Mortality and Readmission of Patients With Heart Failure, Atrial Fibrillation, or Coronary Artery Disease Undergoing Noncardiac Surgery: An Analysis of 38 047 Patients

Study Questions:

What is the postoperative mortality rate of patients with heart failure (HF), atrial fibrillation (AF), or coronary artery disease (CAD) undergoing noncardiac surgery?


The authors created an inception cohort of patients with incident HF, AF, or CAD, by linking the Alberta Inpatient Discharge Abstracts Database, the Alberta Health Care Insurance Registry, and the Ambulatory Care Classification System for the province of Alberta, and then evaluated outcomes during subsequent noncardiac surgery. Subjects were divided into four groups based on cardiac diagnosis: nonischemic HF (NIHF; n = 7,700), ischemic HF (IHF; n = 12,249), CAD (n = 13,786), or AF (n = 4,312). These diagnoses were identified by ICD-9 codes. Subsequent noncardiac surgery occurred between April 1, 1999 and September 31, 2006. Primary outcome was 30-day postoperative mortality.


Unadjusted 30-day mortality was 9.3% for NIHF, 9.2% for IHF, 2.9% for CAD, and 6.4% for AF (p < 0.0001 for each group vs. CAD). Similar results were seen looking at 30-day postoperative mortality after minor surgical procedures: 8.5% for NIHF, 8.1% for IHF, 2.3% for CAD, and 5.7% for AF (p < 0.0001 for each group versus CAD). Multivariable adjustment suggested postoperative mortality was higher for all three groups compared with CAD: (NIHF odds ratio [OR], 2.92; 95% confidence interval [CI], 2.44-3.48); (IHF OR, 1.98; 95% CI, 1.70-2.31); (AF OR, 1.69; 95%, CI 1.34-2.14).


The authors concluded that, although current perioperative risk prediction models place greater emphasis on CAD than HF or AF, patients with HF or AF had significantly higher risk of postoperative mortality than patients with CAD, even after minor surgical procedures.


This interesting study benefits greatly from the existence of comprehensive administrative databases for the province of Alberta, allowing the investigators to make observations on population-wide cohorts for the entire province. This important study highlights the significant perioperative risk faced by patients with a diagnosis of HF or AF undergoing noncardiac surgery. This observation is based, however, on the inclusion of all subjects who had any diagnosis of either HF or AF—a clearly very heterogeneous group. Nonetheless, this serves to remind practitioners that these diagnoses (HF and AF) can be related to significant perioperative risk. The conclusion of this paper that patients with HF or AF face greater mortality risk than CAD patients undergoing noncardiac surgery may be somewhat misleading, however. Perioperative risk prediction models mostly identify only CAD patients who are actively symptomatic or have a history of myocardial infarction as being at increased risk for perioperative complications. This is clearly only a high-risk subset of all patients with an ICD-9 diagnosis of CAD, which was the cohort studied in this paper. Thus, it is not known if existing perioperative risk prediction models may identify patients with CAD-associated perioperative risk that exceeds that observed with a preoperative diagnosis of HF or AF. Further study is needed to parse out the relative incremental perioperative risk associated with additional diagnoses of HF or AF.

Clinical Topics: Heart Failure and Cardiomyopathies, Acute Heart Failure

Keywords: Coronary Artery Disease, Myocardial Infarction, Heart Failure, Surgical Procedures, Minor

< Back to Listings