SOURCES OF BIAS IN PATIENT-REPORTED OUTCOME (PRO) DATA COLLECTION: EVIDENCE GAPS IN THEIR IMPACT ON DATA QUALITY
Author(s)
Kelly Kato, PhD1, Beatriz Antolin Fontes, PhD2, Erieta Bountouva, MSc3, Matthew D. Reaney, PhD4, Julie R. Bailey, MBA5, Katharine S. Gries, PharmD, PhD1, Alexis L. Oldfield, PhD1, Eva G. Katz, PhD1, Patricia S. Delong, MSc6, Lindsay Hughes, PhD5;
1Johnson & Johnson Innovative Medicine, Raritan, NJ, USA, 2IQVIA, Basel, Switzerland, 3IQVIA, Madrid, Spain, 4IQVIA, Reading, United Kingdom, 5IQVIA, New York, NY, USA, 6Johnson & Johnson Innovative Medicine, Horsham, PA, USA
1Johnson & Johnson Innovative Medicine, Raritan, NJ, USA, 2IQVIA, Basel, Switzerland, 3IQVIA, Madrid, Spain, 4IQVIA, Reading, United Kingdom, 5IQVIA, New York, NY, USA, 6Johnson & Johnson Innovative Medicine, Horsham, PA, USA
OBJECTIVES: Patient-reported outcomes (PROs) capture how patients feel and function in clinical trials. Robust, high-quality PRO data can inform health authority (HA) benefit-risk evaluation. However, PRO implementation decisions can inadvertently introduce bias, compromising data reliability, interpretability and HA consideration. This review aimed to identify operational factors that may bias PRO data, assess their impact on data quality, and summarize mitigation strategies.
METHODS: A targeted literature review was conducted across Embase, MEDLINE, and Cochrane databases, supplemented by grey literature and industry sources. Of 830 unique records screened, 31 publications were selected for extraction. Findings were synthesized into two themes: (1) biases associated with PRO planning and administration and (2) biases associated with site and site staff activities.
RESULTS: The most discussed bias contributors associated with PRO planning and administration were variability in administration setting (clinic vs. home), timing (pre- vs. post-treatment or within the site visit), method of PRO administration (paper vs. electronic), mode (self- vs. interviewer-administered), and participation burden. Site-level challenges, such as infrastructure limitations, and lack of standardized procedures for assisting participants and checking for missing data, further compromised PRO quality. While these biases are widely discussed, empirical evidence quantifying their impact is limited; observed effects were generally small, described qualitatively/anecdotally, and context-dependent. Proposed mitigation strategies include early integration of PRO planning in protocol development, standardized administration procedures, comprehensive site training, real-time monitoring, and transparent reporting of deviations. However, most strategies lack empirical validation.
CONCLUSIONS: Implementation decisions in PRO collection can influence data quality, although the magnitude and impact of these remain poorly defined. Future research should prioritize empirical evaluation of these biases and test whether mitigation strategies improve data completeness and reliability. Practical guidance balancing methodological rigor with operational feasibility is essential to support consistent implementation across diverse trial settings.
METHODS: A targeted literature review was conducted across Embase, MEDLINE, and Cochrane databases, supplemented by grey literature and industry sources. Of 830 unique records screened, 31 publications were selected for extraction. Findings were synthesized into two themes: (1) biases associated with PRO planning and administration and (2) biases associated with site and site staff activities.
RESULTS: The most discussed bias contributors associated with PRO planning and administration were variability in administration setting (clinic vs. home), timing (pre- vs. post-treatment or within the site visit), method of PRO administration (paper vs. electronic), mode (self- vs. interviewer-administered), and participation burden. Site-level challenges, such as infrastructure limitations, and lack of standardized procedures for assisting participants and checking for missing data, further compromised PRO quality. While these biases are widely discussed, empirical evidence quantifying their impact is limited; observed effects were generally small, described qualitatively/anecdotally, and context-dependent. Proposed mitigation strategies include early integration of PRO planning in protocol development, standardized administration procedures, comprehensive site training, real-time monitoring, and transparent reporting of deviations. However, most strategies lack empirical validation.
CONCLUSIONS: Implementation decisions in PRO collection can influence data quality, although the magnitude and impact of these remain poorly defined. Future research should prioritize empirical evaluation of these biases and test whether mitigation strategies improve data completeness and reliability. Practical guidance balancing methodological rigor with operational feasibility is essential to support consistent implementation across diverse trial settings.
Conference/Value in Health Info
2026-05, ISPOR 2026, Philadelphia, PA, USA
Value in Health, Volume 29, Issue S6
Code
PCR158
Topic
Patient-Centered Research
Topic Subcategory
Patient-reported Outcomes & Quality of Life Outcomes
Disease
No Additional Disease & Conditions/Specialized Treatment Areas