Changing trends in mechanical circulatory assistance:.

BACKGROUND: The success of long-term implantable ventricular assist devices has led to their increased use in patients previously thought to be unsuitable for mechanical support. As a result, the demographic profile of patients presenting for LVAD support has changed over time. We reviewed our institutional experience to identify emerging risk factors and changing trends in patients who received the HeartMate VE LVAD. METHODS: The clinical records were reviewed of 131 consecutive LVAD recipients between 1996 and 2001. All perioperative data were collected prospectively and entered into an institutional database. All patients received a preoperative risk stratification score based upon published criteria. The cohort was arbitrarily divided into early (n = 45), mid (n = 45), and late groups (n = 41). RESULTS: Overall operative mortality was 25% with no difference between groups. The mean risk score increased significantly over time (early 3.5 +/- 0.4 vs. late 5.3 +/- 0.3, p < 0.05). The proportion of patients at high risk for mortality (score >5) was significantly higher in the late group (51% vs. 29%, p < 0.05). Although ventilation time and ICU stay was similar for all groups, hospital stay was longer in the late group (43 days vs. 23 days, p < 0.05). Mean duration of support fell from 90 to 59 days, but this failed to achieve statistical significance. Out-patient therapy decreased from 73% in the early group to 15% in the late group (p < 0.001). Multivariate analysis identified right heart failure (odds ratio 4.1, 95% CI 2-11) and risk score (OR 1.4, 95% CI 1.2-1.6) as independent predictors of death. CONCLUSIONS: Despite an increasingly high risk patient population, the mortality associated with LVAD therapy has remained constant. Duration of LVAD support has decreased with a trend toward transplantation before hospital discharge. These data continue to support the aggressive institution of mechanical assistance for acute or chronic heart failure.

Revised screening scale to predict survival after insertion of a left ventricular assist device.

BACKGROUND: We previously calculated a risk factor summation score that successfully predicted survival after insertion of a left ventricular assist device. We sought to validate our previous score by using a single center's clinical experience and to determine emerging risk factors for mortality after device insertion. METHODS: The clinical records of 130 consecutive patients who received the HeartMate VE left ventricular assist device (Thoratec Corp, Pleasanton, Calif) at our institution between June 1996 and March 2001 as a bridge to transplantation were reviewed. Univariate and multivariable analyses were performed to determine the predictors of operative mortality after device insertion. Using the relative risks for each identified variable, we devised a new risk factor summation score. The new and old scores were then compared by using linear regression analyses to determine whether the revised score improved statistical accuracy. RESULTS: Overall operative mortality was 25% (n = 33). The old score successfully predicted operative mortality in the current patient population (operative mortality of 38% for score >5 vs 13% for score < or =5). However, the revised score improved risk discrimination (operative mortality of 46% for a score >5 vs 12% for a score < or = to 5). Statistical accuracy was comparable between scores, but the relationship between observed and predicted outcomes was improved with the revised score. CONCLUSIONS: The changing demographic profile and management of patients presenting for mechanical circulatory support has led to a change in the predictors of mortality after device insertion. Periodic remodeling and recalibration of risk indices helps to accurately predict outcomes in high-risk patient groups and identifies emerging risk factors for mortality.