Incidence, Determinants and Mortality of Heart Failure Associated With Medical-Surgical Procedures in Patients ≥ 65 Years of Age (from the Cardiovascular Health Study)

Abstract

Heart failure (HF) and myocardial infarction are serious complications of major noncardiac surgery in older adults. Many factors can contribute to the development of HF during the postoperative period. The incidence of, and risk factors for, procedure-associated heart failure (PHF) occurring at the time of, or shortly after, medical procedures in a population-based sample ≥ 65 years of age have not been fully characterized, particularly in comparison with HF not proximate to medical procedures. This analysis comprises 5,121 men and women free of HF at baseline from the Cardiovascular Health Study who were followed up for 12.0 years (median). HF events were documented by self-report at semi-annual contacts and confirmed by a formal adjudication committee using a review of the participants' medical records and standardized criteria for HF. Incident HF events were additionally adjudicated as either being related or unrelated to a medical procedure (PHF and non-PHF, respectively). We estimated cause-specific hazards ratios for the association of covariates with PHF and non-PHF. There were 1,728 incident HF events in the primary analysis: 168 (10%) classified as PHF, 1,526 (88%) as non-PHF, and 34 unclassified (2%). For those 1,045 participants in whom LV ejection fraction was known at the time of the HF event, it was ≥45% in 89 of 118 participants (75%) with PHF, compared to 517 of 927 participants (55%) with non-PHF (p < 0.001). Increased age, male gender, diabetes, and angina at baseline were associated with both PHF and non-PHF (range of hazard ratios (HR): 1.04-2.05]. Being Black was inversely associated with PHF [HR: 0.46, 95% confidence interval: 0.25-0.86]. Participants with increased age, without baseline angina, and with baseline LVEF<55% were at a significantly lower risk for PHF compared to non-PHF. Among those with PHF, surgical procedures-including cardiac, orthopedic, gastrointestinal, vascular, and urologic-comprised 83.3%, while percutaneous procedures comprised 8.9% (including 6.5% represented by cardiac catheterizations and pacemaker placements). Another group composed of a variety of procedures commonly requiring large fluid volume administration comprised 7.7%. There was a lower all-cause 30-day mortality in the PHF versus the non-PHF group (2.2% vs 5.7%), with a nonsignificant odds ratio of 0.39 in a minimally adjusted model. When individuals with prior myocardial infarction (MI) were excluded in a sensitivity analysis, the proportion of incident HF with concurrent MI was greater for PHF (32.9%) than for non-PHF (19.8%). In conclusion, PHF in older adults is a common entity with relatively low 30-day mortality. Baseline angina, lower age, and LVEF ≥ 55% were associated with a higher risk of PHF compared to non-PHF. Being Black was associated with a lower risk of PHF and PHF as a proportion of HF was lower in Black than in non-Black participants. Compared to non-PHF, PHF more frequently presented with concurrent MI and with preserved LV ejection fraction.