A comparison of research general practices and their patients with other practices - A cross-sectional survey in trent

Background: When interpreting results of studies undertaken by research networks we need to know how representative volunteer practices and their registered patients are of the total population of practices and patients in their locality. Aim: To compare the following in research and non-research general practices in one region: practice and population demography, morbidity and mortality, selected performance indicators, and health outcomes. Design of study: Cross-sectional survey. Setting: Sixty-six Trent Focus Collaborative Research Network general practices and 749 other general practices in Trent, United Kingdom. Method: Practice characteristics and GP contract data were obtained from the NHS Executive, Quarry House, Leeds. The Trent Regional NHS Hospital Admission Database was searched to identify all relevant admissions to hospital from all practices between 1 April 1993 and 31 March 1997. Ward-linked data on cancer were obtained from the Trent Cancer Registry. Results: Of the 815 general practices in Trent Region in the study period, 66 (8%) were in the Trent Focus network. They were more likely to be involved in training GPs and to have a female partner. They tended to be larger, with fewer single-handed doctors and younger GPs. Network practices prescribed a higher proportion of generics (median % prescribed/practice = 70%, versus 51%, Mann-Whitney U = 1615, P<0.0001). There were no clinically important differences between hospital admission rates between the two groups or waiting times for surgical procedures. There was no difference in the incidence of cancer and standardised mortality ratios related to the electoral wards of the GP surgery. Conclusion: Although there were differences in practice structure and some aspects of performance, we found no important differences in the demography of registered patients, nor in morbidity, mortality, or access to or use of secondary care.