Types of studies

Randomised controlled trials (RCTs).  

Types of participants

We included adults and children (any age) diagnosed with mumps by positive signs and symptoms or with laboratory evidence (isolation of the mumps virus, detection of viral nucleic acid, or serodiagnosis). The diagnosis of mumps was made on the basis of TCM diagnostic criteria (Jiang 1994), issued by the State Administration of Traditional Chinese Medicine standards. TCM diagnostic criteria for mumps include a history of exposure to mumps and symptoms beginning with fever and then swelling of the parotid glands, which may be unilateral (one side) or bilateral (both sides). The parotid glands are located in front of the ear, indistinctly outlined, elastic to palpation, painful and tender, with swelling of the parotid gland orifices. The white blood cell count is normal or slightly low and the lymphocyte count is slightly increased.

We also included mumps cases with complications such as epididymo‐orchitis, meningitis, encephalitis, pancreatitis, oophoritis, mastitis, spontaneous miscarriage, arthritis, nephritis, thyroiditis, myocarditis, hepatitis, acalculous cholecystitis, kerato‐uveitis, haemophagocytic syndrome, thrombocytopenia, and diabetes mellitus. We established the diagnostic criteria at the beginning of the trial and any changes in the diagnostic criteria that might result in variability in the clinical features of the participants included and the result obtained. We considered, documented, and explored these differences in a sensitivity analysis. We excluded patients concurrently suffering from other infections (for example, common cold, rotavirus enteritis, or bacterial infection).

Types of interventions

1. Chinese medicinal herbs with or without other routine treatments compared with placebo, no treatment, or other routine care.

2. Chinese medicinal herbs administered by any route.

Types of outcome measures

Electronic searches

For this update we searched the Cochrane Central Register of Controlled Trials (CENTRAL 2015, Issue 1) (accessed 4 February 2015), which includes the Cochrane Acute Respiratory Infections Group's Specialised Register, MEDLINE (March 2012 to January week 4, 2015), EMBASE (March 2012 to February 2015), CINAHL (March 2012 to February 2015), AMED (2012 to April 2014), the Chinese Biomedical Database (CBM) (1980 to February 2015), China National Knowledge Infrastructure (CNKI) (1979 to February 2015), and VIP Information (1989 to February 2015).

Previously we searched the Cochrane Central Register of Controlled Trials (CENTRAL 2012, Issue 4) (accessed 26 April 2012), which includes the Cochrane Acute Respiratory Infections Group's Specialised Register, MEDLINE (1948 to April week 3, 2012), EMBASE (1974 to April 2012), CINAHL (1981 to April 2012), AMED (1985 to April 2012), the Chinese Biomedical Database (CBM) (1980 to May 2012), China National Knowledge Infrastructure (CNKI) (1979 to May 2012), and VIP Information (1989 to May 2012).

We used the search strategy described in Appendix 1 to search MEDLINE and CENTRAL. We did not use a filter to identify RCTs as there were too few results. We adapted the search strategy to search EMBASE (Appendix 2), CINAHL (Appendix 3), AMED (Appendix 4), VIP (Appendix 5), CNKI (Appendix 6), and CBM (Appendix 7). We imposed no language or publication restrictions.

Searching other resources

We searched relevant reference lists of all relevant trials, reviews, conference proceedings, and journals. We contacted organisations (including the World Health Organization (WHO)), experts working in the field, and medical herb manufacturers to obtain information about unpublished or incomplete trials, confidential reports, and raw data from published trials. In addition, we searched the trials registries for completed and ongoing trials (last searched 4 February 2015): World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) search portal, ClinicalTrials.gov, and the Chinese Clinical Trial Registry.

Selection of studies

Two review authors (MS, YW) independently conducted the initial review, applied the inclusion criteria, and reviewed titles and abstracts from the search results. They identified the authenticity of each reported RCT by telephone interviews with the trial authors. We resolved disagreements by discussion with a third review author (YQZ). We excluded trials that were not counted as true RCTs and the reasons are listed in the Characteristics of excluded studies table. There was no inconsistency in the selection of studies.

Data extraction and management

Two review authors (MS, YW) independently conducted data extraction using a standard data extraction sheet. We resolved disagreements through consultation with a third review author (YQZ). We obtained and included in the review any missing data by contacting the trial author. One review author (YW) independently reviewed all data extracted from the original articles to confirm the quality of methods and analysis used. We included the following items on our data extraction sheet.

1. General information: study ID, author, title, published or unpublished, reference or source, contact address, country, language of publication, publication year, duplicate publications, funding, and setting.

2. Trial characteristics: design, method of randomisation, allocation concealment, blinding (participants, people administering treatment, outcome assessors), and duration of follow‐up.

3. Intervention(s): dose, route and timing, names of single medicine or compounds (including any possible specific data of the essential herbs in a compound), and use of placebo.

4. Participants: age (children/adults), total number and number in comparison groups, baseline characteristics, similarity of groups at baseline (including any co‐morbidity), diagnostic criteria, exclusion criteria, assessment of compliance, withdrawals, losses to follow‐up (reasons/description), and subgroups.

5. Outcomes: outcomes specified above, any other outcomes assessed, and quality of reporting of outcomes.

We extracted the number of events and total number in each group for binary outcomes. We extracted the mean standard deviation and sample size of each group for continuous outcomes.

Assessment of risk of bias in included studies

Two review authors (MS, YW) independently assessed the methodological quality of studies to be included. We resolved disagreements by discussion with another review author (YQZ). The specific methodological quality items to be assessed were randomisation, allocation concealment, blinding, treatment adherence, percentage of participation, and comparability of groups on baseline characteristics. According to empirical evidence (Higgins 2011; Juni 2001; Kjaergard 2001; Moher 2001; Schulz 1995; Wu 2007), we assessed the methodological quality of each trial as described by Kjaergard 2001, Wu 2007, and Higgins 2011.

1. Generation of allocation sequence: adequate (computer‐generated random numbers, random number table, or similar); inadequate (other methods or not described); or unclear.

2. Allocation concealment: adequate (central allocation, sealed envelopes, or similar); inadequate (not described or open random allocation schedule or similar); or unclear.

3. Blinding: low risk (the outcome assessors were blinded); high risk (no blinding was performed); unclear.

4. Incomplete outcome data (checking for possible attrition bias through withdrawals, drop‐outs, protocol deviations): low risk (few drop‐outs/losses to follow‐up are noted and an intention‐to‐treat (ITT) analysis is possible); high risk (the rate of exclusion was at least 20%, or wide differences in exclusions between groups whatever ITT analysis was performed); or unclear.

5. Selective reporting bias: low risk (all of the trial's prespecified outcomes and all expected outcomes of interest to the review have been reported clearly); high risk (not all the trial's prespecified outcomes have been reported; one or more reported primary outcomes were not prespecified; outcomes of interest are reported incompletely; study fails to include results of a key outcome that would have been expected to have been reported); or unclear.

6. Other sources of bias: we assessed whether each trial was free of other problems that could put it at risk of bias.

We planned to integrate this assessment into the interpretation of results by carrying out sensitivity analyses and excluding poor quality studies.

Measures of treatment effect

In future we will express dichotomous outcomes as a risk ratio (RR) with 95% confidence intervals (CIs). We will express continuous variables as a mean difference (MD) if reported on the same scale or as a standardised mean difference (SMD) if reported using different continuous scales, with 95% CIs.

Unit of analysis issues

We will analyse data using Review Manager (RevMan 2014). For binary data we will compare outcome measures using a risk ratio (RR). For continuous data, we will use the MD. If continuous data are reported using geometric means, we plan to combine the findings on a log scale and report on the original scale.

Dealing with missing data

If a marked drop‐out of participants is noted, we will undertake an intention‐to‐treat (ITT) analysis. Where data are missing, we will try to contact the trial authors. We will analyse the available data and discuss the impact of the missing data on our results.

Assessment of heterogeneity

We will measure heterogeneity by using the I² statistic (Higgins 2011), which will assess the significance of any differences between trials (P value < 0.05). Where we do not detect heterogeneity, we will perform a fixed‐effect meta‐analysis. Where substantial heterogeneity (I² statistic above 50%) is present, we will consider possible interpretations for this and, where applicable, use a random‐effects model and consider not combining the results, but instead present a descriptive analysis.

Assessment of reporting biases

We will evaluate reporting biases using funnel plots in future updates where we have sufficient trials and consider causes for asymmetry if it is marked.

Data synthesis

We will address the meta‐analysis method, including whether to use a fixed‐effect or random‐effects model. If we do not undertake a meta‐analysis, we will describe systematic approaches to synthesising the findings from multiple studies (Higgins 2011).

Subgroup analysis and investigation of heterogeneity

If we identify a sufficient number of studies, we will perform subgroup analyses to explore effect size differences by group and heterogeneity found in the primary analyses. We will consider subgroup analyses for the following areas.

1. Children/adults.

2. Different formulations.

3. Different administration routes (external, oral, or intravenous).

4. Doses (low and high, based on data).

5. Type of control ‐ placebo/standard clinical care.

Sensitivity analysis

We plan to perform a sensitivity analysis by removing single trials to investigate the extent to which they contribute to heterogeneity, paying close attention to baseline characteristics.