Prehospital care is essential in the stroke chain of survival, yet adherence to evidence-based guidelines by emergency medical services (EMS) remains inconsistent.1 The 2019 American Stroke Association guidelines2 recommend use of a validated stroke scale and blood glucose evaluation by EMS, yet <1% of patients receive fully guideline-concordant prehospital stroke care.
Neighborhood socioeconomic status is associated with stroke incidence3 and outcomes,4 but there is limited research evaluating its impact on prehospital stroke care. Given associations between urbanicity and blood glucose documentation,1 we sought to determine if the Social Vulnerability Index (SVI), incorporating a broader range of social determinants,5 was associated with EMS care for patients with suspected stroke/transient ischemic attack (TIA).
We conducted a retrospective analysis using the 2019 ESO Data Collaborative, containing >8 million annual EMS activations. We included 9–1-1 responses with an EMS field impression of stroke/TIA. Encounters were excluded if they occurred in a hospital/acute care facility, or if data were missing on EMS call type, location, incident Census tract, or SVI. This project was deemed not human subjects research by the Harvard Medical School Institutional Review Board (Protocol #IRB23–0604).
Our primary outcomes were EMS documentation of stroke scale assessment and blood glucose. While stroke scale assessment is included in some telephone dispatch systems, our analysis focused on EMS clinician assessments. We analyzed incident location SVI, patient demographics, level of EMS care, and location type. A masked EMS agency identifier attributed activations to each agency. Mixed effects logistic regression models were fit to assess associations between SVI and documentation of stroke scale and blood glucose. Residual intraclass correlation (ICC) was calculated to assess the extent of EMS agency-level variation in care.
Among 84,138 patients with suspected stroke/TIA (46% male, mean age 71 years), stroke scale was documented in 78% and blood glucose in 88% of cases. Compared to patients in the least vulnerable SVI quartile, those in the most vulnerable quartile had 10% lower odds of documented stroke scale (OR 0.90, 95% CI 0.82–0.99, Figure 1A), but similar odds of documented blood glucose (Figure 1B). Agency-level variation was higher for stroke scale documentation (ICC 72%, 95% CI 69–74) than blood glucose (ICC 22%, 95% CI 20–24).
Figure 1.
Association between Social Vulnerability Index (SVI) and documentation of stroke scale* (panel A) and blood glucose (panel B).
*Stroke scales used included BEFAST, CPSS, EMSA, FAST, FAST-ED, LAMS, LAPSS, MEND, RACE, and VAN. Given differences in local and state EMS protocols across the country, specific stroke scale used varied by EMS agency, with some regional standardization.
** Rates of missing data varied across variables, as follows: age (3.1%), sex (0.9%), race/ethnicity (6.5%), and level of care (9.1%).
Less documentation of stroke scale in the most vulnerable areas may represent failure to conduct the assessment or documentation challenges. The marked agency-level variation suggests that differences in care may be driven by EMS agency-level factors and neighborhood characteristics. Blood glucose assessment was more uniformly documented, suggesting stronger clinical standardization.
Limitations include potential misclassification of race/ethnicity by EMS and reporting bias from differential documentation practices across agencies. Misclassification of cases as stroke/TIA versus not could also introduce bias, though this is unlikely to vary systematically by SVI. Future directions include analyzing hospital pre-notification and EMS response times, which may also be associated with socioeconomic disparities. Given the substantial agency-level variation, agency-level interventions may be a powerful way to improve care delivery, ultimately decreasing disparities in prehospital stroke care.
SOURCES OF FUNDING
Dr. Misra received an American Heart Association Student Scholarship in Cerebrovascular Disease and Stroke. Drs. Cash and Zachrison received grant support from the Society for Academic Emergency Medicine Foundation. Dr. Zachrison reports grants from the National Institute on Aging.
Non-standard Abbreviations and Acronyms
- EMS
Emergency medical services
- SVI
Social Vulnerability Index
- TIA
Transient ischemic attack
- OR
Odds ratio
- ICC
Intraclass correlation
Footnotes
DISCLOSURES
Dr. Crowe is an employee of ESO.
REFERENCES
- 1.Misra AJ, Goldberg SA, Zachrison KS, Crowe RP, Cash RE. EMS Agency Characteristics Associated with Documentation of Prehospital Stroke Scale and Blood Glucose Level. Prehosp Emerg Care. 2025. Jul 22:1–8. [DOI] [PubMed] [Google Scholar]
- 2.Dylla L, Rice JD, Poisson SN, Monte AA, Higgins HM, Ginde AA, Herson PS. Analysis of Stroke Care Among 2019–2020 National Emergency Medical Services Information System Encounters. J Stroke Cerebrovasc Dis. 2022;31(3):106278. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kim Y, Twardzik E, Judd SE, Colabianchi N. Neighborhood Socioeconomic Status and Stroke Incidence: A Systematic Review. Neurology. 2021;96(19):897–907. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Stulberg EL, Twardzik E, Kim S, Hsu C, Xu Y, Clarke P, Morgenstern LB, Lisabeth LD. Association of Neighborhood Socioeconomic Status With Outcomes in Patients Surviving Stroke. Neurology. 2021;96(21):e2599–e2610. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.CDC/ATSDR Social Vulnerability Index (SVI). https://stacks.cdc.gov/view/cdc/131979. Published July 28, 2023. Accessed May 30, 2025. [Google Scholar]