Arnold G, Baranowski R, Duhigg L. — American Board of Internal Medicine
Presented: American Public Health Association Conference, November 2013
Background: Few studies on the effectiveness of planned visits for patients with chronic conditions exist; instead, researchers have focused on the Chronic Care Model overall. Planned visits ensure that clinical teams prepare and provide patients with evidence-based care and condition-specific self-management training. This study investigated whether primary care practices completing components in a planned visit protocol (PVP) had better patient-visit experiences than matched-control practices not using PVP. Better experiences included higher patient ratings on quality of communication, shared decision-making and overall quality of care.
Method: The study included 1,323 internists from practices reporting specific chronic illnesses as the most important condition among their patients. Participants completed ABIM's Practice Improvement Module (PIM) on Communication for Maintenance of Certification. In this PIM, patients completed CG-CAHPS-based surveys. A total of 23,927 eligible respondents made three or more visits per year – 197 physicians used PVP; 197 of 1,126 practices without PVP were controls. Three patient rating scores were created for: 1) physician communication (0-35 score, Cronbach's α = .86); 2) level of shared decision-making (0-9 score; Cronbach's α = 0.77); and 3) overall physician rating (0-10 score; Cronbach's α = 0.82). It was hypothesized that practices using PVP have higher ratings than practices that do not. Propensity-score matching formed case-control pairs. One score matched 28 physician and practice variables. A second score matched 11 patient demographic and survey variables; this weighted score adjusted for patient clustering within physicians. Pairs were matched by shortest Mahalanobis distances. Scores were compared using Wilcoxon signed-rank tests with a two-tailed α = 0.1 significance.
Results: PVP cases had a higher communication mean score than controls, but the difference between means was not significant [mean and (SD)]: cases, 27.7 (2) versus controls, 27.6 (2); difference = 0.09, 90% CI: -.28 to .45, P = .50. The PVP cases had a higher shared decision-making mean than controls, but the difference was not significant: cases, 7.4 (1) versus controls, 7.3 (1); difference = .14, 90% CI: -.07 to .36, P = .20. Mean overall ratings for matched pairs were nearly identical and again not significantly different: cases, 9.4 (0.6) versus controls, 9.4 (0.6); difference < 0.01, 90% CI: -.10 to .10, P = .60.
Conclusion: Practices utilizing the PVP for patients with chronic illnesses show no significant differences in quality ratings, compared with practices not using the protocol. Implementing planned visits alone, however, may not be sufficient to improve the quality of interactions during visits. Whether improved interactions result from planned visits is inconclusive.
For more information about this presentation, please contact Research@abim.org.