Study type, population and duration
This was a prospective hospital-based study among infants less than 6 months of age with SAM, following breastfeeding failure. The study was conducted during the period between January 1st, 2018 through April 19th, 2020 in both domains.
This study was conducted in the Paediatric Department at Alsadaqa Teaching Hospital in Sheikh Othman district, Aden, Yemen. It is a tertiary referral hospital, which serves all referral cases from surrounding provinces of Aden city. The treatment feeding centre (TFC) is located inside the hospital of 21 beds with 4 doctors and 15 nurses. Bed occupancy is 3 patients per month and the mortality rate is < 5% per year.
Sample size, type and data collection
Badi et al has documented the prevalence of malnutrition in Aden as 5.2% among hospitalized children (12). The sample size (n) was calculated using open OpenEpi, Version 3, open-source calculator—SSPropor where the hospitalized population was approximately 12000 children within a confidence interval of 97% and a design effect of 1 and yielded a sample of 97. In this study a total of 108 infants were included.
Inclusion and exclusion criteria
Inclusion criteria were infants less than 6 month of age with SAM due to failure of breastfeeding with SAM who were admitted for feeding in TFCs
. SAM was diagnosed based on the definition of the updated WHO criteria for management of SAM in infants and young children (9). These were weight-for-length less than –3 Z-score, or the presence of bilateral pitting oedema. In addition, all mothers were unable to successful breastfeeding at admission due to failure of lactation owing to stress of arm-conflict.
Infants with known underlying organic diseases as a cause of malnutrition were excluded. Any infant with a known congenital, chronic or metabolic disease as a cause of SAM was excluded. Furthermore, an infant was excluded if he or she had any serious clinical condition or medical complication, recent weight loss, ineffective feeding (attachment, positioning and suckling) directly observed for 15–20 min, any medical or social issue needing more detailed assessment or intensive support (e.g., disability, depression of the caregiver, or other adverse social circumstances) (9).
Infants were chosen randomly using a computer-generated set of random numbers.
A validated questionnaire containing personal and clinical data was filled and labelled with a code number.
All infants included in the study were assessed and evaluated clinically. Data of the main socio-demographic characteristics and clinical attributers were recorded using a data collection sheet. The infants were weighed to the nearest decimal point on daily basis using a beam balance scale. Length was initially measured and then every five days thereafter using locally-made measuring board. Bilateral lower limb pitting edema was confirmed by pressing on the dorsum of the lower limbs.
Supplementary suckling technique (SST) was used in the management of SAM infants using supplementary feeding formulas Diluted F feeding (DF100), a milk formula with higher protein and energy content with cup for those presenting with no oedema, while on the other hand those with oedema were fed on feeding formula F75, a low-protein milk-based formula diet. All mothers were counseled about breastfeeding attachment and positioning according to the WHO guidelines (13). Mothers were then trained on the use of SST in TFC. Nasogastric tube size 8 was used for feeding. The tip of the tube was cut about 1 cm and the end of the tube was put in a cup with supplementary diluted F100 or F75. The tip of the tube adhered to the nipple was put inside the angle of the infant’s mouth during breastfeeding. The cup was initially put 5- 10 cm below the breast, while the child was breastfeeding, the milk from the cup was sucked up through the tube and was taken by the infant initially. The cup was then gradually lowered to approximately 30 cm below the level of the breast so that the milk did not flow too quickly. The volume of F100D in the cup was gradually reduced when the infant gained weight for 2-3 days (at least 20g per day), was free from illness and breast milk flow was evident. Then supplementary milk in the cup (F100D) was reduced by one third and the mother continued breastfeeding for 2 or 3 days. When the weight continued to rise, the amount of supplementary milk (F100 D) was reduced until the milk was no longer needed and the infant showed gain in weight from exclusive breastfeeding without any supplementary milk. If weight gain was not satisfactory when the volume of supplementary milk was reduced then the policy was to increase the volume to the previous level for the next 2 days, with a successive repeat trial.
All infants received a course of antibiotic with amoxicillin. Four infants developed signs of pneumonia and three had gastroenteritis and sepsis who were all then consequently transferred to the inpatient care and received all necessary treatment but succumbed thereafter.
An infant was discharged if he/she was gaining adequate weight on isolated breastfeeding to be followed in the nutritional units at health centers.
These denoted that that infant was cured, died, defaulter or in non-recovery state. Cured denoted an infant who was gaining adequate weight on breastfeeding at the time of discharge. Infants who died were those who were transferred to inpatient care after development of complications at TFC.
The standard indices of nutritional status of children were weight, length, weight-for-length Z score (WLZ) and weight-for-age Z score. Indicators were expressed as Z scores in standard deviation (SD) from the median of the reference population. The WHO Multicentre Growth Reference Study Group growth standards was the reference (14).
The Ministry of Public Health and Population in Yemen and the Faculty of Medicine and Health Sciences, University of Aden, Yemen both provided the ethical approval. Mothers agreed to participate in the study after signing an informed consent form.
The data collected and the variable results were entered into the computer, using SPSS® (Statistical Package for Social Science SPSS Version 24) software for windows. Test of normality (Shapiro-Wilk) was performed on the continuous data and parametric tests were used where the data were normally distributed otherwise nonparametric were used. Ccontinuous variables were expressed as mean (SD) when normally distributed or otherwise median interquartile range (IQR). Hence either the paired student t test or the Wisconsin test was used for comparison of two sets of continuous variables. Dichotomous or otherwise nominal variables were displayed as frequency (%). Chi-square test was used to establish associations between dichotomous variables.
A binary logistic regression model was employed to determine whether the independent covariates (age, sex, vomiting, diarrhea, anemia, pneumonia, seizure, WAZ scores on admission and discharge and duration of treatment) had had any effect on the dependant variable “outcome” as dichotomous variable. The outcome was divided into cured (the infant was gaining weight) and not cured (defaulter, non-recovery or death) and odd ratio (OR) with 95% CI was used for outcome. A p value <0.05 was considered significant.