Secondary Logo

Journal Logo

Selected Summaries

Management of Severe Malnutrition and Diarrhea

Tolboom, Jules J. M.

Author Information
Journal of Pediatric Gastroenterology and Nutrition: March 2000 - Volume 30 - Issue 3 - p 346-348
  • Free

Mortality in severely malnourished children with diarrhoea and use of a standardised management protocol. Ahmed T et al. Lancet 1999;353:1919–22.


Severely malnourished children have high mortality rates. Death commonly occurs during the first 48 hours after hospital admission and has been attributed to faulty case management. A standardized protocol for acute-phase treatment of children with severe malnutrition and diarrhea was developed to reduce mortality.

At the International Centre for Diarrhoeal Diseases Research (ICDDR, B Hospital Dhaka, Bangladesh) the outcomes in two treatment groups of severely malnourished children with diarrhea were compared. The standardized-protocol group (n = 334) included children admitted from January 1 through June 30, 1997, and treated with slow rehydration with emphasis on oral rehydration. The nonprotocol group (n = 293) included children admitted from January 1 through June 30, 1996, and managed by conventional treatment, before implementation of the protocol.

In the standardized protocol, rehydration was mainly managed by using rice-based oral rehydration solution (ORS with 90 mmol/l Na+, 20 mmol/l K+, 80 mmol/l Cl, 10 mmol/l citrate, and 50 g/l rice powder), except in infants less than 4 months of age, who received standard ORS (with 90 mmol/l Na+, 20 mmol/l K+, 80 mmol/l Cl, 10 mmol/l citrate, and 111 mmol/l glucose). The ORS was administered orally or by nasogastric tube if a child was unable to drink. Only in severe dehydration was initial rehydration achieved by intravenous fluids for 2 hours and thereafter by use of ORS, according to World Health Organization (WHO) guidelines (Management of the patient with diarrhoea. Geneva: WHO, 1992).

Feeding was begun immediately after admission, and food was given every 2 hours, using a feed containing (per liter) 40 g whole milk powder, 40 g rice powder, 25 g sugar, 25 g soy oil, 0.5 g magnesium chloride, 1 g potassium chloride, and 2 g calcium lactate (“milk suji”). Children with marasmus and marasmic kwashiorkor were given 10 ml/kg per feeding (80 kcal/kg per day) on day 1 and 12 ml/kg per feeding (96 kcal/kg per day) on days 2 and 3. If no diarrhea occurred, feeding was changed from day 4 onward to 12 ml/kg of a concentrated feed (“milk suji 100”), containing (per liter) 80 g whole milk powder, 50 g rice powder, 50 g sugar, 25 g soy oil, 0.5 g magnesium chloride, and 1 g potassium chloride (providing 144 kcal/kg per day). Children with kwashiorkor were given milk suji, 9 ml/kg per feeding (72 kcal/kg per day), on days 1 to 3. If no diarrhea occurred, from day 4 onward they were given a special milk (9 ml/kg per feeding), containing (per liter) 100 g whole milk powder, 70 g sugar, 30 g soy oil, 25 g egg albumin, 0.5 g magnesium chloride, and 1 g potassium chloride (providing 108 kcal/kg per day). Breast-feeding was continued and non–breast-fed infants younger than 4 months received a noncommercial cow's milk–based infant formula.

In children with signs or symptoms of infection other than diarrhea, intramuscular or intravenous antibiotic therapy with ampicillin (100 mg/kg per day with doses every 6 hours) and gentamycin (5 mg/kg per day in intramuscular doses every 12 hours) was started. If evidence of septicemia was absent, this was changed to amoxicillin (100 mg/kg per day with doses every 8 hours), orally for 3 days. In all other cases, antibiotic therapy was targeted at the suspected or detected infection.

Vitamin A was given according to WHO recommendations (Strategies for the prevention of blindness in national programmes: A primary health care approach. Geneva: WHO, 1997). Further, micronutrient supplementation included elemental zinc (2 mg/kg per day), 1.25 mg folic acid, and 1 ml of a multivitamin supplement (5000 IU vitamin A palmitate, 1000 IU vitamin D, 1.6 mg thiamine hydrochloride, 1 mg riboflavin, 10 mg nicotinamide, 5 mg calcium D-pantothenate, 1 mg pyridoxine hydrochloride, and 50 mg ascorbic acid), administered for 15 days. Intramuscular magnesium sulphate (0.4 mmol/kg) was administered daily for 7 days.

Hypoglycemia (blood glucose <3 mmol/l) was managed with 50 ml 10% glucose, given orally or by nasogastric tube. If a child was unconscious or having convulsions, 2 ml/kg 25% glucose was given intravenously, followed by oral glucose. Severe acidosis (serum CO2 <6 mmol/l) was treated by slow intravenous infusion with sodium bicarbonate (1 ml/kg). For hypokalemia, oral potassium was administered (5 mmol/kg, daily). In cases of parenteral rehydration and serum potassium less than 2 mmol/l, potassium was added to the intravenous fluids (up to 40 mmol/l).

If packed-cell volume was less than 15%, packed cells or whole blood was transfused (10 ml/kg). Hypothermia, abdominal distension, congestive heart failure, and weeping skin lesions were managed according to WHO guidelines (Management of severe malnutrition: A manual for physicians and other senior health workers. Geneva: WHO, 1999).

In the nonprotocol group children were rehydrated within 4 hours, according to WHO guidelines (Management of the patient with diarrhoea. Geneva: WHO, 1992). Severe dehydration was treated with intravenous fluids without dextrose. Antibiotics were given only if indicated. Feeding was delayed until rehydration was complete; only the very sick were fed nasogastrically. Liquid feedings were offered every 2 hours, but intake was not ensured. Older infants were also offered rice porridge two to three times daily. Total daily dietary energy intake was 100 kcal/kg, 80 kcal/kg, and 60 to 70 kcal/kg for children with marasmus, marasmic kwashiorkor, and kwashiorkor, respectively. Regarding micronutrients, only vitamin A was routinely offered; magnesium injections were not given. For hypoglycemia, intravenous glucose was not followed by oral glucose drink. Children with hypokalemia received intravenous fluids with potassium added. In cases of severe abdominal distension, feeding was discontinued and intravenous fluid rationed. If packed-cell volume was less than 20%, whole blood (15–20 ml/kg) was transfused, with or without furosemide. In case of heart failure, digoxin was provided in standard doses.

Characteristics on admission were similar for the two groups, except that more children on standardized protocol had edema, acidosis, and Vibrio cholerae isolated from stools (in 26%). Successful rehydration with oral rehydration solution was accomplished in 199 (59.9%) children in the standardized protocol group, compared with 85 (29%) in the nonprotocol group (P < 0.0001). Use of expensive antibiotics was less frequent in children on standardized protocol than in the nonprotocol group (P < 0.0001). Children treated by standardized protocol had fewer episodes of hypoglycemia than nonprotocol children (15 vs. 30;P = 0.005). Mortality in the standardized protocol group was 9% (30 children) which was significantly lower (odds ratio 0.49, 95% confidence interval [CI] 0.3–0.8;P = 0.003) than the 17% mortality (49 children) in the nonprotocol group.

Compared with nonprotocol management, the standardized protocol resulted in fewer episodes of hypoglycemia, less need for intravenous fluids and a 47% reduction in mortality. This standardized protocol should be considered in all children with severe malnutrition and diarrhea.


Severe protein energy malnutrition (PEM) accounts for approximately 10% of pediatric hospital admissions in developing countries (J Trop Pediatr Environ Child Health 1979;25:42–5;Lancet 1995;345:453;Bull World Health Org 1996;74:223–9). Many children have diarrhea and evidence of infection on admission (Nutr Res 1981;1:33–46;Trop Geogr Med 1986;38:351–8). Even in large and relatively well-equipped hospitals, severe PEM carries unacceptably high case fatality rates of 20% to 30% (Trop Geogr Med 1986;38:351–8;J Trop Pediatr 1994;40:225–30;Lancet 1995;345:453). Generally, in edematous malnutrition (marasmic kwashiorkor and kwashiorkor) mortality is higher than in marasmus (Trop Geogr Med 1986;38:351–8;BMJ 1993;307:710–3;Bull World Health Org 1996;74:223–9). Most deaths occur in the first days of admission, due to hypoglycemia, hypothermia, dehydration and electrolyte imbalance, anemia, heart failure, and generalized infection (BMJ 1993;307:710–3;Trop Geogr Med 1993;45:290–3). Reduction of hospital mortality is considered to largely depend on careful management of complications during the critical first days. This has been pointed out in manuals for the treatment of severe PEM, published by different agencies during the past 25 years (The therapy of the severely malnourished child. Kampala, Uganda: National Food and Nutrition Council, 1973;The treatment and management of severe protein-energy malnutrition. Geneva: WHO, 1981;Nutrition Guidelines. Paris: Médicins Sans Frontières, 1995).

Since 1980, the Children's Nutrition Unit of Save the Children Fund (SCF, also located in Dhaka, Bangladesh) has been achieving low case fatality rates between 4% and 7%, even with domiciliary treatment after 1 week of hospital care or day care (Lancet 1994;344:1728–32). The SCF management protocols included routine prescribing of antibiotics on admission, blood transfusion for severe anemia, withholding iron in the first week, avoidance of intravenous rehydration, and cautious refeeding with low-sodium diets. In 1999, the WHO manual, Management of severe malnutrition: A manual for physicians and other senior health workers, replaced the 1981 edition. Again, special emphasis is on initial treatment, constant monitoring of the child, and the prevention of complications. For treatment of dehydration, use of a special ORS formula is advised, containing less sodium (45 mmol/l) and more potassium (40 mmol/l) than standard ORS, with magnesium (3 mmol/l), zinc (0.3 mmol/l), and copper (0.045 mmol/l) added. Because of the risk of overhydration and heart failure, intravenous treatment of dehydration is only advised if there are definite signs of shock. The role of micronutrients in dietary management, especially during the initial phase of treatment, is well recognized. For the initial phase (lasting 2 to 7 days) and the rehabilitation phase thereafter, the use of two noncommercial formula diets, supplying 75 kcal (“F-75”) and 100 kcal/100 ml (“F-100”), respectively, are advised. The formula diets are dried skimmed milk–based, and also contain sugar, cereal flour, and vegetable oil along with mineral and vitamin mixes. Initial feeding is every 2 hours; thereafter, the frequency of feeding is gradually decreased.

Comparing the standardized protocol group with the nonprotocol group at ICDDR in the study by Ahmed et al., distinct differences in approach may explain the better outcome in the standardized protocol group. First, the standardized protocol puts more emphasis on initial routine treatment with broad-spectrum antibiotics, slow rehydration with ORS by oral or nasogastric route, treatment of hypoglycemia and, most important, supervision of 2-hour feedings, which is begun immediately after admission. Further, in the standardized protocol, volume and energy intakes with liquid foods are more gradually increased, according to the type of severe PEM. Lastly, more strict criteria are used for transfusion, and only packed cells are transfused, which may have prevented some cases of heart failure. In addition, the use of digoxin in the nonprotocol group may have caused cardiac arrhythmia in some cases and contributed to deaths. Contrary to WHO guidelines and the practice at the Children's Nutrition Unit (SCF), both in the standardized and the nonprotocol group, ORS with 90 mmol/l Na+ and 20 mmol/l K+ was used, instead of ORS with 45 mm/l Na+ and 40 mmol/l K+. Retrospectively, this approach was probably more appropriate, in view of the relatively high incidence of cholera (26% in the standardized protocol group).

As others agree (Lancet 1999;354:1142), Ahmed et al. clearly demonstrate the importance of a systemic approach in the initial treatment of children who have severe PEM. However, any protocol is effective only if it can be properly instituted and children are closely monitored, especially during the critical first week of admission. To accomplish that in less privileged conditions is the real challenge for all health staff involved in the treatment of children with severe PEM.

© 2000 Lippincott Williams & Wilkins, Inc.