Children with severe malnutrition who are sick and admitted to hospitals have high mortality, usually because of infection. Malnourished children have more potentially harmful bacteria in their upper intestines than well-nourished children and this may contribute to inflammation in the gut and whole body. These bacteria may cross from the intestines to the bloodstream causing life-threatening infections. A related abnormality among malnourished children is reduction in the digestive enzymes made by the pancreas and the liver. Apart from helping with digestion of food, these enzymes are important in helping the body control bacteria in the upper intestines. It is therefore possible that treatment with digestive enzymes could help reduce the burden of harmful bacteria and thus lower inflammation and the risk of serious infection. One study conducted in Malawi has shown that children with severe malnutrition who were supplemented with pancreatic enzymes had a lower risk of dying. However, this was a small study and although promising, requires validation. No studies of supplementation with bile acids have been done among severely malnourished children. However, bile acids are commonly used to manage patients with liver function abnormalities, something that malnourished children suffer from as well. The investigators want to find out if supplementing these pancreatic enzymes and bile acids among ill children with severe acute malnutrition is safe and reduces the risk of death, deterioration or readmission to hospital.

Severely malnourished children who present with an acute illness have a high risk of mortality. Severe malnutrition is associated with intestinal inflammation and changes in the fecal microbiome ('dysbiosis'). Apart from in the large intestine, dysbiosis is also present in the small (upper) intestine, where increased bacterial density and altered microbial composition can contribute to intestinal inflammation and intestinal dysfunction which may ultimately contribute to the development of sepsis and death. The bacterial density and composition in the small intestine can be altered using antibiotics. However, apart from side effects, antibiotic use contributes to the development of antibiotic resistance, which is very common in hospitalized malnourished patients and can pose a threat to both individual and public health. In addition to intestinal dysbiosis and intestinal inflammation, children with severe malnutrition suffer from impaired exocrine pancreatic and hepatobiliary function, which are important for nutrient digestion and absorption to aid their recovery.

One previous pilot trial showed that treating children with severe malnutrition with pancreatic enzymes led to a reduction in mortality. The investigators hypothesize that supplementing these enzymes exogenously to severely malnourished children will improve their clinical outcome by reducing dysbiosis, intestinal inflammation and sepsis.

The objective of this study is to determine whether treating ill severely malnourished children with pancreatic enzymes or bile acids is safe and improves mortality. The investigators will conduct a double blind, randomized clinical trial in a 2x2 factorial design in hospitalized severely malnourished children. Participants will be treated with pediatric formulations of pancreatic enzymes, bile acids, both or placebo for 21 days. Participants will be followed up daily during their hospital stay and on day 21 and 60 after enrollment.

This trial will be conducted in three stages to allow for careful interim evaluations to assess safety and study progress. After the first and second stage, interim analyses assess safety and likelihood of benefit before enrolling the full sample size to assess mortality as the primary outcome. Secondary outcomes include adverse events, length of hospital stay and growth. Exploratory outcomes will examine intestinal and systemic inflammation and metabolic changes to examine mechanisms affected by the interventions, and costs.

Two sub-studies will be conducted. In Kenya and Bangladesh, daily blood gases, lactate and biochemistry for the first 5 days to assess early clinical progress. In Malawi and Uganda, hydrogen breath testing will be used to evaluate impact on upper small intestinal bacterial overgrowth.

Overall, this trial will determine whether a non-antibiotic, bactericidal intervention given in additional to standard of care management reduces mortality in acutely ill severely malnourished children.

Inclusion Criteria:

• Age 2 to <59 months
• Admitted to hospital with an acute, non-traumatic illness and within 72 hours of admission at the time of enrolment
• Severe malnutrition (weight-for-height <-3 z scores of the median WHO growth standards and/or mid upper arm circumference <115mm (<110mm age below 6 months), or symmetrical oedema of at least the feet related to malnutrition (not related to a primary cardiac or renal disorder)
• Able to feed orally in usual state of health.
• Accompanied by care provider who provides written informed consent
• Primary caregiver plans to stay in the study area during the duration of the study
• Presence of two or more features of severity as specified below. If a child meets two criteria, they may be enrolled before further criteria are assessed (e.g. a child may be eligible on clinical signs before the complete blood count results are known):

Respiratory distress "Subcostal indrawing" or "nasal flaring" or "head-nodding" Oxygenation "Central cyanosis" or SaO2 <90% Circulation Limb temperature gradient or capillary refill >3 seconds Conscious level AVPU < "A" Pulse > 180 per min Haemoglobin < 7g/dl Blood glucose < 3mmol/L White blood cells < 4 or > 17.5 x 109/L Temperature <36 or >38.5oC Very low MUAC MUAC <11cm

Exclusion Criteria:

Anonymised data may be shared through application to DGC@kemri-wellcome.org. Datasets will be uploaded on the HARVARD DATAVERSE for managed access

Sequential study numbers will be computer generated according to a blocked randomization list of random block sizes allocated to each site before the study begins.

Identical study investigation medical product (IMP) packs will be labelled with sequential study numbers according to a prepared blocked randomization list before the trial begins. An independent off-site trial-pharmacist will implement labeling of the IMP with unique study numbers according to the randomization list.

Study numbers will be issued in consecutive order at each site.