correlation between trust measure and quality measure as a tool to improve trust through quality improvement

Context: It is clear that the trust patients have for their providers impacts health outcomes. In particular, mis-trust patients have for providers can lead to significant health inequity. We are seeking ways to improve trust patients have for providers leveraging tools for quality improvement (QI) by finding QI measures that correlate with trust patients have for providers.

Objective: Understand if there are quality improvement measures which correlate with trust patients have for providers.

Study Design and Analysis: Pilot study that compares the results of two validated survey-based measures in a population of adults.

Setting or Dataset: Patients recruited from a research registry set in an academic health center. Dataset is demographic information about the 111 adult participants along with the results from the Patient Centered Primary Care Measure (PCPCM) and the Health Care Relationship Trust Scale (HCR-TS) for each participant.

Population Studied: diverse adult volunteers

Intervention/Instrument: Survey using two established instruments: PCPCM and HCR-TS

Outcome Measures: Correlation between the survey results as a whole, and individual questions on the instruments were evaluated

Results: The correlation between the measures is evident with a slope of 0.98, an r^2 of 0.53.

Conclusions: This pilot work demonstrates that study of the relationship between trust and the PCPCM is feasible and can be carried out using online survey tools in a rapid cost effective way.

Further investigation of this correlation, especially with a focus on selected demographic groups, will help inform how robust the correlation will be. Given the small numbers in various demographic groups in this study, it is premature to speculate what associations might be present in different groups.

The subscore correlation of the PCPCM measure elements with the overall trust measure allow the construction of a “surrogate trust measure” that may actually correlate more with trust than the PCPCM. We share an example of a model using the three PCPM elements with the highest individual correlation integrated into a metric. In this example the r^2 correlation goes up to 0.65.

After further study results indicating that PCPCM is a measure of trust can support the use of PCPCM as a critical measure which will provide information about the trust people have for their providers. Interventions around improving PCPCM should improve trust and vice versa.