PT - JOURNAL ARTICLE AU - Trevor Duke AU - Francis Pulsan AU - Doreen Panauwe AU - Ilomo Hwaihwanje AU - Martin Sa'avu AU - Magdalynn Kaupa AU - Jonah Karubi AU - Eleanor Neal AU - Hamish Graham AU - Rasa Izadnegahdar AU - Susan Donath TI - Solar-powered oxygen, quality improvement and child pneumonia deaths: a large-scale effectiveness study AID - 10.1136/archdischild-2020-320107 DP - 2020 Oct 16 TA - Archives of Disease in Childhood PG - archdischild-2020-320107 4099 - http://adc.bmj.com/content/early/2020/10/15/archdischild-2020-320107.short 4100 - http://adc.bmj.com/content/early/2020/10/15/archdischild-2020-320107.full AB - Background Pneumonia is the largest cause of child deaths in low-income countries. Lack of availability of oxygen in small rural hospitals results in avoidable deaths and unnecessary and unsafe referrals.Method We evaluated a programme for improving reliable oxygen therapy using oxygen concentrators, pulse oximeters and sustainable solar power in 38 remote health facilities in nine provinces in Papua New Guinea. The programme included a quality improvement approach with training, identification of gaps, problem solving and corrective measures. Admissions and deaths from pneumonia and overall paediatric admissions, deaths and referrals were recorded using routine health information data for 2–4 years prior to the intervention and 2–4 years after. Using Poisson regression we calculated incidence rates (IRs) preintervention and postintervention, and incidence rate ratios (IRR).Results There were 18 933 pneumonia admissions and 530 pneumonia deaths. Pneumonia admission numbers were significantly lower in the postintervention era than in the preintervention era. The IRs for pneumonia deaths preintervention and postintervention were 2.83 (1.98–4.06) and 1.17 (0.48–1.86) per 100 pneumonia admissions: the IRR for pneumonia deaths was 0.41 (0.24–0.71, p<0.005). There were 58 324 paediatric admissions and 2259 paediatric deaths. The IR for child deaths preintervention and postintervention were 3.22 (2.42–4.28) and 1.94 (1.23–2.65) per 100 paediatric admissions: IRR 0.60 (0.45–0.81, p<0.005). In the years postintervention period, an estimated 348 lives were saved, at a cost of US$6435 per life saved and over 1500 referrals were avoided.Conclusions Solar-powered oxygen systems supported by continuous quality improvement can be achieved at large scale in rural and remote hospitals and health care facilities, and was associated with reduced child deaths and reduced referrals. Variability of effectiveness in different contexts calls for strengthening of quality improvement in rural health facilities.Trial registration number ACTRN12616001469404.