Cluster randomised controlled trial dataset: Medical oncologist and cancer patient consultations and associated clinical data

Brief description

This dataset relates to a cluster randomised controlled trial in which the intervention was an information and decision making resource: the Australian Cancer Trials (ACT) web site (www.australiancancertrials.gov.au). The cluster randomised controlled trial investigated the impact of the intervention on doctor-patient discussions about clinical trial awareness and participation. The ACT website is intended for patients, doctors and carers, and contains information about cancer trials, including clinical trials that are currently recruiting participants within Australia. The website also contains decision making support tools, in order to help patients decide, in consultation with their specialist, whether to join a clinical trial. Information on the ACT website is drawn from two cancer trial registries, the Australian New Zealand Clinical Trials Registry (ANZCTR) and ClinicalTrials.gov.

The primary goal of the trial was to identify whether doctors and patients who had access to the intervention (website) discussed clinical trials more than the control group, who did not have access. Secondary outcomes included clinical trial recruitment, patient knowledge about clinical trials, and whether the intervention had an effect on the patient’s ability to make a decision to join a clinical trial.

A cluster sample of thirty medical oncologists and 493 patients were recruited from 30 metropolitan and rural sites in New South Wales and Victoria, Australia. A baseline questionnaire was used to collect patient demographic and other clinical information (including cancer type and stage). This data was recorded in a Microsoft Access database. Consultations between medical oncologists and their patients were audio-recorded and transcribed into text files, and linked to demographic and clinical information. A follow-up questionnaire was administered, measuring patient knowledge of clinical trials and patient decision conflict about whether to join a clinical trial. Transcripts were de-identified, and coded in Nvivo according to a pre-defined coding schema.

The dataset comprises approximately 500 audio files (Windows Media Audio format), 500 associated de-identified interview transcripts (MS Word), an Access database of clinical data, coded interview data (MS Excel), and two master spreadsheets (MS Excel). Coded interview and questionnaire responses, transcript ID numbers, doctor ID numbers, patient ID numbers and associated clinical data were merged into two master spreadsheets by biostatistician Dr Kevin McGeechan. For further information regarding the project, methodology and findings please refer to the publications below.

Project contributors included Professor Alexandra Barratt; Professor David Currow; Professor Martin Tattersall; Professor Phyllis Butow; Associate Professor Lisa Askie; Dr Kevin McGeechan; Ms Sally Crossing (Chair of Cancer Voices NSW) and Dr Rachel Dear.