Characteristics and outcomes of patients with advanced gastric cancer who declined to participate in a randomized clinical chemotherapy trial.

PURPOSE: There is insufficient data to verify whether participation in clinical trials in itself can lead to better clinical outcomes. We have analyzed the characteristics and outcomes of patients who declined to participate in a randomized trial in comparison with those who participated in the trial. PATIENTS AND METHODS: A randomized trial for naive advanced gastric cancer was offered to 286 patients. The trial investigated the superiority of irinotecan plus cisplatin and the noninferiority of S-1 compared with continuous fluorouracil infusion. We retrospectively reviewed the characteristics and outcomes for both participants and nonparticipants in this trial.
RESULTS: Of the 286 patients, 98 (34%) declined to participate in the trial. The rate of declining was significantly higher among younger patients (P = .003), and it varied significantly between attending physicians (range, 23% to 58%; P = .004). There were no other significant correlations between rate of declining and patient characteristics. No significant differences were observed in the clinical outcomes between the participants and nonparticipants, for whom the median survival times were 367 versus 347 days, respectively. The hazard ratio for overall survival, adjusted for other confounding variables, was 1.21 (95% CI, 0.91 to 1.60). No interaction was observed between participation and the various regimens.
CONCLUSION: There was no difference in clinical outcomes between participants and nonparticipants. However, the patient's age and the doctor-patient relationship may have an effect on patient accrual to randomized trials.

RCT Entities:   Population Interventions Outcomes