This was a prospective cross-sectional study, conducted on malnourished children recruited from September 2017 to August 2018, over one year period at Department of Paediatrics, Institute of Medical Sciences, Banaras Hindu University, Uttar Pradesh. The inclusion criteria were children aged 6 to 60 months fulfilling the WHO criteria of SAM [7]. The study was approved by the Ethics committee of the college (EC/454/2015-16) and informed consent was obtained from the parents of all patients before the study. We collected data on demography, anthropometry, history and clinical examination and relevant investigations were done, it included CBC, Arterial blood gas analysis, Serum electrolytes, Serum albumin, Blood sugar. Children were managed as per WHO SAM protocol [8]. Outcome was defined as discharge or death.

Triage includes rapid recognition of emergency or priority signs present in sick children. Presence of emergency signs needs simultaneous stabilisation of sick children; however presence of priority signs needs close observation and further evaluation of child. Triage of SAM children was done in paediatric emergency by resident on duty and A (airway), B (breathing), C (circulation, coma and convulsion), D (severe dehydration) model was used to identify the presence of emergency sign [8]. The emergency sign [6] are: (i) Obstructed and absent breathing; (ii) Severe respiratory distress; (iii) Central cyanosis; (iv) Signs of shock i.e., cold extremities with capillary refill time >3 sec and weak & fast pulse; (v) Coma, reduced level of consciousness; (vi) Seizures; (vii) Severe dehydration. Presence of priority signs in children warrants prompt assessment and treatment. They are: (i) Tiny baby (<2 months); (ii) Bleeding; (iii) Pallor (severe); (iv) Malnutrition: Visible severe wasting; (v) Respiratory distress; (vi) Trauma or other urgent surgical condition; (vii) Referral (urgent); (viii) oedema of both feet; (ix) Temperature <36.5°C; or >38.5°C (x) Restless, continuously irritable or lethargy; (xi) Poisoning; (xii) Burns (major).

All data were entered into excel sheet and analysed by using SPSS version 18. The demographic characteristics, clinical information was presented as frequencies with percentages. Two tailed Fisher-exact test (when row or column had value less than 5) or Chi-square test was used to compare categorical variables. The critical level of significance of the results were considered at 0.05 levels i.e., p<0.05 was considered significant.

Triage is a process of segregation of sick children on their arrival to health facility. In present study, triage was done as per WHO-ETAT on SAM children and it was found that 154 emergency signs and 293 priority signs were present at the time of admission. Therefore, out of 122 total children, 108 children who were analysed, at the time of presentation had 1 or 2 emergency and 2 or 3 priority signs; 18 children who died had 3 or 4 emergency signs at time of admission.

Demographic characteristic as shown in [Table/Fig-1]. Out of 122, 90 (73.77%) were discharged from hospital, 18 (14.75%) died and 14 (11.47%) Leave Against Medical Advice (LAMA, excluded from analysis) . Almost two third of the children (63.2%) were male and most common affected age group was less than 3 year (63%). SAM was found to be more prevalent in lower middle class (58%) as compared to upper lower class.

As shown in [Table/Fig-2] the most common emergency sign was hypoxemia (37.9%) followed by dehydration (32.4%), hypoglycaemia (21.2%), shock (18.5%), severe respiratory distress (18.5%) convulsion (10.1%). Coma and apnea were the least common signs (1.8% each).

[Table/Fig-3] depicts the most common priority signs at the time of admission; respiratory distress (64.8%) followed by edema of both feet (60.1%), severe pallor (42.5%), lethargy (33.33%), visible wasting (27.7%), hyperthermia (27.7%), hypothermia (14.8%). No children with SAM presented with burn, trauma and poisoning.

[Table/Fig-4] reveals that SAM children who presented with severe respiratory distress (Odd-30.82), dehydration (Odd-7.3), shock (Odd-5.5), hypoxia (Odd-20.8), severe pallor (Odd-4.49) and hypoglycaemia (Odd-3.69) had significant risk of death, on univariate analysis. However, child with severe respiratory distress had much higher mortality (65%) as compared to respiratory distress (5.7%). Children presenting with dehydration, shock, coma and apnea had 77.7%, 55.5%, 100%, 100% mortality, respectively. Twenty eight percent deaths were found in SAM children admitted with severe pallor, however no significant difference in outcome was found in SAM children admitted with lethargy and edema. Children with fatal outcome were 11.29 times more likely to have shock (0.001), 10.2 times more likely to have dehydration (<0.001),7.2 times more likely to have severe respiratory distress (<0.001) [Table/Fig-4].

In this present study, we triaged the SAM children at the time of admission and determined the association of emergency or priority signs with mortality. The common emergency signs were dehydration, severe respiratory distress, shock, convulsion, desaturation and hypoglycaemia, and the common priority signs were respiratory distress (tachypnoea), oedema, pallor, lethargy, visible wasting, hyperthermia and hypothermia. In the present study SAM children with fatal outcome were 11.29 times more likely to have shock (0.001),10.2 times more likely to have dehydration (<0.001),7.2 times more likely to have severe respiratory distress (<0.001). Similar observation was seen by Barungi NN et al., and found that SAM children presenting with respiratory distress has 5 time odds of death (p=0.01) and hypoxia has two and half time odds of death (p=0.01) [10]. Retrospective analysis by Girum T et al., found that SAM children presenting with shock has 3 time odds of death (p<0.0001) and anaemia has two and half time odds of death (p=0.02) [11]. Wagnew F et al., reported similar observation and concluded that SAM children presenting with shock has 8 time odds of death (p<0.0001) and anaemia has two time odds of death (p=0.02) [12]. To our best knowledge this was the first attempt to study triage of SAM children as per WHO-ETAT and to assess effect of emergency or priority signs on outcome at admission and it was found that not only emergency signs but presence of priority signs on arrival to facility like, Severe Pallor (Odd-4.49) and hypothermia (Odd-4) have significant increased risk of death in SAM children. Hypoxia (SpO₂ <95%) and hypoglycaemia (glucose <54 mg/dL) may not manifest clinically in SAM children however they co-exist with emergency or priority signs. And in present study children with fatal outcome were 20 times more likely to have hypoxia (p<0.0001), 3.6 times more likely to have hypoglycemia (p=.001). Addition of pulse oximetry to detect hypoxia and blood glucose to detect hypoglycaemia would affect the outcome of SAM children on triage, as SAM triage protocol did not consider oxygen saturation and blood glucose while segregation of sick children in emergency room.

The strength of present study is that it was a clinical study, and we determined independent signs associated with mortality during hospitalization. Clinical variables associated with poor outcome were easily identified by peripheral health workers.

The sample size was small, children who opted for LAMA were excluded, and the effect of sepsis, electrolyte, and acidosis mortality were not studied. Large sample size study would be needed to find out association of emergency or priority signs with mortality at 24 hours of hospitalisation.

Early recognition of clinical signs of shock, severe respiratory distress and dehydration would identify SAM children who are at risk of early death on arrival to facility.