Background: In-hospital mortality has been proposed as an indicator of health care quality. It is also a fact that should be used for epidemiological monitoring and health planning. The aim of this work is to analyse the mortality in Algeciras Hospital, belonging to the Andalusia Health Authority.
Methods: A descriptive study was carried out of the patients discharged from hospital between January 1995 and December 1996, using discharge data and/or their medical record. The structure of the joint basic minimum data report on hospital discharge was used. The periods of hospitalisation were grouped into the diagnosis-related groups (DRGs). Following an initial analysis of all the periods of hospitalisation, the discharges of patients under 80 years of age were analysed and the risk of death in hospital in line with the reason for admission was set.
Results: In Algeciras Hospital between January 1995 and December 1996 a total of 24,194 episodes of hospitalisation and a 4.1% death rate were recorded. A higher death rate was observed for males (5.3%), as opposed to females (3.2%). (P < 0.01). There were 750 in-hospital deaths in patients under 80 years of age, with a continued higher mortality rate for males. The most frequent causes of deaths in this age group are: 66 deaths caused by cerebro-vascular disease (the probability of dying in this age group if the patients were admitted to hospital because of cerebro-vascular disease IC95% 0.12-0.19), 58 deaths due to AIDS (IC95% 0.09-0.15), 51% due to bronchial and lung neoplasias (IC95% 0.18-0.30), 49 deaths from acute myocardial infarction (IC95% 0.12-0.21), 39% as a result of chronic obstructive pulmonary disease (IC95% 0.07-0.14).
Conclusions: The abundant information supplied by hospital mortality statistics is confirmed, especially if the age and illness of the patients involved is taken into account. There is an evident need for health policies to prevent HIV/AIDS, cerebro-vascular disease, lung cancer and ischaemic heart disease to be adopted in the Algeciras catchment area. Thus, there is a need to monitor and draw up new indicators of hospital mortality and to expand the list of unnecessarily premature and avoidable deaths with strictly hospital causes.