Use of simple clinical signs to predict pneumonia in young Gambian children: the influence of malnutrition.

Introduction Acute respiratory infections (ARI) are estimated to be responsible for about 4 million childhood deaths per year, most of them in developing countries[1]. Malnutrition is common in most developing countries, and it has been shown to increase both the frequency and severity of ARI episodes [2]. Indeed, malnutrition is an important determinant of mortality due to ARI[3]. The current WHO strategy for the control of mortality due to ARI relies heavily on standardized case management using simple clinical signs which have been shown to predict pneumonia.(a) Several studies have shown that fast breathing and lower chest wall indrawing are the best predictors of pneumonia in children with a cough or breathing difficulty [4-7]. However, these are signs which indicate the greater workload of breathing for a young child with pneumonia and it is possible that malnourished children may not have the strength to manifest some of these physical signs in the same manner as well nourished children. The study described below was designed to evaluate the power of various widely used clinical signs as predictors of pneumonia in children presenting for the first time with malnutrition and respiratory symptoms, and to compare this with the predictive power of the same signs in well nourished children presenting with cough or breathing difficulty. Patients and methods The study was carried out in the outpatient department of the Medical Research Council (MRC) Hospital at Fajara in the Gambia. This clinic serves a large, predominantly urban population located about 15 km from Banjul, the capital. During the period from November 1990 to March 1992, all children aged 3 months to 5 years who arrived at the clinic for whatever reason were screened for malnutrition by being weighed and examined for oedema. All the children whose weight was less than 70% of the United States National Center for Health Statistics (NCHS) mean or who had oedema were referred to the study physician (AGF) for possible inclusion in the study. From April to November 1992 all children who presented at the same clinic with clinically evident malnutrition were referred directly to the study physician by the nurse in charge of the outpatient department. Children with oedema underwent urinalysis and those with the nephrotic syndrome or acute nephritis were excluded. Children who had already undergone treatment for malnutrition were also excluded. All the children underwent a full history and examination and had a chest radiograph taken. Only children whose parents stated, in response to questioning, that their children had a cough, fast breathing, breathing difficulty or chest pain were included in the analysis. During the same period, children in the same age group who presented at the MRC Hospital with a cough or breathing difficulty and who were not malnourished, were recruited into a parallel study. On a specified day, usually once every two weeks, the first seven eligible children presenting were enrolled. They were assessed clinically using the same format as that used for the malnourished children. All the children with any auscultatory findings or fast breathing (defined as a respiratory rate of [greater than or equal to]50 per minute in children under 12 months and [greater than or equal to] in children aged 12 months to 5 years), and 10% of the children without these findings had a chest radiograph taken. Respiratory rate was measured by observation of the child's chest for exactly one minute with the child either asleep, awake and quiet, or while breastfeeding. If this was interrupted by coughing the measurement was started again. Chest wall indrawing was classified as suprasternal (indrawing of the soft tissues above the clavicles and sternum with inspiration), intercostal (indrawing of the intercostal tissues with inspiration), and lower chest wall (indrawing of the bony structures of the lower chest wall with inspiration). …