A retrospective analysis of emergency medical records found that witnessed non-shockable out-of-hospital cardiac arrests were associated with higher odds of return of spontaneous circulation, as were arrests whose first recorded rhythm was pulseless electrical activity, or PEA, rather than asystole, the other rhythm studied. The same broad pattern appeared when researchers looked at circulation that lasted through the end of EMS care. The findings describe associations in a registry, not proof that any treatment caused a better result.
ROSC was uncommon: 166 of 1,198 patients, or 13.9%, had any documented ROSC, and 146, or 12.2%, had sustained ROSC through the end of EMS care. That endpoint was not survival to hospital discharge.
Witnessed status and initial rhythm stood out.
The researchers used a deidentified Polk County Fire Rescue CARES registry export to examine which patient and event factors were associated with any ROSC and sustained ROSC in non-shockable out-of-hospital cardiac arrest. The analysis was retrospective and evaluated age, sex, whether the arrest was witnessed, the initial rhythm, epinephrine, CPR timing, field hypothermia care and AED use before EMS arrival.
Of 1,365 source records, 167 were excluded, leaving 1,198 patients whose first monitored rhythm was PEA or asystole. PEA accounted for 250 patients, or 20.9%, while 948, or 79.1%, had asystole. The median age was 66, with an interquartile range of 53 to 78, and 716 patients, or 59.8%, were male.
Before adjustment, patients with any ROSC more often had witnessed arrests and PEA and had shorter time to first CPR than patients without ROSC.
The treatment links need more caution.
Researchers reported adjusted odds ratios, which compare groups after accounting for the other recorded variables. For any ROSC, witnessed arrest was associated with an aOR of 2.65, with a reported 95% confidence interval from 1.80 to 3.90. PEA was associated with an aOR of 1.90, with an interval from 1.29 to 2.79, and field hypothermia care with an aOR of 1.99, with an interval from 1.37 to 2.90. Male sex was associated with lower odds, with an aOR of 0.60 and an interval from 0.42 to 0.86.
For sustained ROSC, the corresponding aORs were 2.29 for witnessed arrest, with a 95% confidence interval of 1.50 to 3.49; 1.97 for PEA, with an interval of 1.32 to 2.94; and 2.03 for field hypothermia care, with an interval of 1.37 to 3.02. Male sex again had lower odds, at 0.59, with an interval of 0.40 to 0.87. Each additional minute to first CPR was associated with lower odds, with an aOR of 0.989 per minute and an interval of 0.979 to 0.999.
Epinephrine produced a more difficult-to-interpret signal. Its adjusted association was not statistically significant for any ROSC, with an aOR of 0.43, a reported interval of 0.19 to 1.01 and p=0.053, but it was inversely associated with sustained ROSC, with an aOR of 0.34, an interval of 0.14 to 0.87 and p=0.025. AED application before EMS arrival was close to the null in both models, with aORs of 1.04 for any ROSC and 1.00 for sustained ROSC.
The authors caution that the epinephrine and field hypothermia findings should not be read as treatment effects. They point to likely time-dependent confounding and selection bias, while the observational design prevents the analysis from establishing treatment efficacy.
The registry leaves important questions open.
The complete-case regression models included 893 patients. The analysis treated 281 spreadsheet error values for time to first CPR as missing, and field hypothermia status was missing in 35 records. No imputation, winsorization or categorization was applied.
About 25% of the cohort was therefore excluded from the complete-case models. The retrospective design limits causal conclusions, and results from a single county may not represent other EMS systems.
The time-to-first-CPR measure contained extreme values and possible data-entry or time-calculation errors. Important confounders and process variables were not modeled. Because the outcomes stopped at ROSC during EMS care, the study did not assess hospital-discharge survival or neurologic function.
Taken together, the results point to witnessed status, PEA and earlier CPR as factors associated with better ROSC measures in this registry. They do not establish treatment effects for epinephrine or field hypothermia care, and they cannot be extended to survival after hospital care.
Paper data and sources
Original title: Prehospital factors associated with return of spontaneous circulation after nonshockable out-of-hospital cardiac arrest: a retrospective analysis of the Polk County Fire Rescue CARES registry.
Authors: Hammad Elahi, Latha Ganti, Paul R Banerjee
Journal/Repository: International journal of emergency medicine
Status: Peer-reviewed
First online: 2026-08-20
DOI: 10.1186/s12245-026-01347-2
Original paper · Full text