OBJECTIVES:

Evidence is unclear on whether individuals with Limited English Proficiency (LEP) and access to a trained language interpreter have greater or less health care utilization compared to similar individuals without access to an interpreter. This paper seeks to examine the health care utilization of individuals with LEP at Mayo Clinic who used an interpreter compared to those who did not.

METHODS:

The first model examined intensive-care-unit (ICU) length-of-stay (LOS) using a negative binomial model. The second model examined where an individual was discharged to using an ordered probit model. The final model examined whether an individual was readmitted to hospital using a probit model. Each model used robust standard errors and controlled for a range of covariates including interpreter use.

RESULTS:

There were 3,835 hospitalizations in the dataset from 2008-2017. 1,532 hospitalizations had use of an interpreter and 2,303 did not. The incremental effect of an individual having used an interpreter compared to an individual not having used an interpreter is 0.57 days (p<0.01) in the ICU ceteris paribus. If an individual had used an interpreter, they were 11-percentage points less likely (p<0.01) to be discharged to home; three-percentage points more likely (p<0.01) to be discharged to further care; seven-percentage points more likely (p<0.01) to die compared to an individual who did not use an interpreter ceteris paribus. If an individual had used an interpreter, they were six-percentage points less likely (p<0.01) to have been readmitted to hospital compared to an individual who did not use an interpreter ceteris paribus.

CONCLUSIONS:

Individuals with LEP who use an interpreter may be more resource intensive initially but become resource-saving through their lower likelihood of hospital readmission compared to those without use of an interpreter. These findings can provide policymakers with insight on the benefit of interpreters and may provide the impetus for hiring more.