UBC Faculty Research and Publications

Practice patterns of physicians with two year residency versus one year internship training : do both… Schechter, Martin T.; Sheps, Samuel Barry; Grantham, P.; Sizto, Ronnie.; Finlayson, Niall D. C. Sep 30, 1988

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

Item Metadata


52383-Schechter_MT_et_al_Practice_patterns.pdf [ 540.43kB ]
JSON: 52383-1.0048521.json
JSON-LD: 52383-1.0048521-ld.json
RDF/XML (Pretty): 52383-1.0048521-rdf.xml
RDF/JSON: 52383-1.0048521-rdf.json
Turtle: 52383-1.0048521-turtle.txt
N-Triples: 52383-1.0048521-rdf-ntriples.txt
Original Record: 52383-1.0048521-source.json
Full Text

Full Text

PRACTICE PATTERNS OF PHYSICIANS WITH TWO YEAR RESIDENCYVERSUS ONE YEAR INTERNSHIP TRAINING:DO BOTH ROADS LEAD TO ROME?HPRU 88:9Health Policy Research UnitDivision of Health Services Research and DevelopmentOffice of the Coordinator of Health SciencesThe John F. McCreary Health Sciences CentreThe University of British ColumbiaVancouver, B.C.V6T 126M.T. SchechterS.B. ShepsP. GranthamN. FinlaysonR. SiztoSeptember 1988AcknowledgementsThis work was supported by a grant from the College of FamilyPractice of Canada and by a National Health Research Scholar Award to Dr.Schechter from the National Health Research Development Programme of theDepartment of National Health and Welfare of Canada. The authors areindebted to colleagues at the British Columbia Medical Association foraccess and prepara.tion of the necessary data.Page 1INTRODUCTIONIn Canada, the traditional postgraduate training route for generalpractitioners has been the hospital-based junior rotating internship. Until themid 1970's, all provincial licensing authorities had granted unrestrictedlicenses to practice after one year of prelicensure training,The College of Family Physicians of Canada was created by theCanadian Medical Association (CMA) in 1954 and was committed to upgradethe quality of medical care in general practice through education. Initially,this was carried out through continuing medical education of practitionersand later with the institution of the 2 year family practice residency. In 1966,the first family practice residency programmes were established and accreditedby the College of Family Physicians. At present, all 16 Canadian medicalschools offer family practice residency training.Discussion of the relative merits of one versus two years ofprelicensure training date as far back as the early 1930's in North America".In 1969, Millis advocated the integration of internship and residency trainingfor physicians entering into practice in Canada-. The Noakes Committee onPostgraduate Medical Education and Licensure, in 1974, recommended that by1980 "the minimum requirement for independent licensure be 2 years ofresidency training, one year of which should be a rotating type internship,plus a further year which ... should qualify for credit towards fulfillment ofthe requirements for a family practice or specialty certification'<. Althoughthis recommendation was not supported by the Federation of ProvincialMedical Licensing Authorities of Canada (FPMLAC) which had formed thePage 2Noakes Committee, a similar recommendation was put forwardindependently by the Committee on Goals and Priorities of the NationalBoard of Medical Examiners and supported by the Association of CanadianMedical Colleges'[. In 1981, the Royal College of Physicians and Surgeons ofCanada considered this issue and although an ad hoc committee supported a2 year prelicensure requirement, the Position Paper on PrelicensureRequirements recommended that "graduates should be able to complete theprelicensure requirements in one year">.Because of the continuing controversy, the CMA in 1983 created theTask Force on the Provision of Primary Care, chaired by Dean L. Wilson, andsometimes referred to as the Wilson Task Force. Its mandate was to reviewtraining for general/family practice in Canada. In 1985, this grouprecommended a 2 year prelicensure requirement similar to what had beenrequired in Alberta since 1975. Subsequent deliberation by the CoxCommittee supported this and further recommended that the 2 years be spentin a residency-like experience. Finally, in 1987, the Federation of ProvincialMedical Licensing Authorities of Canada (FPMLAC) reversed their earlierposition and recommended a 2 year requirement. As of July 1988, Quebecbecame the first province requiring a university-based 2-year family practiceresidency prior to licensure. Other provinces are expected to follow suit.The controversy surrounding the 2 year prelicensure requirement andthe evolution of its acceptance highlight the need to evaluate theseprogrammes. Moreover, such evaluations must take into account a variety ofperspectives. For example, trainees are concerned whether the lengthening oftheir professional education and the attendant opportunity costs are justified.Page 3Professional bodies and licensing authorities are primarily interested inwhether the additional year significantly improves physician competence andquality of care. From the public's perspective, the issue is whether theadditional year significantly improves patient satisfaction, quality of care, andaccessibility. Finally, governments are oriented toward the manpower andcost implications of the extra training. Some of the variables in the latterequation include the cost of providing all the necessary second year residencypositions, the manpower effects of deferring entry into medical practice byone year, and the effects of the additional training on patterns of practice.From each of these perspectives, the essential question is the relative costversus benefit of the extra year of training.Although each of these questions can be addressed, it is our feeling thatthe most reasonable starting point for any evaluation centres on establishingwhether or not patterns of practice are influenced by type of training. Webelieve this is a critical piece of evidence to establish the benefit side of theequation. Indeed, as Corley has noted, "the validation of an educationprogram lies in the professional practices of its graduates'v.Although a number of articles have explored the broader issuesdiscussed above1,7-14, it is surprising how little comprehensive data areavailable from Canada which explicitly address the question of whether the 2­year family practice residency significantly alters patterns of practice. Wecould only locate three Canadian studies which objectively compared thesetwo types of training15-17. Unfortunately, these suffered from a number ofmethodologic problems including failure to match physicians on importantPage 4confounding variables15,17, small sample size16, and lack of formal statisticalanalysisl".In order to address the central question of practice patterns, weundertook a study designed to answer the following question: "Are theredifferences in patterns of practice between actively practicing physicians whohave been certified after a two year family practice residency versus a matchedgroup without certification who have completed the standard one yearin ternship?"Page 5MATERIALS AND METHODSData SourcesAll data for this study were obtained from the British ColumbiaMedical Association (BCMA) billing files prepared by the Medical ServicesPlan of British Columbia. The initial file used to select the physicians for thisstudy contained data pertaining only to physician characteristics (birth year,medical school and year of graduation, type of training, category of billing,region, and billing status). Type of training is defined as two year trainingwith CCFP certification, one year training (internship) with CCFPcertification, or one year training without CCFP certification. Category ofbilling is defined as solo practice, group practice with individual billingnumber, or group practice with common billing number. Region refers toone of 12 BCMA geographic areas in the province. In addition, the BCMAdefines billing status as 'active' if the physician is billing at least 0.75 of a fulltime equivalent (FTE). The FTE level for a physician is determined by themean billings (in dollars) of all comparable physicians in the same region.This FTE does not include sessional payments.Inclusion CriteriaPhysicians were eligible for selection into the study or control group ifthey were: 1) currently practicing in BC; 2) had either obtained certificationfrom the CCFP after a two year family practice residency (the study group), orhad completed a one year internship without certification (the control group);Page 63) had graduated from a Canadian medical school during 1975-83; and 4) hadmaintained at least 0.75 FTE billings (active status) for the years 1984-87inclusive. The requirement of active status for the latter 4 years was meant toensure that physicians in the study were well established in their practicesand that they spent most of their professional time in primary care activities.We specifically excluded the small number of physicians who completed onlya 1 year internship but subsequently were allowed to sit the CCFP exam­ination.Selection of Study and Control GroupsThe study group consisted of all eligible physicians who were certifiedby the CCFP following a 2 year family practice residency. A total of 347physicians were eligible as controls. To construct the control group, 2physicians were matched to each study group physician on the followingvariables: category of billing (as defined above); region of the province (to thisend, the 12 BCMA regions were collapsed to 3, i.e. Lower Mainland,Vancouver Island, and other); year of graduation (1975-77, 1978-80, or 1981-83);and medical school. When more than 2 physicians were eligible for matchingto a given study group physician, the selection was made randomly. Sex wasnot used as a matching criterion since it was felt that the main differencebetween male and female physicians would be in the amount of time workedand this would be adequately controlled for with the active statusrequirement. It should be noted that the selection of study and controlgroups was undertaken using only the initial data file and thus withoutknowledge of physician practice patterns. Once the study and control groupsPage 7were chosen, their identification numbers were sent to the BeMAwhereupon individual physician's practice variables were abstracted from thebilling file.Study Period and Practice VariablesThe two groups of physicians were compared for each of the fiscal years1984-85, 1985-86, and 1986-87. All available practice variables were compared.It should be noted that most data in the billing file is categorized as 'personal'or 'referred-out'. The former refers to a service performed by the study orcontrol group physician him or herself while the latter refers to any serviceperformed by the first level of consultants arising from direct referrals madeby the study or control group physician. Thus, services (and costs) generatedas a result of subsequent referrals (that is arising from consultants) would notbe included in 'referred-out' calculations. For each study or control groupphysician, we also computed sex-specific age-adjusted costs per patient byapplying the physician's sex and age-specific costs to the age distribution of thepooled patients of all study and control group physicians. In addition, sixspecific practice services which were thought a priori to be influenced by thefamily practice residency program were compared; these were the number ofpatients and visits for counselling, house calls, institutional visits, maternitycare, and non-minor and minor surgical procedures respectively. For each ofthese services it was hypothesized that the study group would have asignificantly higher number of both patients and visits. Because of changes incoding, these special services were only compared for 1986-87 sincecomparable codes could not be reliably defined for the previous 2 fiscal years.Page 8AnalysisStatistical analysis utilized Student's t-test and chi-square test. Giventhe large number of variables examined in this study, we adopted a priori theconvention concerning p-values that those less than 0.001 were to beconsidered significant, those greater than 0.05 were to be considered non­significant, and those intermediate to these values were to be consideredsuggestive and worthy of further study.Page 9RESULTSAccording to the BCMA files, there were 146 family practice graduates in'active status' practice in British Columbia in fiscal 1986-87. Of these, 13 weregraduates of foreign medical schools and 5 graduated outside the period 1975­83. Of the remaining 128 physicians, a total of 65 had maintained 'activestatus' throughout the period 1984-87 and were thus eligible for the studygroup. Accordingly, 130 physicians were chosen for the control group. Theproportions of males in the study and control groups were 75.4% and 83.8%respectively (p=O.22). The median year of graduation for both groups was 1978and the median age of both groups as of 1987 was 35 years with ranges of 30-46for the study group and 28-49 for the control group. The geographicdistribution of physicians in the study was as follows: 13.3 per cent(Vancouver Island); 49.7 per cent (Lower Mainland including GreaterVancouver); and 36.9 per cent (remainder of BC). The distribution ofphysicians as to category of billing was: solo practice (41.5 per cent); grouppractice with individual billing number (17.4 per cent); and group practicewith common billing number (41.0 per cent).Although we examined data for 3 fiscal years, for the sake of brevity wepresent data only for fiscal 1986-87. Table 1 presents comparisons of selectedpractice variables between the study and control groups for that year. As canbe seen, there were no significant differences with regard to a wide array ofpractice variables. For example, the study group and control group treatedapproximately the same number of patients (1888 vs 1842), billed forapproximately the same number of personal services (7265 vs 7173), billed forvirtually the same number of personal services per patient (3.9 vs 3.9),Page 10received essentially the same amount of funds for personal services ($140,192vs $140,100), and received approximately the same number of dollars perpatient for personal services ($77 vs $79). Although we have discussedservices performed 'personally', it can be seen from the table that strikingsimilarities were also present with regard to services 'referred out'.Moreover, when we examined the proportion of each practice referred outduring the year, no difference was detected (51 per cent vs 56 per cent; p =0.37).TABLE 1: COMPARISON OF SELECTED PRACTICE VARIABLES BETWEEN STUDYAND CONTROL GROUP PHYSICIANS FOR FISCAL 1986-87Practice Variable Study Group Control Group p-value(mean± sd) (mean± sd)Total patients treated 1888± 606 1842±616 0.62Total number of services paid* 11938 ± 3324 11878± 3398 0.91Total number of personal services paid 7265± 2253 7173± 2265 0.79Total number of referred-out services paid 4832± 1735 4938 ± 1822 0.69Number of services per patient" 6.6 ± 1.7 7.0 ± 2.8 0.23Number of personal services per patient 3.9 ± 0.9 3.9 ± 1.1 0.88Number of referred-out services per patient 2.7 ± 1.0 3.1±2.5 0.14Total $ paid" 236371 ± 66134 237839 ± 66347 0.88Total $ paid for personal services 140192 ± 41878 140100± 41337 0.98Total $ paid for referred-out services 96118 ± 32247 98370± 34603 0.66Total per patient $* 131±33 137±41 0.21Total per patient $ for personal services 77± 18 79±22 0.44Total per patient $ for referred-out services 103± 18 109± 28 0.08* includes services performed by physician (denoted personal) and by first level consultants towhom patients were referred (denoted referred-out)Page 11Turning our attention to other services, we found that the number oflaboratory services per patient were similar in the study and control groups(2.4 vs 2.7; p = 0.26) and that the mean cost of laboratory services per patientwas also similar ($22.32 vs $25.66; p = 0.11). The number of X-ray services perpatient did not differ between the groups (0.25 vs 0.26; p = 0.45) and the meancost of X-ray services per patient were similar ($7.96 vs $8.27; p=0.57).Table 2 provides a further analysis of practice variables by sex of patient.As seen in the table, the total number of male and female patients treated, thetotal number of services for these patients, the total earnings derived fromthese patients, and the age-adjusted cost per male and female patient werevirtually identical in the study and control groups.TABLE 2: COMPARISON OF PRACTICE VARIABLES FOR MALES AND FEMALEPATIENTS BETWEEN STUDY AND CONTROL GROUP PHYSICIANS (INCLUDINGONLY PERSONAL SERVICES) FOR FISCAL 1986-87Praetiee Variable Study Group Control Group p-value(mean± sd) (mean± sd)Total number of male patients treated 795 ± 333 786 ± 332 0.85Total number of female patients treated 1028 ± 301 986±315 0.37Total number of services for male patients 2514 ± 1174 2567± 1085 0.76Total number of services for female patients 4325± 1241 4099 ± 1328 0.25Total earnings for male patients ($) 50137 ± 22769 52255 ± 21265 0.52Total earnings for female patients ($) 85019± 23256 82038 ± 24957 0.42Age-adjusted cost per male patient ($) 63.32 ± 14.60 67.61 ± 19.10 0.09Age-adjusted cost per female patient ($) 84.20 ± 19.22 85.27 ± 20.87 0.73Page 12With regard to the age- and sex-specific costs per patient for the study andcontrol physicians, Figure 1 presents these graphically for male and femalepatients across 8 age groups.Figure 1. Mean age-specific costs by sex and type of physician160140-fF>~....::: 120<lJ.......Oi0..... 100<lJ0......<Jl0tj~ 80<lJ::E6040<1 1-4 5-14 15-24 25-44 45-64 65-74Age Group75+Legend for Figure 1:xmale patients of study group physiciansmale patients of control group physiciansL\. female patients of study group physicians+ female patients of control group physiciansNo significant differences were detected between the physician groupswith the exception of the category of patients who were female aged 75 ormore for which a difference of intermediate significance was present ($123.80vs $144.08; P =0.022).Page 13Table 3 presents data for the 6 pre-specified services. Although nodifferences were seen for counselling, home visits, institutional visits, andminor and non-minor surgery, we did detect a non-significant difference inthe mean number of female patients receiving maternity care (62 vs 50; p =0.05) and a marginally significant difference in the number of maternityservices billed (341 vs 249; p =0.001).TABLE 3: COMPARISON OF SPECIFIC SERVICES BETWEENSTUDY AND CONTROLGROUP PHYSICIANS (ONLY PERSONAL SERVICES) FOR FISCAL 1986-87Specific Service Study Group Control Group p-value(mcan±sd) (mean± sd)Counselling-number of patients served 125± 103 129±94 0.79Counselling-number of services billed 164± 156 172 ± 142 0.74Home visits-number of patients served 14± 16 18±22 0.17Home visits-number of services billed 31 ±47 40±67 0.33Institutional visits-number of patients served 120±48 114± 55 0.44Institutional visits-number of services billed 542± 294 629 ±421 0.10Maternity care-number of women served 62±32 50±47 0.05Maternity care-number of services billed 341± 186 249± 164 0.001Surgery (non-minorl-number of patients served 48±29 55±36 0.20Surgery (non-minorl-number of services billed 68±43 76±55 0.31Surgery (minor)-number of patients served 92±56 92±82 0.97Surgery (minori-number of services billed 133±95 134± 117 0.98Page 14As noted above, similar analyses were carried out for the two fiscal years1984-85 and 1985-86. The data are presented in Appendices A and B. It is clearthat for the variables studied, no differences between the groups weredetected.Page 15DISCUSSIONSupporters of 2-year residency training can cite the findings of Brennanand Stewart15 who compared trainees from both types of programmes at theUniversity of Western Ontario and found that family practice graduates weremore satisfied with practice, placed greater importance on emotional factorsin illness, conducted more psychotherapy, spent more time with patients, andprovided more non-institutional care. Unfortunately, these groups were notmatched on several important potential confounding variables makinginterpretation of the data difficult. On the other hand, supporters of the 1year internship can cite the results of Curry16 who conducted a similarcomparison of trainees from Dalhousie University and found no differencesin the proportion of medical services billed in each of 15 service classes.While the latter investigation did match the study groups on the basis ofpractice location, age, and gender, it unfortunately suffered from very smallsample sizes (main study groups of 11 and 23 physicians) making acceptanceof its negative results tenuous.The ideal study to compare these educational interventions is arandomized controlled trial with medical graduates randomly assigned toeither of the 2 types of programmes. This would remove the confoundingarising from the self-selection of individuals into one type of program or theother. Given the fact that such a trial is not feasible, we conducted anobservational study comparing graduates of 2-year family practice residencieswith I-year internship trainees. In order to make the comparisons as valid aspossible, we took care to match on what we considered to be critical potentialPage 16confounders including year and school of graduation, category of billing, andregion.Given the various perspectives from which these two types of trainingcould be compared, we felt an assessment of patterns of practice to be the mostreasonable starting point for several reasons. First, data with which to makethese comparisons were readily available. More important, it seems plausiblethat if the two year programme produces measurable improvements inquality of care, these should at least be reflected in practice pattern differences.We therefore undertook a study comparing patterns of practice of trainees ofboth types of programmes. In assessing our results, one cannot help but besurprised at the striking similarity of two groups of physicians with clearlydivergent training experiences. We detected no differences whatsoever withregard to a wide range of practice parameters; indeed, only one measurement,services for maternity care, approached a statistically significant difference.Such an observation is entirely within the realm of chance given the numberof comparisons made.There are several alternative explanations which are worthy ofdiscussion. In any negative study, one must always consider the possibility oftype II error, that is that a true difference was missed due to sampling error.This, however, was unlikely given the size of our study. For example, therewas 80% power to detect a $10.00 (or 15%) reduction in the mean age-adjustedcost per male patient even with a stringent confidence level of alpha=O.OOl.Moreover, one would have to postulate the simultaneous occurrence ofseveral type II errors to explain our data and this is extremely unlikely.Page 17A second possibility is that by matching on year of graduation and typeand location of practice, we artifactually created similar groups with regard topractice patterns. This phenomenon is known as 'over-matching'. Toaddress this, it is important to separate the effects of training on two types ofdecisions. The first set of decisions pertains to the type of practice (ie solo vsgroup) and location (ie urban vs rural) which a physician chooses. Thesecond set relates to clinical decision-making within the context of patientcare such as whether or not a physician chooses to make a particular referralor order a given laboratory test. Since both type of practice and location mayinfluence the available options within clinical decision-making and thusindirectly affect patterns of care, any analysis of these patterns must take thesevariables into account; we chose to do this by matching our groups. It ispossible, however, that the effects of the 2 year residency are mediated solelythrough influencing the first set of decisions regarding practice type andlocation. Indeed, when the 2 year graduates were compared to all 1 yeartrainees rather than just to the matched group, several trends were apparent.First, the residency graduates were younger as expected since the residency is arecent phenomenon. In addition, the residency graduates were somewhatmore likely to be in a group practice and to practice in rural areas but theseeffects were inconsistent across age groups. This suggests that temporalfactors such as recent practice saturation of the urban areas and the recenttrend toward group practices were equally if not more influential than anyintrinsic preference of the physicians themselves. Indeed, despite the lack ofmatching, Brennan and Stewart's prediction that family practice graduateswould more likely choose group practice was not borne out in their data15.Simply put, it would appear that the 1 year graduates having entered practicePage 18earlier, had greater flexibility in their decisions regarding practice type andlocale.At any rate, proponents of the 2 year programme have generally notargued that its main benefits are on extrinsic decisions regarding practice typeand location but rather on clinical decision making and hence on practicepattcrnsU. Our data suggest that all other things being equal in terms of age,practice type and practice location, this argument does not appear to besubstantiated.Several advantages of our study over previous Canadian reports includethe use of a provincial rather than a programme focus thus samplinggraduates of both types of programmes from all across the country rather thanfrom a single institution. Thus, these results may be more generalizable thanstudies derived from a single university. It is noteworthy that we studiedonly 65 out of 146 residency trained physicians currently practicing in BritishColumbia. While at first glance, this might suggest the possibility of selectioneffect, it should be recalled that this was due mostly to our restriction tophysicians defined to be in active practice over the entire study period basedon the prespecified B.C.M.A. criterion. This was an explicit decision becausewe believe it is most appropriate to make these comparisons once full timepractice is established when the data are more likely to reflect longer termpractice patterns rather than in the first one or two years of practice whenearly but transient effects are more likely to be observed.An important caveat with regard to the interpretation of this study isthat the issue of quality of care could not be addressed. Indeed, it is possiblePage 19that there are significant differences in quality of care between these groups interms of appropriateness, patient satisfaction, and physician satisfaction thatwe were unable to measure within the context of this study.How can one explain the striking similarity we observed between thesetwo groups in the face of the differences in their training. The most plausibleexplanation, we would speculate, is that the primary determinants of aphysician's patterns of practice are environmental including patientexpectation and demand, epidemiological parameters, institutionalrequirements, economic factors, medicolegal issues, and the patterns ofpractice of peer practitioners in the same community. If there are differenteffects arising from different training routes, we speculate these are likely tobe transient; any differences are likely to be rapidly overwhelmed byenvironmental factors as the physician is assimilated into the local practicecommunity.Having demonstrated that there are no differences in practice patterns, itis now critical to address the issue of quality of care. However, whether or notthe graduates of the programmes under study differ in terms of the quality ofcare, it is unlikely, based on the present data, that these differences will bereflected in cost savings or in decreased utilization of health care resources.Page 20REFERENCES1. McDougall GM, Lockyear JM, Anholt LM, Cockroft L. Two years ofprelicensure training: Observations from the Alberta experience. CanFam Physician 1983; 29: 1224-35.2. Millis J. The objectives of residency training. Can Med Assoc J 1969; 100:599-602.3. Noakes JA. Report on Postgraduate Medical Education and Licensure.Presented to the Royal College of Physicians and Surgeons of Canada.Federation of Provincial Medical Licensing Authorities, Ottawa 1974.4. Hatcher JD, Spooner J. Report to the Association of Canadian MedicalColleges. Association of Canadian Medical Colleges Forum 1981; 14: 19­20.5. Royal College of Physicians and Surgeons of Canada. Position Paper onPrelicensure Requirements. Ottawa, May 1981.6. Corley JB. In-training residency evaluation. J Fam Pract 1976; 3: 499-504.7. Hines RM, Curry DJ. The consultation process and physiciansatisfaction: Review of referral patterns in 3 urban family practice units.Can Med Assoc J 1978; 118: 1065-73.8. Brock C. Consultation and referral patterns of family physicians. J FamPract 1977;4: 1129-34.9. Biehn J. Family medicine residency: Does it make a difference? CanFam Physician 1983; 29: 1031-34.Page 2110. Cherkin OC, Rosenblatt RA, Hart LG, et al. A comparison of the patientsand practices of recent graduates of family practice and general internalmedicine residency programs. Med Care 1986; 24: 1136-50.11. McCranie EW, Hornsby JL, Calvert [C. Practice and career satisfactionamong residency trained family physicians: A national survey. J FamPract 1982;14: 1107-14.12. Kane RL, Gardner J, Wright DD, et al. Differences in the outcomes ofacute episodes of care provided by various types of family practitioners. JFam Pract 1978; 6: 133-8.13. Hueston WJ. The influence of a family practice residency on the costs ofin-patient diagnostic testing. J Fam Pract 1986; 23: 559-63.14. Allingham JD, Heaton CJ, Wakefield J, Longhurst M. A decline ingeneralism for family medicine? First-practice profile ideals and realities.Can Fam Physician 1988; 34: 289-93.15. Brennan M, Stewart M. Attitudes and patterns of practice: A comparisonof graduates of a residency program in family medicine and controls. JFam Pract 1978; 7: 741-48.16. Curry L. Post-graduate training route and content of subsequent practice.Can Fam Physician 1985; 31: 1417-20.17. Frenette J. What do family practice graduates do? A comparative study.Can Fam Physician 1984; 30: 2281-88.Appendix ATABLE lA- COMPARISON OF SELECTED PRACTICE VARIABLES BETWEEN STUDYAND CONTROL GROUP PHYSICIANS FOR FISCAL 1984-85Practice Variable Study Group Control Group p-value(mean± sd) (mean± sd)Total patients treated 1816 ± 556 1845± 613 0.76Total number of services paid" 10736 ± 3619 11010± 3326 0.60Total number of personal services paid 6403± 1973 6572± 2112 0.59Total number of referred-out services paid 4334± 2011 4438± 1931 0.73Number of services per patient* 6.2 ± 2.2 6.4 ± 2.2 0.61Number of personal services per patient 3.7± 1.1 3.7 ± 1.1 0.81Number of referred-out services per patient 2.5 ± 1.3 2.7 ± 1.5 0.53Total $ paid* 199402 ± 64633 208581 ± 62508 0.34Total $ paid for personal services 121933± 36556 126965 ± 36869 0.37Total $ paid for referred-out services 77468 ± 32692 81616± 35725 0.43Total per patient $* 115±39 121± 42 0.33Total per patient $ for personal services 70±20 72±20 0.39Total per patient $ for referred-out services 94±22 101 ± 28 0.05* includes services performed by physician (denoted personal) and by first level consultants towhom patients were referred (denoted referred-out)TABLE 2A- COMPARISON OF PRACTICE VARIABLES FOR MALES AND FEMALEPATIENTS BETWEEN STUDY AND CONTROL GROUP PHYSICIANS (INCLUDINGONLY PERSONAL SERVICES) FOR FISCAL 1984-85Practice Variable Study Group Control Group p-value(mean ±sd) (mean ± sd)Total number of male patients treated 736± 289 762 ± 316 0.58Total number of female patients treated 961 ± 279 955 ±306 0.89Total number of services for male patients 2160± 814 2330± 934 0.21Total number of services for female patients 3760± 1306 3713 ± 1280 0.81Total earnings for male patients ($) 41104± 15667 45177 ± 16898 0.11Total earnings for female patients ($) 71507± 24097 71377± 22826 0.97Age-adjusted cost per male patient ($) 60.05 ± 15.80 64.73 ± 17.55 0.08Age-adjusted cost per female patient ($) 76.02± 20.94 78.17 ± 21.70 0.51Appendix BTABLE 1B- COMPARISON OF SELECTED PRACTICE VARIABLES BETWEEN STUDYAND CONTROL GROUP PHYSICIANS FOR FISCAL 1985-86Practice Variable Study Group Control Group p-value(mean ± sd) (mean± sd)Total patients treated 1882± 603 1862±601 0.83Total number of services paid* 11334± 3211 11314± 3145 0.97Total number of personal services paid 7100± 2129 7043 ± 2276 0.88Total number of referred-out services paid 4234± 1570 4271 ± 1864 0.90Number of services per patient" 6.3 ± 1.7 6.5± 2.1 0.57Number of personal services per patient 3.9± 1.0 3.9 ± 1.0 0.95Number of referred-out services per patient 2.4 ± 1.0 2.6 ± 1.4 0.36Total $ paid* 218028 ± 60111 221942 ± 59720 0.67Total $ paid for personal services 135082± 38876 135547± 38487 0.94Total $ paid for referred-out services 82946± 29188 86395± 33963 0.47Total per patient $* 122±33 127± 40 0.33Total per patient $ for personal services 75± 18 76±20 0.57Total per patient $ for referred-out services 98 ± 18 105± 26 0.03* includes services performed by physician (denoted personal) and by first level consultants towhom patients were referred (denoted referred-out)TABLE 2B- COMPARISON OF PRACTICE VARIABLES FOR MALES AND FEMALEPATIENTS BETWEEN STUDY AND CONTROL GROUP PHYSICIANS (INCLUDINGONLY PERSONAL SERVICES) FOR FISCAL 1985-86Practice Variable Study Group Control Group p-value(mean ±sd) (mean± sd)Total number of male patients treated 774 ± 320 778 ±316 0.93Total number of female patients treated 1008± 303 976 ±307 0.50Total number of services for male patients 2430± 1094 2539± 1073 0.51Total number of services for female patients 4236± 1260 4013 ± 1340 0.27Total earnings for male patients ($) 46069 ± 20085 49194± 19631 0.30Total earnings for female patients ($) 80521 ± 23277 76757± 22793 0.28Age-adjusted cost per male patient ($) 63.90 ± 15.97 68.21 ± 19.64 0.14Age-adjusted cost per female patient ($) 82.09 ± 20.04 81.97 ± 20.74 0.97


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items