|
|
108, 687 Isolated Aortic Valve Replacements Over 10 years in North America:Changing Risk, Valve type and Outcome in the Society of Thoracic Surgeons Database
j. M. brown1, S. O'Brien2, J. Sikora1, B. P. Griffith1, J. S. Gammie1. 1University of maryland, baltimore, MD, 2Duke Clinical Research Institute, Durham ,North Carolina, NC,
Background
Aortic valve replacement (AVR) is performed more than 200,000 times per year worldwide. The United States elderly population is expanding. There will soon be more than 200,000 octogenarians with severe Aortic stenosis. Our purpose was to describe the changes which have occurred in the US population of patients receiving aortic valve replacement over 10 years and analyze the outcome from AVR with regard to age , gender, and ejection fraction.
Methods
The society of thoracic surgeons database was queried to identify all isolated AVR performed between January 1, 1997 and December 31st , 2006. After excluding endocarditis (N = 6,372) and patients missing age or gender data(N = 104), a sample of 108,687 isolated AVR was analyzed. The frequency of in-hospital mortality, permanent stroke, and the median post-operative length of stay were compared across patient subgroups based on gender, age (≤75 vs. >75), and ejection fraction (<30 vs >30). Time related trends were assessed by comparing the distribution of risk factors, valve types, and outcomes in 1997 vs. 2006. Differences in case mix over time were summarized by comparing the average predicted mortality risk in 1997 vs. 2006 using a logistic regression model that included 10 patient explanatory variables. Differences across subgroups and time were assessed using chi-square tests for categorical variables and Wilcoxon tests for continuous variables.
Results
Female gender, age over 75 and ejection fraction less than 30 all were related to more frequent mortality, stroke rate, and longer post-op length of stay ( table 1). Over time, there was a dramatic shift toward use of a bioprosthetic valve ( figure 1). In addition,when compared to 1997, AVR patients in 2006 were older ( mean age 65.9 vs 67.9; p<0.001) and had higher predicted risk for operative mortality (2.75 vs. 3.25; p<0.001). By contrast both observed mortality and the rate of permanent stroke fell by 24 % and 27 % respectively during the decade. There was a 39 % reduction in the rate of death in the setting of preoperative renal failure.
| mortality , permanent stroke and length of stay in subgroups | | female | males | p value | age > 75 | age < 75 | p value | ejection fraction <30 | ejection fraction>30 | p value | | In hospital Mortality (%) | 3.97 | 2.80 | <.001 | 4.85 | 1.96 | <.001 | 5.52 | 2.67 | <.001 | | Permanent Stroke (%) | 1.84 | 1.34 | <.001 | 2.43 | 1.10 | <.001 | 2.00 | 1.51 | <.001 | | Post-op Length of stay (median) | 7.80 | 7.40 | <.001 | 9.23 | 7.06 | <.001 | 9.53 | 7.67 | <.001 |
| A decade of change in Aortic valve replacement | | 1997 | 2006 | p value | | number of AVRs | 9497 | 15397 | NA | | % bioprostheses | 45.2 | 78.7 | <.001 | | % patients >75 | 29.8 | 37.3 | <.001 | | % patients with preop renal failure | 4.3 | 5.2 | <.001 | | pedicted mortality (%) | 2.75 | 3.25 | <.001 | | observed mortality (%) | 3.37 | 2.55 | <.001 | | Ratio observed:predicted mortality | 1.15 | 0.85 | <.001 | | permanent stroke % | 1.83 | 1.33 | <.001 | | mortality with preop renal failure | 13.91 | 8.48 | <.001 |
Conclusions
1. There has been a reduction in the morbidity and mortality of isolated AVR, despite a gradual increase in age and overall risk profile of the population having surgery.
2. There has been a shift toward the implantation of tissue, rather than mechanical prostheses among patients having AVR.
3. Women, patients over 75, and patients with ejection fraction under 30 have worse outcome from isolated AVR regarding mortality, stroke , and length of stay.
Back to 2008 Annual Meeting
Back to Program Outline
|
