Measuring quality in general practice. Pilot study of a needs, process and outcome measure.

1. As part of the Scottish Shadow Fundholding Evaluation (1990-92), quality of care was assessed in 6 practices with 49 general practitioners using a pre-consultation health needs questionnaire, consultation length as a process measure (previously shown to be a proxy measure for quality) and a post-consultation satisfaction/outcome measure which contained a subset of six questions assessing whether patients felt enabled by their consultation. This report describes secondary analysis of the available dataset undertaken to explore whether the approach used to evaluate quality of care for patients with specific clinical problems could be extended to the generality of general practice consultations. 2. Chapters 1 and 2 of the report describe earlier work developing both the concepts and instruments used in the Shadow Evaluation, and general findings already reported. The reliability and the construct validity of the measure of enablement are examined and found to be satisfactory. 3. Strong correlations between more time at consultations and more enablement for more patients are found at population level for patients with psychological problems, with social problems and with physical problems. More complex problems require more time to achieve equal benefit. 4. Mean consultation length and mean enablement score correlate well with each other and can be used as summary statistics of quality. Where trends require explanation or exploration, other measures of the use of time and the level of benefit gained are more helpful; both sets of analyses can be derived from the same datasets (Chapter 3). 5. Analyses at practice level show that practices which spend more time at consultations enable patients more whatever the nature of problems presented. The rank orders of time spent at consultation and of enablement are highly correlated (Chapter 4). 6. Analyses at doctor level show that doctors who spend more time at consultations enable patients more and that those who spend less time enable patients less. The numbers of patients available for study were not sufficient to explore this association within subgroups of clinical presentations. As in previous studies, we found that doctors who take longer time are likely to be more patient centred, and those who take less time are likely to be less patient centred. Case-mix varies between doctors, but seems to be independent of whether a doctor is more or less patient centred (Chapter 5). 7. The methods developed in this study give useful insights into the definition and delivery of quality of care in general practice (Chapter 6). The measures now need to be tested in different clinical, cultural and organizational settings and results compared with those found using routinely available NHS data on prescribing and achievement of other clinical and management targets (Chapter 7).