Comparing the quality of care in physician networks to usual care for elderly patients in three German regions: a quasi-experimental cohort study

Objectives: Patients in Germany have free choice of physicians in the ambulatory care sector and can consult them as often as they wish without a referral. This can lead to inefficiencies in treatment pathways. In response, some physicians have formed networks to improve the coordination and quality of care. This study aims to investigate whether the care provided by these networks results in better health and process outcomes than usual care.

Study design: This was a quasi-experimental cohort study.

Methods: We analysed claims data from 2017 to 2018 in Bavaria, Brandenburg, and Westphalia-Lippe. Our study population includes patients aged 65 years or older with heart failure (n = 267,256), back pain (n = 931,672), or depression (n = 483,068). We compared condition-specific and generic quality indicators between patients treated in physician networks and usual care. Ambulatory care-sensitive emergency department cases were used as a primary outcome measure. Imbalances between the groups were minimized using propensity score matching.

Results: Rates of ambulatory care-sensitive emergency department cases yielded insignificant differences between networks and usual care in the depression and heart failure subgroups. For back pain patients, rates were 0.17 percentage points higher (P < 0.01) in network patients compared with usual care. Among network patients, generic indicators for prevention and coordination showed significantly better performance. For instance, the rate of completed vaccination against influenza is 3.03 percentage points higher (P < 0.01), and the rate of specialist visits after referral is 1.6 percentage points higher (P < 0.01) in heart failure patients, who are treated in physician networks. This is accompanied by higher rates of polypharmacy. Furthermore, the results for condition-specific indicators suggest that for most indicators, a greater proportion of the care provided by physician networks adhered to national treatment guidelines.

Conclusions: Our findings suggest that physician networks in Germany do not reduce rates of ambulatory care-sensitive emergency department cases but perform better than usual care in terms of care coordination and prevention. Further research is needed to confirm our findings and explore the implications of the potentially higher rates of polypharmacy seen in physician networks.