Search

Purpose: Patient-reported outcomes (PROs) can promote patient-centered care, but previous research has documented interpretation challenges among clinicians and patients. We engaged stakeholders to improve formats for presenting individual-level PRO data (for patient monitoring) and group-level PRO data (for reporting comparative clinical studies)., Methods: In an iterative process, investigators partnered with stakeholder workgroups of clinicians and patients to address previously identified interpretation challenges. Candidate approaches were then tested in semi-structured, one-on-one interviews with cancer patients and clinicians. Interpretation issues addressed included conveying score meaning (i.e., what is good/bad) and directional inconsistency (whether higher scores are better/worse). An additional issue for individual-level PROs was highlighting potentially concerning scores and, for group-level PROs, identifying important between-group differences (clinical, statistical)., Results: One-on-one interviews in a purposive sample of clinicians (n = 40) and patients (n = 39) provided insights regarding approaches to address issues identified. For example, adding descriptive labels to the Y-axis (none, mild, moderate, severe) helps address directional inconsistency and aids interpretation of score meaning. Red circles around concerning data points or a threshold line indicating worse-than-normal scores indicate possibly concerning scores for individual-level PRO data. For group-level PRO data, patients and some clinicians are confused by confidence limits and clinical versus statistical significance, but almost all clinicians want p values displayed., Conclusions: Variations in interpretation accuracy demonstrate the importance of presenting PRO data in ways that promote understanding and use. In an iterative stakeholder-driven process, we developed improved PRO data presentation formats, which will be evaluated in further research across a large population of patients and clinicians.

Background: Patient-reported outcomes (PROs) promote patient-centered care by using PRO research results ("group-level data") to inform decision making and by monitoring individual patient's PROs ("individual-level data") to inform care. We investigated the interpretability of current PRO data presentation formats., Method: This cross-sectional mixed-methods study randomized purposively sampled cancer patients and clinicians to evaluate six group-data or four individual-data formats. A self-directed exercise assessed participants' interpretation accuracy and ratings of ease-of-understanding and usefulness (0 = least to 10 = most) of each format. Semi-structured qualitative interviews explored helpful and confusing format attributes., Results: We reached thematic saturation with 50 patients (44 % < college graduate) and 20 clinicians. For group-level data, patients rated simple line graphs highest for ease-of-understanding and usefulness (median 8.0; 33 % selected for easiest to understand/most useful) and clinicians rated simple line graphs highest for ease-of-understanding and usefulness (median 9.0, 8.5) but most often selected line graphs with confidence limits or norms (30 % for each format for easiest to understand/most useful). Qualitative results support that clinicians value confidence intervals, norms, and p values, but patients find them confusing. For individual-level data, both patients and clinicians rated line graphs highest for ease-of-understanding (median 8.0 patients, 8.5 clinicians) and usefulness (median 8.0, 9.0) and selected them as easiest to understand (50, 70 %) and most useful (62, 80 %). The qualitative interviews supported highlighting scores requiring clinical attention and providing reference values., Conclusions: This study has identified preferences and opportunities for improving on current formats for PRO presentation and will inform development of best practices for PRO presentation. Both patients and clinicians prefer line graphs across group-level data and individual-level data formats, but clinicians prefer greater detail (e.g., statistical details) for group-level data.