Quality of life after patient-initiated vs physician-initiated response to symptom monitoring: the SYMPRO-Lung trial

Abstract

Abstract Background Previous studies using patient-reported outcomes measures (PROMs) to monitor symptoms during and after (lung) cancer treatment used alerts that were sent to the health-care provider, although an approach in which patients receive alerts could be more clinically feasible. The primary aim of this study was to compare the effect of weekly PROM symptom monitoring via a reactive approach (patient receives alert) or active approach (health-care provider receives alert) with care as usual on health-related quality of life (HRQOL) at 15 weeks after start of treatment in lung cancer patients. Methods The SYMPRO–Lung trial is a multicenter randomized controlled trial using a stepped wedge design. Stage I-IV lung cancer patients in the reactive and active groups reported PROM symptoms weekly, which were linked to a common alerting algorithm. HRQOL was measured by the EORTC QLQ-C30 at baseline and after 15 weeks. Linear regression analyses and effect size estimates were used to assess mean QOL–C30 change scores between groups, accounting for confounding. Results A total of 515 patients were included (160 active group, 89 reactive group, 266 control group). No differences in HRQOL were observed between the reactive and active group (summary score: unstandardized beta [B] = 0.51, 95% confidence interval [CI] = -3.22 to 4.24, Cohen d effect size [ES] = 0.06; physical functioning: B = 0.25, 95% CI = -5.15 to 4.64, ES = 0.02). The combined intervention groups had statistically and clinically significantly better mean change scores on the summary score (B = 4.85, 95% CI = 1.96 to 7.73, ES = 0.57) and physical functioning (B = 7.00, 95% CI = 2.90 to 11.09, ES = 0.71) compared with the control group. Conclusions Weekly PRO symptom monitoring statistically and clinically significantly improves HRQOL in lung cancer patients. The logistically less intensive, reactive approach may be a better fit for implementation.