Malnutrition remains one of the main health problems prevalent among Ethiopian children [5]. This was also observed in our pediatric surgical patients with a prevalence rate of 45%. To our knowledge, there is no other similar study in the country to compare our results. Although the observed prevalence of wasting, stunted, and underweight was lower in comparison to the Ethiopian demographic and health survey, a comparable estimate of malnutrition was seen in community-based studies done in a rural part of Ethiopia,48.5% by Endris N et al. [17].

Studies done on pediatric surgical patients in other low- and middle-income countries have revealed high prevalence rates of malnutrition. Adigun and Ogun Doyin from Nigeria reported a 46.2% prevalence, and from India, Pooja and Dave et al. reported a 46% prevalence [12, 18]. These results are higher than the reported prevalence from higher-income countries like Germany (6·1%), France (11%), and Brazil (6·9%) [11]. The difference could be explained by good economic status with good healthcare systems in high-income countries.

Infants were more likely to be malnourished than older children similar to reports by Barutçu A. and Barutçu S [19]. Similarly, the Nigerian study showed that children aged 1 year and below are 4·28 times more likely to suffer from malnutrition than older ones [12]. Likewise, in the study done by Ross et al., on the cardiac surgical patient the infants were found to have higher rates of malnutrition than neonates and older children [20]. On the contrary, from community-based studies of prevalence as the age of a child increases the risk of being malnourished increases [17].The difference between the community report and our result could be because 78.3% of the infants were having gastrointestinal anomalies and disease which result in poor intake and malabsorption affecting their nutritional status.

Factors that could affect the prevalence of malnutrition in other studies like maternal education level, and the economic status of the family didn’t show any association in our study [17, 21]. This could be because 73·6% of the patients came from urban settings, and were a homogeneous group in terms of the wealth index. even though more female patients were malnourished compared with male patients, their number is small to conclude.

On the assessment of the impact of malnutrition on postoperative complications, there are controversial results with clear associations in some and no correlation in others. Two systemic reviews were done to answer and clarify the effect of nutritional status on postoperative outcomes. Wessner et al., were trying to determine if nutritional assessment impact clinical outcome [7]. They found weak evidence of preoperative nutritional assessment being predictive of adverse clinical outcomes in pediatric surgical patients. However, this review only identified six studies in total, five of which were undertaken in a pediatric cardiac surgical patient, limiting the validity of these findings to other non-cardiac surgical patients. The other review was done by R. Hill et al., on the prognostic effect of undernutrition on infectious complications in children undergoing surgery. They analyzed twelve studies of which only four of them deal with general surgical cases. Even though there was some evidence showing a relationship between undernutrition and the risk of developing any infection-related complication again the evidence was weak [22]. One of the challenges faced by both reviews was the uniformity of nutritional assessment as well as the heterogeneity of the disease conditions. Hence, they failed to conclude.

In our study, even if infectious complications were seen more in malnourished patients, it was not statistically significant (p = 0·06). This result is consistent with the study done in Nigeria [12]. However, Secker et al. reported a significantly higher postoperative infection rate in the malnourished population versus the well-nourished group (p = 0·02) [8]. They used the Subjective Global Nutritional Assessment tool to determine nutritional status. But when they use the objective measures, no association was observed except height-for-age and postoperative length of stay. With the high prevalence of malnutrition in our patients and the immunosuppressive effect on surgical patients, it would be expected to have a high rate of infection. Nonetheless, the lower sample size in our study may be the reason for the non-significant findings.

Assessing the effect on the length of hospital stay was important not only to estimate the hospital cost and economic impact but also to determine the turnover rate of our patients because of the long waiting list. However malnourished patients did not have significantly longer hospital stay compared with well-nourished patients. And also, the biochemical profiles and the subjective assessment method failed to predict the effect on the length of postoperative stay. The limitations of the study were, the biochemical profile was not assessed for all patients, the lack of a validated subjective nutritional assessment tool suitable for our setup, and heterogeneity of the disease may affect the outcome measurement as some disease conditions may need longer hospital stay without any complications.