شناسایی اثرات پزشکی همسالان بر همسالان با استفاده از داده حرکت بیمار

In this paper, we identify and quantify peer-to-peer effects using physician prescription data and patient movement data between physicians. We categorize the movements into three types: 1) primary care physician (PCP) to specialist and back, 2) specialist to specialist, and 3) PCP to PCP. In-depth physician interviews and surveys reveal different reasons for these movements: PCP to PCP is purely patient-generated; PCP to specialist is mostly physician-generated; and specialist to specialist is a mix of patient- and physician-generated movements. We estimate a simultaneous equations model on these three types of movements and find that in the purely patient-generated movement sample (PCP to PCP), the physicians have a significantly negative effect on each other's prescription behavior due to observational learning and congestion effects. In contrast, in the PCP to specialist sample and the specialist to PCP sample, we find that the specialist has a significantly positive effect on the PCP but not vice versa. This result suggests an opinion leader effect. Specialist to specialist movement is a mixed case, and the effect is insignificant in most cases. Based on model estimates, we calculate the social multiplier to quantify the effect of opinion leaders on other physicians in the sample. We find focal specialists who are high prescribers are more likely to be opinion leaders.

Interest in tapping customer networks to increase revenues in mature markets has been growing exponentially. Marketers have tried to exploit the potential of online peer-to-peer networks using viral marketing campaigns (De Bruyn & Lilien, 2008). Firms track how consumers interact in online social networks to target these consumers better (Steel, 2010). There has also been an increase in loyalty programs and ‘referral’ marketing, which involves a deliberate, structured program of soliciting and rewarding referrals from current customers. The key to better utilizing consumers' social networks is to understand their structure and how their members influence each other. In this research, we try to quantify peer-to-peer effects using observable transaction data. We use the context of the pharmaceutical industry for this research. The identification of opinion leaders in physician networks is a substantial opportunity for pharmaceutical firms that, faced with declining returns, are making several changes to the traditional sales channels. First, pharmaceutical firms have streamlined the traditional ‘one-size fits all’ model. The overall size of the industry's U.S. sales force declined 10% to about 92,000 in 2008 from a peak of about 102,000 in 2005. All major US pharmaceutical firms have made significant reductions in sales force headcount since 2006, with potentially more reductions to come (Pettypiece & Alesci, 2009). In addition, firms are pulling back on Direct to Consumer spending. For example, GSK plans to cut back advertising on TV in 2009 (Whalen, 2009). Second, pharmaceutical firms are making refinements to how the sales model works at the regional level by giving greater autonomy to regional sales forces. ‘Local peer-to-peer networks’ offer a significant opportunity for pharmaceutical firms to improve the effectiveness of marketing to physicians. While the national key opinion leaders are well respected as academicians/thought leaders who publish in leading journals, physicians have a stronger referral relationship with local opinion leaders. In addition, physicians’ interactions with national key opinion leaders are not sustained in nature because these leaders are not local. Identifying peer-to-peer networks at the local territory/district level is a key hurdle limiting impact and adoption of a localized sales model. Information typically resides in the field and is not compiled for usage because there is limited bandwidth and there are no consistent sets of criteria applied by sales representatives. Additionally, there is no one-stop information source to provide detailed network information at the local territory/district level. To complicate matters further, these networks of influence differ by therapeutic area. The focus of this paper is on identifying the ‘market leaders’ who are closely connected to the local physician population and not the ‘clinical leaders’ who are identifiable at the national level ( Stremersch & Van Dyke, 2009).

This paper models the peer-to-peer effect among physicians in the local region using patient movement data combined with physician prescription and targeted marketing data. We categorize the patient movements as patient-generated and mostly physician-generated based on information from a physician survey. We find that the groups we categorize as having mostly physician-generated patient movements (SP–PCP or PCP–SP) reflect the physician network so that specialists’ prescriptions have the opinion leader effect on the PCPs’ prescriptions. The patient-generated movements (PCP–PCP) reveal private information about the starting physician to the ending physician, leading to observational learning by the ending physician, whose prescriptions are negatively influenced by the starting physician's prescriptions. The starting physician's prescriptions are also negatively influenced by the ending physician due to congestion effect.