Treatment of Severe Malnutrition in Children: Experience in Implementing the World Health Organization Guidelines in Turbo, Colombia

Abstract

Objective: 

To evaluate the implementation of the World Health Organization guidelines for the treatment of children with severe acute malnutrition in a class I hospital.

Paitents and Methods: 

Descriptive and prospective study of 335 children under the age of 6, admitted between 2001 and 2005 for severe acute malnutrition (83%) and moderate acute malnutrition associated with illness (17%). The care of the children was provided by clinicians and medical staff trained under World Health Organization guidelines.

Results: 

Kwashiorkor was a common result in children with severe acute malnutrition (60.8%); 58% were younger than 1 year old. Complications upon admission were diarrhea (68.4%) and anemia (51.1%), and the most common complication during hospital stay was sepsis (9%). Overall, 61.7% attained −1 standard deviation of weight for height after an average stay of 3 weeks in the hospital; 5.1% were sent to tertiary care hospitals. Overall, mortality during the 5 years was 5.7%, with sepsis the most common cause. In the first year, the mortality rate was 8.7%; it decreased to 4.0% in the last year. The mortality rate was significantly higher in children with edema. Children with moderate acute malnutrition had similar complications and mortality when compared with children with severe acute malnutrition (5.3%).

Conclusions: 

Once the World Health Organization guidelines were implemented, low mortality rates were achieved in children with severe acute malnutrition in class I hospitals.

Malnutrition is associated with half of all of the deaths in children younger than 5 years of age in the world ^(1,2). In the last 50 years, significant progress has been made to improve the recovery of children with this condition. In the 1990s, interventions and development efforts decreased the undernourished population from 177 million to 147 million ⁽³⁾. Despite this reduction, malnutrition continues to be a severe problem in the Americas, and has increased in some areas ⁽⁴⁾.

Schofield and Ashworth ⁽⁵⁾ demonstrated that mortality in children with severe malnutrition remained unchanged in the last 50 years (20% and 30%), with even higher rates (50% and 60%) in children with malnutrition edema (kwashiorkor), even taking into consideration the new knowledge about therapy for malnutrition. The authors hypothesized that these high mortality rates could be associated with misuse of diets high in protein, energy, and sodium and with low micronutrients; inadequate rehydration; improper antibiotic use; and administration of diuretics in children.

In 1999 the World Health Organization (WHO) published a set of guidelines for the treatment of severe malnutrition in children ⁽¹⁾. With the implementation of this protocol, mortality was reduced from 30% to 5% in nutritional centers around the world ⁽⁶⁾. Proper care for hospitalized malnourished children is a priority issue across the globe ⁽⁷⁾.

In 2000 the malnutrition mortality rate in the Antioquia Department was a reported 20.33/100,000 children younger than 5 years old. In Turbo, a town in Antioquia, the rate was nearly 3 times higher: 59.56 ⁽⁸⁾. Malnourished children in Turbo are treated at the Universidad de Antioquia and the Francisco Valderrama Hospital, using the Integrated Management of Childhood Illness (IMCI) strategy. Nutritional disorders are treated complying with WHO guidelines for severe acute malnutrition therapy ⁽¹⁾. Results of the first 53 children treated with WHO guidelines were published previously ⁽⁹⁾, showing a mortality rate of 7.5%. Our study objective is to evaluate the response of 355 children with severe acute malnutrition and moderate acute malnutrition to treatment under the WHO guidelines in a class I hospital ⁽¹⁾.

PATIENTS AND METHODS

This is a prospective and descriptive study consisting of consecutive children under the age of 6 with acute moderate or severe malnutrition treated between 2001 and 2005.

The coordinating team adopted the WHO guidelines and trained personnel to comply with them. Children were treated by staff physicians, pediatric residents, nutritional practitioners, and nursing assistants directed by a chief medical doctor. The turnover of nursing assistants was reduced to a minimum. The family and caregivers were encouraged and motivated, and information was registered in WHO standard forms.

Diagnostics and classification of malnutrition were based on WHO criteria ⁽¹⁾. Marasmus: weight for height (W/H) less than −3 median standard deviations (SDs) from the National Center for Health Statistics/WHO reference values ⁽¹⁰⁾. Kwashiorkor: symmetrical edema affecting at least the lower extremities. Mixed: simultaneous marasmus and kwashiorkor criteria present. Moderate acute malnutrition was diagnosed in children with weight/height between −2 and −3 SDs. Exclusion criteria included secondary malnutrition associated with pathologies and children with edema due to renal, cardiac, or endocrine causes.

The anthropometrical evaluation was performed by nursing assistants following standardized methodology. Length was measured using an infant meter with resolution of 1 mm. Weight was measured on a 10-g Health-o-Meter mechanical weighing scale (Precision Weighing Balances, Bradford, MA). Children were weighed daily at the same time of day, before their 8 AM breakfast. The change in weight was calculated daily (in grams per kilogram per day).

Children diagnosed with severe malnutrition in the outpatient clinic were admitted for immediate care. Children with moderate acute malnutrition and associated complications requiring hospitalization received the same protocol therapy.

Treatment was divided into 2 phases. In the initial phase, the goals were to treat and/or prevent hypoglycemia, hypothermia, and dehydration; initiate feeding and administration of oligoelements and vitamins; and treat infections and other complications such as severe anemia or heart failure. In cases of slight or moderate dehydration, oral rehydration solution for children with severe malnutrition (ReSoMaL) or the standard WHO oral rehydration salts solution were given.

All of the children were given antibiotics following the WHO guidelines; however, specific antibiotic therapy was administered depending on the diagnosis, severity, and clinical course of the infection. Children older than 2 years of age were treated with parasite control according to the protocol.

Upon admission, feeding was started immediately with the administration of the F-75 (75 kcal/100 mL and 0.9 g protein/100 mL) formula, given consecutively every 2 hours, including day and night, according to the frequency indicated in the guide ⁽¹⁾. The F-75 formula replaces cereal with cooked green plantain (Musa species) commonly consumed and produced in this region. F-75 and F-100 were prepared with mineral solution recommended by the guidelines. In addition to the formula, all of the children received vitamin A, folic acid, and multivitamins without iron according to the guidelines. Anemia was diagnosed when hemoglobin was below 9.4 g/dL and severe anemia below 4.1 g/dL.

The rehabilitation phase started when the childrens' appetites improved and infections were controlled. F-75 was replaced with an equivalent quantity of F-100 plus plantain and reduced sugar concentration to decrease the osmolarity of the solution. A modified F-100 formula with less kidney solute load was used in children under 6 months old (Appendix 1).

Soup and solid food were administered between the F-100 intakes. Energy consumption was measured by adding the energy units of the formula and the energy units from complementary daily food ingested in a subsample of 153 children with 15 days of follow-up. In this phase, iron supplement administration began, vitamins and oligoelements continued to be given, and psychoaffective stimulation was provided. The stated goal for discharge was a weight/height of −1 SD ⁽¹⁰⁾, and initial resolution of edema.

Ethical Aspects

This study was approved by the ethics committee of the University of Antioquia Medical Research Center. Informed consent of each child's parents or guardians was obtained before enrollment.

Statistical Analysis

The statistical analysis was done using SPSS version 10 software (SPSS, Chicago, IL). Differences across groups in parametric variables were analyzed using analysis of variance, Scheffé test, and Pearson product moment correlation. For the nonparametric variables, the Kruskal-Wallis test, Mann-Whitney U test, and the Spearman rank correlation test were used. The qualitative variable analysis used chi-square and Fisher exact test. The energy consumption and weight increase curves were compared with the multivariate analysis of variance for repeated measurements. The level of significance for all of the tests was P < 0.05. All of the data are reported as means ± SD, unless otherwise specified.

RESULTS

Overall, 335 children were surveyed, 278 with severe acute malnutrition (83%) and 57 with moderate malnutrition (17%). The most frequent type of malnutrition was kwashiorkor (60.8%), followed by marasmus (30.9%) and mixed malnutrition (8.3%). There were no sex-based differences in frequency. Average age was 14.8 months (SD, 11.5), and 88.4% of the children were younger than age 2 years. The children with kwashiorkor were significantly younger than the children with marasmus, mixed malnutrition, and moderate acute malnutrition (P = 0.001) (Tables 1 and 2).

TABLE 1: Sex distribution according to type of malnutrition

TABLE 2: Age distribution according to type of malnutrition

The majority of children presented chronic malnutrition with height for age values between −2 and −3 SD. There were no statistically significant differences between groups of malnutrition diagnosis. The majority of the children presented global malnutrition (W/A < −3 SD) except for those with kwashiorkor. This is likely because edema alters weight (Table 3).

TABLE 3: z Scores of weight/age (W/A), weight/height (W/H), and height/age (H/A) according to type of malnutrition

The most common sign of infection at admission was fever (26.3%), followed by hypothermia (2.4%) and hypoglycemia (3.6%). The most common associated illness at admission was diarrhea (68.4%) with a significantly higher frequency in children with mixed and moderate malnutrition (P = 0.003). Among the children with diarrhea, 31.5% were dehydrated and 10.5% presented dysentery, with a higher prevalence in children with marasmus or moderate malnutrition (P = 0.001) (Table 4).

TABLE 4: Frequency of diarrhea, dehydration, and dysentery at admission according to type of malnutrition

Other complications at admission were skin infections, which were more common in children with edema (Table 5). Among the 335 children, 99.4% received antibiotics. The initial antibiotic treatment had to be changed in 29.3% of the cases. Hemoglobin was measured in 296 children; 56.1% of them had slight anemia and 3% had severe anemia. Children with edema had a lower average hemoglobin concentration than the other groups (marasmus, 9.4 g/dL; moderate malnutrition, 9.1 g/dL; kwashiorkor, 8.4 g/dL; and mixed, 8.1 g/dL; P = 0.001).

TABLE 5: Other complications at admission according to type of malnutrition

The initial phase lasted an average of 5.1 days (SD, 2.8) without significant differences between different types of malnutrition (P = 0.124). On average, the children's consumption of F-75 formula was 103 mL/kg on the first day of therapy with no significant difference between the groups (P = 0.105).

The rehabilitation phase, on average, lasted 14.2 days (SD, 8.3). This phase was significantly longer in the marasmus and mixed groups (18.8 ± 8.8 SD and 19.9 ± 7.1 SD days, respectively) compared with kwashiorkor (10.9 ± 7.0 SD days) and moderate groups (13.8 ± 6.9 SD days) (P = 0.001).

The average F-100 consumption on the first day of rehabilitation was 130.9 mL/kg without significant statistical difference between groups (P = 0.496). At discharge, the consumption of F-100 had increased to 176.5 mL/kg without significant difference between groups (P = 0.680). In the rehabilitation phase, energy consumption including complementary feeding reached >200 kcal · kg⁻¹ · day⁻¹. Iron supplementation was started on average on day 8 of hospitalization. The majority of children received folic acid (99.7%) and vitamin A (97.9%) supplements.

Sepsis was the most frequent hospital complication, affecting 30 children (9.0%). Other complications were pneumonia in 26 children (7.7%), diarrhea in 21 (6.3%) of which 8 (2.4%) became dehydrated, and heart failure in 15 (4.5%) children. None of these complications showed statistical differences between types of malnutrition. However, 10 of the 15 children with heart insufficiency had kwashiorkor.

Relative to weight gain, 200 children (59.7%) reached the goal of −1 SD for weight/height. The percentage of children with edema reaching this goal was significantly higher than those with marasmus and moderate malnutrition (P = 0.001) (Table 6). Data were missing for 10 children.

TABLE 6: Outcome variables according to type of malnutrition

Seventeen children (5.1%) were transferred to secondary- and tertiary-care centers; no difference in referrals was found between the types of malnutrition (Table 6). Leading causes for the referral were sepsis (9 cases) and severe anemia (4 cases).

From the study population, 19 children (5.7%) died. During the 4-year study, the mortality rate decreased from 7.8% in the first year to 4.0% in the last year. Thirteen died in the hospital and 6 died in the institution to which they were sent. A higher number of deaths occurred in the kwashiorkor group (14 children), although the difference was not significant when compared with the other groups (P = 0.138) (Table 6). When considering the 16 children with severe acute malnutrition who died, 15 of them had edematous malnutrition (kwashiorkor or mixed) and 1 had marasmus. In this case, the difference was statistically significant (P = 0.027).

The most common cause of death in the study population was sepsis ⁽¹¹⁾, followed by heart failure ⁽⁴⁾. Of the 13 deaths in the hospital, 7 occurred during the day and 6 during the night. Of the total deaths in the study, 4 occurred on the first day of hospitalization, 7 between the second and third day, and 8 after the fourth day. Of the total children studied, 14 (4.2%) were readmitted with a malnutrition diagnosis during the study.

DISCUSSION

The results of the implementation of the WHO guidelines for severe acute malnutrition therapy in children appear to be promising, even though high indices of mortality are recorded in other areas around the world ⁽¹²⁾. Mortality in this study was relatively low (5.7%), with a significant decrease as the personnel gained experience and compliance with the guidelines. This is demonstrated by the decrease in the mortality rate from 7.8% in the first year to 4.0% in the final year of the study. Previously, malnutrition was not considered a cause of death; therefore, it was not possible to compare the mortality rate from this study with previous rates in this hospital.

A study in Bangladesh reported a 17% mortality rate in children before the implementation of the WHO guidelines, and a 9% mortality rate after they were officially introduced ⁽¹¹⁾. A similar trend in South Africa revealed a reduction from 30% to 20% in the initial mortality rate, with a further reduction to 6% when WHO guidelines became more rigorously introduced ⁽¹³⁾.

Comparison of mortality rates between different studies is difficult because of differences in applying guidelines and other factors such as child selection, associated complications, and diseases. However, these 3 studies have demonstrated that strict application of WHO guidelines will reduce mortality rates ⁽¹⁴⁾.

Factors contributing to low mortality in this study include early diagnostics, prompt start of therapy, continuity and training of personnel, availability of vitamins and oligoelements, adaptation of formula to the regional nutritional resources, and presence of family.

The study also included children with acute moderate malnutrition and associated life-threatening complications (ie, persistent diarrhea, dysentery, and severe anemia). These children were at the same risk for complications as children with acute severe malnutrition, suggesting that they also needed specialized treatment.

The majority of the children were younger than 2 years old, and 19.7% were younger than 6 months. It is striking to observe that the presentation of kwashiorkor occurs before marasmus, even in children younger than 6 months of age. This presentation could be associated with other causes, such as the early weaning from breast-feeding, the inappropriate introduction of complementary nutrition, and a deficit of micronutrients.

In addition to acute malnutrition (starvation), children also presented with chronic moderate malnutrition (stunting). In the cases above, the mothers did not identify malnutrition as a health problem because they usually only seek attention for diarrhea, skin infections, respiratory infections, and edema.

The most common complication in this pediatric population was diarrhea. More than half of the children suffered diarrhea at admission and others developed diarrhea during therapy. Khanum et al ⁽¹³⁾ found that 60% of acutely malnourished children have diarrhea, supporting the association between diarrhea and malnutrition.

Children with dehydration were treated with ReSoMaL. When ReSoMaL was not available, the standard WHO oral rehydration salts solution was used. For dehydrated children, the guidelines recommend ReSoMaL because acutely malnourished children show an increase in the corporal total sodium and a potassium deficit. The WHO guidelines caution against high sodium solutions because they can trigger heart failure ⁽¹⁵⁾. However, Bhan et al ⁽⁶⁾ argue there is insufficient evidence to demonstrate that using ReSoMaL is advantageous over the standard solution. This suggests that more controlled clinical studies are needed. A later study by Alam et al ⁽¹⁶⁾ comparing ReSoMaL with oral rehydration salts solution containing 90 mEq/L sodium demonstrated that children receiving ReSoMaL improved significantly from hypokalemia while their serum sodium decreased. It is therefore necessary to conduct studies to determine the most adequate composition for the oral hydration solution for children with severe malnutrition. A possible alternative is to adjust the new WHO solution formula consisting of 75 mEq/L sodium concentration and 20 mEq/L potassium concentration to 40 mEq/L potassium ⁽¹⁷⁾.

It is important to note that in the majority of cases, diarrhea improved after the start of F-75, which is low in lactose and has low osmolarity. The early administration of micronutrients had a potential positive effect on the structural and functional integrity of the gastrointestinal tract. Carbohydrate intolerance was only tested using sugar reduction in a few cases of severe and persistent diarrhea. Lactose-free formulas were used temporarily in <5% of the children with diarrhea. In some patients, diarrhea began or was aggravated when F-100 was administered. This formula has a relatively high osmolarity, and it was necessary to decrease the amount of sugar, substituting plaintain.

Antibiotics are routinely administered to acutely malnourished children because these children may not present signs or symptoms of infection. Bhan et al ⁽⁶⁾ suggest that there is insufficient evidence to support this recommendation. In the hospital, 99.7% of the children received antibiotics, but only 26.4% presented with fever. Unfortunately, considering the severe condition of these children, it is not possible to demonstrate adequate recovery without the use of antibiotics.

The increasing severity of the biochemical imbalance in malnourished children is enhanced by the deficit of vitamins and oligoelements. Therefore, high vitamin A, folic acid, other vitamins, and oligoelement supplements, dosed at the start of therapy, are fundamental in obtaining a successful outcome ⁽¹⁸⁾.

On average, the children's appetite improved after hospital day 5, when they attained the minimal necessary metabolic and physiological requirements. Following this, children began the rehabilitation phase, including iron supplements according to the guidelines. Previous studies by Manary et al ⁽¹⁹⁾ suggest that iron supplements should not be started until children are metabolically and physiologically stable. Additionally, edematous malnourished children present with ‘free’ iron in serum, low transferrin, and high ferritin concentrations, which makes them prone to oxidative stress, major edema, and infection. Taking this into consideration, the iron supplements were administered around hospital day 8, when the children recovered their appetite and infections were controlled ^(20,21).

Adaptation of the F-75 and F-100 formulas to the economic and sociocultural conditions of the region was crucial. It eased the preparation and acceptance by children and their mothers because it was closer to their cultural taste. Recent research points out the inconvenience of high protein intakes during nutritional recovery ⁽²²⁾. The F-100 is a high-protein formula for children younger than 6 months old (2.9 g/100 mL). We therefore modified the F-100 formula to contain less protein (1.9 g/mL) for children younger than 6 months to decrease the kidney solutes load (Appendix 1). This amount of protein proved to be sufficient to promote the growth and recovery at the same rate in children older than 6 months of age. For the duration of the study, the children's caregivers developed the ability to diagnose heart failure at earlier stages, when it is often misinterpreted as pneumonia.

The average hospital stay was 3 weeks, consistent with the guidelines of 2 to 6 weeks. The ideal discharge should match the goal of −1 SD of weight/height, but this was not always possible; 125 children were dismissed before attaining −1 SD. This occurs due to cultural differences in the treatment of disease or socioeconomical needs of the children's families. Some children were dismissed after a prolonged hospitalization, attaining between −1 SD and −2 SD for weight/height.

Acute severe malnutrition has an overall low prevalence in Colombia, but it has a high prevalence in some areas. Acute severe malnutrition is not recognized as a disease by the Colombian health system, and hospitalization is not justified by the system. Therefore, outside this study, children are hospitalized for the treatment of diarrhea, and respiratory and skin infections, but are rarely admitted for the diagnosis of malnutrition. By implementing WHO guidelines, the study aim was to obtain financial support for hospitalization expenses for malnutrition (eg, vitamins and oligoelements not covered by the national health system). Following its first year, the study obtained continued financial sponsorship from Hospital de Turbo, University of Antioquia, and the Department of Antioquia. This work became a pilot study for the treatment of children with acute malnutrition in the Department of Antioquia, and similar projects adopting the WHO guidelines for nutritional recovery centers. Along with the application of the guidelines, an ambulatory program was established that includes basic health care, nutritional education, home visits, complementary alimentation, and micronutrients. This stage of the study is under evaluation.

The daily hospitalization cost per child was approximately US $20 with an average stay of 19 days for a total of US $380 per child for the entire recovery time. According to Schofield and Ashworth ⁽⁵⁾, the mortality rate in children with acute severe malnutrition is between 30% and 50%. Therefore, assuming a middle value of 40% as the mortality rate without the application of the WHO guidelines, 134 of 335 children would have died during the length of the study. Nevertheless, the number of deaths in our study was 19, meaning that by applying the WHO guidelines, 115 children were saved. The cost per child saved was calculated as the number of children that were saved in our study (316) times the amount of the average treatment (US $380), divided by the difference in children that were saved with and without the guidelines (115). Therefore, the cost per child saved in our study was US $1050.

CONCLUSIONS

It is crucial during the stabilization phase that children are hospitalized because severe acute malnutrition often comes with complications. After the stabilization phase, in some cases it may be possible to manage the rehabilitation phase as an outpatient with home visits and nutritional and micronutrient supplementation.

Our results indicate that strict and careful implementation of the WHO guidelines for acutely malnourished children may reduce mortality to <5%. This study also highlights the importance of treating children with moderate acute malnutrition because the risk of complications and death is similar to that of children diagnosed with severe acute malnutrition. These results were obtained in a level I hospital without laboratory support, with limited technology and specialized care, by providing training of health care personnel and by strictly following the WHO guidelines.

Acknowledgments

The authors thank the University of Antioquia Schools of Medicine, Nursing, Nutrition and Dietetics; the Programa de Regionalización, Hospital Francisco Valderrama; the government of Antioquia; the doctors, the nursing, nutrition, psychology, and bacteriology students, the nursing assistants, the children's caregivers, and the children and their mothers involved in the study; and Megan Baerny, Carlos Mantilla, and Miguel Bernal from the Mayo Clinic, and Christopher Drasbek of the Pan American Health Organization for reviewing and editing the manuscript.