A new method for identification of drowning-related out-of-hospital cardiac arrest (Danish Drowning Formula): A retrospective cohort study with 30-day follow-up

Abstract

Background Accurate, reliable, and sufficient data is required to reduce the burden of drowning by targeting preventive measures and improving treatment. Today’s drowning statistics are informed by various methods sometimes based on data sources with questionable reliability. These methods are likely responsible for a systematic and significant underreporting of drowning. This study’s aim was to assess the 30-day survival of patients with out-of-hospital cardiac arrest (OHCA) identified in the Danish Cardiac Arrest Registry (DCAR) after applying the Danish Drowning Formula. Methods This nationwide, cohort, registry-based study with 30-day follow-up used the Danish Drowning Formula to identify drowning-related OHCA with a resuscitation attempt from the DCAR from January 1st, 2016, through December 31st, 2021. The Danish Drowning Formula is a text-search algorithm constructed for this study based on trigger-words identified from the prehospital medical records of validated drowning cases. The primary outcome was 30-day survival from OHCA. Data were analyzed using multiple logistic regression. Results In total, 30,215 OHCA were registered in the DCAR. The Danish Drowning Formula identified 707 potential drowning related OHCA. Of these, 374 were non-drowning, and 16 were excluded because of irreversible death resulting in 317 (1%) patients with drowning related OHCA compared to 29,882 patients with OHCA from other causes. The 30-day survival for patients with drowning-related OHCA was higher compared to OHCA from other causes (33% vs 14%, p < 0.001). Adjusted OR for 30-day survival for drowning-related OHCA and other causes of OHCA was 2.3 [1.7–3.2], p < 0.001. Conclusions This study found higher 30-day survival among drowning-related OHCA compared to OHCA from other causes. This study proposed that a text-search algorithm (Danish Drowning Formula) could explore unstructured text fields to identify drowning persons. This method may present a low-resource solution to inform the drowning statistics in the future. Registration: This study was registered at ClinicalTrials.gov before analyses (NCT05323097).