Arts LPJ, et al. J Med Internet Res 2020.
BACKGROUND: Randomized controlled trials (RCTs) often provide accurate estimates of the internal validity of an intervention but lack information on external validity (generalizability). We conducted an RCT on the effectiveness of a self-management intervention among patients with lymphoma in a population-based setting.
OBJECTIVE: The objectives of the current study were to describe the proportion of RCT participants compared to all patients invited to participate, and compare sociodemographic and clinical characteristics of RCT participants with all respondents, all patients invited to participate, and all patients selected from the Netherlands Cancer Registry (NCR) to determine the reach of the intervention. An additional objective was to assess differences on RCT outcome variables between RCT and paper respondents.
METHODS: Patients with lymphoma or chronic lymphocytic leukemia ≥18 years old at diagnosis from 13 hospitals in the Netherlands were selected from the population-based NCR, which routinely collects data on sociodemographic and clinical characteristics. Eligible patients were invited to participate in an RCT and complete a questionnaire. Web-based completion determined RCT enrollment, whereas paper respondents were followed observationally.
RESULTS: A total of 1193 patients were selected from the NCR, 892 (74.77%) of whom were invited to participate in the trial by their hematologist after verifying eligibility. Among those invited, 25.4% (227/892) completed the web-based questionnaire and were enrolled in the RCT. The RCT participants were younger and there was a higher proportion of men than nonparticipants (P<.001). In addition, 25.7% (229/892) of those invited opted to participate in the paper-based observational follow-up study. Compared with paper respondents, RCT participants were younger (P<.001), with a higher proportion of men (P=.002), and had higher education levels (P=.02). RCT participants more often wanted to receive all available information on their disease (P<.001), whereas paper respondents reported higher levels of emotional distress (P=.009).
CONCLUSIONS: From a population-based sample of eligible patients, the participation rate in the RCT was approximately 25%. RCT participants may not be representative of the target population because of different sociodemographic and clinical characteristics. Since RCT participants represent a minority of the target population, RCT results should be interpreted with caution as patients in the RCT may be those least in need of a self-management intervention.
TRIAL REGISTRATION: Netherlands Trial Register NTR5953; httpss://www.trialregister.nl/trial/5790.