Information exchange networks of health care providers and evidence-based cardiovascular risk management: an observational study

Abstract

Background: Although a wide range of preventive and clinical interventions has targeted cardiovascular risk management (CVRM), outcomes remain suboptimal. Therefore, the question is what additional determinants of CVRM and outcomes can be identified and addressed to optimize CVRM. In this study, we aimed to identify new perspectives for improving healthcare delivery and explored associations between information exchange networks of health care providers and evidence-based CVRM.

Methods: This observational study was performed parallel to a randomized clinical trial which aimed to improve professional performance of practice nurses in the Netherlands. Information exchange on medical policy for CVRM ("general information networks") and CVRM for individual patients ("specific information networks") of 180 health professionals in 31 general practices was measured with personalized questionnaires. Medical record audit was performed concerning 1620 patients in these practices to document quality of care delivery and two risk factors (systolic blood pressure (SBP) and LDL cholesterol level). Hypothesized effects of five network characteristics (density, frequency of contact, centrality of CVRM-coordinators, homophily on positive attitudes for treatment target achievement, and presence of an opinion leader for CVRM) constructed on both general and specific information exchange networks were tested and controlled for practice and patient factors using logistic multilevel analyses.

Results: Odds for adequate performance were enhanced in practices with an opinion leader for CVRM (OR 2.75, p < .05). Odds for achievement of SBP targets were reduced in practices who had networks with low homophily on positive attitudes for SBP and LDL targets (homophily for SBP targets OR 0.57, p < .05 and OR 0.60, p < .05, homophily for LDL targets OR 0.59, p < .05 and OR 0.61, p < .05 in general and specific information networks, respectively). No effects of network characteristics on cholesterol were found.

Conclusions: Delivery of evidence-based CVRM is associated with homophily of clinical attitudes and presence of opinion leaders in primary care teams. These results signal the potential of social networks to be taken into account in further attempts to improve the implementation of evidence-based care for CVRM. Future research is needed to identify and formulate optimal strategies for using opinion leaders to improve CVRM. Future interventions may be more effective if they target a common vision on CVRM within practices.

Keywords: Cardiovascular disease; Evidence-based practice; Implementation science; Primary care; Social network analysis.

Fig. 1

Hypothesized relations