Outcomes Research in Review

Declining Adverse Event Rates Among Patients With Cardiac Conditions But Not With Pneumonia or Surgical Conditions

Wang Y, Eldridge N, Metersky ML, et al. National trends in patient safety for four common conditions, 2005–2011. N Engl J Med 2014;370:341–51.



Study Overview

Objective. To examine changes in adverse event rates among Medicare patients with common medical conditions and conditions requiring surgery hospitalized in acute care hospitals.

Design. Retrospective review utilizing the Medicare Patient Safety Monitoring System (MPSMS) [1], a large database of information abstracted from medical records of a random sample of hospitalized patients in the United States. The database was established in by the Centers for Medicare and Medicaid Services in 2001 to track adverse events in hospitals among Medicare patients, with data collected from every year thereafter except for 2008. The MPSMS tracks 21 indicators of safety that can be reliably abstracted from medical records. Among these are inpatients falls, hospital-acquired pressure ulcers, catheter-associated urinary tract infections, selected hospital-acquired infections, selected adverse events related to high-risk medications, operative events and postoperative events for certain conditions.

Setting and participants . Medicare patients aged 65 and older who had been hospitalized for acute myocardial infarction, congestive heart failure, pneumonia, or conditions requiring surgery from 2005 to 2007 and 2009 to 2011. A total of 61,523 patients were included in the final study sample—11,399 with acute myocardial infarction, 15,374 with congestive heart failure, 18,269 with pneumonia, and 16,481 with conditions requiring surgery from a total of 4372 hospitals.

Main outcome measures. The rate of occurrence of adverse events for which patients were at risk, the proportion of patients with 1 or more adverse events, and the number of adverse events per 1000 hospitalizations.

Statistical analysis. Outcome rates were described and reported in 2-year intervals: 2005–2006, 2007–2009, and 2010–2011. Trends in the number of adverse events per 1000 hospitalizations were modeled using a linear mixed-effects model with Poisson link function. Other composite outcomes were also modeled using linear mixed models for trend analysis.

Main results. Adverse event rates among patients with myocardial infarction and congestive heart failure declined significantly. Among patients with myocardial infarction, rate of adverse event among patients at risk for events declined from 5% to 3.7% (rate difference 1.3%; 95% confidence interval [CI], 0.7 to 1.9) and among patients with congestive heart failure, the rate declined from 3.7% to 2.7% (rate difference 1%; 95% CI, 0.5 to 1.4). Proportion of patients with 1 or more adverse events declined by 6.6% (95% CI, 3.3 to 10.2) among patients with myocardial infarction, and 3.3% (95% CI, 1.0 to 5.5) among patients with congestive heart failure. Number of adverse events per 1000 hospitalizations also declined by 139.7 among patients with myocardial infarction and by 68.3 among patients with congestive heart failure. On the other hand, among patients admitted for pneumonia or for conditions requiring surgery, adverse events rates remained the same. Rate of adverse events among patients admitted for pneumonia remained the same at 3.4% in 2005–2006 and 3.5% in 2010–2011; and for patients admitted for conditions requiring surgery, rate of adverse events remained the same at 3.2% in 2005–2006 and 3.3% in 2010–2011. Similarly, proportion of patients with 1 or more events in the hospital also remained the same in patients with pneumonia (a proportion of 17.1% in 2005–2006 and 17.5% in 2010–11) and conditions requiring surgery (a proportion of 21.6% in 2005–2006 and 22.7% in 2010–2011). Number of events per 1000 hospitalizations also did not change over time. When accounting for patient characteristics and geographic differences in the models, the results also did not substantially change.

Conclusions. In a large nationally representative sample of older adults aged 65 and above, adverse event rates declined among patients admitted for cardiac conditions, including myocardial infarction and congestive heart failure, but did not decline among patients admitted for other medical (pneumonia) or surgical conditions.


Patient safety in inpatient hospital care is of paramount importance, and the Affordable Care Act has placed significant emphasis on improving patient safety by aligning incentives and disincentives with patient outcomes on the hospital level [2,3].These measures, including adverse event rates, are reported publicly in reports such as Hospital Compare [3–5].The current study reports on the recent national trends in safety and adverse events using data abstracted from medical records among older Medicare patients with 4 common conditions. The demonstration of the trends in adverse events represent an important first step towards understanding the current environment and trends in patient safety. The finding that in-hospital adverse event rates have improved in patients admitted for cardiac conditions is reassuring given that there were substantial nationwide efforts in promoting patient safety in hospitals, but the lack of progress in other conditions both medical and surgical is rather disappointing.

There is good quality evidence suggesting how hospitals may make changes to improve patient safety; these steps may include adopting care practices and protocols such as pressure ulcer monitoring and prevention protocols, fall prevention protocols, safety checklists, models for older adults inpatient care such as Mobile Acute Care of Elderly teams [6] and Acute Care for the Elderly models [7], quality improvement initiatives, and incorporation of information systems for data tracking and reporting, to name a few. How hospitals adopt different practices for the care of patients with different conditions may explain the study findings. The challenge is to figure out why noncardiac conditions do not have improving trends in patient safety and to demonstrate what works (and what doesn’t) on the hospital level. Understanding how care is delivered on the hospital level and correlating hospital level practices with patient outcomes from databases such as MPSMS may yield clues as to what specific steps hospitals have taken that have yielded changes in patient safety.

Applications for Clinical Practice

This study highlights trends in adverse events among hospitalized older adults that demonstrated improvements for patients with cardiac conditions but not for others. Future studies need to focus on understanding what works and what doesn’t so that hospitals can adopt safety practices that improve outcomes for older hospitalized patients.

—William Hung, MD, MPH

Next Article: