Construction of a GP integration model
Introduction
Complex, multi-dimensional social concepts or phenomena deemed to have positive impacts on community health (such as ‘community empowerment’, ‘service responsiveness’) are in widespread use. Surprisingly, there have been few attempts to examine their meaning closely or to identify their dimensions. Perhaps their very complexity deters such efforts. However, study of the concept would have value in demonstrating that it has, or does not have a stable reality, judged by agreement about its meaning and existence of stable patterns of observations implying such a construct.
The complex, multi-dimensional concept studied in this paper is ‘integration of general practitioners with the health care system’. In the words of the emergent realists, the paper investigates whether ‘GP integration’ is something ‘real’ not merely a convenient shorthand term to describe a class of programs (Cronbach, 1989). Given the lack of preexisting theory, its methodology is exploratory and iterative and inevitably involves judgements of the investigators which are influenced by their presuppositions and values. There is however no alternative approach save abandoning the task.
In this paper, Concept Mapping is used to specify a theoretical model, which can be subsequently tested by confirmatory factor analysis, using structural equation modelling (SEM) in an appropriate data set. This hypothesis generation and testing approach contrasts with the traditional hypothesis generating step only of exploratory factor analytic (EFA) approach (McArdle, 1996). In the latter, a convenience sample of subjects or researchers is typically used to generate statements or questions for which no model is specified. Factors are subsequently derived from these statements in a data set using EFA.
Concept Mapping is a nominal group technique that permits the rapid development of a conceptual framework for any given topic (Trochim, 1989a). It is expressed entirely in the language of the participants and yields a graphic or pictorial product which simultaneously shows all major ideas and their interrelationship. To construct the map, ideas (statements) have first to be described or generated and the interrelationship between them articulated. Multi-variate statistical techniques are then applied to this information and the results are depicted in map form.
The paper describes a measurement model of GP integration as distinct from a structural causal model which would investigate the relationship between the whole (or indeed parts) of the GP integration model with other concepts and variables. It is necessary for the measurement model to be tested and refined before detailed exploration of causal models is possible.
Section snippets
The role of the general medical practitioner (GP) in the Australian health care system
The Australian health care system is a mixed public/private system. It is further fragmented by a federal system of government where the Commonwealth Government's role is to fund public health services within a large number of programs and to underwrite private care, and the States and Territories’ role is to be responsible for the delivery of public health services. Universal medical insurance covers medical practitioner services but fee rebates for these services are only obtainable if the
Developing a conceptual schema to describe GP integration
To begin, the Macquarie Dictionary (1997) defines integration as ‘the act of bringing together parts (into a whole)’. Coordination, by comparison, is defined as ‘the act of arranging according to a plan especially one drawing together a number of different events, organisations, people, etc.’ and ‘working together smoothly in combination’.
Descriptions of integration of health services (not GP integration) that exist in the research literature are primarily used for purposes of classification.
Specifying a model of GP integration through application of Concept Mapping techniques in small groups
Concept Mapping was used to specify the model of GP integration to be subsequently tested using SEM. It is a nominal group technique that permits the rapid development of a conceptual framework for any given topic (Trochim, 1989a). It is expressed entirely in the language of the participants and yields a graphic or pictorial product which simultaneously shows all major ideas and their interrelationship. To construct the map, ideas (statements) have first to be described or generated and the
Confirming and re-specifying GP integration models using SEM in a national GP survey data set
Two variants of the hypothesised models for the integration process factors and the enabling factors were proposed. The simpler without higher order factors is shown in Fig. 3 above. A more complex variant of the integration process factor model (not diagrammed) encompassed the simple variant plus three mediating higher order factors between the higher order factor integration and the five individual factors. These three higher order factors represented several though not all of the five
Scoring the index
Each factor was scored from 1 to 5, these scores being derived by averaging the scores on the five-point Likert scale of the constituent items. This adjustment was necessary as factors varied in the number of constituent items and some missing data existed. Higher order factors were scored by averaging the scores of the contributing scales rather than averaging the scores of all items within the scales. This was done to avoid weighting the score in favour of scales with more items.
In principle
Discussion
To what extent then has the aim of the study been realised? What is the content and what are the boundaries of the complex, multi-dimensional social concept GP integration with the health care system. Does it have a stable reality and describe a stable pattern of events? In the words of the emergent realists, is ‘integration’ real and not merely a convenient shorthand term to describe a class of programs (Cronbach, 1989)?
The results firstly indicated that GP integration could be conceptualised
Acknowledgements
We thank all GPs, consumer representatives, allied health professionals, nurses, specialist doctors, hospital administrators and associated professionals participating in Concept Mapping sessions. We also thank all GPs completing the national mail questionnaire surveys. The members of the GP Steering Group comprised A/Profs. Phillip Hegarty, Remi Guibert, Drs. Peter Waxman, Michael Montalto, Stephen Wales and Mr. Richard Smith. The members of the project's Reference Group comprised Drs. Rob
References (37)
A pattern matching approach to assess the construct validity of an evaluation instrument
Evaluation Program and Planning,
(1989)An introduction to concept mapping for planning and evaluation
Evaluation Program and Planning
(1989)Outcome pattern matching and program theory
Evaluation Program and Planning
(1989)Self-efficacy mechanism in human agency
American Psychology
(1982)EQS structural equations program manual
(1995)- et al.
EQS 5.4 for Windows
(1995) - Commonwealth Department of Health & Community Services (1995). Meeting people's needs better. A discussion paper...
- Cronbach, L.J. (1989). Construct validation after thirty years. In Linn, R.L. (Ed.), Intelligence: Measurement, theory,...
Issues in factorial invariance
- et al.
Primary care: core values developing primary careGate keeping, commissioning, and managed care
British Medical Journal
(1998)
Cluster analysis
Case management for special populationsMoving beyond categorical distinctions
Journal of Case Management
Using LISREL to evaluate measurement models and scale reliability
Educational and Psychological Measurement
Concept mapping for planning and evaluation of a Big Brother/Big Sister program
Evaluation Program and Planning
Improving general practitioner involvement in urban hospitals. Departments or divisions of general practice
Medical Journal of Australia
PRELISA program for multivariate data screening and data summarization: A preprocessor for LISREL
Cited by (29)
The use of concept mapping in measurement development and evaluation: Application and future directions
2017, Evaluation and Program PlanningCitation Excerpt :First, a set of studies focused on the creation of an instrument to be used in the evaluation of an intervention (Ciciriello, Buchbinder, Osborne, & Wicks, 2014; Osborne, Elsworth, & Whitfield, 2007; Rosas & Camphausen, 2007). Second, several studies emphasized the purpose of the research was in response to a need to better conceptualize and measure complex phenomenon (Battterham et al., 2002; Behfar, Mannix, Peterson, & Trochim, 2011; Conrad, Iris, Ridings, Langley, & Anetzberger, 2011; Conrad, Iris, Ridings, Langley, & Wilber, 2010; Iris, Conrad, & Ridings, 2014; Jordan et al., 2013; Wolfinbarger & Gilly, 2003). Third, a set of studies focused on developing more comprehensive instruments that addressed the limitations (e.g., coverage, content, scope) found in previous efforts (Osborne, Batterham, Elsworth, Hawkins, & Buchbinder, 2013; Rosas, Behar, & Hydacker, 2014; Van Haitsma et al., 2012; van Nieuwenhuizen, Schene, Koeter, & Huxley, 2001).
The Health Literacy Management Scale (HeLMS): A measure of an individual's capacity to seek, understand and use health information within the healthcare setting
2013, Patient Education and CounselingCitation Excerpt :In addition to in depth interviews, the conceptual model development was guided by concept mapping, a structured group process that integrates qualitative and quantitative approaches to derive a graphical representation of major ideas about the phenomenon of interest [34]. The authors have used this technique previously in the development of questionnaires [35–38]. It involved five steps:
A systematic grounded approach to the development of complex interventions: The Australian WorkHealth Program - Arthritis as a case study
2010, Social Science and MedicineCitation Excerpt :Participants were invited purposefully to ensure national representation, breadth of representation within the categories and interest in contributing to the process (Patton, 2002). A concept mapping process was undertaken to help specify program principles, aspects of service delivery and constituent elements (Batterham et al., 2002; Trochim & Linton, 1986; Trochim, Milstein, Wood, Jackson, & Pressler, 2004). Concept Mapping is a structured, systematic, computer-assisted qualitative method of obtaining and organising the largest possible breadth of ideas on a topic.
Individual and area factors associated with general practitioner integration in Australia: A multilevel analysis
2006, Social Science and MedicineOvert and covert barriers to the integration of primary and specialist mental health care
2005, Social Science and Medicine