The prevalence of hypoxaemia in paediatric and adult patients in health-care facilities in low-income and middle- income countries: a systematic review and meta-analysis

Background Hypoxaemia (low oxygen saturation in blood) is a key predictor of in-hospital mortality, affecting people of all ages with many different conditions. Early detection and treatment of hypoxaemia are critical, but there are few data to quantify hypoxaemia burden outside the child pneumonia population. We aimed to estimate hypoxaemia prevalence for adults and children with acute illness attending health facilities in low-income and middle-income countries (LMICs). Methods We conducted a systematic review and meta-analysis, searching MEDLINE, PubMed, Embase, Cumulated Index in Nursing and Allied Health Literature, Index Medicus, and Google Scholar for studies reporting hypoxaemia prevalence among patients attending health facilities. We included articles with original data on peripheral blood oxygen saturation (SpO2), from an LMIC, published between Jan 1, 1998, and Jan 10, 2023. We included studies in acutely unwell people of any age and with any condition, but excluded those admitted to intensive care units, receiving perioperative care, or attending hospital for preventive or chronic care. We assessed study quality using Joanna Briggs Institute's Checklist for Prevalence Studies. Two reviewers independently conducted title and abstract screening, full- text review, data extraction, and quality assessment, requesting summary data from authors. We reported pooled prevalence of hypoxaemia (typically defined as SpO2 <90%) overall and by condition, using a random-effects meta- analysis model. This study is registered with PROSPERO, CRD42019136622. Findings We identified 9173 unique records from searches and included 213 in meta-analyses involving 601 757 participants. The majority of studies were from the World Bank regions of sub-Saharan Africa (108 [51%] of 213) or south Asia (58 [27%]). The pooled prevalence of hypoxaemia among admitted patients was 24<middle dot>5% (95% CI 19<middle dot>9-29<middle dot>4) for neonates (aged 0-28 days), 12<middle dot>1% (10<middle dot>0-14<middle dot>4) for children (aged 1 month-17 years), and 10<middle dot>8% (4<middle dot>9-18<middle dot>7) for adults (aged >= 18 years). Hypoxaemia prevalence was highest in neonatal and primary respiratory conditions but still common in many other conditions. Hypoxaemia was associated with 4<middle dot>84 (95% CI 4<middle dot>11-5<middle dot>69) times higher odds of death than no hypoxaemia. Interpretation Hypoxaemia is common across all age groups and a range of primary respiratory and other critical illnesses and is strongly associated with death. These estimates will inform oxygen-related strategies and programmes, and integration of pulse oximetry and oxygen into clinical guidelines, service structures, and strategies for maternal, neonatal, child, adolescent, and adult health.