It’s good to talk: social network analysis as a method for judging the strength of integrated care

Last year I was undertaking an interview with a healthcare manager in England who was bemoaning the lack of care integration between primary care (general practice) and secondary care (hospitals) in the management of people with diabetes. When I asked her when she might know that the system had improved in the future she answered—‘when GPs routinely pick up the telephone and talk to the consultant about a patient’s diagnosis or care options … and actually work together to support the patient in the care they receive’.

The quote encapsulates everything that has gone wrong in England in terms of care fragmentation, especially for people with long-term care needs. The English system does not adequately value care co-ordination as a marker of quality. Indeed, care has become commodified to the extent that hospital consultants are sometimes actively discouraged from developing professional relationships with their primary care colleagues where this is not remunerated. Without any constructive dialogue between supposed partners in care there is clearly little chance to achieve better care integration.

In many ways it seems obvious that examining the extent to which people are communicating with each other can be used as a proxy for success in integrated care. For example, a systematic review of different strategies to co-ordinate care within primary care, and between primary care and other health-related services, showed that improved communication between providers was a pre-requisite for successful care integration [1]. However, there are several challenges to testing the value of care co-ordination emprically: the complexity of the multiple linkages that exist; the challenges in adjusting for patient-related and external factors influencing outcomes; and many aspects of prolonged, coordinated, interpersonal care and informed self-management that are difficult to measure quantitatively.

These issues were brought home to me at this year’s Annual Integrated Care Conference in Finland [2]. In particular, a paper by Professor Mike Martin from Newcastle University, UK, described how supporting connectivity in information exchange and dialogue between care providers was potentially more likely to yield better results as a strategy than the more traditional focus on developing organizationally-based solutions. He demonstrated how the design and assessment of care integration might be better developed and examined through the lens of social network analysis.

Social network analysis measures the relationships and ties between people and/or groups within a network and so can generate a lot of useful information—for example, how many people are in the network?; who is (or is not) talking to who?; who lies at the centre or edge of the network?; who is driving the agenda?; have cliques formed?; is the network cohesive or are there ‘holes’ to be bridged?; and so forth. In other words, it is measuring the degree of social contact between individuals and can be used to measure social capital—the value that individuals get from being in a network which is often reported as important in building partnerships across health and social care.

For studies of integrated care there seems to be clear value in adopting social network analysis as a key analytical approach in our scientific field of enquiry. There have already been some attempts at this. For example, two recently published IJIC papers by Wiktorowicz et al. [3] and Holmesland et al. [4] have both sought to interpret care integration in this way. However, given the sophistication in the range of analytical tools that are now used in social network analysis in other scientific disciplines, this is an innovative methodological approach that might significantly advance our ability to ‘measure’ key aspects of care integration in a way that has not been developed before. The growth in electronic records and communication systems in healthcare would also support the practicality of such analysis.

Of course, whilst good communication between providers is a pre-requisite for integrated care, it does not tell us directly whether care (as experienced by patients) is any better or worse. In particular, there have been many examples over the years of strategies bringing different professionals and organizations together that end up as ‘talking shops’—discussing new forms of integrated care but delivering very little. For example, in 2008, a report of a study tour of two innovative integrated care projects in Sweden noted that these were characterized by a healthy and open culture for discussion and debate between care staff and managers but that there was relatively little focus on developing new ways of working [5].

In conclusion, it is clearly ‘good to talk’ since effective dialogue and communication is a pre-requisite of integrated care. It can help build social capital and so predispose individuals and groups to work collaboratively. Embracing analytical techniques such as social network analysis might enable us to measure and assess the nature and strength of these relationships in a way that has not been done before. This is important since the ‘hard to measure’ elements of integrated care need to be better captured such that they begin to demonstrate their value in supporting improvements in care. Those tasked with the design, implementation and evaluation of integrated care schemes might then seek to focus as much on the ‘inner workings’ of care integration as with the organisational structures, governance arrangements and incentive frameworks that provide the overarching framework in which it takes place.