gms | German Medical Science

Background and current state of (inter)national research: The market entry for new drugs in Germany is regulated and the decisions about reimbursement and price setting depend on the benefit assessment of the GBA within the AMNOG process.

Sacubitril/Valsartan (Entresto^®) was admitted to the market in January 2016 and was approved to have a considerable benefit for patients with heart failure. The market price at entry was about 50 times higher than the comparative therapy with ACE-inhibitors. The national guideline recommends treatment with Entresto^® only if standard medication therapy is ineffective.

In the context of drug diffusion, earlier analyses have shown that beside physicians’ characteristics, their social setting and direct patients’ demand influence their prescription and thereby adoption decision of new drugs.

Research questions and objectives: The study aims to analyze the diffusion process of Entresto^® among ambulatory physicians in Germany. We are interested in peer effects in patient-sharing networks that influence the adoption process, the role of repeat prescriptions, as well as differences between recommended and not recommended prescriptions.

Methods or hypothesis: We analyze routine data of three regional statutory health insurances (AOK) from four federal states from the years 2016 and 2017. We identify all patients receiving Entresto^® and their prescribing physicians. For each physician we examine the date of first prescription as time of adoption and determine her peer network resulting from commonly treated patients. We apply survival analysis methods to identify factors, influencing time of adoption.

We hypothesize that physicians who share patients with physicians who are already prescribing Entresto^® are more likely to prescribe it as well and will adopt earlier. Additionally, we assume that patients might accelerate the diffusion process of the new drug because of repeat prescriptions.

Results: In total about 7,200 physicians prescribed Entresto^® at least once and 65% of these physicians initiated a medication with this drug for one patient. About 2,000 physicians prescribed the drug initially to patients for whom it was presumably not recommended following the guideline.

The survival analysis confirmed an existing positive influence of peer effects in patient-sharing networks and between colleagues within a practice. There also exists an accelerating effect, if physicians treat patients who already got a prescription before.

Discussion: The study showed that even though the analyzed drug is significantly more expensive than the comparative medication, there exists a considerable amount of physicians prescribing it, in some cases not following guideline recommendations. Beside known factors, peer effects in patient-sharing networks exist and patients’ premedication impact physicians’ adoption time.

Practical implications: The analyses indicate that also in the context of efficiency evaluation, influences on physicians’ prescribing behavior are multifactorial. The results indicate a patient impact on the diffusion of new drugs in the ambulatory sector but also interdependencies in the adoption process among physicians.