A Queensland study suggests that oxygen therapy is still being overused in patients admitted to hospital with COPD and hypercapnia.
The study identified 157 eligible admissions to the Prince Charles Hospital in Brisbane during 12 months from June 2016-2017. Mean PaCO2 was 61 mmHg.
Oxygen prescriptions were documented for most patients (84%).
But concerningly, over-oxygenation – ≥10% of individual patient nursing observations recording oxygen delivery with oxygen saturation >92% – was observed in 62% of admissions.
Over-oxygenation was found to occur more often in non-respiratory ward admissions (76% v 56%; p=0.02) and in admissions without any form of oxygen prescription (84% v 58%; p=0.01).
In a sub-group of only first presentation admissions, not readmissions, 22% experienced an adverse event including a MET response or new acidosis event.
The study found a non-significant increase in the risk of an adverse event in patients exposed to over-oxygenation (HR: 2.1; 95% CI: 0.7-6.3; p=0.18).
“Although not statistically significant at the 5% level, effect estimates from modelling were consistent with a markedly increased risk an adverse event (in-hospital respiratory failure needing assisted ventilation, emergency response calls or death) after exposure to over-oxygenation; the lack of statistical significance was likely due to insufficient statistical power,” the study said.
“Over half of at-risk COPD patients were exposed to potentially harmful oxygen. This occurred in spite of longstanding evidence of potential for harm and recent efforts to increase healthcare provider awareness of the importance of strict oxygen titration to at-risk COPD patients.”
The investigators, led by Dr James Anderson from the University of Queensland and the Sunshine Coast University Hospital, said it appeared over-oxygenation of patients had not improved in almost 30 years.
“Monitoring over-oxygenation in the ward setting may be inherently difficult due to intermittently recorded vital signs and potential for interval change in respiratory status and oxygenation,” they said.
“Intermittent oximetry restricts nursing alteration to oxygen delivery dynamically or in “real time”, as is the case in high dependency areas when continuous oximetry is available.”
They said wearable continuous oximetry monitoring and automated oxygen delivery systems were potential solutions to over-oxygenation in the ward environment.
“The major drawbacks to automated systems are cost, inappropriate alarming, poor signal quality and bypassing of human clinical judgement, particularly relevant for deteriorating patients.”
Staff training, specification of oxygen targets (SpO2 range 88-92% as per guidelines) and auditing of oxygen practices may also have a place in reducing inappropriate oxygen delivery.