Table 2 Quantitative data sources

 1. To assess effectiveness of the intervention on the trial outcomes within sites, data from electronic medical records were collected using a validated extraction tool at baseline, end of trial and end of post-trial phase as part of the TORPEDO trial.

 2. To assess the support requirements provided by the project staff, support time was calculated based on contact time logged by both the technical helpdesk and the research team. Support time varied depending on availability and number of staff, staff requests and technology-related troubleshooting.

 3. To assess acceptability and fidelity of the intervention, staff were invited to complete three surveys toward the end of cRCT:

  i. An end of study mail survey for general practitioners who were part of the intervention sites was developed by the research team to assess acceptability and fidelity of the intervention. The questions were focused on satisfaction with the intervention components, recommendations of evidence-based guidelines, the intervention’s effect on the quality of care, and frequency of use. In addition, there were questions about the practice characteristics and personal use of information technology. It was reviewed for content validity by the PWG. Although we had intended to look at usage analytics to look at intervention fidelity, due to technical problems with the software database, we were unable to generate accurate usage logs and therefore had to rely on staff self-report. Nine GPs within the six cases completed the survey, and 23 GPs from 15 non-case intervention sites completed the survey. The findings from this survey have been published and used in this paper as complementary data [20].

  ii. Drawing on the NPT sub-domain of ‘collective action’ in which team members work together to incorporate innovation into practice, a team climate inventory (TCI)* survey was administered. This is a 44-item questionnaire which assesses team vision (11 items), participative safety (12 items), task orientation (7 items), support for innovation (8 items) and social desirability (6 items) with each item rated on a 5-point Likert scale.

  iii. In order to assess if job satisfaction may be an influential factor in driving outcomes, the Warr-Cook Wall Job Satisfaction survey* was administered. Based on previous work, this 10-item questionnaire assesses physical work conditions, income, amount of responsibility given, freedom in the job, variety, work colleagues, opportunity to use abilities, recognition and hours of work. It was adapted for use with general practices and ACCHSs using a 7-point Likert scale.

* The TCI and job satisfaction surveys were either distributed together by mail or in person during the end of trial data collection period. Sites were followed up 1 week later by telephone on expected completion timeframe. For surveys not received within the month, a second attempt to follow-up was made. The TCI and job satisfaction surveys were completed by 68 health professionals from the six cases, 113 health professionals from 18 non-case intervention sites and 65 health professionals from 15 control sites.