Invited commentary

Abstract

The authors have presented the results of an analysis from a unique dataset, the United Kingdom Heart Valve Registry. They have analyzed mortality rates in an extremely large cohort of valvular surgery patients operated on over a 15-year time period. The purpose was to determine if one-year mortality is a more meaningful outcome statistic for patients than 30-day mortality. Unfortunately this question cannot be answered without interviewing patients and measuring their reactions to different expressions of postoperative mortality rates. However, the authors have succeeded in illustrating the fact that 30-day mortality rates significantly underestimate the early high-risk hazard phase of heart valve surgery.The most important finding from the current study is that one-year mortality after heart valve surgery is approximately twice that of 30-day mortality, a finding that has not been previously reported in such a large cohort of patients. It has been well described that there is an early (less than 30 days postoperatively) high-risk hazard phase, followed by a low-risk constant phase after valvular surgery. Indeed, this is true of most surgical procedures. However, Figure 1 of the current study reveals that there is an intermediate-risk hazard phase between 30 days and one year postoperatively. The reasons for this are not clear since the authors lack detailed information regarding preoperative risk factors and the precise causes of death. Possible explanations include preoperative impaired ventricular function or significant comorbidities, as well as postoperative complications resulting in delayed death (eg. stroke, renal failure).The authors present 30-day mortality rates as a ratio of one-year mortality for several different patient subgroups (Table 1), but the utility of this measurement is unclear. They also present risk factors for early and intermediate mortality in Table 2. However, the UK Heart Valve Registry does not contain information on several important predictors of mortality (eg. left ventricular function, heart failure classification) making the interpretation of this analysis difficult. In addition, it has been well documented that the results for valve surgery are improving over time and therefore the hazard profiles may have changed over the 15-year time period of this study.In summary, the authors have successfully demonstrated that 30-day mortality rates significantly underestimate the early, increased risk phase of heart valve surgery. Longer followup periods may be necessary when reporting the early results of valvular operations. The authors have also supplied us with important, generalizable mortality data on a very large number of heart valve patients, making it likely that this paper will be frequently scrutinized and referenced in the future. The authors have presented the results of an analysis from a unique dataset, the United Kingdom Heart Valve Registry. They have analyzed mortality rates in an extremely large cohort of valvular surgery patients operated on over a 15-year time period. The purpose was to determine if one-year mortality is a more meaningful outcome statistic for patients than 30-day mortality. Unfortunately this question cannot be answered without interviewing patients and measuring their reactions to different expressions of postoperative mortality rates. However, the authors have succeeded in illustrating the fact that 30-day mortality rates significantly underestimate the early high-risk hazard phase of heart valve surgery. The most important finding from the current study is that one-year mortality after heart valve surgery is approximately twice that of 30-day mortality, a finding that has not been previously reported in such a large cohort of patients. It has been well described that there is an early (less than 30 days postoperatively) high-risk hazard phase, followed by a low-risk constant phase after valvular surgery. Indeed, this is true of most surgical procedures. However, Figure 1 of the current study reveals that there is an intermediate-risk hazard phase between 30 days and one year postoperatively. The reasons for this are not clear since the authors lack detailed information regarding preoperative risk factors and the precise causes of death. Possible explanations include preoperative impaired ventricular function or significant comorbidities, as well as postoperative complications resulting in delayed death (eg. stroke, renal failure). The authors present 30-day mortality rates as a ratio of one-year mortality for several different patient subgroups (Table 1), but the utility of this measurement is unclear. They also present risk factors for early and intermediate mortality in Table 2. However, the UK Heart Valve Registry does not contain information on several important predictors of mortality (eg. left ventricular function, heart failure classification) making the interpretation of this analysis difficult. In addition, it has been well documented that the results for valve surgery are improving over time and therefore the hazard profiles may have changed over the 15-year time period of this study. In summary, the authors have successfully demonstrated that 30-day mortality rates significantly underestimate the early, increased risk phase of heart valve surgery. Longer followup periods may be necessary when reporting the early results of valvular operations. The authors have also supplied us with important, generalizable mortality data on a very large number of heart valve patients, making it likely that this paper will be frequently scrutinized and referenced in the future.