Medical oxygen service readiness and service coverage in seven countries in Africa and Asia: a cross-sectional survey and clinical audit.

Publication date: Oct 01, 2025

The COVID-19 pandemic revealed long-standing deficits in access to lifesaving medical oxygen, especially in low-income and middle-income countries (LMICs). We aimed to evaluate availability and use of pulse oximetry and medical oxygen in seven LMICs in Asia and Africa to establish a baseline to inform the Medical Oxygen Project (MOXY) and national oxygen plans, and to allow for a future pre-post intervention evaluation. We conducted a cross-sectional facility survey and retrospective clinical audit of pulse oximetry and medical oxygen availability and use in government primary, secondary, and tertiary health facilities within selected high-burden subnational regions of Cambodia, Ethiopia, Laos, Liberia, Nigeria, Rwanda, and Uganda (all part of MOXY) during 2022-23. We either surveyed all facilities within the selected regions, or a purposive sample, depending on the country and level of facility, and all countries except Liberia had ward-level assessments (in all wards or a purposive sample selected according to greatest oxygen demands). Clinical audits were conducted in all countries except Uganda; data on patient numbers, sex, age, presenting symptoms, vital signs, diagnosis, treatments, and outcomes were retrieved from admissions registers and medical records. Primary outcomes were pulse oximeter and oxygen availability and use at facility and ward levels, including documented pulse oximetry measurement on admission, documented oxygen therapy for hypoxaemia, and whether minimum standards for availability were met. We reported individual country data and produced pooled estimates using random-effects meta-analyses. Facility surveys were conducted in 567 (235 [41%] primary, 285 [50%] secondary, and 47 [8%] tertiary) facilities across the seven countries between Jan 9, 2022, and Feb 16, 2023. Clinical audits included 13 847 admissions: 180 (1.3%) in primary, 10 185 (73.6%) in secondary, and 3482 (25.1%) in tertiary facilities; with 1003 (7.2%) neonates, 3331 (24.1%) children, and 9264 (66.9%) adults. 8248 (59.6%) of all admitted patients were female, 5551 (40.1%) were male, and sex data were missing for 48 (0.3%). 24% (95% CI 14-35) of primary, 74% (61-88) of secondary, and 83% (73-93) of tertiary facilities had at least one functional pulse oximeter. Oxygen was provided by 33% (22-43) of primary facilities, 90% (81-100) of secondary facilities, and 96% (91-100) of tertiary facilities. Ward-level equipment availability was highest in operating theatres, high-dependency units, and tertiary wards. Predefined minimum standards for the ratio of pulse oximeters to beds were met by fewer than half of secondary and tertiary acute wards surveyed, apart from emergency department wards. Pulse oximetry measurements on admission were documented for 79% (64-94) of neonates, 61% (38-84) of children, and 55% (31-79) of adults. Of patients with hypoxaemia, oxygen therapy was documented for 88% (79-97) of neonates, 68% (48-89) of children, and 65% (44-86) of adults. Children with hypoxaemia admitted to wards meeting minimum standards for pulse oximeters and oxygen outlets were more likely to receive oxygen (adjusted odds ratio [aOR] 18.4 [95% CI 2.9-117.5] if meeting oximeter standards and aOR 5.0 [1.1-21.8] if meeting oxygen standards) than those admitted to wards not meeting these minimum standards. We found that availability and use of pulse oximetry and oxygen in these seven LMICS (which had implemented national medical oxygen plans) were low at primary level facilities, and that in hospitals, oxygen was more consistently available than pulse oximeters, with major gaps in all areas when assessed against predefined minimum standards. We highlight strategic opportunities to continue incremental progress in medical oxygen access. Our data show that collecting oxygen metrics at ward level and patient level is possible and helps reveal inequities in access. The Gates Foundation and ELMA Philanthropies to the Clinton Health Access Initiative.

Open Access PDF

Semantics

Original Article