RESUMO
BACKGROUND: Racial-ethnic disparities are pervasive in health care. One mechanism that may underlie disparities is variation in shared decision-making (SDM), which encompasses high-quality clinician-patient communication, including deliberative discussions about treatment options. OBJECTIVES: To determine whether SDM has causal effects on outcomes and whether these effects are stronger within racial-ethnic concordant clinician-patient relationships. RESEARCH DESIGN: We use instrumental variables to estimate the causal effect of SDM on outcomes. SUBJECTS: A total of 60,584 patients from the 2003-2017 Integrated Public Use Microdata Series Medical Expenditure Panel Survey. Years 2018 and 2019 were excluded due to changes in the Medical Expenditure Panel Survey that omitted essential parts of the SDM index. MEASURES: Our key variable of interest is the SDM index. Outcomes included total, outpatient, and drug expenditures; physical and mental health; and the utilization of inpatient and emergency services. RESULTS: SDM lowers annual total health expenditures for all racial-ethnic groups, but this effect is only moderated among Black patients seen by Black clinicians, more than doubling in size relative to Whites. A similar SDM moderation effect also occurs for both Black patients seen by Black clinicians and Hispanic patients seen by Hispanic clinicians with regard to annual outpatient expenditures. There was no significant effect of SDM on self-reported physical or mental health. CONCLUSIONS: High-quality SDM can reduce health expenditures without negatively impacting overall physical or mental health, supporting a business case for health care organizations and systems to improve racial-ethnic clinician-patient concordance for Black and Hispanic patients.
Assuntos
Gastos em Saúde , Hispânico ou Latino , Humanos , Tomada de Decisões , Tomada de Decisão Compartilhada , Grupos Raciais , População Branca , Negro ou Afro-AmericanoRESUMO
This national US study determined the causal impact of shared decision making (SDM) on pain outcomes, including any overall pain and the subcategories of any acute pain and any chronic pain. We additionally examined whether the causal impact of SDM on overall pain is moderated by gender, race-ethnicity, clinician-patient racial-ethnic concordance, and clinician-patient gender concordance. We used national US data from the 2003-2017 Medical Expenditure Panel Survey, which were externally valid to the US national non-institutionalized population, employed a standard measure of shared decision making, and applied an internally-valid two-stage least squares approach that used the peer SDM behavior of similar clinicians as an instrument. The instrument was sufficiently strong and statistically uncorrelated with patient characteristics. We found a large impact of SDM on both female and male chronic pain outcomes, where the effect for females was approximately 50% larger than for males, with a 10 percentage point increase in SDM quality resulted fewer females experiencing chronic pain that interfered with everyday activities (-24.8 percentage points; 95% confidence interval [CI]: 43.3, -6.4) than males (-16.5 percentage points; 95% CI: 32.9, -1.0). We estimated that a 10 percentage point national increase in the SDM index would thus result in 10.1 million fewer females and 5.7 million fewer males in the US experiencing chronic pain that interfered with their daily activities. Policy implications include both increasing the quality of SDM for all patients and educating male patients on the value of SDM.