ÌÇÐÄvlog

Key Points

QuestionÌý How are septal myectomy and alcohol septal ablation used in the United States, and what are the in-hospital outcomes after these procedures by institutional procedural volume?

FindingsÌý In US hospitals from January 1, 2003, through December 31, 2011, most centers that provide septal reduction therapy performed few septal myectomy and alcohol septal ablation procedures annually, which is below the threshold recommended by the 2011 American College of Cardiology Foundation/American Heart Association Task Force Guideline for the Diagnosis and Treatment of Hypertrophic Cardiomyopathy. Low septal myectomy volume was associated with worse in-hospital outcomes, including higher mortality.

MeaningÌý More efforts are needed to encourage referral of patients to high-volume centers of excellence for septal reduction therapy according to the guidelines.

ImportanceÌý Previous data on septal myectomy (SM) and alcohol septal ablation (ASA) in obstructive hypertrophic cardiomyopathy have been limited to small, nonrandomized, single-center studies. Use of septal reduction therapy and the effect of institutional experience on procedural outcomes nationally are unknown.

ObjectiveÌý To examine in-hospital outcomes after SM and ASA stratified by hospital volume within a large, national inpatient database.

Design, Setting, and ParticipantsÌý This study analyzed all patients who were hospitalized for SM or ASA in a nationwide inpatient database from January 1, 2003, through December 31, 2011.

Main Outcomes and MeasuresÌý Rates of adverse in-hospital events (death, stroke, bleeding, acute renal failure, and need for permanent pacemaker) were examined. Multivariate logistic regression analysis was performed to compare overall outcomes after each procedure based on tertiles of hospital volume of SM and ASA.

ResultsÌý Of 71 888 761 discharge records reviewed, a total of 11 248 patients underwent septal reduction procedures, of whom 6386 (56.8%) underwent SM and 4862 (43.2%) underwent ASA. A total of 59.9% of institutions performed 10 SM procedures or fewer, whereas 66.9% of institutions performed 10 ASA procedures or fewer during the study period. Incidence of in-hospital death (15.6%, 9.6%, and 3.8%; P < .001), need for permanent pacemaker (10.0%, 13.8%, and 8.9%; P < .001), and bleeding complications (3.3%, 3.8%, and 1.7%; P < .001) after SM was lower in higher-volume centers when stratified by first, second, and third tertiles of hospital volume, respectively. Similarly, there was a lower incidence of death (2.3%, 0.8%, and 0.6%; P = .02) and acute renal failure (6.2%, 7.6%, and 2.4%; P < .001) after ASA in higher-volume centers. The lowest tertile of SM volume among hospitals was an independent predictor of in-hospital all-cause mortality (adjusted odds ratio, 3.11; 95% CI, 1.98-4.89) and bleeding (adjusted odds ratio, 3.77; 95% CI, 2.12-6.70), whereas being in the lowest tertile of ASA by volume was not independently associated with an increased risk of adverse postprocedural events.

Conclusions and RelevanceÌý In US hospitals from 2003 through 2011, most centers that provide septal reduction therapy performed few SM and ASA procedures, which is below the threshold recommended by the 2011 American College of Cardiology Foundation/American Heart Association Task Force Guideline for the Diagnosis and Treatment of Hypertrophic Cardiomyopathy. Low SM volume was associated with worse outcomes, including higher mortality, longer length of stay, and higher costs. More efforts are needed to encourage referral of patients to centers of excellence for septal reduction therapy.

Hypertrophic cardiomyopathy (HCM) is one of the most commonly inherited cardiac conditions, affecting more than 700 000 people in the United States.¹ Left ventricular outflow tract (LVOT) obstruction in HCM is associated with greater propensity to develop heart failure symptoms, exertional syncope, or sudden cardiac death.² Surgical septal myectomy (SM), when performed in experienced centers, is considered to be the preferred strategy for relieving LVOT in patients with HCM who are refractory to medical therapy and are good surgical candidates.^3,4 Previous studies^5,6 have revealed the efficacy of SM in patients with severe LVOT obstruction and drug-refractory symptoms of heart failure, resulting in relief of heart failure symptoms, improved quality of life, and possibly long-term survival. In experienced surgical centers, SM is associated with low periprocedural mortality of less than 1%; however, such results are limited to few dedicated HCM centers with extensive SM experience.^7,8 Alcohol septal ablation (ASA) is a minimally invasive alternative to surgery performed via a percutaneous catheter technique by using absolute alcohol to induce a targeted septal myocardial infarction. Alcohol septal ablation is performed in patients with certain anatomical criteria and is favored in those with multiple comorbidities and in those at high risk for surgical intervention.^9,10 Similar to SM, ASA is associated with significant improvement in LVOT gradient, symptoms of heart failure, exercise capacity, and low periprocedural morbidity and mortality.^11,12

Multiple single- and multicenter studies^11-15 from high-volume centers have revealed the efficacy and safety of SM and ASA. The 2011 American College of Cardiology Foundation and American Heart Association Task Force Guideline for the Diagnosis and Treatment of Hypertrophic Cardiomyopathy recommend that septal reduction therapy be performed only by experienced operators in dedicated HCM clinical programs.⁴ However, it is unknown whether patients are being referred to these centers of excellence for SM or ASA and whether patients with higher surgical risk profiles are indeed being referred preferentially to experienced centers for ASA as recommended by the guidelines. In addition, higher-volume centers overall tend to have better surgical outcomes compared with low-volume centers in a variety of cardiovascular procedures, such as coronary artery bypass surgery, carotid endarterectomy, heart transplantation, and acute aortic dissection repair.^16-19 However, given the low prevalence of septal reduction procedures, data on the association between procedural experience and outcomes after SM and ASA are limited. Therefore, we sought to evaluate the trends, characteristics, and in-hospital outcomes after SM and ASA and, importantly, to examine the association between institutional procedural volume and outcomes after each procedure in a large, comprehensive national database of hospital discharges from January 1, 2003, through December 31, 2011.

Data were obtained from Agency for Healthcare Research and Quality Healthcare Cost and Utilization Project Nationwide Inpatient Sample (NIS) files from January 1, 2003, through December 31, 2011.²⁰ The NIS is a 20% stratified sample of all nonfederal US hospitals and in 2011 contained deidentified information for 38 590 733 discharges from 1049 hospitals in 46 states. Weill Cornell Medical College determined that institutional review board approval and informed consent were not required for this study because this study uses a deidentified administrative database. Discharges are weighted based on the sampling scheme to permit inferences for a nationally representative population.²⁰ Each record in the NIS includes all procedure and diagnosis International Classification of Diseases (ICD) codes recorded for each patient’s hospital discharge. From January 1, 2003, through December 31, 2011, hospitalizations that led to SM and ASA were selected by searching for the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) procedure codes 37.33 for SM and 37.34 for ASA in any of the 15 procedure fields in the database. Only adults with HCM were included by selecting for ICD-9 code 425.1 in the primary diagnosis.

Patient-level and hospital-level variables were included as baseline characteristics (eTable in the Supplement). Hospital-level data elements were derived from the American Hospital Association Annual Survey Database. The Agency for Healthcare Research and Quality comorbidity measures based on the Elixhauser methods were used to identify comorbid conditions.²¹ The outcome measures examined included in-hospital all-cause mortality, need for permanent pacemaker (PPM), stroke, bleeding, and acute renal failure (ARF). Stroke was identified using ICD-9 codes 997.02, 362.31, 368.12, 781.4, 433.11, 435, and 434. Major bleeding was identified by ICD-9 codes 430 to 432, 578.X, 719.1X, 423.0, 599.7, 626.2, 626.6, 626.8, 627.0, 627.1, 786.3, 784.7, and 459.0. Need for PPM was selected by ICD-9-CM procedure codes 37.80 to 83. Acute renal failure was identified by ICD-9 code 584.

For descriptive analyses, we compared baseline patient and hospital characteristics of those undergoing SM and ASA. Continuous variables are presented as medians; categorical variables are expressed as frequencies (percentages). To compare baseline characteristics, Mann-Whitney and Wilcoxon nonparametric tests or t tests were used for continuous variables, and Pearson χ² tests were used for categorical variables. We calculated the procedure rates for SM and ASA as the weighted number of procedures divided by 20% of the total number of US adults during the same periods. Estimates of the US adult population from January 1, 2003, through December 31, 2011, were obtained from the US Census Bureau.²² Trends in the annual rates of SM and ASA were assessed using time series modeling (autoregressive integrated moving average model).

Unadjusted rates of in-hospital outcomes were calculated for subgroups within SM or ASA cohorts stratified according to tertiles of hospital procedural volume. Hospital procedural volume was used to divide institutions into tertiles based on the volume of SM and ASA performed from January 1, 2003, through December 31, 2011. The first tertile was defined as centers with the lowest procedural volume. Within SM and ASA cohorts, weighted multivariate logistic regression modeling was performed to compare subgroups stratified according to tertiles of hospital volume (with the third tertile as the reference), adjusting for all univariate predictors of outcomes (P < .10). Furthermore, we performed a sensitivity analysis by excluding any concomitant operations, such as aortic valve surgery, mitral valve repair or replacement, or coronary artery bypass surgery, when performed together with SM and then comparing subgroups stratified according to tertiles of hospital SM volume using multivariable logistic regression analysis. For all regression analyses, the Taylor linearization method with replacement design was used to compute variances. All statistical tests were 2-sided, and P < .05 was set a priori to be statistically significant. All statistical analyses were conducted using SAS statistical software, version 9.2 (SAS Institute Inc) and SPSS statistical software, version 20 (SPSS, Inc).

For each year from 2003 through 2011, the NIS data set included discharges from all hospitals that performed SM and ASA. Of 71 888 761 discharge records reviewed from 2003 through 2011, a total of 11 248 patients underwent septal reduction procedures, of whom 6386 (56.8%) underwent SM and 4862 (43.2%) underwent ASA. The annual rate of SM decreased by 24.5% from 2.00 procedures per million people per year in 2003 to 1.51 procedures per million adults per year in 2011. On the contrary, the annual rate of ASA increased by 56.2% from 1.60 procedures per million people per year in 2003 to 2.49 procedures per million adults per year in 2011. The trends in overall rates of SM or ASA procedures during the study period, however, were not significant. The median numbers of cases for SM and ASA were 1.0 and 0.7 per year per institution, respectively. Figure 1 shows the frequency of SM and ASA procedures performed across institutions during the 9-year period. Four institutions performed 35.9% of all isolated SM operations (1132 of 3157 cases), whereas 6 institutions performed 24.1% of all ASA procedures (1173 of 4862 cases). A total of 215 (59.9%) of 359 institutions performed 10 SM procedures or fewer, 282 institutions (78.6%) performed 20 SM procedures or fewer, and 338 institutions (94.2%) performed 50 SM procedures or fewer. Similarly, 164 (66.9%) of 245 institutions performed 10 ASA procedures or fewer, 195 institutions (79.6%) performed 20 ASA procedures or fewer, and 225 institutions (91.8%) performed 50 ASA procedures or fewer during the study period.

Table 1 and Table 2 compare baseline characteristics stratified by hospital volumes (in tertiles) for SM and ASA. Patients undergoing SM in highest-volume centers were more likely to be younger and undergo concomitant coronary artery bypass graft or valve operations. Highest-volume SM centers (third tertile) were more likely to be larger and teaching institutions. Length of stay in high-volume SM centers (median, 7.0 vs 8.0 days; P < .001) and cost of hospitalization were lower compared with lower-volume SM centers. Patients having ASA performed in lowest- vs highest-volume centers (first quartile vs third quartile) were older and more likely to have anemia, diabetes mellitus, chronic pulmonary disease, congestive heart failure, chronic renal failure, and liver disease. Highest-volume centers were more likely to be a teaching institution and a large hospital. Similarly to SM, undergoing ASA in highest-volume centers was associated with shorter length of stay and lower cost of hospitalization. Median length of hospitalization was significantly longer with SM vs ASA (7.0 vs 3.0 days), with median hospital costs being 2.3 times greater for SM vs ASA.

The unadjusted incidence of in-hospital death (15.6%, 9.6%, and 3.8%; P < .001), need for PPM (10.0%, 13.8%, and 8.9%; P < .001), and bleeding complications (3.3%, 3.8%, and 1.7%; P < .001) after SM was higher in lower-volume centers when stratified by first, second, and third tertiles of hospital volume, respectively (Figure 2). Similarly, there was a higher incidence of death (2.3%, 0.8%, and 0.6%; P = .02) and ARF (6.2%, 7.6%, and 2.4%; P < .001) after ASA in lower-volume centers when stratified by first, second, and third tertiles of hospital volume, respectively (Figure 2). Being an SM center in the lowest-volume hospital tertile (first vs third tertile of volume) was associated with significantly greater odds of all-cause mortality (adjusted odds ratio [OR], 3.11; 95% CI, 1.98-4.89; P < .001) and bleeding complications (adjusted OR, 3.77; 95% CI, 2.12-6.70; P < .001) after adjustment with multivariate propensity score–adjusted logistic regression analysis (Table 3). A sensitivity analysis, excluding any concomitant operations performed together with SM, confirmed that SM centers in the lowest-volume hospital tertile (first vs third tertile of volume) were associated with significantly greater odds of all-cause mortality (adjusted OR, 2.50; 95% CI, 1.69-3.69; P < .001) using multivariable logistic regression analysis. However, lower ASA procedural volume (first vs third tertile of volume) was not associated with increased odds of any in-hospital adverse events after multivariable adjustment (Table 4).

There are several important findings in this nationally representative sample of US hospital discharge records examining outcomes after SM and ASA from January 1, 2003, through December 31, 2011. First, most US institutions perform few SM and ASA cases annually (defined as <10 per year). Second, undergoing SM in centers that rarely perform septal reduction operations was associated with worse postprocedural outcomes compared with high-volume centers, with a 3-fold higher mortality (in first- vs third-tertile institutions by volume). On the other hand, undergoing ASA in lower-volume institutions was not associated with worse outcomes after ASA compared with high-volume ASA centers.

Our study is the first, to our knowledge, to report a nationally representative experience with SM and ASA across a representative spectrum of more than 1000 hospitals. Previous studies^11-15 examining outcomes after these 2 procedures were primarily retrospective from experienced single-center institutions. In the early reports of SM, perioperative mortality was relatively high (≥5%), whereas more recent surgical results have markedly improved, with reported operative mortality at less than 1%.^6,23 However, such improved outcomes may be limited to only a few high-volume centers. We found significant variability in in-hospital mortality after SM when hospital volume was taken into account (3.8% in the highest tertile vs 15.6% in the lowest volume tertile). Interestingly, our findings indicate overall mortality after SM is much higher (5.2%) across US institutions, with most institutions having limited experience with this operation. However, when the first tertile of experience is excluded, mortality more closely mirrored that seen at major referral centers. A previous study²⁴ found an effect of an operator or institution learning curve, emphasizing the importance of surgical operator volume. Septal myectomy can be a technically demanding operation, with few centers currently having extensive surgical expertise in the procedure. Most surgical centers perform 1 to 2 SM operations per year as opposed to other commonly performed cardiac operations, such as coronary artery bypass and valve operations.²⁵ Thus, previous reports^4,26 of less than 1% postoperative mortality after SM are likely seen in a relatively small number of high-volume HCM centers that perform SM in carefully selected patients. Furthermore, this report suggests that small-volume institutions are performing SM on older patients with more comorbidities while using fewer concomitant valvular and coronary artery bypass graft operations. The American College of Cardiology and American Heart Association guidelines suggest that operators and institutions should aim to achieve mortality rates of less than 1% and major complication rates of less than 3%.⁴ Although it is difficult to define minimum annual experience for the operators required to perform SM, current guidelines recommend (class I indication) that only experienced operators with cumulative operator volume of at least 20 procedures should perform SM. However, our study reveals a potentially concerning practice pattern in which approximately 80% of SMs were performed in hospitals where the cumulative volume of SM was fewer than 20 cases during the 9-year study period. This finding is particularly important given that SM procedures performed at low-volume centers were associated with in-hospital mortality that far exceeds the recommended threshold. Therefore, given the 3- to 10-fold higher mortality observed in the lowest-volume centers, referral to centers of excellence for SM would better serve this challenging patient population.

Our postprocedural outcomes after ASA, with a 0.7% mortality rate, are similar to a previous report¹³ from the Mayo Clinic (1.4%) and the Multicenter North American Registry (<1%). Improvements in overall outcomes (ie, mortality) and decreased need for PPM after ASA may be related to the evolution and refinements of the ASA technique, which were seen from 2000 through 2010. These refinements included use of myocardial contrast echocardiography to target the septal perforator, reduction in ethanol volume and rate of injection, and more judicious ASA case selection. During the initial experience with ASA, the rates of conduction abnormalities with ASA were quite high, with approximately 20% to 25% of patients requiring PPMs.¹³ The need for PPM after ASA in our study (11.9%) is similar to the rates of PPM implantation seen in more contemporary reports (in the 8%-17% range).^12,27,28 The rate of PPM implantation after SM in our study (9.8%) is in line with the rates of 7.9% to 10% seen in previous studies^29,30 and higher than the approximately 2% to 3% rates seen in early reports of SM.^31,32 Our study suggests that outcomes after ASA in this population are similar to those seen in the published literature from experienced centers. In contrast to the observation made in the SM cohort, after multivariate adjustment, there was no clear association between hospital volume of ASA and procedural outcomes, although there was a numerically higher rate of mortality in the lowest-tertile group. This finding could be a reflection of a significantly steeper learning curve associated with SM and the relative ease of adapting ASA by operators with experience in catheter-based therapy.

Our study also suggests that the annual rate of ASA is increasing while the annual rate of SM is slowly decreasing. Given the prevalent use of catheter-based coronary revascularization in the United States, it is possible that more operators are performing ASA over time. The American College of Cardiology Foundation and American Heart Association 2011 Hypertrophic Cardiomyopathy Guideline recommends that, in those patients who are acceptable surgical candidates, SM should generally be preferred (class IIa) over ASA (class IIb), whereas in those patients who are not acceptable candidates for surgical intervention, ASA would be the favored treatment option (class IIa).⁴ Further studies are necessary to determine whether the guidelines are being applied in clinical practice and that patients with HCM are properly delegated to appropriate therapy.

Our study examines a nationally representative sample of outcomes of SM and ASA during a period of roughly 10 years and the challenges we face in applying guidelines to clinical practice. Given higher procedural mortality seen in low-volume centers for SM and ASA, we need to increase the efforts of the cardiology community to refer patients to centers of excellence. Septal myectomy remains the criterion standard for treatment of refractory obstructive HCM. However, for geographic areas with limited access to surgical centers with expertise in SM, ASA remains a viable option in anatomically and clinically appropriate patients. However, it is important to refer these patients to centers of excellence in ASA as suggested by the guidelines. This directive is especially important to acknowledge given the technical challenges associated with SM and the clear association of experience and outcomes after SM.²⁷ In fact, ASA is increasingly used in many European centers because of several factors, including physician and patient preference; minimally invasive; catheter-based nature of the procedure; local availability of experienced ASA operators; and lack of regional expertise in SM. Recent European Society of Cardiology guidelines support this practice, with SM and ASA receiving a class I recommendation.³³ Nevertheless, in experienced HCM centers with a high volume of SM operations, SM as the primary treatment option would seem appropriate given the long-term data supporting favorable outcomes after SM.^5,10

There are several limitations present in our study. First, this is a retrospective study based on an NIS sample, which approximates the national distribution of key hospital characteristics. Our estimates for SM and ASA were derived from a 20% sample. Individual hospitals may not have provided data for the entire 2003-2011 duration, and it is possible that the overall volume of procedures was underrepresented or overrepresented by the sample. However, NIS has been used extensively to examine national health care trends, and its sampling design has been validated in numerous publications.^34,35 Second, unmeasured confounders could not be accounted for, including preexisting conduction abnormalities, anatomical abnormalities, and current medications. The intent of this study was not to compare SM and ASA given the difficulty in adjusting for confounders not captured in this administrative database or for all factors that lead to referral bias for SM vs ASA. Patients undergoing SM differ from patients undergoing ASA in type and severity of medical comorbidities and requirement for concomitant surgical procedures. Therefore, our finding should not be used for comparison of procedural safety or efficacy between SM and ASA. Third, the NIS database provides only in-hospital outcomes. Therefore, our findings do not address 30-day or long-term hemodynamic and clinical outcomes or need for additional hospitalizations or procedures in the future. This limitation may be particularly important given that ASA is associated with a potential risk of future ventricular arrhythmias due to septal scar development.³⁶ Fourth, our study does not have data for patients with obstructive HCM who were not treated with an invasive strategy, potentially introducing an additional selection bias. Finally, the NIS does not allow assessment of procedural success, including reduction in LVOT gradient or improvement in heart failure symptoms.

In the United States from 2003 through 2011, most institutions performed few SM and ASA cases annually, which are fewer than recommended by the guidelines. Importantly, low-volume centers were associated with worse in-hospital outcomes, including higher mortality, longer length of stay, and higher hospital cost. More efforts by the cardiology community are needed to encourage referral of patients with HCM to centers of excellence for septal reduction therapy.

Accepted for Publication: February 13, 2016.

Correction: This article was corrected on October 12, 2016, to fix a typographical error in the text and on January 4, 2017, for incorrect ICD-9 codes and percentages reported in tables.

Corresponding Author: Luke K. Kim, MD, Division of Cardiology, Department of Medicine, Weill Cornell Medical College, New York Presbyterian Hospital, 520 E 70th St, Starr Pavilion, 4th Floor, New York, NY 10021 (luk9003@med.cornell.edu).

Published Online: April 27, 2016. doi:10.1001/jamacardio.2016.0252

Author Contributions: Dr Kim had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Kim, Swaminathan, Minutello, Wong, Naidu, Charitakis, Feldman.

Acquisition, analysis, or interpretation of data: Kim, Looser, Minutello, Bergman, Naidu, Gade, Singh, Feldman.

Drafting of the manuscript: Kim, Looser, Minutello, Feldman.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Kim, Minutello.

Obtained funding: Kim.

Administrative, technical, or material support: Kim, Wong, Gade, Charitakis.

Study supervision: Kim, Swaminathan, Naidu, Singh, Feldman.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Feldman reported receiving consulting/speaker’s fees from Eli Lilly, Daiichi-Sankyo, Abbott Vascular, Pfizer, and Bristol-Myers Squibb. Dr Naidu reported receiving consulting fees from Abiomed, Abbott Vascular, Gilead, The Medicines Company, and Maquet. No other disclosures were reported.

Funding/Support: This work was supported by grants from the Michael Wolk Heart Foundation and the New York Cardiac Center Inc (Dr Kim).

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.