Search

Calmette is a national university hospital with 220 adult beds. It has emergency, surgical, medical and gynecology and obstetrics departments, along with a radiology unit, a laboratory for medical analyses, a central pharmacy and an outpatient clinic. This hospital has an unusual statute, with managerial autonomy and a system of cost recovery that currently provides 64% of the hospital's income. Since 1994, it has benefited from a French cooperation program. The French NGO, Médecins du Monde, has been present at Calmette since 1990, providing support forMedicine B, the indigent sector of the medical department. The aim of the Medical Information System (SIM) is to develop a simple, reliable and reproducible system so that, for every action undertaken at the hospital (hospitalization, day hospital and outpatient clinic) the following pieces of information are recorded: 1) the disease; 2) the type of patient; 3) the type of management; 4) the means used to treat the patient; 5) the cost. Data are collected and analyzed using programs created with EPIINFO software (CDC, WHO), using the EPIGLUE module. In 1998, 10,814 admissions were recorded at Calmette Hospital, 7,811 (72.2%) of which were to the Emergency Department and 3,003 (27.2%) of which were direct admissions to other wards. We analyzed 10,603 (95%) computerized medical summaries (RMI). About 50% of beds were occupied in the maternity and gynecology ward whereas almost 90% of beds were occupied in the surgical and emergency wards. AIDS and tuberculosis were the conditions most frequently treated by the medical department, despite a marked increase in more specialized areas of medicine such as cardiology and diabetology. The surgical department reflected the concentration on emergency services of the hospital, with cranial traumatism the primary reason for admission for the hospital as a whole. The mean age of patients was 27 years for the maternity ward and 49 years for the medicine A ward. The mortality rate was about 5% for the medical wards (mainly due to AIDS) and almost 50% in the emergency department (cerebrovascular neurologic disease, cranial traumatism). The proportion of nonpaying patients was high (about 40% in terms of stays in hospital and about 50% of all days spent in hospital). The training of a Cambodian manager for the SIM is a key priority. The point of the SIM is to use the treated data it produces to improve management and decision-making. The data it produces should be used to define the profile of the patients treated, both from a medical point of view and in terms of their ability to pay. This is a fundamental step towards identifying activities that should receive priority as part of a development strategy for a structure evolving in a highly competitive environment. The SIM data are also invaluable for the short-term management of the hospital through the contribution they make to the development of effective analytical accounting, making it possible to evaluate costs and to adjust charges appropriately. Finally, the involvement of the SIM in the setting up and functioning of the Comité de Lutte Contre les Infections Nosocomiales (CLIN; the Hospital-Acquired Infections Committee) in 1999 to 2000 is not utopia, it is the logical continuation of improvements in the overall quality of care It involves, in particular, the training of nurses and head nurses, initiated by nurses acting as technical assistants in the French cooperation program. The definition of the role of the hygiene nurse and the selection of such a nurse from the trained head nurses are also part of this process.

The obstetrical referral patients admitted to the Sanon Souro hospital in Bobo-Dioulasso (Burkina Faso) during 1991 were retrospectively reviewed. During the study period 741 women were admitted. The median distance between the referring centre and Bobo-Dioulasso was 33 km. Median age was 22 years and median parity was 3. Indications for referral were dominated by obstructed labour, haemorrhages and infections. The proportion of maternal mortality in the study was 5.4% and was not related to maternal age or parity. However, there was a linear trend of increasing mortality according to the distance between the referring centre and Bobo-Dioulasso (X2 for trend: 7.56; p = 0.006). The perinatal mortality rate of 420 p. thousand was not related to maternal age, but a parity exceeding 4 was a risk factor for the occurrence of perinatal death (relative risk: 1.4; 95% C.I.: 1.2-1.7). As for maternal mortality, there was a linear increase of perinatal mortality with the distance from the referring centre (X2: 20.8; p = 0.00001). There was no marked seasonal variation in the number of referrals nor in the mortality rates per month.