Use and Reporting of Patient-Reported Outcomes in Trials of Palliative Radiotherapy: A Systematic Review

Abstract

Importance: Approximately 50% of all patients with cancer have an indication for radiotherapy, and approximately 50% of radiotherapy is delivered with palliative intent, with the aim of alleviating symptoms. Symptoms are best assessed by patient-reported outcomes (PROs), yet their reliable interpretation requires adequate reporting in publications.

Objective: To investigate the use and reporting of PROs in clinical trials of palliative radiotherapy.

Evidence review: This preregistered systematic review searched PubMed/Medline, EMBASE, and the Cochrane Center Register of Controlled Trials for clinical trials of palliative radiotherapy published from 1990 to 2020. Key eligibility criteria were palliative setting, palliative radiotherapy as treatment modality, and clinical trial design (per National Institutes of Health definition). Two authors independently assessed eligibility. Trial characteristics were extracted and standard of PRO reporting was assessed in adherence to the Consolidated Standards of Reporting Trials (CONSORT) PRO extension. The association of the year of publication with the use of PROs was assessed by logistic regression. Factors associated with higher CONSORT-PRO adherence were analyzed by multiple regression. This study is reported following the PRISMA guidelines.

Findings: Among 7377 records screened, 225 published clinical trials representing 24 281 patients were eligible. Of these, 45 trials (20%) used a PRO as a primary end point and 71 trials (31%) used a PRO as a secondary end point. The most prevalent PRO measures were the Numeric Rating Scale/Visual Analogue Scale (38 trials), European Organization for the Research and Treatment of Cancer Quality of Life Questionnaire C30 (32 trials), and trial-specific unvalidated measures (25 trials). A more recent year of publication was significantly associated with a higher chance of PROs as a secondary end point (odds ratio [OR], 1.04 [95% CI, 1.00-1.07]; P = .03) but not as primary end point. Adherence to CONSORT-PRO was poor or moderate for most items. Mean (SD) adherence to the extension adherence score was 46.2% (19.6%) for trials with PROs as primary end point and 31.8% (19.8%) for trials with PROs as a secondary end point. PROs as a primary end point (regression coefficient, 9.755 [95% CI, 2.270-17.240]; P = .01), brachytherapy as radiotherapy modality (regression coefficient, 16.795 [95% CI, 5.840-27.751]; P = .003), and larger sample size (regression coefficient, 0.028 [95% CI, 0.006-0.049]; P = .01) were significantly associated with better PRO reporting per extension adherence score.

Conclusions and relevance: In this systematic review of palliative radiotherapy trials, the use and reporting of PROs had room for improvement for future trials, preferably with PROs as a primary end point.

Figure 1.. Patient-Reported Outcomes (PRO) in Published Trials of Palliative Radiotherapy Over Time

Logistic regression analysis of the use of a PRO end point (1 indicates yes; 0, no) as dependent variable and year of publication as independent variable. A, Use of PROs as primary end points among 145 trials clearly stating their endpoint. B, Use of PROs as secondary end points among all 225 assessed trials (n = 225). Dots indicate data points; lines, mean; shading, 95% CI.

Figure 2.. Patient-Reported Outcome Measures (PROMs) Used in Published Trials of Palliative Radiotherapy (N = 116)

Multiple measures could be used per trial. CLAS indicates Cancer Linear Analogue Scale; EORTC, European Organization for Research and Treatment of Cancer; FACT, Functional Assessment of Cancer Therapy; HADS, Hospital Anxiety and Depression Scale; IAEA, International Atomic Energy Agency; LCSS, Lung Cancer Symptom Scale; MRC, Medical Research Council; NRS/VAS, numeric rating scale/visual analogue scale; RSCL, Rotterdam Symptom Checklist.