Differences in Cancer Screening Patterns Between Urban and Suburban Caregivers Accompanying Patients to Radiation Oncology

Abstract

Bias in efficacy reporting is defined as reporting that the treatment is beneficial based on secondary endpoints despite a statistically non-significant difference in primary endpoint. Bias in toxicity reporting is defined as not reporting toxicity findings in the abstract or discussion or results table. Logistic regression multivariate models were used to determine the predictors of biased reporting. Results: We identified 323 and 279 RCTs eligible for assessment of bias in reporting of efficacy and toxicity, respectively. Forty-two of 323 trials (13%) and 97 of 279 trials (35%) were judged to have bias in the reporting of efficacy and toxicity, respectively. Multivariate analysis showed that non-cooperative group trials were a significant predictor of biased reporting of efficacy (odds ratio (OR) 2.08, 95% confidence interval (CI) 1.04 to 4.17, P Z.04) while trials not listed in clinicaltrials.gov were a significant predictor of biased reporting of toxicity (OR 3.12, 95% CI 1.33-7.69, P Z.009). Industry sponsorship, trial setting (curative versus palliative), sample size, and journal impact factor were not predictors of bias in the reporting of efficacy and toxicity. There was no association between bias in the reporting of efficacy and toxicity. Conclusion: Bias in reporting of efficacy in radiation therapy trials is low and more common for non-cooperative group trials. Reporting of toxicity is poor, especially for trials not listed in clinicaltrials.gov.