Balancing data quality and participant burden: a comparative analysis of abbreviated vs extended symptom diaries in the CanTreatCOVID trial

Abstract Background Symptom diaries are widely used in acute respiratory infection trials to capture patient-reported symptom severity and recovery. Longer questionnaires may provide a more complete clinical picture but can increase participant burden and reduce adherence. Evidence directly comparing long and short formats within the same trial is limited. Objective To compare adherence, symptom trajectories, agreement between recovery measures, and predictive performance for recovery-related outcomes between a short and long symptom diary in an outpatient SARS-CoV-2 trial. Methods This secondary analysis of the CanTreatCOVID trial compared a 9-item Abbreviated Diary and a 34-item FLU-PRO Plus Diary over 14 days in non-hospitalized participants with confirmed SARS-CoV-2 infection. Outcomes included diary initiation, completion, completion rate, compliance, symptom trajectories, agreement between recovery outcomes, and predictive performance. Analyses used logistic regression, generalized estimating equations, survival models, and predictive modelling. Results Of 712 participants, 638 used the Abbreviated Diary and 74 the FLU-PRO Plus Diary. Baseline characteristics were similar between groups. Diary type was not significantly associated with diary initiation or full 14-day compliance, whereas treatment assignment was associated with higher adherence ( p < 0.0001). Completion rates were slightly higher in the Abbreviated group (88.3% vs. 84. 8%), but differences were not statistically significant. Agreement between “feeling recovered” and “return to usual health” was strong to excellent (κ = 0.83–0.88), while agreement with “return to usual activities” was moderate (κ = 0.52–0.65). Predictive models performed well for both diaries (AUCs 0.87–0.94), with only marginal gains from including the extended FLU-PRO Plus items. Conclusion In this outpatient SARS-CoV-2 trial, no statistically significant differences in adherence, symptom trajectories, or predictive performance were detected between abbreviated and extended symptom diaries. These analyses were not designed or powered as equivalence tests; the absence of statistically significant differences should not be interpreted as evidence of equivalence between diary formats. Future research should extend follow-up beyond 14 days to capture longer-term patterns and test diary performance in more diverse and digitally underserved populations.