Deployment of a Cloud-Based Speech Translation Tool in a Saudi Emergency Medical Dispatch Center: A Descriptive Implementation Study

Background Emergency medical dispatch depends on spoken communication, and that dependency becomes acute during mass gatherings. The Saudi Red Crescent Authority (SRCA) 997 dispatch center in Al-Madinah receives calls from pilgrims and residents who speak languages for which no on-shift interpreter is routinely available. A cloud-based speech translation tool, SMART TRANSLATION, was introduced into the dispatch workflow to support these calls. We describe what the tool is, how it was used, and what the operational record can and cannot establish about it. Materials and methods We conducted a retrospective descriptive analysis of an operational log of translator-assisted 997 calls and two anonymous cross-sectional dispatcher surveys. SMART TRANSLATION (versions 11 and 12) went live on 18 May 2025. The merged log held 269 records. We analyzed the 79 records that carry a complete calendar date in 2025, which span 19 May to 14 October 2025 (first to last valid record), and excluded 190 records with no date (169), a day and month without a year (16), or a year of 2015 or 2016 (5). A retrospective survey of pre-deployment experience and a post-deployment survey were administered on 14 to 16 October 2025, after deployment, and were completed by every eligible dispatcher (15 of 15 and 10 of 10). Analysis was descriptive and scripted in Python 3.10.11. Results All 79 analyzed records came from Al-Madinah. Six named caller languages appear, led by Urdu (44 records, 55.7%) and English (nine records, 11.4%). A further 13 records (16.5%) carry no usable language label. A dispatcher-coded field recorded instructions as given in 63 records (79.7%) and not given in 16 (20.3%). This field is a documentation category, not a measure of instruction delivery, and it was coded inconsistently. Identical free-text descriptions were coded both ways in 20 of the 79 records, and the recorded rate moved from 69.6% to 93.9% between two date strata, in a period when staff were asked to log every case. Out of the 16 records coded as no instruction given, 11 (68.8%) describe calls in which the chief complaint was confirmed through translation and no instruction phase was recorded, and one (6.3%) records difficulty understanding the caller. The two surveys use different response bins for the understanding-time item and are not compared. All 10 post-deployment respondents chose an improvement option on an item that offered no option for no change or for worsening, and seven recommended continued use. Conclusions A cloud speech translation tool was in documented operational use at one Saudi dispatch center, and the dated record describes the languages of the translator-assisted calls that dispatchers chose to document. These preliminary, setting-specific findings do not establish translation accuracy, call-handling efficiency, patient safety, or patient benefit, and we make no such claim. The contribution is a characterized deployment, an account of the governance gaps that accompanied it, and an explicit list of the measurements a prospective evaluation would require.

Authors

Publication Details

Journal
Cureus
Published
2026-11-16
DOI
https://doi.org/10.7759/cureus.119773
Primary Topic
Interpreting and Communication in Healthcare
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Deployment of a Cloud-Based Speech Translation Tool in a Saudi Emergency Medical Dispatch Center: A Descriptive Implementation Study

Bader Alharbi, Hameed Almarwani
2 links
Cureus
Interpreting and Communication in Healthcare
article

Deployment of a Cloud-Based Speech Translation Tool in a Saudi Emergency Medical Dispatch Center: A Descriptive Implementation Study

Bader Alharbi, Hameed Almarwani
article en

Abstract

Background Emergency medical dispatch depends on spoken communication, and that dependency becomes acute during mass gatherings. The Saudi Red Crescent Authority (SRCA) 997 dispatch center in Al-Madinah receives calls from pilgrims and residents who speak languages for which no on-shift interpreter is routinely available. A cloud-based speech translation tool, SMART TRANSLATION, was introduced into the dispatch workflow to support these calls. We describe what the tool is, how it was used, and what the operational record can and cannot establish about it. Materials and methods We conducted a retrospective descriptive analysis of an operational log of translator-assisted 997 calls and two anonymous cross-sectional dispatcher surveys. SMART TRANSLATION (versions 11 and 12) went live on 18 May 2025. The merged log held 269 records. We analyzed the 79 records that carry a complete calendar date in 2025, which span 19 May to 14 October 2025 (first to last valid record), and excluded 190 records with no date (169), a day and month without a year (16), or a year of 2015 or 2016 (5). A retrospective survey of pre-deployment experience and a post-deployment survey were administered on 14 to 16 October 2025, after deployment, and were completed by every eligible dispatcher (15 of 15 and 10 of 10). Analysis was descriptive and scripted in Python 3.10.11. Results All 79 analyzed records came from Al-Madinah. Six named caller languages appear, led by Urdu (44 records, 55.7%) and English (nine records, 11.4%). A further 13 records (16.5%) carry no usable language label. A dispatcher-coded field recorded instructions as given in 63 records (79.7%) and not given in 16 (20.3%). This field is a documentation category, not a measure of instruction delivery, and it was coded inconsistently. Identical free-text descriptions were coded both ways in 20 of the 79 records, and the recorded rate moved from 69.6% to 93.9% between two date strata, in a period when staff were asked to log every case. Out of the 16 records coded as no instruction given, 11 (68.8%) describe calls in which the chief complaint was confirmed through translation and no instruction phase was recorded, and one (6.3%) records difficulty understanding the caller. The two surveys use different response bins for the understanding-time item and are not compared. All 10 post-deployment respondents chose an improvement option on an item that offered no option for no change or for worsening, and seven recommended continued use. Conclusions A cloud speech translation tool was in documented operational use at one Saudi dispatch center, and the dated record describes the languages of the translator-assisted calls that dispatchers chose to document. These preliminary, setting-specific findings do not establish translation accuracy, call-handling efficiency, patient safety, or patient benefit, and we make no such claim. The contribution is a characterized deployment, an account of the governance gaps that accompanied it, and an explicit list of the measurements a prospective evaluation would require.

Cureus
Interpreting and Communication in Healthcare
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon