After hospital discharge children may look better from their illness, but remain weak and susceptible to infection. CHAIN found half of all study deaths happened after discharge.


Follow your doctor's advice about when it's the right time to leave hospital to help prevent deaths after discharge. When leaving, ask your doctor to schedule a follow-up visit for your child, and how to recognize danger signs before going home:

  • Drowsiness                  Not being able to feed
  • Breathing difficulty       Persistent vomiting

Background  |  Findings from the CHAIN Study  |  What the CHAIN results mean  |  CHAIN Information


Post-discharge mortality (deaths that occur after discharge from hospital) is a widely recognized phenomenon in higher- and lower-income settings. Current WHO and most national guidelines provide little guidance for post-discharge care or specific recommendations for the care of children with challenging social circumstances. Most studies that have attempted to evaluate risk factors for post-discharge mortality have focused on clinical features present at admission, many or all of which may have resolved or changed by the time of discharge. Identifying vulnerable children at the time of discharge and targeting effective post-discharge interventions to those at highest risk may be a highly effective strategy to accelerate progress towards the achievement of the child mortality targets of the Sustainable Development Goals.1

Reported post-discharge fatality rates range widely, from 1-2% to over 20% in low and middle-income countries. While some of this variability is due to differences in underlying populations and environment, up to now, studies have also had widely varying methods, including differences in inclusion criteria, follow up duration and loss to follow up.1-6 The CHAIN cohort study aimed to identify reliable figures and risk factors for mortality through a harmonised study across multiple different communities and settings in Africa and South Asia.

To the top

Findings from the CHAIN cohort

CHAIN enrolled 3,101 children across nine sites in six countries. Overall, 11% of the children enrolled in the CHAIN cohort-study died. Half of these deaths (52%) occurred during admission and half (48%) occurred in the post-discharge period. Two-thirds (67%) of all child deaths, both inpatient and post-discharge, occurred within 30-days of the child being admitted to hospital.

Nutritional status was an important predictor of post-discharge mortality risk. Children without evidence of wasting had lower risk of death in the post-discharge period (1.6% died), while those with moderate wasting (4.3% died), and severe wasting or kwashiorkor (11% died) were more likely to die. The proportion of deaths happening post-discharge was similar across all nutritional strata (see Table 1 below). About half of all post-discharge deaths (54%) happened at home and half (46%) happened after the child returned to hospital. Home deaths were more common among wasted children than among those without wasting (P=0.04).

Table 1. Place of post-discharge deaths.



All deaths
N (%)

Place of death

Deaths during hospital readmission
N (% of deaths)

Deaths in community
N (% of deaths)



168 (5.9)

78 (46)

90 (54)

Not wasted


17 (1.6)

11 (65)

6 (35)

Moderately wasted


30 (4.1)

17 (57)

13 (43)

Severely wasted/Kwashiorkor


121 (11)

50 (41)

71 (59)

Factors associated with 30-day mortality: Children who were severely wasted, were younger, were more severely ill at admission, HIV-exposed and/or infected, had underlying pre-existing medical conditions, had adverse caregiver characteristics, (e.g., maternal mental and physical health, education, employment) and household-level exposures (e.g., food insecurity, unimproved toilet and water source), and had limited access to healthcare were more likely to die in the 30-days after hospital admission.

Factors associated with mortality within 180 days after discharge: Children who were severely wasted or had oedematous malnutrition (kwashiorkor) at admission to hospital, those with adverse caregiver characteristics, those with signs of illness severity at discharge, those who were discharged against medical advice, and those who were HIV-exposed but uninfected had higher risk of mortality within the 180 days following discharge from hospital.

Duration of hospital admission was not directly associated with post-discharge mortality, which may suggest that keeping children in hospital longer than necessary is unlikely to reduce mortality.

At discharge, many children (40%) had a very low risk of dying in the subsequent 180 days (less than 1% mortality). However, using the factors noted above, we were able to identify a group of children (20%), who had an 18% risk of dying. These findings suggest that risk assessment at discharge may be an important step in preventing post-discharge mortality.

To the top   

What the CHAIN results mean

CHAIN found that children’s clinical features at discharge were more strongly associated with post-discharge death than admission features. Post-discharge mortality is as predictable as inpatient mortality when children’s admission, discharge, and social circumstance information is utilized, and more than half of post-discharge deaths occurred at home without readmission to hospital.

Despite being highly predictable, post-discharge care is not addressed in current guidelines. To further reduce childhood mortality, a fundamental shift to risk-based approaches for post-discharge management is needed. Risk stratification among paediatric admissions may help identify low-risk children who could be discharged earlier using formal criteria. In turn, this may reduce the burden of financial and non-financial costs for families and for the health system and may enable reallocation of staff and resources for children at higher risk of death.

Suggested interventions at discharge to improve mortality:

  • ‘Down-referral’ of high-risk children to a named/known community health worker
  • Structured phone and short messaging services (SMS) contact with health professionals to deliver health advice and identify children in need of medical evaluation
  • At discharge, empower health care staff with a checklist based on clinical features to identify children at increased risk pf post-discharge mortality
  • Train mothers and family members to recognize clinical danger signs
  • Facilitate priority emergency room access for recently discharged children (e.g., facilitate travel costs, avoid queueing and waive hospital costs) to help overcome families’ hesitancy in re-presenting to hospital.
  • Early identification of families who may leave against medical advice and addressing their concerns
  • Interventions to improve maternal wellbeing, agency, and financial independence are needed

To the top

  1. Nemetchek B, English L, Kissoon N, et al. Paediatric postdischarge mortality in developing countries: a systematic review. BMJ Open 2018;8(12):e023445. doi: 10.1136/bmjopen2018-023445
  2. Bwakura-Dangarembizi M, Dumbura C, Amadi B, et al. Risk factors for postdischarge mortalityfollowing hospitalization for severe acute malnutrition in Zimbabwe and Zambia. The American journal of clinical nutrition 2021;Jan 20;nqaa346. doi: 10.1093/ajcn/nqaa346. Online ahead of print doi: 10.1093/ajcn/nqaa346 [published Online First: 2021/01/21]
  3. Chisti MJ, Graham SM, Duke T, et al. Post-discharge mortality in children with severe malnutritionand pneumonia in Bangladesh. PloS one 2014;9(9):e107663. doi: 10.1371/journal.pone.0107663 [published Online First: 2014/09/17]
  4. Moisi JC, Gatakaa H, Berkley JA, et al. Excess child mortality after discharge from hospital in Kilifi,Kenya: a retrospective cohort analysis. Bulletin of the World Health Organization 2011;89(10):725-32, 32A. doi: 10.2471/BLT.11.089235
  5. Ngari MM, Fegan G, Mwangome MK, et al. Mortality after Inpatient Treatment for Severe Pneumonia in Children: a Cohort Study. Paediatr Perinat Epidemiol 2017;31(3):233-42. doi: 10.1111/ppe.12348
  6. Chhibber AV, Hill PC, Jafali J, et al. Child Mortality after Discharge from a Health Facility followingSuspected Pneumonia, Meningitis or Septicaemia in Rural Gambia: A Cohort Study. PloS one 2015;10(9):e0137095. doi: 10.1371/journal.pone.0137095


CHAIN is continuing to examine laboratory markers of mortality mechanisms and risks to better identify new treatments that can be tested in clinical trials. CHAIN investigators are working in Kenya, Malawi, Uganda, Burkina Faso, Bangladesh and Pakistan. The project is coordinated from the KEMRI/Wellcome Trust Research Programme, Nairobi, Kenya, and the Department of Global Health, University of Washington, Seattle, USA. Collaborating centres are in the UK, USA, Canada and the Netherlands. CHAIN is supported by the Bill & Melinda Gates Foundation as a grant to the Centre for Tropical Medicine & Global Health, University of Oxford, UK.

To the top

Acute Care   |  Discharge   |  Leaving Against Medical Advice   |  Stigma and Wasting

Partner Organisations: