Clinician Perceptions on Using Decision Tools to Support Prediction-Based Shared Decision Making for Lung Cancer Screening

Abstract

Background: Considering a patient's full risk factor profile can promote personalized shared decision making (SDM). One way to accomplish this is through encounter tools that incorporate prediction models, but little is known about clinicians' perceptions of the feasibility of using these tools in practice. We examined how clinicians react to using one such encounter tool for personalizing SDM about lung cancer screening (LCS).

Design: We conducted a qualitative study based on field notes from academic detailing visits during a multisite quality improvement program. The detailer engaged one-on-one with 96 primary care clinicians across multiple Veterans Affairs sites (7 medical centers and 6 outlying clinics) to get feedback on 1) the rationale for prediction-based LCS and 2) how to use the DecisionPrecision (DP) encounter tool with eligible patients to personalize LCS discussions.

Results: Thematic content analysis from detailing visit data identified 6 categories of clinician willingness to use the DP tool to personalize SDM for LCS (adoption potential), varying from "Enthusiastic Potential Adopter" (n = 18) to "Definite Non-Adopter" (n = 16). Many clinicians (n = 52) articulated how they found the concept of prediction-based SDM highly appealing. However, to varying degrees, nearly all clinicians identified challenges to incorporating such an approach in routine practice.

Limitations: The results are based on the clinician's initial reactions rather than longitudinal experience.

Conclusions: While many primary care clinicians saw real value in using prediction to personalize LCS decisions, more support is needed to overcome barriers to using encounter tools in practice. Based on these findings, we propose several strategies that may facilitate the adoption of prediction-based SDM in contexts such as LCS.

Highlights: Encounter tools that incorporate prediction models promote personalized shared decision making (SDM), but little is known about clinicians' perceptions of the feasibility of using these tools in practice.We examined how clinicians react to using one such encounter tool for personalizing SDM about lung cancer screening (LCS).While many clinicians found the concept of prediction-based SDM highly appealing, nearly all clinicians identified challenges to incorporating such an approach in routine practice.We propose several strategies to overcome adoption barriers and facilitate the use of prediction-based SDM in contexts such as LCS.

Keywords: academic detailing; lung cancer screening; shared decision making.

Figure 1

Output from the DecisionPrecision encounter-based decision tool (example for a high-benefit patient. (1) Risk factor inputs. (2) Patient’s screening eligibility status. (3) Patient’s risk of dying from lung cancer in the next 6 y. (4) Visual representation of where the patient falls on the risk spectrum for the target population of screening-eligible patients. (5) Notes on how to discuss screening options with patient.