Interpreters were requested for 12.1% of pediatric ED patients in 2016, up from ~2% in 2002

Description #

Across the 2016 CHOA pediatric ED cohort, interpreter services were requested for at least one ED encounter in 18,572 (12.1%) of 152,945 patients, of whom the large majority were Latino/Hispanic (83.9%) and Spanish-speaking (79.5%). The authors note this 12.1% LEP/interpreter-request prevalence rose from approximately 2.0% in 2002, and that 83.3% of patients with a primary language other than English were documented as having requested an interpreter.

"Interpreter services were requested for at least one ED encounter in 18,572 (12.1%) patients. Of these patients, 15,582 (83.9%) were Latino or Hispanic, and 14,763 (79.5%) listed Spanish as their preferred language." (Greenky, 2019, p. 2)

"in 2016 12.1% of patients requested an interpreter and deemed LEP compared to 2.0% in 2002." (Greenky, 2019, p. 5)

Methods Context #

What? #ⓘ

The observable: whether an interpreter was requested during a patient's ED encounter (used both as the operational definition of LEP and as a utilization measure), recorded from the EPIC EMR "interpreter requested" field.

"Patients were categorized as LEP if there was an interpreter requested during their encounter. The interpreter may have been requested by the patient, the patient's family, or the healthcare provider." (Greenky, 2019, p. 2)

How? #ⓘ

Retrospective cohort study; interpreter-request status pulled from the EPIC EMR and summarized as frequencies/percentages at the patient level (N=152,945).

"Baseline demographic summaries were calculated at the patient-level (N=152,945 records), whereas clinical and hospital values were computed at the encounter-level (N=232,787 records)." (Greenky, 2019, p. 2)

Who? #ⓘ

All patients aged 0–18 presenting to three CHOA pediatric EDs in Atlanta during 2016 (152,945 patients / 232,787 encounters); a child is considered LEP when their caregivers are LEP. Dead-on-arrival and unlisted-language patients excluded.

"It is important to note that, in our system, a child is considered LEP when their caregivers are LEP, even if the child is proficient in English." (Greenky, 2019, p. 1)

Other Notes #

The next five preferred languages after Spanish/English were Burmese, Amharic, Vietnamese, Portuguese, and Arabic (each ≤0.13%), illustrating a long tail of less-resourced LEP languages relevant to the distribution-of-need sub-question.

Caveats #

• Interpreter-request EMR field is an imperfect proxy for LEP with substantial undocumented ascertainment LEP status and interpreter use were both operationalized from a single EMR "interpreter requested" field, which the authors acknowledge is an imperfect proxy. Only 83.3% of patients with a non-English primary language were documented as having requested an interpreter, leaving 16.7% whose interpreter use is unknown; there is no EMR field confirming an interpreter was actually offered or used, it is unclear who requested the interpreter (patient, family, or provider), and unclear who was LEP (patient, caregiver, or both). Patients who declined an interpreter but would have benefited are also misclassified. This measurement uncertainty could bias the small/null group differences in the ED outcome comparisons.