLQC-30 The E O R T C Q L Q C - 3 0 has been described in Chapter One. This instrument was chosen given its excellent psychometric properties and the experience of using it in other cancer 26  populations. Written approval was obtained to use the E O R T C Q L Q C - 3 0 in developing a module for esophageal cancer, to be used in conjunction with the E O R T C Q L Q C - 3 0 .  2.2.2.2 M O S SF-20 The M O S SF-20 is a shorter version of the M O S SF-36. W e anticipated that the amount of time needed for the patient interviews would be approximately 30 minutes in order to adequately identify items from the list of 195 items and to then assign importance ratings to each of the identified items. The M O S SF-20 has proven psychometric properties. Therefore, we felt that it would be appropriate for use in the item generation phase of this study due to it's shorter length.  2.2.3 Patient Population Our objective was to develop a quality of life questionnaire for patients with potentially curable carcinoma of the esophagus to be used along with the E O R T C Q L Q C - 3 0 , for use in clinical trials comparing different treatment options. Only patients with potentially curable carcinoma were included. Patients receiving palliative care only were excluded. Patients were enrolled as they presented to the study surgeons in two centres - Vancouver Hospital and Health Sciences Centre ( V H H S C ) in Vancouver, British Columbia; and the T o m Baker Cancer Centre ( T B C C ) in Calgary, Alberta. A third centre was included after the study began - the British Columbia Cancer Agency, Victoria, British Columbia. Signed informed consent was obtained for all patients. (Appendix I)  Since the final questionnaire w i l l be used on patients with squamous cell carcinoma of the esophagus or adenocarcinoma of the esophagus or gastroesophageal junction, both neoplasms are included. Both are similar in natural history, symptoms, management problems, treatment and survival. Since the purpose of the questionnaire is its use in clinical trials comparing treatment options, the preliminary work included patients undergoing therapies which would be assessed in future studies.  27  2.2.4 Item Generation It is important to begin with a pool of items which encompasses all the ways in which the life of a patient with esophageal carcinoma is affected. Interviews with particularly articulate patients are key in this process, since we are interested in assessing the quality of life of the patient - a concept which is very subjective and w i l l often be different among patients with the same clinical severity of disease. The aim is to capture the individual's assessment of the impact of their disease on the quality of their life. Interviews with health professionals, care givers and family members are all useful in generating items of importance in this phase. A review of the existing medical literature as well as reviewing previously established generic instruments w i l l also prove beneficial in generating items to be tested. (Appendix II, III, IV)  In the Item Generation phase we conducted interviews (Appendix V) with: 20 articulate patients; five family members; health care professionals - four thoracic surgeons, four radiation oncologists, two medical oncologists, three gastroenterologists, four nursing staff members and one dietician. This process yielded a total of 195 items as follows:  • symptoms  55  • emotions  53  • physical functioning  17  • activities of daily living  48  • leisure/social activities  22  2.2.5 Item Reduction In the second phase, item reduction, patients are asked whether or not an item has affected them over a specified period of time and then asked to rate the importance of the item using a Likert Scale. (Appendix VI) Once patients have been interviewed, an impact score for each item is calculated as the product of the item frequency times the mean item importance.  28  Clinical judgement is also applied. For example, item responsiveness - its ability to detect change over time is an important aspect which must be considered when reducing the initial list of items. There may be an item which has no possibility of changing over time due to an intervention or because of the natural history of the disease. For example, an item indicating that the patient is doing "more inactive pastimes" is unlikely to change with any of the treatments available. In an evaluative instrument, such as ours, we would exclude such an item since it would only add to the measurement noise and not provide any insight into the quality of life of the patient population. Here is where clinical judgement must play a role. The same item may be deemed important by the patients completing the questionnaire in the this phase of the study and one must consider whether it may be a future target for therapy in deciding whether to include the item.  In addition to the impact scores and clinical judgement the following must also be considered:  i)  the possibility of excluding items which cannot demonstrate change over time given the interventions which are currently available  ii)  inclusion of items related to the specific goals of the intervention  iii)  inclusion of an adequate number of items for each dimension to be measured  iv)  exclusion of redundancies  Once the items have been reduced, they w i l l be grouped into domains and the item responses w i l l be developed. Usually in an evaluative instrument, Likert Scales are used - either five or seven point scales. Usually, one is equal to no impairment and seven (or five) is equal to total impairment. In this study, we reversed the numbers to correspond to the scale used in the E O R T C - Q L Q 30, so one is equal to total impairment and 7 is equal to none. Visual analogue scales may also be used but are more difficult to administer and not as easy to interprets as Likert scales.  29  One final point of importance, the time period of evaluation must be specified very clearly. Period of time w i l l depend on the given disease and the type of intervention taking place, but it must also consider the limitations of the patient's memory. T w o weeks is a time frame which is often used. A s k i n g patients to recall how they have been feeling, for the purposes of evaluation, over a longer period of time could cause difficulties.  2.2.5.1 Inclusion Criteria Patients were included in the item reduction phase of the study i f they met the following criteria:  • squamous or adenocarcinoma of the esophagus or gastroesophageal junction • clinical staging T l - 3 , N X , M O •candidate for potentially curable therapy • 18 years of age or older • able to give written, informed consent • speak and read English • no other major illness affecting quality of life  2.2.5.2 Patient Sample Patients were enrolled in the Item Reduction phase of the study as they presented to the study surgeons either in the clinical office or in hospital. Demographic data was also recorded for each patient by the research assistant. (Appendix VII) Thirty-eight patients were interviewed in the following treatment categories:  • pretreatment and recently diagnosed  (n=10)  • chemoradiotherapy as primary treatment  (n=8)  • neo-adjuvant therapy and surgical resection  (n=9)  • surgical resection alone  (n=H)  30  2.2.5.3 Questionnaire A d m i n i s t r a t i o n A l l interviews were conducted in a clinical setting by two trained research assistants in Vancouver and one trained research nurse in Calgary. (Appendix VI) Patients were first asked to spontaneously identify all the ways that their lives had been affected by their disease in the previous month, first in general and then in each of the domains. Their responses were indicated on the item response sheet (Appendix VII) as "spontaneous". After all spontaneous items were identified, the patient was asked i f any of the remaining items also posed problems for them. These responses were identified on the item response sheet as "elicited". For example, we asked: "in thinking about how your  esophageal cancer has affected your life over the past month, which of the following items have also been a problem for you". The items were then read out by the interviewer who paused briefly after each one allowing the patient time to indicate i f it had been a problem. Once all items were identified, the interviewer asked the patient to rate the importance of each identified item, not the severity of the item, but the importance. The patient was asked: "how much does it bother you that  affects you at all? " The patient would then give a rating using the following five point Likert scale.  1  this only bothers me a little  2  this bothers me somewhat  3  this bothers me quite a bit  4  this bothers me a lot  5  this bothers me very, very much  Items which were not identified as a problem for the patient received an importance rating of zero. The impact score was calculated for each item by multiplying the item frequency (number of patients identifying the item as important) by the mean item importance.  31  Patients also completed the M O S 20 Item Short-Form Health Survey (SF-20) after the item reduction interview and before leaving the clinic. (Appendix VIII)  2.2.5.4 Statistical M e t h o d s Mean and standard deviations were calculated for demographic variables. Impact scores were calculated as the item frequency X mean item importance. For example, if the frequency of a given item was 17 and the mean importance was 1.47, then the impact score would be 24.9. For consistency across all patient groups, the impact score was converted to a number out of 100. Patient treatment groups were analyzed separately in order to capture differences in patients importance ratings due to type of treatment received. Items with impact scores less than 20/100 were excluded. O f the remaining items, impact scores by patient treatment group and domain were examined, clinical judgement applied and the number of items was reduced to 22 (QOL22).  If the 22 items were measuring what they were suppose to measure and adequately represented the domains of quality of life for patients with esophageal cancer, then we would expect correlations between the Q O L 2 2 and the corresponding scores on the M O S SF-20 to be moderately correlated (between 0.35 to 0.5). Crude cumulative scores were calculated for each patient on the Q O L 2 2 by simply summing the importance rating to get a total score out of 100 and domain scores. Pearson's Correlation Co-efficient was calculated, first between the SF-20 and the Q O L 2 2 in full, then as follows:  i)  SF-20 total score vs Q O L 2 2 total score  ii)  SF-20 physical score vs Q O L 2 2 physical score  iii)  SF-20 role function score vs Q O L 2 2 activities of daily living score  iv)  SF-20 mental score vs Q O L 2 2 emotional score  v)  SF-20 social score vs Q O L 2 2 social score  32  2.2.6 Pre-Testing Patients diagnosed with esophageal carcinoma were approached as they presented to the study surgeon in Vancouver and asked to participate in the pre-testing of the questionnaire. Patients first completed the E O R T C Q L Q C - 3 0 Questionnaire. Immediately following they completed the 15 item questionnaire. The questionnaires were completed without prompting from the research assistant. Once the questionnaires were complete, the research assistant recorded patients' responses (Figure 3) when asked i f they understood the questionnaires, if there were any words or concepts which were difficult to understand, and finally, i f they thought there were any areas of importance that were not covered in the questionnaires. F i g u r e 3 Pre-Testing F o r m DEVELOPMENT OF A QUALITY OF LIFE QUESTIONNAIRE FOR PATIENTS WITH E S O P H A G E A L CANCER PRE-TESTING INTERVIEW Study ID#:  Name:  Date:  /  /.  Please take a few minutes to complete this questionnaire - EORTC QLQC-30 which is a general cancer quality of life questionnaire. When you have finished, please complete the next 15 item questionnaire. Please keep in mind that there are no right or wrong answers to your questions and that all information will be kept confidential. Now that you have completed the questionnaires, I'd like to ask you a few questions about your impression of them. 1. Overall, did you find the 15 item questionnaire easy or difficult to understand?  2. Were there any concepts that you found difficult to understand? v  3. Did you find the wording used in the questionnaire easy or difficult to understand?  4. In your opinion, do you think this questionnaire covered the areas of interest to you because of your illness, or were there areas that were missing that should have been included?  Thank vou for taking the time to complete the questionnaires todav.  33  2.3.1 Objectives - Validation Phase The objective of the validation phase of the study was to assess the instrument, developed in the previous phase, for its: reliability - is it reproducible on a stable population; responsiveness - its ability to detect small, clinically important, change over time; and validity - whether it is actually measuring what it is suppose to measure.  2.3.2  Choice of Instruments  A l o n g with our newly developed Esophageal Cancer Quality of Life Questionnaire ( E Q O L ) , the data collection for this phase of the study included the following questionnaires.  2.3.2.1 EORTC QLQC-30 The E O R T C Q L Q C - 3 0 is described in Chapter 1. Chapter 2 explains why it was chosen for this study and our intent to develop a module for use with the E O R T C Q L Q C - 3 0 . The rationale holds for the validation phase of the study as well.  2.3.2.2 MOS SF-36 In the validation phase of the study, patients completed the M O S SF-36 Health Survey. In the earlier development stage of the study, patients completed the M O S SF-20 item questionnaire. The domains of the M O S SF-20 and the domains of the E Q O L were moderately correlated, which is what we would expect given that our questionnaire is a disease specific one. However, we felt that the M O S SF-36 would be more appropriate for the validation phase of the study. Not only are there more items overall, but the items are better distributed within the domains of interest. The M O S SF-36 is described on page 13. It is a multi-dimensional questionnaire assessing overall health, functional status and well being. The M O S SF-36 has a history of high response rate and excellent 85  test-retest reliability.  34  2.3.2.3 Global Rating of Change Questionnaire For each time interval between evaluations, each patient is categorized either as having "stayed stable" or as having "changed". This categorization can come from clinical indices or from a global rating of change. W e chose to use a Global Rating of Change Questionnaire (Appendix VI) as described by Juniper et a l . This global rating asks 62  whether the patient has changed, i f that change was worse or better, and how much the patient changed - using a 7 point Likert scale.  2.3.3  Time Sequence  2.3.3.1 Permission to use the instruments Permission to use the E O R T C - Q L Q 3 0 had been obtained prior to the development phase of the study and was valid through to completion. Permission was also obtained to use the M O S SF-36 and the scoring manual and additional material was received. The Global Rating of Change Questionnaire was obtained from Dr. Gordon Guyatt at McMaster University.  2.3.3.2 Orientation of Study Personnel A l l study surgeons and co-ordinators were oriented to the study procedures - enrolment, inclusion and exclusion criteria, questionnaire administration and data collection - either by telephone or in person. The research assistants overseeing the patient completion of the forms, was also instructed not to influence patient's answers to the questions but to indicate that patients should pick the most appropriate answer from their own perspective.  2.3.3.3 Preparation of the Questionnaire Study Pack The questionnaire study pack contained the consent form, demographic form, written interview, final version of the E Q O L , the E O R T C - Q L Q 3 0 , M O S SF-36 and the Global Rating of Change Questionnaire. The study pack was couriered to each study centre.  35  2.3.4  Patient Population  Patients enrolled in the validation study met the same inclusion/exclusion criteria as patients in the development phase (refer to page 28). Patients with either squamous cell carcinoma or adenocarcinoma of the esophagus or gastroesophageal junction were included. Patients were referred to the study surgeons either from their Family Physicians or from a Respirologist. Signed informed consent was obtained for all study patients. (Appendix I X ) Four study centres participated: Vancouver Hospital and Health Sciences Centre ( V H H S C ) in Vancouver British Columbia; T o m Baker Cancer Centre ( T B C C ) in Calgary, Alberta; Winnipeg General Hospital in Winnipeg, Manitoba, London Health Sciences Centre in London, Ontario. The study surgeon assessed the patient, determined a diagnosis of esophageal cancer and the absence of any signs or symptoms of metastatic disease. The patient was then seen by a trained Research Assistant who explained the study and obtained signed, informed consent.  2.3.5 Patient Sample Demographic data was collected and recorded for all patients. ( A p p e n d i x X ) Sixtyseven patients were enrolled in the validation study, however one patient died before completing any of the study forms and a second patient died after completing the baseline forms only. Therefore, sixty-five patients were followed in the study, shown by study centre in F i g u r e 4.  F i g u r e 4 Patient Sample by Study Centre  • • • •  n=7  Vancouver London Calgary Winnipeg  n=5  36  F i g u r e 5 shows the patient treatment groups in the validation phase. A l l potentially curable treatment groups are represented with the largest group in the validation phase being the surgical resection alone group.  F i g u r e 5 Patient Sample by Treatment G r o u p Study Population by Treatment Group  • Chemoradiotherapy • Adj u va nt/S u rg e ry • Surgery Alone • Palliative  2.3.6 Questionnaire A d m i n i s t r a t i o n The final version of the E Q O L is self-administered. Patients were seen in clinic and given the questionnaire package by a trained research assistant and instructed to complete all questions, to pick the most appropriate answer, and to complete the questionnaire themselves, according to how they perceived their quality of life had changed since the previous visit. (Appendix X I ) Patients were also reminded that there were no right or wrong answers to any of the questions, that we wanted to get their opinion of how their health status. If patients were from out of town, the questionnaire packs were mailed to them in stamped, return addressed envelopes with a covering letter. One patient was unable to come to the clinic office and was seen at home by the research assistant and one patient completed the forms over the phone with the research assistant. F i g u r e 6 illustrates the time line for patients entering the study and all subsequent Q O L assessments. Questionnaire packets were completed at baseline, one week after baseline but prior to treatment, one month after treatment began, three months after and six months after. The baseline questionnaire packet included the E Q O L (Appendix X I I ) , E O R T C - Q L Q - C 3 0 (Appendix X I I I ) , the and the M O S SF-36 ( A p p e n d i x X I V ) . A l l subsequent packets were the same with the addition of the Global Rating of Change Questionnaire (Appendix X V ) . 37  2.3.7 Statistical Methods  2.3.7.1 Demographics Mean and standard deviations were calculated for the demographic variables including age, gender, employment status, interview location, patient treatment groups, patient symptom classification, cancer type, stage of disease and surgical approach.  2.3.7.2 Global Rating of Change The Global Rating of Change Questionnaire was used to categorize patients in one of two groups - Group A (stable) or Group B (changed). Patients were asked i f there had been any change in their overall quality of life and asked i f there had been any change in each of the five domains - symptoms, physical function, emotional function, activities of daily living, and social. Patients who indicated that they had changed were then asked to rate their change on a seven point Likert scale indicating better change and a separate seven point Likert scale indicating worse change. Patients who indicated that they were about the same, or who indicated that they had changed but rated that change as 1 (very small change, not important) were placed in Group A and considered to be stable. Patients who indicated that they had changed for the better or for the worse and then further rated that change between 2 to 7, were placed in Group B and considered to have changed their health status since the previous visit.  2.3.7.3 EORTC-QLQC 30 Scoring The E O R T C - Q L Q C 3 0 includes five functional scales, three symptom scales, a global health status/QOL scale and six single items. A l l of the scales and single items range in score from 0 to 100, in final score format, with a high scale score representing a higher response level. In other words, a high score for a functional scale represents a high or healthy level of functioning, whereas a low score would indicate a low level of functioning. For the single item symptom scales, however, a high score indicates a high level of symptomatology or problems, whereas a low scale indicates a lower level of symptomatology and hence less problem associate with that particular symptom. Scoring  38  of this questionnaire followed the principles outlined in the scoring manual  which was  obtained directly from the E O R T C . First, the raw score was obtained by calculating the mean for the domain as: Raw Score = (Ii + I + I3 + .... I )/n 2  n  Next, a linear transformation was performed to obtain a score, S, between 0 to 100: Functional scales -  S = { 1 - (RS - 1) }x 100 Range  For symptom scales - S = {(RS - 1) / range } x 100 A n d global status Range is the difference between the maximum possible value of the raw score and the minimum possible value. A l l items in a scale take the same range of values, hence the range of the raw score equals the range of the item values. For example, most items are scored from 1 to 4, giving a range of 3. The exceptions are the items of global status/QOL which are 7 point scales with a range of 6, and the yes/no items which have a range of 1. Table 1 illustrates the domains of the questionnaire and the scoring principal including the number of items in a scale and the range.  The E O R T C Q L Q - C 3 0 scoring manual suggests that less than 2% of patient data w i l l be missing and in our study we encountered very few missing data points. The small amount of missing data was handled as follows: if at least half of the items in the scale were answered, an assumption was made that the missing items have values equal to the average of those items which are present for that patient.  39  F i g u r e 6 V a l i d a t i o n Phase - Patient T i m e L i n e  Patient referral: Family Physician or Respirologi logist  I  Surgeon assessment: - diagnosis of potentially curable esophageal cancer - no signs or symptoms of metastatic disease - Further clinical investigations if needed  Assessment of Q O L - 1 week: (prior to treatment) - EORTC - QLQ30 - EQOL - MOS SF-36 Global rating of change questionnaire  Informed consent: obtained by trained research assistant and Baseline forms completed: - EORTC - QLQ30 - EQOL - MOS SF-36  ^ chemoradiotherapy as primary treatment  —  Treatment:  adjuvant radiotherapy + surgical resection  -  surgery alone  palliative (indicated after enrolment)  Assessment of Q O L - one month: - EORTC - QLQ30 - EQOL - MOS SF-36 Global rating of change questionnaire  • Assessment of Q O L - three months: - EORTC - QLQ30 - EQOL - MOS SF-36 Global rating of change questionnaire  • Assessment of Q O L - 6 months: - EORTC - QLQ30 - EQOL - MOS SF-36 Global rating of change questionnaire  40  2.3.7.4 E Q O L Scoring The E O R T C Q L Q - C 3 0 suggests that for those who are writing their own modules, simply prepare a similar table and calculate scores accordingly (Table 2). The E Q O L consists of five domains: symptom, emotional function, physical function, activities of daily living and social function. The fifteen items in the questionnaire used a seven point Likert scale with 1 representing the worse score and 7 representing the best score. The E Q O L uses three different response options depending on the domain and type of question asked (Figure 7). For the symptom domain, the response options range from 1 = a very great deal of distress and discomfort to 7 = no distress or discomfort. For the  Table 1 Scoring the EORTC QLQ-C30 Scale  #of Items  Item Range  Item #'s  QL  2  6  29,30  PF RF EF CF SF  5 2 4 2 2  1 1 3 3 3  1-5 6,7 21 - 2 4 20,25 26,27  FA NV PA DY SL AP CO DI FI  3 2 2 1 1 1 1 1 1  3 3 3 3 3 3 3 3 3  10,12,18 14,15 9,19 8 11 13 16 17 28  Function Scales  Global health status/QOL Global health status/QOL  Functional Scales Physical functioning Role functioning Emotional functioning Cognitive functioning Social functioning  F F F F F  Symptom scales/items Fatigue Nausea and vomiting Pain Dyspnoea (SOB) Insomnia Appetite loss Constipation Diarrhea Financial difficulties  emotional function domain, the response options range from 1 = all of the time to 7 = none of the time. For the physical function domain there are two response options - one  41  as in the symptom domain addressing distress and discomfort and the other ranging from 1 = totally limited to 7 = not at all limited. Activities of daily living and the social domain both use the "limited" response option as outlined for the physical function domain. Therefore a high score on the E Q O L would represent a higher level of functioning/health, except in the symptom domain in which a high score would represent a higher symptomatology.  A s in the E O R T C Q L Q - C 3 0 , raw scores were first calculated as follows: R S = (Ii + I + . . . . I n ) / n 2  A linear transformation was calculated using a similar formula as the E O R T C Q L Q - C 3 0 , but in reverse. This is necessary since a high raw score on the E O R T C Q L Q - C 3 0 actually represents a lower level of functioning/health, while a high raw score on the E Q O L represents a higher level of functioning/health. After the linear transformations, however, the scores for both questionnaires go in the same direction.  The linear transformation, therefore, for the E Q O L is: Function scales  Score = {(RS - 1) / range} x 100  Symptom scale  Score = { 1 - ( R S - 1 ) } x 100 Range  Missing data points were handled the same as in the E O R T C Q L Q - C 3 0 . If at least half of the items in the scale were answered, an assumption was made that the missing items have values equal to the average of those items which are present for that patient. Again, there were very few missing data points.  2.3.7.5 M O S SF-36 Scoring The scoring manual for the M O S S F - 3 6  87  was obtained directly from the Medical  Outcomes Study office. Again, raw scores were calculated which then underwent a linear transformation to produce a score from 0 to 100. The scale for the items varied in  42  Table 2 Scoring the EQOL Scale  #of Items  Item Range  Item #'s  Function Scales  Functional Scales Physical functioning Emotional functioning Activities of daily living ( A D L ' s ) Social functioning  PF EF ADL SF  3 2 3 2  6 6 6 6  8,10,15 6,11 3,5,9 4,14  F F F F  Symptom scales Symptoms  SS  5  6  1,2,7,12,13  Figure 7 E Q O L Response Options Symptom Domain  Emotional Domain  1 a verv great deal of distress or discomfort 2 a great deal of distress or discomfort 3 a good deal of distress or discomfort 4 a moderate amount of distress or discomfort 5 some distress or discomfort 6 verv little distress or discomfort 7 no distress or discomfort  1 all of the time 2 most of the time 3 a good bit of the time 4 some of the time 5 a little bit of the time 6 hardlv anv of the time 7 none of the time  Physical, ADL, Social Domains 1 totallv limited 2 extremely limited 3 very limited 4 moderately limited 5 some limitation 6 a little limitation 7 not at all limited  direction - some item scales had a higher number indicating better quality of life and others had a low score indicating better quality of life. Raw scores were calculated by 43  first converting the item scoring so that all item scores were in the same direction. Once the linear transformation was complete, higher scores indicated a higher level of functioning and lower scores indicated a lower level of functioning or worse quality of life. Once all scoring was complete, all three questionnaires - the E O R T C - Q L Q C 3 0 , the E Q O L and the M O S SF-36, all had higher scores indicating better quality of life.  2.3.7.6 Reliability Analysis Reliability in an evaluative instrument is expressed as the within-patient change prior to any forms of treatment or change due to natural history of the disease. (Figure 8) Before we can expect an instrument to measure what it is suppose to measure, we must first obtain evidence that it is measuring something in a reproducible fashion. In other words, we must first obtain evidence that the measurements obtained by the instrument, whether it be measurements of individuals on different occasions, or by different observers, or by similar or parallel tests, is in fact producing the same or similar results.  69  In this case, we  are testing the reproducibility of results on individuals at a specified time interval. F o r each domain of the E Q O L , a paired samples f-test was conducted between the baseline and the one week administration of the questionnaires. Patients completed the questionnaire at one week, prior to receiving any treatment. One would expect this group of patients to be relatively stable. A t the one week administration, patients also completed the Global Rating of Change Questionnaire in which they were asked i f there health status had changed since their last visit (one week earlier). If patients answered that they had not changed, they were placed in Group A . (Table 3) If they answered that they had changed and rated that changed between 2 to 7, they were placed in Group B. Patients who rated their change as 1, were placed in Group A . Therefore, one would expect that the patients in Group A provided a group of stable patients. In the reliability analysis, it was the patients in this stable group - Group A , that were included. If the instrument is reliable, one would expect to see a high correlation between the scores at baseline and at one week and expect to see a small magnitude of change from baseline to one week.  44  Table 3 Global Change Groups Group A  Group B  Interval  Baseline to 1 week  Global change response Worse change score Better change score Reliability testing Responsiveness part 1 Responsiveness part 2 Responsiveness index  'about the same' 1 1 Yes No Yes Yes  Baseline to 1 month 1 month to 3 months 3 months to 6 months 'better' or 'worse' 2 to 7 2 to 7 No Yes Yes yes  2.3.7.7 Responsiveness Analysis There are three aspects to responsiveness that must be tested i n validating a quality of life instrument.  62  First we must ask, "in those who truly change their health status, can the  EQOL measure this change". Secondly, "is EQOL able to distinguish between those who changed and those who remain stable? Does the magnitude of change differ between changed and stable? Finally, the responsiveness index is determined by calculating the minimal important difference and the pooled standard deviation of patients.  1. In those who truly change their health status, can the EQOL measure this change? In other words can the E Q O L measure changes where it truly exists. In order to answer this question, patient scores from Group B were included. Group B is comprised of those patients who stated that they had changed i n their health status from the previous visit, and who rated that change between 2 to 7. Since patients were asked i f their change was better or worse, their scores must remain i n the category they indicated. For example, if a patient said they were worse since the previous visit, their scores were multiplied by - 1 , regardless of whether the scores indicated a direction change of better or worse. Paired samples Mests were conducted using patient scores from Group B for the time interval of baseline to one month, from one month to three months and from three months to six months. If the E Q O L is, indeed, able to measure change where it truly exists, one would  45  expect to see a significant difference (p<0.05) between the mean scores at each time interval.  Figure 8 Testing the Reliability, Responsiveness and Validity of the EQOL  Reliability What: Expressed as within-patient change prior to treatment. How: Paired t-test in stable patients between baseline/1 week Result: High correlation i f reliable  Responsiveness 'What: How: Result:  What: How: Result:  2  3  'What: How: Result:  E Q O L measure change that truly exists? Paired Mest by domain between intervals Significant change in mean scores  E Q O L distinguish between changed and unchanged Independent samples Mest Significant magnitude of change Responsiveness Index Mean change at 1 month for Global Rating Score of - 3 , -2, +2, +3 M i n i m a l important difference  Validity 1  2  3  What: How: Result:  Face validity Experts opinion of content, understandability, wording, presentation High assessment rating if valid  What: How: Result:  Criterion validity Compare to gold standard - used SF- 36 pseudo gold standard High correlation i f valid, but difficult i f no gold standard  What: How: Result:  Construct Validity A prior predictions between E Q O L and other constructs High correlation, i f valid  46  2.1s the EQOL able to distinguish between those who change and those who remain stable? Does the magnitude of change differ between changed and stable? Patient scores from both Groups A and B are used in this analysis. What we are measuring in order to answer these questions, is the magnitude of change between scores in Group A patients (stable group) and the magnitude of change between scores in Group B (changed group). The objective in developing and validating the E Q O L is to have an instrument which can measure the quality of life associated with the various treatment modalities for esophageal cancer.  Since the effect of treatment is largest in approximately the first  month or so after treatment has begun, we are most interested in the magnitude of change between the baseline visit and the one month visit. However, for comparison, the magnitude of change was also analysed for the intervals between one month and three months and between three months and six months. W e were also interested in seeing if the magnitude of change differed between those patients who indicated that they were better than at the previous visit or for those patients who indicated that they were worse.  In order to compare the magnitude of change, an Independent Samples f-test was conducted. Since it is the magnitude of change that we are interested in to answer this question, and not the direction of change, no transformations were done to the data before running the analysis. So, four groups of analysis were run in order to assess the magnitude of change. The groups were between: i)  those who indicated change and those who indicated no change  ii)  those who indicated no change and those indicated they were worse  iii)  those who indicated no change and those indicated they were better  iv)  between those who indicated worse and those who indicated better  3.Responsiveness Index The responsiveness index is derived from the minimal important difference and the pooled standard deviation of patients in both Group A and Group B. The responsiveness index and minimal important difference is key in understanding the interpretability of quality of life measures.  The minimal clinically important difference ( M C I D ) was  47  calculated for each domain of the E Q O L . Again, since we are most interested in the time period between the baseline visit and the one month visit, we calculated the M C I D for that period. Patients have indicated in other studies  that changes in Global Rating of  Change scores of two or greater, are important to them. Therefore, patients who score 88  two or three are considered to have experienced a minimal change.  Patients who scored  four or five on the Global Rating of Change Questionnaire are considered to have experienced a moderate change and those scoring six or seven to have experienced a large change. For patients in Group B (changed group) who rated their change as either 2 or 3 we calculated the magnitude of change from the baseline visit to the one month visit. If a patient said they were worse than the previous visit the score was multiplied by - 1 , in other words, when the change was negative, the sign was reversed.. The magnitude of change was calculated by domain but expressed as a number per question. For example, the emotional domain has 5 items, therefore, the difference in item scores, from baseline to one month, divided by five is the minimal clinically important difference for the emotional function domain. The M C I D is the smallest difference in score in a domain of interest which patients perceive as beneficial or important and which could result in a change in patient management.  89  The minimal effect can be expressed as a range, as can  the moderate and large effect. Information gathered from this responsiveness index can be used in planning future trials in calculating the sample size which w i l l depend in part, on the magnitude of difference that investigators consider to be clinically important. 2.3.7.8 Face and Content Validity Face validity refers to whether an instrument appears to be measuring what it is intended to measure. Content validity refers to the extent that a domain is comprehensively sampled by the items in the instrument. According to Feinstein, quantitative testing of face and content validity is rarely attempted. Hence, Feinstein developed a reformulation of these aspects of validity - face and content - by suggesting criteria for what he calls "sensibility".  90  Oxman and Guyatt validated these criteria further in assessing review  articles. These criteria used in our study to assess face and content validity are: 91  48  i)  applicability of the questionnaire  ii)  questionnaire clarity and simplicity  iii)  likelihood of bias  iv)  comprehensiveness  v)  redundancy items  The following steps were performed to apply these criteria:  1. Five patients were pretested. Patients were asked to complete the questionnaire. When complete, they were asked, by the research assistant, i f it covered all the areas that were important to them because of their esophageal cancer. They were also asked i f any of the wording was difficult to understand or i f the questions were difficult to understand in terms of content.  2. Expert opinion was sought from three thoracic surgeons, one thoracic surgery resident, and one thoracic surgery nurse. Each were asked to review the questionnaire using the above criteria.  2.3.7.9 Criterion Validity Criterion validity is the psychometric property that is assessed when a gold standard exists. Studies sometimes assess criterion validity in the absence of a gold standard, by using a "pseudo" gold standard. However, the results w i l l not be conclusive and in fact, may not be very useful. W e decided to conduct a criterion validity assessment using the M O S S F - 36 Health Survey as the "pseudo" gold standard. Pearson's Correlation C o Efficient was calculated between corresponding domains of the E Q O L and the M O S SF36 as follows: i) E Q O L physical  and  SF-36 physical function  ii) E Q O L activities of daily living  and  SF-36 role function  iii) E Q O L emotional  and  SF-36 emotional role function  iv) E Q O L social  and  SF-36 social function 49  Correlations are assessed as follows : r > .5  strong correlated  r = .35 to .5  moderate correlation  r = .20 to .35 poor correlation  The symptom domain of the E Q O L could not be compared to the SF-36 as no symptom domain exists, since it is a general qol questionnaire designed to evaluate overall qol across patient populations. A high correlation between the domains of the E Q O L and the SF-36 would indicate no need for the E Q O L . A poor correlation would indicate that the E Q O L is missing something important. Therefore, ideally, we would like to see a moderate correlation between the domains of the E Q O L and the SF-36. However, given that the SF-36 is a stand-in gold standard, it may be difficult to get an accurate assessment. Therefore, the results should be looked at with some scepticism.  2.3.7.10 Construct Validity Construct validity is the most vigorous approach to establishing validity. A construct is a theoretically derived notion of the domain of interest.  91  One's understanding of the  domain's construct produces expectations, or a priori predictions, about how the instrument should behave, in conjunction with other objective measures such as clinical indices, i f it is indeed valid. For the E Q O L , an evaluative instrument, construct validity is demonstrated by showing that changes in the E Q O L correlate with changes in other related measures within the theoretical construct. A n example, the H R Q L , an evaluative qol questionnaire for chronic lung disease was predicted to have moderate correlations with changes in walk test scores. The prediction was supported in the validation study.  92  For patients with esophageal cancer, due to the nature of the disease and its presentation, there are no true clinical indices of change after treatment, with which to compare the E Q O L , except recurrence of disease. Therefore we chose to use pseudo indices, or surrogates.  50  In assessing construct validity we developed a set of a priori predictions.  1. Since the vast majority of esophageal cancer patients present with severe symptoms of dysphagia, one would expect that at one month post treatment, dysphagia would be improved, therefore, we predicted: an improvement in dysphagia from baseline to one month follow-up in all treatment groups. Paired £-tests were performed between the baseline raw dysphagia scores and the one month follow-up, first for all treatment groups, then for the chemoradiotherapy as primary treatment group, the adjuvant therapy followed by surgical resection group, and finally, the surgery alone group.  2. Given the amount of morbidity associated with surgical treatment in the immediate post-operative period, we predicted that: i)  at one month follow-up, the chemoradiotherapy group would have better scores in the physical function domain than the group treated with adjuvant therapy followed by surgery, or the surgery alone group  but, given that the chemoradiotherapy regimen has a longer duration than surgical treatment and given our preliminary findings in the development phase which indicated that patients receiving chemoradiotherapy had worse emotional function scores, we predicted that: ii)  at one month follow-up, the chemoradiotherapy group would have worse scores in the emotional function domain than the group treated with adjuvant therapy followed by surgery, or the surgery alone group  For i) above, descriptive statistics were produced, stem and leaf plots were generated and paired M e s t s were performed at one month, first for all treatment groups together, then for the chemoradiotherapy group, then for the adjuvant therapy followed by surgery group, and finally for the surgery alone group. Independent samples M e s t s were also performed at one month as follows: chemoradiotherapy group vs both surgery groups together; chemoradiotherapy group vs adjuvant/surgery group; chemoradiotherapy group  51  vs surgery alone group. For ii) above, the same analysis was conducted with data from the emotional function domain of the E Q O L .  3.Stage of disease in this patient population is closely related to survival. The earlier stages have a greater survival while the later stages of the disease have a very low five year survival. In this case, stage of disease could be considered a surrogate for recurrence of disease. Therefore, we predicted that: patients with Stage III or Stage IV disease at baseline have worse scores at three months and six months post treatment (due to the likelihood of recurrence) than patients with Stage I or Stage II disease at baseline in all domains - symptom, emotional, physical, adl 's, and social Descriptive statistics were generated as well as stem and leaf plots at three months post treatment and as six months post treatment. Paired samples r-tests were conducted for each domain between the baseline and three month follow-up visits for all stages combined, then for early stage (stage I and II), then for late stage (stage III and IV). Finally an Independent Samples r-test was performed between early stage and late stage at the three month follow-up visit. The analysis was repeated for each of the domains of the E Q O L .  4.Patients were asked at baseline to categorize their symptoms, in their opinion, as mild, moderate or severe. For those patients who categorized their symptoms as mild (again an indicator of early stage disease), we predicted: a better score at baseline and at three months in all domains, than patients who rated their symptoms as severe In order to assess this construct, descriptive statistics were calculated and an Independent samples r-test between mild and severe symptoms at baseline and at three months was performed. Stem and leaf plots were also generated.  52  CHAPTER THREE - RESULTS  3.1 D E V E L O P M E N T P H A S E R E S U L T S  3.1.1 Item Reduction Results There were 29 men (76.3%) and 9 women (23.7%) participating in the Item Reduction Phase of the study. This gender distribution adequately represents the approximate gender distribution for this disease. Mean age was 63 (st.dev. 13.1), minimum age was 36 and maximum age was 82. Demographic data is summarized in Table 4. There were 10 patients (26.3%) in the pretreatment group; 8 (21.2%) in the chemoradiotherapy group; 9 (23.7%) in the adjuvant therapy followed by surgical resection group; and 11 (29.0%) in the surgical resection alone group. Nineteen (50%) patients rated their esophageal cancer symptoms as mild; 11 (28.9%) as moderate; and 8 (21.1%) as severe. Patients were asked if they had experienced any positive change in their outlook since the diagnosis of esophageal cancer with the following results: 15 (39.5%) not at all; 9 (23.7%) a little; 8 (21.0%) somewhat; 5 (13.2%) a lot.  W e included on our item reduction list, items which appear on the E O R T C Q L Q C - 3 0 , as they were identified through the item generation phase as being important. However, we excluded those items from the analysis since our questionnaire is to be used along side the E O R T C Q L Q C - 3 0 in assessing quality of life for these patients. Once those items were excluded, 151 items remained out of the initial list of 195. Forty-four items had an impact score greater than or equal to 20.0. T a b l e 5 displays the five items with the highest impact scores in each patient group. It is important to note that the highest scoring item in the pre-treatment group was food sticking (dysphagia) with a score of 30.8 and that this item disappears from most patient groups after treatment. This result is promising as dysphagia is the most prominent symptom of esophageal cancer and often the most difficult to alleviate or control. W e can see that the majority of items of importance to the post surgical groups are associated with eating difficulties, which makes clinical sense given the type of procedure performed. The 44 items with an impact  53  score of 20 or greater were further reduced to 22 items (Table 6) when examining the impact scores by domain. T A B L E 4 D E M O G R A P H I C S - Item Reduction Phase MEAN AGE GENDER  MALE FEMALE  63 (S.D. 13.11) 29 (76.3%) 9 (23.7%)  E M P L O Y M E N T STATUS Employed outside the home Unemployed *  15 (39.5%) 23 (60.5%)  INTERVIEW L O C A T I O N Vancouver Calgary Victoria  20 (52.6%) 10 (26.3%) 8(21.1%)  PATIENT GROUPS Pretreatment Chemoradiotherapy Adjuvant/surgery Surgery alone  10 8 9 11  PATIENT S Y M P T O M CLASSIFICATION Mild Moderate Severe  19 (50.0%) 11 (28.9%) 8 (21.1%)  CANCER TYPE Adenocarcinoma esophagus Adenocarcinoma cardia Squamous cell carcinoma Other  9 18 10 1  (26.3%) (21.1%) (23.7%) (29.0%)  (23.7%) (47.4%) (26.3%) ( 2.6%)  * either retired or homemakers, none were unemployed due to disease status  Total scores and domain scores were calculated on this preliminary 22 item questionnaire, with high scores indicating a worse quality of life, and low scores indicating a better quality of life. W e found that patients who classified their symptoms as mild had lower overall total scores with a mean of 18.3 (sd 15.27) and a median of  54  Table 5 Top Five Impact Scores by Treatment Group T R E A T M E N T GROUP  QOL I T E M  Pretreatment  Food sticking (dysphagia) Fear of cancer Difficulty swallowing solid food Fear of recurrence Concern not knowing time left Concern not knowing time left Burping Anxiety Fear of recurrence Feeling helpless N o items with impact scores >20 Eat smaller meals during day Difficulty visiting friends/relatives Feeling full quickly Inability to eat sufficient food Fear of recurrence Need to elevate head of bed Food sticking Need to eat smaller meals Difficulty bending/leaning over Burping Regurgitation when bending/lean Change number of meals/snacks Feel full quickly Difficulty swallowing Inability to eat sufficient food Need to eat smaller meals Unable to do activities phys exert Feeling full quickly Dining out at restaurants Fear of cancer More inactive pastimes Decreased exercise tolerance Exercise less Difficulty bending/leaning over Unable to do activities phys exert Decreased exercise tolerance Upset at reducing activity Discourage with slow recovery Need to eat smaller meals Angry  Chemoradiotherapy Post-Op lmo  Chemoradiotherapy Post-op 3mo Adjuvant Therapy/Surgery lmo  Adjuvant Therapy/Surgery 3mo  Surgery Alone Post-op lmo  Surgery Alone Post-op 3mo  IMPACT SCORE 30.8 28.8 26.4 22.0 20.4 66.7 46.7 40.0 35.3 31.3 41.6 38.4 32.0 31.2 28.8 45.0 33.5 30.0 26.0 26.0 26.0 26.0 20.0 20.0 20.0 33.3 30.3 30.3 30.3 26.7 25.0 22.3 52.0 35.2 35.2 32.0 31.2 31.2 28.8 28.8  55  Table 6 Twenty-two Items a n d T h e i r Impact Scores IMPACT SCORE & TREATMENT GROUP ITEMS BY DOMAIN  Symptoms (7) Burping Need to elevate head of bed at night Feeling full quickly  Intolerance to certain foods Increased sensitivity to cold Food sticking Difficulty swallowing  46.7 45.0 32.0 20.0 30.0 26.4 31.3 30.8 33.5 26.4 20.0  chemoradiotherapy l m o post-op adjuvant therapy and surgery 3mo post-op adjuvant therapy and surgery l m o post-op adjuvant therapy and surgery 3mo post-op surgery only l m o post-op adjuvant therapy and surgery chemoradiotherapy l m o post-op pretreatment adjuvant therapy/surgery 3mo post-op pretreatment adjuvant therapy/surgery 3mo post-op  28.8 26.7 22.0 35.3 28.8 20.0 22.7 22.7 28.8 20.0 21.6  pretreatment surgery only l m o post-op pretreatment chemoradiotherapy l m o post-op adjuvant therapy/surgery l m o post-op pretreatment chemoradiotherapy l m o post-op chemoradiotherapy l m o post-op adjuvant therapy/surgery l m o post-op surgery alone l m o post-op surgery alone 3mo post-op  26.0 35.2 31.2 20.0 22.3 32.0 20.0 26.0 30.3 24.0  adjuvant therapy/surgery 3mo surgery only 3mo post-op adjuvant therapy/surgery l m o adjuvant therapy/surgery 3mo surgery alone l m o post-op surgery alone 3mo post-op surgery alone l m o post-op adjuvant therapy/surgery 3mo surgery alone l m o post-op surgery alone 3mo post-op  28.8 41.6 30.0 33.3 28.8  adjuvant therapy/surgery l m o post-op adjuvant therapy/surgery l m o post-op adjuvant therapy/surgery 3mo post-op surgery alone l m o post-op surgery alone 3mo post-op  25.6 30.3 38.4 52.0  adjuvant therapy/surgery l m o post-op surgery alone l m o post-op adjuvant therapy/surgery l m o post-op surgery alone 3mo post-op  Emotions (4) Fear of cancer Fear of recurrence  Fear of unpleasant treatment options Frustrated  Physical Functioning (6) Difficulty leaning/bending over Inability to eat sufficient food Decreased exercise tolerance Difficulty exercising Regurg when bending/leaning over Unable - activities with phys exertion  post-op post-op post-op  post-op  Activities of Daily Living (2) A v o i d certain foods and beverages Need to eat smaller meals in the day  Leisure/Social (3) Difficulty dining out at restaurants Difficulty visiting friends/ relatives Exercise less  56  10.9, than those who classified their symptoms as moderate (mean 28.6, sd 17.9, median 22.7) or severe (mean 33.5, sd 19.6, median 30.0).  Pearson's Correlation Coefficients were calculated on this 22 item questionnaire as follows: i)  SF-20 total score vs Q O L 2 2 total score  .45  ii)  SF-20 physical score vs Q O L 2 2 physical score  .44  iii)  SF-20 role function score vs Q O L 2 2 activities of daily living score .33  iv)  SF-20 mental score vs Q O L 2 2 emotional score  .75  v)  SF-20 social score vs Q O L 2 2 social score  .42  Using the correlations representing strong, moderate and poor described under criterion validity in the previous chapter, we can see that these results reflect a moderate correlation in all but the mental/emotional domain which has a relatively high correlation. These results w i l l be further tested in the validation phase of the study.  W e then applied clinical judgement to the remaining 22 items. The items were reviewed to see if there were any redundancies in concepts and to determine if the items make clinical sense. Seven items were removed: • Intolerance to certain foods • Difficulty swallowing • Fear of cancer • Frustrated • Difficulty leaning/bending over • Decreased exercise tolerance • Exercise less Two items, fear of cancer and frustrated, were eliminate as it was felt that neither of these items was likely to change given the current treatments available. The other five items were redundant. Intolerance to certain foods was eliminated, while avoiding certain foods and beverages was maintained, difficulty swallowing was eliminated, but food sticking was maintained. Difficulty leaning/bending over was elminated since it was felt 57  that regurgitation when bending/leaning over specifically addressed the issue and it was retained. Decreased exercise tolerance and exercise less were both removed in favour of keeping difficulty exercising.  The remaining 15 items along with the 30 items in the E O R T C questionnaire w i l l be selfadministered and easily completed within approximately 20 minutes or less.  3.1.2 Pre-testing Results Pre-testing was conducted on five patients using the interview format shown in F i g u r e 3 on page 31. A l l five patients found the 15 item questionnaire to be easy to understand both in wording and in concepts presented. A l l the patients felt that their concerns re their health, status due to esophageal cancer, were well represented in the questionnaire and fully addressed the issues of importance to them in their day-to-day lives. Given these positive results, no further pre-testing was conducted.  3.2 V A L I D A T I O N P H A S E R E S U L T S  3.2.1 Study P a r t i c i p a t i o n Over the period of February 1999 to March 2002, 67 patients with potentially curable esophageal carcinoma, were enrolled in the Validation Study for the Esophageal Cancer Quality of Life Questionnaire. O f these patients, one died prior to completing any questionnaires and one died after completing the baseline questionnaire pack only, leaving 65 patients enrolled in the validation study. Patients were enrolled in four study centres: Vancouver, British Columbia; Calgary, Alberta; Winnipeg, Manitoba; and London, Ontario.  O f the 65 patients enrolled in the validation study, some patients did not complete all intervals, for example, of the 17 patients who completed the six month questionnaire ovservation in Vancouver, five missed the one week administration and one missed the one month administration. T a b l e 7 shows the total number of patients completing all questionnaire administrations and the percentage for each centre. 58  T a b l e 7 - N u m b e r of Patients Completing Study Study Centre Vancouver Calgary Winnipeg London TOTAL  Enrolment 41 7 6 13  Complete 6 Months 17 4 2 9  % Completed 41.5 57.1 33.3 69.2  67  32  47.8  There were 48 patients who completed the one week questionnaire pack and hence participated in the reliability study, as this questionnaire administration was prior to any treatment. Forty-seven patients completed the one month pack, 45 completed the three month pack and 32 completed the six month pack. In total, seven patients died from the Vancouver centre during the course of the study. A n additional 12 withdrew because they were too i l l to continue and may have subsequently died. In Calgary, no patients died during the course of the study, but three patients withdrew due to their illness. In Winnipeg, one patient died prior to completing the packs and three additional patients withdrew during the course of the study. In London, one patient died prior to completing the questionnaire packs and three additional patients withdrew from the study.  3.2.2 Demographics The demographic characteristics of the study population are shown in Table 8.  The  average age of the patients participating in the validation phase of the study was 62.7 with the youngest being 29 and the oldest being 84. Fifty-seven patients were male representing 87.7% of the study population and 8 were female. Twenty-four patients were still employed outside the home, but 41 (63.1%) were unemployed either because of retirement or because they were housewives. None were unemployed due to their illness. The various patient groups were represented with nine patients from the chemoradiotherapy as primary treatment group, 14 from the adjuvant therapy and surgical resection group, and 35 or 53.8% from the surgical resection alone group. Seven patients were considered palliative on further investigation after enrolment into the study.  59  T A B L E 8 - D E M O G R A P H I C S - V a l i d a t i o n Phase PATIENTS IN STUDY* MEAN A G E  65 62.7 st.dev. 11.9  GENDER  57 (87.7) 8(12.3)  MALE FEMALE  EMPLOYMENT STATUS Employed outside the home Unemployed ** INTERVIEW SITE Vancouver London Calgary Winnipeg PATIENT GROUPS Chemoradiotherapy Adjuvant/surgery Pre-surgery Post-surgery Surgery alone Palliative (on further investigation) PATIENT SYMPTOM CLASSIFICATION Mild Moderate Severe CANCER T Y P E Adenocarcinoma esophagus Adenocarcinoma cardia Squamous cell carcinoma Other STAGE OF DISEASE Stage 1 Stage 2a Stage 2b Stage 3 Stage 4 Not Staged SURGICAL APPROACH Transhiatal esophagectomy (THE) L thoracotomy Lap + R thoraco L thoraco-abdominal Three hole  24 (36.9) 41 (63.1) 41 (63.1) 12(18.5) 7(10.8) 5 ( 7.7) 9(13.8) 14(21.5) 5 9 35 (53.8) 7 (10.8) 28 (43.1) 24 (35.2) 13 (20.0) 35 (53.8) 14(21.5) 14(21.5) 2( 3.0) 9(13.8) 14(21.5) 5 ( 7.7) 25 (38.5) 8(12.3) 4 ( 6.2)  38 (58.5) 0( 0.0) 2( 3.1) 4 ( 6.2) 5 ( 7.7)  * one patient died before completing any forms and one died after completing the baseline form only Unemployed were either homemakers or retired, none were unemployed due to disease  Patients were asked at the baseline visit if, in their opinion, the symptoms they were experiencing because of the esophageal cancer, were mild, moderate or severe. Twentyeight (43.1%) stated that they were experiencing mild symptoms, 24 (35.2%) felt their symptoms were moderate, and 13 (20.0-%) said they were experiencing severe symptoms. Stage of disease is represented in Table 8. W e can see that there was a fairly equal distribution between early and late stage disease. Twenty-eight patients had early stage disease, defined as stage 1, 2a, and 2b. Thirty-three patients had late stage disease, defined as stage 3 or 4. Four patients were not staged. A l l fell within the neo-adjuvant therapy followed by surgical resection group. Often, when patients received preoperative radiotherapy, the tumour is sterilized and pathological staging is difficult.  3.2.3 Reliability Paired samples r-tests were run for all domains within the E Q O L - symptom, emotional, physical functioning, activities of daily living and the social domain. Paired Mests were conducted using the symptom scores at the baseline visit and at the one week visit. Table 9 summarizes the paired samples correlation and the paired differences.  Table 9 Reliability Results Domain  N  Symptoms Emotions Physical Adl's Social  39 36 41 39 40  Mean Baseline Score 27.09 70.37 85.41 67.14 76.04  Mean 1 wk Score 25.89 71.77 80.85 64.78 70.42  Mean Difference  Standard Deviation  Pvalue  1.20 -1.4 4.5 2.3 5.6  9.9 16.6 12.5 14.6 20.1  .454 .617 .025 .319 .085  Paired Samples Correlation .889 .733 .871 .880 .749  Pvalue .000 .000 .000 .000 .000  1. Symptom domain The mean difference between the two visits was 1.20 on a scale of 0 to 100. This mean difference was not significant (p=.454). The paired samples correlation was .889 and did reach significance (p<.001).  61  2. Emotional domain The paired Mest resulted in a mean difference of -1.4 (p=/617) The paired samples correlation was .733 (p<0.001). 3. Physical domain The paired r-test for the physical domain between baseline and one week showed a mean difference of 4.56 (p=.025) with a high correlation at .871 (p<.001). 4. Activities of daily living The mean difference in this domain between baseline and one week was 2.36 (p=.319) with a paired samples correlation of .880 (p<0.001). 5. Social domain The paired r-test resulted in a mean difference of 5.62 (p=.085) with a correlation of .749 (p=.0O0).  3.2.4  Responsiveness Analysis  3.2.4.1 P a r t I - Change i n G r o u p B Part I of a responsiveness analysis looks at the difference in scores for patients who indicated that their health state had changed (Group B) since the previous visit. Paired ttests were conducted for these patients in each domain of the E Q O L from baseline to the one month visit, from the one month visit to the three month visit, and from the three month visit to the six month visit. For patients who stated that they were worse, the scores were multiplied by - 1 .  1. Symptom domain - T a b l e 10 summarizes the results for the symptom domain. Table 10 - Change i n S y m p t o m D o m a i n - G r o u p B Questionnaire Interval  N  Mean Score  1 month  25  40.00  Mean FU Score 26.26  Mean Difference  Standard Deviation  Pvalue  13.74  23.01  21.95  11.97  21.82  5.45  Pvalue  .006  Paired Samples Correlation .197  14.58  .004  .726  .001  12.20  .018  .780  .000  .346  baseline  3 months  17  6 months  22  33.91 1 month 27.27 3 moths  62  A t the one month visit, the mean difference in score from the baseline visit for the symptom domain was 13.74 (p=.006), with a mean baseline score of 40.00 and a mean score at one month of 26.26. The paired samples correlation was . 197 and did not reach significance. (p=.346). A t the three month visit, the mean difference in score from the one month visit was 11.96 (p=.004) with a mean score of 33.91 at one month and 21.95 at three months. The paired samples correlation was fairly high at .726 (p=.001). A t the six month visit, the mean difference in score from the three month visit was 5.45 (p=.048) with a mean score of 27.27 at three months and 21.81 at six months. The paired samples correlation was again fairly high at.780 (p<0.001).  2. Emotional function domain - Table 11 summarizes emotional function domain results. Table 11 - Change i n E m o t i o n a l F u n c t i o n D o m a i n - G r o u p B Questionnaire Interval  N  Mean Score  1 month  24  3 months  14  6 months  10  76.03 baseline 73.21 1 month 78.33 3 months  Mean FU Score 62.85 63.69 78.33  Pvalue  .011  Paired Samples Correlation .449  13.43  .020  .819  .000  6.82  1.00  .945  .000  Mean Difference  Standard Deviation  Pvalue  13.19  23.30  9.52 .000  .028  At the one month visit, the mean difference in score from baseline was 13.19, with a mean of 76.03 at baseline and a mean of 62.85 at one month. The paired samples correlation was .449 (p=.028). A t the three month visit, the mean difference from the one month score was 9.52 with a mean score at one month of 73.21 and 63.69 at three months. The paired samples correlation was .819 (p<0.001). A t the six month visit, the mean difference from the three month score was .000 with a mean score at both the three and six month visit of 78.33. (p=1.0). The paired samples correlation was .945 (p<0.001).  63  3. Physical function domain - Table 12 summarizes physical function domain results. A t the one month visit, the mean difference i n score from baseline was 20.28 (p=.013) with a mean baseline score of 77.55 and a mean score at one month of 56.97. The paired samples correlation was -.049 (p=.803).  Table 12 - Change in Physical Function Domain - Group B Questionnaire Interval  N  Mean Score  1 month  28  3 months  19  6 months  13  77.55 baseline 67.09 1 month 72.69 3 months  Mean FU Score 56.97  Mean Difference  Standard Deviation  Pvalue  Pvalue  .013  Paired Samples Correlation -.049  20.58  41.06  48.95  18.14  20.77  .001  .724  .000  60.51  12.18  17.26  .026  .892  .000  .803  A t the three month visit, there was a mean difference of 18.14 (p=.001) from the one month visit. The mean one month visit was 67.09 and the mean three month visit was 48.95. The paired samples correlation was .724 (p<0.001). A t the six month visit, the mean difference in score from three months was 12.18 (p=.026) with a mean three month score of 72.69 and 60.51 at six months.  4. Activities of daily living domain - Table 13 summarizes the A D L domain results  Table 13 - Change in Activities of Daily Living Domain - Group B Questionnaire Interval  N  Mean Score  1 month  27  3 months  16  6 months  16  72.16 baseline 73.54 1 month 79.99 3 months  Mean FU Score 55.02  Mean Difference  Standard Deviation  Pvalue  Pvalue  .017  Paired Samples Correlation .220  17.13  35.01  62.81  10.73  24.96  .106  .599  .014  61.98  18.01  19.13  .002  .768  .001  .270  A t the one month visit, there was a mean difference in score from baseline of 17.13 (p=.017) with a mean baseline score of 72.16 and 55.02 at one month. The paired samples correlation was .220.  64  A t the three month visit, there was a mean difference in score from one month of 10.73 (p=.106) with a mean one month score of 73.54 and a mean three month score of 62.81.  A t the six month visit, there was a mean difference in score of 18.01 (p=.002). The mean three month score was 79.99 and the mean six month score was 61.98. The paired samples correlation was .768 (p=0.001).  5. Social domain - Table 14 summarizes the social domain results  Table 14 - Change in Social Domain - Group B Questionnaire Interval  N  Mean Score  1 month  25  3 months  16  6 months  16  69.67 baseline 70.31 1 month 72.92 3 months  Mean FU Score 48.33  Mean Difference  Standard Deviation  Pvalue  Pvalue  .016  Paired Samples Correlation .181  21.34  41.32  50.53  19.79  26.86  .010  .684  .003  53.64  19.28  27.68  .014  .686  .003  .388  At the one month visit, the mean difference in score from baseline was 69.67 (p=.016), with a mean baseline score of 69.67 and 48.33 at one month. The paired samples correlation was .181 (p=.388). A t the three month visit, the mean difference in score from one month was 19.79 (p=.010), with a mean one month score of 70.31 and a mean three month score of 50.53. The paired samples correlation was .684 (p=.003). A t the six month visit, the mean difference in score from three months was 19.28 (p=.014) with a mean three month score of 72.92 and 56.64 at six months. The paired samples correlation was .686 (p=.003).  3.2.4.2 Part II - Change between Group A and Group B Part II of the responsiveness analysis looks at the magnitude of change between those patients who said they remained the same (Group A ) and those patients who said their health state had changed (Group B ) since the prior visit. A n Independent Samples t- test was conducted for each domain between the magnitude of change in group A and the 65  magnitude of change in group B, at each consecutive visit. In the changed vs unchanged or stable group, none of the comparisons show as statistically significant in any of the domains. (Table 15) In the better vs worse group, those who indicated that they had changed, the mean difference reached statistical significance in all but 4 out of 15 instances. In the symptom domain and in the emotional domain, the mean difference was significant at one and three months. In the physical function domain, the only significant difference was seen at 3 months. In the A D L domain and in the social domain, the mean difference between worse vs better was significant at all three time intervals - one, three and six months.  In the symptom domain, a higher score equals a higher symptomatology, or more severe symptoms. A higher score at three months would indicate worsening of symptoms and would appear as a - change. In all other domains, a higher score indicates better quality of life and a lower score indicates a worsening of quality of life.  In the symptom domain, at the one month visit, the magnitude of change was statistically significant between the better vs stable group (p=0.012) and between the worse vs better group (p=0.006). A t the three month visit, the magnitude of change was statistically significant between the worse vs stable group (p=.0009) and between the worse vs better group (p=0.001). A t the six month visit, there was no statistical significance between any of the groups, however, the worse vs better group approached statistical significance (p=0.064).  In the emotional function domain, at the one month visit, the magnitude of change was statistically significant between the better vs worse group (p=0.009). A t the three month visit, it was statistically significant between both the better vs stable group (p=0.022) and between the worse vs better group (p=0.004). The magnitude of change was not significant in any of the groups at the six month visit.  66  In the physical function domain, the magnitude of change was only significant at three months in the worse vs stable group (p=0.015) and in the worse vs better group (p=0.001). The better vs worse group approached significance (p=0.062).  In the activities of daily living domain, the magnitude of change reached statistical difference in the worse vs better group at one month (p=0.034) at three months (p=0.014) and at six months (p=0.007). The worse vs stable group approached significance with a p-value of 0.060. In the social domain, at the one month visit, the magnitude of change was significant in the better vs stable group (p=0.048) and in the worse vs better group (p=0.042). At three months it was significant in the worse vs stable group (p=0.002) and in the worse vs better group (p<0.001). Finally, at six months, it was significant in the worse vs better group (p=0.017). Table 15 S y m p t o m D o m a i n - M e a n Differences Between G r o u p s  Changed Vs Stable 1 mo - 5.40 5.21 3 mo 6 mo 3.91  Worse Vs Stable - 9.12 -26.58 (p=.<)09) - 9.84  Better Vs Stable -18.80 <p=.012) 0.12 -4.500E-02  Worse Vs better -27.91 (p=.0()6) -26.46 (p=.001)  Emotions  1 mo 3.92 3 mo - 8.43 6 mo - 0.37  -17.11 -17.78 0.37  - 9.26 -17.68 (p=.022) - 0.37  -26.38 (p=.009) -35.46 (p=.()04) -5.92 IE-16  Physical  1 mo 9.59 3 mo - 2.17 6 mo - 3.65  -16.40 -24.54 (p=.015) - 6.26  - 3.97 -13.17 (p=.062) - 7.90  -20.36 -37.71 (p=.()01) -14.16  ADL's  1 mo 3 mo 6 mo  1.49 9.05 0.62  -11.94 -37.26 -23.12 (p=.060)  -19.43 1.22 -10.62  -31.38 (p=.034) -36.04 ( =.014) -33.75 (p=.()07)  1 mo - 6.75 4.52 3 mo 6 mo - 3.78  - 4.89 -46.53 (p=.<>02) -17.53  -30.02 ( =.()48) - 9.48 -16.20  Domain  Symptom  Social  Time  - 9.88 (p=.061)  P  P  -34.92 ( =.042) -56.02 (p-c.001) -33.73 ( =.017) P  P  67  3.2.4.3 P a r t III - Responsiveness Index Data from patients who rated their change as important on the Global Rating of Change questionnaire were included in the calculation of the responsiveness index.  A rating of  one was considered not to be an important change. Therefore, a rating of two or three was considered to be a minimal change, of four or five to be a moderate change, and a rating of six or seven to be a large change. Table 16 displays the results of the responsiveness index analysis.  The mean change in score per question corresponding to the minimal clinically important difference is consistently around 0.5: symptom domain is 0.53; emotional domain 0.5; physical domain 0.57; A D L domain 0.40; and social domain 0.64. Effect sizes are:  • M i n i m a l effect  0.40 to 0.64  • Moderate effect  0.64 to 1.17  • Large effect  1.3 to 2.8  3.2.5 Face a n d Content V a l i d i t y In addition to the five patients who participated in the pre-testing face, we assessed face and content validity by asking five experts to review the final questionnaire - three thoracic surgeons, one in Vancouver, two in Calgary; one thoracic resident in Vancouver; and one thoracic nurse in Vancouver. A l l felt that the major areas of concern for patients with esophageal cancer were covered adequately, that the questionnaire was applicable to this patient population. They saw no apparent likelihood of bias, nor any redundancy of items. A l l felt that the questionnaire represented the important issues of quality of life in a comprehensive manner.  68  Table 16 Responsiveness Index Results Minimal change Symptom (5 items) N=8 Mean 2.63 Std.dev 4.96 Minimum -5.00 Maximum 11.00 Median 3.00 Change/item 0.53*  Moderate change N=9 3.00 7.59 -15.00 11.00 6.00 0.64  Large change  Emotional (2 items) Mean Std.dev Minimum Maximum Median Change/item  N=8 0.00 1.97 - 4.00 2.00 1.00 0.50  N=ll 2.09 3.05 - 2.00 6.00 1.00 1.05  N=5 3.00 2.65 - 1.00 6.00 4.00 1.50  Physical (3 items) Mean Std.dev Minimum Maximum Median Change/item  N=7 1.71 6.45 - 6.00 13.00 0.00 0.57  N=17 2.24 6.98 -14.00 13.00 4.00 0.75  N=6 8.00 8.32 - 2.00 17.00 9.00 2.7  Adl's (3 items) Mean Std.dev Minimum Maximum Median Change/item  N=7 1.14 4.52 - 7.00 7.00 2.00 0.40  N=16 2.25 5.42 -10.00 11.00 2.50 0.70  N=6 8.33 6.68 - 3.00 15.00 10.50 2.80  Social (2 items) Mean Std.dev Minimum Maximum Median Change/item  N=10 1.29 2.90 - 3.00 6.00 2.00 0.64  N=12 2.33 4.85 - 7.00 9.00 3.0 1.17  N=6 5.33 3.20 2.00 10.00 4.5 2.70  N=9 6.56 7.37 - 5.00 17.00 6.00 1.30  *the minimal clinically important difference expressed as change/item is displayed in red  3.2.6 Criterion Validity According to Streiner and N o r m a n  69  the traditional definition of criterion validity is the  correlation of a scale with some other measure of the trait or disorder under study, ideally, a gold standard which has been used and accepted in the field. In the absence of a gold standard, criterion validity can be conducted using a "psuedo" gold standard, however, the results may not be conclusive. N o gold standard, meeting the above definition, exists for esophageal carcinoma. Knowing the difficulty using a "psuedo" gold standard, we decided to proceed with criterion validity using the SF-36 Health Survey. Although, to my knowledge, it has not been validated on patients suffering from esophageal cancer, it has demonstrated, repeatedly, a high response rate and excellent test-retest reliability.  75  The Pearson Correlation Coefficient for the corresponding  domains of the E Q O L and the SF-36 are illustrated i n Table 17. W e used the following scale to categorize correlations for testing the criterion validity:  Strongly correlated  r>0.5  Moderately correlated  r = .35 to .5  Poorly correlated  r = .20 to .35  The following domains were compared: physical function domain of both the E Q O L and the SF-36; activities of daily living domain of the E Q O L and the role function domain of the SF-36; emotional function domain of the E Q O L and the emotional role function of the SF-36; and the social function domain of both the E Q O L and the SF-36. Sizes of the domains varied between the two questionnaires as illustrated in Table 18.  Table 17 Criterion Validity - Correlations Over Time Domain Physical Function Activities of Daily L i v i n g Emotional Function Social Function  Baseline  1 Month  3 Month  6 Month  .57 .46 .56 .65  .57 .35 .45 .75  .74 .39 .32 .77  .76 .54 .44 .84  70  A t the baseline administration of the questionnaires, all but the activities of daily living domain are highly correlated. A t one month after treatment has begun, the physical function domain and the social function domain both have strong correlation. The activities of daily living and the emotional function domains are moderately correlated. A t three months, activities of daily living and the emotional function have not increased in correlation, with activities of daily living being moderately correlated, and the emotional function domain decreasing from .45 to .32 to poor correlation. However, both the physical function domain and the social function domain are highly correlated. Finally, at the six month administration of the questionnaires, all domains are either moderately (emotional function) or strongly (physical, A D L , social) correlated.  Table 18 Number of Domain Questions in EQOL and SF-36 Questionnaire  Physical  ADL/Role  Emotional  Social  EQOL SF-36  3 10  3 4  2 3  2 2  3.2.7  Construct Validity  Four a priori predictions were developed to assess construct validity. Correlations are assessed as listed under criterion validity with r>.5 indicating a strong correlation between the E Q O L and the construct being tested; r between .35 to .5 indicating a moderate correlation; and r between .20 to .35 indicating a poor correlation.  3.2.7.1 Dysphagia 1. improvement in dysphagia from baseline to one month follow-up in all treatment groups One question in the E Q O L addressed dysphagia and asked patients to indicate the amount of distress or discomfort due to solid food sticking by using a seven point Likert scale: 1 = a very great deal of distress or discomfort 7 = no distress or discomfort When looking at all treatment groups combined, the mean raw dysphagia score at baseline for all 65 patients in the study was 4.02 (s.d. 1.95) with a median of 4.0. Forty71  three or 66% of the 65 patients rated their dysphagia as 4 (moderate distress and discomfort) or worse. (Figure 9) O f the 65 patients who completed baseline questionnaires, 47 went on to complete one month questionnaires. The baseline and one month descriptive statistics for these 47 patients are shown in Figure 10. The mean at baseline for these 47 patients was 4.30 (s.d 1.99) and at one month 4.77 (s.d. 1.99) The median at baseline was 4.00, however, at one month the median was 6.0. Only 36% of patients rated their dysphagia as causing moderate or worse distress or discomfort at the one month follow-up, compared to 66% of all patients at baseline. O f the 47 patients who completed both baseline and one month followup questionnaires, 59.6% had rated their dysphagia as moderate or worse at baseline.  Paired f-tests were performed using data from the 47 patients. Table 19 displays data for all groups combined and for each of the three treatment groups. The change in dysphagia score did not reach statistical significance in any of the treatment groups. The paired samples correlation for all groups combined was .188.  Figure 9 Dysphagia Frequency at Baseline for 65 Patients Dysphagia at Baseline N Valid Missing Mean Median Std. Deviation  very great deal of distress 2 3 4 5 6 no distress Total  65 0 4.02 4.00 1.95  Frequency  Percent  8  12.3  8 9 18 6 4 12 65  12.3 13.8 27.7 9.2 6.2 18.5 100.0  Valid Cumulative Percent Percent 12.3 12.3  12.3 13.8 27.7 9.2 6.2 18.5 100.0  24.6 38.5 66.2 75.4 81.5 100.0  72  Figure 10 Dysphagia Frequency at Baseline and One Month for 47 Patients Dysphagia at Baseline N Valid Missing Mean Median Std. Deviation  Frequency at Baseline Frequency Percent *1 2 3 4 5 6 7 Total  5 5 5 13 5 3 11 47  Dysphagia at 1 Month N Valid Missing Mean Median Std. Deviation  47 0 4.30 4.00 1.99  10.6 10.6 10.6 27.7 10.6 6.4 23.4 100.0  Valid Cumulative Percent Percent 10.6 10.6 10.6 21.3 10.6 31.9 27.7 59.6 10.6 70.2 6.4 76.6 23.4 100.0 100.0  1  47 0 4.77 6.00 1.99  I  1 1 11  Frequency at 1 Month Frequency Percent *1 2 3 4 5 6 7 Total  5 3 6 3 6 15 9 47  10.6 6.4 12.8 6.4 12.8 31.9 19.1 100.0  Valid Cumulative Percent Percenl 10.6 10.6 6.4 17.0 12.8 29.8 6.4 36.2 48.9 12.8 80.9 31.9 19.1 100.0 100.0  */ =very great deal;=no 7 distress  Table 19 Mean Change in Dysphagia Score by Treatment Groups Mean  baseline 1 month Mean change Significance (p)  All groups 4.30 (2.0) 4.77 (2.0) 0.47 (2.5) 0.21  Chemoxrt 4.50(1.2) 4.90(1.5) 0.38(1.3) 0.44  Adj/Surg 4.33 (2.3) 4.78 (2.4) 0.44 (3.5) 0.72  Surg Alone 4.23 (2.1) 4.73 (2.1) 0.50 (2.5) 0.29  Although the change in mean dysphagia scores did not show a statistical difference between baseline and one month, it is important to note that the median changed from 4.0 (moderate distress/discomfort) at baseline to 6.0 (very little distress/discomfort) at one month - an improvement of two categories. Another factor affecting the mean scores is the number of patients who dropped out of the study either due to death or progressive disabling disease.  73  I  1  1  3.2.7.2 Chemoradiotherapy  2.i) at one month follow-up, the chemoradiotherapy group will have better scores in the physical function domain than the groups treated with surgical resection  Seven patients received chemoradiotherapy as primary treatment; nine received adjuvant therapy and surgical resection; and 27 received surgical resection alone. Descriptive statistics and paired f-tests for each treatment group in the physical function domain from baseline to the one month follow-up were calculated. Table 20 summarizes the mean change from baseline to one month follow-up. A Stem and Leaf Plot illustrates the mean at one month between treatment groups. (Figure 11).  F i g u r e 11 Stem a n d L e a f Plot - Physical F u n c t i o n at O n e M o n t h 120 100  CO  — I  CO  > >  I CL chemoxrt  adjuvant/surgery  surgery alone  GROUP  The mean change in score in the physical function domain from baseline to one month , for all groups combined was 26.5 (p<0.001); in the group of patients receiving chemoradiotherapy as primary treatment -2.6 (p=0.72); in the adjuvant therapy and  74  surgery group the mean difference was 45.9 (p=0.001); and in the surgical resection alone group the mean change was 27.65 (p<0.001).  Table 20 Physical Function Change Baseline to One Month by Treatment Groups All Groups N Mean Median @ 1 month Mean Change Significant (p value)  43 baseline 85.5 (22.5) 1 month 58.9 (28.0) 26.5 (31.8) < 0.001  Chemoxrt  Adj/Surg  7 83.1 (23.8) 85.7(18.5) 95.0 -2.6 0.72  9 92.0(11.8) 46.1 (30.6) 38.3 45.9 0.001  Surg Alone 27 83.9 (25.0) 56.2 (25.4) 55 27.7 (30.9) < 0.001  ( ) standard deviation  Independent Samples t-tests were performed between the chemoradiotherapy treatment group and the surgery groups comparing mean change i n physical function score at one month follow-up. The mean change i n score in the chemoradiotherapy group was -2.6 (s.d. 19.3) compared to 32.2 (s.d. 30.8) (p=0.006) in the surgery groups combined; in the adjuvant/surgery group 45.9 (s.d. 27.9) (p=0.001); and in the group receiving surgical resection alone 27.6 (s.d. 30.9) (p=0.019). A n Independent Samples  Mest  was also  performed between the chemoradiotherapy group and the surgery groups comparing mean physical function scores at one month. (Table 21) A l l surgically treated groups showed a significant change in scores from the chemoradiotherapy group at one month. As well, the surgical groups combined had worse physical function scores at one month than the chemoradiotherapy group, p=0.001.  Table 21 Mean Physical Function Scores At One Month - Treatment Groups Mean score St. Dev. P-value  Chemo-radiotherapy 83.96 17.85  Surgery Groups 59.92 26.35 O.C101  75  2.U) at one month follow-up, the chemoradiotherapy group will have worse scores in the emotional function domain than the groups treated with surgical resection Descriptive statistics and paired Mests for each treatment group in the physical function domain from baseline to the one month follow-up were calculated. A Stem and Leaf Plot illustrates the mean at one month between treatment groups. (Figure 12). Table 22 summarizes the mean change from baseline to one month follow-up.  Figure 12 Stem a n d L e a f Plot - E m o t i o n a l F u n c t i o n at O n e M o n t h  120 T  GROUP  Table 22 E m o t i o n a l F u n c t i o n Change at O n e M o n t h by Treatment G r o u p s  N Mean  baseline 1 month Median @ 1 month Mean Change Significant (p value)  A l l Groups  Chemoxrt  Adj/Surg  Surg Alone  45 64.9 (26.4) 73.8 (23.3)  8 59.4 (29.0) 76.0 (28.3) 79.2 -16.7 (31.8) 0.182  9 69.4 (26.4) 62.0 (23.2) 58.3 7.4(19.7) 0.292  28 64.6(26.1) 77.1 (21.6) 83.3 -12.5 (25.2) 0.014  -8.9 (27.9) 0.035  () standard deviation 76  Independent Samples Mests were performed between the chemoradiotherapy treatment group and the surgery groups comparing mean change in physical function score at one month follow-up. The mean change in score in the chemoradiotherapy group was -16.7 (s.d. 31.8) compared to -7.6 (s.d. 25.3) (p=0.569) in the surgery groups combined; in the adjuvant/surgery group 7.4 (s.d. 19.7) (p=0.284); and in the group receiving surgical resection alone the mean change was -12.5 (s.d. 25.2) (p=0.705). When all groups are combined, there is a statistically significant improvement in emotional function scores from baseline to one month (p=0.035). The adjuvant/surgery group was the only one to show a decrease in emotional functioning at one month, with a change of 7.4, but it did not reach statistical significance. Although the chemoradiotherapy showed an improved in scores at one month, it was not significant, however, the surgery alone group showed a statistically significant improvement in scores, p=0.014).  A n Independent Samples f-test compared emotional function scores at one month, (Table 23) between the chemoradiotherapy group and the surgical groups combined. Athough the mean score for the surgery groups was lower, it did not reach statistical significance, p=0.812).  Table 23 Mean Score At One Month Emotional Function By Treatment Group Mean score St. Dev. P-value  Chemo-radiotherapy 76.04 28.32  Surgery Groups 73.42 22.64 0.8.12  3.2.7.3 Stage of Disease 3. Patients with Stage III or Stage IV disease at baseline have worse scores at three months and six months post treatment (due to likelihood of recurrence) than patients with Stage I or Stage II disease at baseline, in all domains - symptom, emotional, physical, activities of daily living, social Seven patients had Stage I disease, 15 had Stage II disease (13 had Ha, 2 had lib), 16 had Stage III disease and four had Stage 4 disease. (Figure 13). .  77  Figure 13 Stage of Disease  Stage o fD i s e a s e N=42  Symptom Domain Paired r-tests were performed between the baseline and three month follow-up. Mean symptom scores across all stages of disease were, at baseline 25.8 (s.d. 17.7), and at three month 26.0 (s.d. 17.8), p=.94. Table 24 shows the mean symptom change from baseline to three months by early and late stage disease. Paired Mests were also performed within the early stage group and within the late stage group. Within the early stage group, the mean symptom baseline score was 20.30 (s.d. 17.61) and the mean three month score was 24.55 (s.d. 19.46), with a mean change in score between the two time intervals of -4.25 (s.d. 19.75), p=0.324. Within the late stage group, the mean baseline symptom score was 33.01 (s.d. 15.22) and the mean three month score was 27.83 (s.d. 16.70), p=0.255. Figure 14 is a stem and leaf plot showing the difference between early and late stage symptom scores at three months.  An Independent Samples r-Test was conducted to compare the mean change in symptom domain at three months between the early and late stage patients. The mean change in the early stage patients was 4.25 (s.d. 19.75), which differs from Table 26 as the sample changed by one. The mean change in the late stage group of patients was 5.18 (s.d. 1975). There was no statistical difference between the two means. However, it is useful 78  to note that the median at three months in the early stage group was 18.35 compared to 26.84 in the late stage group.  T a b l e 24 S y m p t o m Change Baseline T o Three M o n t h s - E a r l y V s L a t e Stage  A l l Stages 43 25.81 (17.68) 26.04 (17.85)  N Mean  baseline 3 months Median @ 3 months Mean Change @ 3 months Significance (p-value)  -2.30 (20.08) 0.940  Early (Stage 1,2) 22* 20.30(17.61) 24.55 (19.46) 18.35 -4.25 (19.75) 0.240  Late (Stage 3,4) 20 33.01 (15.22) 27.83 (16.70) 26.84 5.18(19.75 0.255  ( ) standard deviation, *one patient was not staged due to tumor sterilization  Figure 14 Stem a n d L e a f Plot Symptoms at 3 M o n t h s E a r l y V s L a t e Stage  80  60  40  20  O  2  o  CO  CO -20 N =  22  early stage (1,2)  20  late stage (3,4)  STAGE2  An Independent Samples /-test comparing the symptom scores at three months was also conducted (Table 25). The mean symptom score in the late stage group was worse than the early stage group, but it was not statistically significant, p=0.442. 79  Table 25 Symptom Score At Three Months Early Vs Late Stage  Mean Score St. dev. Significance (p-value)  Early Stage 24.93 19.10  Late Stage 29.20 17.45 0.4142  Paired r-tests were also performed between the baseline and six month follow-up. Mean symptom scores across all stages of disease were, at baseline 26.17 (s.d. 17.70), and at six months 24.50 (s.d. 17.37), p=.583. Table 26 shows the mean symptom change from baseline to six months by early and late stage disease. A stem and leaf plot illustrates the differences in mean six month symptom scores between the early and late stage groups.  (Figure 15)  Within the early stage group, the mean six month symptom score was 21.51 (s.d. 19.51) with a mean change between baseline and six months of -1.37 (s.d. 18.77), p=0.736. Scores for the early stage group were slightly worse at one month than at baseline. Within the late stage group, the mean six month symptom score was 28.82 (s.d. 14.13) with a mean change between intervals of 6.49 (s.d. 19.38), p=0.186. These scores were lower than at baseline indicated less symptomatology. However, the early stage mean, as seen above in the stem and leaf plot, remains lower (less symptomatology) than the late stage group. Higher scores in the symptom domain indicate higher symptomatology or worse symptoms.  Table 26 Symptom Change Baseline To Six Months - Early Vs Late Stage N Mean  baseline 6 months Median @ 6 months Mean Change @ 6 months Significance (p-value)  A l l Stages 40 26.17 (17.70) 24.50(17.37) 1.67 (19.10) 0.583  Early (Stage 1,2) 22* 20.14(17.83) 21.51 (19.51) 18.35 -1.37 (18.77) 0.736  Late (Stage 3,4) 17 35.31 (13.24) 28.82 (14.13)) 30.00 6.49 (19.38) 0.186  () standard deviation *one patient was not staged due to tumor sterilization 80  F i g u r e 15 Stem a n d L e a f Plot Symptoms at 6 M o n t h s E a r l y V s L a t e Stage  An Independent Samples f-Test was conducted to compare the mean change in symptom domain at six months between the early and late stage patients. The mean change in the early stage patients was 1.37 (s.d. 18.77). The mean change in the late stage group of patients was -6.48 (s.d. 19.38). There was no statistical difference between the two means. Again, it is useful to note that the median for the early stage group at six months is 18.35 and for the late stage group 30.00.  An Independent Samples r-test between the mean symptom scores at six months for early and late stage disease was also performed. (Table 27) Again, although the mean symptom scores at six months were worse in the late stage group, it was not statistically significant.  81  Table 27 S y m p t o m Scores A t S i x M o n t h s E a r l y V s L a t e Stage  Late Stage 28.82 14.13 0.183  Early Stage 21.51 19.51  Mean Scores St. dev. Significance (p-value) Physical Function Domain  Paired r-tests were performed between baseline and three month scores for all stages of disease combined. Mean baseline score was 85.98 (s.d. 22.23) and at three months 70.70 (s.d. 24.94). The mean difference in scores between the time intervals was 15.27 (s.d. (27.21), p=0.001. (Table 28) A stem and leaf plot illustrates the difference in means between the early and late stage groups at three months in the physical function domain. (Figure 16) The mean score in the early stage group is higher indicating better physical functioning than the mean score in the late stage group. Figure 16 Stem a n d L e a f Plot Physical F u n c t i o n at 3 M o n t h s E a r l y V s L a t e Stage 120  1  1  •  •  •  N =  22  20  early stage (1,2)  late stage (3,4)  stage  82  Paired r-tests were also performed between baseline and three months for the early stage disease group and between baseline and three months for the late stage disease group. In the early stage group the mean difference from baseline to three months was 17.35 (s.d. 23.35), p=0.002; and in the late stage group the mean difference was 13.16 (s.d. 31.68), p=0.08.  Table 28 Physical Function Change Baseline To Three Months - Early Vs Late Stage N Mean  baseline 3 months Median @ 3 months Mean Change @ 3 months Significance (p-value)  A l l Stages 43 85.97 (22.23) 70.70 (24.94) 15.27 (27.21) 0.001  Early (Stage 1,2) 22 90.37 (15.30) 73.02 (25.79) 78.30 17.35 (23.75) 0.003  Late (Stage 3,4) 20 80.42 (27.72) 67.26 (24.59) 64.20 13.16(31.68) 0.080  () standard deviation A n Independent Samples r-test was performed between the change in scores at three months for the early and late stage groups but it did not reach statistical significance (p=0.628).  A n Independent Samples t-test comparing the symptom scores at three months was also conducted (Table 25) but was not statistically significant. The late stage group had scores slightly lower than the early stage group.  Table 29 Physical Function Scores At Three Months Early Vs Late Stage  Mean Scores St. dev. Significance (p-value)  Early Stage 69.84 29.44  Late Stage 64.05 28.10 0.51  A Paired Mest was performed between baseline and six month scores for all stages of disease combined (Table 30). The mean baseline score for the physical function domain  83  was 85.21 (s.d. (22.42) and the mean six month score was 71.63 (s.d. 25.24). The mean difference was 13.85 which was statistically significant (p=0.004). A stem and leaf plot illustrates the difference between the six month physical function score in the early stage group, which is higher, indicating a higher physical function than in the late stage group. (Figure 17)  F i g u r e 17 Stem a n d L e a f Plot Physical F u n c t i o n at 6 M o n t h s E a r l y V s L a t e Stage 120 i  1  100 H  •  •  •  N =  22  17  early stage (1,2)  late stage (3,4)  stage  Paired r-tests were also performed between baseline and six months in the early stage group and between baseline and six months in the late stage group (Table 30). The mean baseline score for the early stage group was 92.85 (s.d. 9.44) and the mean six month score for the early stage group was 83.68 (s.d. 19.08). The mean difference was 9.17 (s.d. 16.32), p=0.056. For the late stage group, the mean baseline score was 77.65 (s.d. 28.41) and the mean six month score was 62.36 (s.d. 25.52). The mean difference wasl5.29, p=0.84.  84  Table 30 Physical Function Change Baseline To Six Months - Early Vs Late Stage N Mean  baseline 6 months Median @ 6 months Mean Change @ 6 months Significance (p-value)  A l l Stages 40 85.21 (22.42) 71.63 (25.24) 13.58 (27.81) 0.004  Early (Stage 1,2) 22* 90.37 (15.30) 77.72 (23.51) 85.80 12.66 (23.22) 0.018  Late (Stage 3,4) 17 77.65 (28.41) 62.36 (25.52) 71.70 15.29 (34.17) 0.084  () standard deviation * one patient was not staged due to tumor sterilization  A n Independent Samples Mest was performed using the change in score between the early and late stage disease groups. The mean difference in the early stage was 12.66 and in the late stage 15.29, p=.221.  In addition, an Independent Samples t-test comparing the mean physical function scores at six months was performed (Table 31) and approached statistical significance (p=0.059).  Table 31 Physical Function Change At Six Months Early Vs Late Stage  Mean Scores St. dev. Significance (p-value)  Early Stage 77.72 23.51  Late Stage 62.37 25.52 0.063  Emotional Function Domain Paired r-tests were performed between baseline and three month scores for all stages of disease combined. Mean baseline score was 65.50 (s.d. 25.56) and at three months 74.81 (s.d. 23.75). The mean difference i n scores between the time intervals was -9.30 (s.d. (32.85), p=0.070. (Table 32)  85  Table 32 Emotional Function Change Baseline To Three Months By Stage N Mean  baseline 3 months Median @ 3 months Mean Change @ 3 months Significance (p-value)  A l l Stages 43 65.50 (25.56) 74.81 (23.75) -9.30 (32.85) 0.070  Early (Stage 1,2) 22* 72.35 (21.73) 76.13(27.74) 83.30 -3.78 (24.08) 0.469  Late (Stage 3,4) 20 56.25 (27.15) 72.09(18.79) 75.00 -15.84 (40.73) 0.098  () standard deviation, * one patient was not staged due to tumour sterilization Paired t-tests were also performed between the baseline and three month mean scores i n the early stage group and between the baseline and three month mean scores in the late stage group. Twenty-two patients were in the early stage group. The mean baseline emotional function score was 72.35 (s.d. (21.73) and the mean three month emotional function score was 76.13 (s.d. 27.74) with a mean change o f - 3 . 7 8 (s.d 24.08), p=0.469. Twenty patients were in the late stage group. The mean baseline emotional function score was 56.25 (s.d. 27.153) and the mean three month emotional function score was 72.09 (s.d. 18.79). The mean difference was -15.84 (s.d. 40.73), p=0.098. A stem and leaf plot illustrates the difference between emotional function scores at three months for the early and the late stage groups. (Figure 18) The early stage group has a slightly higher emotional function at three months than the late stage group.  A n Independent Samples t-test was performed between the change at three months for the early vs late stage groups. The mean differences are -3.78 for the early stage group and -15.84 for the late stage group, p=0.060.  A n Independent Samples t-test comparing mean emotional scores at three months between early vs late stage (Table 33) disease was not statistically significant, p=0.82. The mean scores in the late stage group were only slightly worse than in the early stage group.  86  Figure 18 Stem a n d L e a f Plot E m o t i o n a l F u n c t i o n 3 M o n t h s E a r l y V s Late Stage 120  UJ  0J N =  ,  ,  22  I  20  early stage (1,2)  late stage (3,4)  STAGE2 Table 33 E m o t i o n a l F u n c t i o n Scrores A t Three M o n t h s E a r l y V s Late Stage  Mean Score St. dev. Significance (p-value)  Early Stage 72.82 31.41  Late Stage 71.03 18.93 0.822  Paired /-tests were also performed between baseline and six month scores for all stages of disease combined. Mean baseline score was 67.71 (s.d. 24.91) and at six months 73.96 (s.d. 25.33). The mean difference in scores between the time intervals was -6.25 (s.d. (28.54), p=0.1.74. (Table 34)  Paired f-tests were also performed within the early stage group between baseline and six month emotional function scores, and within the late stage group between baseline and six month emotional function scores. Within the early stage group, the mean baseline score was 72.78 (s.d. 23.03) and the mean six month emotional function score was 78.41 (s.d. 25.92), with a mean change of-5.69 (s.d. 23.61), p=0.271. Within the late stage group, the mean baseline emotional function score was 59.31 (s.d. 25.32) and the mean 87  Table 34 E m o t i o n a l F u n c t i o n Change Baseline T o Six M o n t h s B y Stage  N Mean  baseline 6 months Median @ 6 months Mean Change @ 6 months Significance (p-value)  A l l Stages 40 67.71 (24.91) 73.96 (25.33) -6.25 (28.54) 0.174  Early (Stage 1,2) 22 72.78 (23.03) 78.41 (25.92) 87.50 -5.69 (23.61) 0.271  Late (Stage 3,4) 17 59.31 (25.32) 66.67 (23.79) 75.00 -7.35 (35.35) 0.404  () standard deviation * one patient was not staged due to tumour sterilization  six month score was 33.37 (s.d.23.79), p=0.404. Figure 19 is a stem and leaf plot illustrating the higher emotional function in the early stage group at six months over the late stage group.  F i g u r e 19 Stem a n d L e a f Plot E m o t i o n a l F u n c t i o n at 6 M o n t h s E a r l y V S L a t e Stage 120  i  1  •  100  80  60  40  20  early stage (1,2)  late stage (3,4)  STAGE2  An Independent Samples f-test was performed between the early and late stage change in scores at six months. As reported above, the mean change at six months in the early stage group was -5.69 (s.d. 23.61) and in the late stage group -7.35 (s.d. 35.35), p=0.162.  88  A n Independent Samples /-test comparing mean emotional function scores at six months (Table 35) between early and late stage disease was not statistically significant, p=0.15 However, the late stage group did have lower scores than the early stage group.  Table 35 Emotional Function Scores At Six Months Early Vs Late Stage  Mean Score St. dev. Significance (p-value)  Early Stage 78.14 25.92  Late Stage 66.67 23.58 0.148  Activities of Daily L i v i n g ( A D L ' s ) Paired /-tests were performed between baseline and three months follow-up for all stages combined. The mean baseline A D L score was 66.71 (s.d. 28.18) and the mean three month score was 70.89 (s.d. 27.95). The mean difference was -4.18, p=0.394. (Table  36)  Paired /-tests were also performed within the early stage group between baseline and three month A D L scores, and within the late stage group between baseline and three month A D L scores. Within the early stage group, the mean baseline score was 73.19 (s.d. 24.47) and the mean three month A D L score was 77.95 (s.d.24.52), with a mean change of -4.16 (s.d.29.21), p= 0.511. Within the late stage group, the mean baseline A D L score was 57.25 (s.d. 29.70) and the mean three month score was 62.50 (s.d.30.36), p= 0.518. Figure 20 is a stem and leaf plot illustrating the difference between early and late stage disease. The early stage group has a higher mean score i n activities of daily living than the late stage group at three months.  Table 36 Activities of Daily Living Change Baseline To Three Months By Stage N Mean  baseline 3 months Median @ 3 months Mean Change @ 3 months Significance (p-value) () standard deviation  A l l Stages 43 66.71 (28.18) 70.89 (27.95) -4.18(31.82) 0.394  Early (Stage 1,2) 22* 73.19(24.47) 77.95 (24.52) 85.80 -4.16(29.21) 0.511  Late (Stage 3,4) 20 57.25 (29.70) 62.50 (30.36) 58.35 -5.25 (35.66) 0.518  * one patient was not staged due to tumour sterilization  89  F i g u r e 20 Stem a n d L e a f Plot - A D L 3 M o n t h E a r l y V s . L a t e Stage 120  An Independent Samples /-test was performed between the early and late stage change in scores at three months. As reported above the mean change at three months in the early stage group was -4.16 (s.d. 29.21) and in the late stage group -5.25 (s.d. 35.66), p=0.914.  An Independent Samples /-test comparing mean A D L scores at three months between early and late stage disease (Table 37) reached statistical significance (p=0.038). Table 37 Activities of Daily L i v i n g Scores A t Three M o n t h s E a r l y V s Late Stage  Mean Scores St. dev. Significance (p-value)  Early Stage 78.72 23.98  Late Stage 60.31 31.23 O.C•38  Paired /-tests were performed between baseline and six month follow-up for all stages combined. (Table 38) The mean baseline A D L score at six months was 67.08 (s.d. 28.41) and the mean six month score was 73.21 (s.d. 25.91). The mean difference was -6.13, p=0.319. 90  . Table 38 Activities of D a i l y L i v i n g Change Baseline T o Six M o n t h s B y Stage N Mean  baseline 6 months Median @ 6 months Mean Change @ 6 months Significance (p-value)  A l l Stages 40 67.08 (28.41) 73.21 (25.91) -6.13 (38.36) 0.319  Early (Stage 1,2) 22* 75.08 (23.90) 77.27 (26.12) 85.80 -2.19 (33.55) 0.767  L  L a t e (Stage 3,4) 17 54.79 (30.03) 67.06 (25.77) 71.70 -12.27 (44.39) 0.271  () standard deviation * one patient was not staged due to tumour sterilization Paired Mests were also performed within the early stage group between baseline and six month A D L scores, and within the late stage group between baseline and six month A D L scores. Within the early stage group, the mean baseline score was 75.08 (s.d. 23.90) and the mean six month A D L score was 77.27 (s.d.26.12), with a mean change of -2.19 (s.d.33.55), p= 0.767. Within the late stage group, the mean baseline A D L score was 54.79 (s.d. 30.03) and the mean six month score was 67.06 (s.d.25.77), with a mean change o f - 1 2 . 2 7 (44.39), p= 0.271. F i g u r e 21 is a stem and leaf plot illustrating the difference between early and late stage disease at six months for the A D L domain. The patients with early stage disease had a higher mean activities of daily living score at six months than the patients with late stage disease.  A n Independent Samples Mest was performed between the early and late stage change in scores at six months in the A D L domain. A s reported above the mean change at six months in the early stage group was -2.19 (s.d. 34.21) and in the late stage group -12.27 (s.d. 44.39), p=0.225  A n Independent Samples Mest comparing mean A D L scores at six months between early and late stage disease (Table 39) was not statistically significant. The mean score for the early stage group was 77.27 and for the late stage group it was 67.06. Although there it did not reach statistical significance, the late stage group did have worse scores.  91  F i g u r e 21 Stem and L e a f Plot - A D L at 6 M o n t h s E a r l y V s Late Stage Disease 120  early stage (1,2)  late stage (2,3)  STAGE2  Table 39 Activities of D a i l y L i v i n g Scores A t Three M o n t h s E a r l y V s Late Stage  Mean Score St. dev. Significance (p-value)  Early Stage 77.27 26.12  Late Stage 67.06 25.77 0.231  Social Domain Paired  Mests  were performed between baseline and three month follow-up for all stages  combined. The mean score at baseline was 73.26 (s.d. 28.04) and at three months it was 68.80 (s.d. 32.08). The mean difference was 4.46 (s.d. 32.55), p=0.374. (Table 40) A stem and leaf plot for the social domain at three months illustrates the higher social function in the early stage group over the late stage group. (Figure 22) A n Independent Samples  Mest  was performed for mean change in social scores between the early and late  stage group at three months. The mean difference in score in the early stage group was 9.48 (32.87) and in the late stage group -.84 (32.98), p=0.317.  92  Figure 22 Stem a n d L e a f Plot Social D o m a i n at 3 M o n t h s E a r l y V s L a t e Stage 120  Table 40 Social F u n c t i o n Change Baseline T o Three M o n t h s B y Stage  N Mean  baseline 3 months Median @ 3 months Mean Change @ 3 months Significance (p-value)  A l l Stages 43 73.26 (28.04) 68.80 (32.08) 4.46 (32.55) 0.374  Early (Stage 1,2) 22* 84.10(23.84) 74.62 (31.76) 83.30 9.48 (32.87) 0.191  Late (Stage 3,4) 20 59.99 (27.39) 60.83 (31.66) 58.30 -8.35 (32.98) 0.911  () standard deviation * one patient was not staged due to tumour sterilization  An Independent Samples Mest was also performed comparing the mean social function scores at three months between early and late stage disease, (Table 41) but it did not reach statistical significance. The mean score in the early stage group was 71.27 and in the late stage group 58.73.  93  Table 41 Social Function Scores At Three Months Early Vs Late Stage  Mean Score St. dev. Significance (p-value)  Late Stage 58.73 32.33 0.219  Early Stage 71.37 34.71  Paired /-tests were performed between baseline and six month follow-up for all stages combined. The mean baseline social score at six months was 74.38 (s.d. 27.77) and the mean six month score was 71.68 (s.d. 31.40). The mean difference was 2.70 (s.d. 39.11), p=0..665. (Table 42) Paired Mests were also performed within the early stage group and within the late stage group. The mean social function score at baseline in the early stage group was 83.34 (s.d. 24.67) and the six months 71.68 (31.70). The mean change in score was 2.70 (39.11), p=0.665. In the late stage group, the mean baseline score was 61.27 (s.d. 27.32) and at six months 65.21 (s.d. 30.93) with a mean change in score of -3.94 (s.d. 45.95), p=0.728. A stem and leaf plot illustrates the mean scores for the early and late stage group with the early stage having a higher social function than the late stage group at six months follow-up. (Figure 23)  Table 42 Social Function Change Baseline To Six Months By Stage N Mean  baseline 6 months Median @ 6 months Mean Change @ 6 months Significance (p-value)  A l l Stages 40 74.38 (27.77) 71.68 (31.70 2.70 (39.11) 0.665  Early (Stage 1,2) 22* 83.34 (24.67) 75.38 (31.86) 94.70 7.96 (34.17) 0.287  Late (Stage 3,4) 17 61.27 (27.32) 65.21 (30.93) 75.00 -3.94 (45.95) 0.728  () standard deviation, *one patient was not staged due to tumour sterilization  A n independent Samples Mest was performed comparing the mean change in scores between the early stage group and the late stage group. The mean change in social function scores at six months in the early stage group was 7.96 (s.d.34.17) and in the late stage group -3.94 (s.d.45.95), p=0.075.  94  F i g u r e 23 Stem a n d L e a f Plot Social F u n c t i o n at 6 M o n t h s E a r l y V s L a t e Stage 1  1201  tills  22  17  early stage (1,2)  late stage (3,4)  STAGE2 An Independent Samples f-test comparing six month mean social function scores (Table 43) was not statistically significant, although the mean for the early stage group was higher at 75.38 than the late stage group at 65.21.  Table 43 Social F u n c t i o n Scores A t Six M o n t h s E a r l y V s L a t e Stage  Mean Scores St. dev. Significance (p-value)  Early Stage 75.38 31.86  Late Stage 65.21 30.93 0.322  3.2.7.4 S y m p t o m R a t i n g 4. For patients who rated their symptoms as mild at the baseline visit, a better score is predicted at both baseline and at three month follow-up than patients who rated their symptoms as severe at the baseline visit, in all domains - symptoms, physical function, emotional function, activities of daily living and social function This prediction was based on the knowledge that severe symptoms are associated with late stage disease. Our goal was to see if the E Q O L could distinguish between mild and 95  severe symptoms of the disease itself, and not necessarily from treatment, which has its own associated symptoms or side effects, hence three months was the time period we chose. Twenty-eight patients rated their symptoms as mild at the baseline visit, 23 as moderate, and 13 as severe.  Physical Function Domain Mean, standard deviation and median for the physical function domain are displayed in Table 44 for both the baseline and the three month follow-up visits.  Table 44 Baseline to 3 Month Mean Physical Function Scores by Symptom Severity  N Mean St.dev. Median  *  Moderate Mild 3 Month Baseline 3 Month Baseline 17 21 23 28 63.92 85.14 92.68 72.61 19.61 27.46 12.98 31.16 66.70 100.00 83.30 88.30 p=0.007 Baseline score mild vs severe p=0.382 Three month score mild vs severe  Severe 3 Month Baseline 7 13 63.07* 61.19** 32.75 22.88 61.70 66.70  Figure 24 and Figure 25 illustrate the mean physical function scores by symptom severity at baseline and three month follow-up.  A l l three categories of symptoms - mild, moderate and severe have lower scores indicating a worsening of physical function. However, the group who rated their symptoms as severe had consistently lower scores than the group who rated their symptoms as mild. The difference between baseline physical function scores is statistically significant (p=0.007) for those patients who rated their symptoms as mild vs those who rated their symptoms as severe. A n Independent Samples t-test showed no difference between the three months scores (p=0.382), however the mean for the severe group remains lower at 61.19 than the mild group at 72.61.  96  F i g u r e 24 Stem a n d L e a f Plot Physical F u n c t i o n Baseline Scores- S y m p t o m Severity 120  I—  28  23  13  mild symptoms  moderate symptoms  severe symptoms  SEVERITY  F i g u r e 25 Stem and L e a f Plot Physical F u n c t i o n 3 M o n t h Scores - S y m p t o m Severity 120  21  17  7  mild symptoms  moderate symptoms  severe symptoms  SEVERITY  97  Emotional Function Domain Mean, standard deviation and median for the physical function domain are displayed in Table 45 for both the baseline and the three month follow-up visits.  Table 45 Baseline to 3 M o n t h M e a n E m o t i o n a l F u n c t i o n Scores by S y m p t o m Severity  Mild Moderate 3 Month Baseline 3 Month Baseline 21 23 17 28 60.14 73.52 80.55 66.67 23.47 23.43 25.69 24.43 68.52 83.30 83.30 58.30 Baseline score mild vs severe p=0.020 Three month score mild vs severe p=0.125  N Mean St.dev. Median * **  Baseline 13 51.28 33.13 58.30  Severe 3 Month 7 63.10 30.39 66.70  Figure 26 and Figure 27 illustrate the mean emotional function scores by symptom severity at baseline and three month follow-up.  Figure 26 Stem and L e a f Plot E m o t i o n a l F u n c t i o n Baseline Scores by S y m p t o m Severity 120  28  23  13  mild symptoms  moderate symptoms  severe symptoms  SEVERITY  98  F i g u r e 27 Stem a n d L e a f Plot E m o t i o n a l F u n c t i o n 3 M o n t h Scores by S y m p t o m Severity 120  100  80  60  40  20  o f2  o  O UJ  -20 N =  21  mild symptoms  17  7  moderate symptoms severe symptoms  SEVERITY  A l l three categories of symptoms - mild, moderate and severe have higher scores indicating better emotional function at the three month follow-up visit. However, the group who rated their symptoms as severe had consistently lower scores than the group who rated their symptoms as mild. The difference between baseline emotional function scores is statistically significant (p=0.020) for those patients who rated their symptoms as mild and those who rated their symptoms as severe. The difference between the three months scores is not statistically significant (p=0.125), however the mean for the severe group remains lower at 63.10 than the mild group at 80.55. Activities of Daily Living Domain Mean, standard deviation and median for the physical function domain are displayed in Table 46 for both the baseline and the three month follow-up visits.  99  Table 46 Baseline to 3 M o n t h M e a n Activities of D a i l y L i v i n g Scores by S y m p t o m Severity Mild 3 Month Baseline  21 72.21 30.77 83.30  Moderate 3 Month Baseline  Severe 3 Month Baseline  23 57.32 26.19 61.70  13 35.13 28.93 28.30  Mean St.dev. Median  28 79.35 17.95 80.80  * **  Baseline score mild vs severe Three month score mild vs severe  N  17 67.84 26.15 71.70  7 68.34 31.02 71.70  p <0.001 p= 0.776  Figure 28 a n d F i g u r e 29 illustrate the mean activities of daily living scores by symptom severity at baseline and three month follow-up.  Figure 28 Stem a n d L e a f Plot Activities of D a i l y L i v i n g Baseline Scores by S y m p t o m Severity 120 100  LU CO  <  CO _l Q < 28  mild symptoms  23  13  moderate symptoms severe symptoms  SEVERITY  100  F i g u r e 29 Stem a n d L e a f Plot Activities of D a i l y L i v i n g 3 M o n t h Scores by S y m p t o m Severity 120 T  o 2  CO _li  <  1  40 •  ^ ^ H ^ ^ B  20 •  |  — I —  o.  "20  |  J  ,  .  .  N=  21  17  7  mild symptoms  I  moderate symptoms severe symptoms  SEVERITY  Patients who indicated that they had mild symptoms at baseline actually saw a decrease of 7.14 in the mean activities of daily living score at three months follow-up. Both the moderate and the severe group increased their mean scores at three months, with the moderate group increasing by 10.56 and the group who rated their symptoms as severe at baseline had an increase of 33.21 in their activities of daily living at three months. The difference between baseline activities of daily living scores is statistically significant (p<0.001) for those patients who rated their symptoms as mild and those who rated their symptoms as severe. The difference between the three months scores is not statistically significant (p=0.776), however the mean for the severe group remains lower at 68.34 than the mild group at 72.21 at the three month follow-up visit. Social Function Domain Mean, standard deviation and median for the social function domain are displayed in Table 47 for both the baseline and the three month follow-up visits.  101  Table 47 Baseline to 3 M o n t h M e a n Social F u n c t i o n Scores by S y m p t o m Severity  Moderate Mild 3 Month Baseline 3 Month Baseline 21 23 17 28 75.80 66.67 61.27 84.23 17.77 33.94 26.82 32.13 91.70 100.00 66.70 66.70 Baseline score mild vs severe f ) =0.008 Three month score mild vs severe p= 0.077  N Mean St.dev. Median * **  Severe 3 Month Baseline 13 7 48.80 49.35 32.43 39.17 33.30 41.70  Figure 30 and Figure 31 illustrate the mean social function scores by symptom severity at baseline and three month follow-up.  Figure 30 Stem a n d L e a f Plot Social F u n c t i o n Baseline Scores by S y m p t o m Severity 120  N-  28 mild s y m p t o m s moderate  severe symptoms  symptoms  S E V E R I T Y  102  Figure 31 Stem a n d L e a f Plot Social Function 3 M o n t h Scores by S y m p t o m Severity 120 100  o co O O  00 N =  21  mild symptoms  17  moderate symptoms  7  severe symptoms  SEVERITY  A l l categories of baseline symptom classification saw a decrease in mean social function scores at three months. The difference between baseline social function scores is statistically significant (p<0.008) for those patients who rated their symptoms as mild and those who rated their symptoms as severe. The difference between the three months scores is not statistically significant (p=0.077), however the mean for the severe group remains lower at 48.80 than the mild group at 75.82 at the three month follow-up visit.  Symptom Domain Mean, standard deviation and median for the symptom domain are displayed in Table 48 for both the baseline and the three month follow-up visits. Table 48 Baseline to 3 M o n t h M e a n S y m p t o m Scores by S y m p t o m Severity  N Mean St.dev. Median  **  Moderate Mild 3 Month 3 Month Baseline Baseline 17 21 23 28 26.85 26.03 32.75 18.80 15.60 15.79 21.38 13.91 23.30 23.30 33.30 21.65 Baseline score mild vs severe p <0.001 Three month score mild vs severe p= 0.690  Severe 3 Month Baseline 7 13 45.64 29.53 13.80 20.39 26.70 50.00  103  Figure 32 and Figure 33 illustrate the mean symptom scores by symptom severity at baseline and three month follow-up.  Figure 32 Stem a n d L e a f Plot S y m p t o m Baseline Scores by S y m p t o m Severity  UJ  10  23  mild symptoms  13  moderate symptoms severe symptoms  SEVERITY  Scores in the symptom domain are opposite to the other domains of the EQOL. A higher score in the symptom domain indicates higher symptomatology or worsening of symptoms. Moderate and severe baseline symptom classification saw a decrease in mean symptom scores, indicating less symptomatology, at the three month visit than at the baseline visit. However, those who classified their symptoms as mild at baseline actually saw an increase in symptom severity of 7.23 at the three month follow-up visit. The difference between baseline symptom scores is statistically significant (p<0.001) for those patients who rated their symptoms as mild and those who rated their symptoms as severe. The difference between the three months scores is not statistically significant (p=0.690), however the mean for the severe group remains worse (higher) at 29.53 than the mild group at 26.03 at the three month follow-up visit.  104  F i g u r e 33 Stem a n d L e a f Plot S y m p t o m 3 M o n t h Scores by S y m p t o m Severity  SEVERITY  3.3 S U M M A R Y  3.3.1 R E L I A B I L I T Y  The baseline and one week completion of the questionnaires correlated strongly in all domains of the EQOL. Correlations ranged from .733 to .889. These paired samples correlations were statistically significant (p<0.001). Change in scores between baseline and one week were not statistically significant except the physical function domain which had a mean difference of 5.4 (p=0.025). Therefore, since the mean scores between baseline and one week in four of the five domains in the reliability study, did not reach statistical significance, we can consider the E Q O L to have good reliability properties.  105  3.3.2 RESPONSIVENESS  3.3.2.1 Part I - Change In Patients Who Stated They Had Changed Twenty-five patients in the symptom domain stated that their quality of life had changed from baseline to one month; 17 stated that they had changed from one month to three months; and 22 stated that they had changed from three months to six months. The mean difference in scores at each of these time intervals, reached statistical significance, indicating the E Q O L was responsive to change when patients stated they had experienced change.  Twenty-four patients, in the emotional function domain, stated that their qol had changed from baseline to one month; 14 from one month to three months; and 10 from three months to six months. The mean change in score at one month and at three months reached statistical significance, however, not at six months.  Twenty-eight patients, in the physical function domain, stated that their qol had changed from baseline to one month; 19 from one month to three months; and 13 from three to six months. The mean difference in scores was statistically significant at all three time intervals, suggesting that the E Q O L was responsive to change.  Twenty-seven patients indicated a change in health state at one month in the activities of daily living domain; 16 at three months; and 16 at six months. The mean difference in scores from baseline to one month was statistically significant (p=0.17) and from three months to six months (p=0.002). However, the mean difference between one and three months was not statistically significant (p=0.106).  Twenty-five patients in the social domain indicated change at one month; 16 at three months; and 16 at six months. The mean differences in social function all reached statistical significance: 0.016 at one month; 0.010 at three months; and 0.014 at six months, suggesting that the E Q O L was responsive to change.  106  3.3.2.2 Part II - Magnitude of Change Between Stable and Changed Four groups were analyzed: i) changed vs unchanged (stable) None of the mean differences between any of the time intervals, reached statistical significance in the changed vs unchanged group. ii) worse vs unchanged Between those who stated that they were worse and those who stated they hadn't changed, the mean differences at three months in the symptom, physical function and social function domain reached statistical significance (p=0.009; 0.015; 0.002) as well as the six month mean difference in the A D L domain (p=0.060). The remaining 11 time intervals did not have statistically significant mean differences. iii) better vs unchanged In this group, the mean difference at one month in the symptom and social function domain was statistically significant (p=0.012; 0.048) and the three month mean difference in the emotional function domain was statistically significant (p=0.022), with the three month mean difference in the physical function domain approaching significance (p-0.062). The remaining 11 time intervals had no significant mean difference. iv) worse vs better In this group, all but three time intervals were statistically significant. A l l three time intervals in the symptom, A D L and social function domains were statistically significant. The mean difference in the emotional function domain at six months did not reach significance and the one month and six month mean differences in the physical function domain did not reach significance.  3.3.2.3 Part III - Responsiveness Index In a previous paper, Jaeschke et a l , hypothesized that the minimal clinically important 1 9  difference approximated a mean change of 0.5 per item in each domain of the quality of life questionnaire they were testing. Their results showed that the minimal clinically important difference in each of their domains of interest were consistently around 0.5 (0.5, 0.43, 0.64) and they surmised that it would be interesting to test whether these 107  results hold true for other quality of life questionnaires using a seven point likert scale. In the E Q O L the minimal clinically important difference per item was: symptom domain 0.53; emotional function domain 0.50; physical function domain 0.57; activities of daily living domain 0.40; and social domain 0.64. The M C I D for the E Q O L is also consistently around 0.5. Using data from the minimal, moderate and large change we can construct effect sizes for esophageal cancer. A minimal effect is 0.40 to 0.64; a moderate effect is 0.64 to 1.17 and a large effect is 1.3 to 2.8.  3.3.3 Criterion Validity The E Q O L physical function domain was strongly correlated with the SF-36 physical function domain at baseline, one month, three months and six months as were the social domains for the two questionnaires. Activities of daily living in the E Q O L and the role functioning domain of the SF-36 was strongly correlated at six months, moderately correlated at baseline and three months and poorly correlated at one month. The emotional function domains were strongly correlated at baseline, moderately correlated at one and six months and poorly correlated at three months. Only two time intervals out of the 16 had poor correlations, suggesting that the E Q O L does have good criterion validity.  3.3.4 Construct Validity W e made four a priori predictions in this study. The first prediction was that there would be an improvement in dysphagia from baseline to one month in all of the treatment groups. This prediction was based on the premise that most patients who present with this disease w i l l have moderate to severe dysphagia. Since there was only one question on dysphagia we used the raw score, from one to 7. Although the change in dysphagia score from baseline to one month did not reach statistical significance in any of the treatment groups, the score was higher in all groups. A high score of 7 represents no distress/discomfort from dysphagia with a low score of 1 representing a very great deal of distress/discomfort. Therefore, our prediction that there would be an improvement in dysphagia at one month was confirmed.  108  Our second a priori prediction was that patients receiving chemoradiotherapy would have better scores in the physical function domain and worse scores in the emotional function domain than patients receiving surgical resection. A l l three treatment groups had mean physical function scores greater than 80.0 at baseline (chemxrt 83.1, adj/surg 92.0, surg alone 83.9). A t one month only the chemoradiotherapy group had a mean physical function score over 80.0. The mean for this group was 85.7. The mean for the adjuvant/surgery group was 46.1 and for the surgery alone group 56.2. The change in scores from baseline to one month for the surgical groups was statistically significant, but the change in the chemoradiotherapy group was not.  The change in the  chemoradiotherapy group from baseline to one month was -2.6 (the mean scores improved) while the change in the adjuvant therapy plus surgery group was 45.9 (mean scores were worse) and in the surgery alone group, it was 27.7 (mean scores were worse). Therefore, our prediction that the chemoradiotherapy group would have better physical function at one month was confirmed.  The second part of the prediction was that the chemoradiotherapy group would have worse emotional functioning at one month than the surgical groups. This prediction did not hold. In fact, the mean scores in the chemoradiotherapy group improved more from baseline to one month than in the other groups. The chemoradiotherapy group had a mean improvement of 16.7; the surgery groups combined a mean improvement of 7.9; adj/surgery group had a mean change for the worse of 19.7; and the surgery alone group had a mean improvement of 12.5.  The third prediction was that patients with late stage disease (Stage III, IV) w i l l have worse mean scores in all domains at three and at six months than patients with early stage disease (Stage 1,11). The mean scores in the symptom, physical function, emotional function, activities of daily living, and social domains were all worse at both three and six months for patients with late stage disease than for patients with early stage disease. Hence our prediction about late stage disease was confirmed.  109  The final prediction dealt with symptom severity. W e predicted that patients who rated their symptoms as mild at baseline would have better mean scores in all domains, at baseline and three months than patients who rated their scores as severe. In the physical domain all groups - mild, moderate and severe - showed a worsening of scores, but the severe group remained lower than the mild group. In the emotional domain, all groups showed improvement but the severe group remained lower in score than the m i l d group. In the A D L ' s the group with mild symptoms at baseline was worse at three months, while the moderate and severe groups were better. However, the severe group still had lower mean scores than the mild group. In the social domain all the mean scores were worse with the severe group lower than the mild group, and in the symptom domain the mild group had worse mean scores at three months than at baseline and the moderate and severe groups showed improvement. But still, the mild group had better symptomatology than the severe group. Our prediction re symptom severity was confirmed. These results indicate that the E Q O L has good construct validity.  110  CHAPTER FOUR - DISCUSSION AND CONCLUSIONS  4.1 INTRODUCTION This chapter w i l l provide a discussion of the results of this study. First, I w i l l discuss the instrument itself - pre-testing, reliability, responsiveness, and validity. Next I w i l l discuss the study population in terms of gender, employment status, and treatment groups. Other measures of quality of life, quality of life in patients with esophageal carcinoma, and use of the E Q O L w i l l be discussed next. Limitations of the study w i l l follow and then conclusions and future studies.  4.2 PRE-TESTING Five patients participated in the pre-testing of the questionnaire. A l l five patients completed the questionnaires but the results were not used further since the purpose of the pre-testing was to ensure that the wording and concepts were understandable. Surprisingly, the first set of patients in the pre-testing phase all felt that the E Q O L covered the areas of importance to them because of their esophageal cancer. None of the patients had difficulty with the wording or the concepts represented in the questionnaire. Their ease with the wording could be due to the fact that great care was taken in the development phase to ensure that all wording was in lay language, based on experience in previous quality of life studies. For example, the word dysphagia does not appear on the questionnaire, but instead it is described as "solid food sticking". Their ease with the content of the questionnaire is probably due to the fact that only patients with the disease were interviewed in the item reduction phase.  4.3 RELIABILITY • The E Q O L was shown to be reliable in all but the physical function domain, in which the difference between scores at baseline and one week was statistically significant (p=0.025).  Ill  This result is surprising and cannot simply be attributed to the small number of items in the domain. The physical function domain had three items as did the activities of daily living domain. The emotional function and social function domains each had two items and the symptom domain had five items. The statistically significant difference between baseline and the one week administration of the questionnaire could, perhaps, be suggestive that the domain itself is not specific enough and is open to more interpretation than the other domains. For example, one item in the physical function domain deals with the patient's "ability to participate in activities that require physical exertion". A more specific item might be ability to climb a specified number of stairs. Or perhaps, the items in this domain were not as pertinent at this stage of disease as they may be after treatment. The questions raised by the lack of establishing reliability for the physical function domain should be addressed in future studies in which reliability can again be tested.  4.4 RESPONSIVENESS 4.4.1 Responsiveness To Change When Patients State They Have Changed • The symptom domain, physical function domain and the social domain of the E Q O L were responsive to change, using data from patients in Group B (those who stated that they had changed), at all three time intervals - one month, three months and six months. • The emotional function domain was responsive to change at one month and three months, however at six months it was not. • The activities of daily living domain was responsive at the one month and six month visits, but not at the three month visit.  Although the sample size was the same for each domain at a given time interval, the numbers were greatly reduced by the six month follow-up and was, perhaps, not large enough to detect a difference in emotional function.  112  4.4.2 Responsiveness - Magnitude of Change Between Changed and Stable • The magnitude of change between those who stated they had changed and those who stated they had not, was not statistically significant at any visit. The magnitude of change ranged from 0.37 to 9.59 using a score out of 100. • The magnitude of change between those who said they were worse and those who said they had not changed was statistically significant in the symptom domain at three months, in the physical function domain at three months and in the social domain at three months. A t six months in the A D L domain, the magnitude approached statistical significance. The magnitude of change ranged from 0.37 to 46.53. • Between those who stated they were better compared to those who stated they were unchanged, the magnitude reached statistical significance in the symptom domain at one month, in the emotional domain at three months and in the social domain at one month, with the three month visit in the physical domain approaching significance. • Between those who stated they were better and those who stated they were worse, the magnitude of change reached statistical difference in all domains at all visits except in the symptom domain at six months (p=0.061), in the emotional domain at six months and in the physical domain at six months.  The magnitude of change between the group that said they had changed and the group that said they had not changed was not significant.  However, between those who stated  the direction of change, the magnitude was significant. It is possible that with this particular sample, the reality was, that there really was not a big difference between those who stated that they had changed and those who stated that they had not. This population is an older one with a mean age of 62 and a median age of 63, ranging from 29 to 84. Some patients had other age related conditions which could have affected whether they felt that they had changed or not. A s well, this population is quite i l l with high morbidity, which could also be a factor in their subjective opinion of whether they had changed and how much they had changed. However, for those patients who stated that they had changed and the direction of change, the magnitude was significant. One would expect  113  the differences to be greater between these two polar groups i f the instrument was responsive.  4.4.3 Responsiveness Index • The minimal clinically important difference was consistently around 0.5. The minimal effect range was 0.40 to 0.64; the moderate effect range was 0.64 to 1.17; and the large effect range was 1.3 to 2.8.  Although we cannot suggest with certainty, a single best estimate of the minimal clinically important difference, our data does suggest a range within which the M C I D may fall. A s well, we have established ranges that correspond to a moderate change and to a large change. This information w i l l be useful in interpreting quality of life scores for this patient population and it can be used in calculating sample size for new trials, as sample size depends on the magnitude of the difference that investigators consider to be clinically important.  4.5 F A C E A N D C O N T E N T V A L I D I T Y • Good face and content validity was established for the E Q O L . Experts' opinions were sought in this stage and surprisingly, none of those interviewed had changes, rather, all felt that the the content was appropriate.  4.6 C R I T E R I O N V A L I D I T Y • A t baseline, all but the activities of daily living domain are highly correlated. • A t the one month visit, physical function and social function are highly correlated; emotional function and activities of daily living are moderately correlated. • A t the three month visit, physical function and social function are highly correlated; activities of daily living is moderately correlated; and emotional function is poorly correlated. • A t the six month visit, all domains are highly correlated except the emotional function domain which is moderately correlated. 114  Although we used a pseudo gold standard, we were still able to attain moderate to strong correlations between the corresponding domains of the E Q O L and the SF-36 in 14 of the 16 possible time intervals suggesting that the E Q O L does have good criterion validity.  4.7 C O N S T R U C T V A L I D I T Y Construct validity posed a challenge since there are no true clinical indices with which to compare esophageal cancer except recurrence of disease. Recurrence is strongly associated with late stage disease and so we used stage of disease as a surrogate in our third a priori prediction.  • The prediction that there would be an improvement in dysphagia from baseline to one month in all treatment groups was confirmed, although the change did not reach statistical significance. • The prediction that the chemoradiotherapy group would have better scores in the physical function domain at one month than the surgical groups was confirmed and reached statistical significance. • The prediction that the chemoradiotherapy group would have worse emotional function scores than the surgical groups was not confirmed. The chemoradiotherapy and surgery alone group had similar scores at one month. The adjuvant therapy and surgery group combined had a much lower score. • The prediction that patients with Stage III or Stage IV disease at baseline would have worse scores at three months and at six months than patients with Stage I or Stage II disease was confirmed for all domains at both three and six months follow-up. • The prediction that patients who rated their symptoms as mild at baseline would have better scores in all domains at three months than patients who rated their symptoms as severe was confirmed. The five domains of the E Q O L each had three time intervals for a total of 15 intervals. A l l but two of these intervals had a mean difference in score that was statistically significant when comparing the change in scores between time intervals  115  for those patients who stated that they had experienced change. These results indicate that the E Q O L is responsive to change.  4.8 S T U D Y P O P U L A T I O N Patients enrolled in the development phase of the questionnaire were a representative sample of the esophageal cancer population in type of cancer, natural history, symptoms, management problems, treatment and survival. Both squamous cell as well as adenocarcinoma were included. Twenty patients participated in the item generation phase of the study and an additional 38 patients participated in the item reduction phase of the study. O f these 38, 76.3% were male and 23.7% were female. The mean age was 63 (36-82).  W e originally planned to enrol 100 patients in the validation phase of the study. Unfortunately, we were not successful for a couple of reasons. First, we underestimated the number of new cases that would be seen in each of the study centres and second, we had not anticipated the relatively large number of withdrawals from the study. In terms of consent, there were only two patients approached in the Vancouver centre who did not consent to be in the study. A l l other patients consented and were subsequently enrolled. However, the impact of the disease for many of the patients meant that they felt they could no longer participate. Our final enrolment number in this phase was 65. The mean age of the patients was 62.7 (29-84). There were 57 male and 8 female patients which represents the general distribution of this disease.  4.8.1 Gender A separate analysis comparing results of males and females in the study was not conducted. The sample is already small with a total of 65 enrolling in the study. The percentage of females was 12.3% in the validation phase. Although this reflects the reality of the disease distribution, it would be difficult to detect differences between the two population subsets with such small numbers. W e can, however, note that in the item reduction phase, there did not appear to be differences between the items rated as  116  important by males and those rated as important by females. Future studies with a larger sample size could be useful in adding information about any differences in the way that esophageal cancer impacts on gender.  4.8.2 E m p l o y m e n t Status Esophageal cancer, in this sample population, did not impact on patients' employment status. None of the patients in either the development phase of the validation phase of this study were unemployed because of their esophageal cancer.  O f those who enrolled in the development phase 39.5% continued to be employed outside the home, in the validation phase 36.9% were employed. None of the unemployment was due to the disease, but rather due to retirement or by choice (i.e. housewife). This high number of retired patients is not surprising given the mean age of 62.7, but it is interesting to note that among patients who had not yet reached retirement age, none were unemployed because of the disease.  4.8.3 Treatment G r o u p s Three treatment groups were included in each phase - chemoradiotherapy as primary treatment, adjuvant therapy and surgery, and surgical resection alone. These treatment modalities were included in the study as they represent the treatment options available to patients in Canada. The large percentage of patients in the surgery alone group and the small percentage in the chemoradiotherapy group reflect the actual treatment practice across Canada. Patients who receive chemoradiotherapy as primary treatment w i l l often go on to have surgery at a later date, however, that was not the case with the nine patients from that treatment group during this study.  In the development phase of the questionnaire we noted some interesting differences between the treatment groups. A t the one month post-op completion of questionnaires, patients in the chemoradiotherapy group place more importance on items from the emotional function domain, with four out of the top five items for that treatment group  117  coming from the emotional function domain. Impact scores for these four items ranged from 31.3 to 66.7 out of a total of 100. None of the top five items for this group dealt with food or eating. In the adjuvant therapy and surgery group, only one item out of the top five was from the emotional function domain and the impact score was 28.8 out of a possible 100. Three of the items dealt with food or eating. In the surgical resection alone group, one item was from the emotional function domain with an impact score of 26.7 and three dealt with food or eating.  A s well, we noted that the overall impact scores in the item reduction phase differed between treatment groups. The top five items in the chemoradiotherapy group ranged from 31.3 to 66.7, in the adjuvant therapy and surgery group from 28.8 to 41.6, and in the surgery alone group from 22.3 to 33.3. A s we move away from chemotherapy or radiation therapy, the impact scores decrease. These results suggest that there is a higher impact on quality of life with chemoradiotherapy than with the surgically treated groups, and confirms the objective of developing this questionnaire for use in future clinical trials to measure the quality of life associated with the various treatment options.  Severity of disease was also balanced in the validation study population with 28 early stage patients and 33 late stage patients. Corresponding to the early and late stage breakdown, there were 28 patients who classified their symptoms as mild and 37 who classified their symptoms at baseline as severe. M i l d symptoms usually are indicative of early stage disease while severe symptoms are indicative of late stage disease.  Finally, 38 of the 49 surgical patients in the validation phase underwent transhiatal esophagectomy (THE), also representative of thoracic surgery practice across Canada.  4.9 OTHER MEASURES OF QUALITY OF LIFE N o other disease specific quality of life questionnaire currently exists for this patient population. The E O R T C has developed an esophageal cancer module and are in the validation process. Their module was developed using a similar methodology to ours  118  except that it placed a higher emphasis on input from health professionals. In the item reduction phase of their study, they included a number of health professionals. In our study, we gathered input from health professionals in generating the items, but included only patients with esophageal cancer in the item reduction phase since it is the patients' perspective on how their quality of life is affected, that we are interested in measuring.  The E O R T C module was developed using a patient population in England which could have different characteristics than our patient sample. In addition, different treatment modalities are used in England than in Canada and the associated quality of life for each treatment option may differ.  Our instrument, the E Q O L , was designed specifically for potentially curable patients with esophageal cancer, whereas the E O R T C module included patients who had palliative disease in all phases of the study. W e enrolled patients who did not have signs or symptoms of metastatic disease, but it was subsequently discovered in five of the study patients after enrolment.  The two modules have been developed on different populations, with different treatment modalities, for different purposes. It may be useful in future studies, once the E O R T C module is validated, to have patients complete both modules allowing for a comparison of the reliability, responsiveness and validity of the two.  4.10 Q U A L I T Y O F L I F E A N D E S O P H A G E A L  CARCINOMA  In addition to the areas that affect the quality of life of all cancer patients, as measured by the E O R T C Q L Q - C 3 0 , such as feelings of irritability, depression, worrying, tiredness, weakness. Patients suffering from esophageal carcinoma have very specific areas of impact. This patient population has very high morbidity associated with the disease. Not only do many of the patients, by the time they present to the surgeon, have great difficulty swallowing solid food but many can only swallow liquids, hence their quality of life is already greatly impacted by loss of weight, energy loss and concern and worry.  119  Some patients have even progressed to the point of having difficulty swallowing liquids. Although knowledge that dysphagia is the major symptom of this disease is already well known and accepted, the effect that it has on the patient's quality of life is devastating. W e have learned through this study that because of difficulty swallowing, patients withdraw from their normal social lives, they suffer nutritionally, they do not enjoy leisure activities and pastimes that they once did, everything becomes difficult. Other symptoms, such as regurgitation when bending over, has a great affect on their day to day lives - bending to do housework, sport activities such as golf, and gardening- and again, because of these difficulties they tend to simply withdraw.  M a n y of the symptoms of this disease are food-associated. Because of difficulty swallowing, patients are not able to eat the foods that they enjoy. Once treatment begins, they may be able to swallow better, but other associated morbidities come to the fore. Surgically treated patients experience a need to eat smaller meals, are not able to eat sufficient food, and feel full quickly. Patients treated with chemoradiotherapy as primary treatment undergo a much longer period of treatment, up to four months. These patients experience nausea and are unable to eat the foods they like in the quantities that they would like.  Effects of the disease and its treatment permeates patients entire lives. Relationships with family and friends suffer.  4.11 USE OF T H E EQOL The E Q O L in conjunction with the E O R T C Q L Q - C 3 0 should be used in clinical trials comparing treatment modalities. A number of clinical trials have already been conducted, as outlined in the introductory chapter. None of these trials measured quality of life in any formal fashion. The omission of quality of life assessment is a major drawback to the studies, particularly since no statistically significant difference in five year survival has been noted between the various treatments.  120  In a clinical milieu, family physicians may find the E Q O L useful in determining the impact of the disease for their patients. However, once treatment is complete little can be done to change patients quality of life. If future studies discover that certain treatment modalities offer a better quality of life than other treatments, physicians may find the E Q O L useful in determining the type of treatment for their patients.  4.12 L I M I T A T I O N S  4.12.1 Patient N u m b e r s The lower than expected numbers enrolled in this study may have limited the ability to draw definite conclusions regarding the validity of the E Q O L questionnaire. The study numbers were lower than anticipated for the following reasons: • the number of new patients seen by the study surgeons each year were lower than estimated at the beginning of the study • the number of centres involved in the study were lower than expected at the beginning of the study • the number of patients who missed completing questionnaires due to the severity of their illness at various time intervals was unexpected • the number of patients who withdrew completely from the study was unexpected  In order to have met our original projected sample size of 100 would have required a longer enrolment period into the study. Sixty-seven patients were enrolled in the validation phase over a period of three years. In order to reach 100 patients, enrolment would need to continue for an additional year, Taking into account the numbers who withdrew or died from their disease, that period would need to be extended by at least Vi year. Extension of the study was not feasible due to funding deadlines and willingness of centres to participate for a much longer period of time.  Investigators in future quality of life studies, might consider a longer time line for recruitment.  A s well, application for increased resources could provide incentives to 121  new centres joining in the study thereby enabling the study to recruit the projected sample size.  4.12.2 Rate of D r o p o u t Unfortunately, the numbers of patients who dropped out of the study either because of death or increased morbidity associated with their disease, did so at an uneven rate. For example patients in the mild symptom rating group had 25% dropout by the three month visit, in the moderate symptom rating group 26% and in the severe symptom rating group 46% dropout. This large and uneven distribution of patients withdrawing from the study could affect the results since those who drop out or die can be assumed to have worse quality of life than those who remain in the study. Therefore, the results likely underestimate the degree to which quality of life is compromised over time, particularly in the severe symptom group.  4.12.3 C o n f o u n d i n g P r o b l e m In the original study inclusion/exclusion criteria, patients with other conditions affecting their quality of life were excluded. Due to the slower recruitment, this exclusion criteria was changed to exclude patients with other conditions affecting quality of life that are indistinguishable from the effects of their esophageal cancer. A s well, this study had a fairly high proportion of older patients (the mean age was 62.7, st.dev. 11.9), hence a number of patients had age related conditions such as cardiac problems, arthritis, or memory loss. These confounding affects may have decreased the responsiveness of the EQOL.  4.13 C O N C L U S I O N S The present study designed a self-administered quality of life questionnaire for potentially curable patients with esophageal carcinoma. Rigorous validation testing was performed to discover whether the questionnaire was suitable for use in clinical trials.  The E Q O L was found to have face and content validity. It was acceptable to both  122  patients suffering from the disease as well as to health professionals. The questionnaire is in a practical format, is self-administered and can be completed in less than 20 minutes.  The E Q O L was shown to be reliable in all five domains, with paired sample correlations between .733 to .899. The only domain that may be in question in terms of reliability is the physical function domain in which the mean difference was significant at 0.025.  The responsiveness analysis showed that the E Q O L is responsive to change where change exists. A l l but two of the possible 15 time intervals (three time intervals X five domains) had a statistically significant mean difference in score from the previous visit. In addition, when patients stated the direction of their change - better or worse - the change in mean scores between time intervals reached statistical significance in all domains. However, it was not responsive in testing the magnitude of change between those who stated they had changed and those who said they had not experienced change.  The minimal clinically important difference was calculated for each of the five domains and all were consistently around 0.5 which corresponds to earlier studies by other investigators researching quality of life for other medical conditions. Ranges for minimal, moderate and large effect were also established and w i l l be useful in calculating sample size in future clinical studies.  Although we used a psuedo gold standard, good criterion validity was established for the E Q O L . Only two time intervals out of a possible 16 had poor correlations with the SF36.  O f the four a priori predications we made to test for construct validity, only the second part of one prediction did not hold true - in the emotional function domain. The failure to meet our prediction could be due to the E Q O L or it could be due to a misconception on our part in terms of emotional function, therefore, it would be useful to further test the construct validity of the emotional domain in future clinical trials. 123  Overall, this study established that the E Q O L has good face and content validity; good reliability, except in one domain; good responsiveness to change when the direction of change is stated; and relatively good construct validity.  4.14 F U T U R E STUDIES Future studies can be used to further validate the E Q O L in the following areas: • reliability of the physical function domain • magnitude of change between those who state change and those who state no change • construct validity of the emotional domain In addition, a comparison of the quality of life between male and female patients with this disease could be done to ascertain if differences exist due to gender.  Studies are currently underway assessing the various treatment modalities for esophageal cancer with the main outcome being survival. N e w studies are being planned and quality of life should now be included as an additional outcome. The E Q O L should be used to assess this outcome.  124  CHAPTER FIVE - REFERENCES  National Cancer Institute of Canada. Canadian Cancer Statistics 1994. Toronto, 1994. 2. Blot W J , Devesa SS, Kneller R W , Fraumeni JF. Rising incidence of adenocarcinoma of the esophagus and gastric cardia. J A M A 1991;265:1287-9. 3. Devesa SS, Blot W J , Fraumeni JF. Changing patterns in the incidence of esophageal and gastric carcinoma in the United States. Cancer 1998;83(10):20492053. 4. National Cancer Institute of Canada. Canadian Cancer Statistics 2001. Cited2-82001; Available from: Internet Communication http://66.59.133.166/stats/tables/table.html 5. Earlam R, Cunha-Melo JR. Oesophogeal squamous cell carcinoms: II. A critical view of radiotherapy. Br.J.Surg. 1980;67(7):457-61. 6. Launois B, Delarue D, Campion JP, Kerbaol M . Preoperative radiotherapy for carcinoma of the esophagus. Surg.Gynecol.Obstet. 1981;153(5):690-2. 7. C o i a L, Paul A , Engstrom P. Nonsurgical management of esophageal cancer: report of a study of combined radiotherapy and chemotherapy. J C l i n Oncol 1987;5:1783-90. 8. Gignoux M , Roussel A , Paillot B, et al. The value of preoperative radiotherapy in esophageal cancer: results of a study of the E.O.R.T.C. World J.Surg. 1987;ll(4):426-32. 9. Wang M , G u X Z , Y i n W B , Huang G J , Wang L J , Zhang D W . Randomized clinical trial on the combination of preoperative irradiation and surgery in the treatment of esophageal carcinoma: report on 206 patients. Int.J.Radiat.Oncol.Biol.Phys. 1989;16(2):325-7 10. Coia L, Paul A , Engstrom P. Nonsurgical management of esophageal cancer: report of a study of combined radiotherapy and chemotherapy. J C l i n Oncol 1987;5:1783-90. 11. C o i a L R , Engstrom PF, Paul A R , Stafford P M , Hanks G E . Long-term results of infusional 5-FU, mitomycin-C and radiation as primary management of esophageal carcinoma. Int.J.Radiat.Oncol.Biol.Phys. 1991;20:29-36 12. Herskovic A , Martz K, Al-Sarraf M , Leichman L, Brindle J, Vaitkevicius V , et al. Combined chemotherapy and radiotherapy compared with radiotherapy alone in patients with cancer of the esophagus. N E J M 1992;326(24): 1593-98 13. Gelfand G A J , Finley R J , Nelems B, Inculet R, Evans K, Fradet G . Transhiatal esophagectomy for carcinoma of the esophagus and cardia. A r c h Surg 1992;127:1164-8 14. Arnott SJ, Duncan W , Kerr G R , et al. L o w dose preoperative radiotherapy for carcinoma of the oesophagus: results of a randomized clinical trial. Radiother. Oncol. 1992;24(2): 108-13 15. Nygaard K, Hagen S, Hansen H S , et al. Pre-operative radiotherapy prolongs survival in operable esophageal carcinoma: a randomized, multicenter study of L  125  pre-operative radiotherapy and chemotherapy. The second Scandinavian trial in esophageal cancer. World J.Surg. 1992; 16(6): 1104-9 16. Fok M , McShane J, Law S, and W o n g J. Prospective randomised study in the treatment of oesophageal carcinoma. Asian Journal of Surgery 1994; 17 223-9. 17. Walsh T N , Noonan N , Hollywood D, K e l l y A , Keeling N, Hennessy TP. A comparison of multimodal therapy and surgery for esophageal adenocarcinoma. N E J M 1996;335(7):462-7 18. Bosset JF, Gignoux M , Triboulet JP, Tiret E, Mantion G , Elias D, Lozach P, et al. Chemoradiotherapy followed by surgery compared with surgery alone in squamous cell cancer of the esophagus. N E J M 1997;337(3): 161 19. Graham A J , Finley RJ, Clifton JC, Evans K G , Fradet G . Surgical management of adenocarcinoma of the cardia. Amer J Surg 1998 May;175(5):418-21. 20. Finley R J . Adenocarcinoma of the esophagus and esophagogastric junction. In: Esophageal Surgery ( 2 Edition), Churchill Livingston, Editors: F. G . Pearson, J.D. Cooper, J Deslauriers, R.J. Ginsberg, C A . Hiebert, G . Alexander Patterson, H.C. Urschel Jr. 2002;Chapter 45:725-734. 21. Ruol A , Merigliano S, Baldan N . Santi S, Petrin G F , Bonavina L, Ancona E, Peracchia A . Prevalence, management and outcome of early adenocarcinoma ( p T l ) of the esophago-gastric junction. Comparison between early cancer in Barrett's esophagus (type 1) and early cancer of the cardia (type II). Dis Esophagus 1997 Jul;10(3): 190-5. 22. Skinner D. Choices in treating carcinoma of the esophagus. C J S 1989;32(6):420423 23. Finley RJ, Lamy A , Clifton J, Evans K G , Fradet G , Nelems B. Gastrointestinal function following esophagectomy for malignancy. American Journal of Surgery 1995;169:471-475. 24. Earlam R, Cunha-Melo JR. Oesophageal squamous cell carcinoma. I. A critical review of surgery. B r J S u r g 1980;76:381-90. 25. Idem. Oesophageal squamous cell carcinoma. II. A critical review of radiotherapy. B r J S u r g 1980;67:457-61 26. Rosenberg JC, Lichter A S , Leishman L P . Cancer of the esophagus. In: DeVita V T Jr, Hellman S, Rosenberg S A , eds. Cancer: principles and practice of oncology. 3 ed V o l 1 Philadelphia: J.B. Lippincott, 1989:725-64 27. Schlag P M . Randomized trial of preoperative chemotherapy for squamous cell cancer of the esophagus. The Chirurgische Arbeitsgemeinschaft Fuer Onkologie der Deutschen Gesellschaft Fuer Chirurgie Study Group. Arch.Surg. 1992; 127(12): 1446-50. 28. Maipang T, Vasinanukorn P, Petpichetchian C, et al. Induction chemotherapy in the treatment of patients with carcinoma of the esophagus. J.Surg. Oncol. 1994;56(3): 191-7. 29. L a w S, Fok M , Chow S, Chu K - M , W o n g J. Preoperative chemotherapy versus surgical therapy alone for squamour cell carcinoma of the esophagus: a prospectice randomized trial. J.Thorac.Cardiovasc.Surg. 1997;114:210-7. 30. K o k T C , van Lanschot J, Siersema P D , et al. Neoadjuvant chemotherapy in operable esophageal squamous cell cancer: final report of a phase III multicenter nd  r d  126  randomized controlled trial (Meeting abstract). [Abstract] Proc.Annu.Meet.Am.Soc.Clin.Oncol 1997;16:A984. 31. Kelsen D P , Ginsberg R, Pajak TF, et al. Chemotherapy followed by surgery compared with surgery alone for localized esophageal cancer. N.Engl. J.Med. 1998;339(27): 1979-84 32. Bonn D. Medical research council ( M R C ) randomised phase III trial of surgery with or without preoperative chemotherapy in resectable cancer of the oesophagus. Lancet Oncology 2000;May:5. 33. Karnofsky D A , Burchenal J H . The clinical evaluation of chemotherapeutic agents in cancer. In: evaluation of chemotherapeutic agents, C M MacLead, editor. N e w York, Columbia University Press, 1949:191-205. 34. W H O . Constitution of the world helath organization: basic documents. Geneva, Switzerland: W o r l d Health Organization 1948. 35. Priestman TJ, Baum M . Evaluation of quality of life in patients receiving treatment for advanced breast cancer. Lancet 1976;i:899-901. 36. Fitzpatrick R, Ziebland S, Jenkinson C, Mowat A . A generic health status instrument in assessment of rheumatoid arthritis. British Journal of Rheumatol, 1992;31:87-90. 37. Bergner M , Bobbitt R A , Carter W B , Gilson B S . The sickness impact profile: development and final revision of a health status measure. Medical Care 1981;19:787. 38. Torrance G W . Measurement of health state utilities for economic appraisal. J Health Econom 1986;5:1. 39. Bulpitt C J , Fletcher A E . Quality of life in hypertensive patients on different antijypertensive treatments; rationa for methods employed in a multicentre randomized controlled trial. Journal of Cardiovascular Pharmacology 1985;7(suppl): 137-45. 40. Laing M H , Robb-Nichoson C. Health status and utility measurements viewed from the right brain; experience from rheumatic disease. Journal of Chronic Diseases 1987;40:579-83. 41. Guyatt G H , Berman L B , Townsend M , Pugsley A O , Chambers L W . A measure of quality of life for clinical trials in chronic lung disease. Thorax 1987;42:773. 42. Guyatt G H , Mitchell A , Irvine E J , et al. A new measure of health status for clinical trials in irritable bowel disease. Gastroenterology 1989;96:804-10. 43. Schipper H, Clinch J, Powell V . Definitions an conceptual issues. In Quality of Life Assessments in Clinical Trials, B. Spilker, editor. N e w York, Raven Press, 1990; 11-24. 44. Spilker B. Introduction. In Quality of Life Assessments in Clinical Trials. Spilker B, editor. N e w York, Raven Press 1990:3-9. 45. G i l l T M , Feinstein A R . A critical appraisal of quality of life measurements. J A M A 1994;272:619-26. 46. Nelson E, Berwick D. The measurement of health status in clinical practise. Medical Care 1989;27:S77-90. 47. McHorney C A . Generic health measurement: past acomplishments and a measurement paradigm for the 2 1 century. Annals of Internal Medicine 1997;127(8S):743-750. st  127  48. Patrick D L , Bush J W , Chen M M . Toward an operational definition of health. J Health Soc Behav. 1973;14:6-21. 49. Gilson B S , Gilson JS, Bergner M , Bobbit R A , Kressel S, Pollard W E , et al. The sickness impact profile. Development of an outcome measure of health care. A m J Public Health 1975;6:1304-10. 50. Chambers L W , Sacket D L , Goldsmith C H , MacPherson A S , M c A u l e y R G . Development and application of an index of social function. Health Serv Res 1976;11:430-41. 51. Hunt S M , M c E w e n J. The development of a subjective health indicator. Sociol Health Illn. 1980;2:231-46. 52. Dupuy H J . The psychological general well-being ( P G W B ) index. In: Wenger N K , Mattson M E , Furberg C D , Elinson J (eds). Assessment of quality of life in clinical trials of cardiovascular therapies. N e w York: L e Jacq Publishing, 1984:184-188. 53. Ware J E Jr. Scales for measuring general health perceptions. Health Services Research 1976;11(4):396-415. 54. Ware JE, Sherbourne C D . A 36 item short form health survey (SF-36 conceptual framework and item selection). Medical Care 1986;30:473-83. 55. Stewart A L , Hays R D , Ware J E Jr. The M O S Short-Form General Health Survey. Reliability and validity in a patient population. Medical Care 1988;26:724-35. 56. Ware J Jr, Kosinski M , Keller S D . A 12-item short-form health survey: construction of scales and preliminary tests of reliability and validity. Medical Care 1996;34:220-33. 57. Brazier J, Harper R, Jones N M B , O'Cathain A , Thomas K J , et al. Validating the SF-36 health survey questionnaire: new outcome measure for primary care. British Medical Journal 1992;305(6846): 160-4. 58. Aaronson N, et al. The European Organization for Research and Treatment of Cancer Q L Q - C 3 0 : a quality of life instrument for usin in international clinical trials oncology. Journal of the National Cancer Institute 1993;85(5):365-376. 59. Aaronson N K , Bullinger M , Ahmedzai S. A modular approach to quality of life assessment in cancer clinical trials. Recent results. Cancer Research 1988;111:231-249. 60. Aaronson N K , Ahmedzai S, Bullinger M , et al. The E O R T C core quality of life questionnaire: interim results of an international field study. In Effect of Cancer on Quality of Life (Osoba E, ed). Boston C R C Oressk, 1991:285-203. 61. Guyatt G H , Bombardier C, Tugwell P. Measuring disease specific quality of life in clinical trials. C M A J 1986;134:889-895. 62. Juniper EF, Guyatt G H , Jaeschke R. H o w to develop and validate a new healthrelated quality of life instrument. In: Quality of Life and Pharmacoeconomics in Clinical Trials. Edited by Spilker. Lippincott-Raven Publishers, Philadelphia. 1996;49-56. 63. Juniper EF, Guyatt G H , Streiner D L , K i n g D R . Clinical impact versus factor analysis for quality of life questionnaire construction. Journal of Clinical Epidemiology 1997;50(3):233-238.  128  64. Guyatt G H , Townsend M , Berman L B , Keller JL. A comparison of Likert and visual analogue scales for measuring change in function. Journal of Chronic Disease 1987;40(12): 1129-1133. 65. Jaeschke R, Singer J, Guyatt G H . A comparison of seven-point and visual analogue scales: data from a randomized trial. Controlled Clinical Trials 1990;11:43-51. 66. Rust J, Golombok S. Modern psychometrics; the science of psychological assessment. London and new york: Routledge. 1989. 67. Aday JA. Designing and conducting health surveys: a comprehensive guide. Jossey-Bass Publishers ( 2 edition). 1996 68. Guyatt G H , Kirshner B, Jaeschke R. Measuring health status: what are the necessary measurement properties? Journal of Clinical Epidemiology 1992;45(12):1341-1345. 69. Streiner D L , Norman G R . Health measurement scales: a pracitcal guide to their development and use. Oxford University Press. 2 edition. 1995. 70. Hays R D , Anderson R, Revicki D. Psychometric considerations in evaluating health related quality of life measures. Quality of Life Research 1993;2:441-449. 71. Cronbach L J . Coeficient alpha and the internal structure of tests. Psychometrika. 1951;16:297-334. 72. Ware JE. (Ed) The general health rating index. In assessment of Quality of Life in clinical triasl of cardiovascular therapies. Wenger N K , Watson N K , Furberg C D , Elinson J editors) 1984. 73. Feinstein A R . Clinimetrics. New Haven: Conn: Yale University Press. 1987 74. Juniper E F , Guyatt G H , Ferrie PJ, Griffith L E . Measuring quality of life in asthma. A m Rev Respir Dis 1993;147:832-838. 75. V i s i c k A H . A study of the failures after gastrectomy. Annals of the Royal College of Surgeons England 1948;3:266-284. 76. Karnofsky D A , Burchenal J H . The clinical evaluation of chemotherapeutic agents in cancer. In: M a c L e a d C M (ed): Evaluation of chemotherapeutic agents. New York, Columbia University Press, 1949:191-205. 77. Zubrod C G , Schneiderman M , Frei E, Brindley C, G o l d G L , et al. Appraisal of methods for the study of chemotherapy of cancer in man: comparative therapeutic trial of nitrogen mustard and triethylene thoiphosphoramide. Journal of Chronic Disease 1960;11:7-33. 78. M i l l e r A B , Hoogstraaten B, Staquet M , Winkler A . Reporting results of cancer treatment. Cancer 1981;47:207-214. 79. Hutchinson T A , B o y d N F , Feinstein A R . Scientific problems in clinical scales as demonstrated in K P S . Journal of Chronic Disease 1979;32:661-6. 80. Spitzer W O , Dobson A J , Hall J, Chesterman E, L e v i J, et al. Measuring the quality of life of cancer patients. Journal of Chronic Disease 1981;34:40-43. 81. Selby P, Robertson B. Measurement of quality of life in patients with cancer. Cancer Surveys 1987;6:521-543. 82. Schipper H, Clinch J, M c M u r r a y A , Levitt M . Measuring the quality of life of cancer patients. The functional living index-cancer: development and validation. Journal of Clinical Oncology 1984;2:472-483. nd  n d  129  83. Priestman TJ, Baum M . Evaluation of quality of life in patients receiving treatment for advanced breast cancer. Lancet 1976;1:899-901. 84. Gelfand G A . Finley R J . Quality of life with carcinoma of the esophagus. W o r l d Journal of Surgery 1994;18(3):399-405. 85. Brazier J. The short form 36 (SF-36) health survey and its use in pharmacoeconomic evaluation. Pharmacoeconomics 1995;7(5):403-415. 8.6. E O R T C Scoring Manual 87. Ware JE. Jr with Snow K K , Kosinski M , Gandek B. SF-36 Health Survey: Manual & Interpretation Guide. Medical outcomes trust. 1993 88. Juniper EF, Guyatt G H , W i l l i a n A , Griffith L E . Determining the minimal important change in a disease-specific quality of life questionnaire. J C l i n Epidemiol 1994;47:81-87. 89. Jaeschke R, Singer J, Guyatt G . Measurement of health status; ascertaining the minimal clinically important difference. Controlled Clinical Trials 1989;l-:407415. 90. Feinstein A R . Clinimetrics. New Haven, C T : Yale University Press, 1987:141166. 91. Gordon H. Guyatt, Roman Jaeschke, David H. Feeny, and Donald L. Patrick. Measurements in clinical trials: choosing the right approach. In: Quality of Life and Pharmacoeconomics in Clinical Trials. Edited by Spilker. Lippincott-Raven Publishers, Philadelphia. 1996;41-48. 92. Guyatt G H , Berman L B , Townsend M , Pugsley SO, Chambers L W . A measure of quality of life for clinical trials in chronic lung disease. Thorax 1987;42:773778.  130  APPENDIX II DEVELOPMENT OF A QUALITY OF LIFE QUESTIONNAIRE FOR PATIENTS WITH ESOPHAGEAL CANCER ITEM GENERATION PHASE - Health Professionals Input  Date:  Name: Profession:  l=thoracic surgeon 2=radiation oncologist 3=medical oncologist 4=gastroenterologist  /  /  5=psychologist 6=nursing staff 7=dietician  Please indicate all the ways in which you feel the lives of patients with esophageal cancer would be affected in the following domains: i) physical functioning/symptoms:  ii) unpleasant emotions:  iii) self-care functions and activities of daily living:  iv) social/leisure activities:  v) other/comments:  133  APPENDIX III DEVELOPMENT OF A QUALITY OF LIFE QUESTIONNAIRE FOR PATIENTS WITH ESOPHAGEAL CANCER ITEM GENERATION PHASE - Family Input Name: Relationship to Patient: Patient Category:  Physician: 1 = spouse; 2 = sibling 3 = parent 4 = other If other, specify:  Date:  /  /  Main Caregiver:  l=yes; 2=no  1= newly diagnosed (within one month) and prior to treatment 2 = treated with chemoradiotherapy alone, one month post-treatment 3 = treated with chemoradiotherapy alone, three months post-treatment 4 = pre-operative adjunctive therapy and resection, one month post-op 5 = pre-operative adjunctive therapy and resection, three months post-op 6 = resection only, one month post-op 7 = resection only, three months post-op  1. Please tell me about all the ways in which you feel the patient's life has been affected by his/her condition - cancer of the esophagus, (give time to respond spontaneously without prompting)  2. (Once the family member has identified all the ways in which they think the patient's life has been affected, ask them to add any additional items they can think of which affect the patient's life in the following domains) i) physical functioning/symptoms:  ii) unpleasant emotions:  iii) self-care functions and activities of daily living  iv) social/leisure activities 134  APPENDIX IV DEVELOPMENT OF A QUALITY OF LIFE QUESTIONNAIRE FOR PATIENTS WITH ESOPHAGEAL CANCER ITEM GENERATION PHASE - Patient Input Physician:  Patient Name: Patient Category:  1 2 3 4 5 6 7  = = = = = = =  Date:  /  /  recently diagnosed (within one month) and prior to treatment treated with chemotherapy alone, one month post-treatment treated with chemotherapy alone, three months post-treatment pre-operative adjunctive therapy and resection, one month post-op pre-operative adjunctive therapy and resection, three months post-op resection only, one month post-op resection only, three months post-op  1. Please tell me about all the ways in which you feel your life has been affected by your condition - cancer of the esophagus, (give the patient time to respond spontaneously without prompting)  2. (Once the patient has identified all the ways in which they think their lives have been affected, go following domains and ask them to add any additional items they can think of which affect their lives) i) physical functioning/symptoms:  ii) unpleasant emotions:  iii) self-care functions and activities of daily living  iv) social/leisure activities 135  patgen.frm  APPENDIX V Page: 1 of 2  ESOPHAGEAL CANCER QUALITY OF LIFE QUESTIONNAIRE VALIDATION DEMOGRAPHICS FORM  Patient Study ID: Patient Initials:  Physician:  Interviewer Initials:  Interview Date: ( D D / M M / Y Y ) Interview Location:  /  /_  (1 = Vancouver 2=Calgary 3 = V i c t o r i a 4 = London 5 = Other) (1 = clinic/hospital 2=home 3=other)  Date of Birth: ( D D / M M / Y Y )  /  Sex:  (1 =male 2 = female)  Patient Group:  1= pre-treatment 2 = chemoradiotherapy alone 3=pre-operative adjuvant therapy & surgical resection 4 = surgical resection alone  /.  Surgical Approach: (for patient groups 3 or 4) l=non-thoracotomy ( T H E ) 2 = L thoracotomy 3 = laparotomy + R thoracotomy 4 = L thoraco-abdominal incision  Date of Diagnosis:  5=three hole technique [Rt thorac, L t neck incision (cervical)] 6 = other / /  Treatment Start Date: ( D D / M M / Y Y )  /  /  Treatment E n d Date: ( D D / M M / Y Y )  /  /  Type of Cancer:  _  1= adenocarcinoma esophagus (Barrett's) 2 = adenocarcinoma primary gastric 3 = squamous cell carcinoma  Stage of Disease:  (Clinical)  T  N  M  (Pathological)  T  N  M  136  Page: 2 of 2 Investigations:  CXR  (attach copies of reports)  C T Chest  C T Abdomen  L i v e r Function Tests  Other  Pathology Report Attached:  Yes  No  1. D o you live more than one hour from treatment centre:  Yes  No  Yes  No  If yes, how long:  City of Residence:  2. A r e you employed outside the home: If yes, what is your occupation: If no, why not?  1 = retired 2=unemployed 3=homemaker  4=related to esophageal cancer 5 = other medical  3. What medication are you currently taking?  4. H o w would you classify your condition due to your esophageal cancer or its treatment, in the past two weeks: ( l = m i l d 2=moderate 3 = severe) 5. Do you have any other medical problems or conditions which affect the quality of your life? For example, cause symptoms, limit or make activies difficult.  Yes  No  If yes, list conditions:  137  APPENDIX VI ESOPHAGEAL CANCER QUALITY OF LIFE ITEM REDUCTION PHASE INTERVIEWER ADMINISTERED QUESTIONNAIRE  Patient Name: Interview: ( D D / M M / Y Y )  Study ID: /  /  This questionnaire is designed to find out how your esophageal cancer related symptoms affect your life. Please take a few moments to think about all the ways in which your symptoms have affected you during the past month. For example: - physical symptoms which you think are related to your condition - things you can't do because o f your esophageal cancer and things that you would like to be able to do - difficult or unpleasant feelings or emotions caused by your condition - work, home responsibilities, self-care functions (eating, bathing, dressing) social or recreational activities (the things you like to do in your spare time) which are limited or not possible because o f your esophageal cancer. Now, please take your time and tell me about the ways in which you feel the problems due to your symptoms have affected your life during the past month. I w i l l be making notes of the things you mention. Remember that there are no right or wrong answers to any of the questions I w i l l ask you, but rather, it is how you feel your life has been affected by your condition. Everything you say will be confidential and w i l l not affect your medical care in any way.  TO INTERVIEWER: PATIENTS WILL BEGIN TO LIST ASPECTS OF THEIR LIVES, WHICH IN THEIR OPINION, ARE AFFECTED BY ESOPHAGEAL CANCER. CIRCLE THE "S" CORRESPONDING TO ANY ITEM MENTIONED WHICH IS ON THE GENERAL RESPONSE SHEET, INDICATING THAT THESE ITEMS WERE IDENTIFIED SPONTANEOUSLY. IF THE ITEM IS NOT LISTED ON THE RESPONSE SHEET, WRITE IT AT THE BOTTOM OF THE LIST  138  WHEN THE PATIENTS APPEAR FOLLOWING QUESTIONS.  TO HAVE FINISHED,  RECORD ITEMS FROM THE FOLLOWING RESPONSE SHEET BY CIRCLING THE "S".  CONTINUE  QUESTIONS  ON  THE  WITH THE  GENERAL  1.  Have you experienced any other symptoms or problems which you think are associated with your esophageal cancer, during the past month?  2.  Axe there any other unpleasant feelings or emotions which you have experienced in the past month as a result of your condition?  3.  A r e there any other physical activities which you have not been able to do, or in which you have been limited as a result of your condition, during the past month? (i.e. day to day responsibilities, self-care functions - eating, bathing, dressing)  4.  A r e there any other leisure time activities which are limited or not possible because of your esophageal cancer?  5.  A r e there any other ways in which your condition has affected your work or occupation in the past month?  6.  Are there any problems with the treatment or management o f your esophageal cancer that have bothered you in the past month?  7.  A r e there any other ways you can think of in which your condition has affected your life in the past month?  BEFORE PROCEEDING TO THE NEXT SECTION, ENSURE THAT YOU HAVE CIRCLED THE "S" ACROSS FROM EACH ITEM ON THE GENERAL RESPONSE SHEET TO INDICA TE THA T IT WAS A SPONTANEOUS ITEM.  N o w that you have listed a number of ways in which your esophageal cancer has affected your life during the past month, I am going to read a list o f additional items which affect some people like yourself. Some o f the items might not apply to you at all, and others might. Remember that there are no right or wrong answers to any of the questions and all of your answers w i l l be kept confidential. A s I read the list, I w i l l pause briefly after each one and you can tell me i f you experienced it at all in the past month. D o you have any questions before we begin?  139  READ THROUGH THE ITEMS ON THE GENERAL RESPONSE SHEET, PAUSING BRIEFLY AFTER EACH ONE. DO NOT REPEAT ITEMS WHICH THE PATIENT HAS ALREADY VOLUNTEERED SPONTANEOUSLY. THIS TIME, IN GOING THROUGH THE LIST, CIRCLE THE "E" ACROSS FROM EACH ITEM IDENTIFIED TO INDICATE THA T IT WAS AN ELICITED ITEM.  Thank you, now that you have identified all the ways in which you feel the problems with your esophageal cancer have affected your life over the past year, I would like you to tell me how I M P O R T A N T each item is to you in your day to day life, not how severe it is, but how important it is. To do this, I w i l l read each item you have identified and I would like you to indicate the importance o f each item to you. In other words, how much each item "bothers" you or "interferes" with your life. (HAND IMPORTANCE C A R D TO PATIENT) Please tell me the number on this card which matches how important you feel the item is to you. The choices are: 1 = this only bothers me a little 2 = this bothers me somewhat 3 = this bothers me quite a bit 4 = this bothers me a lot 5 = this bothers me very, very much  READ ONLY THE ITEMS WHICH WERE IDENTIFIED, BOTH SPONTANEOUS AND ELICITED, INCLUDING THOSE WHICH WERE IDENTIFIED BUT DID NOT APPEAR ON THE GENERAL RESPONSE SHEET. CIRCLE THE NUMBER RESPONSE SHEETS.  IN THE IMPORTANCE  COLUMN  ON THE  GENERAL  We have reached the end o f the questionnaire. Thank you very much for your time and assistance in answering these questions for us. Patient Comprehension: ( ) completely Patient Compliance: ( ) completely  ( ) somewhat ( ) somewhat  ( ) not at all ( ) not at all  140  APPENDIX  VII  RESPONSE SHEET SECTION 1: SYMPTOMS Item 1 ABDOMINAL CRAMPS 2 ABDOMINAL PAIN 3 DECREASED APPETITE 4 A S P I R A T I O N (inhale fluid into airways) 5 B A C K PAIN 6 BURPING 7 C H A N G E in V O I C E 8 CHEST PAIN 9 CHOKING 10 C O N S T I P A T I O N 11 C O U G H 12 D I A R R H E A 13 D I Z Z I N E S S after eating 14 need to E L E V A T E H E A D OF B E D A T N I G H T 15 F A I N T I N G / S W O O N I N G after eating 16 F E E L I N G F U L L quickly 17 G A S 18 H A R D E N I N G of V E I N S 19 H E A R T B U R N 20 H I C C U P S 21 H O A R S E N E S S 22 I N D I G E S T I O N 23 I N F E C T I O N of surgical site 24 I N T O L E R A N C E to certain F O O D S 25 I N T O L E R A N C E to A L C O H O L 26 I N T O L E R A N C E to S M O K I N G 27 I N T O L E R A N C E to C O F F E E 28 L O O S E S T O O L after eating 29 N A U S E A 30 P A I N in hands and/or feet 31 P E E L I N G on hands and/or feet 32 R E C U R R E N T R E S P I R A T I O R Y I N F E C T I O N 33 R E G U R G I T A T I O N of food stuffs 34 R E G U R G I T A T I O N of liquids 35 increased S E N S I T I V I T Y to C O L D 36 S H O R T N E S S O F B R E A T H 37 S O R E T H R O A T 38 difficulty S P E A K I N G 39 SPITTING up B L O O D 40 need to SPIT up S A L I V A often 41 food S T I C K I N G 42 S T R I C T U R E requiring D I L A T I O N 43 increased S W E A T I N G 44 S W E A T I N G after meals 45 difficulty S W A L L O W I N G liquids  Simultaneous/ Elicited SE S E S E S E S E S E SE S E S E S E S E S E SE S E S E S E S E SE S E S E S E S E S E SE S E SE SE S E S E S E SE S E SE SE S E S E S E SE SE S E S E SE SE S E S E  Rating 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1  2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2  3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3  4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4  5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 141  46 difficulty S W A L L O W I N G solid food 47 painful S W A L L O W I N G 48 S W E L L I N G in throat area 49 food doesn't T A S T E good anymore 50 T I R E D N E S S 51evelopULCER 52 V O M I T I N G 53 W E A K N E S S 54 W E A K N E S S after eating 55 W E I G H T L O S S  S S S S S S S S S S  E E E E E E E E E E  1 1 1 1 1 1 1 1 1 1  2 2 2 2 2 2 2 2 2 2  3 3 3 3 3 3 3 3 3 3  4 4 4 4 4 4 4 4 4 4  5 5 5 5 5 5 5 5 5 5  s s s s  E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E  1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1  2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2  3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3  4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4  5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5  SECTION 2: EMOTIONS 56 A N G E R 57 A N G E R at self 58 A N X I E T Y 59 difficulty C O N C E N T R A T I N G 60 C O N C E R N about not knowing how much time is left 61 D I S A P P O I N T E D 62 D E P R E S S I O N 63 act D I S A G R E E A B L E with family or friends 64 D I S B E L I E F 65 D I S C O U R A G E D with slow recovery 66 D I S S A T I S F I E D with your life 67 E D G Y 68 E M B A R R A S S E D 69 don't E N J O Y E A T I N G anymore 70 unrealistic E X P E C T A T I O N S re treatment 71 very E M O T I O N A L about everything 72 F E A R of body changes 73 F E A R of cancer 74 F E A R of choking 75 F E A R of decreased longevity 76 F E A R of dependence on others 77 F E A R of dying/death 78 F E A R of pain 79 F E A R of recurrence 80 F E A R of unpleasant treatment options 81 F R U S T R A T E D 82 G U I L T Y over having caused the cancer 83 feeling of H E L P L E S S N E S S 84 feeling of H O P E L E S S N E S S 85 I M P A T I E N T 86 I N T O L E R A N T of others 87 I R R I T A B L E 88 sense of I S O L A T I O N 89 L A C K of D E S I R E to do anything 90 L I F E IS M I S E R A B L E 91 L O N E L Y 92 L O S E T E M P E R easily 93 P R E O C C U P I E D with condition  S  s s S S S  s s s S S S  s s s s s s s s s s s s s s s s s s s s s s  1 1 1 1 1 1 1 1 1 1 1 1 1 1  94 S E L F - C O N S C I O U S 95 low S E L F - E S T E E M 96 feel S O R R Y F O R Y O U R S E L F 97 S H O C K 98 T E N S E 99 T H I N G S are getting me down 100 T R A U M A T I Z E D 101 U N H A P P Y 102 U N P L E A S A N T E M O T I O N S related to certain foods or smells 103 U P S E T at having to reduce activity 104 feeling of U S E L E S S N E S S 105 W O R R I E D 106 W O R R I E D about telling family 107 W O R R I E D about the effects of the disease on family or friends 108 W O R R I E D about getting things done  s s s s s s s s S S  s s S S S  E E E E E E E E E E E E E E F.  1 1 1 1 1 1 1 1 1 1 1 1 1 1 1  2 2 2 2 2 2 2 2 2 2 2 2 2 2 2  3 3 3 3 3 3 3 3 3 3 3 3 3 3 3  4 4 4 4 4 4 4 4 4 4 4 4 4 4 4  5 5 5 5 5 5 5 5 5 5 5 5 5 5 5  E E E E E E E E E E E E E E E E E  1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1  2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2  3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3  4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4  5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5  SECTION 3: PHYSICAL FUNCTIONING 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125  BEDRIDDEN difficulty B E N D I N G / L E A N I N G over difficulty D I G E S T I N G F O O D inability to E A T sufficient food decreased E N D U R A N C E lack of E N E R G Y decreased E X E R C I S E T O L E R A N C E difficulty E X E R C I S I N G difficulty G E T T I N G T O S L E E P difficulty G E T T I N G a G O O D N I G H T ' S S L E E P increased symptoms L A Y I N G F L A T on back difficulty M O V I N G around must R A I S E H E A D of B E D at night R E G U R G I T A T I O N when bending/leaning over R E G U R G I T A T I O N upon exertion increased S E N S I T I V I T Y to certain smells/tastes difficulty having S E X  S S S S  s s s s s s s s s s s s s  SECTION 4: ACTIVITIES OF DAILY LIVING Do your symptoms from esophogeal cancer cause difficulty with any o f the following activities: 126 C A R I N G for your family 127 128 129 130 131 132  D R E S S I N G and/or U N D R E S S I N G DRIVING E A T or D R I N K before going to bed GROOMING doing H O U S E W O R K . . L O O K I N G after C H I L D R E N  S E  1 2 3 4 5  S S S S S S  1 1 1 1 1 1  E E E E E E  2 2 2 2 2 2  3 3 3 3 3 3  4 4 4 4 4 4  5 5 5 5 5 5  143  133 134 135 136 137  L O O K I N G after Y O U R S E L F R E L A T I O N S H I P with P A R T N E R SHOPPING W A S H I N G yourself doing Y A R D W O R K  S S S S S  E E E E E  1 1 1 1 1  2 2 2 2 2  3 3 3 3 3  4 4 4 4 4  5 5 5 5 5  S S S S S S S S S S S S S S S S S S  E E E E E E E E E E E E E E E E E E  1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1  2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2  3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3  4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4  5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5  Have you had to/ have you been: 138 A V O I D certain foods and beverages 139 require A S S I S T A N C E with self-care functions 140 have homecare nurse to C H A N G E D R E S S I N G S 141 C H A N G E the N U M B E R of M E A L S / S N A C K S in a day 142 need help looking after D A I L Y C H O R E S , etc 143 E A T M O R E S L O W L Y due to difficulty swallowing 144 E A T M O R E S L O W L Y due to regurgitation 145 unable to E A T O R A L L Y 146 E A T S M A L L E R M E A L S during the day 147 experience increased F I N A N C I A L D I F F I C U L T Y due to disease 148 eat F O O D S U P P L E M E N T S , S P E C I A L D I E T 149 unable to do activities that require L E A N I N G / B E N D I N G 150 unable to do any L I F T I N G 151 take extra time and effort to P R E P A R E M E A L S 152 need more R E S T during day 153 be A B S E N T from work at times 154 work S H O R T E R H O U R S 155 limited in carrying out W O R K P L A C E T A S K S  SECTION 5: LEISURE/SOCIAL ACTIVITIES Does your esophageal cancer cause you difficulty with any of the following leisure or social activities: 156 A E R O B I C S  S E  1 2 3 4 5  157 B I C Y C L I N G 158 C A M P I N G 159 participation in C O M M U N I T Y ACTIVITIES/volunteer groups 160 D A N C I N G 161 D I N I N G O U T at restaurants 162 D I N I N G O U T at other's homes 163 G A R D E N I N G 164 G O L F I N G 165 H I K I N G 166 H O B B I E S (crafts, carpentry, sewing, etc) 167 going to the M O V I E S / T H E A T R E 168 P L A Y I N G a musical instrument 169 R A C K E T S P O R T S 170 R A I S I N G pets 171 participation in S P O R T S I N V O L V I N G R U N N I N G  S E S E S E S E S E S E S E S E S E S E S E S E S E S E S E  1 1 1 1 1 1 1 1 1 1 1 1 1 1 1  2 2 2 2 2 2 2 2 2 2 2 2 2 2 2  3 3 3 3 3 3 3 3 3 3 3 3 3 3 3  4 4 4 4 4 4 4 4 4 4 4 4 4 4 4  5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 144  172 173 174 175 176 177  SINGING SWIMMING TALKING V A C A T I O N S , holidays V I S I T I N G friends or relatives W A L K I N G (unable to go for long walks)  S E s E s E s E S E S E  1 1 1 1 1 1  2 2 2 2 2 2  3 3 3 3 3 3  4 4 4 4 4 4  5 5 5 5 5 5  s E  s s s s s s s s  E E E E E E E E E E E E  1 1 1 1 1 1 1 1 1 1 1 1 1  2 2 2 2 2 2 2 2 2 2 2 2 2  3 3 3 3 3 3 3 3 3 3 3 3 3  4 4 4 4 4 4 4 4 4 4 4 4 4  5 5 5 5 5 5 5 5 5 5 5 5 5  s s s s s  E E E E E  1 1 1 1 1  2 2 2 2 2  3 3 3 3 3  4 4 4 4 4  5 5 5 5 5  Do any of the following apply to you: 178 want to be left A L O N E 179 C O N F I N E D to home or immediate surroundings 180 activity limited due to increased DISTANCE from family/ friends  181 182 183 184 185 186 187 188  E N T E R T A I N less often unable to E N T E R T A I N or spend time with family or friends E X E R C I S E less hesitant to go out with F A M I L Y tired of having F O C U S on oneself and condition no F O O D C E N T E R E D socialization doing more I N A C T I V E P A S T I M E S (TV, reading, etc) restricted social/leisure activities due to L A C K of F I N A N C E S  189 NERVOUS in CROWDS, post-op (incisions, exposure to colds, etc.)  190 preference for Q U I E T T I M E 191 unable to participate in activities that require PHYSICAL EXERTION 192 no desire to S O C I A L I Z E 193 need more S U P P O R T / C O N T A C T with others 194 T I R E D after going out 195 don't go VISIT people very often  S S S S  APPENDIX VIII  MOS 2 0 I T E M SHORT-FORM H E A L T H SURVEY (SF-20) PATIENT QUESTIONNAIRE  Patient Name: Interview: (DD/MM/YY) D D  DD  LZD  Study ID:  Thank you for filling out this questionnaire. 1.  Please start right away and fill out as much as you can before your visit.  2.  Keep the questionnaire with you and finish it after your visit, but before you leave the office.  3.  When you are finished, seal the questionnaire in the envelope and leave it with the receptionist on your way out. STATEMENT O F CONFIDENTIALITY  All information that would permit identification of clinicians or their patients will be regarded a strictly confidential, will be used only for the purposes of operating and evaluation of the study, an will not be disclosed or released for any other purposes without prior consent, except as required by law.  DIRECTIONS PLEASE USE A NO.2 PENCIL ONLY. Do not use ink, felt-tip or ballpoint pen. Fill the circle entirely. Erase errors completely. Avoid stray marks or notes. (There's a space at the end for your comments.) EXAMPLE: o  o  right  •  o  o  o  o  o  wrong  146  1.  In general, would you say your health is: O O O O O  2  excellent very good good fair poor  For how long (if at all) has your health limited you in each of the following activities? (mark one circle on each line) Limited for more than 3 months  Limited for 3 months or less  Not limited at all  The kinds or amounts of vigourous activities you can do, like lifting heavy objects, running or participating in strenuous sports  O  O  O  The kinds or amounts of moderate activities you can do, like moving a table, carrying groceries or bowling  O  O  O  4.  Walking uphill or climbing a few flights of stairs . . . O  O  O  5.  Bending, lifting or stooping  O  O  O  6.  Walking One block  O  O  O  7.  Eating, dressing, bathing, or using the toilet  O  O  O  8.  How much bodily pain have you had during the past 4 weeks?  2.  3.  O O O O O O  none very mild mild moderate severe very severe  Does your health keep you from working at a job, doing work around the house or going to school? O O  O  Y E S , for more than 3 months Y E S , for 3 months or less NO  Have you been unable to do certain kinds or amounts of work, housework or schoolwork because of your health? O O  O  Y E S , for more than 3 months Y E S , for 3 months or less NO  147  For each of the following questions, please mark the circle for the one answer that come closest to the way you have been feeling during the past month. . (mark one circle on each line)  11.  12.  13.  14.  15.  16.  All of the time  Most of the time  A good bit of the time  How much of the time during the past month, has your health limited your social activities (like visiting with friends or close relatives? O ..  Some of the time  A little of the time  None of te time  O ....  O  O  O  O  How much of the time, during the past month, have you been a very nervous person?  O ..  O  O  O  O  o  During the past month, how much of the time have you felt calm and peaceful?  O ..  O  o  o  o  o  How much of the time, during the past month, have you felt downhearted and blue?  O ..  O  o  o  o  o  During the past month, how much of the time have you been a happy person?  O ..  O  o  o  o  o  How often, during the past month, have you felt so down in the dumps that nothing could cheer vou up? O ..  O  O  o  o  o  Please mark the circle that best describes whether each of the following statements is tr or false for you. (mark one circle on each line) Definitely True  Mostly True  Not Sure  Mostly False  Definitely False  17.  I am somewhat ill  O  O  O .  O  O  18.  I am as healthy as anybody I know . . . . O  O  O .  O  O  19.  My health is excellent  O  o  o  .  o  o  20.  I have been feeling bad lately  O  O  o  .  o  o  COMMENTS:  THE UNIVERSITY  OF BRITISH  APPENDIX IX  COLUMBIA  Department of Surgery Faculty of Medicine 910 West 10th Avenue, Room 3100 Vancouver, B.C. Canada V5Z 4E3 Telephone: 604-875-4545 Fax: 604-875-4036  VALIDATION OF A QUESTIONNAIRE TO ASSESS THE QUALITY OF LIFE OF PATIENTS WITH CARCINOMA OF THE LOWER ESOPHAGUS AND GASTRIC CARDIA Investigators: Vancouver, B . C . - D r . R.J. Finley, J . Clifton Calgary, Alta - D r . G . Gelfand London, Ont - D r . R. Inculet, D r . Richard Malthaner Winnipeg, M B - D r . Lawrence Tan, D r . Helmut Unruh Y o u have been asked to participate in a study to learn about the ways in which your symptoms due to esophageal cancer and its treatment affect your life. W e previously interviewed patients like yourself, who suffer from this condition, and asked them to tell us about all the ways i n which their day-to-day lives were affected by their esophageal cancer. This information was used to construct the Quality of Life Questionnaire for Esophageal Cancer. The purpose of the current research is to see i f the questionnaire is doing what its suppose to do. It can then be used in future research studies and by general practitioners to help in understanding the degree of problems caused by esophageal cancer, and whether different treatments help different patients with this condition. We will be asking you to complete four questionnaires related to the problems you experience i n your day-to-day life because of your esophageal cancer or its treatment. The questionnaires are: the above mentioned Quality o f Life Questionnaire for Esophageal Cancer; Global Rating Questionnaire; the E O R T C Q L Q - 3 0 , a general cancer quality of life questionnaire; and the M O S 36, the Medical Outcomes Study questionnaire. W e w i l l ask you to complete these questionnaires at five separate visits - at an initial visit; one week after the initial visit but still before any treatment is begun; one month after treatment; three months after treatment; and six months after treatment. Each visit w i l l require approximately 30 minutes of your time and w i l l not pose any health risks or discomfort, nor cause any immediate, direct benefits to you. There will be no physical examination. A l l information communicated during the interviews and received by questionnaire w i l l be confidential. Participation in this project is strictly voluntary and there will be no financial remuneration. Participation i n the project w i l l not affect your medical care i n any way. Y o u are free to withdraw from the study at any time, with no consequences whatever. ....1 149  APPENDIX X Page: 1 of 2  ESOPHAGEAL CANCER QUALITY OF LIFE QUESTIONNAIRE VALIDATION DEMOGRAPHICS FORM  Patient Study I D : Patient Initials:  Physician:  Interviewer Initials:  Interview Date: ( D D / M M / Y Y ) Interview Location:  /  /  (1 = Vancouver 2=Calgary 3 = Victoria 4=London 5=Other) (1 = clinic/hospital 2=home 3=other)  Date o f Birth: ( D D / M M / Y Y )  /  Sex:  (1 = male 2=female)  Patient Group:  1= pre-treatment 2=chemoradiotherapy alone 3= pre-operative adjuvant therapy & surgical resection 4=surgical resection alone  /  Surgical Approach: (for patient groups 3 or 4) l=non-thoracotomy ( T H E ) 2 = L thoracotomy 3=laparotomy + R thoracotomy 4 = L fooraco-abdominal incision  Date o f Diagnosis:  5=three hole technique [Rt thorac, L t neck incision (cervical)] 6= other / /  Treatment Start Date: ( D D / M M / Y Y )  /  /  Treatment E n d Date: ( D D / M M / Y Y )  /  /  Type o f Cancer:  Stage o f Disease:  1 = adenocarcinoma esophagus (Barrett's) 2—adenocarcinoma primary gastric 3=squamous cell carcinoma (Clinical) (Pathological)  T  N  M  T _  N  M  151  Page: 2 of 2 Investigations:  CXR  (attach copies of reports)  C T Chest  C T Abdomen  L i v e r Function Tests  Other  Pathology Report Attached:  Yes  No  1. Do you live more than one hour from treatment centre: If yes, how long: City o f Residence:  Yes  No  2. A r e you employed outside the home: If yes, what is your occupation:  Yes  No  If no, why not?  1 = retired 2=unemployed 3=homemaker  4=related to esophageal cancer 5=other medical  3. What medication are you currently taking?  4. H o w would you classify your condition due to your esophageal cancer or its treatment, in the past two weeks: ( l = m i l d 2=moderate 3=severe) 5. D o you have any other medical problems or conditions which affect the quality of your life? For example, cause symptoms, limit or make activies difficult.  Yes  No  If yes, list conditions:  152  APPENDIX ESOPHAGEAL CANCER QUALITY OF LIFE QUESTIONNAIRE - VALIDATION PHASE  XI  SELF-ADMINISTERED QUESTIONNAIRE  Patient Initials:  Study I D :  Interview Date (dd/mm/yy):  /  /  P r e - O p Visits:  Baseline:  1 Week:  Post-Op Visits:  1 Month:  3 Months:  6 Months:  This questionnaire has been designed to find out how your problems due to your esophageal cancer and its treatment have affected your life in the past two weeks (one week if the second pre-treatment interview). Please think about how difficult it has been for you in the following areas: physical symptoms related to your cancer or its treatment; difficult or unpleasant feelings or emotions associated with your cancer; physical functioning; activities of your day to day life; and leisure or social activities, because of your esophageal cancer. Remember that there are no right or wrong answers to any of the questions on the attached form, but rather, it is how you feel that your life has been affected by your condition in the past two weeks (one week if the second pre-treatment interview). Please complete the EORTC QLQ-30 which is a general cancer quality of life questionnaire - and let me know when you have finished. Now, I would like you to complete the Esophageal Cancer Quality of Life Questionnaire. Please circle the best response to each question and let me know when you have finished. Please complete the Global Rating of Change questionnaire. When you have finished, please continue on to complete the last questionnaire - MOS SF-36, which is a general health questionnaire. We will complete this same process again in one week, one month after you complete your treatment, three months after your treatment and six months after your treatment. Thank you for participating in our study. 153  E S O P H A G E A L QUALITY  OF  LIFE Q U E S T I O N N A I R E QUESTION  1.  C A N C E R -  APPENDIX  VALIDATION  XII  P H A S E  SHEET  H o w m u c h d i s t r e s s o r d i s c o m f o r t h a v e y o u h a d b e c a u s e o f b u r p i n g in t h e p a s t t w o w e e k s , b e c a u s e of y o u r e s o p h a g e a l c a n c e r or its t r e a t m e n t ? 1 a v e r y g r e a t d e a l of d i s t r e s s or d i s c o m f o r t 2 a g r e a t d e a l of d i s t r e s s or d i s c o m f o r t 3 a g o o d deal of d i s t r e s s or d i s c o m f o r t 4 a m o d e r a t e a m o u n t of d i s t r e s s or d i s c o m f o r t 5 s o m e d i s t r e s s or d i s c o m f o r t 6 v e r y little d i s t r e s s or d i s c o m f o r t 7 no d i s t r e s s or d i s c o m f o r t  2.  H o w m u c h distress or d i s c o m f o r t h a v e y o u had b e c a u s e of solid f o o d s t i c k i n g , in t h e p a s t t w o  weeks?  1 a v e r y g r e a t d e a l of d i s t r e s s or d i s c o m f o r t 2 a g r e a t d e a l of d i s t r e s s or d i s c o m f o r t 3 a g o o d deal of d i s t r e s s or d i s c o m f o r t 4 a m o d e r a t e a m o u n t of d i s t r e s s or d i s c o m f o r t 5 s o m e d i s t r e s s or d i s c o m f o r t 6 v e r y little d i s t r e s s or d i s c o m f o r t 7 no d i s t r e s s or d i s c o m f o r t 3.  H o w l i m i t e d h a v e y o u b e e n in t h e p a s t t w o w e e k s b e c a u s e y o u n e e d e d t o a v o i d c e r t a i n f o o d s or b e v e r a g e s b e c a u s e of y o u r c o n d i t i o n ? 1 totally limited 2 extremely  limited  3 very limited 4 moderately 5 some  limited  limitation  6 a little l i m i t a t i o n 7 n o t a t all l i m i t e d 4.  H o w l i m i t e d h a v e y o u b e e n in t h e p a s t t w o w e e k s  visiting friends or relatives?  1 totally limited 2 extremely  limited  3 very limited 4 moderately 5 some  limited  limitation  6 a little l i m i t a t i o n 7 n o t a t all l i m i t e d  154  5.  H o w limited h a v e y o u been in the past t w o w e e k s b e c a u s e y o u w e r e unable to eat sufficient f o o d b e c a u s e of your condition? 1 totally limited 2 extremely limited 3 very limited 4 moderately limited 5 s o m e limitation 6 a little limitation 7 not at all limited  6.  H o w often in the past t w o  w e e k s have y o u felt fearful of c a n c e r recurrence?  1 all of the time 2 m o s t of the time 3 a g o o d bit of the time 4 s o m e of the time 5 a little bit of the time 6 hardly any of the time 7 n o n e of the time 7.  H o w m u c h distress or d i s c o m f o r t have y o u had in the past t w o w e e k s b e c a u s e of regurgitation w h e n bending/leaning over? 1 a very great deal of distress or d i s c o m f o r t 2 a great deal of distress or d i s c o m f o r t 3 a g o o d deal of distress or d i s c o m f o r t 4 a m o d e r a t e a m o u n t of distress or d i s c o m f o r t 5 s o m e distress or d i s c o m f o r t 6 very little distress or d i s c o m f o r t 7 no distress or d i s c o m f o r t  8.  H o w limited h a v e y o u been in the past t w o w e e k s b e c a u s e y o u n e e d e d to elevate the head of your bed at night b e c a u s e of your condition? 1 totally limited 2 extremely limited 3 very limited 4 m o d e r a t e l y limited 5 s o m e limitation 6 a little limitation 7 not at all limited  155  9. H o w limited h a v e y o u been in the last t w o w e e k s b e c a u s e y o u f o u n d it n e c e s s a r y to eat smaller meals during the day b e c a u s e of your condition? 1 totally limited 2 extremely limited 3 very limited 4 moderately limited 5 s o m e limitation 6 a little limitation 7 not at all limited 10. H o w m u c h distress or d i s c o m f o r t have y o u experienced in the past t w o w e e k s b e c a u s e of difficulty exercising? 1 a very great deal of distress or d i s c o m f o r t 2 a great deal of distress or d i s c o m f o r t 3 a g o o d deal of distress or d i s c o m f o r t 4 a m o d e r a t e a m o u n t of distress or d i s c o m f o r t 5 s o m e distress or d i s c o m f o r t 6 very little distress or d i s c o m f o r t 7 no distress or d i s c o m f o r t 1 1 . H o w often in the past t w o w e e k s have y o u felt fear of unpleasant treatment options? 1 aH of the time 2 m o s t of the time 3 a g o o d bit of the time 4 s o m e of the time 5 a little bit of the time 6 hardly any of the time 7 n o n e of the time 1 2. H o w m u c h distress or d i s c o m f o r t have y o u had in the past t w o w e e k s b e c a u s e of feeling full quickly? 1 a very great deal of distress or d i s c o m f o r t 2 a great deal of distress or d i s c o m f o r t 3 a g o o d deal of distress or d i s c o m f o r t 4 a m o d e r a t e a m o u n t of distress or d i s c o m f o r t 5 s o m e distress or d i s c o m f o r t 6 very little distress or d i s c o m f o r t 7 no distress or d i s c o m f o r t  156  1 3 . H o w m u c h distress or d i s c o m f o r t have y o u e x p e r i e n c e d in the past t w o w e e k s b e c a u s e of i n c r e a s e d sensitivity to cold b e c a u s e of your e s o p h a g e a l c a n c e r ? 1 a very great deal of distress or d i s c o m f o r t 2 a great deal of distress or d i s c o m f o r t 3 a g o o d deal of distress or d i s c o m f o r t 4 a m o d e r a t e a m o u n t of distress or d i s c o m f o r t 5 s o m e distress or d i s c o m f o r t 6 very little distress or d i s c o m f o r t 7 no distress or d i s c o m f o r t 14. H o w limited h a v e y o u been in the last t w o w e e k s dining out at restaurants b e c a u s e of your condition? 1 totally limited 2 extremely limited 3 very limited 4 m o d e r a t e l y limited 5 s o m e limitation 6 a little limitation 7 not at all limited 1 5 . H o w limited h a v e y o u been in the past t w o w e e k s b e c a u s e y o u are unable to participate in activities that require physical exertion, b e c a u s e of your e s o p h a g e a l c a n c e r or its treatment? 1 totally limited 2 extremely limited 3 very limited 4 m o d e r a t e l y limited 5 s o m e limitation 6 a little limitation 7 not at all limited  Thank  you  for taking  the time  to complete  our  questionnaire.  157  APPENDIX XII  E O R T C QLQ-C30  W e are interested in some things about you and your health. Please answer all of the questions yourself by circling the number that best applies to you. There are no "right" or " w r o n g " answers. The information that you provide w i l l remain strictly confidential. Please fill in your initials: Y o u r birthdate (Day, Month, Year): Today's date (Day, Month, Year): No  Yes  D o you have any trouble doing strenuous activities, L i k e carrying a heavy shopping bag or a suitcase?  1  2  2.  Do you have any trouble taking a long walk?  1  2  3.  Do you have any trouble taking a short walk outside of the house?  1  2  4.  Do you have to stay in a bed or a chair for most of the day?  1  2  5.  D o you need help with eating, dressing, washing yourself or Using the toilet?  1  2  A r e you limited in any way in doing either your work or doing Household jobs?  1  2  A r e you completely unable to work at a job or to do household jobs?  1  2  1.  6.  7.  D u r i n g the past week:  N o t at All  A Quite Little a Bit  Very Much  8.,  Were you short of breath?  1  2  3  4  9.  Have you had pain?  1  2  3  4  10.  D i d you need to rest?  1  2  3  4  11.  Have you had trouble sleeping?  1  2  3  4  12.  have you felt weak?  1  2  3  4  13.  have you lacked appetite?  1  2  3  4  14.  Have you felt nauseated?  1  2  3  4  15.  Have you vomited?  1  2  3  4  16.  Have you been constipated?  1  2  3  4 158  Please go on to the next page. During the past week:  Not at A Quite Very AH Little a Bit Much  17.  Have you had diarrhea?  2  3  4  18.  Were you tired?  2  3  4  19.  D i d pain interfere with your daily activities?  2  3  4  20.  Have you had difficulty i n concentrating on things, like reading a newspaper or watching television?  2  3  4  21.  D i d you feel tense?  2  3  4  22.  D i d you worry?  2  3  4  23.  D i d you feel irritable?  2  3  4  24.  D i d you feel depressed?  2  3  4  25.  Have you had difficulty remembering things?  2  3  4  26.  Has your physical condition or medical treatment Interfered with your family life?  27.  Has your physical condition or medical treatment Interfered with your social activities?  28.  Has your physical condition or medical treatment caused you financial difficulties?  2  3  4  For the following questions please circle the number between you. 29.  H o w would you rate your overall physical condition during the past week? 1 Very poor  30.  and 7 that best applies to  2 Excellent  H o w would you rate your overall quality of life during the past week? 1 Very poor  2  Copyright 1992 EORTC Study Group on Quality of Life. All rights reserved.  Excellent  159  SF-36 HEALTH S U R V E Y  INSTRUCTIONS: This survey asks for your views about your health. This information will help keep track of how you feel and how well you are able to do your usual activities.  Answer every question by marking the answer as indicated. If you are unsure about how to answer a question, please give the best answer you can.  1.  In general, would you say your health is: (circle one)  2.  Excellent  1  Very good  2  Good  3  Fair  4  Poor  5  Compared to one year ago, how would you rate your health in general now? (circle one) Much better now than one year ago  1  Somewhat better now than one year ago  2  About the same as one year ago  ;  Somewhat worse now than one year ago  '  Much worse now than one year ago  !  1 Copyright 01992 Medical Outcomes Trust. All rights reserved. (SF-36 Standard U.S. Version 1.0)  161  3.  The following items are about activities you might do during a typical day. Does your health now limit you in these activities? If so, how much?  ACTIVITIES a.  Vigorous activities, such as running, lifting heavy objects, participating in strenuous sports  b.  Moderate activities, such as moving a table, pushing a vacuum cleaner, bowling, or playing golf  c.  Lifting or carrying groceries  d.  Climbing severalflightsof stairs  e.  Climbing oneflightof stairs  f.  Bending, kneeling, or stooping  g- Walking more than a mile Walking several blocks i.  Walking one block  j-  Bathing or dressing yourself  4.  Yes, I Limited | A Little  No, Not Limited At All  2  3  1  2  3  1  2  3  1  2  3  1  2  3  1  2  3  1  2  3  Yes, Limited A Lot  I I I I I I I ^ "I ^  j  1 1  2  "1  2  1  *  3  I  3  "  During the past 4 weeks, have you had any of the following problems with your work or other regular daily activities as a result of your physical health? (circle one number on each line) YES a.  Cut down on the amount of time you spent on work or other activities .  b.  Accomplished less than you would like  c.  Were limited in the kind of work or other activities  d.  Had difficulty performing the work or other activities (for example, it took extra effort)  Copyright 01992 Medical Outcomes Trust. Allrightsreserved. (SF-36 Standard U.S. Version 1.0)  NO  T  T  1  2  T  162  5  During the past 4 weeks, have you had any of the following problems with your work or other regular daily " H h H t i U »resultof anv emotional problems (such as feeling depressed or anxtous)?  '  a  R  (circle one number on each line)  a  Cut down the amount of time you spent on work or other activities  b  Accomplished less than you would like  c  Didn't do work or other activities as carefully as usual  6.  YES  NO  1  2  1  2  1  2  During the past 4 weeks, to what extent has your physical health or emotional problems interfered with your normal social activities with family, friends, neighbors, or groups? (circle one)  1  Not at all Slightly  7.  ;  2  Moderately  3  Quite a bit  4  Extremely  5  How much bodily pain have you had during the past 4 weeks? (circle one) None  1  Very mild  *  Mild  •  -  Moderate  '  Severe Very severe  3 Copyright C1992 Medical Outcomes Trust. Allrightsreserved. (SF-36 Standard U.S. Version 1.0)  16"3  During the past 4 weeks, how much did pajn interfere with your normal work (including both work outside the home and housework)?  ( d r c | e  Q n e )  1  Not at all  .2  A little bit  .3  Moderately.  .4  Quite a bit..  .5  Extremely..  These Questions are about how you feel and how things have been with you during the past 4 weeks. J ^ ^ S ^ p l e a s e give one answer that comes closest to the way you have been feel.ng. How much of the time during the past 4 weeks (circle one number on each line) A Little A Good of the None Bitot Some Most All Time of the the Time of the of the of the Time Time Time Time  L  a.  Did you feel full of pep?  T  Have you been a very nervous person? c.  Have you felt so down in the dumps that nothing could cheer you up?  d.  Have you felt calm and peaceful?  e.  Did you have a lot of energy?  f.  Have you felt downhearted and blue?  g.  Did you feel worn out?  h.  Have you been a happy person? Did you feel tired?  Copyright 01992 Medical Outcomes Trust. All rights reserved. (SF-36 Standard U.S. Version 1.0)  164  10.  During the past 4 weeks, how much of the time has your physical health or emotional problems interfered with your social activities (like visiting with friends, relatives, etc.)? (circle one) All of the  time  1  Most of the time  2  Some of the time  3  A little of the time  ,  4  None of the time  11  5  How TRUE or FALSE is each of the following statements for you?  Definitely True  Mostly True  (circle one number on each line) Definitely Mostly Dont False False Know  T  a.  I seem to get sick a little easier than other people  b.  I am as healthy as anybody I know  c.  I expect my health to get worse  2  3  4  5  d.  My health is excellent  T  T  T  ~5  Copyright 01992 Medical Outcomes Trust. Alt rights reserved. (SF-36 Standard U.S. Version 1.0)  165  STUDY ID: VISIT#:  GLOBAL RATING QUESTIONNAIRE - O V E R A L L  1. Overall, has there been any change in your quality o f life because of your cancer or its treatment, since you last saw us about 4 weeks ago? Please indicate i f there has been any change in your quality of life by choosing one of the following options. Has your quality of life been: 1 2 3  worse about the same better  if 1, go to question 3 if 2, go to question 2 if 3, go to question 4  2. W o u l d you say that the change in your quality of life has been large enough to be important to you, or so small that it is not important? 1 2  important not important  3. H o w much worse would you say your quality of life has been since the last time you saw us? Please choose one of the following options: 1 2 3 4 5 6 7  very small change, not important a little worse, but large enough to be important somewhat worse, still a small change but large enough to be important moderately worse, an important change for the worse a good deal worse, an important change for the worse a great deal worse, an important change for the worse a very great deal worse, very important change for the worse  4. H o w much better would you say your quality of life has been since the last time you saw us? Please choose one of the following options: 1 2 3 4 5 6 7  very small change, not important a little better, but large enough to be important somewhat better, still a small change but large enough to be important moderately better, an important change for the better a good deal better, an important change for the better a great deal better, an important change for the better a very great deal better, very important change for the better  166  GLOBAL RATING QUESTIONNAIRE - SYMPTOMS  1. Overall, has there been any change in your symptoms because of your cancer or its treatment, since you last saw us about 4 weeks ago? Please indicate i f there has been any change in your symptoms by choosing one of the following options. Have your symptoms been: 1 2 3  worse about the same better  if 1, go to question 3 if 2, go to question 2 if 3, go to question 4  2. Would you say that the change in your symptoms have been large enough to be important to you, or so small that it is not important? 1 2  important not important  3. H o w much worse would you say your symptoms have been since the last time you saw us? Please choose one of the following options: 1 2 3 4 5 6 7  very small change, not important a little worse, but large enough to be important somewhat worse, still a small change but large enough to be important moderately worse, an important change for the worse a good deal worse, an important change for the worse a great deal worse, an important change for the worse a very great deal worse, very important change for the worse  4. H o w much better would you say your symptoms have been since the last time you saw us? Please choose one of the following options: 1 2 3 4 5 6 7  very small change, not important a little better, but large enough to be important somewhat better, still a small change but large enough to be important moderately better, an important change for the better a good deal better, an important change for the better a great deal better, an important change for the better a very great deal better, very important change for the better  167  GLOBAL RATING QUESTIONNAIRE - EMOTIONS  1. Overall, has there been any change in your emotions because of your cancer or its treatment, since you last saw us about 4 weeks ago? Please indicate if there has been any change in your emotions by choosing one of the following options. Have your emotions been: 1 2 3  worse about the same better  if 1, go to question 3 if 2, go to question 2 if 3, go to question 4  2. Would you say that the change in your emotions has been large enough to be important to you, or so small that it is not important? 1 2  important not important  3. H o w much worse would you say your emotions has been since the last time you saw us? Please choose one of the following options: 1 2 3 4 5 6 7  very small change, not important a little worse, but large enough to be important somewhat worse, still a small change but large enough to be important moderately worse, an important change for the worse a good deal worse, an important change for the worse a great deal worse, an important change for the worse a very great deal worse, very important change for the worse  4. H o w much better would you say your emotions has been since the last time you saw us? Please choose one of the following options: 1 2 3 4 5 6 7  very small change, not important a little better, but large enough to be important somewhat better, still a small change but large enough to be important moderately better, an important change for the better a good deal better, an important change for the better a great deal better, an important change for the better a very great deal better, very important change for the better  168  GLOBAL RATING QUESTIONNAIRE - PHYSICAL FUNCTIONING  1. Overall, has there been any change in your physical functioning because of your cancer or its treatment, since you last saw us about 4 weeks ago? Please indicate if there has been any change in your physical functioning by choosing one of the following options. Has your physical functioning been: 1 2 3  worse about the same better  if 1, go to question 3 if 2, go to question 2 if 3, go to question 4  2. Would you say that the change in your physical functioning has been large enough to be important to you, or so small that it is not important? 1 2  important not important  3. H o w much worse would you say your physical functioning has been since the last time you saw us? Please choose one of the following options: 1 2 3 4 5 6 7  very small change, not important a little worse, but large enough to be important somewhat worse, still a small change but large enough to be important moderately worse, an important change for the worse a good deal worse, an important change for the worse a great deal worse, an important change for the worse a very great deal worse, very important change for the worse  4. H o w much better would you say your physical functioning has been since the last time you saw us? Please choose one of the following options: 1 2 3 4 5 6 7  very small change, not important a little better, but large enough to be important somewhat better, still a small change but large enough to be important moderately better, an important change for the better a good deal better, an important change for the better a great deal better, an important change for the better a very great deal better, very important change for the better  169  G L O B A L RATING QUESTIONNAIRE - DAY TO DAY ACTIVITIES  1. Overall, has there been any change in your activities of daily living because of your cancer or its treatment, since you last saw us about 4 weeks ago? Please indicate if there has been any change in your activities of daily living by choosing one of the following options. Has your daily living been: 1 2 3  worse about the same better  if 1, go to question 3 if 2, go to question 2 if 3, go to question 4  2. Would you say that the change in your activities of daily living has been large enough to be important to you, or so small that it is not important? 1 2  important not important  3. H o w much worse would you say your activities of daily living has been since the last time you saw us? Please choose one of the following options: 1 2 3 4 5 6 7  very small change, not important a little worse, but large enough to be important somewhat worse, still a small change but large enough to be important moderately worse, an important change for the worse a good deal worse, an important change for the worse a great deal worse, an important change for the worse a very great deal worse, very important change for the worse  4. H o w much better would you say your activities of daily living has been since the last time you saw us? Please choose one of the following options: 1 2 3 4 5 6 7  very small change, not important a little better, but large enough to be important somewhat better, still a small change but large enough to be important moderately better, an important change for the better a good deal better, an important change for the better a great deal better, an important change for the better a very great deal better, very important change for the better  170  G L O B A L RATING QUESTIONNAIRE - SOCIAL ACTIVITIES  1. Overall, has there been any change in your social activities because of your cancer or its treatment, since you last saw us about 4 weeks ago? Please indicate if there has been any change in your social activities by choosing one of the following options. Have your social activities been: 1' worse 2 about the same 3 better  if 1, go to question 3 if 2, go to question 2 if 3, go to question 4  2. W o u l d you say that the change in your social activities has been large enough to be important to you, or so small that it is not important? 1 2  important not important  3. H o w much worse would you say your social activities has been since the last time you saw us? Please choose one of the following options: 1 2 3 4 5 6 7  very small change, not important a little worse, but large enough to be important somewhat worse, still a small change but large enough to be important moderately worse, an important change for the worse a good deal worse, an important change for the worse a great deal worse, an important change for the worse a very great deal worse, very important change for the worse  4. H o w much better would you say your social activities has been since the last time you saw us? Please choose one of the following options: 1 2 3 4 5 6 7  very small change, not important a little better, but large enough to be important somewhat better, still a small change but large enough to be important moderately better, an important change for the better a good deal better, an important change for the better a great deal better, an important change for the better a very great deal better, very important change for the better  171  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0091183/manifest

Comment

Related Items