Types of studies

There are two parts to this review; the main review concerning evidence from randomized trials, and a narrative review of safety data from non‐randomized studies. The full methods for the narrative review of safety data from non‐randomized studies is contained within Appendix 1, and summarized briefly at the end of the Methods section. The methods detailed below relate to the review of randomized controlled trials (RCTs).

Studies were eligible if they were of randomized controlled design. Cluster‐randomized controlled trials would have been eligible; however, none were identified. Trials with a cross‐over design were not eligible.

Types of participants

Humans with mpox virus infection confirmed by reverse transcription polymerase chain reaction (RT‐PCR), inclusive of participants with confirmed mpox infection as a subset population within a study, were eligible for inclusion.

Studies conducted in any country or setting were eligible for this review, including inpatient and outpatient settings.

Types of interventions

All therapeutics used for the treatment of human mpox were eligible for this review. This included, but was not limited to, tecovirimat, brincidofovir, cidofovir, NIOCH‐14, immunomodulators, and vaccine immune globulin.

Types of outcome measures

The main review addresses patient‐important outcomes as determined by the WHO Clinical Management and Infection Prevention and Control guideline development group (GDG), described below. These are outcomes of interest, and we did not use them as criteria for study inclusion.

Electronic searches

We searched the following databases using the search terms detailed in Appendix 2: MEDLINE (OVID; 1946 to 25 January 2023), Embase (OVID; 1947 to 25 January 2023), Biosis previews (Web of Science; 1926 to 25 January 2023), CAB Abstracts (Web of Science; 1910 to 25 January 2023), and Cochrane Central Register of Controlled Trials (CENTRAL) (Issue 1 of 12, January 2023). There were no date or language limits placed on the search.

We also searched ClinicalTrials.gov and the WHO ICTRP for trials in progress, on 25 January 2023 using the terms shown in Appendix 2.

Searching other resources

We asked experts from the WHO Clinical Management and Infection Prevention and Control guideline development group for details of potentially relevant studies or ongoing trials to be considered for inclusion in the review.

Selection of studies

Three review authors (Tilly Fox, Rebecca Kuehn, Naveena Princy) independently screened the titles and abstracts of the search results for potentially relevant articles using the inclusion criteria. If there was any doubt over the potential for inclusion of a study, it was included for full‐text screening.

Two review authors (TF, RK) independently reviewed the full text of studies identified through title and abstract screening; any disagreements were resolved by discussion with all authors.

Data extraction and management

Two review authors (TF, RK) independently extracted data using a pre‐piloted standardized data extraction form. We contacted the study authors to obtain missing data if applicable. At each step of data extraction, we resolved any discrepancies through discussion between all review authors.

We extracted the following information.

• General information: author, title, publication date, country, date(s) of study, funding details, conflict of interest statement

• Study characteristics: study setting (inpatient or outpatient); study design; dates of recruitment; eligibility criteria; length of follow‐up; loss to follow‐up; and adherence to assigned treatment.

• Participant characteristics: number of participants (recruited, allocated, and evaluated); source of participants; age; sex; disease severity; vaccination status; concurrent treatments; pregnancy; and comorbidities (e.g. skin disease, immunosuppression).

• Interventions: type; route of administration; dosage; timing; frequency; duration of treatment; and duration of follow‐up.

• Control: type (placebo/active treatment); route of administration; dosage; timing; frequency; duration of treatment; and duration of follow‐up.

• Outcomes: data on the prespecified outcomes in both the intervention and control arms as follows: for dichotomous outcomes, number of events and participants; for continuous outcomes, mean, standard deviation (SD), and total number of participants; and for time‐to‐event outcomes, hazard ratios (HRs).

Assessment of risk of bias in included studies

Two review authors (TF, RK) planned to independently assess the risk of bias using the Cochrane risk of bias 2 tool (RoB 2) (Cochrane 2022; Higgins 2022; Sterne 2019). Any disagreements would be settled by a third review author. We planned to justify judgements made in the risk of bias tables. The effect of interest was to be the effect of assignment to the intervention at baseline, regardless of whether the interventions were received as intended (the ‘intention‐to‐treat effect’). We intended to manage the assessments using the RoB 2 Excel tool for randomized trials and assess risk of bias for the critically important outcomes included in the analyses. We planned to use the following domains to assess bias:

• bias arising from the randomization process;

• bias due to deviations from intended interventions;

• bias due to missing outcome data;

• bias in measurement of the outcome; and

• bias in selection of the reported result.

For trials that had randomized clusters, we planned to assess an additional component, namely bias arising from identification or recruitment of individual participants within clusters.

We planned to answer signalling questions as either yes; probably yes; probably no; no; or no information. We would then use these to determine the overall risk of bias for each domain (high, some concerns, low) and later, the overall risk of bias for each primary outcome from the included studies (high, some concerns, low). We planned to judge study outcomes to have an overall low risk of bias if all domains were at low risk, some concerns if any domain had some concerns, and high if we assessed any domain to be at high risk of bias.

Measures of treatment effect

For dichotomous outcomes (serious adverse events, development of disease‐related complications, admission to hospital, detection of virus, mortality, scarring, adverse maternal outcomes, adverse perinatal outcomes, and mental health outcomes), we planned to use the risk ratio (RR) with the corresponding 95% interval (CI) as the effect measure.

For continuous outcomes (pain, time to healing of all skin lesions, clinical disease progression, and duration of illness) when outcomes were measured in the same way between trials, we planned to use mean differences (MD) with the corresponding 95% interval (CI) as the effect measure. We planned to use the standardized mean difference (SMD) to combine trials that measured the same outcome with different methods.

For time‐to‐event data we planned to use the hazard ratio (HR) with the corresponding 95% interval (CI).

Unit of analysis issues

The unit of analysis for this review was planned to be the individual randomized participant. If multi‐arm trials were identified, to enable individual pairwise comparisons, we planned to either select relevant arms for inclusion in our analyses, or if more than two arms were relevant to the review, we planned to combine intervention arms to allow a single comparison if appropriate (e.g. study arms with different doses of the same therapeutic). If it would not be reasonable to combine intervention groups, we planned to split the 'shared' comparator group to avoid double‐counting of participants.

We did not anticipate that any cluster‐randomized studies would meet the inclusion criteria for this review. However, if any studies that met the inclusion criteria were identified, we planned to undertake analyses at the individual level while accounting for the clustering in the data.

Dealing with missing data

We planned to contact study authors to obtain missing study characteristics, missing outcomes, missing summary data, and missing individual data.

We planned to assess the risk of reporting bias due to missing studies and missing outcomes as described in the Assessment of reporting biases section.

If we were unable to obtain missing summary data, we planned to calculate or estimate the required data from other reported statistics using the formulae specified in the Cochrane Handbook for Systematic Reviews of Interventions.

If we were unable to obtain missing individual data, we planned to assess the risk of bias using the RoB 2 tool (Higgins 2022; Sterne 2019). In the first instance, we planned to conduct a complete‐case analysis, and we could have performed analyses to investigate the impact of missing data. For example, we could have varied the event rate within missing individuals from intervention and control groups within plausible limits as suggested by clinical experts, or we could have excluded studies thought to be at risk of bias due to missing data (rated as some concerns or high risk of bias in the 'Bias due to missing outcome data' for RoB2) from our meta‐analyses.

Assessment of heterogeneity

We planned to present results of the included studies in forest plots, which we would have inspected visually to assess heterogeneity (that is, non‐overlapping CIs generally signify statistical heterogeneity). We also planned to use the Chi² test with a P value of less than 0.1 to indicate statistical heterogeneity. Heterogeneity would have been quantified using the I² statistic, which describes the percentage of the variability in effect estimates that is due to heterogeneity rather than sampling error. We would have interpreted this statistic using the following guidance from the Cochrane Handbook:

• 0% to 40%: might not be important;

• 30% to 60%: may represent moderate heterogeneity*;

• 50% to 90%: may represent substantial heterogeneity*;

• 75% to 100%: considerable heterogeneity*.

*The importance of the observed value of I² depends on (1) magnitude and direction of effects, and (2) strength of evidence for heterogeneity (e.g. P value from the Chi² test, or a confidence interval for I²: uncertainty in the value of I² is substantial when the number of studies is small).

Assessment of reporting biases

We searched for ongoing trials that met our eligibility criteria and classified them as ‘ongoing’ until their publication.

If we had included 10 studies in a meta‐analysis, we planned to explore the possibility of small study biases (a tendency for estimates of the intervention effect to be more beneficial in smaller studies) for the primary outcomes using funnel plots. In the case of asymmetry, we planned to consider various explanations such as publication bias, poor study design and the effect of study size.

Data synthesis

We planned to perform all meta‐analyses using random‐effects models. Where performing a meta‐analysis would be considered to be inappropriate due to important clinical or methodological heterogeneity, or if study results differed to the extent that combining them in a pooled analysis would not make sense, we planned to summarize data using other methods outlined in the Cochrane Handbook (McKenzie 2022).

Subgroup analysis and investigation of heterogeneity

We planned to conduct subgroup analysis to explore the following sources of heterogeneity as determined by the WHO Clinical Management and Infection Prevention and Control GDG.

• Disease severity – severe disease versus non‐severe disease. We planned to use the Centre for Disease Control guidance for a severe case: when a patient has conditions such as hemorrhagic disease; a large number of lesions such that they are confluent; necrotic lesions; severe lymphadenopathy that can be necrotizing or obstructing (such as in airways); involvement of multiple organ systems and associated comorbidities (for example, pulmonary involvement with nodular lesions; sepsis; encephalitis; myocarditis; ocular or periorbital infections); or other conditions requiring hospitalization (CDC 2022b). The therapeutic effect of an intervention may differ according to disease severity.

• Days from symptom onset to therapeutic administration – less than seven days versus seven or more days. The effect of an intervention may differ depending on time to therapeutic administration.

We also intend to explore the following subgroups of absolute effects.

• Immunosuppression – immunosuppressed versus non‐immunosuppressed. Immunosuppression for the purposes of this review is defined as a participant having any comorbid condition or taking any medication associated with a reduction of the activation or efficacy of the immune system, as reported by study authors. The therapeutic effect of an intervention may differ according to level of participant immune function.

• Age – less than 60 years old versus 60 years old or more. The therapeutic effect of an intervention may differ due to altered physiological changes in older persons compared to younger persons.

• Skin disease – pre‐existing skin disease versus no pre‐existing skin disease. Pre‐existing skin disease is defined as any skin condition diagnosed in a participant before their diagnosis with mpox, as reported by study authors. Persons with pre‐existing skin disease may modify the effect of a therapeutic, including for the outcomes of time to healing of skin lesions, pain, and scarring, compared to persons with no pre‐existing skin disease.

Sensitivity analysis

We planned to perform sensitivity analyses to investigate the impact of missing data. For example, we may have assessed varying the event rate within missing patients from intervention and control groups within plausible limits, or we may exclude studies thought to be at high risk of attrition bias from our meta‐analyses.

Summary of findings and assessment of the certainty of the evidence

We planned to present the main results of the review in summary of findings tables including our rating of the certainty of evidence based on the GRADE approach. An example summary of findings table is displayed in Appendix 3. We planned to follow current GRADE guidance as recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Schünemann 2022).

Two review authors (TF, RK) planned to assess the certainty of the evidence, considering the domains of risk of bias, inconsistency, imprecision, indirectness, and publication bias.

We planned to construct one summary of findings table for each therapeutic comparison. The summary of findings tables were planned to include all critical outcomes, outlined below.

• Serious adverse events (up to 28 days)

• Development of disease‐related complications (up to 28 days or the longest period as reported by study authors)

• Admission to hospital for non‐hospitalized participants (up to 28 days)

• Pain, as judged by any visual or numerical pain scale (up to 28 days)

• Level of virus detected in clinical samples (as an indicator of infectiousness) (up to 28 days)

• Time to healing of all skin lesions (up to 28 days, or the longest period as reported by study authors)

• Mortality (up to 28 days)