Abstract
Background
The epidemiology of aortic dissection (AD) has not been well-described among older persons in the United States. It is not known whether advancements in AD care over the last decade have been accompanied by changes in outcomes.
Methods and Results
The Inpatient Medicare data from 2000 to 2011 were used to determine trends in hospitalization rates for AD. Mortality rates were ascertained through corresponding vital status files. A total of 32,057 initial AD hospitalizations were identified between 2000 and 2011. The overall hospitalization rate for AD remained unchanged at 10 per 100,000 person-years. For 30-day and 1-year mortality associated with AD, the observed rate decreased from 31.8% to 25.4% (difference, 6.4%; 95% confidence interval [CI], 6.2–6.5; adjusted, 6.4%; 95% CI, 5.7–6.9) and from 42.6% to 37.4% (difference, 5.2%; 95% CI, 5.1–5.2; adjusted, 6.2%; 95% CI, 5.3–6.7) respectively. For patients undergoing surgical repair for type A dissections, the observed 30-day mortality decreased from 30.7% to 21.4% (difference, 9.3%; 95% CI, 8.3–10.2; adjusted, 7.3%; 95% CI, 5.8–7.8) and the observed 1-year mortality decreased from 39.9% to 31.6% (difference, 8.3%; 95% CI, 7.5–9.1%; adjusted, 8.2%; 95% CI, 6.7 – 9.1). The 30-day mortality decreased from 24.9% to 21% (difference, 3.9%; 95% CI, 3.5–4.2; adjusted, 2.9%; 95% CI, 0.7–4.4) and 1-year decreased from 36.4% to 32.5% (difference, 3.9%; 95% CI, 3.3–4.3; adjusted, 3.9%; 95% CI, 2.5–6.3) for surgical repair of type B dissection.
Conclusions
While AD hospitalization rates remained stable, improvement in mortality was noted, particularly in patients undergoing surgical repair.
Keywords: aorta dissection, epidemiology, mortality, surgery
Introduction
Aortic dissection (AD) is a life threatening condition associated with morbidity and mortality. According to the Centers for Disease Control and Prevention, diseases of the aorta and its branches account for 43,000 to 47,000 deaths annually in the United States.1 Most autopsy studies suggest that the presentation of thoracic aortic disease is often death due to aortic dissection and rupture.1, 2 For those with acute thoracic aortic disease who manage to obtain medical care, the mortality is quite high with in-hospital mortality reported to be 25%.3, 4
Few studies have studied the hospitalization rates of AD nationally in recent times. Previous studies were limited insofar as they included only a limited geographic area5, 6 or data from selected high-volume centers of excellence,3 or were not conducted recently.4, 6 We do not know if there have been recent changes in the hospitalization rates for AD given recent improvements in important risk factors such as blood pressure control.
Additionally, data on recent outcomes associated with care of patients in the real world is unknown. Especially since several innovations in diagnostic techniques and management of AD have been adopted in the past decade that have promising improved clinical outcomes for those who survive long enough to receive medical care.7–10 Thus data on the recent epidemiology and outcomes associated with AD could potentially assist in quantifying the present burden of AD, effectiveness of newer interventions as well as provide valuable data for benchmarking performance of future technologies.
Accordingly, in this study, we describe national trends in hospitalization rates of patients with AD over the last decade. We also assess short- and long-term outcomes of patients who received different management strategies, such as surgical repair, thoracic endovascular aortic repair (TEVAR) and medical therapy. To do so, we analyzed data from all Medicare Fee-for-Service (FFS) beneficiaries from 2000 to 2011.
Methods
Data Sources
We used the Medicare beneficiary denominator file from the Centers for Medicare & Medicaid Services (CMS) to identify beneficiaries aged 65 years or older who were enrolled in the Fee-For-Service plan for at least one month from January 1, 2000 to December 31, 2011. We calculated person-years for each beneficiary to account for new enrollment, disenrollment, or death for each year of the study. We then link this person-years file to the Medicare inpatient standard analytical file from CMS to identify all Medicare FFS beneficiaries excluding patients receiving other forms of combined coverage i.e. Medicare Advantage who were hospitalized for AD from January 1, 2000, to December 31, 2011. These administrative claims included information on patient demographics (age, sex, and race), admission and discharge dates, and up to 10 discharge diagnoses and six procedures in 2000 and increasing up to 25 diagnoses and procedure codes in 2011 (as coded by the International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM]).11, 12 However, we restricted the number of diagnoses and procedures for the 2011 data to the same as the 2010 and older years data (i.e. 10 diagnoses and six procedure codes) to make the calculation of Hierarchical Condition Categories consistent across the study period.13 Aortic dissection hospitalization was defined as discharge from an acute-care hospital with a principal discharge diagnosis of ICD-9-CM code 441.01 or 441.03 (dissection of aorta, thoracic or thoracoabdominal respectively). We restricted all beneficiaries in FFS for at least 12 months and used the 1999 inpatient data for beneficiaries hospitalized in 2000. To differentiate surgical repair performed for dissection of the ascending aorta (Type A) and descending aorta (Type B) we adopted the criteria developed by Sachs et al.14 Hospitalizations associated with surgical repair for type A AD were defined as having a principal discharge diagnosis of 441.01 or 441.03, any one of the following ICD-9-CM procedure codes for surgery of AD: resection of thoracic vessel or aorta with anastomosis (38.34, 38.35), resection of thoracic vessel or aorta with replacement (38.44, 38.45), or repair of blood vessel with synthetic or unspecified patch graft (39.57, 39.58) as well as procedure codes for either cardioplegia (39.63), valve repair (35.00–35.99), or operations on vessels of the heart (36.00–36.99, 37.0, 37.2, 37.31–37.90, 37.93–37.99). To isolate type B AD, the aforementioned diagnosis codes for AD as well as procedure codes for surgery of AD excluding procedure codes for cardioplegia, valve repair and operations on vessels of the heart were used. We defined hospitalizations in which the patient underwent TEVAR as both principal discharge diagnosis of 441.01 or 441.03 and ICD-9-CM procedure code for endovascular implantation of graft in thoracic aorta (39.73). Finally, hospitalizations receiving medical management were identified as having a primary discharge diagnosis of AD (441.01 or 441.03) with no accompanying procedure codes for surgery for AD or TEVAR. We excluded beneficiaries if they were hospitalized for AD outside of the 50 U.S. states, the District of Columbia, Puerto Rico, and other U.S. territories.
Patient Characteristics and Comorbidities
We examined the demographic and clinical characteristics of patients with AD over time. Demographic variables included age (65–74, 75–84, ≥85 years), sex, and race (white, black, and other). We determined race using the Medicare denominator file, which is based on data from the Social Security Administration.15 We identified common comorbidities that are included in the publicly reported CMS 30-day mortality measures for acute myocardial infarction, heart failure, and pneumonia.16–18 Specifically, we identified comorbidities using the secondary diagnosis codes from the index AD hospitalization, specifically including those that did not represent a potential complication of care, as well as principal and secondary diagnosis codes of all hospitalizations up to 1 year before the index AD hospitalization (Appendix 1). The 1999 inpatient data was used to obtain comorbidity information on patients who were hospitalized for AD in 2000. Data on TEVAR are presented from 2005 to 2011, as the ICD-9-CM procedure code for TEVAR (39.73) was first introduced in the fourth quarter of 2005.
Primary Outcomes
There were two main primary outcomes: hospitalization for AD and mortality. Because Medicare FFS beneficiaries were enrolled throughout the year, some beneficiaries were in FFS for the entire year while others were in FFS for a few months. Thus, we calculated person-years for each beneficiary to account for new enrollment, disenrollment, or death during an index year (e.g., if a beneficiary enrolled in FFS for three months, this beneficiary contributed 3/12 person-years into the calculation). All Medicare beneficiaries who had more than one month of enrollment in the FFS plan were included in the person-years calculation.
The hospitalization rate was calculated separately for each year by dividing the total number of AD hospitalizations in a year by the corresponding person-years for that year.
To calculate 30-day and 1-year mortality outcomes, we identified all AD hospitalizations that occurred in a given year. If a patient had more than one AD hospitalization during the study period, the initial hospitalization was selected. We adopted this approach because additional hospitalizations might reflect manifestations of the initial disease process. Additionally, we examined all-cause hospitalizations one year prior to the index hospitalization for AD to rule out prior hospitalizations for AD. The admission date of the AD hospitalization was the “time zero” for 30-day and 1-year mortality analyses. For hospitalizations in which the patient underwent either surgical repair or TEVAR of the dissection, the date of the procedure was considered as “time zero” for determining 30-day and 1-year mortality associated with that hospitalization. The 2012 Medicare enrollment file was used to obtain mortality information for patients who hospitalized for AD in 2011.
Statistical Analysis
Changes in patient characteristics and primary outcomes (AD hospitalization and mortality rates) were examined over time. We used the Mantel-Haenszel Chi-squared test to evaluate whether these changes over time were statistically significant. We stratified trends in hospitalization rate by age, sex, and race, and trends in both hospitalization and mortality by choice of management strategy.
We fitted a linear mixed-effects model with a Poisson link function and state-specific random intercepts to assess rates of hospitalization for AD adjusted for age, sex, and race. We considered the AD hospitalization rate during 2000 as the reference and calculated the incidence rate ratio for each subsequent year by including dummy variables for the subsequent years in the mixed-effects model.
To obtain annual mortality rates adjusted for patient demographics and comorbidities, we fitted a linear mixed-effects model with a logit link function and hospital-specific random intercepts. Using data from 2000 as the referent and dummy variables for each subsequent year, we calculated the adjusted odds ratio for mortality for subsequent years. Using the method described by Zhang and Yu, we converted the odds ratio values to risk ratio estimates.19 We then multiplied the risk ratio for each year by the mortality rate in the baseline year (2000) to calculate the adjusted mortality rates across years.
We conducted the analyses using SAS version 9.3 64-bit Windows version (SAS Institute Inc., Cary, North Carolina). All statistical testing was 2-sided, and p<0.05 was considered statistically significant. Institutional review board approval was obtained through the Yale University Human Investigation Committee.
Results
Patient Characteristics
The final sample consisted of 353,442,679 observations, representing 79,278,075 beneficiaries aged 65 years or older with at least one month of enrollment in Medicare FFS during the study period (336,781,989 person-years) leading to 32,057 hospitalizations for AD from 2000–2011. Demographic and clinical characteristics of the patients hospitalized with AD are presented in (Table 1). The mean age of hospitalized AD patients increased slightly from 77.1 years in 2000 to 77.4 years in 2011. In addition, there was a slight increase in the proportion of female patients (47.4% to 51.9%) and decrease in the proportion of white patients (87.0% to 83.2%) over time. The prevalence of many of the comorbidities remained relatively stable over the study period. However, certain conditions such as hypertension (65.0% to 71.5%), diabetes mellitus (8.9% to 13.9%), dementia (3.8% to 7.3%), renal failure (3.0% to 9.2%), pneumonia (8.5% to 12.6%), respiratory failure (2.1% to 5.1%), and depression (3.3% to 6.5%) increased. Prior literature demonstrates that the profile of patients with chronic cardiovascular conditions such as heart failure has become sicker with significant increase in the age and proportion of chronic comorbidities over the last decade.12 Hence the aforementioned temporal changes in comorbidities are most likely real versus more intense coding practice patterns.
Table 1.
Characteristics of Medicare Fee-for-Service Beneficiaries Hospitalized with Aortic Dissection, 2000–2011
| 2000 | 2001–2002 | 2003–2004 | 2005–2006 | 2007–2008 | 2009–2010 | 2011 | |
|---|---|---|---|---|---|---|---|
|
Aortic Dissection Hospitalizations, No.* |
2507 | 5262 | 5527 | 5546 | 5267 | 5146 | 2482 |
| Demographics, % | |||||||
| Age, mean (SD), y | 77.1 (7.0) | 77.0 (7.0) | 77.1 (7.1) | 77.3 (7.3) | 77.5 (7.6) | 77.5 (7.7) | 77.4 (7.7) |
| Female | 47.4 | 48.8 | 50.4 | 50.8 | 51.6 | 52.1 | 51.9 |
| White | 87.0 | 85.8 | 84.7 | 83.3 | 84.1 | 84.7 | 83.2 |
| Black | 8.8 | 10.1 | 10.3 | 11.1 | 10.8 | 9.9 | 11.6 |
| Other race** | 4.2 | 4.1 | 5.0 | 5.6 | 5.1 | 5.4 | 5.3 |
| Cardiovascular conditions, % | |||||||
| Hypertension | 65.0 | 67.2 | 67.6 | 67.1 | 69.8 | 69.1 | 71.5 |
| Diabetes mellitus | 8.9 | 10.0 | 10.8 | 12.2 | 12.6 | 12.6 | 13.9 |
| Atherosclerotic disease | 34.9 | 34.5 | 34.6 | 34.0 | 31.4 | 30.5 | 31.2 |
| Unstable angina | 3.7 | 3.4 | 2.9 | 2.1 | 1.8 | 1.7 | 1.5 |
| Prior MI | 2.3 | 2.1 | 2.5 | 2.3 | 2.1 | 2.3 | 2.1 |
| Prior HF | 9.9 | 9.7 | 10.3 | 10.8 | 9.9 | 9.7 | 8.3 |
| Peripheral vascular disease | 12.5 | 13.6 | 13.9 | 15.5 | 14.4 | 13.2 | 13.2 |
| Stroke | 1.9 | 1.6 | 1.6 | 1.7 | 1.3 | 1.6 | 1.5 |
| Cerebrovascular disease other than stroke | 4.1 | 3.6 | 3.9 | 3.9 | 3.8 | 3.5 | 3.5 |
| Geriatric conditions, % | |||||||
| Malnutrition | 2.8 | 2.8 | 3.3 | 4.3 | 6.2 | 7.7 | 7.7 |
| Dementia | 3.8 | 5.2 | 6.0 | 6.4 | 6.6 | 7.4 | 7.3 |
| Functional disability | 2.8 | 2.3 | 2.5 | 2.4 | 2.7 | 2.9 | 3.4 |
| Other conditions, % | |||||||
| Renal failure | 3.0 | 4.0 | 4.6 | 6.9 | 9.4 | 9.0 | 9.2 |
| COPD | 24.3 | 25.8 | 27.1 | 27.0 | 23.8 | 20.6 | 21.9 |
| Pneumonia | 8.5 | 9.4 | 10.4 | 11.1 | 11.4 | 12.0 | 12.6 |
| Respiratory failure | 2.1 | 2.4 | 2.7 | 3.6 | 4.3 | 4.6 | 5.1 |
| Liver disease | 0.4 | 0.7 | 0.7 | 0.7 | 0.7 | 0.6 | 1.1 |
| Cancer | 5.9 | 5.9 | 6.0 | 6.1 | 6.0 | 6.4 | 6.3 |
| Depression | 3.7 | 4.7 | 4.8 | 5.2 | 5.3 | 6.0 | 6.5 |
| Other psychiatric disorder | 2.1 | 1.5 | 1.7 | 1.8 | 1.9 | 1.9 | 2.0 |
| Trauma in past year | 4.5 | 5.0 | 5.4 | 6.1 | 6.3 | 5.8 | 5.8 |
Data for the initial and final year of the study i.e. 2000 and 2011, has been presented individually while data for all other years have been combined in two year intervals for brevity
Other race includes Asian, Hispanic, North American Native, or other not specified COPD, chronic obstructive pulmonary disease; HF, heart failure; MI, myocardial infarction
Overall, patients who underwent an invasive repair procedure were, on average, younger than other AD patients receiving medical therapy. The mean age of patients undergoing surgical repair for type A dissection was 73.9 years in 2000 and 75.0 years in 2011, 74.4 years in 2000 and 74.7 in 2011 for surgical repair of type B dissection and 74.7 years in 2005 and 75.7 years in 2011 of patients undergoing TEVAR compared to mean ages of 78.2 and 78.9 years in 2000 and 2011 among hospitalized AD patients receiving medical therapy (Table 2). Fewer women received surgical repair for type A dissection (43.5% in 2011) or TEVAR (46.9% in 2011) compared with men across the study period, despite an increase in the proportion of female AD patients over time (Table 2). Finally, patients receiving medical therapy were also noted to have significantly more comorbidities (hypertension, prior history of unstable angina, myocardial infarction, congestive heart failure, chronic obstructive pulmonary disease, stroke) compared to patients receiving an invasive procedure (Appendices 2–5).
Table 2.
Demographic Characteristics of Patients Hospitalized for Aortic Dissection Stratified by Intervention Strategy
| 2000 | 2001–2002 | 2003–2004 | 2005–2006 | 2007–2008 | 2009–2010 | 2011 | |
|---|---|---|---|---|---|---|---|
| Type A surgical repair, % | |||||||
| Age, mean (SD), y | 73.9 (5.2) | 74.2 (5.4) | 74.5 (5.7) | 74.6 (5.7) | 75.0 (6.2) | 74.6 (5.9) | 75.0 (6.2) |
| Females | 40.6 | 40.6 | 42.3 | 39.0 | 43.2 | 44.4 | 43.5 |
| Type B surgical repair, % | |||||||
| Age, mean (SD), y | 74.4 (5.6) | 74.4 (5.6) | 74.6 (5.6) | 74.5 (5.6) | 74.7 (6.2) | 74.7 (6.2) | 74.7 (6.2) |
| Females | 39.6 | 43.2 | 46.1 | 45.6 | 44.9 | 44.9 | 50.6 |
| Thoracic endovascular aortic repair, % | |||||||
| Age, mean (SD), y | – | – | – | 74.7 (6.3) | 76.0 (6.4) | 75.9 (6.7) | 75.7 (6.5) |
| Females | – | – | – | 47.9 | 45.7 | 48.0 | 46.9 |
| Medical therapy, % | |||||||
| Age, mean (SD), y | 78.2 (7.3) | 78.0 (7.2) | 78.2 (7.4) | 78.5 (7.6) | 78.7 (7.8) | 79.1 (8.0) | 78.9 (8.2) |
| Females | 50.4 | 51.4 | 52.7 | 54.2 | 55.1 | 56.0 | 55.5 |
Hospitalization for Aortic Dissection
The overall AD hospitalization rate remained stable at 10 per 100,000 person-years across the study period (Table 3). Trends in hospitalization rates were relatively flat for all age, sex, and race subgroups. Over time, AD hospitalization rates remained higher in older persons (particularly those aged 75–84 years), men, and blacks. Hospitalization rates based on individual management strategies have been highlighted in table 3. The hospitalization rate for different management strategies as a percentage of AD hospitalizations have been presented in Figure 1. In summary, percentage of hospitalizations for surgical repair of type A AD remained constant while hospitalizations for type B AD were increasingly supplanted by TEVAR starting 2005 (Figure 1). Results were substantively similar after adjusting for age, sex, and race.
Table 3.
Hospitalization and Readmission Rates for Aortic Dissection in Medicare Fee-for-Service, 2000–2011*
| 2000 | 2001 | 2003 | 2005 | 2007 | 2009 | 2011 | |
|---|---|---|---|---|---|---|---|
| Person-years | 26,768,087 | 27,553,904 | 28,821,487 | 29,157,293 | 27,899,732 | 27,343,436 | 27,958,093 |
| No. of patients | |||||||
| Overall | 2507 | 2630 | 2780 | 2774 | 2735 | 2609 | 2482 |
| Overall Rate per 100,000 Person-years | 9 | 10 | 10 | 10 | 10 | 10 | 9 |
| By Age, y | |||||||
| 65–74 | 8 | 8 | 8 | 8 | 8 | 7 | 7 |
| 75–84 | 12 | 12 | 12 | 12 | 12 | 12 | 11 |
| ≥85 | 9 | 9 | 10 | 11 | 12 | 12 | 11 |
| By Sex | |||||||
| Male | 12 | 12 | 12 | 11 | 11 | 10 | 10 |
| Female | 8 | 8 | 8 | 9 | 9 | 9 | 8 |
| By Race | |||||||
| White | 9 | 10 | 10 | 9 | 10 | 9 | 9 |
| Black | 11 | 13 | 12 | 13 | 14 | 12 | 13 |
| Other** | 7 | 6 | 7 | 9 | 8 | 7 | 7 |
| By Management Strategy*** | |||||||
| Surgical repair for type A dissection subgroup | 2 | 2 | 2 | 2 | 2 | 2 | 2 |
| Surgical repair for type B dissection subgroup | 2 | 2 | 2 | 1 | 1 | 1 | 1 |
| Thoracic endovascular aortic repair subgroup | 0 | 0 | 0 | 0 | 1 | 1 | 1 |
| Medical management subgroup | 7 | 7 | 7 | 7 | 7 | 6 | 5 |
| 30-day Readmission rates by different management strategies, % | |||||||
| Surgical repair for type A dissection subgroup | 23.7 | 24.1 | 27.0 | 30.4 | 25.9 | 23.7 | 21.8 |
| Surgical repair for type B dissection subgroup | 21.1 | 28.6 | 28.6 | 24.9 | 26.5 | 28.1 | 25.5 |
| Thoracic endovascular aortic repair subgroup | – | – | – | 15.0 | 25.4 | 20.3 | 22.3 |
| Medical management subgroup | 20.7 | 21.9 | 20.4 | 19.3 | 19.5 | 19.2 | 19.8 |
Data has been presented individually for the years 2000 and 2011 i.e. initial and final years of the study and in 2-year intervals for the intervening period
Other race includes Asian, Hispanic, North American Native, or other not specified
Note: All rates reported per 100,000 person-years
All the hospitalizations were counted in the hospitalization analysis but only one hospitalization (the initial one) was counted for those patients who had more than one hospitalization in the mortality analysis to avoid multiple counting of deaths
Figure 1.
Trends in hospitalization rates across the study period based on different treatment modalities
Mortality Outcomes
Trends in observed mortality for all treatment arms are presented in (Table 4). Overall, for patients hospitalized with AD, there were substantial declines in observed mortality. Between 2000 and 2011, 30-day mortality declined 9.3% (95% CI, 8.3–10.2), from 31.8% to 25.4%, and 1-year mortality declined 6.4% (95% CI, 5.7–6.9), from 42.6% to 37.4%. Reductions in mortality were also observed in the subgroup of patients who underwent surgical repair for type A dissections by day 30 (9.3%; 95% CI, 8.3–10.2) and by 1 year after surgery (8.3%; 95% CI, 7.5–9.1), for surgical repair of type B dissections by day 30 (3.9%; 95% CI, 3.5–4.2) and 1-year (3.9%; 95% CI, 3.3–4.3) and finally for medical management subgroup as well by 30-day (3.9%; 95% CI, 3.8–4.1) and 1-year (2.4%; 95% CI, 2.3–2.7).
Table 4.
Unadjusted Mortality Rates after Aortic Dissection Hospitalization in Medicare Fee-for-Service, 2000–2011*
| 2000 | 2001 | 2003 | 2005 | 2007 | 2009 | 2011 | |
|---|---|---|---|---|---|---|---|
| Overall aortic dissection cohort, % (95% C.I.) | |||||||
| 30-day mortality | 31.8 (29.9–33.6) |
31.3 (29.6–33.1) |
30.0 (28.3–31.7) |
27.9 (26.3–29.7) |
27.5 (25.9–29.3) |
28.6 (26.8–30.3) |
25.4 (23.7–27.1) |
| 1-year mortality | 42.6 (40.6–44.5) |
41.0 (39.1–42.9) |
40.9 (39.0–42.7) |
38.6 (36.8–40.5) |
38.4 36.6–40.3) |
38.7 (36.8–40.6) |
37.4 (35.5–39.3) |
| Type A surgical repair subgroup, % (95% C.I.) | |||||||
| 30-day mortality | 30.7 (26.2–35.4) |
27.5 (23.3–32.0) |
32.5 (28.3–36.9) |
26.2 (22.3–30.3) |
27.8 (23.8–32.1) |
21.8 (18.3–25.7) |
21.4 (17.9–25.2) |
| 1-year mortality | 39.9 (35.1–44.9) |
37.7 (33.0–42.5) |
42.7 (38.2–47.2) |
35.7 (31.4–40.1) |
36.9 (32.6–41.5) |
31.0 (27.0–35.2) |
31.6 (27.6–35.8) |
| Type B surgical repair subgroup, % (95% C.I.) | |||||||
| 30-day mortality | 24.9 (21.0–29.2) |
24.9 (20.8–29.3) |
24.8 (20.9–29.0) |
21.2 (17.4–25.5) |
20.7 (16.9–24.9) |
19.3 (15.3–24.0) |
21.0 (16.8–25.7) |
| 1-year mortality | 36.4 (31.9–41.1) |
37.0 (32.4–41.8) |
35.7 (31.3–40.2) |
31.9 (27.3–36.6) |
30.9 (26.5–35.6) |
29.8 (24.9–35.0) |
32.5 (27.6–37.8) |
| Thoracic endovascular aortic repair subgroup, % (95% C.I.) | |||||||
| 30-day mortality | – | – | – | 9.5 (2.66–22.6) |
6.3 (3.68–9.82) |
11.4 (8.18–15.3) |
13.9 (10.7–17.7) |
| 1-year mortality | – | – | – | 16.7 (6.97–31.4) |
18.8 (14.3–23.9) |
22.2 (17.8–27.0) |
25.8 (21.6–30.4) |
| Medical management, % (95% C.I.) | |||||||
| 30-day mortality | 32.8 (30.7–35.1) |
32.6 (30.5–34.8) |
30.0 (28.0–32.1) |
29.2 (27.3–31.3) |
29.8 (27.7–32.0) |
33.8 (31.5–36.1) |
28.9 (26.6–31.3) |
| 1-year mortality | 44.0 (41.7–46.4) |
41.6 (39.4–43.8) |
41.0 (38.8–43.2) |
40.3 (38.2–42.5) |
40.9 (38.7–43.2) |
43.9 (41.5–46.3) |
41.6 (39.0–44.1) |
Data has been presented individually for the years 2000 and 2011 i.e. initial and final years of the study and in 2-year intervals for the intervening period
In contrast, among the limited number of patients who underwent TEVAR, 30-day mortality changed from 9.5% in 2005 to 13.9% in year 2011 (p=0.4 for trend), and 1-year mortality changed from 16.7% in 2005 to 25.8% in year 2011 (p=0.3 for trend) respectively. Though an increase in mortality was observed from 2005 to 2011, the changes did not appear to be meet criteria for significance.
The declines in 30-day and 1-year mortality noted in the overall group hospitalized with AD, undergoing surgical repair and receiving medical therapy remained significant after adjusting for age, sex, race, and comorbidities. Between 2000 and 2011, adjusted 30-day mortality rates decreased by 5.2% (95% CI, 5.1–5.2) among all patients hospitalized with AD, by 7.3% (95% CI, 5.8–7.8) among patients undergoing surgical repair for type A dissection, by 2.9% (95% CI, 0.7–4.4) among patients undergoing surgical repair for type B dissection and by 4.5% (95% CI, 3.4–5.2%) in the subgroup of patients undergoing medical management (Figure 2). Similarly, adjusted 1-year mortality rates decreased by 6.2% (95% CI, 5.3–6.7) among all hospitalized AD patients, by 8.2% (95% CI, 6.7 – 9.1) in the surgical repair for type A dissection subgroup, by 3.9% (95% CI, 2.5–6.3) in the subgroup of patients who underwent surgical repair for type B dissection and by 4.0% (95% CI, 2.6–4.9) in the subgroup of patients who underwent medical management (Figure 3). Adjusted rates for 30-day and 1- year mortality after TEVAR could not be reliably calculated because of small numbers.
Figure 2.
Trends* in adjusted** 30-day mortality rates across the study period for the overall aortic dissection cohort, type A open surgical repair subgroup, type B open surgical repair subgroup and medical therapy subgroup
*A fitted trend line was estimated from a local regression model
**Adjusted for age, sex, race and comorbidities
Note: Adjusted rates for 30-day mortality after TEVAR could not be reliably calculated because of small numbers
Figure 3.
Trends* in adjusted** 1-year mortality rates across the study period for the overall aortic dissection cohort, type A open surgical repair subgroup, type B open surgical repair subgroup and medical therapy subgroup
*A fitted trend line was estimated from a local regression model
**Adjusted for age, sex, race and comorbidities
Note: Adjusted rates for 1-year mortality after TEVAR could not be reliably calculated because of small numbers
Discussion
Using national data for all Medicare FFS beneficiaries, we found that the hospitalization rate for AD has remained steady over the last decade. There has been a continued decline in both short-term mortality as well as long-term mortality rates in the overall group of patients hospitalized with AD, in the subgroup of patients undergoing surgical repair and being managed medically. Thoracic endovascular aortic repair has become an increasingly used therapeutic option across the study period. There has been no significant change in short- and long-term mortality associated with TEVAR.
Previous studies have reported AD incidence rates of 2–5 per 100,000 persons per year in samples of all age groups.4–6 However, given that AD is predominantly a condition of the elderly, the hospitalization rate nationally in older persons as measured in our study was approximately 10 per 100,000 person-years and remained stable across the study period. Hypertension has traditionally been recognized as an important risk factor for AD. Although, nationally the prevalence of hypertension has remained stable during this period, there has been improved control of blood pressure with antihypertensive therapy.20 Yet, a concomitant decrease in hospitalization rates for AD was not noted in our study. This may partly be due to advancements in noninvasive diagnostic technology, such as the development of multi-detector CT resulting in rapid high definition CT angiography from neck to abdomen, leading to increased detection of dissection.8, 21 Alternatively, increased prevalence of aortic dissection could be indicative of the contribution of other important etiologies of aortic dissection namely congenital cardiovascular conditions such as bicuspid aortic valve, familial conditions affecting the connective tissue like Marfan’s Syndrome, Ehler-Danlos Syndrome, Turner Syndrome or increased drug use e.g. cocaine/amphetamine.22 Males were at higher risk of hospitalization compared with females. Women were less likely to receive invasive procedures compared with men. These findings are consistent with prior literature, and potential explanation for this difference could be the increased age of women and comparatively sicker clinical presentation compared with men that could preclude the use of invasive procedures in women.23 Even in our study, the subgroup being managed medically was older and had a higher percentage of females compared to the subgroup receiving invasive management (Table 2). Finally, black patients were more likely to be hospitalized with a dissection compared with other racial groups. The reason for this increased susceptibility among this group is unclear and could be from higher prevalence and severity of uncontrolled hypertension or fundamental differences in their genetic profile.20, 22 With regards to management strategy, the hospitalization rate associated with surgical repair for type A dissection remained relatively stable throughout the study period. However, hospitalization rates for surgical repair for type B dissection were noted to decrease during the study period concomitant with an increase in hospitalization rate for TEVAR. This is likely secondary to the widespread uptake of endovascular repair as a viable therapeutic alternative for patients who were previously considered to be high risk for an open surgical intervention.
We observed improvements in both short-term and long-term outcomes among patients undergoing surgical repair (for type A and type B dissections) during the study period. In the setting of a type A dissection, surgical repair is the treatment of choice and is usually performed as an emergent, life-saving procedure. In contrast, surgical repair for type B dissections is only indicated in patients with complicated presentations (impending rupture, hypotension, signs of malperfusion or refractory hypertension).24 While a singular technological breakthrough in the operative management of this condition has not taken place in the last decade, subtle improvements on various fronts have been made. Advancements such as use of sophisticated neuroprotection techniques, development of improved intraoperative and postoperative pharmacological therapy, use of operative algorithm and standardized cannulation techniques, as well development of care teams could all have possibly contributed towards improving mortality rates.1, 3, 7, 9, 10, 25–28 Additionally, part of the improvement in mortality seen with type B dissection could be possible due to the increasing use of endovascular approach in the treatment of surgically high risk patients.
On the other hand, 70% of type B dissections are managed medically with intensive anti-hypertensive therapy. For complicated cases endovascular repair has evolved as a novel therapeutic alternative to surgical repair.24 Multiple small clinical trials at high-volume centers as well as analysis of observational data from multi-center registries have been performed evaluating the effectiveness of this as an alternative to surgical management.14, 29, 30 Based on the mortality benefit observed in these studies, a recent expert consensus statement has recommended TEVAR as initial therapy for acute complicated type B dissections.1, 24, 31 Thus, the availability of a less invasive modality for treatment of acute complicated dissection of the descending aorta, perhaps explains the increasing adoption of this procedure nationally in spite of the absence of large, multi-site randomized clinical trials, which have been difficult to conduct in this high-risk patient population. The 30-day mortality for TEVAR ranged from 6.2% to 13.9% in our study. This is consistent with the 30-day mortality rate (9.1%) reported by Conrad et al.32 from a similar patient population, early mortality rate (10.2%) reported by Frattoria et al.24 in the recent multi-disciplinary expert consensus document on the management of type B dissection and the 30-day mortality rate (10.8%) reported by White et al.33 in their report produced in collaboration with the United States Food and Drug Administration for benchmarking future performance of this new technology.
Previous data on long-term outcomes after TEVAR are scarce34 with five-year survival rate in the Medicare population noted to be approximately 58%.32 The 1-year mortality rate after this procedure, as observed in our study, ranged from 16.7% to 25.8% during the study period. Since the TEVAR mortality was unadjusted given low procedure numbers, it is difficult to comment on the cause of the year-to-year fluctuations in mortality rate. These observations could be confounded by the expansion of indications for TEVAR from thoracic aneurysm repair to repair of acute high-risk AD patients recently. Other possible explanations of changing mortality over time could be indicative of rapid adoption of the procedure outside of high-volume centers of excellence to centers with less experience, although we acknowledge this is speculative.
Finally improved short and long-term mortality outcomes were observed in the subgroup of patients receiving medical therapy. Based on the known epidemiology of AD, majority of the patients in this subgroup would appear to have a dissection in the descending aorta and are being managed medically. It also includes patients with dissections in the ascending aorta where surgery is contraindicated given high surgical risk. Hence, lack of procedure could be marker of higher risk. Based on the presence of comorbidities, patient receiving medical therapy consistently appeared to be older and had a sicker profile compared to subgroup of patients receiving invasive therapy. However, lack of information regarding the kind of medical therapy administered as well as precise clinical picture on presentation results in difficulties interpreting the results of this subgroup of patients with certainty.
There are several limitations to the current study. As our analysis was limited to Medicare FFS beneficiaries, we cannot comment on trends in patients <65 years of age or patients enrolled in Medicare Advantage programs. Although we adjusted for available potential confounders, it is possible that changing characteristics of the FFS population were not accounted for by available data. A Cox model may have been an appropriate alternative to logistic regression for the 1-year outcome analyses. However, the large sample size of our data (e.g., more than 3000 hospitals) limited our ability to fit a Cox model with hospital-specific intercepts using SAS. Because 1-year mortality was counted from the date of admission and all patients were in the fee-for-service program during the follow-up period (i.e., no patient was lost except a death event), we took an alternative approach, which was to first fit a mixed-effects model with a logit link function and hospital-specific intercepts to obtain adjusted odds ratios of annual changes in mortality rates and then convert the adjusted odd ratios to risk ratios using the Zhang and Yu method. Although this method has been used widely over the last decade, it may be an oversimplification35{McNutt, 1999 #44}{McNutt, 1999 #44} and alternative methods36, 37 could be used when data does not fit well with the Zhang and Yu’s method. Using claims data, we were unable to differentiate if the presentation of AD was acute versus sub-acute (i.e. incidental finding on radiographic imaging) or if TEVAR was being performed for type A versus type B dissection. This is important since it could affect our outcome estimates. Additionally, we do not account for the impact of patients crossing over from one treatment strategy to the other (i.e. medical to TEVAR/surgical repair) on outcomes such as mortality. However, this number was noted to be less than 2% (Appendix 6).
Conclusion
Thus with our study, we have attempted to define the current landscape of AD in an elderly population. While the hospitalization rates for the disease have been stable over the last decade, a consistent decrease in mortality was observed with management of AD. Improvement in outcomes has been particularly marked in the sub-group undergoing surgical repair (both type A and type B repair) and medical therapy across the study period.
Supplementary Material
Acknowledgments
Funding Source: This work was supported by grant U01 HL105270-03 (Center for Cardiovascular Outcomes Research at Yale University) from the National Heart, Lung and Blood Institute.
Disclosures: Dr. Krumholz reports that he is the recipient of a research grant from Medtronic, Inc. through Yale University and is chair of a cardiac scientific advisory board for UnitedHealth.
References
- 1.Hiratzka LF, Bakris GL, Beckman JA, Bersin RM, Carr VF, Casey DE, Jr, Eagle KA, Hermann LK, Isselbacher EM, Kazerooni EA, Kouchoukos NT, Lytle BW, Milewicz DM, Reich DL, Sen S, Shinn JA, Svensson LG, Williams DM. 2010 ACCF/AHA/AATS/ACR/ASA/SCA/SCAI/SIR/STS/SVM guidelines for the diagnosis and management of patients with thoracic aortic disease. A Report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines, American Association for Thoracic Surgery, American College of Radiology, American Stroke Association, Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, Society of Interventional Radiology, Society of Thoracic Surgeons, and Society for Vascular Medicine. Circulation. 2010;55:e27–e129. doi: 10.1016/j.jacc.2010.02.015. [DOI] [PubMed] [Google Scholar]
- 2.Svensson LG, Rodriguez ER. Aortic organ disease epidemic, and why do balloons pop? Circulation. 2005;112:1082–1084. doi: 10.1161/CIRCULATIONAHA.105.564682. [DOI] [PubMed] [Google Scholar]
- 3.Hagan PG, Nienaber CA, Isselbacher EM, Bruckman D, Karavite DJ, Russman PL, Evangelista A, Fattori R, Suzuki T, Oh JK, Moore AG, Malouf JF, Pape LA, Gaca C, Sechtem U, Lenferink S, Deutsch HJ, Diedrichs H, Marcos y Robles J, Llovet A, Gilon D, Das SK, Armstrong WF, Deeb GM, Eagle KA. The International Registry of Acute Aortic dissection (IRAD): New insights into an old disease. JAMA. 2000;283:897–903. doi: 10.1001/jama.283.7.897. [DOI] [PubMed] [Google Scholar]
- 4.Olsson C, Thelin S, Stahle E, Ekbom A, Granath F. Thoracic aortic aneurysm and dissection: Increasing prevalence and improved outcomes reported in a nationwide population-based study of more than 14 000 cases from 1987 to 2002. Circulation. 2006;114:2611–2618. doi: 10.1161/CIRCULATIONAHA.106.630400. [DOI] [PubMed] [Google Scholar]
- 5.Clouse WD, Hallett JW, Jr, Schaff HV, Spittell PC, Rowland CM, Ilstrup DM, Melton LJ., 3rd Acute aortic dissection: population-based incidence compared with degenerative aortic aneurysm rupture. Mayo Clin Proc. 2004;79:176–180. doi: 10.4065/79.2.176. [DOI] [PubMed] [Google Scholar]
- 6.Meszaros I, Morocz J, Szlavi J, Schmidt J, Tornóci L, Nagy L, Szép L. Epidemiology and clinicopathology of aortic dissection. Chest. 2000;117:1271–1278. doi: 10.1378/chest.117.5.1271. [DOI] [PubMed] [Google Scholar]
- 7.Geirsson A, Bavaria JE, Swarr D, Keane MG, Woo YJ, Szeto WY, Pochettino A. Fate of the residual distal and proximal aorta after acute type a dissection repair using a contemporary surgical reconstruction algorithm. Ann Thorac Surg. 2007;84:1955–1964. doi: 10.1016/j.athoracsur.2007.07.017. [DOI] [PubMed] [Google Scholar]
- 8.Mark DB, Berman DS, Budoff MJ, Carr JJ, Gerber TC, Hecht HS, Hlatky MA, Hodgson JM, Lauer MS, Miller JM, Morin RL, Mukherjee D, Poon M, Rubin GD, Schwartz RS. ACCF/ACR/AHA/NASCI/SAIP/SCAI/SCCT 2010 expert consensus document on coronary computed tomographic angiography: A report of the American College of Cardiology Foundation Task Force on Expert Consensus Documents. Circulation. 2010;55:2663–2699. [Google Scholar]
- 9.Moizumi Y, Motoyoshi N, Sakuma K, Yoshida S. Axillary artery cannulation improves operative results for acute type a aortic dissection. Ann Thorac Surg. 2005;80:77–83. doi: 10.1016/j.athoracsur.2005.01.058. [DOI] [PubMed] [Google Scholar]
- 10.Sedrakyan A, Wu A, Sedrakyan G, Diener-West M, Tranquilli M, Elefteriades J. Aprotinin use in thoracic aortic surgery: Safety and outcomes. J Thorac Cardiovasc Surg. 2006;132:909–917. doi: 10.1016/j.jtcvs.2006.06.021. [DOI] [PubMed] [Google Scholar]
- 11.Birman-Deych E, Waterman AD, Yan Y, Nilasena DS, Radford MJ, Gage BF. Accuracy of ICD-9-CM codes for identifying cardiovascular and stroke risk factors. Medical care. 2005;43:480–485. doi: 10.1097/01.mlr.0000160417.39497.a9. [DOI] [PubMed] [Google Scholar]
- 12.Wong CY, Chaudhry SI, Desai MM, Krumholz HM. Trends in comorbidity, disability, and polypharmacy in heart failure. Am J Med. 2011;124:136–143. doi: 10.1016/j.amjmed.2010.08.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Pope GC, Kautter J, Ellis RP, Ash AS, Ayanian JZ, Lezzoni LI, Ingber MJ, Levy JM, Robst J. Risk adjustment of Medicare capitation payments using the CMS-HCC model. Health Care Financ Rev. 2004;25:119–141. [PMC free article] [PubMed] [Google Scholar]
- 14.Sachs T, Pomposelli F, Hagberg R, Hamdan A, Wyers M, Giles K, Schermerhorn M. Open and endovascular repair of type B aortic dissection in the Nationwide Inpatient Sample. J Vasc Surg. 2010;52:860–866. doi: 10.1016/j.jvs.2010.05.008. [DOI] [PubMed] [Google Scholar]
- 15.Eicheldinger CR, Bonito A. More accurate racial and ethnic codes for Medicare administrative data. Health Care Financ Rev. 2008;29:27–42. [PMC free article] [PubMed] [Google Scholar]
- 16.Bratzler DW, Normand SL, Wang Y, O’Donnell WJ, Metersky M, Han LF, Rapp MT, Krumholz HM. An administrative claims model for profiling hospital 30-day mortality rates for pneumonia patients. PLoS One. 2011;6:e17401. doi: 10.1371/journal.pone.0017401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Krumholz HM, Wang Y, Mattera JA, Han LF, Ingber MJ, Roman S, Normand SL. An administrative claims model suitable for profiling hospital performance based on 30-day mortality rates among patients with an acute myocardial infarction. Circulation. 2006;113:1683–1692. doi: 10.1161/CIRCULATIONAHA.105.611186. [DOI] [PubMed] [Google Scholar]
- 18.Krumholz HM, Wang Y, Mattera JA, Lein FH, Ingber MJ, Roman S, Normand SLT. An administrative claims model suitable for profiling hospital performance based on 30-day mortality rates among patients with heart failure. Circulation. 2006;113:1693–1701. doi: 10.1161/CIRCULATIONAHA.105.611194. [DOI] [PubMed] [Google Scholar]
- 19.Zhang J, Yu KF. What’s the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. JAMA. 1998;280:1690–1691. doi: 10.1001/jama.280.19.1690. [DOI] [PubMed] [Google Scholar]
- 20.Yoon SS, Ostchega Y, Louis T. Recent trends in the prevalence of high blood pressure and its treatment and control, 1999–2008. NCHS data brief. 2010:1–8. [PubMed] [Google Scholar]
- 21.Ledbetter S, Stuk JL, Kaufman JA. Helical (spiral) CT in the evaluation of emergent thoracic aortic syndromes. Traumatic aortic rupture, aortic aneurysm, aortic dissection, intramural hematoma, and penetrating atherosclerotic ulcer. Radio Clin North Am. 1999;37:575–589. doi: 10.1016/s0033-8389(05)70112-3. [DOI] [PubMed] [Google Scholar]
- 22.LeMaire SA, Russell L. Epidemiology of thoracic aortic dissection. Nat Rev Cardiol. 2011;8:103–113. doi: 10.1038/nrcardio.2010.187. [DOI] [PubMed] [Google Scholar]
- 23.Nienaber CA, Fattori R, Mehta RH, Richartz BM, Evangelista A, Petzsch M, Cooper JV, Januzzi JL, Ince H, Sechtem U, Bossone E, Fang J, Smith DE, Isselbacher EM, Pape LA, Eagle KA. Gender-related differences in acute aortic dissection. Circulation. 2004;109:3014–3021. doi: 10.1161/01.CIR.0000130644.78677.2C. [DOI] [PubMed] [Google Scholar]
- 24.Fattori R, Cao P, De Rango P, Czerny M, Evangelista A, Nienaber C, Rousseau H, Schepens M. Interdisciplinary expert consensus document on management of type B aortic dissection. J Am Coll Cardiol. 2013;61:1661–1678. doi: 10.1016/j.jacc.2012.11.072. [DOI] [PubMed] [Google Scholar]
- 25.Bavaria JE, Brinster DR, Gorman RC, Woo YJ, Gleason T, Pochettino A. Advances in the treatment of acute type A dissection: An integrated approach. Ann Thorac Surg. 2002;74:S1848–S1852. doi: 10.1016/s0003-4975(02)04128-0. [DOI] [PubMed] [Google Scholar]
- 26.Raivio P, Suojaranta-Ylinen R, Kuitunen AH. Recombinant factor VIIa in the treatment of postoperative hemorrhage after cardiac surgery. Ann Thorac Surg. 2005;80:66–71. doi: 10.1016/j.athoracsur.2005.02.044. [DOI] [PubMed] [Google Scholar]
- 27.Tritapepe L, De Santis V, Vitale D, Nencini C, Pellegrini F, Landoni G, Toscano F, Miraldi F, Pietropaoli P. Recombinant activated factor VII for refractory bleeding after acute aortic dissection surgery: a propensity score analysis. Crit Care Med. 2007;35:1685–1690. doi: 10.1097/01.CCM.0000269033.89428.B3. [DOI] [PubMed] [Google Scholar]
- 28.Pasic M, Schubel J, Bauer M, Yankah C, Kuppe H, Weng YG, Hetzer R. Cannulation of the right axillary artery for surgery of acute type A aortic dissection. Eur J Cardiothorac Surg. 2003;24:231–236. doi: 10.1016/s1010-7940(03)00307-5. [DOI] [PubMed] [Google Scholar]
- 29.Brunt ME, Egorova NN, Moskowitz AJ. Propensity score-matched analysis of open surgical and endovascular repair for type B aortic dissection. Int J Vasc Med. 2011;2011:364046. doi: 10.1155/2011/364046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Eggebrecht H, Nienaber CA, Neuhauser M, Baumgart D, Kische S, Schmermund A, Herold U, Rehders TC, Jakob HG, Erbel R. Endovascular stent-graft placement in aortic dissection: A meta-analysis. Eur Heart J. 2006;27:489–498. doi: 10.1093/eurheartj/ehi493. [DOI] [PubMed] [Google Scholar]
- 31.Grabenwoger M, Alfonso F, Bachet J, Bonser R, Czerny M, Eggebrecht H, Evangelista A, Fattori R, Jakob H, Lonn L, Nienaber CA, Rocchi G, Rousseau H, Thompson M, Weigang E, Erbel R. Thoracic Endovascular Aortic Repair (TEVAR) for the treatment of aortic diseases: a position statement from the European Association for Cardio-Thoracic Surgery (EACTS) and the European Society of Cardiology (ESC), in collaboration with the European Association of Percutaneous Cardiovascular Interventions (EAPCI) Eur J Cardiothorac Surg. 2012;42:17–24. doi: 10.1093/ejcts/ezs107. [DOI] [PubMed] [Google Scholar]
- 32.Conrad MF, Ergul EA, Patel VI, Paruchuri V, Kwolek CJ, Cambria RP. Management of diseases of the descending thoracic aorta in the endovascular era: A Medicare population study. Ann Surg. 2010;252:603–610. doi: 10.1097/SLA.0b013e3181f4eaef. [DOI] [PubMed] [Google Scholar]
- 33.White RA, Miller DC, Criado FJ, Dake MD, Diethrich EB, Greenberg RK, Piccolo RS, Siami FS. Report on the results of thoracic endovascular aortic repair for acute, complicated, type B aortic dissection at 30 days and 1 year from a multidisciplinary subcommittee of the Society for Vascular Surgery Outcomes Committee. J Vasc Surg. 2011;53:1082–1090. doi: 10.1016/j.jvs.2010.11.124. [DOI] [PubMed] [Google Scholar]
- 34.Tsai TT, Fattori R, Trimarchi S, Isselbacher E, Myrmel T, Evangelista A, Hutchison S, Sechtem U, Cooper JV, Smith DE, Pape L, Froehlich J, Raghupathy A, Januzzi JL, Eagle KA, Nienaber CA. Long-term survival in patients presenting with type B acute aortic dissection: Insights from the International Registry of Acute Aortic Dissection. Circulation. 2006;114:2226–2231. doi: 10.1161/CIRCULATIONAHA.106.622340. [DOI] [PubMed] [Google Scholar]
- 35.McNutt LA, Hafner JP, Xue X. Correcting the odds ratio in cohort studies of common outcomes. JAMA. 1999;282:529. doi: 10.1001/jama.282.6.529. [DOI] [PubMed] [Google Scholar]
- 36.Kleinman LC, Norton EC. What’s the risk? A simple approach for estimating adjusted risk measures from nonlinear models including logistic regression. Health Serv Res. 2009;44:288–302. doi: 10.1111/j.1475-6773.2008.00900.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.McNutt LA, Wu C, Xue X, Hafner JP. Estimating the relative risk in cohort studies and clinical trials of common outcomes. Am J Epidemiol. 2003;157:940–943. doi: 10.1093/aje/kwg074. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.