Review section

Change

Description of the intervention

We added the paragraph below on 'nonclassical' hyperbaric oxygen therapy:

In addition to the 'classical' hyperbaric oxygen therapy defined by the Undersea and Hyperbaric Medical Society (UHMS 2016), some ASD trials have used nonclassical hyperbaric oxygen therapy (Granpeesheh 2010; Rossignol 2009). In these trials, nonclassical hyperbaric oxygen therapy consisted of a chamber pressurized to greater than one ATA with less than 100% oxygen concentration. In one study, both classical and nonclassical hyperbaric oxygen therapies produced significant improvement in children with autism, as evidenced by normal levels of oxidative stress and inflammation markers (Rossignol 2007a). As outlined in our protocol (Xiong 2014), we assessed only the use of classical hyperbaric oxygen therapy in this review.

Types of participants

We revised this section as follows:

Participants of any age with a diagnosis of autism spectrum disorder (ASD) based on the criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM‐5) (APA 2013); or individuals with a diagnosis of one of the four pervasive developmental disorders from fourth edition text revision version of the DSM (DSM‐IV‐TR) (APA 2000), including autistic disorder, Asperger syndrome, or pervasive developmental disorder not otherwise specified (PDD‐NOS); or from theInternational Classification of Mental and Behavioural Disorders, 10th Edition (ICD‐10; WHO 1993). We accepted diagnoses that were derived following use of assessment tools, such as the Autism Diagnostic Observation Scale (ADOS) (Lord 1997) and the Autism Diagnostic Interview ‐ Revised (ADI‐R) (Lord 1994).

Electronic searches

We have specified that, when searching online clinical trial registries such as ClinicalTrials.gov and WHO ICTRP, we checked the publication status of each trial identified and contacted study authors for results of finished unpublished trials.

Searching other resources

We have clarified that we handsearched reference lists of relevant studies. We also searched the gray literature from the Internet using the academic search engine "Baidu Scholar."

Measures of treatment effect

We have clarified how we will handle final values and changes from baseline data, as follows:

When final values and changes from baseline data are available in included trials, we shall analyse them separately. Apart from analysing those values separately, we will combine final values and changes from baseline data using the MD when both types of data are available for the same scale. We will not incorporate skewed data in future analyses.

Unit of analysis issues

We explained how we would handle multiple intervention groups. See Appendix 2.

Assessment of heterogeneity

We regraded the degree of heterogeneity, as follows (Deeks 2011).

1. 0% to 40%: might not be important.

2. 30% to 60%: may represent moderate heterogeneity.

3. 50% to 90%: may represent substantial heterogeneity.

4. 75% to 100%: may represent considerable heterogeneity.

We added the following information:

Studies have shown that different estimation methods may lead to different results and conclusions. For example, the DerSimonian and Laird (DL) estimator, which is currently widely used by default to estimate between‐study variance, has been long challenged (Veroniki 2016). The DL estimator can lead to erroneous conclusions (Cornell 2014) or can largely underestimate the true value for dichotomous outcomes (Novianti 2014). For continuous data, the restricted maximum likelihood estimator is a better alternative for estimating between‐study variance when compared with other estimators (Veroniki 2016).

We also specified that:

We plan to assess heterogeneity by comparing the estimated magnitude of the heterogeneity variance with the empirical distribution of Turner 2012 for dichotomous data and Rhodes 2015 for continuous data.

Data synthesis

We have clarified when we will report results of the fixed‐effect model and when we will report results of the random‐effects model, as follows:

If no significant heterogeneity is present, we will report the results of the fixed‐effect model only. If significant heterogeneity or severe asymmetry of the funnel plot is observed, we will report the results of the random‐effects model.