This retrospective study of 9,386 adult non-traumatic OHCA patients identified multiple independent predictors of ROSC and developed a nomogram prediction model with an AUC of 0.775 and negligible overfitting that showed good calibration and favorable net clinical benefit.
The overall ROSC rate among adult non-traumatic OHCA patients in this study was 7.2%.
Shockable initial cardiac rhythm was the strongest single predictor of ROSC, associated with nearly a 3-fold increase in odds.
Endotracheal intubation was independently associated with significantly higher odds of ROSC.
Longer onset-to-ambulance intervals were progressively and independently associated with lower odds of ROSC.
Epinephrine administration exceeding 5 mg was independently associated with lower odds of ROSC.
Public location of cardiac arrest onset was independently associated with higher odds of ROSC compared to non-public locations.
Physician characteristics including sex, years of practice, and professional title were independently associated with ROSC outcomes.
The developed nomogram prediction model demonstrated good discrimination, minimal overfitting, good calibration, and favorable net clinical benefit.
This research suggests that many factors — ranging from where and when a cardiac arrest happens, to how quickly emergency services respond, to the specific treatments provided and the experience of the treating physician — all meaningfully affect whether an out-of-hospital cardiac arrest patient achieves return of spontaneous circulation (ROSC), meaning their heart starts beating again before reaching the hospital. The study analyzed nearly 9,400 cardiac arrest cases treated by emergency services in Tianjin, China over two years, finding that only about 1 in 14 patients achieved ROSC. Key factors linked to better outcomes included having a 'shockable' heart rhythm (the type that can be treated with a defibrillator), receiving endotracheal intubation (a breathing tube), having a cardiac arrest in a public place, and having a more experienced physician respond. Factors linked to worse outcomes included longer delays between collapse and ambulance arrival, and receiving high doses of epinephrine (adrenaline), which likely reflects prolonged and difficult resuscitation attempts. Using these findings, the researchers built a predictive scoring tool called a nomogram that can estimate the probability of ROSC for an individual patient based on these factors. The model performed well statistically, with an accuracy score (AUC) of approximately 0.775 in both the group used to build it and a separate test group, and showed very little overfitting — meaning it is likely to generalize to new patients rather than just fitting the original data. This research suggests that pre-hospital emergency systems might improve outcomes by minimizing response times, prioritizing airway management, and accounting for physician experience in dispatch decisions. The prediction model could potentially help clinicians and emergency systems identify which patients are more or less likely to benefit from continued resuscitation efforts.
Zhang Q, Pang Z, Yan X, Tang F, Li S, Shao X, et al.. (2026). Influencing factors analysis and prediction model development for ROSC in adult non-traumatic OHCA: a retrospective study based on pre-hospital emergency settings.. Annals of medicine. https://doi.org/10.1080/07853890.2026.2740275