A31-06 Lessons Learned From Designing a Patient-Centered Smoking History Survey for Lung Cancer Screening

Rationale Accurate identification of lung cancer screening (LCS)-eligible individuals is often hindered by the need for detailed smoking history, especially pack-year calculation. Collecting smoking history directly from patients through surveys may facilitate determination of LCS eligibility and personalized risk assessments. Methods We developed two prototypes of smoking history surveys that differed mainly in their approach to pack-year determination (Table). The short survey asked about total duration of smoking and average daily cigarette consumption. The long survey collected detailed history for each distinct period of smoking, beginning with smoking initiation and incorporating changes over time by tailoring follow-up questions according to prior responses. Both surveys were designed to be accessed via a link sent through text message, allowing patients to respond on their smartphones. Usability testing of both prototypes and cognitive interviews using think-aloud methods were performed with patients potentially eligible for LCS. Interview transcripts were analyzed qualitatively to examine participants’ comprehension, response process, and preferences between the survey prototypes. Results We conducted cognitive interviews with 29 participants (median age 67); 13 reported current smoking, and 11 had a high school degree or less. Participants found the survey language clear. Many used major life events (e.g., career changes or childbirth) as reference point to recall their smoking habits. Calculating total duration and average cigarette consumption in the short survey was frequently challenging for patients and inaccurate. Despite general preference for brevity, many recommended the approach used in the long survey for three reasons: (1) it captured more complete and nuanced data (“I like this long one better than the short one. The short one was easier, but the long one was more detailed.”) (2) it minimized errors related to self-calculation (“I don’t know if you’re going to get all accurate information on that short one with the math.”) and (3) its detailed questions conveyed a sense of personalized care by reflecting changes in smoking habits over the lifespan (“I kind of liked the long survey a little better because it seemed a little bit more personal. The short one almost implies to me that you don’t really care.”) Conclusions Despite the brevity and widespread use of the questions from the short survey in clinical settings, LCS websites, and decision aids, patients perceived them as less accurate and less patient-centered. In contrast, patients endorsed the long survey approach for its ease of response, completeness, accuracy, and patient-centeredness over the shorter survey approach. This abstract is funded by: NCI