Brief commentary on “Evaluating integrated health care: a model for measurement”
International Journal of Integrated Care, 31 August 2005 - ISSN 1568-4156
Guest editorial
Brief commentary on “Evaluating integrated health care: a model for measurement”
Overview

“How integrated are we?” This is an important question for healthcare managers. Answering the question requires a method of evaluation. Ahgren and Axelsson record an explorative study for providing Swedish managers with a useable and useful model for the measurement of integration that has considerable potential.

The model is a further step towards the scientific assessment of functional clinical integration. In this respect, the development of a ratio scale and the use of data based on actual integration represent notable achievements. The authors are careful to point out that these positivist attributes should lead to the judicious informing of managerial actions. The cycle of evaluation involves group self-assessment, data analysis and feedback in a way that ensures comparison of actual and optimal integration.

Taking forwards the Ahgren and Axelsson model

Two issues are of particular interest. First, there are considerations for replicating the work of Ahgren and Axelsson. The second and connected issue is that of precise terminology in the integrated care field.

In describing their explorative study, the authors provide readers with much substantive information, especially the instructions to research participants contained in the appendix. However, if other researchers are to test the reliability and validity of the model, then further details are required. The authors themselves allude to some respondent difficulties in deriving integration ranks. To itemise specific facets that require clarification:

  • In the appendix, one side bar refers to “How to derive the integration level?” Do the authors really mean ‘level’? I think they mean ‘rank’. Am I confused in the same way as respondents?

  • The calculation of rank is a two-step procedure. First, it is required to specify the highest degree of integration between two healthcare units, e.g. presence of network managers. Second, ‘fine tuning’ of this degree is achieved by calculating the relative number of cases in a category. For example, presumably, if all network manager cases are shared then a black square is the measured output. I say ‘presumably’ because I don't think the description is clear enough.

  • Further, in respect to data elicitation and collection, more information about the guidance given to participants would be valuable.

  • Twenty-eight ‘health care units’ are recorded in the appendix but only 20 appear in the data record. Do we need to know about the missing 8 units?

  • The previous point is raised because data were collected dyadically with the implications of a network arrangement. Do the recorded figures represent assessments from both parties of the dyad? Or, do they contain some one-sided assessments due to 8 of the 28 failing to contribute data? It would be interesting to see the raw data as set out in the 20×20 or 28×28 matrix.

  • The paper specifies a number of key terms in the integrated care lexicon, for example: degree of integration and level of integration. The authors also use the terms integration rank and integration scope whose definitions are embedded within the general text. Distinct and formal definitions of key terms are essential for future development of the potential Ahgren and Axelsson have demonstrated.

This final bullet point leads naturally on to ‘terminology’. I suggest that if researchers in integrated care are to establish a truly scientific approach to the discipline then we require greater terminological precision. For example, ‘vertical’ and ‘horizontal’ are two terms in frequent use. In a supply chain sense, we can describe a primary care physician with a contractual linkage to a secondary provider as ‘vertically integrated’.

Implicitly, we adapt our sense of meaning of ‘vertically integrated’ to the context. But in the case of 28 healthcare units, what is meant by ‘vertical’?

To give another example. ‘Cooperation’ is the term nominally given to the degree of integration between ‘coordination in networks’ and ‘full integration’. But don't ‘patient referrals’, a lower degree of integration, also require cooperation?

This blurring of meaning might well be negotiable in the managerial world. However, as research scientists, we have a peculiar duty regarding precision. In the ‘Discussion’ section of their paper, Ahgren and Axelsson remind us of ‘… the importance of using terms like “degree of integration” or “integration rank” in the development of integrated health care, instead of a general term like “integration”.’ Perhaps the time is right for an agreed glossary of key defined terms in integrated care.

Nick Gould

Research fellow Welsh Institute for Health and Social Care (WIHSC), University of Glamorgan, United Kingdom

Morton Warner

Professor and Director Welsh Institute for Health and Social Care (WIHSC), University of Glamorgan, United Kingdom