Ozone therapy for patients with COVID-19 pneumonia: Preliminary report of a prospective case-control study

Background

In the early stages of the COVID-19 pandemic, clinicians had few proven options for treating the severe pneumonia that could follow infection, and interest turned to therapies that might calm the excessive inflammation and oxidative stress seen in the sickest patients. Ozone therapy, most commonly delivered as ozonated autohaemotherapy - where a patient's own blood is briefly mixed with a medical oxygen-ozone gas mixture and then reinfused - had long been used in some settings for its proposed antioxidant and immune-modulating effects. A team of Spanish and North American researchers set out to test, in a real hospital setting, whether adding this therapy to standard care could speed recovery in patients admitted with severe COVID-19 pneumonia.

Study design

The study followed 18 hospitalised adults with confirmed, severe COVID-19 pneumonia, split evenly into two groups of nine. One group received ozonated autohaemotherapy alongside standard care; the other received standard care alone, which at the time included oxygen support and the medicines then in common use, such as hydroxychloroquine, lopinavir/ritonavir, corticosteroids and antibiotics. The ozone group had 200 millilitres of their own blood withdrawn, mixed with 200 millilitres of an oxygen-ozone gas mixture at a concentration of 40 micrograms per millilitre, and reinfused twice daily for a median of four days. Patients were not randomly allocated to a group, and outcome assessors were not blinded to treatment. The main measure was the time from hospital admission to clinical improvement, defined as a two-point drop on a standard six-point clinical severity scale. The researchers also tracked inflammatory blood markers, how quickly patients tested negative for the virus, days free from ventilation, and death within 28 days.

Findings

Patients who received ozonated autohaemotherapy reached clinical improvement in a median of 7 days, compared with 28 days in the control group (p=0.04). By day 14, 88.8% of the ozone group had improved, against 33.3% of the control group (p=0.01). After adjusting for baseline differences between the groups, ozone therapy was associated with clinical improvement occurring 11.3 days sooner on average, though the confidence interval around this estimate was wide (95% CI -22.25 to -0.42). The ozone group also showed faster two-fold falls in several inflammatory markers - C-reactive protein, ferritin, D-dimer and lactate dehydrogenase - and cleared the virus on PCR testing sooner. There was no statistically significant difference between the groups in ventilator-free days or in mortality at 28 days (11.1% in the ozone group versus 22.2% in the control group, p=1).

What the authors concluded

The authors concluded that ozonated autohaemotherapy was associated with a significantly shorter time to clinical improvement in patients with severe COVID-19 pneumonia, and suggested it may be a useful adjunct to standard care. They were careful to frame this as preliminary evidence, stating that the small sample size and study design meant the findings would need to be confirmed in larger, randomised controlled trials before any firm conclusions could be drawn.

Limitations of this study

This was a small, single-centre study of only 18 patients, with just nine in each group, so individual patients could heavily influence the results and the confidence intervals around key findings were wide. Patients were not randomly assigned to treatment, and outcome assessors knew which group each patient was in, both of which can introduce bias. The control group was, on average, slightly older and heavier than the ozone group at baseline, which may have affected recovery times independently of the treatment given, although the authors reported that additional sensitivity analyses supported their main findings. The observational design also meant the researchers could not measure cytokine levels in detail. Finally, the standard care used for comparison reflects treatment protocols from mid-2020, including medicines such as hydroxychloroquine and lopinavir/ritonavir that are no longer part of routine COVID-19 care, so the findings may not translate directly to how the illness is managed today.