Language Access in Healthcare
EvidenceE-0305Initial AI draft

After Spanish, Chinese (29 percent) and Arabic (23 percent) were the most requested non-English languages with significant regional variation at BMT centers

2026-06-052 out · 0 in

Source

Idossa (2019). Access to Linguistically Appropriate Information for Blood and Marrow Transplant Patients: Results from Transplant Center Staff Survey. Journal of Cancer Education.

Description #

After Spanish, the next most frequently requested non-English languages among BMT-center patients were Chinese (29.3%), Arabic (23.2%), Vietnamese (11.0%), and Portuguese (8.5%) (Table 2). Requests for several of these languages varied significantly by U.S. region: Chinese was more frequently requested in the Northeast and West (P = 0.0245), Vietnamese in the West (P = 0.0211), and Portuguese in the Northeast (P = 0.0007). Arabic requests did not vary significantly by region (P = 0.0832). This documents a long, regionally patterned tail of language demand beyond Spanish.

"More than 90% of respondents indicated that Spanish was the most frequently requested non-English language across all geographic regions, followed by Chinese (29.3%), Arabic (23.2%), Vietnamese (11.0%), and Portuguese (8.5%) (Table 2). Chinese was more frequently requested in the Northeast and West regions; Vietnamese more frequently in the West, and Portuguese was requested most often in the Northeast (P < 0.05, respectively)." (Idossa, 2019, p. 1034)

Methods Context #

What? #

The observable: which non-English languages transplant-center patients most often requested and whether that pattern differed across U.S. census regions.

"Most often requested languages (in general)" (Idossa, 2019, p. 1034, Table 2)

How? #

Cross-sectional web-based survey; language-request frequency was collapsed into "often"/"not often" categories to meet Chi-square assumptions and compared across regions with Chi-square or Fisher's exact tests.

"In order to meet Chi-square test assumptions, Likert scales were collapsed into two categories, Boften^ and Bnot often,^ for comparison of language service requests and provider-level/system-level barriers by geographic region and patient population." (Idossa, 2019, p. 1033)

Who? #

The 77 transplant-center-staff respondents with a known U.S. census region (Midwest n = 22, Northeast n = 17, South n = 22, West n = 16).

"Regional variation in requests and difficulty in meeting needs by language (N = 77)" (Idossa, 2019, p. 1034, Table 2)

Other Notes #

The pdftotext rendering "Boften^"/"Bnot often^" reflects the source's curly quotation marks around "often"/"not often" (a known verbatim-audit artifact).

Caveats #

  • Language categories were not disaggregated by dialect or written-form preference within a language The survey collected language demand at the level of broad language labels and did not disaggregate dialect differences within a language (for example, Mandarin vs. Cantonese) or preference for the written form (for example, formal/traditional vs. simplified Chinese). Reported request frequencies for a language such as "Chinese" therefore aggregate distinct spoken and written needs that would matter for actually delivering concordant interpretation or translated materials.
  • Needs are aggregate staff-reported perceptions with no patient-level denominator of LEP patients served or excluded The reported language needs and demand distributions are aggregate staff perceptions, not patient-level measurements. The survey did not determine the exact number of LEP patients each participating center serves, and the authors had no data on LEP patients who never access BMT. The reported prevalences of most-requested and most-difficult languages therefore reflect staff-perceived frequency rather than a denominator-based count of LEP patients, and may not capture the needs of patients who are screened out or never reach a transplant center.