Among vulnerable populations, high trust in emergency department physicians associated with higher primary care-related ED use

Previous research suggests that perceived and actual barriers to usual source care (USOC) providers increase emergency department (ED) utilization for primary care-related (PCR) conditions – those preventable or treatable with appropriate primary care. Less is known about how patients' perceived quality of care, particularly trust in providers, influences PCR-ED use. Using a sample of primarily low-income, uninsured/underinsured Black and Hispanic patients in Houston, TX, this study aimed to: 1) identify patient characteristics associated with provider trust and 2) examine relationships between mean PCR-ED visits (outcome of interest) and provider trust based on patients' responses to the statements “I trust the doctors in the ER to provide better overall care than doctors at other places” (ED physician trust) and “I trust the hospital to provide better overall care than other places” (hospital trust). Binary variables were coded: high trust=1 if patients responded “strongly agree”/“agree”; high trust=0 if patients responded “strongly disagree”/“disagree”/“neither/neutral”. We calculated univariate odds ratios (ORs) to examine associations between patient characteristics and ED physician trust and hospital trust; employed Poisson models to estimate the effect of high trust on mean PCR-ED visits; predicted mean PCR-ED visits for each group (high trust=1/0) and tested the differences for significance at alpha=0.05. Our sample included patients who visited one of three Houston-area EDs from February-August 2012 for PCR-ED reasons (n=445). Participants' PCR-ED use averaged 2.6 visits. The majority of participants were female (65.9%); Hispanic (45.2%) and Black (33.9%); educated at a high school (HS) or less level (56.6%); low income (75.5% had incomes <$5,000); and uninsured (48.5%). Sixty-two percent had no USOC provider. Hispanic race/ethnicity was associated with increased odds of high trust in ED physicians (OR=2.27, p-value<.05), but not hospitals. HS (OR=0.60, p-value<.05) and more than HS (OR=0.46, p-value<.01) education levels were associated with decreased odds of high trust in ED physicians. More than HS education was associated with lower odds of high trust in hospitals (OR=0.56, p-value<.05). Having a USOC provider was associated with decreased odds of high trust in ED physicians (OR=0.57, p-value<.01) and hospitals (OR=0.37, p-value<.001). Mean PCR-ED visits were significantly higher among patients reporting high trust in ED physicians (+0.3, p-value<.05) but not among patients reporting high trust in hospitals. Our findings suggest that high trust in ED physicians may increase PCR-ED use among some patients. ED physicians should be involved in education strategies to transition PCR-ED users to more appropriate care settings.