Associations of admission- and transfer criteria with clinical outcomes of infants (6 - 23 months) treated for severe acute malnutrition in Ghanaian referral hospitals - the SAMAC Study

Abstract

INTRODUCTION: Worldwide, severe acute malnutrition (SAM) is an important risk factor for morbidity and mortality, contributing to more than half of deaths in children under-five years. Inasmuch as the World Health Organization (WHO) has provided diagnostic criteria to be used for admitting into inpatient care for the management of SAM in children aged 6 - 59 months, these are not based on strong evidence. Furthermore, there are limited studies on the relationships of WHO admission and transfer criteria in relation to clinical outcomes, such as mortality rate, length of stay (LOS) in the hospital and daily weight gain in children aged 6 - 59 months with SAM when categorised into different age groups. This study was therefore undertaken to determine the associations of admission diagnostic and transfer criteria with clinical outcomes of infants aged 6 - 23 months diagnosed with SAM, admitted, treated and subsequently discharged. METHODS: A collection of already existing data was carried out at three referral hospitals in Ghana namely the Komfo Anokye Teaching Hospital, the Princess Marie Louis Children's Hospital and the Tamale Teaching Hospital. The medical records of 399 infants aged 6 - 23 months who were diagnosed with SAM, admitted, treated and subsequently discharged at the three study sites between January 2013 and June 2018 were included in the final analysis. Data on demographic, anthropometric, clinical signs and complications, recovery, LOS in the hospital and death were collected. Anthropometric and clinical characteristics on admission were assessed in relation to mortality, daily weight gain, admission z-scores and improvement in mid-upper arm circumference (MUAC). Data was analysed using SAS version 9.4 and linear regression analysis was used to determine the association between admission characteristics and the clinical outcomes. RESULTS: At admission, 89.1% (n = 229) had a weight-for-age z-score (WAZ) < -3 standard deviation (SD), 81.4% (n = 131) had a MUAC < 115 mm, 83.2% (n = 79) had a weight-for-length z-score (WLZ) < -3 SD, 34.8% (n = 69) had oedema and 6.5% (n = 26) had MUAC < 115 mm complicated by oedema. Among infants with clinical signs, complications or co-morbidities, 51% (n = 208) presented with diarrhoea; 51% (n = 203) presented with vomiting while 8% (n = 33) tested positive for HIV. Overall, 15.8% (n = 63) of the infants died with 65.1% (n = 41) of the deaths occurring within 92 hours of admission. The median LOS in the hospital and weight gain was 11 days (Interquartile range (IQR): 7, 17) and 6.8 g/kg/day (IQR: 0.8, 11.8), respectively. With the exception of admission WLZ, a non-statistically significant difference was observed between admission anthropometric criteria in infants who died compared to those who survived. Infants with a WAZ < -3 SD had 5.05 (95% Cl: 1.80, 14.17; p = 0.002) higher odds of weight gain of > 10 g/kg/day than those with a WAZ ≥ -3 SD. Infants with an admission MUAC < 115 mm had 4.64 (95% Cl: 1.64, 13.12; p = 0.02) higher odds of being hospitalised for more than 7 days when compared to those with a MUAC ≥ 115 mm. Infants with convulsions and shock at admission had 6.76 (95% CI: 2.57, 17.83; p = 0.0001) and 5.18 (95% CI: 1.49, 18.08; p = 0.01) higher odds of mortality than those without convulsion and shock, respectively. The odds of mortality for infants with HIV infection was 4.04 (95% CI: 1.86, 8.81; p = 0.0004) while that for those with dermatitis was 3.46 (95% CI: 1.77, 6.75; p = 0.0003). CONCLUSIONS AND RECOMMENDATIONS: There was a statistically significant difference between admission WLZ for infants who died and those who survived but it was not statistically significantly associated with death in infants aged 6 - 23 months. Although non-statistically significant, it was observed that, infants with MUAC < 115 mm at admission were two times more likely to die compared to those with a MUAC < 115 mm at admission. With the exception of recovery rate (defined as resolution of oedema) and LOS in the hospital, mortality and daily weight gain were observed to be outside the minimum range recommended by Sphere for the inpatient management of SAM. Therefore, it is recommended that the Ministry of Health in collaboration with Ghana Health Service (GHS) should try and improve the diagnostic skills of health workers/clinicians in the area of SAM by incorporating weight, height and MUAC measurements into paediatric clinics as well as the provision of appropriate child growth standard charts. There is also the need to update clinicians in the recognition and management of SAM, especially children with SAM who present with complications and co-morbidities such as convulsions, shock and HIV infection for better treatment outcomes.