Aim: To assess the value of maintaining a death register in a general practice with particular reference to monitoring quality of care. Design of study: Observational study. Setting: Inner London general practice. Method: The practice maintained a manual death register, retained medical records of all deceased patients, and requested information on cause of death from health authorities and coroners for 15 years. Main outcome measures: Number and causes of deaths; 3 yearly age standardised death rates; proportion of deaths formally notified to the practice; place of death; source of cause of death information. Results: During the study period 578 patients died. Practice age standardised death rates fell significantly from 35.59 to 27.12/1000. 498 (86.2%) deaths were formally notified to the practice, 392 within 7 days of death. Of 143 deaths reported to the coroner, only 45 coroners' reports were received. 360 (64.1%) died in hospital, 139 (24.8%) at home, and 38 (6.8%) in a hospice. Death certificate cause of death information was obtained from patients' records in 33.6% (n=194) of cases and from health authority sources for 50% (n=289). The pattern of ascertained causes of deaths was similar to the national pattern. Conclusion: A death register can examine trends in practice deaths by age and place of death and comparisons undertaken with nationally published mortality data. An accurate picture of cause of death cannot be generated from routine data flows alone. There is delay in informing GPs of patient deaths. Meaningful and timely monitoring of deaths cannot be undertaken by individual practices. National Statistics should provide routine analysis of GP death certificate information. P atterns and rates of mortality provide important evidence of healthcare outcomes. Since the first published analysis of deaths in a UK general practice, 1 subsequent studies have examined several aspects of death registers in primary care. Primary care trusts (PCTs), charged with establishing systems for clinical governance that monitor effectively the quality of care provided to their populations, 15 are currently examining whether practice level information on mortality can provide reliable markers of primary health care quality (box 1). The recent conviction for multiple murders of the GP Harold Shipman and subsequent enquiries into how his crimes remained undetected over a 23 year period bring new urgency to monitoring deaths in general practice (box 2). [16][17][18] In this paper we present findings from an analysis of a general practice death register maintained over a 15 year period and examine its potential to contribute to clinical governance at practice and PCT level. Whether a register of deaths offers relevant information for monitoring quality of care depends on whether it can be compiled from routinely collected data that allow a practice to undertake meaningful clinical governance in relation to practice mortality (box 3).
METHODSThe study was carried out between August 1985 and July 2000 in an inner London pra...