To investigate the clinical effects of time window on hyperbaric oxygen treatment (HBOT) in patients with disorders of consciousness (DOC).

All the clinical research literature regarding HBOT for DOC published between January 2000 and November 2020 were retrieved from China National Knowledge Infrastructure (CNKI), Wanfang Standards Database (WFSD) and VIP Database using Chinese key words disorders of consciousness, the vegetable state, minimally conscious state, or hyperbaric oxygen followed by a comprehensive meta-analysis.

The query gave rise to 348 results, in which 21 articles were eventually selected for meta-analysis. Among the selected 21 articles, 18 articles involved a time window comparison. All the patients were classified into < 60- (718 patients) and ≥ 60- (374 patients) day groups depending on the number of days from HBOT initiation. The Jadad scores for the included datasets were relatively low in general with 2 points as the highest score. Comparable baseline data were demonstrated in all of the articles. Datasets from different sources were pooled and analyzed, and the results suggested that the clinical curative effect rate in the treatment group was significantly higher compared with that in the control group (curative effect rate: 69.86% versus 42.30%; Z = 11.28, P = 0.000, odds ratio = 3.80, 95% CI = 3.02-4.80). Additionally, the adverse reaction rate of the < 60-day group was found to be significantly lower compared with that of the ≥ 60-day group (Z = 10.01, P = 0.000, odds ratio = 4.82, 95% CI = 3.54-6.56). The funnel diagram in articles related to curative effect analysis and time window evaluation is inverted and symmetrical, indicating that publication bias was not significant.

The clinical curative effect of the HBOT group is higher compared with that of the control group. However, the conclusions based on meta-analysis are limited because of the methodological problems of some studies. Therefore, the clinical efficacy needs to be further tested using carefully designed large sample trials (multicenter, randomized, controlled, and double-blind).