Social network analysis in primary care: The impact of interactions on prescribing behaviour
Introduction
In the last 20 years healthcare reforms in Western Europe have reshaped primary care in general and the role of GPs in particular. Such reforms have broaden GPs’ service portfolio [1], introduced innovative financial incentive schemes [2], [3], [4], attributed to GPs a leading role in purchasing and/or coordinating providers, especially in Beveridge systems, [1], and promoted cooperative arrangements across the interface [1], [5]. All these measures are changing the organizational role of GPs and the content of their professional and managerial activities. A major common theme across countries is the promotion of cooperation among GPs as a means to spread knowledge, facilitate accountability and, in the end, improve patients care with limited resources [6], [7], [8]. In this paper, we use social network analysis to study whether the collaboration initiatives launched by a local health authority (LHA) in Italy between 2001 and 2004 have had any effect on individual and district-level GP performance, as measured in terms of meeting drug expenditure targets.
In the Italian National Health Service, LHAs are responsible for providing comprehensive healthcare to the resident population of their area and are divided into subunits called districts where initiatives to coordinate GPs work and other community care activities are generally managed. Among other activities, LHAs manage general practitioners, who are family doctors working for LHAs as independent contractors and acting as gatekeepers to higher levels of care. Traditionally, GPs in Italy have worked in solo practices without any auxiliary staff or institutional links to other GPs. However, organisational models and management systems have significantly changed over the past 10 years [9]. Several LHAs have introduced budgeting systems and financial incentives to make GPs more accountable for their prescribing and referral behaviour, and have invested in sophisticated information systems to provide detailed reports on GPs prescribing patterns.
Many LHAs are also trying to reshape the traditional model of primary care by encouraging GPs to participate in collaborative arrangements such as group practices in which GPs share practice space and other resources. The basic idea behind these initiatives is that organisational and professional development in primary care requires cooperation and the sharing of resources and knowledge. A vast body of research has investigated the impact of financial incentives, other interventions and individual factors on GP behaviour [10], [11]. Instead, the present study focuses on the implications of relationships created through collaboration initiatives launched by an LHA. We investigate determinants of medical practice variation in primary care from a supply perspective, which states that physician characteristics and the environment in which physicians operate may shape patterns of healthcare utilisation [12]. Within this perspective, we adopt a relational (or network) framework that can be seen as complementary to economic models of incentives created by payment systems [2], [13], as well as to clinical explanations based on physicians’ response to uncertainty [13], [14].
The various collaborative arrangements in which GPs participate are forging sophisticated primary care networks, a development that raises two important questions: what are the consequences of these relationships, and how are these relationships influencing GP behaviour?
We argue that the social network paradigm represents a useful approach to answering these questions. Social network analysis has been variously used to describe the structure of relationships, to investigate the consequences of a given structure [15], and to understand individual outcomes as a function of an actor's social network characteristics (see Borgatti and Foster [16] for a review of this literature). Indeed, recent contributions in primary care have started to use Social Network Analysis as a tool to describe and analyze patters of relationships among physicians in primary care [17], [18]. We follow this line of research and explore the structure and consequences of collaborative arrangements in an Italian LHA that launched a number of GP collaboration initiatives.
Section snippets
Theoretical framework and hypotheses
The social network perspective is based on the premise that actors are embedded within networks of social relationships [19], [20] and that these networks have ‘important behavioural, perceptual, and attitudinal consequences’ [21]. There are two fundamental theoretical strands on the mechanisms underlying the behavioural and attitudinal consequences of networking: (1) the social capital perspective, and (2) the social influence perspective [16].
The social capital perspective views networks of
Materials and methods
The network data used in this paper were provided by a LHA of Empoli in the Tuscany region of Italy, which comprises two districts (henceforth designated as districts A and B) with a total population of 227,038. Data on GP activities, including information about formal networks, are collected regularly and archived digitally by the LHA for its operational activities. From these archives, we obtained data about the membership of individual GPs in different types of collaboration initiatives from
Results
In the present analysis, the dependent variable was expressed by the difference between an individual GP's per capita pharmaceutical expenditure and the LHA's assigned target. Within the social capital framework, Hypothesis 1 posited that the more central a GP is to his or her network, the more the GP is able to meet the target assigned by the LHA. As shown in Table 2, centrality has a very small and insignificant effect, which means that Hypothesis 1 is not supported by our data.
Grounded in
Discussion
In this study we investigated the individual and district-level impact of the collaboration initiatives launched by an Italian LHA between 2001 and 2004. In line with recent research in primary care [17], [18], social network analysis provided us with the theoretical and instrumental framework for identifying specific hypotheses, measures, and empirical models. The main contribution of our study is its comparison of two mechanisms – i.e. social capital and social influence – by which different
Conclusions
A first contribution made by this study is in the provision of a preliminary evaluation of the consequences of an LHA's initiatives on GP performance, expressed in terms of GP prescribing behaviour. Analyses of formal ties created by collaboration initiatives are important because policy makers and managers have some control over the development of these relationships, as opposed to the control over informal ties, which is weaker. Many LHAs are considering or are already implementing
References (63)
- et al.
Mutual influences of general practitioners in partnership
Social Science and Medicine
(2003) - et al.
Professional uncertainty and the problem of supplier-induced demand
Social Science & Medicine
(1982) - et al.
The network paradigm in organizational research: a review and typology
Journal of Management
(2003) The art and science of clinical knowledge: evidence beyond measures and numbers
Lancet
(2001)- et al.
Hierarchies and cliques in the social networks of health care professionals: implications for the design of dissemination strategies
Social Science & Medicine
(1999) - et al.
Dissemination of effectiveness and outcomes research
Health Policy
(1995) - et al.
Organizational restructuring in European health system: the role of primary care
Social Policy and Administration
(2004) - et al.
Effects of financial incentives on medical practice: results from a systematic review of the literature and methodological issues
International Journal for Quality in Health Care
(2000) - et al.
Managing primary care behaviour though payment system and financial incentives
- et al.
A literature review