Does de-implementation of low value care impact the patient-clinician relationship? A mixed methods study

Context: In the U.S., provision of LVC is prevalent (up to 20% of total health services), costly ($350 billion annually), and associated with patient harm. Concern about a negative impact on the patient-clinician relationship has been cited by primary care clinicians as a barrier to reducing LVC. Study Design: Mixed-methods study. Participants completed the Patient-Doctor Relationship Questionnaire (PDRQ-9) after reading one of three vignettes about a clinician declining to provide a requested LVC service (antibiotics for sinusitis, screening EKG, or screening vitamin D test) or a comparison vignette. A different sample of participants was asked to imagine that their own clinician did not order requested antibiotics or screening EKG and then to respond to structured interview questions about satisfaction and trust. Setting: Primary care patients of a southwest Virginia health system. Population: Adults (n= 232 questionnaire/n= 24 interview). Outcome Measures: Participant demographics data, PDRQ-9 score for each vignette (higher score = greater relationship integrity), and thematic analysis of interview responses. Results: Among questionnaire participants, a lower PDRQ-9 score was associated with the vignette about not providing LVC vitamin D screening (31.2) compared with antibiotics (38.9), EKG screening (37.5), and the comparison vignette (36.4) (p<0.05). There was a statistically significant, but weak, correlation between education and PDRQ-9 score (r=0.2, p<0.01). More than half of interviewees believed that their satisfaction and trust would not be negatively impacted by not receiving the LVC service, citing the strength of their relationship with their clinician and faith in their guidance. Some even felt that not providing the service would increase their satisfaction and trust. Participants who believed their satisfaction and trust may be impacted seemed to recognize the complexity of the scenario, discussing medical necessity, potential harm, insurance, and the option to go elsewhere (ex: urgent care or a new doctor). However, most emphasized that negative impacts could be mitigated if the clinician listened to them, spent time with them, and offered understandable advice. Conclusions: Findings emphasize prioritization of the patient-clinician relationship in LVC de-implementation interventions and suggest minimal impact of such interventions on the patient-clinician relationship. Evidence of service-specific differences was observed.