Photo Credit: DC Studio

The following is a summary of “Nonresponse Bias against the underserved in emergency department patient experience surveys: A Cross-Sectional analysis,” published in the August 2024 issue of Emergency Medicine by Zitek et al. 

A small fraction of patients typically respond to patient experience surveys to assess emergency physicians. No prior studies have identified patient characteristics that predict responses to emergency department (ED) patient experience surveys.

Researchers conducted a retrospective study to determine patient characteristics associated with experience survey response. 

They analyzed a random sample of 2,500 patients from the hospital system who had been discharged from an ED between January 1 and June 30, 2022. The hospital system included 1 hospital-based ED and 2 freestanding EDs, 1 in an economically disadvantaged area. Each randomly selected patient reviewed charts to collect data on the ED visit, patient age, gender, race, preferred language, primary residence, time of arrival, marital status, health insurance, and the emergency severity index for that visit. They used multivariable logistic regression to determine which variables were associated with responses to the patient experience survey.

The results showed that out of 2,500 patients, 207 (8.3%) responded. The lowest response rates were observed among patients from a freestanding ED in an economically disadvantaged area (5.9%), those who were uninsured (4.9%), and those who were homeless (2.1%). Multivariable analysis revealed the following adjusted odds ratios for survey response: 0.54 (95% CI 0.33–0.84) for patients from the economically disadvantaged ED, 0.56 (95% CI 0.34–0.90) for uninsured patients, and 0.30 (95% CI 0.02–1.44) for homeless patients.

They concluded that multiple markers of lower socioeconomic status were associated with decreased patient experience survey response rates, giving a nonresponse bias against these individuals.

Source: jem-journal.com/article/S0736-4679(24)00271-3/abstract#%20