Clin Res Cardiol (2021)
DOI DOI https://doi.org/10.1007/s00392-021-01843-w

Influence of physicians’ networks on the prescription of a new ingredient combination for the treatment of heart failure
C. Arnold1, J. Koetsenruijter1, M. Wensing1
1Abteilung Allgemeinmedizin und Versorgungsforschung, Universitätsklinikum Heidelberg, Heidelberg;

Background: Because of increased life expectancy and improved treatment, the prevalence of heart failure (HF) is rising.
Drug therapy is critical in the treatment of HF. In 2016 the new combination of
sacubitril and valsartan became available for heart failure treatment in Germany.

The purpose of the study was to determine the influence of the physicians’ networks in shape of linkages to physicians who are already prescribing the new drug on the prescription of sacubitril/valsartan.

Methods: We conducted a cross-sectional analysis based on claims data in 2016 from health insurance company AOK (‘Allgemeine Ortskrankenkasse’) Baden-Wurttemberg.
The characteristics of physician’s networks were derived from a social network analysis, which considered the networks of physicians who were connected by at least five shared patients.

Physicians working in Baden-Wurttemberg with the following specialist titles were included in the analysis: general practitioners, internists, cardiologists, angiologists, pulmonologists and nephrologists.
Physicians who have prescribed sacubitril/valsartan once in the first year of approval (2016) were defined as ‘prescribers’.

Next, a binary logistic regression analysis with outcome ‘prescribing of sacubitril/valsartan’ (yes or no) was carried out, using number of linkages to prescriber as independent variable and sociodemographic characteristics as predictors.

Results: 8,371 physicians were included in the network analysis and they had 143,648 connections (mean 34.3 (SD 34.2) per physician).
Prescribers reported on average 47.8 (SD 40.7) connections while the non-prescribers showed 32.2 (SD 32.6) connections.
In 2016, 1,127 (13.5 %) physicians prescribed sacubitril/valsartan: 37.7 % among cardiologists and 12.7 % among primary care physicians.
Compared to non-prescribers (mean 4.5 (SD 5.0)), prescribers had 6.4 (SD 6.1) linkages to physicians who have already prescribed the new drug.
The regression analysis demonstrated a significant yet small influence of the linkages to prescribers of the new medication on individual prescribing (OR 1.05 (p < 0.001)).

Discussion: Physicians who share patients with other physicians who prescribe sacubitril/valsartan are more likely to prescribe this new medication as well.
In further analyses we will examine the influence of other network characteristics on the medication prescribing behaviours of physicians.

 


https://dgk.org/kongress_programme/jt2021/aP1603.html