Methods

\label{methods}
The study will be a two-arm, parallel, open, randomised controlled trial with a 1:1 allocation ratio.

Eligibility

\label{eligibility}

Inclusion criteria

A Cochrane Schizophrenia systematic review published in the Cochrane Library for which there is a relevant specific Wikipedia page, and in which there is at least one Summary of Findings table. The Summary of Findings tables are largely created though GradePro \cite{GRADEpro_GRADEpro}, are incorporated into most Cochrane reviews and are a succinct summary of the key outcomes of the review.

Exclusion criteria

If the review does not have a Wikipedia page which is judged to have a title specifically relevant to the content of a Cochrane Schizophrenia review. For example, a general page such as ‘hypnotherapy’ may only briefly mention its subject matter’s relevance to schizophrenia/mental illness. This would be ineligible. However, should there be a page ‘hypnotherapy for schizophrenia’ then that would be eligible. If the Cochrane review has no Summary of Findings table it will be ineligible.

Randomisation

\label{randomisation}
Reviews will be stratified according to type of intervention (drug or other) and amount of access activity prior to baseline (low or high, according to median split) and allocated to the intervention or control arm using a computer-generated random number sequence. Allocation will be conducted using unique code numbers for each review, rather than review title, to avoid risk of selection bias.

Interventions

\label{interventions}

Experimental group

Reviews in the intervention group will have a referenced table/s inserted on to their relevant Wikipedia page. This will consist of a summary of the intervention’s efficacy using data from, but entirely re-writing, the Cochrane Summary of Findings table. The original table contains information that may not be relevant to most Wikipedia readers and is not in plain language. The inserted tables will be referenced (and hyperlinked) to both the review’s summary page (entirely free access) and the full review (variable levels of full access) on the Cochrane Library website. All reviews in the intervention group will have the referenced tables inserted onto the relevant Wikipedia page on the same day. This then provides a convenient consistent ‘start time’ (time zero) for all reviews in the trial, including those allocated to the control group.

Control intervention

The control group will not have a table or reference placed onto the relevant Wikipedia page.

Source of data

Google Analytics, originally called Urchin, before it was signed over to Google in March 2005, will be used as one data source for outcomes. Google Analytics, focusing on web-page activity, is mainly used by businesses to identify customers’ needs and how those needs are being met. We will use it to track activity on the Cochrane Summaries web pages. With the production of various data reports in real-time, this system can answer questions about who has visited, and when and where they had visited, a site, as well as how they ‘arrived’ at that site. Google Analytics is a good source of user-relevant data about websites accessible with the standard (free) account.
In addition we will use the routine data collected by John Wiley and Sons Ltd., on the activity relevant to the full Cochrane review. These data report full text downloads, and Altmetric scores; a measure attributed to scholarly articles based on the number of online mentions in various digital outputs including Twitter, Facebook, news outlets, blogs, and peer reviews.

Outcomes

\label{outcomes}
Outcomes will be measured at 12 months.
There are two outcome of primary interest:
  1. The number of visits to the free summary page (All page views)
  2. The number of full text downloads
Secondary outcomes will be divided into activity on the free to all summary page, and outcomes relating to activity on the Cochrane Library’s full review. Both sources of data are routinely collected. The former is available using Google Analytics. The Cochrane Library is published by John Wiley and Sons Ltd. and data are routinely acquired by the company for monitoring purposes. The Cochrane office of John Wiley have agreed to supply relevant outcome data.

Statistical considerations

\label{statistical-considerations}

Power

The size of the study is constrained by limited funding and the number of reviews meeting eligibility criteria, of which we expect there to be around 100. However we can estimate the magnitude of the detectable between-group difference in the primary outcome. With 5% two-sided alpha and a sample size of 50 reviews per arm, an effect size of 0.57 standard deviations is detectable with 80% power.

Data analysis

We will compare characteristics of the intervention and control arms at baseline using appropriate descriptive statistics. The primary between-group comparison will analyse reviews as randomised, regardless of how long the Wikipage held the table, and will estimate the difference in mean number of visits per year, and difference in mean number of full text downloads per year and 95% confidence interval using analysis of covariance. This will be implemented using multivariable linear regression, and will include in the model baseline number of visits/downloads. Secondary outcomes will be analysed similarly.
As a secondary analysis of the primary outcome, we will conduct a pre-specified subgroup analysis to investigate whether any effect of the intervention differs according to baseline activity. We will do this by including an appropriate interaction term in the primary regression model.
We anticipate that the intervention will be implemented fully as planned, and that there will not be any missing primary outcome data. However in the event that either of these assumptions is untrue, we will consider sensitivity analyses to investigate the effect of the receiving the intervention as intended, and of imputing missing outcome data.

Ethics

\label{ethics}
This study will employ inanimate Cochrane systematic reviews as participants and collect routine data from electronic systems for outcomes. We have enquired of the local Ethics Committee and have been advised that ethical approval is not required.