Original ArticleMorbidity severity classifying routine consultations from English and Dutch general practice indicated physical health status
Introduction
In the United Kingdom [1] and the Netherlands, most of the population are registered with a general practitioner (GP), and in an average British practice there are an estimated 50,000 consultations per year [2]. Population-based consultation data provide an estimate of morbidity and have been used to study health needs relating to specific conditions [2], [3], and to assess health care use as an outcome of primary care interventions [4], [5]. The focus of attention in health care tends to be on arbitrarily defined chronic diseases, major life-threatening disorders such as cancer, or diseases that result in hospital admissions. Little attention has been paid to the much broader spectrum of morbidity which people present to primary care. The importance of this spectrum lies in the fact that it relates to the whole health experience of patients, which is inclusive of symptoms and chronic health states, and the need for the comprehensive health care which general practice attempts to address. In this spectrum, however, individual morbidities may vary in the extent to which they impair overall health, and it is possible that the cumulative effect of consultations for different morbidities over time may provide a more powerful and useful severity measure of individual health status than single and specific diagnostic labels, and yet could still be derived from routinely collected data. The practical usefulness of allocating morbidity to a “severity” measure from the GP viewpoint is that it might enable the grouping of consulters based on overall health need (of single and multiple morbidity consultations) and for planning health care delivery. An alternative approach is to use health-related quality-of-life questionnaires, but these can be cumbersome. Questionnaires provide estimates of health status that are fixed in time unless repeated, but are also time and cost-intensive with concomitant ethical and patient recruitment issues.
The morbidity that patients' present is now routinely recorded by GPs in consultations, particularly in networks dedicated to collecting clinical data. Simple measures of contact frequency have already shown that patient's exhibit specific patterns of consultation [6]. Previous attempts to add a measure of morbidity severity to this basic picture have been based on additional assessment of severity in the consultation, focused on nonprimary care settings [7], [8] or have used a restricted number of morbidities [9], [10] that have not been validated fully. Such measurement of “severity” has taken two forms: (i) based on a priori classification for routinely collected morbidity data, that is, morbidity severity classified relative to one another [7], [9], [10] or (ii) based on classification of the severity of morbidity in each patient [8]. We have worked with a group of UK GPs and used their clinician constructs of morbidity severity to produce a classification of the first type. Because GPs see the fullest range of morbidities, they are perhaps best placed to provide clinical constructs that can order the individual morbidities according to likely health severity. So in the North Staffordshire (English) setting, four scales of morbidity severity based on routine consultation data were first developed with GPs. The objective of the work described in this paper was to validate these severity scales against two constructs: self-reported physical health status and variation with individuals' sociodemographic characteristics.
In the development of any epidemiological measure, it is important that face validation is combined with other tests of validity such as content, construct, and predictive validity. An additional important component is the external validity or generalizability of the tool in different populations' [11]. So in a two-step process, we first validated the scales in general practices from England and then repeated the same validation process in general practices from the Netherlands. Dutch general practice provides an appropriate external reference to English general practice, as they have similar health care systems with general practice as the main access point to health services [12]. The consulting populations selected for study in both countries had also participated in health surveys, thus enabling consultation morbidity data, health status, and sociodemographic data to be linked.
Section snippets
Design
In preliminary work in England, we first carried out focus group meetings to develop scales of morbidity severity that were then applied to a list of morbidities for classification by GPs using consensus methods. In the main study, we applied the severity classification to routine consultation data for a 12-month period from six English general practices that were linked to individual patient health survey data, and to consultation data for the same time period from general practices in the
Statistical analysis
For each of the four scales separately, patients were categorized by the most severe grading for which they had a morbidity consultation in the study year, and two analyses were performed.
First, associations were estimated, for each scale separately, between severity category and (i) age categorized in four 10-year bands and 80+ for older English and Dutch consulters, and for younger Dutch consulters aged 18–34 and 35–49 years, (ii) gender and (iii) deprivation status (Townsend data and Dutch
Results
In the 12-month study period, 9,003 (80.2%) of the 11,232 older English sample had consulted for at least one coded morbidity. In the Dutch sample of 9,664, 3,757 (83.7%) of the “older” patients and 3,996 (77.2%) of the “younger” patients had consulted.
Study findings
We have developed four new scales to classify morbidity severity that can be applied to routinely collected general practice consultation data. Morbidity severity classified on four separate scales was associated with older age, female gender, socioeconomic deprivation, and poor physical health. Our study provides evidence that clinical morbidity across the spectrum seen in general practice can be ordered by severity, using clinical constructs developed by GPs. This framework is a measure of
Acknowledgments
U.T.K. has been supported by a Medical Research Council (UK) Training Fellowship in Health Services Research (G106/1035). Other funding support was by Claire Wand Fund, North Staffordshire Primary Care Research Consortium, MRC Programme grant (G9900220), and NHS R&D funding. We wish to thank all patients, North Staffordshire GP Consortium group, GPs, and the Primary Care Network and Surveys teams at Keele University, and staff at EMGO (Amsterdam) and NIVEL (Utrecht).
References (36)
- et al.
Use of a cross-sectional survey to estimate outcome of health care: the example of anxiety and depression
J Clin Epidemiol
(2001) - et al.
Comparison of physical treatments versus a brief pain-management programme for back pain in primary care: a randomised clinical trial in physiotherapy practice
Lancet
(2005) - et al.
Duke Case-Mix System for ambulatory health care
J Clin Epidemiol
(1997) - et al.
A comparison of two consensus methods in a single professional group showed the same outcomes
J Clin Epidemiol
(2006) - et al.
Data quality of general practice electronic health records: the impact of a program of assessments, feedback, and training
J Am Med Inform Assoc
(2004) - et al.
Frequent attenders in general practice care: a literature review with special reference to methodological considerations
Public Health
(2005) - et al.
Which chronic conditions are associated with better or poorer quality of life?
J Clin Epidemiol
(2000) Research methods in health
(1997)- et al.
Morbidity Statistics from General Practice: Fourth National Study 1991–92. Series MB5 No. 3. Office of Population Censuses and Surveys
(1995) United Kingdom back pain exercise and manipulation (UK BEAM) randomised trial: effectiveness of physical treatments for back pain in primary care
BMJ
(2004)