Improved oxygen systems for childhood pneumonia: a multihospital effectiveness study in Papua New Guinea

Summary

Background

In rural hospitals of developing countries, oxygen supplies are poor and detection of hypoxaemia is difficult. Oxygen concentrators and pulse oximeters might help to manage the disease; however, use of such technology in developing countries needs comprehensive assessment. We studied the effect of an improved oxygen system on death rate in children with pneumonia in Papua New Guinea.

Methods

We installed an improved oxygen system in five hospitals in Papua New Guinea, and assessed its use in more than 11 000 children with pneumonia (2001–07) and compared case-fatality rates. Admissions between January, 2001, and December, 2004, formed the pre-intervention group, and those between July, 2005, and October, 2007, formed the post-intervention group. Oxygen concentrators and pulse oximeters were introduced in the five hospitals, and a protocol for detection of hypoxaemia and clinical use of oxygen was supplied. All children admitted had their oxygen saturation measured; if it was less than 90%, oxygen was delivered via nasal prongs at a starting flow rate of 0·5–1 L/min. We recorded all costs associated with the establishment and maintenance of this system. The study was approved by the Medical Research Advisory Committee of Papua New Guinea, number MRAC 04.02.

Findings

Before the use of this system, 356 of 7161 children admitted in the five hospitals for pneumonia died (case-fatality rate 4·97% [95% CI 4·5–5·5]), whereas 133 of 4130 children died in the 27 months after the introduction of the system (3·22% [2·7–3·8]). After the improved system was introduced, the risk of death for a child with pneumonia was 35% lower than was that before the project began (risk ratio 0·65 [0·52–0·78], p<0·0001). Mortality rates varied between hospitals. The estimated costs of this system were US$51 per patient treated, US$1673 per life saved, and US$50 per disability-adjusted life-year (DALY) averted.

Interpretation

Pulse oximetry and oxygen concentrators can alleviate oxygen shortages, reduce mortality, and improve quality of care for children with pneumonia in developing countries. The cost-effectiveness of this system compared favourably with that of other public-health interventions.

Funding

The Papua New Guinea National Department of Health; WHO, Papua New Guinea and Western Pacific Regional Office; AirSep corporation, Buffalo, NY, USA; the Ross Trust, VIC, Australia; AusAID; Jacques Gostelli, Switzerland; and a grant from the University of Melbourne.

Introduction

Acute respiratory infection is the cause of death of 2 million children per year worldwide, 98% of whom are in developing countries.1, 2 In Papua New Guinea, the major cause of death in children under the age of 5 years is pneumonia. Hypoxaemia is the most serious complication of pneumonia,³ and is also a complication of other common diseases, particularly in newborn babies.⁴ WHO recommended treatment of children with severe pneumonia includes antibiotics and oxygen,⁵ but in Papua New Guinea, similar to many developing countries, oxygen shortages are common, especially in rural and remote areas because of the high cost and complex logistics of transporting oxygen cylinders. Oxygen concentrators—machines that generate concentrated oxygen from ambient air—have been proposed in an attempt to improve the reliability of oxygen supplies and reduce mortality rates.

Hypoxaemia is difficult to detect through clinical signs, especially in heavily pigmented children and in poorly illuminated health facilities, where clinical signs, including cyanosis, are difficult to identify.3, 4, 6 Pulse oximetry is the most reliable, non-invasive way of diagnosis for hypoxaemia,⁷ and its relevance to pneumonia management in developing countries has been emphasised by the poor sensitivity and specificity of clinical predictors of hypoxaemia, and by the increasing affordability of pulse oximeters.

Previous studies in Papua New Guinea showed that the frequency of hypoxaemia was greater in highland hospitals (40% of all patients) than was in coastal hospitals (10% of all patients). This difference is mainly due to the relative contribution of acute respiratory tract infection to overall childhood illness in every region. On the day of admission, oxygen was not available for 22% of children admitted to five hospitals.⁸ Scarce research exists on clinical efficacy and cost-effectiveness of oxygen in the treatment of pneumonia in children admitted to hospitals in developing countries.3, 9, 10, 11, 12 Most cost-comparison analyses of oxygen systems have focused on equipment costs only, and estimations have been made by theoretical modelling.¹³ However, the true cost of improving technology in developing countries is more than that needed for equipment, and countries should know all the costs of implementation and the feasibility of sustaining such a system.

We did a clinical study using oxygen concentrators and pulse oximeters in five hospitals in Papua New Guinea to assess whether installation of a reliable, sufficient, and affordable source of oxygen, together with the use of pulse oximetry, would reduce case-fatality rates in children with pneumonia.