PCN321 PRO Conceptual Models in LUNG Cancer: The VALUE of Multiple DATA Sources

Abstract

Conceptual models (CM) attempt to describe the patient experience of medical conditions/treatments using data from a variety of primary and secondary sources, including direct elicitation of information from patients themselves. Certain conditions, though not rare, pose challenges to patient recruitment; researchers might rely solely on literature to develop a CM. A case in point is lung cancer (LC). Recently, only two publications in the ISPOR database were devoted to CMs of LC, both based entirely on published literature. How do these literature-based models compare to the results of studies in which patients provided data directly? PubMed was queried for LC CMs, resulting in 47 publications; 1 provided an additional model for comparison, it too was based on literature review. In addition, a search of IQVIA’s COA Accelerator Database uncovered 16 pCODR reports containing LC patient-level research, and a search of the FDA VOP reports found an LC report containing patient input. These were not included in any of the CMs and therefore served as an alternative basis for comparison. Across the data sources, there was general agreement that cough, dyspnea, pain, and fatigue are the symptoms of greatest concern. The burden of illness due to these symptoms was well characterized, impacting physical, social/emotional, and cognitive functioning. However, the CMs based upon only literature reviews may have overemphasized certain measurement concepts that, in fact, are not commonly reported by patients, such as constipation, sweats/flushing, and hemoptysis. This comparison of findings across data sources demonstrates the importance of multiple streams of information when developing a CM. Not surprisingly, direct patient input provided nuances into the patient experience highlighting what is most important and relevant. Finding patient reports in non-traditional sources, increasingly in social media, is becoming a critical step towards understanding the lived patient experience of hard-to-recruit conditions.