Incidence of treated cardiac arrest in hospitalized patients in the United States

Abstract

Objective: The incidence and incidence over time of cardiac arrest in hospitalized patients is unknown. We sought to estimate the event rate and temporal trends of adult inhospital cardiac arrest treated with a resuscitation response.

Design: Three approaches were used to estimate the inhospital cardiac arrest event rate. First approach: calculate the inhospital cardiac arrest event rate at hospitals (n = 433) in the Get With The Guidelines-Resuscitation registry, years 2003-2007, and multiply this by U.S. annual bed days. Second approach: use the Get With The Guidelines-Resuscitation inhospital cardiac arrest event rate to develop a regression model (including hospital demographic, geographic, and organizational factors), and use the model coefficients to calculate predicted event rates for acute care hospitals (n = 5445) responding to the American Hospital Association survey. Third approach: classify acute care hospitals into groups based on academic, urban, and bed size characteristics, and determine the average event rate for Get With The Guidelines-Resuscitation hospitals in each group, and use weighted averages to calculate the national inhospital cardiac arrest rate. Annual event rates were calculated to estimate temporal trends.

Setting: Get With The Guidelines-Resuscitation registry.

Patients: Adult inhospital cardiac arrest with a resuscitation response.

Measurements and main results: The mean adult treated inhospital cardiac arrest event rate at Get With The Guidelines-Resuscitation hospitals was 0.92/1000 bed days (interquartile range 0.58 to 1.2/1000). In hospitals (n = 150) contributing data for all years of the study period, the event rate increased from 2003 to 2007. With 2.09 million annual U.S. bed days, we estimated 192,000 inhospital cardiac arrests throughout the United States annually. Based on the regression model, extrapolating Get With The Guidelines-Resuscitation hospitals to hospitals participating in the American Hospital Association survey projected 211,000 annual inhospital cardiac arrests. Using weighted averages projected 209,000 annual U.S. inhospital cardiac arrests.

Conclusions: There are approximately 200,000 treated cardiac arrests among U.S. hospitalized patients annually, and this rate may be increasing. This is important for understanding the burden of inhospital cardiac arrest and developing strategies to improve care for hospitalized patients.

Figure 1. Event rate of in-hospital cardiac arrest in hospitals contributing data to the Get With the Guidelines-Resuscitation Registry for consecutive years 2003–2007

This figure depicts the annual event rate of in-hospital cardiac arrest for hospitals participating in the Get With the Guidelines-Resuscitation registry. The circular points represent the event rates and the corresponding bars represent the interquartile range.