SHM Converge Scientific Abstract ePoster Gallery-Preferred Language and Inpatient Mortality: Every Voice Counts

Pdf Summary

This poster examines whether missing language data in the electronic health record affects analyses of inpatient mortality. Prior literature has found no adjusted mortality difference between English and non-English speakers, but patients with missing language information are often excluded. In the Mount Sinai health system, missing language is recorded as “null,” “language not listed,” “need to collect,” or “unable to communicate.”<br /><br />The authors reviewed 2022 medicine admissions and found that patients with missing language data had higher mortality than those with a recorded preferred language. They also observed mortality differences within the missing-language subgroup itself. Because of this, they conducted chart reviews focused on documented interpreter use to retrospectively identify the preferred language for these patients.<br /><br />Preliminary results show that 294 unique patients accounted for 507 admissions with missing language data. So far, the preferred language has been identified for 138 patients. Early findings suggest that missing language status often reflects severity of illness at presentation rather than simply incomplete documentation. English remains the most common preferred language, but the proportion of non-English speakers appears higher than initially recorded (28% vs. 17%).<br /><br />The main conclusion is that equity analyses must include patients with missing language data, since excluding them may underestimate language-related disparities and bias mortality findings. Next steps include reassessing mortality after reclassifying patients by preferred language and examining whether disparities exist around the hospitalization experience.

Asset Subtitle

Patricia Dharapak

Meta Tag

Author List Benjamin Dempsey, Beth G. Raucher, Ian Yu, Mary Ellen Dupont, Patricia Dharapak

Category Innovations

Concept Preferred Language Data

Concept Inpatient Mortality

Concept Health Equity

Concept Incomplete Data

Concept Electronic Health Record

Distinguished Non-Finalist

Presenter Organization Mount Sinai Beth Israel

Presenting Author Patricia Dharapak

Track Quality Improvement

Keywords

language data

electronic health record

inpatient mortality

missing data

preferred language

health equity

interpreter use

chart review

mortality disparities

non-English speakers

Preferred Language Data

Inpatient Mortality

Health Equity

Incomplete Data

Electronic Health Record