Measuring the Quality of Medical Care

Researchers have attempted to measure the quality of medical care for a considerable time. Perhaps the most scholarly work on the subject was undertaken by Avedis Donabedian in the 1970s who considered that the quality of medical care and its performance should be analysed in terms of three factors: structure, process and outcome, and each of these could be measured. Outcome measurements had already been used in a number of studies, particularly in the investigation of maternal mortality. In the United Kingdom, the Royal College of Obstetricians and Gynaecologists, together with the Ministry of Health, had started a confidential enquiry into maternal mortality as early as 1928. This was developed further in 1948. Obstetricians reported confidentially on every maternal death, giving details of what had been done during the care of the individual mother. These reports were reviewed by two expert obstetricians who commented on what the failures had been, if any, in the care of individual mothers who had died. The results were fed back to the obstetricians involved, in confidence. George Godber gives an account both of the history and methods of this Enquiry. Confidentiality was essential, and the causes of maternal death were classified by a number of criteria. He considers that consultation with regional assessors led to changes in practice and states ‘I have no doubt that the Confidential Enquiry was an important factor in saving maternal lives’. Irvine Loudon, in his historical analysis of maternal mortality, states ‘It is impossible to know what would have happened without this system of continuous audit but the reports certainly give the impression that they identified the avoidable maternal deaths and led to ways of preventing such deaths’. WHO in a report describing the various methods of investigating maternal mortality and morbidity states that even though there is no formal proof of the effectiveness of such enquiries ‘the lessons derived will enable health care practitioners and health planners to learn from the past’. Similar methods were used in the United States, where the New York Academy of Medicine investigated every maternal death in New York City. Facts were carefully collected and the circumstances reviewed and evaluated by a group of outstanding obstetricians. Application of confidential enquiries was followed by sharp decreases in maternal mortality both in the United Kingdom, and in New York. A further example of looking at vital statistics in relationship to the quality of care in hospitals was undertaken by Paul Lembcke. He showed not only variability in the frequency and outcomes from routine operations in different hospitals but, in contrast to many other studies, calculated rates based on defined populations rather than only on hospital discharges. Building on this and other contributions, David Rutstein and his collaborators examined the causes of death and decided which ones could be used as quality of care indices. The principle was to identify a series of indicators of avoidable disease, disability and death, i.e. sentinel adverse health events. They assumed that if everything had gone well, the condition would have been prevented or managed successfully. Of course, they realized that the chain of responsibility was very long, so it was difficult from their work to identify who was actually responsible for any treatment or system failure. The conditions chosen were divided into three different types. First, those that were clear cut and could be used as immediate quality of care indices, consisting of over a hundred conditions further divided into unnecessary disease, unnecessary disability and unnecessary, untimely death. The majority of conditions in this category were infectious diseases such measles, rubella and yellow fever. There were also a number of chronic conditions including some cancers, such as neoplasm of the cervix, lung and trachea, gout, hypervitaminosis, glaucoma and infections of the skin. The second group of conditions chosen were considered of limited use as indicators of quality of care, Walter W Holland CBE, Emeritus Professor of Public Health Medicine, LSE Health, London School of Economics and Political Science, Houghton Street, London WC2A 2AE, UK.