BACKGROUND
High-frequent patient-reported outcome (PRO) assessments are used to measure patients’ symptoms after surgery for surgical research; however, quality of those longitudinal PRO data has seldom been discussed.
OBJECTIVE
To describe errors, to identify factors influencing the data quality, and to profile error trajectories of data longitudinally collected via paper-and-pencil (P&P) or web-based-assessment (ePRO) after thoracic surgery.
METHODS
We extracted longitudinal PRO data from two prospective clinical studies. PROs were assessed by the MD Anderson Symptom Inventory Lung Cancer Module and single-item Quality of Life Scale before surgery and then daily after surgery until discharge or up to 14 days of hospitalization. Patient compliance and data error were identified and compared between P&P and ePRO. Generalized estimating equations models and two-piecewise models were used to describe trajectories of error incidence over time and to identify the risk factors.
RESULTS
Among 629 patients with at least 2 PRO assessments, 440 completed 3347 P&P assessments and 189 completed 1291 ePRO assessments. In total, 49.44% of patients had at least 1 error, including 1) missing items (64.69%), 2) modifications without signatures (27.99%), 3) selection of multiple options (3.02%), 4) missing patient signatures (2.54%), 5) missing researcher signatures (1.45%) and 6) missing completion dates (0.3%). ePRO patients had fewer errors than P&P patients (30.16% vs. 57.73%, p <0.0001). Compared with ePRO patients, those using P&P were older, less educated and sicker. Common risk factors of having errors were with a lower education level (P&P, OR=1.39, 95%CL=1.20-1.62, p<.0001; ePRO, OR=1.82, 95%CI=1.22-2.72, p=0.0032), treated in a provincial hospital (P&P, OR=3.34, 95%CI=2.10-5.33, p<.0001; ePRO, OR=4.73, 95%CI=2.18-10.25, p<.0001) and with severe disease (P&P, OR=1.63, 95%CI=1.33-1.99, p<.0001; ePRO, OR=2.70, 95%CI=1.53-4.75, p=0.0006). Errors peaked on postoperative day (POD) 1 for P&P, and on POD 2 for ePRO.
CONCLUSIONS
ePRO might be superior to P&P in terms of data quality. However, sampling bias needs to be considered for studies using longitudinal PROs as major outcomes.