EuroREACH Case Study on Diabetes Care
- Cross-country assessment of performance requires substantial effort to prepare information and the use data.
- Lack of comparable data must be addressed by developing a common protocol to be used by all collaborators.
- Data pre-processing must be carefully conducted to guarantee technical quality.
- Regulations on privacy and confidentiality must be considered in cross-country comparisons of performance.
Numerous policy papers discuss how to compare quality or efficacy by using an "ideal" dataset. However, such ideal data is neither collected nor available to the degree of detail necessary for evaluating performance of health systems or health service providers. Read more
Comprehensive performance assessment of health systems in management of chronic diseases calls for nationally representative data on incidence and prevalence over time and utilization of preventive and curative health services at person level in order to link the health outcomes to treatment provided and the resources used. The most obvious advantages of administrative health data are that it covers representative populations and long time-frame and is already collected.
Another major strength of using linkable administrative databases over other methods of data collection is that those provide data for long-term follow-up of patients, which is essential for drawing solid conclusions on distant outcomes of any interventions, whether preventive or curative.
The EuroREACH project set up a study to maximize information from the data already collected and available for use and to illustrate the efforts required to make this happen. This study provides information about which research questions could be addressed and which questions can be reliably answered when individual level administrative data on health service use is collected.
For the purposes of this study anonymous person-level data from three countries (Estonia, Finland and Israel) was used. To mitigate legal, ownership, confidentiality and privacy concerns, a stepwise decentralized approach was developed.
This coordinated approach allowed for internationally comparable data in calculating many relevant performance indicators which are novel in cross-country comparison of individual level data. Read more
- All-cause mortality in patients with diabetes.
- Outcomes and complications of diabetes treatment – (cumulative) proportion and frequency of patients' hospitalized and treated for coronary revascularizations, myocardial infarction, stroke, lower limb amputations, end-stage renal disease (dialysis or with a kidney transplant), and/or eye complications (retinopathy or cataract).
- Visit frequency and proportion of patients visiting primary care doctors and/or specialists, calculated as no gap more than 365 or 730 days, and not visiting a doctor for 8 years.
- Extent and consistency of pharmaceutical care as proportion of diabetic patients using insulin and different oral antidiabetic drugs regularly (at least 2 separate purchases per year), as well as proportion of patients using statins, antihypertensives and antidepressants.
- Regular monitoring for hyperglycemia (HbA1c), cholesterol (LDL) and regular testing for microalbumin or albumin/creatinine, calculated as no gap more than 365 or 730 days, and never tested during 8 years of follow-up.
- Cumulative cost-of-treatment of the diabetes patient cohort and use of resources, presented as cumulative bed-days, doctor visits and medication purchased per 1000 patient-years.
- Unit costs as average (cost weighted) use of resources per patient treated.
Case study article in Health Policy:
Kiivet R, Sund R, Linna M, Silverman B, Pisarev H, Friedman N. (2013) Methodological challenges in international performance measurement using patient-level administrative data. Health Policy (2013).