Death due to use of traditional medicines in Africa: a preventable cause of neonatal and infant mortality.

Abstract

To the Editors: In South Africa, 60% to 80% of the population use traditional medicines (muti), some of which are toxic. The only comprehensive studies of toxicity in in-patients in Southern Africa are those from Harare (1) and Medunsa (SA) (2). We conducted a study on patients admitted to hospitals in Johannesburg over one year, with ethical clearance from our University. Seventy seven patients were included, all with a definite history of muti administration. The identity of the muti was unknown to patients or relatives. Sixty six percent were less than one year old (median 8 m). The female: male ratio was 1:1.4. HIV, hepatitis B and C status were determined, with informed consent, when necessary. The route of administration was known in 52 cases: 54% oral, 23% rectal, and 23% both routes. The major findings on admission are shown in Table 1. The clinical features attributable to herbal intoxication may be superimposed on the symptoms of the pre-existing disease. Fifty three percent had gastro-intestinal problems, with dehydration in 59%. Fifty three percent had symptomatic liver disease. In the 94% with renal dysfunction as determined by creatinine concentration, the urea was raised in only 41%. This may have been due to the malnourished state of many patients, of whom 22% were clinically malnourished and 5 patients were suffering from kwashiorkor. Forty eight percent of patients were HIV positive. The overall in-hospital mortality was 33%, with the malnourished patients showing a similar percentage to the remainder.TABLE 1: Details of pathologies observed in 77 patients with a history of ingested traditional remedies *Neo- and perinatal mortality is high in most of Africa with reported figures of up to 13% in Nigeria. Preventable factors, mainly infections, account for most of the mortality. In South Africa the overall figure for infant mortality in 2000 was 4%. HIV makes an increasingly important contribution, although in a study in Malawi, with a perinatal death rate of 13.6%, HIV was found not to be the main determinant of infant mortality. The contribution of traditional medicine has not been addressed in most studies of infant mortality. In Nigeria mortality was 9.1% among patients who had been given traditional medicine compared with 6.1% for the non-users (3) and in Niger, traditional practitioner consultation was a significant factor in infant mortality (4). In South Africa, Moore (5) reported 50 cases with a mortality of 28% (43% for herbal and 21% for chemical enemas). Children are especially vulnerable to muti intoxication, possibly due to the difficulties in adjusting doses. The mortality rate of one third in our patients was much higher than the overall in-hospital mortality for this age group, which is less than 5%. How do we reduce these deaths in Africa? It is over 20 years since Joubert (6) called for the instigation of epidemiological studies of muti toxicity in developing countries. Action is usually a response to a clear cause and the manner in which infant mortality statistics are reported can have a major effect on the perceived epidemiology. We believe that the possible contribution of traditional medicines to deaths that follow their administration should be notified in mortality statistics that are collected in African countries. *Vanessa Steenkamp *Michael J. Stewart †Michele Zuckerman