Cost-Effectiveness of Transcatheter Aortic Valve Replacement Compared With Standard Care Among Inoperable Patients With Severe Aortic Stenosis
2012; Lippincott Williams & Wilkins; Volume: 125; Issue: 9 Linguagem: Inglês
10.1161/circulationaha.111.054072
ISSN1524-4539
AutoresMatthew R. Reynolds, Elizabeth A. Magnuson, Kaijun Wang, Yang Lei, Katherine Vilain, Joshua Walczak, Susheel Kodali, John M. Lasala, William W. O’Neill, Charles J. Davidson, Craig R. Smith, Martin B. Leon, David J. Cohen,
Tópico(s)Cardiac pacing and defibrillation studies
ResumoHomeCirculationVol. 125, No. 9Cost-Effectiveness of Transcatheter Aortic Valve Replacement Compared With Standard Care Among Inoperable Patients With Severe Aortic Stenosis Free AccessResearch ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessResearch ArticlePDF/EPUBCost-Effectiveness of Transcatheter Aortic Valve Replacement Compared With Standard Care Among Inoperable Patients With Severe Aortic StenosisResults From the Placement of Aortic Transcatheter Valves (PARTNER) Trial (Cohort B) Matthew R. Reynolds, MD, MSc, Elizabeth A. Magnuson, ScD, Kaijun Wang, PhD, Yang Lei, MSc, Katherine Vilain, MPH, Joshua Walczak, MS, Susheel K. Kodali, MD, John M. Lasala, MD, PhD, William W. O'Neill, MD, Charles J. Davidson, MD, Craig R. Smith, MD, Martin B. Leon, MD and David J. Cohen, MD, MSc Matthew R. ReynoldsMatthew R. Reynolds From the Harvard Clinical Research Institute, Boston, MA (M.R.R., J.W.); Boston VA Healthcare System, Boston, MA (M.R.R.); Saint Luke's Mid America Heart & Vascular Institute, University of Missouri–Kansas City, Kansas City, MO (E.A.M., K.W., Y.L., K.V., D.J.C.); Columbia-Presbyterian Hospital, New York, NY (S.K.K., C.R.S., M.B.L.); Washington University School of Medicine, St. Louis, MO (J.M.L.); University of Miami School of Medicine, Miami, FL (W.W.O.); and Northwestern University School of Medicine, Chicago, IL (C.J.D.). , Elizabeth A. MagnusonElizabeth A. Magnuson From the Harvard Clinical Research Institute, Boston, MA (M.R.R., J.W.); Boston VA Healthcare System, Boston, MA (M.R.R.); Saint Luke's Mid America Heart & Vascular Institute, University of Missouri–Kansas City, Kansas City, MO (E.A.M., K.W., Y.L., K.V., D.J.C.); Columbia-Presbyterian Hospital, New York, NY (S.K.K., C.R.S., M.B.L.); Washington University School of Medicine, St. Louis, MO (J.M.L.); University of Miami School of Medicine, Miami, FL (W.W.O.); and Northwestern University School of Medicine, Chicago, IL (C.J.D.). , Kaijun WangKaijun Wang From the Harvard Clinical Research Institute, Boston, MA (M.R.R., J.W.); Boston VA Healthcare System, Boston, MA (M.R.R.); Saint Luke's Mid America Heart & Vascular Institute, University of Missouri–Kansas City, Kansas City, MO (E.A.M., K.W., Y.L., K.V., D.J.C.); Columbia-Presbyterian Hospital, New York, NY (S.K.K., C.R.S., M.B.L.); Washington University School of Medicine, St. Louis, MO (J.M.L.); University of Miami School of Medicine, Miami, FL (W.W.O.); and Northwestern University School of Medicine, Chicago, IL (C.J.D.). , Yang LeiYang Lei From the Harvard Clinical Research Institute, Boston, MA (M.R.R., J.W.); Boston VA Healthcare System, Boston, MA (M.R.R.); Saint Luke's Mid America Heart & Vascular Institute, University of Missouri–Kansas City, Kansas City, MO (E.A.M., K.W., Y.L., K.V., D.J.C.); Columbia-Presbyterian Hospital, New York, NY (S.K.K., C.R.S., M.B.L.); Washington University School of Medicine, St. Louis, MO (J.M.L.); University of Miami School of Medicine, Miami, FL (W.W.O.); and Northwestern University School of Medicine, Chicago, IL (C.J.D.). , Katherine VilainKatherine Vilain From the Harvard Clinical Research Institute, Boston, MA (M.R.R., J.W.); Boston VA Healthcare System, Boston, MA (M.R.R.); Saint Luke's Mid America Heart & Vascular Institute, University of Missouri–Kansas City, Kansas City, MO (E.A.M., K.W., Y.L., K.V., D.J.C.); Columbia-Presbyterian Hospital, New York, NY (S.K.K., C.R.S., M.B.L.); Washington University School of Medicine, St. Louis, MO (J.M.L.); University of Miami School of Medicine, Miami, FL (W.W.O.); and Northwestern University School of Medicine, Chicago, IL (C.J.D.). , Joshua WalczakJoshua Walczak From the Harvard Clinical Research Institute, Boston, MA (M.R.R., J.W.); Boston VA Healthcare System, Boston, MA (M.R.R.); Saint Luke's Mid America Heart & Vascular Institute, University of Missouri–Kansas City, Kansas City, MO (E.A.M., K.W., Y.L., K.V., D.J.C.); Columbia-Presbyterian Hospital, New York, NY (S.K.K., C.R.S., M.B.L.); Washington University School of Medicine, St. Louis, MO (J.M.L.); University of Miami School of Medicine, Miami, FL (W.W.O.); and Northwestern University School of Medicine, Chicago, IL (C.J.D.). , Susheel K. KodaliSusheel K. Kodali From the Harvard Clinical Research Institute, Boston, MA (M.R.R., J.W.); Boston VA Healthcare System, Boston, MA (M.R.R.); Saint Luke's Mid America Heart & Vascular Institute, University of Missouri–Kansas City, Kansas City, MO (E.A.M., K.W., Y.L., K.V., D.J.C.); Columbia-Presbyterian Hospital, New York, NY (S.K.K., C.R.S., M.B.L.); Washington University School of Medicine, St. Louis, MO (J.M.L.); University of Miami School of Medicine, Miami, FL (W.W.O.); and Northwestern University School of Medicine, Chicago, IL (C.J.D.). , John M. LasalaJohn M. Lasala From the Harvard Clinical Research Institute, Boston, MA (M.R.R., J.W.); Boston VA Healthcare System, Boston, MA (M.R.R.); Saint Luke's Mid America Heart & Vascular Institute, University of Missouri–Kansas City, Kansas City, MO (E.A.M., K.W., Y.L., K.V., D.J.C.); Columbia-Presbyterian Hospital, New York, NY (S.K.K., C.R.S., M.B.L.); Washington University School of Medicine, St. Louis, MO (J.M.L.); University of Miami School of Medicine, Miami, FL (W.W.O.); and Northwestern University School of Medicine, Chicago, IL (C.J.D.). , William W. O'NeillWilliam W. O'Neill From the Harvard Clinical Research Institute, Boston, MA (M.R.R., J.W.); Boston VA Healthcare System, Boston, MA (M.R.R.); Saint Luke's Mid America Heart & Vascular Institute, University of Missouri–Kansas City, Kansas City, MO (E.A.M., K.W., Y.L., K.V., D.J.C.); Columbia-Presbyterian Hospital, New York, NY (S.K.K., C.R.S., M.B.L.); Washington University School of Medicine, St. Louis, MO (J.M.L.); University of Miami School of Medicine, Miami, FL (W.W.O.); and Northwestern University School of Medicine, Chicago, IL (C.J.D.). , Charles J. DavidsonCharles J. Davidson From the Harvard Clinical Research Institute, Boston, MA (M.R.R., J.W.); Boston VA Healthcare System, Boston, MA (M.R.R.); Saint Luke's Mid America Heart & Vascular Institute, University of Missouri–Kansas City, Kansas City, MO (E.A.M., K.W., Y.L., K.V., D.J.C.); Columbia-Presbyterian Hospital, New York, NY (S.K.K., C.R.S., M.B.L.); Washington University School of Medicine, St. Louis, MO (J.M.L.); University of Miami School of Medicine, Miami, FL (W.W.O.); and Northwestern University School of Medicine, Chicago, IL (C.J.D.). , Craig R. SmithCraig R. Smith From the Harvard Clinical Research Institute, Boston, MA (M.R.R., J.W.); Boston VA Healthcare System, Boston, MA (M.R.R.); Saint Luke's Mid America Heart & Vascular Institute, University of Missouri–Kansas City, Kansas City, MO (E.A.M., K.W., Y.L., K.V., D.J.C.); Columbia-Presbyterian Hospital, New York, NY (S.K.K., C.R.S., M.B.L.); Washington University School of Medicine, St. Louis, MO (J.M.L.); University of Miami School of Medicine, Miami, FL (W.W.O.); and Northwestern University School of Medicine, Chicago, IL (C.J.D.). , Martin B. LeonMartin B. Leon From the Harvard Clinical Research Institute, Boston, MA (M.R.R., J.W.); Boston VA Healthcare System, Boston, MA (M.R.R.); Saint Luke's Mid America Heart & Vascular Institute, University of Missouri–Kansas City, Kansas City, MO (E.A.M., K.W., Y.L., K.V., D.J.C.); Columbia-Presbyterian Hospital, New York, NY (S.K.K., C.R.S., M.B.L.); Washington University School of Medicine, St. Louis, MO (J.M.L.); University of Miami School of Medicine, Miami, FL (W.W.O.); and Northwestern University School of Medicine, Chicago, IL (C.J.D.). and David J. CohenDavid J. Cohen From the Harvard Clinical Research Institute, Boston, MA (M.R.R., J.W.); Boston VA Healthcare System, Boston, MA (M.R.R.); Saint Luke's Mid America Heart & Vascular Institute, University of Missouri–Kansas City, Kansas City, MO (E.A.M., K.W., Y.L., K.V., D.J.C.); Columbia-Presbyterian Hospital, New York, NY (S.K.K., C.R.S., M.B.L.); Washington University School of Medicine, St. Louis, MO (J.M.L.); University of Miami School of Medicine, Miami, FL (W.W.O.); and Northwestern University School of Medicine, Chicago, IL (C.J.D.). and on behalf of the PARTNER Investigators Originally published3 Feb 2012https://doi.org/10.1161/CIRCULATIONAHA.111.054072Circulation. 2012;125:1102–1109is corrected byCorrectionOther version(s) of this articleYou are viewing the most recent version of this article. Previous versions: January 1, 2012: Previous Version 1 AbstractBackground—In patients with severe aortic stenosis who cannot have surgery, transcatheter aortic valve replacement (TAVR) has been shown to improve survival and quality of life compared with standard therapy, but the costs and cost-effectiveness of this strategy are not yet known.Methods and Results—The PARTNER trial randomized patients with symptomatic, severe aortic stenosis who were not candidates for surgery to TAVR (n=179) or standard therapy (n=179). Empirical data regarding survival, quality of life, medical resource use, and hospital costs were collected during the trial and used to project life expectancy, quality-adjusted life expectancy, and lifetime medical care costs to estimate the incremental cost-effectiveness of TAVR from a US perspective. For patients treated with TAVR, mean costs for the initial procedure and hospitalization were $42 806 and $78 542, respectively. Follow-up costs through 12 months were lower with TAVR ($29 289 versus $53 621) because of reduced hospitalization rates, but cumulative 1-year costs remained higher ($106 076 versus $53 621). We projected that over a patient's lifetime, TAVR would increase discounted life expectancy by 1.6 years (1.3 quality-adjusted life-years) at an incremental cost of $79 837. The incremental cost-effectiveness ratio for TAVR was thus estimated at $50 200 per year of life gained or $61 889 per quality-adjusted life-year gained. These results were stable across a broad range of uncertainty and sensitivity analyses.Conclusions—For patients with severe aortic stenosis who are not candidates for surgery, TAVR increases life expectancy at an incremental cost per life-year gained well within accepted values for commonly used cardiovascular technologies.Clinical Trial Registration—URL: http://www.clinicaltrials.gov. Unique identifier: NCT00530894.IntroductionValvular aortic stenosis occurs most commonly among the elderly and, in the absence of definitive treatment, leads to progressive symptoms, functional decline, and death.1,2 Nonetheless, many patients with severe aortic stenosis do not undergo surgical valve replacement because of both cardiovascular and noncardiovascular comorbidities that result in unacceptable surgical risk.3–5 Recently, the Placement of Aortic Transcatheter Valves (PARTNER) trial reported that in a cohort of patients who were unsuitable for surgical valve replacement (cohort B), transcatheter aortic valve replacement (TAVR), compared with standard nonsurgical care, resulted in a 20% reduction in mortality at 12 months, as well as improved functional status and a reduction in hospital admissions for aortic stenosis.6Editorial see p 1076Clinical Perspective on p 1109New technologies are often cited as a major contributor to increasing healthcare costs.7 Before a new technology or clinical strategy is widely adopted, it is therefore important to understand the clinical and economic benefits that any increased up-front expenditures may yield. Given the advanced age and multiple comorbid conditions that characterize patients with high surgical risk for surgical valve replacement, the question of whether TAVR can provide meaningful health benefits to the population at an acceptable cost is particularly germane. To address these questions, we conducted a preplanned health economic study alongside the PARTNER trial, with the goal of understanding the incremental costs and cost-effectiveness of TAVR compared with standard therapy among inoperable patients with severe aortic stenosis.MethodsStudy PopulationOne-year clinical results from the PARTNER trial (cohort B) have been published previously.6 Briefly, the trial enrolled adults with severe aortic stenosis, New York Heart Association functional class ≥2, and high surgical risk based on the Society for Thoracic Surgeons risk score8 or other anatomic or technical factors. These patients were determined not to be suitable surgical candidates on the basis of evaluation by at least 2 surgical investigators and the trial's executive committee. Patients were randomized to TAVR via the transfemoral route (n=179) or standard nonsurgical therapy (n=179), which could include balloon aortic valvuloplasty at the discretion of the treating physician. The study was approved by each enrolling center's institutional review board, and all patients provided written informed consent. Of the 358 randomized patients, 234 (65%) enrolled at 17 US centers additionally consented to the collection of hospital billing data.Analytic OverviewAll randomized subjects were included in the present study and analyzed according to intention to treat. Our analysis was performed from the perspective of the US healthcare system (ie, a modified societal perspective) and consisted of 2 main components. Data on survival, quality of life, healthcare resource use, and hospital charges were collected through the first 12 months of follow-up (the minimum follow-up duration for the trial) for all patients and were used to calculate survival, quality-adjusted survival, and costs for the trial period. The empirical 12-month data for costs and quality of life, along with all of the available data on survival (up to a maximum of 30 months), were then used to project outcomes beyond the trial, from which estimates of life-years, quality-adjusted life-years (QALYs), and lifetime costs were developed for each patient who survived the trial period. These estimates were then aggregated to calculate average costs and benefits (and their associated distributions) at the treatment-group level.Determination of Medical Care CostsMedical care costs were assessed from the perspective of the US healthcare system by use of a combination of resource-based accounting and hospital billing data, as described previously,9,10 and are reported in 2010 US dollars. Costs from years before 2010 were converted to 2010 dollars with the medical care component of the Consumer Price Index.TAVR Procedure CostsFor the initial TAVR procedure, study sites recorded procedure duration and counts of major items consumed, such as support wires, guiding catheters, valvuloplasty balloons, Edwards SAPIEN valve systems, temporary pacing catheters, and vascular closure devices. Costs for each procedure were calculated by multiplying item counts by their respective unit prices, determined by the average acquisition costs at a sample of US hospitals. An estimated US commercial price for the Edwards SAPIEN valve system of $30 000 was used for the primary analysis.Other Index Hospital CostsCosts for the remainder of each initial hospital stay for TAVR were derived from hospital bills, which were available for 121 of the 175 patients who underwent an attempted TAVR procedure (97% of patients who agreed to participate in billing data collection from 16 US study hospitals). After the exclusion of charges for care received before randomization and charges for the index TAVR procedure itself, all remaining hospital charges were converted to costs by use of cost-center–specific cost-to-charge ratios obtained from each enrolling hospital's Medicare cost report.11 When bills were unavailable, the costs of hospital care were estimated with a linear regression model derived from the patients with complete billing data (model R2=0.84). Covariates included in the model included total intensive care unit (ICU) and non-ICU length of stay, in-hospital death, in-hospital acute renal failure, and major vascular complication. Use of alternative models, including linear regression of log-transformed costs (with retransformation to natural units), yielded results that were virtually identical.Follow-Up Hospital CareSites collected information on follow-up hospital admissions for any cause at scheduled follow-up visits (1, 6, and 12 months) and on learning of adverse events. Costs for subsequent hospital admissions were calculated from billing data with hospital and cost-center–specific cost-to-charge ratios when bills were available (54% of admissions). When bills were not available (generally because of admission to nonstudy hospitals or to hospitals that do not produce standard billing data), diagnosis, procedure, and adverse event information from the study database were used to assign each admission to a unique Medicare Severity-Adjusted Diagnosis Related Group (MS-DRG). Average reimbursements for each respective MS-DRG, based on 2008 Medicare Provider Analysis and Review (MedPAR) data,12 were used as the proxy for admission costs in these cases.Physician FeesEstimated physician fees for the index TAVR procedure were taken from the Medicare fee schedule and included a primary operator (current fees for surgical aortic valve replacement were used for this unknown value), plus fees for a surgical assistant, cardiac anesthesia (based on measured procedure duration), and intraoperative transesophageal echocardiography. Physician fees for initial consultation and daily care during the remainder of the initial hospital stay and for any additional cardiovascular procedures performed during the index hospitalization (eg, vascular surgery, endovascular stenting) were also taken from the Medicare fee schedule. For follow-up hospitalizations, physician fees were estimated based on the DRG for each admission as described previously.13Other CostsData on rehabilitation facility stays, nursing home stays, and outpatient resource use (emergency room visits, physician office visits, outpatient cardiac testing) were collected by the enrolling sites at each study follow-up visit. These measures of resource use were converted to costs using national average per diem rates for residential care and Medicare reimbursement rates for outpatient care based on the Medicare Fee Schedule.Cost-Effectiveness AnalysisWe evaluated cost-effectiveness over a lifetime horizon in terms of both cost per year of life gained (primary analysis) and cost per QALY gained (secondary analysis). These analyses required the projection of life expectancy, quality-adjusted life expectancy, and costs over the anticipated life expectancy of each patient who remained alive at the completion of the trial.Life Expectancy EstimationSurvival analyses were performed with a locked data set as of September 28, 2010, with a minimum follow-up duration of 12 months, a maximum follow-up duration of 30 months, and mean follow-up duration among survivors of 18 months. To estimate life expectancy for each surviving patient, we used parametric survival models to extrapolate survival probabilities beyond the follow-up time of the trial. Survival curves were fitted separately for the TAVR and control groups by use of exponential, Weibull, log-normal, log-logistic, logistic, and normal models. Covariates included age, sex, and medical history such as diabetes mellitus, coronary artery disease, peripheral vein disease, myocardial infarction, stroke/transient ischemic attack, prior percutaneous coronary intervention, and prior coronary artery bypass graft. To improve the model fit for the TAVR group and to optimize the resulting survival projections, we conditioned the model on survival at 3 months to reduce the influence of periprocedural events not expected to affect long-term survival. Exponential models were identified as optimal for both treatment groups based on the Akaike Information Criterion and Schwarz's Bayesian Criterion and were used for the primary cost-effectiveness analysis. Alternative models were used as the basis for sensitivity analyses (see Statistical Analysis).From the final survival models, patient-level survival probabilities over time were generated until the estimated survival probability was <1%. Individual survival duration was then calculated as the integral of the survival probability versus time function.Quality-Adjusted Life ExpectancyQuality of life was assessed directly from patients at baseline, 1, 6, and 12 months with the EuroQOL (EQ-5D) health status instrument and converted to population-level utility weights with a published algorithm developed for the US population.14 Utility weights are measures of a person's strength of preference for his or her state of health on the basis of a scale from 0 to 1, where 0 represents the worst possible health state (usually death) and 1 represents ideal health. Quality-adjusted life expectancy was calculated for each patient as the time-weighted average of his or her utility values, with the midpoint between assessments used as the transition between health states.15 Missing utility values were estimated by multiple imputation techniques, taking into account baseline patient characteristics, clinical events, number of hospitalizations, and previous utility values. Quality-adjusted life expectancy beyond the first year of follow-up was calculated as the product of projected life expectancy multiplied by the last available utility value for that individual.Long-Term CostsMonthly healthcare costs (including hospital costs, physician fees, outpatient services, and chronic care/rehabilitation costs) beyond the trial period were estimated on the basis of the last 6 months of observed costs for each surviving patient by multiplying these cost estimates by each patient's projected survival duration beyond the trial.Statistical AnalysisCategorical data are reported as frequencies, and continuous data are reported as mean±SD. Discrete variables were compared by Fisher exact test. Normally distributed continuous variables were compared by Student t test, and nonnormally distributed data were compared by the Wilcoxon rank-sum test. Cost data are reported as both mean and median values and were compared by t tests, which are appropriate given the large sample size and our focus on comparing mean costs between groups (rather than the underlying distributions).16 All probability values were 2-sided.For the purposes of the cost-effectiveness analyses, future costs, life expectancy, and quality-adjusted life expectancy were discounted at 3% per year, consistent with current guidelines.17 Incremental cost-effectiveness ratios were calculated as the difference in mean discounted lifetime costs divided by the difference in mean discounted life expectancy or quality-adjusted life expectancy. Bootstrap resampling18 (5000 replications) was used to assess the joint distribution of lifetime cost and survival differences and to generate cost-effectiveness acceptability curves to explore the probability that TAVR would be economically attractive at any given cost-effectiveness threshold.In addition to the primary analysis, we performed a number of sensitivity analyses to explore the impact of key analytic and structural assumptions on the results of our study. These analyses included plausible variations in the discount rate and the acquisition cost of the transcatheter valve; exclusion of all noncardiovascular care costs; exclusion of the costs of balloon valvuloplasty procedures; and an assessment of QALYs with the assumption of no improvement in quality of life from baseline for either group. We also considered alternative hazard functions for the model used to project survival and estimate life expectancy after TAVR. These hazard functions included Weibull, Gompertz, and an "accelerated" Gompertz function in which the shape parameter was increased to "force" survival for the TAVI group to be 1% at 10 years. We also examined results obtained by truncating our base case analysis at 5 and 10 years. Finally, to address uncertainty in our long-term cost projections, we repeated our analysis after inflating and deflating all TAVR group costs beyond 12 months by 25%.ResultsBetween May 2007 and March 2009, a total of 358 patients with inoperable aortic stenosis were enrolled at 21 centers (17 US, 3 Canadian, 1 European) and randomized to either TAVR (n=179) or standard therapy (n=179). Of the 179 patients randomized to TAVR, 175 underwent a TAVR procedural attempt. Two patients died before their scheduled procedure, and in 2 other cases, the aortic annulus diameter was found to be unsuitable for TAVR by intraoperative transesophageal echocardiography, and the patients were instead treated with balloon aortic valvuloplasty.TAVR Procedural Resource Use and Index Hospitalization CostsResource use and costs for the initial TAVR procedures and their associated hospital stays are summarized in Tables 1 and 2. With few exceptions, the initial procedures used a single valvuloplasty balloon and a single Edwards-Sapien valve. In 21 patients, 1 or more unplanned procedures were performed, most commonly a surgical or catheter-based peripheral arterial intervention. The mean TAVR procedural cost, excluding physician fees, was $42 806 (median $38 706), and the mean cost for the initial TAVR admission, including physician fees, was $78 542 (median $67 551). Mean length of stay was 10.1 days, of which 8.6 days were after the procedure.Table 1. TAVR Procedural Resource Use and CostResourceUseUnit Cost, $Procedure duration, min150±8425.52/minTAVR devices, n (%)30 000 1164 (93.7) 210 (5.7) 31 (0.6)Valvuloplasty balloons, n1.3±0.6462Guiding catheters, n2.7±2.251Radiographic contrast, mL132±810.14/mLArterial site closure, n (%) Surgical146 (83)N/A Closure device33 (19)215Procedural costs, $ (median) Devices35 400±14 572 (31 631)… Room/overhead/personnel7406±2134 (7018)… Total42 806±15 206 (38 706)…TAVR indicates transcatheter aortic valve replacement; N/A, not applicable.Table 2. Resource Use and Costs for TAVR Hospitalizations (n=175)Mean±SD (Median)Length of stay, d ICU4.0±7.0 (2) Non-ICU6.1±5.4 (5) Postprocedure8.6±9.8 (6) Total10.1±10.1 (7)Costs, $ TAVR procedure42 806±15 206 (38 706) Room and ancillary costs30 757±27 484 (22 150) Physician fees4979±1697 (4521) Total for initial hospitalization78 542±33 799 (67 551)TAVR indicates transcatheter aortic valve replacement; ICU, intensive care unit.Follow-Up Resource Use and CostsFollow-up resource use and costs for the 2 treatment groups are summarized in Table 3. Over the first 12 months of follow-up, the mean number of hospital admissions per patient was reduced from 2.2 per patient for the control group to 1.0 per patient for the TAVR group (P<0.001), driven entirely by a reduction in cardiovascular hospitalizations. As a result, mean costs for follow-up hospital care were higher in the control group by $26 025 per patient ($44 099 versus $18 074, P<0.001). The total numbers of days spent in rehabilitation and skilled nursing facilities were each higher in the TAVR group, such that mean 12-month costs for residential care were ≈$2500 per patient higher in the TAVR group, although these differences were not statistically significant. Including the initial TAVR admissions, total 12-month medical care costs were approximately $52 000 per patient (95% confidence interval [CI] $40 635 to $64 275) higher in the TAVR group than in the control group ($106 076 versus $53 621, P<0.001).Table 3. Cumulative 1-Year Resource Use and CostsTAVR Group (n=179)Control Group (n=179)Difference (TAVR−Control) (95% CI)PFollow-up hospitalizations1.0±1.32.2±1.5−1.1 (−0.8, −1.4)<0.001 Cardiovascular0.5±0.81.7±1.2−1.2 (−1.0, −1.4)<0.001 Noncardiovascular0.5±0.80.5±0.80.1 (−0.1, 0.2)0.43Rehabilitation days4.6±21.33.8±18.70.7 (−3.5, 4.9)0.75SNF days14.4±56.57.9±39.56.5 (−3.6, 16.7)0.21Follow-up hospitalization costs, $18 074±35 32045 093±46 943−27 018 (−35 654, −18 383)<0.001Rehabilitation costs, $4674±21 8253951±19 209723 (−3551, 4997)0.74SNF costs, $4142±16 2252270±11 3441, 872 (−1038, 4782)0.21Other outpatient costs, $2400±25842308±271591 (−460, 642)0.74Total follow-up costs, $29 289±48 54253 621±53 301−24 331 (−34 929, −13 735)<0.001Total 12-mo costs, $106 076±60 20653 621±53 30152 455 (40 635, 64 275)<0.001TAVR indicates transcatheter aortic valve replacement; CI, confidence interval; and SNF, skilled nursing facility.All data are presented as mean±SD.EQ-5D ScoresMean baseline EQ-5D utility scores were 0.59 in the TAVR group and 0.57 in the control group. These increased to 0.71 at 30 days and 0.72 at 6 and 12 months in the TAVR group. Among surviving patients in the control group, EQ-5D scores also increased to 0.64 at 30 days, 0.66 at 6 months, and 0.62 at 1 year. The between-group differences in utility weights were statistically significant (P<0.05) at each follow-up time point.Projections Beyond 12 MonthsAs reported previously, 12-month survival was 70% for the TAVR group versus 50% for the control group, an absolute survival advantage of 20% that was preserved through 2.5 years of follow-up.6 Observed survival duration through a maximum of 30 months was 1.25 years with TAVR (95% CI, 1.15–1.36) and 0.88 years (95% CI, 0.78–0.97) with standard therapy, a difference of 0.36 years (95% CI 0.23–0.50). An exponential hazard function best approximated observed survival data for each treatment group based on model goodness-of-fit statistics. Projected survival based on several different hazard functions is displayed along with observed survival in Figure 1.Download figureDownload PowerPointFigure 1. Survival probability projections with alternative modeling approaches. Open circles and triangles are observed data from the transcatheter aortic valve replacement (TAVR) and control gro
Referência(s)