UBC Faculty Research and Publications

Evaluating quality of care for patients with type 2 diabetes using electronic health record information… Pérez-Cuevas, Ricardo; Doubova, Svetlana V; Suarez-Ortega, Magdalena; Law, Michael; Pande, Aakanksha H; Escobedo, Jorge; Espinosa-Larrañaga, Francisco; Ross-Degnan, Dennis; Wagner, Anita K Jun 6, 2012

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

Item Metadata

Download

Media
52383-12911_2011_Article_509.pdf [ 214.71kB ]
Metadata
JSON: 52383-1.0074668.json
JSON-LD: 52383-1.0074668-ld.json
RDF/XML (Pretty): 52383-1.0074668-rdf.xml
RDF/JSON: 52383-1.0074668-rdf.json
Turtle: 52383-1.0074668-turtle.txt
N-Triples: 52383-1.0074668-rdf-ntriples.txt
Original Record: 52383-1.0074668-source.json
Full Text
52383-1.0074668-fulltext.txt
Citation
52383-1.0074668.ris

Full Text

RESEARCH ARTICLE Open AccessEvaluating quality of care for patients with type 2diabetes using electronic health recordinformation in MexicoRicardo Pérez-Cuevas1*, Svetlana V Doubova2, Magdalena Suarez-Ortega2, Michael Law3, Aakanksha H Pande4,Jorge Escobedo5, Francisco Espinosa-Larrañaga6, Dennis Ross-Degnan4 and Anita K Wagner4AbstractBackground: Several low and middle-income countries are implementing electronic health records (EHR). In thenear future, EHRs could become an efficient tool to evaluate healthcare performance if appropriate indicators aredeveloped. The aims of this study are: a) to develop quality of care indicators (QCIs) for type 2 diabetes (T2DM) inthe Mexican Institute of Social Security (IMSS) health system; b) to determine the feasibility of constructing QCIsusing the IMSS EHR data; and c) to evaluate the quality of care (QC) provided to IMSS patients with T2DM.Methods: We used a three-stage mixed methods approach: a) development of QCIs following the RAND-UCLAmethod; b) EHR data extraction and construction of indicators; c) QC evaluation using EHR data from 25,130 T2DMpatients who received care in 2009.Results: We developed 18 QCIs, of which 14 were possible to construct using available EHR data. QCIs comprisedboth process of care and health outcomes. Several flaws in the EHR design and quality of data were identified. Theindicators of process and outcomes of care suggested areas for improvement. For example, only 13.0% of patientswere referred to an ophthalmologist; 3.9% received nutritional counseling; 63.2% of overweight/obese patientswere prescribed metformin, and only 23% had HbA1c <7% (or plasma glucose ≤130 mg/dl).Conclusions: EHR data can be used to evaluate QC. The results identified both strengths and weaknesses in theelectronic information system as well as in the process and outcomes of T2DM care at IMSS. This information canbe used to guide targeted interventions to improve QC.BackgroundThe worldwide prevalence of type 2 diabetes (T2DM) ison the rise, especially in low and middle-income coun-tries (LMIC), reaching up to 14% of the adult populationin Mexico [1], twice the current global prevalence [2].T2DM poses considerable challenges to healthcare sys-tems because it is among the leading causes for ambula-tory care, hospital admissions and death [3]. In Mexico,quality of T2DM care has been suboptimal [4,5], healthoutcomes poor, [6] and costs of T2DM care and its com-plications are on the rise [7].Healthcare for patients with diabetes has both individ-ual and health system goals. At the individual level, thegoal is to achieve the best possible health outcomes. Atthe health system level, the goal is to provide accessible,high quality, effective, and efficient care. Informationabout progression of the disease and processes of carefor patients with T2DM over time is required to assesswhether goals are being accomplished.The electronic health record (EHR) is a tool that canprovide necessary input for dynamic improvement ofsystems for chronic care by a) supporting clinical deci-sions at the point of care; b) providing routine data toevaluate the quality of care (QC); and c) facilitating lon-gitudinal systems research and efficient informationsharing for dynamic care improvement [8,9]. In publiclyfinanced health care systems, EHR may also help im-prove transparency and accountability.LMIC are beginning to use EHR [10]. Argentina, CostaRica and Peru use EHR in circumscribed settings, such* Correspondence: rperez@iadb.org1Division of Social Protection and Health, Inter-American Development Bank,Washignton, USAFull list of author information is available at the end of the article© 2012 Perez-Cuevas et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of theCreative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use,distribution, and reproduction in any medium, provided the original work is properly cited.Pérez-Cuevas et al. BMC Medical Informatics and Decision Making 2012, 12:50http://www.biomedcentral.com/1472-6947/12/50as individual clinics or hospitals, or for research pur-poses [11]. Mexico was the first Latin American countrythat introduced an EHR on a large-scale at the MexicanInstitute of Social Security (IMSS). IMSS is a nation-wide institution providing social security and healthcarebenefits to approximately 47% of the 112 million Mexi-can people. In 2003, IMSS began to introduce its EHRand evidence-based clinical guidelines to improve per-formance of its family medicine clinics [12]. The IMSSEHR became a key component of the institutional infor-mation system, which also includes administrative,member enrollment, pharmacy and other databases.The IMSS EHR comprises information that could beused to improve the quality, efficiency and continuity ofcare; however, it has not been fully used for these pur-poses. Little is known about the quality of the data that areroutinely captured and what relevant quality of care indi-cators (QCIs) can be constructed based on routine data.T2DM is a complex and dynamic problem that requiresa steady flow of information for healthcare providers toguide their clinical and managerial decision-making. Theobjectives of this project were: a) to develop QCIs forT2DM in the IMSS health system; b) to determine thefeasibility of constructing QCIs using IMSS EHR data; andc) to evaluate the QC provided to IMSS patients withT2DM.MethodsDesignWe used data from four large family medicine clinicslocated in Mexico City with a catchment area of 585,536people. Each clinic had between 15 and 30 physicianoffices, a laboratory, and a pharmacy.IMSS’ EHR consists of several linked databases whichinclude information on appointments; medical history;physical examination; clinical encounter notes; aspects ofcare and progress for patients with specific conditions(diabetes, antenatal care, hypertension); social work, occu-pational health, dietary, and preventive care services. Clin-ical encounter notes capture the reason for the visit,symptoms, signs, diagnoses (ICD-10 coded), orders for la-boratory tests, electronic prescriptions, and disability andreferral information. Data fields contain free-text, nu-meric, or pre-coded, menu-driven entries. Electronic pre-scriptions (by menu-driven generic product name) arelinked to the IMSS essential list of medicines and theclinic’s pharmacy inventory. The laboratory database islinked to the menu-driven, coded catalog of IMSS labora-tory exams and contains laboratory test orders. Laboratoryexamination results reside in different, unlinked databases.Stage 1: Development of QC IndicatorsTo design and validate T2DM QCIs, we used the modi-fied version of the RAND/UCLA AppropriatenessMethod, which comprises scientific evidence and expertopinion [13,14].First, we performed a systematic search and review ofthe literature on QCIs for T2DM using Marshall’s pro-posed method [15]. We searched the following databases:Medline, Ovid, Cochrane Library, National GuidelineClearinghouse, CMA Infobase: Clinical Practice Guide-lines, TRIP database, Institute for Clinical System Im-provement, ACP Guideline website, American Academyof Family Physicians, NHS Evidence - National Library ofGuidelines, and IMSS-Clinical Guidelines [16]. The key-words were “type 2 diabetes mellitus”, “quality of careindicators”, “clinical guidelines,” and “family medicine” or“primary care services,” “diagnosis” and “treatment”.The review focused on evidence-based care processesshown to increase the likelihood of achieving the best pos-sible clinical outcomes, following the criteria proposed bySaslow et al. [17]. The review included indicators fromthe Healthcare Effectiveness Data and Information Set(HEDIS) [18], as well as those developed by the RANDHealth Research and Institute for Clinical System Im-provement (ICSI) teams. Indicators for four basic compo-nents of care for T2DM patients were defined a) timelyscreening for complications and comorbidity; b) non-pharmacological treatment; c) pharmacological treatment;and d) health outcomes.Next, we assembled a panel of 8 experts. Panel membershad experience in treating T2DM patients, in conductingclinical and health systems research, and in developingclinical guidelines. Each panelist received informationabout the study objectives, methodology, literature reviewand set of preliminary indicators. Panelists rated indicatorson validity and feasibility according to Shekell’s criteria[19] by assigning a value from 1 to 9 (1= definitely notvalid or not feasible and 9=definitely valid or feasible). Anindicator was valid when it assessed a process of care forwhich there was enough scientific evidence to support therelationship between the specific process and a potentialhealth benefit or the assumption that health care profes-sionals who perform specific care aspects provide betterquality of care than those who do not. An indicator wasfeasible when there was a high probability that therequired information to construct the indicator would beavailable in a typical health record, or, if not, the lack of in-formation would represent poor quality of care. Panelistsused these criteria to rate each proposed indicator. An in-dicator was valid and feasible if its mean score was≥ 7 oneach domain. Panelists participated in two e-mail roundsof ranking and two face-to-face panel meetings.Stage 2: Extraction of routine EHR data to constructthe QCIsTo evaluate whether the indicators defined in Stage 1could be constructed, we extracted data from a cohortPérez-Cuevas et al. BMC Medical Informatics and Decision Making 2012, 12:50 Page 2 of 10http://www.biomedcentral.com/1472-6947/12/50of all continuously enrolled patients in the four studyclinics. All patients with a diagnosis of T2DM accordingto the International Classification of Diseases, Tenth Re-vision (ICD-10) codes E111-E119, E140-E149, and E14Xin one of three EHR diagnosis fields, who were olderthan 19 years of age and visited a family doctor at leastonce during 2009 were eligible.For the cohort of patients with T2DM, we extractedEHR and laboratory data from the participating FamilyMedicine Clinics (FMCs) for calendar year 2009 and cre-ated an analytic dataset with the following variables:a. Variables from the EHR: general patientcharacteristics (age, sex, schooling, marital status,membership (subscriber/dependent and employmentstatus); medical history (age at onset and duration ofT2DM, history of hypertension, other cardiovasculardiseases, or dyslipidemia; chronic complications[peripheral vascular disease, nephropathy,retinopathy, and peripheral neuropathy]; physicalmeasurements (weight, height, and blood pressure);nutritional status (at the beginning and at the end ofthe year) as measured by body mass index (BMI)(underweight [BMI <18.5 kg/m2], normal weight[BMI 18.5 - 24.9 kg/m2], overweight [BMI 25.0 -29.9 kg/m2], obese [BMI ≥30.0 kg/m2]); clinical care(number of visits to the family doctor, referral to anophthalmologist and/or dietitian); prescriptions(insulin and oral hypoglycemic drugs [metformin,glibenclamide, acarbose, rosiglitazone orpioglitazone]; other drugs [acetylsalicylic acid,statins and angiotensin converting enzyme inhibitorsor angiotensin-receptor blockers]); laboratory testsordered.b. Variables from the laboratory database: dates andmeasured values of hemoglobin A1c (HbA1c),fasting blood glucose, and total cholesterol.We queried the relational EHR and laboratory databasetables of each clinic using standard queries in structuredquery language (SQL) to retrieve data from tables and cre-ate a new analytic database. We then assessed the data forcompleteness, consistency and accuracy. We predefinednon-plausible lowest values for the following variables:blood pressure (systolic blood pressure <50 mmHg or>250 mmHg and diastolic blood pressure <40 mm Hg or>200 mmHg [20]); height (<130 cm or >250 cm), weight(<30 kg [21] or >200 kg), HbA1C (<3.0), fasting bloodglucose (<37 mg/dl); and total cholesterol (<100 mg/dl).All non-plausible values were excluded from the analysis.We used the SAS statistical package (V9.2) to constructcomplex variables from the extracted data. We validatedthe extracted data by comparing a sample of 80 records inthe EHR system with our final dataset. The comparisonswere made reviewing data from individual EHR with theextracted data.Stage 3: Evaluation of T2DM quality of careWe described QC indicators by family medicine clinic.We also characterized clinics by available health personnel(number of medical doctors, nutritionists, social workers,and nurses) and number of examination rooms.The IMSS’ National Commissions of Research and Ethicsreviewed and approved the study protocol.ResultsStage 1: Development of QC IndicatorsTable 1 shows the results of the indicator evaluationprocess. Scores ranged from 5.3 to 8.7. Based on the pre-defined threshold of 7, the panel discarded 5 and accepted18 indicators; some indicators scored high in validity, butwere not feasible and were then modified. For example,HbA1c test results were not available in all clinics. There-fore, the outcome indicator was modified from “T2DMpatients who had HbA1c <7%” to: “T2DM patients whohad HbA1c <7%, or fasting blood glucose ≤130 mg/dlin the last three measurements”. Similarly, the LDL-cholesterol indicator was replaced by an indicator basedon total cholesterol, while the indicator of screening formicroalbuminuria was discarded.Stage 2: Extraction of routine EHR data to construct QCIsThe final list (Table 1) shows that 14 indicators wereprogrammable with data from the EHR and laboratorydatabases. We decided that four indicators could not bereliably constructed because information was only avail-able in free-text fields with wide variability in entries.There was a high percentage of missing data for employ-ment, schooling, marital status, and duration of diabetes.Also, weight, height and blood pressure had between 5%and 24% non-plausible values.Stage 3: Evaluation of T2DM quality of careTable 2 presents the main characteristics of the four par-ticipating FMCs. There were 204 family doctors workingin the clinics; 67 registered nurses, 86 auxiliary nurses, 6dietitians, and 36 social workers. Three clinics had a sep-arate department to conduct medical education activities.In 2009, the clinics covered 585,536 members and depen-dents, with 437,417 >20 years old; approximately 174,266patients visited the clinics at least once during 2009, ofwhom 25,130 (14% of patients) had a diagnosis of T2DM.Table 3 presents the characteristics of the T2DMpatients. Most (59%) were women. The average age was62.3 years. Nearly two in five patients had low education(39% illiterate or primary school education only); almosthalf were married or lived with their partner. Nearly 30%were housewives; 17% were employed and 10% werePérez-Cuevas et al. BMC Medical Informatics and Decision Making 2012, 12:50 Page 3 of 10http://www.biomedcentral.com/1472-6947/12/50Table 1 Results from evaluation of candidate indicatorsIndicator Validity Feasibility Result of thevalidation processProgrammable inthe EHRI. Process indicators AverageA. Timely detection of T2D complications and comorbidity in the last year1. At least one measurement of HbA1c 8.7 7.8 Accepted Programmable2. Comprehensive foot evaluation 8.7 8.2 Accepted Programmable3. Referral to the ophthalmologist 8.3 7.7 Accepted Programmable4. Screening for microalbuminuria through the ratio albumin/creatinine 8.5 5.7 Discarded5. Measurement of creatinine and rate of glomerular filtration 8.7 5.3 Discarded6. Screening for dyslipidemia by measuring total cholesterol in patientswithout previous diagnosis of dyslipidemia8.5 7.8 Accepted ProgrammableB. Non pharmacological treatment in the last year7. Smoking cessation counseling for current smokers 8.2 8 Accepted Not Programmable8. Advise to moderate alcohol consumption 6.0 5.8 Discarded9. Nutritional counseling provided by the nutrition service 7.8 7.8 Accepted Programmable10. Advise to practice aerobic physical exercise of moderate intensity,at least 150 minutes per week, unless contraindicated8.2 8.0 Accepted Not Programmable11. Registration of adherence to dietary recommendations 8.5 7.0 Accepted Not Programmable12. Registration of adherence to aerobic physical exercise 8.5 7.0 Accepted Not ProgrammableC. Pharmacological treatment in the last three visits13. Overweight/obese (BMI≥ 25 kg/m2) patients who received metformin,unless contraindicated*8.7 8.5 Accepted Programmable14. Patients with HbA1c≥ 8, or with an average fasting blood glucoseof ≥140 mg/dl in the last 3 months, who had registered: a) recommendationsto modify their diet and physical exercise, b) referral to a social work group;c) modification in their scheme of treatment.8.7 6.3 Discarded15. Patients with HbA1c≥ 8, or with an average fasting blood glucose of≥140 mg/dl in the last 3 months, who had registered adherence tothe pharmacological treatment.8.5 6.8 Discarded16. Patients with hypertension receiving inhibitors of angiotensin convertingenzyme or angiotensin-receptor blocker, otherwise contraindicated**8.5 8.5 Accepted Programmable17. Patients> 40 years of age with one or more of the following risk factors:smoking, hypertension, dyslipidemia, receiving 75-150 mg/day of acetylsalicylic acid,unless contraindicated***8.5 8.2 Accepted Programmable18. Patient with total cholesterol >200 mg/dl and were prescribed statins,unless contraindicated****8.2 8.3 Accepted ProgrammableII. Health outcomes indicators19. HbA1c <7% or fasting glucose ≤130 mg/dl in the last 3 measurements 8. 7 8.3 Accepted Programmable20. Total cholesterol levels< 200 mg/dl in the last measurement 8. 7 8.2 Accepted Programmable21. Blood pressure <130/80 mmHg in the last 3 measurements 8. 7 8.5 Accepted Programmable22. Overweight/obese (BMI≥ 25 kg/m2) patients who lost ≥5% body weight inthe last year8.0 7. 7 Accepted ProgrammableComposed indicator of health outcomes23. Patients with HbA1c <7%, or fasting glucose ≤130 mg/dl, total cholesterollevels< 200 mg/dl and blood pressure <130/80 mmHg in the last 3 measurements8.5 7.5 Accepted Programmable*Contraindications to metformin: a) Renal failure b) respiratory or advanced liver failure c) congestive heart failure, coronary artery disease or advancedatherosclerosis; d) pregnancy; e) intolerance to metformin.**Contraindications to inhibitors of angiotensin converting enzyme: intolerance and/or prior treatment failure.***Contraindications to acetylsalicylic acid in doses of 75-150 mg/day: history of hypersensitivity to aspirin, peptic ulcer disease, and hemophilia.****Contraindications to statins: hypersensitivity to any component of the drug, active liver disease or unexplained persistent elevations of serum transaminases,pregnancy and lactation.Pérez-Cuevas et al. BMC Medical Informatics and Decision Making 2012, 12:50 Page 4 of 10http://www.biomedcentral.com/1472-6947/12/50retired. Most (76%) patients were dependents of insuredmembers.Table 4 lists the clinical conditions and health carecharacteristics of the included patients. More than 16%of those reporting had diabetes for over 15 years; how-ever, disease duration was mostly missing (44%-77%).Hypertension was the most frequent comorbidity (60%).Approximately 31% of patients had a chronic complica-tion, mostly peripheral vascular disease (14%) followedby diabetic nephropathy (11%). Only 15% had normalweight; 36% were overweight and 33% obese.Patients with T2DM were regular users of health careservices; on average, they had 8.8 visits in 2009. Between10% and 20% of patients did not receive any hypoglycemicprescription; most (77%) received 1 or 2 drugs. Metformin(37%) and glibenclamide (46%) were the most frequentlyprescribed medications.Table 5 reports results for the quality of care indicators.There were wide variations among clinics. Two clinics hadno HBA1c measurements, while the other two had limitedavailability. Indicators of timely detection of complicationsvaried widely: between 5%-52% of patients received com-prehensive foot evaluations; 5%-22% were referred to anophthalmologist; and 46%-70% were screened for dyslipi-demia. Only 4% received nutritional counseling. Acrossclinics, metformin was prescribed (63%) to a majority ofoverweight or obese patients; 57% of patients withhypertension were prescribed ACE inhibitors; 43% ofpatients with risk factors for cardiovascular events receivedprophylactic acetylsalicylic acid; and 48% of those withhypercholesterolemia received a statin.Approximately 62% of patients had registered results ofblood glucose and total cholesterol tests during 2009; amongthose without this information, 22% did not have a glucoseand cholesterol test orders and 16% had the laboratory orderregistered, but these patients did not attend to the labora-tory. After comparing the characteristics of patients withand without laboratory tests, we did not find statistically sig-nificant differences regarding sex, age, literacy or occupation,or co-morbidity. Differences were observed in relation toTable 3 Type 2 diabetes patients’ general characteristicsCharacteristics FMC A FMC B FMC C FMC D Totaln = 7184%n=6671%n=7256%n=4019%n=25130%Female sex 59.3 57.8 58.6 58.2 58.5Age, years, mean(standard deviation)62.9(12.9)64.3(12.8)61.8(12.8)58.5(12.4)62.3(12.9)Age groups<30 years 0.6 0.3 0.7 0.9 0.630–39 years 3.4 3.1 4.0 5.2 3.840–49 years 11.2 9.7 12.6 16.8 12.150–59 years 23.8 21.0 24.4 31.2 24.460–69 years 28.3 29.6 29.5 26.0 28.6≥70 years 32.7 36.3 28.8 19.9 30.5SchoolingIlliterate 16.1 7.5 14.9 9.5 12.4Primary school 24.3 17.6 37.7 26.9 26.8Secondary school 11.9 11.5 14.3 10.6 12.3High school 14.8 17.5 10.4 5.3 12.7University degree 8.0 12.4 4.3 1.3 7.0Missing data 0.9 33.5 18.3 46.4 28.7Marital statusMarried or partnership 48.3 39.8 54.5 38.0 46.2Single or divorced 11.9 12.6 7.9 4.5 9.7Widow 14.2 12.8 2.9 7.1 13.0Missing data 25.6 34.7 22.3 50.4 31.0Employment statusHousewife 32.7 26.1 35.3 17.7 29.3Employed 19.3 16.3 19.6 12.6 17.5Unemployed 0.1 0.2 0.1 0.4 0.2Retired 11.9 12.0 11.4 3.1 10.4Missing data 36.0 45.4 33.6 66.2 42.6Insurance statusSubscriber 24.0 25.9 20.6 22.4 23.3Dependent 76.0 74.1 79.4 77.6 76.7Source: Family medicine clinics 2009 electronic health record information.Table 2 Population and clinic characteristicsPopulation affiliated withthe family medicine clinicFMC A FMC B FMC C FMC D Totaln n n n nTotal number of members 123,276 149,396 196,513 116,351 585,536Members per family doctor 2,241 2,449 3,388 3,878 2, 828Members≥ 20 years old 95,303 120,579 143,796 77,739 437,417Members≥ 20 years old whoattended the clinic at leastonce in 200945,703 53,370 46,270 28,923 174,266T2DM patients as percentageof members ≥20 years oldwho attended the clinic atleast once in 2009718415.7%667112.5%725615.7%401913.9%2513014.4%Health personnel andexamining roomsFamily doctors 55 61 58 30 204Registered nurses 14 26 14 13 67Ancillary nurses 14 21 35 16 86Dietitian 2 1 1 2 6Social workers 9 11 9 7 36Number of familydoctors’ offices27 29 30 15 101Department ofmedical education1 1 1 0 3Source: Family medicine clinics 2009 electronic health record information.Pérez-Cuevas et al. BMC Medical Informatics and Decision Making 2012, 12:50 Page 5 of 10http://www.biomedcentral.com/1472-6947/12/50the number of medical visits to a family doctor, as a higherproportion of patients without laboratory exams had only 1or 2 visits per year (1.2% overall and 21.5% among patientswithout laboratory results) (Data not included in the table).Among patients with laboratory exams, few hadachieved desired health outcomes: 23% had HbA1c <7%or fasting glucose ≤130 mg/dl in the last 3 measurements;52% had total cholesterol <200 mg/dl in the lastTable 4 Clinical conditions and health care characteristicsMedical history FMC A FMC B FMC C FMC D Totaln = 7184%n=6671%n=7256%n=4019%n=25130%Duration of diabetes<5 years 3.8 2.5 1.6 0.9 2.45–10 years 12.3 6.8 5.3 10.3 8.511–15 years 11.6 7.3 5.2 7.5 7.9>15 years 27.5 14.8 10.6 11.1 16.6Missing data 44.8 68.6 77.3 70.2 64.5Comorbidity and chronic complicationsHypertensive disease 63.3 68.0 59.2 44.5 60.4Other cardio-vascular disease 11.2 13.2 6.1) 4.2 9.2Hyperlipidemia 49.2 44.3 40.2 26.1 41.6Diabetic chronic complications 29.9 29.5 38.1 27.0 31.7Type of chronic complicationPeripheral vascular disease 8.9 6.8 26.8 14.9 14.5Diabetic nephropathy 14.5 10.4 8.1 8.2 10.6Diabetic retinopathy 7.0 12.1 6.8 4.4 7.9Peripheral neuropathy 5.6 6.0 3.6 4.3 4.9Nutritional statusNutritional status at the end of the yearUnder weight (<18.5 kg/m2) 0.3 0.5 0.2 0.3 0.3Normal weight (BMI18.5- 24.9 kg/m2) 16.4 17.9 14.2 13.5 15.7Overweight (BMI 25.0 - 29.9 kg/m2) 35.8 35.4 36.3 36.2 35.9Obesity (BMI ≥30.0 kg/m2) 33.9 31.7 35.3 34.2 33.8Missing data 13.6 14.5 14.0 15.8 14.3Health care characteristicsNumber of visits of the patient to the family doctor,mean (standard deviation)8.5 (4.4) 9.6 (5.0) 8.7 (4.5) 7.9 (4.1) 8.8 (4.6)Hypoglycemic prescriptionsNumber of hypoglycemic prescriptionsNone 20.5 18.8 14.1 9.9 16.51 37.1 40.9 32.3 34.0 36.22 38.1 37.2 43.0 46.9 40.7≥3 4.3 3.1 10.5 9.1 6.5Type of hypoglycemic drugsMetformin 57.2 62.3 65.3 67.1 37.5Glibenclamide 51.8 46.9 58.7 61.7 45.9Acarbose 3.3 3.2 12.5 9.7 7.0Thiazolidinedione 0.0 0.0 1.5 0.0 0.4Insulin 14.1 12.4 12.5 17.4 13.7Source: Family medicine clinics 2009 electronic health record information.Pérez-Cuevas et al. BMC Medical Informatics and Decision Making 2012, 12:50 Page 6 of 10http://www.biomedcentral.com/1472-6947/12/50measurement; 11% had blood pressure <130/80 mmHg inthe last 3 measurements; 13% of overweight or obesepatients had lost ≥5% body weight in the last year; andonly 1% of patients reached the combined therapy goalsfor blood glucose, cholesterol and blood pressure control.DiscussionOur results suggest that it is feasible to a) develop T2DMQCIs applicable in the context of a middle-income coun-try; b) measure the QCIs using routinely collected datafrom a widely implemented EHR system; and c) identifyaspects of care for T2DM patients that are in need of im-provement. Our work highlights the potential of routineEHR data to contribute to chronic care quality improve-ment, as well as the need for attention to the quality andcompleteness of electronic data. These findings are im-portant in the global context of rapidly increasing preva-lence of chronic conditions, particularly in low andmiddle-income countries, which require regular monitor-ing of processes and outcomes of care to ensure wise useTable 5 Indicators of quality of careIndicators FMC A FMC B FMC C FMC D TotalI. Process of care n= 7184%n=6671%n=7256%n=4019%n=25130%A. Timely detection of T2D complications and comorbidity in the last yearAt least one measurement of HbA1c 9.0 16.9 Not available Not available 7.1Comprehensive foot evaluation 51.9 28.2 24.6 5.4 30.3Referral to the ophthalmologist 22.2 14.2 7.0 5.4 13.0Screening for dyslipidemia by measuring total cholesterol inpatients without previous diagnosis of dyslipidemia3653 3714 4336 2971 1467465.4 54.7 45.6 70.0 57.8B. Non-pharmacological treatment in the last year n n n n n% % % % %Nutritional counseling provided by the nutrition service 7184 6671 7256 4019 251301.8 5.5 4.9 3.0 3.9C. Pharmacological treatment in the last three visits n n n n n% % % % %Overweight/obese (BMI≥ 25 kg/m2) patients who received metformin,unless contraindicated5066 4437 5216 2840 1755957.2 63.0 66.7 67.6 63.2Patients with hypertension receiving inhibitors of angiotensin convertingenzyme or angiotensin-receptor blocker, unless contraindicated4545 4536 4298 1787 1517246.0 56.8 58.9 66.1 57.4Patients> 40 years of age with one or more of the following risk factors:smoking, hypertension, dyslipidemia, receiving 75-150 mg/day ofacetylsalicylic acid, unless contraindicated5242 5035 4904 2076 1725745.8 32.0 46.2 58.9 43.4Patient with total cholesterol >200 mg/dl and were prescribed statins,unless contraindicated2436 1998 2197 1233 786455.6 43.9 49.1 36.9 47.9II. Health outcomes n n n n n% % % % %HbA1c <7% or fasting glucose ≤130 mg/dl in the last 3 measurements 4644 3560 4563 2816 1558323.0 32.5 19.2 17.9 23.1Total cholesterol levels< 200 mg/dl in the last measurement 5097 4168 4125 3008 1639852.2 52.1 46.7 59.0 52.0Blood pressure <130/80 mmHg in the last 3 measurements 7088 6587 7247 4011 2493312.3 8.5 14.1 5.7 10.8Overweight/obese (BMI≥ 25 kg/m2) patients who lost ≥5% bodyweight in the last year5066 4437 5216 2840 1755913.4 14.7 12.6 12.5 13.3Patients with HbA1c <7%, or fasting glucose ≤130 mg/dl, total cholesterollevels< 200 mg/dl and blood pressure <130/80 mmHg in the last3 measurements4272 3123 3479 2516 133901.8 1.7 1.2 0.6 1.4Source: Family medicine clinics information from 2009 electronic health record. The analysis included 12 months.Pérez-Cuevas et al. BMC Medical Informatics and Decision Making 2012, 12:50 Page 7 of 10http://www.biomedcentral.com/1472-6947/12/50of scarce resources. They also highlight the potential butless than optimal uses of electronic health informationsystems for routine monitoring of care processes and out-comes [22,23].A) QCIs in the context of a middle-income country healthcare systemMost published QCIs have been constructed for healthcaresystems in high-income countries. Using the same QCIsacross countries would facilitate international comparisons[24]. However, QCIs should be adjusted according to na-tional variations in financing, structure, and provision ofhealthcare services, and differences in the characteristicsof populations across countries. We originally intended toadopt the HEDIS indicators [18] used by more than 90%of America's health plans to measure performance of careand services. However, since HbA1c, LDL-cholesterol testsand nephropathy screening are not established standardsat IMSS family medicine clinics, we needed to developcontextually suitable indicators. Our process for develop-ing, measuring, validating, and reporting QC indicatorsmay be informative in other developing countries.B) measuring QCIs using routinely collected data from awidely implemented EHR systemThe EHR did not contain all data elements needed toidentify the denominator population of members withdiabetes or to construct the identified QCIs. Clinical ser-vices at IMSS’ FMCs are almost paperless; all informa-tion is electronically stored. Our results indicate thatthere are shortcomings in the electronic information,possibly related to flaws in its design and the limited useof some EHR fields. While we generated rules and algo-rithms to identify plausible values and exclude recordswith missing values from the analysis, it is crucial forIMSS to improve its electronic record system for qualitymonitoring to become a routine process. This would in-clude maintaining a reliable registry of patients with im-portant chronic illnesses like diabetes.We used a conservative approach to manage missingdata, by not including cases with missing information inthe analysis. This decision was supported by results fromcomparisons of socio-demographic characteristics ofpatients with and without laboratory results, which didnot have statistically significant differences; we also tookinto account that the size of the sample was large, thusreducing the number of cases would not reduce statis-tical power. However, the validity and generalizability ofour results are limited to the extent that there was biasin exclusion of members without visits during the obser-vation year or without data needed to measure QCIs.To progress to a robust eHealth system, it will be neces-sary to analyze in-depth the functionality and quality of theEHR, develop decision-support systems, and strengthenlinkages between the EHR and other electronic applica-tions, such as laboratory and pharmacy databases.C) identifying aspects of diabetes care in need ofimprovementComplex healthcare institutions require ongoing evalu-ation of their performance. In Mexico, evaluating QCbased on EHR data is novel and promising, given thehigh cost, lengthy process, and lack of scale of more typ-ical paper-based evaluations. The lack of complete datanotwithstanding, we identified important aspects of carethat require improvement. IMSS clinics have low ratesof use of diagnostic tests, including HbA1c, microalbu-minuria screening and LDL-cholesterol. IMSS’ currentmodel of care relies heavily on family medicine doctors.As our results show, patients have limited access to die-titians, social workers, nurses and other specialists, hin-dering the integrated multi-specialty care patients withchronic conditions require to ensure the best possibleoutcomes [25,26].Our results are comparable to those from studies withinMexico and elsewhere. In a study conducted in Mexico,66.3% of T2DM patients did not have good metabolic con-trol, although those with access to healthcare had betterhealth outcomes. The European Core Indicators DiabetesProject included 19 countries [27]; its results showed widevariability in QCIs among them: HbA1c testing rangedfrom 51% (Ireland) to 99% (France and the Netherlands);lipid measurements from 45% (Ireland) to 99% (the Neth-erlands); and eye examinations from 12% (Ireland) to 84%(The Netherlands). In India, only 13% of T2DM patientshad at least one HbA1c measurement, 16.2% had an eyeexamination, 3.1% foot examination, 8.3% nutritionalcounseling, 32.1% serum cholesterol estimation and 17.5%were prescribed aspirin; [28]. Rates of HbA1c> 7% variedfrom 32% (Ireland) to 83% (Cyprus); total cholesterol> 5mmol/l ranged from 14% (Ireland) to 68% (Cyprus); bloodpressure> 140/90 mmHg, from 17% (France) to 46%(Sweden). [27] The QUALIDIAB network [29] in Argen-tina, Brazil, Chile, Colombia, Paraguay, and Uruguayreported that 43% of patients had blood glucose level of<7.7 mmol/L and the proportion of patients reachingcombined treatment goals for blood glucose, cholesteroland blood pressure ranged from 1% in Mexico to almost12% in the United States [30]. These variations should giverise to continued assessment and improvement efforts.While this study included information from only fourclinics, we found important differences between them;for example, two clinics did not have HbA1c data, point-ing to heterogeneity in care processes even within thesame health delivery system. Further research studiesshould examine ways to reduce these differences. Al-though the study clinics may not represent the QC thatPérez-Cuevas et al. BMC Medical Informatics and Decision Making 2012, 12:50 Page 8 of 10http://www.biomedcentral.com/1472-6947/12/50IMSS provides, results from desk-based medical chartreviews (with smaller sample sizes) are similar [4,5].Our results will serve as input to subsequent improve-ments of the IMSS EHR. The health system in which theEHR is used has a heavy clinical workload (20–25 patientsper doctor per 6 hours of consultations, or 15–18 minutesper patient). In this context, the EHR can either hinder orfacilitate care. The EHR was originally designed to gatherroutine clinical information to facilitate the daily work offamily doctors and yield epidemiological information. Ourstudy supports the notion that data from the EHR can beused to evaluate quality of care. Further, the EHR couldpotentially serve other purposes, such as a decision sup-port system to guide clinicians through gathering relevantpatient information, including risk factors and health sta-tus, and provide drug dosing guidelines and support tools.However the IMSS EHR does not yet include decisionsupport systems. These should be developed as part of theadvancement of the EHR. The experience of Mexico inthe use of the EHR to evaluate quality of care can be ofvalue for other low and middle-income countries that aimto improve their e-health capabilities.ConclusionsThis study demonstrated that it was feasible to evaluateQC for T2DM patients by using EHR data. EvaluatingQC using EHR information can identify the performanceof individual clinics or individual providers, guide futureinterventions aimed at improving QC and evaluatewhether these interventions achieve their expected aims.A well designed EHR, which should include rigorousdata-entry and data quality monitoring would improve thecapabilities to evaluate and monitor processes and out-comes of care, which in turn would provide useful feed-back to authorities, healthcare personnel and patients. Italso can serve to make comparisons among clinics, healthcare systems or countries.From the clinicians’ perspective, EHR design and useshould improve its functionality for the daily work, forexample as a clinical decision-making tool, and as a reli-able source to provide feedback to the providers.Competing interestsThe authors declare that they have no competing interests.AcknowledgmentsWe thank the medical directors of the participating clinics for theircollaboration in the study. We appreciate the panel of experts whoparticipated in the rating of quality of care indicators.Author details1Division of Social Protection and Health, Inter-American Development Bank,Washignton, USA. 2Epidemiology and Health Services Research Unit CMNSiglo XXI, Mexican Institute of Social Security, México, DF, Mexico. 3Centre forHealth Services and Policy Research, School of Population and Public Health,The University of British Columbia, Vancouver, Canada. 4Department ofPopulation Medicine and WHO Collaborating Center in PharmaceuticalPolicy, Harvard Medical School and Harvard Pilgrim Health Care Institute,Boston, MA, USA. 5Unidad de Investigación en Epidemiología Clínica,Hospital Regional 1 Carlos MacGregor Sánchez Navarro, IMSS, México, DF,México. 6División de Innovación de la Coordinación de Educación en Salud,IMSS, México, DF, Mexico.Authors’ contributionsRPC conceptualized and designed the study, coordinated the fieldwork andwrote the article. SVD conducted the literature review, coordinated thedevelopment, definitions, and programming of the indicators, conducted thestatistical analysis, and interpreted the data and contributed to drafting thearticle. MSO programmed the indicators and collaborated on the analysis;ML participated in assessing data quality and programming the indicatorsand reviewed critically the results and paper. AHP participated in thedevelopment of the study. JE and FEL participated in the design of theindicators and reviewed the paper for significant intellectual content. DR-Dcollaborated on conceptualizing and designing the study, participated in thedevelopment of the study and critically reviewed the paper. AKWcollaborated on conceptualizing and designing the study, participated in thedevelopment of the study, reviewed the results and contributed to draftingthe article. All authors approved the final manuscript.The opinions expressed are those of the authors and do not necessarilyrepresent the views of their institutions.Received: 30 November 2011 Accepted: 6 June 2012Published: 6 June 2012References1. Villalpando S, Shamah-Levy T, Rojas R, Aguilar-Salinas CA: Trends for type 2diabetes and other cardiovascular risk factors in Mexico from1993–2006. Salud Publica Mex 2010, 52:S72–S79. suppl 1.2. Shaw JE, Sicree RA, Zimmet PZ: Global estimates of the prevalence ofdiabetes for 2010 and 2030. Diabetes Res Clin Pract 2010, 87:4–14.3. Rulla JA, Aguilar-Salinas C, Rojas R, Rios-Torres JM, Gómez-Pérez FJ, Olaiz G:Epidemiology of type 2 diabetes in Mexico. Arch Intern Med 2005, 36:188–196.4. Salinas-Martínez AM, Amaya-Alemán MA, Arteaga-García JC, Núñez-RochaGM, Garza-Elizondo ME: Technical efficiency in primary care for patientswith diabetes. Salud Publica Mex 2009, 51:48–58.5. Díaz-Apodaca BA, de Cosío FG, Canela-Soler J, Ruiz-Holguín R, Cerqueira MT:Quality of diabetes care: a cross-sectional study of adults of Hispanicorigin across and along the United States-Mexico border. Rev PanamSalud Publica 2010, 28:207–213.6. Vázquez-Martínez JL, Gómez-Dantés H, Fernández-Cantón S: Diabetesmellitus en población adulta del IMSS. Resultados de la EncuestaNacional de Salud 2000. Rev Med Inst Mex Seguro Soc 2006, 44:13–26.7. Arredondo A, Zuñiga A: Economic Consequences of EpidemiologicalChanges in Diabetes in Middle-Income Countries: The case of Mexico.Diabetes Care 2004, 27:104–109.8. McGlynn EA, Damberg CL, Kerr EA, Brook RH: Health Information SystemsDesign Issues and Analytic Applications. Santa Monica: Rand Health; 1998.9. Car J, Black A, Anandan C, Cresswell K, Pagliari C, McKinstry B, Procter R, MajeedA, Sheikh A: The impact of eHealth on the quality and safety of healthcare. InA Systemic Overview & Synthesis of the Literature Report for the NHS Connecting forHealth Evaluation Programme.; 2008. Available at: http://www.haps.10. Fernández A, Oviedo E: Salud electrónica en América Latina y el Caribe:avances y desafíos. Santiago de Chile: CEPAL; 2010.11. Williams F, Boren SA: The role of the electronic medical record (EMR) incare delivery development in developing countries: a systematic review.Inform Prim Care 2008, 16:139–145.12. Pérez-Cuevas R, Ruiz B, Reyes H, Pedrote B, Massa R, Vargas L, Sánchez L,Estrada C, Michaus F, Castro A, Muñoz O: Implementation and evaluationof the Family Medicine Improvement Process experimental model. InFamily medicine at the dawn of the 21st Century. Edited by García-Peña C,Muñoz O, Durán L. Mexico: IMSS; 2005:55–74.13. Brook RH, Chassin MR, Fink A, Solomon DH, Kosecoff J, Park RE: A methodfor the detailed assessment of the appropriateness of medicaltechnologies. Inn J Technol Assess Health Care 1986, 2:53–63.14. Campbell SM, Hann M, Hacker J, Roland MO: Quality assessment for threecommon conditions in primary care: validity and reliability of reviewcriteria developed by expert panels for angina, asthma and type 2diabetes. Qual Saf Health Care 2002, 11:125–130.Pérez-Cuevas et al. BMC Medical Informatics and Decision Making 2012, 12:50 Page 9 of 10http://www.biomedcentral.com/1472-6947/12/5015. Marshall M, Roland M, Campbell S, Jirk S, Reeves D, Brook R, Shekelle PG:Measuring general practice. A demonstration project to develop and test a setof primary care quality indicators. London: The Nuffield Trust; 2003.16. Instituto Mexicano del Seguro Social. Dirección de Prestaciones Médicas:Guía de Práctica Clínica para el Diagnóstico y Tratamiento de la Diabetesmellitus tipo 2 en el Primer Nivel de Atención. México DF:; 2009.17. Saslow D, Runowicz CD, Solomon D, Moscicki AB, Smith RA, Eyre HJ, CohenC: American Cancer Society Guideline for the Early Detection of CervicalNeoplasia and Cancer. CA Cancer J Clin 2002, 52:342–362.18. National Committee for Quality Assurance: HEDIS 2010 vol. 2 Technicalspecifications for Physician Measurement. Washington DC: NCQA; 2010.19. Shekelle PG, MacLean CH, Morton SC, Wenger NS: Assessing care ofvulnerable elders: methods for developing quality indicators. Ann InternMed 2001, 135:647–652.20. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL, JonesDW, Materson BJ, Oparil S, Wright JT, Roccella EJ: Seventh report of theJoint National Committee on Prevention, Detection, Evaluation, andTreatment of High Blood Pressure. Hypertension 2003, 42:1206–1252.21. McDowell MA, Fryar CD, Ogden CL, Flegal KM: Anthropometric ReferenceData for Children and Adults: United States, 2003–2006. In Nationalhealth statistics reports; no 10. Hyattsville, MD: National Center for HealthStatistics; 2008.22. O’Connor P, Crain AL, Rush W, Sperl-Hillen J, Gutenkauf J, Duncan J: Impactof Electronic Medical Record on Diabetes Quality of Care. Ann Fam Med2005, 3:300–306.23. Fraser HS, Blaya J: Implementing medical information systems indeveloping countries, what works and what doesn't. AMIA Annu SympProc 2010, 2010:232–236.24. Steel N, Melzer D, Shekelle PG, Wenger NS, Forsyth D, McWilliams BC:Developing quality indicators for older adults: transfer from the USA tothe UK is feasible. Qual Saf Health Care 2004, 13:260–264.25. Pérez-Cuevas R, Reyes-Morales H, Doubova SV, Zepeda Arias M, DíazRodríguez G, Peña Valdovinos A, Muñoz Hernández O: Comprehensivediabetic and hypertensive patient care involving nurses working infamily practice. Rev Panam Salud Publica 2009, 26:511–517.26. Ortiz-Domínguez ME, Garrido-Latorre F, Orozco R, Pineda-Pérez D,Rodríguez-Salgado M: Sistema de Protección Social en Salud y calidad dela atención de hipertensión arterial y diabetes mellitus en centros desalud. Salud Publica Mex 2011, 53:S436–S444.27. EuCID: Final Report European Core Indicators in Diabetes Project.; 2008.Available at: http://ec.europa.eu/health/ph_projects/2005/action1/docs/action1_2005_frep_11_en.pf.28. Nagpal J, Bhartia A: Quality of diabetes care in the middle- and high-income group populace: the Delhi Diabetes Community (DEDICOM)survey. Diabetes Care 2006, 29:2341–2348.29. Gagliardino JJ, de la Hera M, Siri F, Grupo de Investigación de la RedQUALIDIAB: Evaluation of the quality of care for diabetic patients in LatinAmerica. Rev Panam Salud Publica 2010, 10:309–317.30. Gakidou E, Mallinger L, Abbott-Klafter J, Guerrero R, Villalpando S, RidauraRL, Aekplakorn W, Naghavi M, Lim S, Lozano R, Murray CJ: Management ofdiabetes and associated cardiovascular risk factors in seven countries: acomparison of data from national health examination surveys. Bull WorldHealth Organ 2011, 89:172–183.doi:10.1186/1472-6947-12-50Cite this article as: Pérez-Cuevas et al.: Evaluating quality of care forpatients with type 2 diabetes using electronic health record informationin Mexico. BMC Medical Informatics and Decision Making 2012 12:50.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitPérez-Cuevas et al. BMC Medical Informatics and Decision Making 2012, 12:50 Page 10 of 10http://www.biomedcentral.com/1472-6947/12/50

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:
https://iiif.library.ubc.ca/presentation/dsp.52383.1-0074668/manifest

Comment

Related Items