A mixed method multi-country assessment of barriers to implementing pediatric inpatient care guidelines

PLoS One. 2019 Mar 25;14(3):e0212395. doi: 10.1371/journal.pone.0212395. eCollection 2019.


Introduction: Accelerating progress in reducing child deaths is needed in order to achieve the Sustainable Development Goal child mortality target. This will require a focus on vulnerable children-including young children, those who are undernourished or with acute illnesses requiring hospitalization. Improving adherence to inpatient guidelines may be an important strategy to reduce child mortality, including among the most vulnerable. The aim of our assessment of nine sub-Saharan African and South Asian hospitals was to determine adherence to pediatric inpatient care recommendations, in addition to capacity for and barriers to implementation of guideline-adherent care prior to commencing the Childhood Acute Illness and Nutrition (CHAIN) Cohort study. The CHAIN Cohort study aims to identify modifiable risk factors for poor inpatient and post discharge outcomes above and beyond implementation of guidelines.

Methods: Hospital infrastructure, staffing, durable equipment, and consumable supplies such as medicines and laboratory reagents, were evaluated through observation and key informant interviews. Inpatient medical records of 2-23 month old children were assessed for adherence to national and international guidelines. The records of children with severe acute malnutrition (SAM) were oversampled to reflect the CHAIN study population. Seven core adherence indicators were examined: oximetry and oxygen therapy, fluids, anemia diagnosis and transfusion, antibiotics, malaria testing and antimalarials, nutritional assessment and management, and HIV testing.

Results: All sites had facilities and equipment necessary to implement care consistent with World Health Organization and national guidelines. However, stockouts of essential medicines and laboratory reagents were reported to be common at some sites, even though they were mostly present during the assessment visits. Doctor and nurse to patient ratios varied widely. We reviewed the notes of 261 children with admission diagnoses of sepsis (17), malaria (47), pneumonia (70), diarrhea (106), and SAM (119); 115 had multiple diagnoses. Adherence to oxygen therapy, antimalarial, and malnutrition refeeding guidelines was >75%. Appropriate antimicrobials were prescribed for 75% of antibiotic-indicative conditions. However, 20/23 (87%) diarrhea and 20/27 (74%) malaria cases without a documented indication were prescribed antibiotics. Only 23/122 (19%) with hemoglobin levels meeting anemia criteria had recorded anemia diagnoses. HIV test results were infrequently documented even at hospitals with universal screening policies (66/173, 38%). Informants at all sites attributed inconsistent guideline implementation to inadequate staffing.

Conclusion: Assessed hospitals had the infrastructure and equipment to implement guideline-consistent care. While fluids, appropriate antimalarials and antibiotics, and malnutrition refeeding adherence was comparable to published estimates from low- and high-resource settings, there were inconsistencies in implementation of some other recommendations. Stockouts of essential therapeutics and laboratory reagents were a noted barrier, but facility staff perceived inadequate human resources as the primary constraint to consistent guideline implementation.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Africa South of the Sahara
  • Antimalarials / therapeutic use
  • Cohort Studies
  • Delivery of Health Care / trends*
  • Female
  • Guideline Adherence / trends*
  • Health Services Accessibility / standards
  • Health Services Accessibility / trends
  • Hospital Administration
  • Hospitalization
  • Hospitals
  • Humans
  • Infant
  • Inpatients
  • Malaria / epidemiology
  • Male
  • Pediatrics / trends*
  • World Health Organization


  • Antimalarials

Associated data

  • Dryad/10.5061/dryad.898m083

Grants and funding

This study was funded by the Bill and Melinda Gates Foundation (OPP1131320). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.