USING THE CLALIT HEALTH SERVICES ELECTRONIC MEDICAL RECORD DATABASE FOR SAFETY RESEARCH- A FEASIBILITY CASE STUDY
Author(s)
Feldman B1, Hoshen MB1, Brody J2, Balicer RD3
1Clalit Research Institute, Tel Aviv, Israel, 2Kantar Health, New York, NY, USA, 3Clalit Health Services, Tel Aviv, Israel
OBJECTIVES: Previous published criticisms of safety studies based on secondary databases suggest significant methodological flaws, including substantial missing data, biased diagnoses, outcome under-reporting, and poor transparency in data analytics. Clalit Health Services (CHS) is a payer-provider comprehensive health fund, with 4 million members. The administrative and clinically-oriented CHS database encompasses a decade of integrated inpatient, outpatient, laboratory and pharmacy data. Leveraging published safety studies of the association between the risk of venous thromboembolism (VTE) and arterial thromboembolism (ATE) among Oral Contraceptives (OCs) users of differing progestogens, we tested if the comprehensiveness of the CHS’ EMR database could overcome prior criticisms of secondary database-conducted safety studies. METHODS: Data for all CHS female members aged 15-49 were accessed. A retrospective cohort study design 1.2008-6.2013 identified new OC users in 2008, following-up until study period end. VTE and ATE incidence was determined and Cox-regression models estimated time-to-event across OC groups RESULTS: Among 755,824 women in 2008, 94,792 new OC users were identified using comprehensive pharmacy dispensing data. Monthly demographic data and periodically measured clinical data (like BMI and smoking) are recorded following strict quality measures, resulting in only 9.5% missing smoking status and 2.1% missing BMI values among this sample. Community-based and hospital EMRs are centrally collated. Cardiovascular outcomes were validated via patient file review. As VTE-related hospitalization is independent of gynecologist referral we expect minimal bias on this account. Symptomatic effects requiring intervention are fully reported as cost-related procedures. Out-of-pocket copayments are minimal, not predisposing against referrals. Transparency and quality are assured by detailed epidemiology and biostatistics protocols followed explicitly, statistical programming using cross-checked validated code and peer-reviewed statistical analyses. CONCLUSIONS: As a result of unique historical circumstances and policy decisions, Clalit's comprehensive and integrated data warehouse was able to overcome previously identified flaws in secondary databases for use in safety studies.
Conference/Value in Health Info
2014-05, ISPOR 2014, Palais des Congres de Montreal
Value in Health, Vol. 17, No. 3 (May 2014)
Code
PCV10
Topic
Epidemiology & Public Health
Topic Subcategory
Safety & Pharmacoepidemiology
Disease
Cardiovascular Disorders, Reproductive and Sexual Health