- Email: [email protected]
Insurance Type and Solid Organ Transplantation Outcomes: A Historical Perspective on How Medicaid Expansion Might Impact Transplantation Outcomes
Insurance Type and Solid Organ Transplantation Outcomes: A Historical Perspective on How Medicaid Expansion Might Impact Transplantation Outcomes Derek A DuBay, MD, MSPH, FACS, Paul A MacLennan, PhD, Rhiannon D Reed, MPH, Brittany A Shelton, MPH, David T Redden, PhD, Mona Fouad, MD, MPH, Michelle Y Martin, Stephen H Gray, MD, MSPH, Jared A White, MD, Devin E Eckhoff, MD, FACS, Jayme E Locke, MD, MPH, FACS
PhD,
The number of Medicaid beneficiaries has increased under the Affordable Care Act, improving access to solid organ transplantation in this disadvantaged patient cohort. It is unclear what impact Medicaid expansion will have on transplantation outcomes. We performed a retrospective cohort analysis to measure the frequency and variation in Medicaid transplantation and post-transplantation survival in Medicaid patients. STUDY DESIGN: Adult heart, lung, liver, and renal transplant recipients between 2002 and 2011 (n ¼ 169,194) reported to the Scientific Registry of Transplant Recipients were identified. Transplant recipients were classified based on insurance status (private, Medicare or Medicaid). Outcomes measures included 5-year post-transplantation survival, summarized using Kaplan-Meier curves and compared with log-rank tests. Organ-specific Cox proportional hazards models were used to adjust for donor and recipient factors. RESULTS: Medicaid patients comprised 8.6% of all organ transplant recipients. Fewer transplantations were performed than expected among Medicaid beneficiaries for all organs except liver (liver: observed to expected ratio ¼ 1.21; 95% CI, 0.68e1.90; heart: observed to expected ratio ¼ 0.89; 95% CI, 0.44e1.49; lung: observed to expected ratio ¼ 0.57; 95% CI, 0.22e1.06; renal: observed to expected ratio ¼ 0.32; 95% CI, 0.08e0.72). Medicaid transplant recipients were listed with more severe organ failure and experienced shorter transplant wait times. Post-transplantation survival was lower in Medicaid patients compared with private insurance for all organs. Post-transplantation survival in Medicaid patients was similar to Medicare patients for heart, liver, and renal but lower in lung. CONCLUSIONS: Medicaid organ transplant beneficiaries had significantly lower survival compared with privately insured beneficiaries. The more severe organ failure among Medicaid beneficiaries at the time of listing, suggested a pattern of late referral, which might account for worse outcomes. Implementation of the Affordable Care Act gives the opportunity to develop the necessary infrastructure to ensure timely transplantation referrals and improve long-term outcomes in this vulnerable population. (J Am Coll Surg 2016;223:611e620. 2016 by the American College of Surgeons. Published by Elsevier Inc. All rights reserved.) BACKGROUND:
Disclosure Information: Nothing to disclose. Support: This research was funded by National Institutes of Health grant numbers 1 K23 DK091514 (DD), 1 R03 DK106432 (DD), and 1 K23 DK 103918 (JL). Disclaimer: The interpretation and reporting of these data are the responsibility of the author(s) and in no way should be seen as an official policy of or interpretation by Scientific Registry of Transplant Recipients, Organ Procurement and Transplantation Network, or the US Government.
From the Department of Surgery-Transplantation (DuBay, MacLennan, Reed, Shelton, Gray, White, Eckhoff, Locke), School of Public HealthBiostatistics (Redden), Department of Medicine-Preventive Medicine (Fouad), University of Alabama at Birmingham, Birmingham, AL, and Department of Medicine-Preventive Medicine, University of Tennessee, Memphis, TN (Martin). Correspondence address: Derek A DuBay, MD, MSPH, FACS, Department of Surgery-Transplantation, University of Alabama at Birmingham, ZRB 701, 1720 2nd Ave S, Birmingham, AL 35294-0007. email: [email protected]
Received February 3, 2016; Revised July 11, 2016; Accepted July 12, 2016.
ª 2016 by the American College of Surgeons. Published by Elsevier Inc. All rights reserved.
611
http://dx.doi.org/10.1016/j.jamcollsurg.2016.07.004 ISSN 1072-7515/16
DuBay et al
612
Medicaid Expansion and Transplantation Outcomes
Abbreviations and Acronyms
ACA HR MELD O/E SRTR
¼ ¼ ¼ ¼ ¼
Affordable Care Act hazard ratio Model for End-Stage Liver Disease observed to expected ratio Scientific Registry of Transplant Recipients
The implementation of the Affordable Care Act (ACA) has increased the number of Medicaid beneficiaries in the United States substantially. Current estimates suggest that 50% of uninsured Americans have or will gain coverage via Medicaid expansion.1-4 For the 31 states adopting Medicaid expansion, financial eligibility requirements decreased to 138% of the federal poverty level, resulting in 12.3 million new Medicaid beneficiaries.1-4 Overall uninsured rates have decreased from 15.7% before ACA to 9.2% currently.1-4 Access to solid organ transplantation has increased for this previously disadvantaged population. However, it is unclear what impact, if any, implementation of the ACA Medicaid expansion program will have on transplantation outcomes. Non-private insurance status has been demonstrated to be an adverse marker of transplantation outcomes across solid organ transplantation.5-13 With introduction of the ACA, there will be more Medicaid beneficiaries who can access organ transplantation and, given earlier reports of inferior outcomes among this population,5,7-9 transplantation centers might encounter observed survival rates considerably lower than expected. In the context of the current regulatory environment, understanding the impact of the implementation of the ACA on transplantation outcomes is of paramount importance. To this end, we performed a retrospective cohort analysis of solid organ transplant recipients. Our approach was to examine recent historic transplantation outcomes data in Medicaid, Medicare, and private insured recipients. Outcomes measures include frequency of Medicaid transplantation across organ groups; variation in Medicaid transplantation in the United States; and posttransplantation survival among Medicaid beneficiaries compared with patients with private insurance or Medicare.
METHODS Data source This study used data from the Scientific Registry of Transplant Recipients (SRTR). The SRTR system includes data submitted by the members of the Organ Procurement and Transplantation Network on all donors, waitlisted candidates, and transplant recipients in the
J Am Coll Surg
United States. The Health Resources and Services Administration of the US Department of Health and Human Services provides the oversight to the activities of the Organ Procurement and Transplantation Network and SRTR contractors. The University of Alabama at Birmingham IRB reviewed and approved the study (#E121024003). Study population Deceased donor, adult heart, lung, liver, and renal transplant recipients reported to the SRTR between January 1, 2002 and December 31, 2011 were identified (heart, n ¼ 17,239; lung, n ¼ 13,206; liver, n ¼ 48,201; renal, n ¼ 90,548). This time interval pre-dates implementation of the ACA (Table 1). Transplant recipients were classified by insurance type as private insurance, Medicare, or Medicaid, based on SRTR classifications. Insurance status was assigned based on the patient’s primary insurance at the time of transplantation. Outcomes ascertainment Post-transplantation survival was defined as time from transplantation to death or last follow-up, and was censored at 5 years post-transplantation. Death dates were supplemented by information from the Centers for Medicare and Medicaid Services and the Limited Access Death Master File available from the National Technical Information Service through 2013. Other outcomes included measures of access to transplantation as a function of insurance status, including observed to expected transplantation ratios (O/E), waitlist time, and measures of end-stage organ disease at the time of transplantation (heart status,14 lung allocation score,15,16 Model for End-Stage Liver Disease [MELD],17 and cumulative years of dialysis).18 Expected insurance distribution for the O/E ratios was calculated using health coverage status provided by the Kaiser Family Foundation State Health Facts from 2011 at both the national and state levels,19 and was calculated as the proportion of Medicaid patients receiving transplants in each organ group divided by the proportion of the population insured with Medicaid between 2010 and 2011. Other patient covariates, measured at the time of transplantation, included recipient age, sex, race/ethnicity, BMI, education, select comorbidities (diabetes and hypertension), albumin, and renal function (creatinine). Additional outcomes measures included post-transplantation length of stay. Statistical analysis Baseline characteristics among private, Medicare, and Medicaid solid organ recipients (ie heart, lung, liver,
Vol. 223, No. 4, October 2016
DuBay et al
Medicaid Expansion and Transplantation Outcomes
613
Table 1. Baseline Characteristics of Heart, Lung, Liver, and Renal Transplant Recipients (2002 to 2011) Stratified by Public Insurance, Medicare, or Medicaid Characteristic
Sample size, n Organ type, n (%) Liver Renal Heart Lung Age, y, mean (SD) Female sex, n (%) Race/ethnicity, n (%) Caucasian African American Asian Hispanic Other BMI, kg/m2, mean (SD)* Education,y n (%) Less than high school High school Some college 4 y college Diabetes,* n (%) Hypertension,y n (%) Albumin, g/dL, mean (SD)z Creatinine, mg/dL, mean (SD), excluding renal recipientsz
Total
Private
Medicare
Medicaid
169,194
74,800
79,803
14,591
48,201 90,548 17,239 13,206 52.2 61,978
(28.5) (53.5) (10.2) (7.8) (12.3) (36.6)
30,941 25,970 9,925 7,964 51.9 25,817
(41.4) (34.7) (13.3) (10.6) (11.0) (34.5)
10,158 60,217 5,157 4,271 53.6 30,179
(12.7) (75.5) (6.5) (5.4) (13.1) (37.8)
7,102 4,361 2,157 971 46.6 5,982
(48.7) (29.9) (14.8) (6.7) (12.9) (41.0)
101,081 37,366 7,941 20,449 2,355 27.6
(59.7) (22.1) (4.7) (12.1) (1.4) (6.6)
53,303 10,963 3,429 6,367 738 27.5
(71.3) (14.6) (4.6) (8.5) (1.0) (7.0)
40,689 23,260 3,630 10,870 1,353 27.7
(51.0) (29.1) (4.6) (13.6) (1.7) (6.2)
7,089 3,143 882 3,212 264 27.0
(48.6) (21.5) (6.1) (22.0) (1.8) (6.1)
8,781 66,226 34,059 30,978 49,111 89,637 3.8 1.12
(6.3) (47.3) (24.3) (22.1) (29.1) (59.1) (0.7) (0.6)
1,839 25,292 16,157 17,906 19,107 30,810 3.7 1.12
(3.0) (41.3) (26.4) (29.3) (25.6) (46.6) (0.7) (0.6)
5,383 34,084 15,174 12,033 26,484 53,342 3.9 1.15
(8.0) (50.7) (23.4) (17.9) (33.2) (73.5) (0.6) (0.7)
1,559 6,850 2,188 1,039 3,520 5,485 3.6 1.10
(13.4) (58.9) (18.8) (8.9) (24.2) (42.2) (0.8) (0.7)
p Value
<0.001
<0.001 <0.001 <0.001
<0.001 <0.001
<0.001 <0.001 <0.001 <0.001
*Missing for <10%. y Missing for 10% to 20%. z Missing for >20%.
and renal) were compared. Continuous variables were analyzed using ANOVA or Kruskal-Wallis tests, and categorical variables using chi-square tests. The O/E ratios, stratified by organ-specific transplantation, were calculated for private, Medicare, and Medicaid insurance transplant recipients. Number of Medicaid transplants and O/ E ratios were compared for all 50 States and the District of Columbia. Survival analysis Patient survival was summarized using Kaplan-Meier survival curves stratified by insurance group and compared using log-rank tests. Organ-specific Cox proportional hazards models were built using a backwards step-wise selection, with p < 0.05 on bivariate analysis and were used to adjust for donor and recipient factors as described here, with the most parsimonious model chosen through minimization of Akaike’s Information Criteria. Although education is reported, age, race, and sex were the only sociodemographic factors considered in adjusted analyses. Recipient age, transplantation year (2002 to 2011), BMI,
albumin, and creatinine were treated as continuous covariates. Sex and race (coded as white, black, Hispanic, and other) were categorical variables. Heart status was a categorical variable; MELD and years of dialysis were continuous variables. Adjusted survival curves were generated from the fitted Cox models using the direct adjusted survivor function,20 and proportional hazards assumptions were tested in the fitted Cox models with Schoenfeld residuals and time-dependent covariates. Spline regression was used to generate time-dependent hazard ratios (HRs) when the assumptions were violated for insurance status. Patients missing data for variables included in the models were excluded from adjusted analyses. Sensitivity analyses Sensitivity analyses were performed for pre-emptive renal transplantations (transplantation before initiation of dialysis). This analysis was done to account for renal recipients who had no dialysis-associated morbidity. Additional models were run with the variables with significant missingness excluded, and inferences were
614
DuBay et al
Medicaid Expansion and Transplantation Outcomes
consistent with those reported in this article. Correlation between Medicaid transplantation center volume and 1year post-transplantation survival was determined for all organ types using Spearman’s correlation coefficient. All of the tests were 2-sided with statistical significance set at a ¼ 0.05. Analyses were performed using SAS software, version 9.3 (SAS Institute). Additional supporting information can be found in the online version of this article.
RESULTS Study population Only 8.6% of solid organ transplantations were performed for patients with Medicaid insurance (Table 1). Medicaid transplant recipients were younger and more often female. Medicaid patients were more frequently non-Caucasian than Medicare and private. Medicaid transplant recipients were less likely to have a college or beyond education than private or Medicare recipients. The highest frequency of Medicaid transplants was performed for liver (14.7%) and heart (12.5%) with the lowest frequency performed
J Am Coll Surg
for lung (7.4%) and renal (4.8%) (Fig. 1A and eTables 1 to 4, available online). Access to transplantation Fewer transplantations were performed for Medicaid patients than was expected for all organ types except liver (Fig. 1B). The O/E ratios for transplantation in Medicaid patients were liver: 1.21 (95% CI, 0.68e1.90); heart: 0.89 (95% CI, 0.44e1.49); lung: 0.57 (95% CI, 0.22e1.06); and renal: 0.32 (95% CI, 0.08e0.72). The number of transplantations performed in Medicaid patients, by state of residence, for all organ types combined, varied from 0 to 304 (Fig. 2A and eTable 5, available online). The O/E ratio for all organ transplantation in patients insured with Medicaid ranged from 0 to 1.4, but was only 1.0 in 3 states (Fig. 2B and eTable 5, available online). Heart transplantation outcomes Medicaid patients receiving a heart were more likely to be transplanted as a status 1A (46.9%), the highest-priority
Figure 1. (A) Percentage of heart, lung, liver, and renal transplantation stratified by organ type. The percentages correspond to the number of transplantations performed in Medicaid patients divided by the total number of transplantations performed for each organ as seen in Table 1 (ie liver transplant: 7,102 / 48,201 100%). (B) Observed to expected ratios for heart, lung, liver, and renal transplantation stratified by insurance group.
Vol. 223, No. 4, October 2016
DuBay et al
Medicaid Expansion and Transplantation Outcomes
615
Figure 2. (A) Number of solid organ transplantations performed in Medicaid patients for each state. (B) Observed to expected ratios for Medicaid transplantation for each state.
listing, compared with private (44.3%) or Medicare (40.6%) patients (p < 0.001; Table 2). Medicaid heart transplant recipients also had a significantly shorter wait time (66 days) compared with private (75 days) and Medicare (91 days) patients (p < 0.001). Post-heart transplantation adjusted survival was lower in Medicaid patients compared with private, but similar to Medicare (Fig. 3A). Among Medicaid recipients, heart transplantation survival was 0.8% lower than Medicare and 4.7% lower than private at 3 years and 0.9% lower than Medicare and 5.9% lower than private at 5 years. Compared with
private recipients, the adjusted Medicaid mortality HR for heart transplantation was 1.10 in the first 2 years post-transplantation (95% CI, 0.97e1.26), increasing to 1.77 after 2 years (95% CI, 1.49e2.09; Table 3). Similar survival HRs were observed with sensitivity models that included recipient center as a covariate to control for geography. There was no significant correlation between heart transplantation survival and center Medicaid volume (r ¼ 0.137, p ¼ 0.12). Compared with non-Medicaid patients, the median length of stay was 2 days longer for heart transplant recipients with Medicaid insurance (Table 2).
616
DuBay et al
J Am Coll Surg
Medicaid Expansion and Transplantation Outcomes
Table 2. Waiting Time, Organ-Specific Disease Score, Post-Transplantation Length of Stay, and Post-Transplantation Survival Stratified by Organ Type and Insurance Group Variable
Heart, n Waiting time, d, median (IQR) Status at Tx, n (%) Status 1A Status 1B Status 2 Postoperative LOS, d, median (IQR) Follow-up time, y, median (IQR) 5-y survival, %* Lung, n Waiting time, d, median (IQR) Lung allocation score at Tx,y median (IQR) Postoperative LOS, d, median (IQR) Follow-up time, y, median (IQR) 5-y survival, %* Liver, n Waiting time, d, median (IQR) Model for End-Stage Liver Disease at Tx, median (IQR) Postoperative LOS, d, median (IQR) Follow-up time, y, median (IQR) 5-y survival, %* Renal, n Waiting time, d, median (IQR) Dialysis time, y, median (IQR) Postoperative LOS, d, median (IQR) Follow-up time, y, median (IQR) 5-y survival, %*
Total
Private
Medicare
Medicaid
17,239 78 (23e223)
9,925 75 (23e209)
5,157 91 (27e271)
2,157 66 (20e181)
7,495 6,778 2,960 14 3.9
(43.5) (39.3) (17.2) (10e21) (1.8e5.0) 74.3 13,206 107 (29e329.5)
4,394 3,789 1,741 13 4.0
(44.3) (38.2) (17.5) (10e20) (1.9e5.0) 77.2 7,964 108 (30e325.5)
2,090 2,134 930 14 3.1
(40.6) (41.4) (18.0) (10e22) (1.1e5.0) 71.1 4,271 106 (28e342.5)
1,011 855 289 15 3.8
(46.9) (39.7) (13.4) (10e22) (1.8e5.0) 67.9 971 105.5 (31e331)
p Value
<0.001 <0.001
<0.001 <0.001 <0.001 0.8
39.1 (34e48) 15 (10e26) 2.7 (1.0e5.0) 52.0 48,201 72 (15e256)
39.6 (35e49) 15 (10e25) 2.9 (1.1e5.0) 54.9 30,941 71 (14e248)
38.1 (34e47) 15 (10e26) 2.1 (1.0e4.3) 47.6 10,158 90 (20e331)
39.1 (34e49) 17 (12e28) 2.3 (1.0e4.6) 45.0 7,102 51 (10e203)
<0.001 <0.001 <0.001 <0.001
19 (14e27) 10 (7e17) 3.1 (1.1e5.0) 70.1 90,548 704 (302e1222) 3.2 (1.5e5.2) 6 (5e9) 3.7 (1.9e5.0) 80.2
19 (13e27) 10 (7e16) 3.5 (1.3e5.0) 72.3 25,970 527 (223e956) 1.8 (0.5e3.1) 5 (4e8) 4.0 (2.0e5.0) 84.6
18 (13e25) 10 (7e18) 3.0 (1.0e5.0) 64.8 60,217 778 (351e1314) 3.8 (2.2e5.7) 6 (5e9) 3.4 (1.8e5.0) 78.0
21 (15e30) 11 (8e19) 3.0 (1.1e5.0) 67.6 4,361 799 (326e1441) 3.9 (1.9e6.4) 6 (5e9) 3.8 (1.9e5.0) 82.9
<0.001 <0.001 <0.001 <0.001
<0.001
<0.001 <0.001 <0.001 <0.001 <0.001
*Estimated 5-year survival (unadjusted) calculated via Kaplan-Meier methods and compared using log-rank test. y Lung allocation score missing for 50% before implementation in 2006. IQR, interquartile range; LOS, length of stay; Tx, transplant.
Lung transplantation outcomes The long allocation score, a priority lung listing score, for Medicaid recipients (39.1) was less than private (39.6), but more than Medicare (38.1) patients at the time of transplantation (Table 2). There was no difference in waiting time between insurance cohorts for lung transplantations (p ¼ 0.8). Post-lung transplantation adjusted survival in Medicaid recipients was significantly worse compared with Medicare and private insurance recipients (Fig. 3B). Among Medicaid recipients, post-lung transplantation survival was 3.3% lower than Medicare and 6.7% lower than private at 3 years and 3.8% lower than Medicare and 8.0% lower than private at 5 years. Compared with private, the adjusted Medicaid mortality HR for lung transplantation was 1.15 within the first 2 years post-transplantation (95% CI, 1.00e1.33), then
increased to 1.44 after 2 years (95% CI, 1.20e1.72; Table 3). Similar survival HRs were observed with sensitivity models that included recipient center as a covariate to control for geography. There was no significant correlation between lung transplantation survival and center Medicaid volume (r ¼ 0.058; p ¼ 0.64). Compared with non-Medicaid patients, median length of stay was 2 days longer for lung transplant recipients with Medicaid insurance (Table 2). Liver transplantation outcomes The lab-MELD score, a priority liver wait listing score, was highest in Medicaid21 compared with private19 and Medicare18 patients at the time of transplantation (Table 2). Medicaid liver transplant recipients also had shorter wait times (51 days) compared with private (71 days) and
Vol. 223, No. 4, October 2016
DuBay et al
Medicaid Expansion and Transplantation Outcomes
617
Figure 3. Direct adjusted survivor functions for (A) heart, (B) lung, (C) liver, and (D) renal transplant recipients stratified by insurance type.
Medicare (90 days) patients. Post-liver transplantation adjusted survival was lower in Medicaid patients compared with private but similar to Medicare (Fig. 3C). Compared with private recipients, liver transplantation survival was 2.8% lower in Medicaid at 1 year, 4.5% lower at 3 years, and 5.4% lower at 5 years. Compared with private, the adjusted Medicaid mortality HR for liver transplantation was 1.25 (95% CI, 1.18e1.32; Table 3). Similar survival HRs were observed with sensitivity models that included recipient center as a covariate to control for geography. There was no significant correlation between liver transplantation survival and center Medicaid volume (r ¼ 0.065; p ¼ 0.48). Compared with non-Medicaid patients, median length of stay was 1 day longer for liver transplant recipients with Medicaid insurance (Table 2). Renal transplantation outcomes Medicaid renal transplantation patients had the longest time spent on dialysis (3.9 years) compared with private
(1.8 years) and Medicare (3.8 years) patients (Table 2). Post-renal transplantation adjusted survival was lower in Medicaid patients compared with private, but similar to Medicare (Fig. 3D). Compared with private recipients, renal transplantation survival for Medicaid patients was 1.4% lower at 1 year, 3.0% lower at 3 years, and 4.9% lower at 5 years. Similar survival distributions were observed when excluding pre-emptive renal transplantations. Compared with privately insured recipients, the adjusted Medicaid mortality HR for renal transplantation was 1.37 (95% CI, 1.24e1.51; Table 3). Similar survival HRs were observed with sensitivity models that included recipient center as a covariate to control for geography. There was a statistically significant inverse correlation between renal transplantation survival and center Medicaid volume (r ¼ 0.346; p 0.0001). Median length of stay was not different in Medicaid and non-Medicaid renal transplant recipients (Table 2).
618
DuBay et al
Medicaid Expansion and Transplantation Outcomes
Table 3. Multivariable Analysis of Mortality Risk Factors after Transplantation Stratified by Organ Type Organ type
Heart* Within 2 y Medicare vs private Medicaid vs private After 2 yearsy Medicare vs private Medicaid vs private Lungz Within 2 y Medicare vs private Medicaid vs private After 2 yy Medicare vs private Medicaid vs private Liverx Medicare vs private Medicaid vs private Renaljj Medicare vs private Medicaid vs private
Adjusted hazard ratio
95% CI
p Value
1.18 1.10
1.07e1.30 0.97e1.26
<0.001 0.15
1.39 1.77
1.20e1.60 1.49e2.09
<0.001 <0.001
1.10 1.15
1.01e1.19 1.00e1.33
0.03 0.05
1.18 1.44
1.06e1.31 1.20e1.72
0.004 <0.001
1.24 1.25
1.18e1.30 1.18e1.32
<0.001 <0.001
1.36 1.37
1.30e1.42 1.24e1.51
<0.001 <0.001
*Adjusted for recipient age, race, sex, heart status, diabetes, hypertension, albumin, creatinine, and transplantation year. y Model splined at 2 years post-transplantation to account for violation of the proportional hazards assumption by the Medicare variable. Patients contributed time at risk for the first 2 years, at which point they were censored if they were still alive. Those patients who died within 2 years only contributed time at risk until death date and were not included in the post-2 year estimate of risk. The post-2 year estimate includes only those who survived for more than 2 years post-transplantation with time at risk beginning at 2 years and continuing until death or 5 years posttransplantation. z Adjusted for recipient age, race, sex, hypertension, albumin, creatinine, and transplantation year. x Adjusted for age, race, sex, diabetes, hypertension, Model for End-Stage Liver Disease, and transplantation year. jj Adjusted for age, race, sex, years on dialysis, diabetes, hypertension, creatinine, and transplantation year.
Transplantation outcomes in Medicaid expansion states Of all Medicaid transplantations performed, 69.8% were among states that have implemented Medicaid expansion as of February 2015, and 30.2% were among those not currently planning to expand. Among the 31 states (including the District of Columbia) that have implemented Medicaid expansion,21 the mean O/E ratio for Medicaid organ transplantations was similar compared with those not planning to expand (expansion states: 0.61 vs non-expansion states: 0.59; p ¼ 0.89). Among all organs, there was no difference in 5-year survival for Medicaid organ transplantations among expansion states
J Am Coll Surg
compared with non-expansion states (expansion states: 70.6% vs non-expansion states: 70.4%; p ¼ 0.91). There was no difference in 5-year survival for Medicaid heart transplantation (expansion states: 68.5%; non-expansion states: 67.8%; p ¼ 0.96), or Medicaid lung transplantation (expansion states: 45.6% vs non-expansion states: 42.6%; p ¼ 0.72). Five-year survival was significantly lower in Medicaid liver transplantation in expansion states (expansion states: 65.7% vs non-expansion states: 70.4%; p ¼ 0.005). There was no difference in 5-year survival in Medicaid renal transplantation in expansion states (expansion states: 83.5% vs non-expansion states: 82.1%; p ¼ 0.36).
DISCUSSION An objective of Medicaid expansion is to ensure access to quality, culturally competent care for vulnerable populations.22 The Medicaid program currently provides transplantation care to a diverse patient population. Despite providing transplantation care to a higher proportion of vulnerable patients, there remains a large disparity in the frequency of solid organ transplantation in Medicaid patients. The O/E transplantation ratio for Medicaid was <1.0 for all organ types except liver. In addition, only 3 of 50 states performed solid organ transplantations at a frequency equal to the proportion of population insured by Medicaid. This is an especially alarming statistic because higher rates of organ dysfunction and failure are observed in patients with lower socioeconomic status,23-25 the same patients who are over-represented in the Medicaid population. Development of an infrastructure to increase timely referral as a part of the Medicaid expansion program might improve outcomes. Currently, Medicaid transplant recipients were listed with more severe organ failure, experienced shorter transplantation wait times, and longer time on dialysis, hallmarks of a pattern of late referral for transplantation evaluation. It is likely that Medicaid patients in many states will be managed via some form of managed care organization in the future.21 Managed care organization can reduce barriers to transplantation and increase the likelihood of early referral via prompt diagnosis and an established network of transplantation providers. Earlier referral can ameliorate the advanced organ failure observed in the Medicaid population at the time of transplantation. A second challenge for the Medicaid Expansion Program is to improve post-transplantation survival outcomes. Compared with private insurance, post-transplantation survival in Medicaid patients dropped off soon after transplantation and then continued to slowly diverge for all organ types. Similar survival outcomes were observed for both
Vol. 223, No. 4, October 2016
DuBay et al
Medicaid and Medicare patients, despite the fact that these public insurance programs cover vastly different patient populations. Medicare covers the elderly and disabled, and Medicaid covers patients living at or near poverty. Worsened outcomes among Medicaid patients are especially disturbing, given the younger age of the Medicaid cohort when compared with Medicare patients. Although Medicaid patients had higher disease severity scores at the time of transplantation, post-transplantation survival was reduced significantly, even when controlling for this covariate. Studies demonstrate that adverse predictors of posttransplantation survival were overexpressed in the Medicaid population, especially patients with lower education and lower social support.7,8,10,24 We expected that there would be a volume to outcomes relationship, with high-volume Medicaid transplantation centers having a discriminating evaluation process and/or a more rigorous infrastructure for post-transplantation care. However, no statistical relationship between Medicaid transplantation volume and post-transplantation survival was found. There was also no statistically significant difference in post-transplantation survival outcomes between states that adopted Medicaid expansion and those that did not. Achieving good outcomes in this vulnerable population was a challenge for all centers. It might be advisable for transplantation programs to use Medicaid insurance status as a surrogate marker of poorer post-transplantation survival and implement specific post-transplantation care practices in an attempt to ameliorate decreased survival observed in this population. Policies that encourage more frequent long-term followup clinic appointments in Medicaid-insured patients might also reduce the disparity in post-transplantation survival outcomes. Improving long-term survival outcomes is especially important, given the current practice of using measures of utility in the lung allocation system16 and the consideration of adding utility to the renal26 and liver27 allocation systems. Younger Medicaid patients have high utility potential (ie should have more life-years gained with transplantation), but only if improved posttransplantation survival rates can be achieved. This study has several limitations inherent to retrospective analyses of large cohort studies, including limited data granularity and missingness. Not all patients were followed for 5 years for the survival analysis. In addition, these data did not allow for a direct measure of access to transplantation, as they lack referral information. A better measure of transplantation access would be O/E ratio of Medicaid patients referred for transplantation, but SRTR data lack this information. These analyses used past Medicaid outcomes to predict how the ACA will impact
Medicaid Expansion and Transplantation Outcomes
619
transplantation outcomes, although many of those gaining insurance via Medicaid expansion were without insurance previously. Although uninsured would be a better comparison group, this group has not had transplantation coverage and outcomes in this group are not known. In addition, the assumption of these analyses was that past Medicaid outcomes will be predictive of future outcomes, which might or might not be the case, given the change in the population insured by Medicaid expansion efforts. Finally, insurance type was assigned at the time of transplantation, and it is possible that patients might have changed their insurance type during the follow-up period of the study.
CONCLUSIONS A higher proportion of women and minorities with Medicaid insurance received a solid organ transplantation compared with private insurance or Medicare. However, with the exception of liver, Medicaid patients received transplants at a much lower proportion than would be expected for all other organ groups. The ACA affords the opportunity to develop and implement strategies that will increase overall referrals and the likelihood of early referral for transplantation evaluation. In addition, current ACA implementation practices can improve Medicaid practice infrastructure to reduce the significant disparity in post-transplantation survival between Medicaid and private insurance recipients. Author Contributions Study conception and design: DuBay, MacLennan, Reed, Redden, Fouad, Martin, Gray, White, Eckhoff, Locke Acquisition of data: DuBay, MacLennan, Reed, Locke Analysis and interpretation of data: DuBay, MacLennan, Reed, Shelton, Redden, Locke Drafting of manuscript: DuBay, MacLennan, Reed, Locke, Critical revision: DuBay, MacLennan, Reed, Shelton, Redden, Fouad, Martin, Gray, White, Eckhoff, Locke Acknowledgment: The University of Alabama at Birmingham Comprehensive Transplant Institute Outcomes Research Center provided methodological and statistical support. The data reported here have been supplied by the United Network for Organ Sharing as the contractor for the Organ Procurement and Transplantation Network and the Minneapolis Medical Research Foundation as the contractor for the SRTR. REFERENCES 1. Obama Care Facts. Dispelling the myths. Available at: http://obamacarefacts.com/obamacares-medicaid-expansion.php. Accessed December 17, 2013.
620
DuBay et al
Medicaid Expansion and Transplantation Outcomes
2. US Department of Health and Human Services. About the law. Health and Human Services: healthcare. Available at: http://www.hhs.gov/healthcare/rights/. Accessed August 18, 2014. 3. Henry J Kaiser Family Foundation. The coverage gap: uninsured poor adults in states that do not expand Medicaid. Health Reform. Available at: http://kff.org/health-reform/issue-brief/ the-coverage-gap-uninsured-poor-adults-in-states-that-do-notexpand-medicaid/. Accessed October 28, 2014. 4. Tavernise S, Gebeloff R. Millions of poor are left uncovered by health law. The New York Times October 2, 2013. 5. Allen JG, Arnaoutakis GJ, Orens JB, et al. Insurance status is an independent predictor of long-term survival after lung transplantation in the United States. J Heart Lung Transplant 2011;30:45e53. 6. Kemmer N, Zacharias V, Kaiser TE, Neff GW. Access to liver transplantation in the MELD era: role of ethnicity and insurance. Dig Dis Sci 2009;54:1794e1797. 7. Allen JG, Weiss ES, Arnaoutakis GJ, et al. Insurance and education predict long-term survival after orthotopic heart transplantation in the United States. J Heart Lung Transplant 2012;31:52e60. 8. Goldfarb-Rumyantzev AS, Koford JK, Baird BC, et al. Role of socioeconomic status in kidney transplant outcome. Clin J Am Soc Nephrol 2006;1:313e322. 9. Yoo HY, Thuluvath PJ. Outcome of liver transplantation in adult recipients: influence of neighborhood income, education, and insurance. Liver Transpl 2004;10:235e243. 10. Keith D, Ashby VB, Port FK, Leichtman AB. Insurance type and minority status associated with large disparities in prelisting dialysis among candidates for kidney transplantation. Clin J Am Soc Nephrol 2008;3:463e470. 11. Bryce CL, Angus DC, Arnold RM, et al. Sociodemographic differences in early access to liver transplantation services. Am J Transplant 2009;9:2092e2101. 12. Johansen KL, Zhang R, Huang Y, et al. Association of race and insurance type with delayed assessment for kidney transplantation among patients initiating dialysis in the United States. Clin J Am Soc Nephrol 2012;7:1490e1497. 13. Glueckert LN, Redden D, Thompson MA, et al. What liver transplant outcomes can be expected in the uninsured who become insured via the Affordable Care Act? Am J Transplant 2013;13:1533e1540. 14. Mehra MR, Kobashigawa J, Starling R, et al. Listing criteria for heart transplantation: International Society for Heart and Lung Transplantation guidelines for the care of cardiac transplant candidatesd2006. J Heart Lung Transplant 2006;25: 1024e1042.
J Am Coll Surg
15. Final Rule. Department of Health and Human Services. Organ Procurement and Transplantation Network. Fed Reg 42 CFR-Part 121:56649e56661. 16. Egan TM, Kotloff RM. Pro/con debate: lung allocation should be based on medical urgency and transplant survival and not on waiting time. Chest 2005;128:407e415. 17. Kamath PS, Wiesner RH, Malinchoc M, et al. A model to predict survival in patients with end-stage liver disease. Hepatology 2001;33:464e470. 18. Meier-Kriesche HU, Kaplan B. Waiting time on dialysis as the strongest modifiable risk factor for renal transplant outcomes: a paired donor kidney analysis. Transplantation 2002;74:1377e1381. 19. Henry J Kaiser Family Foundation. Health coverage and uninsured. State health facts. Available at: http://kff.org/statecategory/health-coverage-uninsured/. Accessed October 28, 2014. 20. Bharat Thakkar KH, Henderson WG, Oprian C. A method to generate Kaplan-Meier and adjusted survival curves using SAS. Available at: http://www2.sas.com/proceedings/sugi23/Stats/ p226.pdf. Accessed April 4, 2016. 21. Henry J Kaiser Family Foundation. Status of state action on the Medicaid expansion decision. State health facts. Available at: http://kff.org/health-reform/state-indicator/state-activityaround-expanding-medicaid-under-the-affordable-care-act/. Accessed October 28, 2014. 22. US Department of Health and Human Services. Objective E: ensure access to quality, culturally competent care, including long-term services and supports, for vulnerable populations. Available at: http://www.hhs.gov/about/strategic-plan/strategicgoal-1/index.html#obj_e. Accessed December 17, 2013. 23. Singh GK, Hoyert DL. Social epidemiology of chronic liver disease and cirrhosis mortality in the United States, 1935e1997: trends and differentials by ethnicity, socioeconomic status, and alcohol consumption. Hum Biol 2000;72:801e820. 24. Foraker RE, Rose KM, Suchindran CM, et al. Socioeconomic status, Medicaid coverage, clinical comorbidity, and rehospitalization or death after an incident heart failure hospitalization: Atherosclerosis Risk in Communities cohort (1987 to 2004). Circ Heart Fail 2011;4:308e316. 25. Volkova N, McClellan W, Klein M, et al. Neighborhood poverty and racial differences in ESRD incidence. J Am Soc Nephrol 2008;19:356e364. 26. Wolfe RA, McCullough KP, Schaubel DE, et al. Calculating life years from transplant (LYFT): methods for kidney and kidneypancreas candidates. Am J Transplant 2008;8[Pt 2]:997e1011. 27. Mulligan D. Liver Committee Chair on status of liver distribution discussion. UNOS Transplant Pro. Available at: http:// transplantpro.org/liver-committee-chair-status-liver-distributiondiscussion/. Accessed July 9, 2015.
DuBay et al
Vol. 223, No. 4, October 2016
Medicaid Expansion and Transplantation Outcomes
620.e1
eTable 1. Baseline Characteristics of Heart Transplant Recipients (2002 to 2011) Stratified by Public Insurance, Medicare, or Medicaid Characteristic
Sample size, n Age, y, mean (SD) Female sex, n (%) Race/ethnicity, n (%) Caucasian African American Asian Hispanic Other BMI, kg/m2, mean (SD)* Education, n (%)y Less than high school High school Some college 4 years college Diabetes, n (%)* Hypertension, n (%)* Albumin, g/dL, mean (SD)y Creatinine, mg/dL, mean (SD)*
Total
Private
Medicare
Medicaid
p Value
17,239 52.0 (12.5) 4,233 (24.6)
9,925 51.3 (11.8) 2,406 (24.2)
5,157 57.3 (10.8) 1,113 (21.6)
2,157 43.2 (13.8) 714 (33.1)
<0.001 <0.001
12,406 3,033 411 1,242 146 26.9
(71.8) (17.6) (2.4) (7.2) (0.9) (8.4)
7,659 1,358 274 561 73 26.7
(77.2) (13.6) (2.8) (5.7) (0.7) (10.0)
3,653 1,008 83 367 45 27.2
(70.9) (19.5) (1.6) (7.1) (0.9) (4.8)
1,094 667 54 314 28 26.8
(50.7) (30.9) (2.5) (14.6) (1.3) (6.4)
<0.001
530 6,113 3,580 3,601 4,266 6,998 3.7 1.30
(3.8) (44.2) (25.9) (26.1) (24.8) (44.5) (0.7) (0.7)
197 3,065 2,127 2,534 2,280 3,823 3.7 1.30
(2.5) (36.7) (26.8) (32.0) (23.0) (42.4) (0.7) (0.6)
199 2,050 1,021 909 1,535 2,352 3.7 1.36
(4.8) (49.1) (24.3) (21.8) (29.8) (50.2) (0.7) (0.8)
134 998 432 158 451 823 3.6 1.20
(7.8) (58.0) (25.0) (9.2) (20.9) (40.8) (0.7) (0.6)
<0.001
<0.001
<0.001 <0.001 <0.001 <0.001
*Missing for <10%. y Missing for 10% to 20%.
eTable 2. Baseline Characteristics of Lung Transplant Recipients (2002 to 2011) Stratified by Public Insurance, Medicare, or Medicaid Characteristic
Sample size, n Age, y, mean (SD) Female sex, n (%) Race/ethnicity, n (%) Caucasian African American Asian Hispanic Other BMI, kg/m2, mean (SD)* Education, n (%)y Less than high school High school Some college 4 y college Diabetes, n (%)* Hypertension, n (%)y Albumin, g/dL, mean (SD) y Creatinine, mg/dL, mean (SD)* *Missing for <10%. y Missing for 10% to 20%.
Total
Private
Medicare
Medicaid
p Value
13,206 53.3 (13.1) 5,771 (43.7)
7,964 (60.3) 52.1 (12.4) 3,446 (43.3)
4,271 (32.3) 57.8 (11.9) 1,805 (42.3)
971 (7.4) 43.0 (15.6) 520 (53.6)
<0.001 <0.001
11,269 1,061 161 630 85 24.7
(85.3) (8.0) (1.2) (4.8) (0.6) (4.9)
6,853 600 115 349 47 24.8
(86.1) (7.5) (1.4) (4.4) (0.6) (5.1)
3,691 342 36 172 30 25.0
(86.4) (8.0) (0.8) (4.0) (0.6) (4.5)
725 119 10 109 8 23.3
(74.7) (12.3) (1.0) (11.2) (0.8) (5.0)
354 4,903 3,001 3,266 2,318 2,930 3.9 0.90
(3.1) (42.6) (26.0) (28.3) (17.6) (24.8) (0.6) (0.5)
144 2,591 1,876 2,318 1,354 1,655 3.9 0.90
(2.1) (37.3) (27.1) (33.5) (17.0) (23.3) (0.6) (0.5)
144 1,816 932 831 749 1,101 4.0 0.90
(3.9) (48.8) (25.0) (22.3) (17.6) (28.7) (0.6) (0.4)
66 496 193 117 215 174 3.8 0.88
(7.6) (56.9) (22.1) (13.4) (22.1) (19.8) (0.7) (0.8)
<0.001
<0.001 <0.001
<0.001 <0.001 <0.001 <0.001
620.e2
DuBay et al
J Am Coll Surg
Medicaid Expansion and Transplantation Outcomes
eTable 3. Baseline Characteristics of Liver Transplant Recipients (2002 to 2011) Stratified by Public Insurance, Medicare, or Medicaid Characteristic
Sample size, n Age, y, mean (SD) Female sex, n (%) Race/ethnicity, n (%) Caucasian African American Asian Hispanic Other BMI, kg/m2, mean (SD)* Education, n (%)y Less than high school High school Some college 4 years college Diabetes, n (%)* Hypertension, n (%)* Albumin, g/dL, mean (SD)z Creatinine, mg/dL, mean (SD)z
Total
Private
Medicare
Medicaid
48,201 53.1 (10.2) 15,922 (33.0)
30,941 52.4 (9.5) 9,524 (30.8)
10,158 58.1 (9.8) 3,698 (36.4)
7,102 48.9 (10.7) 2,700 (38.0)
35,026 4,451 2,190 6,043 491 28.2
(72.7) (9.2) (4.5) (12.5) (1.1) (6.2)
23,601 2,716 1,405 2,951 268 28.2
(76.3) (8.8) (4.5) (9.5) (0.9) (6.4)
7,277 847 394 1,535 105 28.4
(71.6) (8.3) (3.9) (15.1) (1.1) (5.8)
4,148 888 391 1,557 118 28.0
(58.4) (12.5) (5.5) (21.9) (1.7) (6.1)
2,030 18,315 9,200 8,651 11,706 8,853 3.0 0.90
(5.3) (48.0) (24.0) (22.7) (24.3) (21.5) (0.7) (0.5)
697 10,715 6,399 6,726 7,051 5,452 3.0 0.90
(2.8) (43.7) (26.1) (27.4) (22.8) (20.6) (0.7) (0.5)
642 4,209 1,757 1,443 3,073 2,219 3.0 0.90
(8.0) (52.3) (21.8) (17.9) (30.3) (25.4) (0.6) (0.4)
691 3,391 1,044 482 1,582 1,182 2.9 0.88
(12.3) (60.5) (18.6) (8.6) (22.3) (19.5) (0.7) (0.8)
p Value
<0.001 <0.001 <0.001
<0.001 <0.001
<0.001 <0.001 <0.001 <0.001
*Missing for <10%. y Missing for 10% to 20%. z Missing for >20%.
eTable 4. Baseline Characteristics of Renal Transplant Recipients (2002 to 2011) Stratified by Public Insurance, Medicare or Medicaid Characteristic
Sample size, n Age, y, mean (SD) Female sex, n (%) Race/ethnicity, n (%) Caucasian African American Asian Hispanic Other BMI, kg/m2, mean (SD) Education, n (%)y Less than high school High school Some college 4 y college Diabetes, n (%)* Hypertension, n (%)* Albumin, g/dL, mean (SD)z Creatinine, mg/dL, mean (SD)* *Missing for <10%. y Missing for 10% to 20%. z Missing for >20%.
Total
Private
Medicare
Medicaid
90,548 51.7 (13.2) 36,052 (39.8)
25,970 51.5 (11.8) 10,441 (40.2)
60,217 52.2 (13.5) 23,563 (39.1)
4,361 45.4 (14.1) 2,048 (47.0)
42,380 28,821 5,179 12,534 1,633 27.8
(46.8) (31.8) (5.7) (13.8) (1.9) (6.4)
15,190 6,289 1,635 2,506 350 27.8
(58.5) (24.2) (6.3) (9.7) (1.3) (6.5)
26,068 21,063 3,117 8,796 1,173 27.8
(43.3) (35.0) (5.2) (14.5) (2.0) (6.4)
1,122 1,469 427 1,232 110 26.6
(25.7) (33.7) (9.8) (28.3) (2.5) (5.7)
5,867 36,895 18,278 15,460 30,821 70,856 3.9 8.3
(7.7) (48.2) (23.9) (20.2) (34.1) (85.5) (0.6) (3.4)
801 8,921 5,755 6,328 8,422 19,880 3.9 7.9
(3.7) (40.9) (26.4) (29.0) (32.5) (84.5) (0.6) (3.4)
4,398 26,009 12,004 8,850 21,127 47,670 3.9 8.5
(8.6) (50.7) (23.4) (17.3) (35.2) (86.2) (0.6) (3.4)
668 1,965 519 282 1,272 3,306 3.9 8.5
(19.5) (57.2) (15.1) (8.2) (29.3) (82.2) (0.6) (3.4)
p Value
<0.001 <0.001 <0.001
<0.001
<0.001 <0.001 <0.001 <0.001
Vol. 223, No. 4, October 2016
DuBay et al
Medicaid Expansion and Transplantation Outcomes
620.e3
eTable 5. State-Level Population and Medicaid Data from 2010 to 2011 Used to Create Figure 2A and B
State
Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware District of Columbia Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma Oregon Pennsylvania Rhode Island South Carolina South Dakota Tennessee Texas Utah Vermont
Adult Medicaid population, n
Expected Adult population, n
Medicaid proportion, n
30,2000 42,100 609,900 179,000 3,557,200 316,000 291,600 82,600 99,700 1,137,700 446,400 116,600 73,600 854,900 470,400 228,700 128,100 339,800 348,800 196,600 323,500 911,100 817,900 393,100 249,800 380,500 445,00 80,000 114,700 40,400 502,000 170,600 2,624,800 611,200 24,200 810,200 213,600 286,000 970,000 90,800 356,500 42,600 528,100 1,224,200 119,400 94,800
2,992,600 374,700 3,894,800 1,692,300 21,581,500 3,072,700 2,383,600 604,100 464,200 11,169,700 5,256,000 891,000 901,100 7,798,700 3,924,300 1,971,200 1,684,600 2,625,500 2,540,700 912,000 3,719,600 4,811,600 6,252,900 3,543,000 1,711,400 3,725,400 583,900 1,144,600 1,489,300 859,000 5,394,700 1,154,900 12,747,300 5,575,200 446,900 7,211,500 2,141,400 2,424,300 8,556,600 678,000 2,811,000 496,700 3,913,900 1,313,1600 1,546,700 439,100
0.101 0.112 0.157 0.106 0.165 0.103 0.122 0.137 0.215 0.102 0.085 0.131 0.082 0.110 0.120 0.116 0.076 0.129 0.137 0.216 0.087 0.189 0.131 0.111 0.146 0.102 0.076 0.070 0.077 0.047 0.093 0.148 0.206 0.110 0.054 0.112 0.100 0.118 0.113 0.134 0.127 0.086 0.135 0.093 0.077 0.216
Observed adult transplant recipients Medicaid All Medicaid recipients, n recipients, n proportion, n
14 4 72 18 620 35 31 4 15 233 53 6 1 180 47 26 24 51 63 9 56 75 69 51 26 75 3 22 7 4 60 14 272 57 4 100 22 21 148 0 31 6 52 104 16 7
525 37 707 305 3,959 483 407 145 72 2,604 1,054 117 94 1,540 719 368 343 494 599 93 933 747 1,215 484 400 810 65 227 213 102 1,196 220 2,299 1,367 62 1,182 383 271 1,810 74 623 85 789 2,820 242 78
0.027 0.108 0.102 0.059 0.157 0.072 0.076 0.028 0.208 0.089 0.050 0.051 0.011 0.117 0.065 0.071 0.070 0.103 0.105 0.097 0.060 0.100 0.057 0.105 0.065 0.093 0.046 0.097 0.033 0.039 0.050 0.064 0.118 0.042 0.065 0.085 0.057 0.077 0.082 0 0.050 0.071 0.066 0.037 0.066 0.090
Observed to expected ratio
0.26 0.96 0.65 0.56 0.95 0.70 0.62 0.20 0.97 0.88 0.59 0.39 0.13 1.07 0.55 0.61 0.92 0.80 0.77 0.45 0.69 0.53 0.43 0.95 0.45 0.91 0.61 1.39 0.43 0.83 0.54 0.43 0.57 0.38 1.19 0.75 0.58 0.66 0.72 0 0.39 0.82 0.49 0.40 0.86 0.42 (Continued)
620.e4
DuBay et al
J Am Coll Surg
Medicaid Expansion and Transplantation Outcomes
eTable 5. Continued
State
Virginia Washington West Virginia Wisconsin Wyoming O/E, observed to expected.
Adult Medicaid population, n
360,500 450,100 158,000 484,400 30,400
Expected Adult population, n
Medicaid proportion, n
4,831,600 4,104,800 1,168,400 3,741,900 337,200
0.075 0.110 0.135 0.129 0.090
Observed adult transplant recipients Medicaid All Medicaid recipients, n recipients, n proportion, n
49 43 18 51 3
952 663 217 663 43
0.051 0.065 0.083 0.077 0.070
Observed to expected ratio
0.69 0.59 0.61 0.59 0.77
PhD,
The number of Medicaid beneficiaries has increased under the Affordable Care Act, improving access to solid organ transplantation in this disadvantaged patient cohort. It is unclear what impact Medicaid expansion will have on transplantation outcomes. We performed a retrospective cohort analysis to measure the frequency and variation in Medicaid transplantation and post-transplantation survival in Medicaid patients. STUDY DESIGN: Adult heart, lung, liver, and renal transplant recipients between 2002 and 2011 (n ¼ 169,194) reported to the Scientific Registry of Transplant Recipients were identified. Transplant recipients were classified based on insurance status (private, Medicare or Medicaid). Outcomes measures included 5-year post-transplantation survival, summarized using Kaplan-Meier curves and compared with log-rank tests. Organ-specific Cox proportional hazards models were used to adjust for donor and recipient factors. RESULTS: Medicaid patients comprised 8.6% of all organ transplant recipients. Fewer transplantations were performed than expected among Medicaid beneficiaries for all organs except liver (liver: observed to expected ratio ¼ 1.21; 95% CI, 0.68e1.90; heart: observed to expected ratio ¼ 0.89; 95% CI, 0.44e1.49; lung: observed to expected ratio ¼ 0.57; 95% CI, 0.22e1.06; renal: observed to expected ratio ¼ 0.32; 95% CI, 0.08e0.72). Medicaid transplant recipients were listed with more severe organ failure and experienced shorter transplant wait times. Post-transplantation survival was lower in Medicaid patients compared with private insurance for all organs. Post-transplantation survival in Medicaid patients was similar to Medicare patients for heart, liver, and renal but lower in lung. CONCLUSIONS: Medicaid organ transplant beneficiaries had significantly lower survival compared with privately insured beneficiaries. The more severe organ failure among Medicaid beneficiaries at the time of listing, suggested a pattern of late referral, which might account for worse outcomes. Implementation of the Affordable Care Act gives the opportunity to develop the necessary infrastructure to ensure timely transplantation referrals and improve long-term outcomes in this vulnerable population. (J Am Coll Surg 2016;223:611e620. 2016 by the American College of Surgeons. Published by Elsevier Inc. All rights reserved.) BACKGROUND:
Disclosure Information: Nothing to disclose. Support: This research was funded by National Institutes of Health grant numbers 1 K23 DK091514 (DD), 1 R03 DK106432 (DD), and 1 K23 DK 103918 (JL). Disclaimer: The interpretation and reporting of these data are the responsibility of the author(s) and in no way should be seen as an official policy of or interpretation by Scientific Registry of Transplant Recipients, Organ Procurement and Transplantation Network, or the US Government.
From the Department of Surgery-Transplantation (DuBay, MacLennan, Reed, Shelton, Gray, White, Eckhoff, Locke), School of Public HealthBiostatistics (Redden), Department of Medicine-Preventive Medicine (Fouad), University of Alabama at Birmingham, Birmingham, AL, and Department of Medicine-Preventive Medicine, University of Tennessee, Memphis, TN (Martin). Correspondence address: Derek A DuBay, MD, MSPH, FACS, Department of Surgery-Transplantation, University of Alabama at Birmingham, ZRB 701, 1720 2nd Ave S, Birmingham, AL 35294-0007. email: [email protected]
Received February 3, 2016; Revised July 11, 2016; Accepted July 12, 2016.
ª 2016 by the American College of Surgeons. Published by Elsevier Inc. All rights reserved.
611
http://dx.doi.org/10.1016/j.jamcollsurg.2016.07.004 ISSN 1072-7515/16
DuBay et al
612
Medicaid Expansion and Transplantation Outcomes
Abbreviations and Acronyms
ACA HR MELD O/E SRTR
¼ ¼ ¼ ¼ ¼
Affordable Care Act hazard ratio Model for End-Stage Liver Disease observed to expected ratio Scientific Registry of Transplant Recipients
The implementation of the Affordable Care Act (ACA) has increased the number of Medicaid beneficiaries in the United States substantially. Current estimates suggest that 50% of uninsured Americans have or will gain coverage via Medicaid expansion.1-4 For the 31 states adopting Medicaid expansion, financial eligibility requirements decreased to 138% of the federal poverty level, resulting in 12.3 million new Medicaid beneficiaries.1-4 Overall uninsured rates have decreased from 15.7% before ACA to 9.2% currently.1-4 Access to solid organ transplantation has increased for this previously disadvantaged population. However, it is unclear what impact, if any, implementation of the ACA Medicaid expansion program will have on transplantation outcomes. Non-private insurance status has been demonstrated to be an adverse marker of transplantation outcomes across solid organ transplantation.5-13 With introduction of the ACA, there will be more Medicaid beneficiaries who can access organ transplantation and, given earlier reports of inferior outcomes among this population,5,7-9 transplantation centers might encounter observed survival rates considerably lower than expected. In the context of the current regulatory environment, understanding the impact of the implementation of the ACA on transplantation outcomes is of paramount importance. To this end, we performed a retrospective cohort analysis of solid organ transplant recipients. Our approach was to examine recent historic transplantation outcomes data in Medicaid, Medicare, and private insured recipients. Outcomes measures include frequency of Medicaid transplantation across organ groups; variation in Medicaid transplantation in the United States; and posttransplantation survival among Medicaid beneficiaries compared with patients with private insurance or Medicare.
METHODS Data source This study used data from the Scientific Registry of Transplant Recipients (SRTR). The SRTR system includes data submitted by the members of the Organ Procurement and Transplantation Network on all donors, waitlisted candidates, and transplant recipients in the
J Am Coll Surg
United States. The Health Resources and Services Administration of the US Department of Health and Human Services provides the oversight to the activities of the Organ Procurement and Transplantation Network and SRTR contractors. The University of Alabama at Birmingham IRB reviewed and approved the study (#E121024003). Study population Deceased donor, adult heart, lung, liver, and renal transplant recipients reported to the SRTR between January 1, 2002 and December 31, 2011 were identified (heart, n ¼ 17,239; lung, n ¼ 13,206; liver, n ¼ 48,201; renal, n ¼ 90,548). This time interval pre-dates implementation of the ACA (Table 1). Transplant recipients were classified by insurance type as private insurance, Medicare, or Medicaid, based on SRTR classifications. Insurance status was assigned based on the patient’s primary insurance at the time of transplantation. Outcomes ascertainment Post-transplantation survival was defined as time from transplantation to death or last follow-up, and was censored at 5 years post-transplantation. Death dates were supplemented by information from the Centers for Medicare and Medicaid Services and the Limited Access Death Master File available from the National Technical Information Service through 2013. Other outcomes included measures of access to transplantation as a function of insurance status, including observed to expected transplantation ratios (O/E), waitlist time, and measures of end-stage organ disease at the time of transplantation (heart status,14 lung allocation score,15,16 Model for End-Stage Liver Disease [MELD],17 and cumulative years of dialysis).18 Expected insurance distribution for the O/E ratios was calculated using health coverage status provided by the Kaiser Family Foundation State Health Facts from 2011 at both the national and state levels,19 and was calculated as the proportion of Medicaid patients receiving transplants in each organ group divided by the proportion of the population insured with Medicaid between 2010 and 2011. Other patient covariates, measured at the time of transplantation, included recipient age, sex, race/ethnicity, BMI, education, select comorbidities (diabetes and hypertension), albumin, and renal function (creatinine). Additional outcomes measures included post-transplantation length of stay. Statistical analysis Baseline characteristics among private, Medicare, and Medicaid solid organ recipients (ie heart, lung, liver,
Vol. 223, No. 4, October 2016
DuBay et al
Medicaid Expansion and Transplantation Outcomes
613
Table 1. Baseline Characteristics of Heart, Lung, Liver, and Renal Transplant Recipients (2002 to 2011) Stratified by Public Insurance, Medicare, or Medicaid Characteristic
Sample size, n Organ type, n (%) Liver Renal Heart Lung Age, y, mean (SD) Female sex, n (%) Race/ethnicity, n (%) Caucasian African American Asian Hispanic Other BMI, kg/m2, mean (SD)* Education,y n (%) Less than high school High school Some college 4 y college Diabetes,* n (%) Hypertension,y n (%) Albumin, g/dL, mean (SD)z Creatinine, mg/dL, mean (SD), excluding renal recipientsz
Total
Private
Medicare
Medicaid
169,194
74,800
79,803
14,591
48,201 90,548 17,239 13,206 52.2 61,978
(28.5) (53.5) (10.2) (7.8) (12.3) (36.6)
30,941 25,970 9,925 7,964 51.9 25,817
(41.4) (34.7) (13.3) (10.6) (11.0) (34.5)
10,158 60,217 5,157 4,271 53.6 30,179
(12.7) (75.5) (6.5) (5.4) (13.1) (37.8)
7,102 4,361 2,157 971 46.6 5,982
(48.7) (29.9) (14.8) (6.7) (12.9) (41.0)
101,081 37,366 7,941 20,449 2,355 27.6
(59.7) (22.1) (4.7) (12.1) (1.4) (6.6)
53,303 10,963 3,429 6,367 738 27.5
(71.3) (14.6) (4.6) (8.5) (1.0) (7.0)
40,689 23,260 3,630 10,870 1,353 27.7
(51.0) (29.1) (4.6) (13.6) (1.7) (6.2)
7,089 3,143 882 3,212 264 27.0
(48.6) (21.5) (6.1) (22.0) (1.8) (6.1)
8,781 66,226 34,059 30,978 49,111 89,637 3.8 1.12
(6.3) (47.3) (24.3) (22.1) (29.1) (59.1) (0.7) (0.6)
1,839 25,292 16,157 17,906 19,107 30,810 3.7 1.12
(3.0) (41.3) (26.4) (29.3) (25.6) (46.6) (0.7) (0.6)
5,383 34,084 15,174 12,033 26,484 53,342 3.9 1.15
(8.0) (50.7) (23.4) (17.9) (33.2) (73.5) (0.6) (0.7)
1,559 6,850 2,188 1,039 3,520 5,485 3.6 1.10
(13.4) (58.9) (18.8) (8.9) (24.2) (42.2) (0.8) (0.7)
p Value
<0.001
<0.001 <0.001 <0.001
<0.001 <0.001
<0.001 <0.001 <0.001 <0.001
*Missing for <10%. y Missing for 10% to 20%. z Missing for >20%.
and renal) were compared. Continuous variables were analyzed using ANOVA or Kruskal-Wallis tests, and categorical variables using chi-square tests. The O/E ratios, stratified by organ-specific transplantation, were calculated for private, Medicare, and Medicaid insurance transplant recipients. Number of Medicaid transplants and O/ E ratios were compared for all 50 States and the District of Columbia. Survival analysis Patient survival was summarized using Kaplan-Meier survival curves stratified by insurance group and compared using log-rank tests. Organ-specific Cox proportional hazards models were built using a backwards step-wise selection, with p < 0.05 on bivariate analysis and were used to adjust for donor and recipient factors as described here, with the most parsimonious model chosen through minimization of Akaike’s Information Criteria. Although education is reported, age, race, and sex were the only sociodemographic factors considered in adjusted analyses. Recipient age, transplantation year (2002 to 2011), BMI,
albumin, and creatinine were treated as continuous covariates. Sex and race (coded as white, black, Hispanic, and other) were categorical variables. Heart status was a categorical variable; MELD and years of dialysis were continuous variables. Adjusted survival curves were generated from the fitted Cox models using the direct adjusted survivor function,20 and proportional hazards assumptions were tested in the fitted Cox models with Schoenfeld residuals and time-dependent covariates. Spline regression was used to generate time-dependent hazard ratios (HRs) when the assumptions were violated for insurance status. Patients missing data for variables included in the models were excluded from adjusted analyses. Sensitivity analyses Sensitivity analyses were performed for pre-emptive renal transplantations (transplantation before initiation of dialysis). This analysis was done to account for renal recipients who had no dialysis-associated morbidity. Additional models were run with the variables with significant missingness excluded, and inferences were
614
DuBay et al
Medicaid Expansion and Transplantation Outcomes
consistent with those reported in this article. Correlation between Medicaid transplantation center volume and 1year post-transplantation survival was determined for all organ types using Spearman’s correlation coefficient. All of the tests were 2-sided with statistical significance set at a ¼ 0.05. Analyses were performed using SAS software, version 9.3 (SAS Institute). Additional supporting information can be found in the online version of this article.
RESULTS Study population Only 8.6% of solid organ transplantations were performed for patients with Medicaid insurance (Table 1). Medicaid transplant recipients were younger and more often female. Medicaid patients were more frequently non-Caucasian than Medicare and private. Medicaid transplant recipients were less likely to have a college or beyond education than private or Medicare recipients. The highest frequency of Medicaid transplants was performed for liver (14.7%) and heart (12.5%) with the lowest frequency performed
J Am Coll Surg
for lung (7.4%) and renal (4.8%) (Fig. 1A and eTables 1 to 4, available online). Access to transplantation Fewer transplantations were performed for Medicaid patients than was expected for all organ types except liver (Fig. 1B). The O/E ratios for transplantation in Medicaid patients were liver: 1.21 (95% CI, 0.68e1.90); heart: 0.89 (95% CI, 0.44e1.49); lung: 0.57 (95% CI, 0.22e1.06); and renal: 0.32 (95% CI, 0.08e0.72). The number of transplantations performed in Medicaid patients, by state of residence, for all organ types combined, varied from 0 to 304 (Fig. 2A and eTable 5, available online). The O/E ratio for all organ transplantation in patients insured with Medicaid ranged from 0 to 1.4, but was only 1.0 in 3 states (Fig. 2B and eTable 5, available online). Heart transplantation outcomes Medicaid patients receiving a heart were more likely to be transplanted as a status 1A (46.9%), the highest-priority
Figure 1. (A) Percentage of heart, lung, liver, and renal transplantation stratified by organ type. The percentages correspond to the number of transplantations performed in Medicaid patients divided by the total number of transplantations performed for each organ as seen in Table 1 (ie liver transplant: 7,102 / 48,201 100%). (B) Observed to expected ratios for heart, lung, liver, and renal transplantation stratified by insurance group.
Vol. 223, No. 4, October 2016
DuBay et al
Medicaid Expansion and Transplantation Outcomes
615
Figure 2. (A) Number of solid organ transplantations performed in Medicaid patients for each state. (B) Observed to expected ratios for Medicaid transplantation for each state.
listing, compared with private (44.3%) or Medicare (40.6%) patients (p < 0.001; Table 2). Medicaid heart transplant recipients also had a significantly shorter wait time (66 days) compared with private (75 days) and Medicare (91 days) patients (p < 0.001). Post-heart transplantation adjusted survival was lower in Medicaid patients compared with private, but similar to Medicare (Fig. 3A). Among Medicaid recipients, heart transplantation survival was 0.8% lower than Medicare and 4.7% lower than private at 3 years and 0.9% lower than Medicare and 5.9% lower than private at 5 years. Compared with
private recipients, the adjusted Medicaid mortality HR for heart transplantation was 1.10 in the first 2 years post-transplantation (95% CI, 0.97e1.26), increasing to 1.77 after 2 years (95% CI, 1.49e2.09; Table 3). Similar survival HRs were observed with sensitivity models that included recipient center as a covariate to control for geography. There was no significant correlation between heart transplantation survival and center Medicaid volume (r ¼ 0.137, p ¼ 0.12). Compared with non-Medicaid patients, the median length of stay was 2 days longer for heart transplant recipients with Medicaid insurance (Table 2).
616
DuBay et al
J Am Coll Surg
Medicaid Expansion and Transplantation Outcomes
Table 2. Waiting Time, Organ-Specific Disease Score, Post-Transplantation Length of Stay, and Post-Transplantation Survival Stratified by Organ Type and Insurance Group Variable
Heart, n Waiting time, d, median (IQR) Status at Tx, n (%) Status 1A Status 1B Status 2 Postoperative LOS, d, median (IQR) Follow-up time, y, median (IQR) 5-y survival, %* Lung, n Waiting time, d, median (IQR) Lung allocation score at Tx,y median (IQR) Postoperative LOS, d, median (IQR) Follow-up time, y, median (IQR) 5-y survival, %* Liver, n Waiting time, d, median (IQR) Model for End-Stage Liver Disease at Tx, median (IQR) Postoperative LOS, d, median (IQR) Follow-up time, y, median (IQR) 5-y survival, %* Renal, n Waiting time, d, median (IQR) Dialysis time, y, median (IQR) Postoperative LOS, d, median (IQR) Follow-up time, y, median (IQR) 5-y survival, %*
Total
Private
Medicare
Medicaid
17,239 78 (23e223)
9,925 75 (23e209)
5,157 91 (27e271)
2,157 66 (20e181)
7,495 6,778 2,960 14 3.9
(43.5) (39.3) (17.2) (10e21) (1.8e5.0) 74.3 13,206 107 (29e329.5)
4,394 3,789 1,741 13 4.0
(44.3) (38.2) (17.5) (10e20) (1.9e5.0) 77.2 7,964 108 (30e325.5)
2,090 2,134 930 14 3.1
(40.6) (41.4) (18.0) (10e22) (1.1e5.0) 71.1 4,271 106 (28e342.5)
1,011 855 289 15 3.8
(46.9) (39.7) (13.4) (10e22) (1.8e5.0) 67.9 971 105.5 (31e331)
p Value
<0.001 <0.001
<0.001 <0.001 <0.001 0.8
39.1 (34e48) 15 (10e26) 2.7 (1.0e5.0) 52.0 48,201 72 (15e256)
39.6 (35e49) 15 (10e25) 2.9 (1.1e5.0) 54.9 30,941 71 (14e248)
38.1 (34e47) 15 (10e26) 2.1 (1.0e4.3) 47.6 10,158 90 (20e331)
39.1 (34e49) 17 (12e28) 2.3 (1.0e4.6) 45.0 7,102 51 (10e203)
<0.001 <0.001 <0.001 <0.001
19 (14e27) 10 (7e17) 3.1 (1.1e5.0) 70.1 90,548 704 (302e1222) 3.2 (1.5e5.2) 6 (5e9) 3.7 (1.9e5.0) 80.2
19 (13e27) 10 (7e16) 3.5 (1.3e5.0) 72.3 25,970 527 (223e956) 1.8 (0.5e3.1) 5 (4e8) 4.0 (2.0e5.0) 84.6
18 (13e25) 10 (7e18) 3.0 (1.0e5.0) 64.8 60,217 778 (351e1314) 3.8 (2.2e5.7) 6 (5e9) 3.4 (1.8e5.0) 78.0
21 (15e30) 11 (8e19) 3.0 (1.1e5.0) 67.6 4,361 799 (326e1441) 3.9 (1.9e6.4) 6 (5e9) 3.8 (1.9e5.0) 82.9
<0.001 <0.001 <0.001 <0.001
<0.001
<0.001 <0.001 <0.001 <0.001 <0.001
*Estimated 5-year survival (unadjusted) calculated via Kaplan-Meier methods and compared using log-rank test. y Lung allocation score missing for 50% before implementation in 2006. IQR, interquartile range; LOS, length of stay; Tx, transplant.
Lung transplantation outcomes The long allocation score, a priority lung listing score, for Medicaid recipients (39.1) was less than private (39.6), but more than Medicare (38.1) patients at the time of transplantation (Table 2). There was no difference in waiting time between insurance cohorts for lung transplantations (p ¼ 0.8). Post-lung transplantation adjusted survival in Medicaid recipients was significantly worse compared with Medicare and private insurance recipients (Fig. 3B). Among Medicaid recipients, post-lung transplantation survival was 3.3% lower than Medicare and 6.7% lower than private at 3 years and 3.8% lower than Medicare and 8.0% lower than private at 5 years. Compared with private, the adjusted Medicaid mortality HR for lung transplantation was 1.15 within the first 2 years post-transplantation (95% CI, 1.00e1.33), then
increased to 1.44 after 2 years (95% CI, 1.20e1.72; Table 3). Similar survival HRs were observed with sensitivity models that included recipient center as a covariate to control for geography. There was no significant correlation between lung transplantation survival and center Medicaid volume (r ¼ 0.058; p ¼ 0.64). Compared with non-Medicaid patients, median length of stay was 2 days longer for lung transplant recipients with Medicaid insurance (Table 2). Liver transplantation outcomes The lab-MELD score, a priority liver wait listing score, was highest in Medicaid21 compared with private19 and Medicare18 patients at the time of transplantation (Table 2). Medicaid liver transplant recipients also had shorter wait times (51 days) compared with private (71 days) and
Vol. 223, No. 4, October 2016
DuBay et al
Medicaid Expansion and Transplantation Outcomes
617
Figure 3. Direct adjusted survivor functions for (A) heart, (B) lung, (C) liver, and (D) renal transplant recipients stratified by insurance type.
Medicare (90 days) patients. Post-liver transplantation adjusted survival was lower in Medicaid patients compared with private but similar to Medicare (Fig. 3C). Compared with private recipients, liver transplantation survival was 2.8% lower in Medicaid at 1 year, 4.5% lower at 3 years, and 5.4% lower at 5 years. Compared with private, the adjusted Medicaid mortality HR for liver transplantation was 1.25 (95% CI, 1.18e1.32; Table 3). Similar survival HRs were observed with sensitivity models that included recipient center as a covariate to control for geography. There was no significant correlation between liver transplantation survival and center Medicaid volume (r ¼ 0.065; p ¼ 0.48). Compared with non-Medicaid patients, median length of stay was 1 day longer for liver transplant recipients with Medicaid insurance (Table 2). Renal transplantation outcomes Medicaid renal transplantation patients had the longest time spent on dialysis (3.9 years) compared with private
(1.8 years) and Medicare (3.8 years) patients (Table 2). Post-renal transplantation adjusted survival was lower in Medicaid patients compared with private, but similar to Medicare (Fig. 3D). Compared with private recipients, renal transplantation survival for Medicaid patients was 1.4% lower at 1 year, 3.0% lower at 3 years, and 4.9% lower at 5 years. Similar survival distributions were observed when excluding pre-emptive renal transplantations. Compared with privately insured recipients, the adjusted Medicaid mortality HR for renal transplantation was 1.37 (95% CI, 1.24e1.51; Table 3). Similar survival HRs were observed with sensitivity models that included recipient center as a covariate to control for geography. There was a statistically significant inverse correlation between renal transplantation survival and center Medicaid volume (r ¼ 0.346; p 0.0001). Median length of stay was not different in Medicaid and non-Medicaid renal transplant recipients (Table 2).
618
DuBay et al
Medicaid Expansion and Transplantation Outcomes
Table 3. Multivariable Analysis of Mortality Risk Factors after Transplantation Stratified by Organ Type Organ type
Heart* Within 2 y Medicare vs private Medicaid vs private After 2 yearsy Medicare vs private Medicaid vs private Lungz Within 2 y Medicare vs private Medicaid vs private After 2 yy Medicare vs private Medicaid vs private Liverx Medicare vs private Medicaid vs private Renaljj Medicare vs private Medicaid vs private
Adjusted hazard ratio
95% CI
p Value
1.18 1.10
1.07e1.30 0.97e1.26
<0.001 0.15
1.39 1.77
1.20e1.60 1.49e2.09
<0.001 <0.001
1.10 1.15
1.01e1.19 1.00e1.33
0.03 0.05
1.18 1.44
1.06e1.31 1.20e1.72
0.004 <0.001
1.24 1.25
1.18e1.30 1.18e1.32
<0.001 <0.001
1.36 1.37
1.30e1.42 1.24e1.51
<0.001 <0.001
*Adjusted for recipient age, race, sex, heart status, diabetes, hypertension, albumin, creatinine, and transplantation year. y Model splined at 2 years post-transplantation to account for violation of the proportional hazards assumption by the Medicare variable. Patients contributed time at risk for the first 2 years, at which point they were censored if they were still alive. Those patients who died within 2 years only contributed time at risk until death date and were not included in the post-2 year estimate of risk. The post-2 year estimate includes only those who survived for more than 2 years post-transplantation with time at risk beginning at 2 years and continuing until death or 5 years posttransplantation. z Adjusted for recipient age, race, sex, hypertension, albumin, creatinine, and transplantation year. x Adjusted for age, race, sex, diabetes, hypertension, Model for End-Stage Liver Disease, and transplantation year. jj Adjusted for age, race, sex, years on dialysis, diabetes, hypertension, creatinine, and transplantation year.
Transplantation outcomes in Medicaid expansion states Of all Medicaid transplantations performed, 69.8% were among states that have implemented Medicaid expansion as of February 2015, and 30.2% were among those not currently planning to expand. Among the 31 states (including the District of Columbia) that have implemented Medicaid expansion,21 the mean O/E ratio for Medicaid organ transplantations was similar compared with those not planning to expand (expansion states: 0.61 vs non-expansion states: 0.59; p ¼ 0.89). Among all organs, there was no difference in 5-year survival for Medicaid organ transplantations among expansion states
J Am Coll Surg
compared with non-expansion states (expansion states: 70.6% vs non-expansion states: 70.4%; p ¼ 0.91). There was no difference in 5-year survival for Medicaid heart transplantation (expansion states: 68.5%; non-expansion states: 67.8%; p ¼ 0.96), or Medicaid lung transplantation (expansion states: 45.6% vs non-expansion states: 42.6%; p ¼ 0.72). Five-year survival was significantly lower in Medicaid liver transplantation in expansion states (expansion states: 65.7% vs non-expansion states: 70.4%; p ¼ 0.005). There was no difference in 5-year survival in Medicaid renal transplantation in expansion states (expansion states: 83.5% vs non-expansion states: 82.1%; p ¼ 0.36).
DISCUSSION An objective of Medicaid expansion is to ensure access to quality, culturally competent care for vulnerable populations.22 The Medicaid program currently provides transplantation care to a diverse patient population. Despite providing transplantation care to a higher proportion of vulnerable patients, there remains a large disparity in the frequency of solid organ transplantation in Medicaid patients. The O/E transplantation ratio for Medicaid was <1.0 for all organ types except liver. In addition, only 3 of 50 states performed solid organ transplantations at a frequency equal to the proportion of population insured by Medicaid. This is an especially alarming statistic because higher rates of organ dysfunction and failure are observed in patients with lower socioeconomic status,23-25 the same patients who are over-represented in the Medicaid population. Development of an infrastructure to increase timely referral as a part of the Medicaid expansion program might improve outcomes. Currently, Medicaid transplant recipients were listed with more severe organ failure, experienced shorter transplantation wait times, and longer time on dialysis, hallmarks of a pattern of late referral for transplantation evaluation. It is likely that Medicaid patients in many states will be managed via some form of managed care organization in the future.21 Managed care organization can reduce barriers to transplantation and increase the likelihood of early referral via prompt diagnosis and an established network of transplantation providers. Earlier referral can ameliorate the advanced organ failure observed in the Medicaid population at the time of transplantation. A second challenge for the Medicaid Expansion Program is to improve post-transplantation survival outcomes. Compared with private insurance, post-transplantation survival in Medicaid patients dropped off soon after transplantation and then continued to slowly diverge for all organ types. Similar survival outcomes were observed for both
Vol. 223, No. 4, October 2016
DuBay et al
Medicaid and Medicare patients, despite the fact that these public insurance programs cover vastly different patient populations. Medicare covers the elderly and disabled, and Medicaid covers patients living at or near poverty. Worsened outcomes among Medicaid patients are especially disturbing, given the younger age of the Medicaid cohort when compared with Medicare patients. Although Medicaid patients had higher disease severity scores at the time of transplantation, post-transplantation survival was reduced significantly, even when controlling for this covariate. Studies demonstrate that adverse predictors of posttransplantation survival were overexpressed in the Medicaid population, especially patients with lower education and lower social support.7,8,10,24 We expected that there would be a volume to outcomes relationship, with high-volume Medicaid transplantation centers having a discriminating evaluation process and/or a more rigorous infrastructure for post-transplantation care. However, no statistical relationship between Medicaid transplantation volume and post-transplantation survival was found. There was also no statistically significant difference in post-transplantation survival outcomes between states that adopted Medicaid expansion and those that did not. Achieving good outcomes in this vulnerable population was a challenge for all centers. It might be advisable for transplantation programs to use Medicaid insurance status as a surrogate marker of poorer post-transplantation survival and implement specific post-transplantation care practices in an attempt to ameliorate decreased survival observed in this population. Policies that encourage more frequent long-term followup clinic appointments in Medicaid-insured patients might also reduce the disparity in post-transplantation survival outcomes. Improving long-term survival outcomes is especially important, given the current practice of using measures of utility in the lung allocation system16 and the consideration of adding utility to the renal26 and liver27 allocation systems. Younger Medicaid patients have high utility potential (ie should have more life-years gained with transplantation), but only if improved posttransplantation survival rates can be achieved. This study has several limitations inherent to retrospective analyses of large cohort studies, including limited data granularity and missingness. Not all patients were followed for 5 years for the survival analysis. In addition, these data did not allow for a direct measure of access to transplantation, as they lack referral information. A better measure of transplantation access would be O/E ratio of Medicaid patients referred for transplantation, but SRTR data lack this information. These analyses used past Medicaid outcomes to predict how the ACA will impact
Medicaid Expansion and Transplantation Outcomes
619
transplantation outcomes, although many of those gaining insurance via Medicaid expansion were without insurance previously. Although uninsured would be a better comparison group, this group has not had transplantation coverage and outcomes in this group are not known. In addition, the assumption of these analyses was that past Medicaid outcomes will be predictive of future outcomes, which might or might not be the case, given the change in the population insured by Medicaid expansion efforts. Finally, insurance type was assigned at the time of transplantation, and it is possible that patients might have changed their insurance type during the follow-up period of the study.
CONCLUSIONS A higher proportion of women and minorities with Medicaid insurance received a solid organ transplantation compared with private insurance or Medicare. However, with the exception of liver, Medicaid patients received transplants at a much lower proportion than would be expected for all other organ groups. The ACA affords the opportunity to develop and implement strategies that will increase overall referrals and the likelihood of early referral for transplantation evaluation. In addition, current ACA implementation practices can improve Medicaid practice infrastructure to reduce the significant disparity in post-transplantation survival between Medicaid and private insurance recipients. Author Contributions Study conception and design: DuBay, MacLennan, Reed, Redden, Fouad, Martin, Gray, White, Eckhoff, Locke Acquisition of data: DuBay, MacLennan, Reed, Locke Analysis and interpretation of data: DuBay, MacLennan, Reed, Shelton, Redden, Locke Drafting of manuscript: DuBay, MacLennan, Reed, Locke, Critical revision: DuBay, MacLennan, Reed, Shelton, Redden, Fouad, Martin, Gray, White, Eckhoff, Locke Acknowledgment: The University of Alabama at Birmingham Comprehensive Transplant Institute Outcomes Research Center provided methodological and statistical support. The data reported here have been supplied by the United Network for Organ Sharing as the contractor for the Organ Procurement and Transplantation Network and the Minneapolis Medical Research Foundation as the contractor for the SRTR. REFERENCES 1. Obama Care Facts. Dispelling the myths. Available at: http://obamacarefacts.com/obamacares-medicaid-expansion.php. Accessed December 17, 2013.
620
DuBay et al
Medicaid Expansion and Transplantation Outcomes
2. US Department of Health and Human Services. About the law. Health and Human Services: healthcare. Available at: http://www.hhs.gov/healthcare/rights/. Accessed August 18, 2014. 3. Henry J Kaiser Family Foundation. The coverage gap: uninsured poor adults in states that do not expand Medicaid. Health Reform. Available at: http://kff.org/health-reform/issue-brief/ the-coverage-gap-uninsured-poor-adults-in-states-that-do-notexpand-medicaid/. Accessed October 28, 2014. 4. Tavernise S, Gebeloff R. Millions of poor are left uncovered by health law. The New York Times October 2, 2013. 5. Allen JG, Arnaoutakis GJ, Orens JB, et al. Insurance status is an independent predictor of long-term survival after lung transplantation in the United States. J Heart Lung Transplant 2011;30:45e53. 6. Kemmer N, Zacharias V, Kaiser TE, Neff GW. Access to liver transplantation in the MELD era: role of ethnicity and insurance. Dig Dis Sci 2009;54:1794e1797. 7. Allen JG, Weiss ES, Arnaoutakis GJ, et al. Insurance and education predict long-term survival after orthotopic heart transplantation in the United States. J Heart Lung Transplant 2012;31:52e60. 8. Goldfarb-Rumyantzev AS, Koford JK, Baird BC, et al. Role of socioeconomic status in kidney transplant outcome. Clin J Am Soc Nephrol 2006;1:313e322. 9. Yoo HY, Thuluvath PJ. Outcome of liver transplantation in adult recipients: influence of neighborhood income, education, and insurance. Liver Transpl 2004;10:235e243. 10. Keith D, Ashby VB, Port FK, Leichtman AB. Insurance type and minority status associated with large disparities in prelisting dialysis among candidates for kidney transplantation. Clin J Am Soc Nephrol 2008;3:463e470. 11. Bryce CL, Angus DC, Arnold RM, et al. Sociodemographic differences in early access to liver transplantation services. Am J Transplant 2009;9:2092e2101. 12. Johansen KL, Zhang R, Huang Y, et al. Association of race and insurance type with delayed assessment for kidney transplantation among patients initiating dialysis in the United States. Clin J Am Soc Nephrol 2012;7:1490e1497. 13. Glueckert LN, Redden D, Thompson MA, et al. What liver transplant outcomes can be expected in the uninsured who become insured via the Affordable Care Act? Am J Transplant 2013;13:1533e1540. 14. Mehra MR, Kobashigawa J, Starling R, et al. Listing criteria for heart transplantation: International Society for Heart and Lung Transplantation guidelines for the care of cardiac transplant candidatesd2006. J Heart Lung Transplant 2006;25: 1024e1042.
J Am Coll Surg
15. Final Rule. Department of Health and Human Services. Organ Procurement and Transplantation Network. Fed Reg 42 CFR-Part 121:56649e56661. 16. Egan TM, Kotloff RM. Pro/con debate: lung allocation should be based on medical urgency and transplant survival and not on waiting time. Chest 2005;128:407e415. 17. Kamath PS, Wiesner RH, Malinchoc M, et al. A model to predict survival in patients with end-stage liver disease. Hepatology 2001;33:464e470. 18. Meier-Kriesche HU, Kaplan B. Waiting time on dialysis as the strongest modifiable risk factor for renal transplant outcomes: a paired donor kidney analysis. Transplantation 2002;74:1377e1381. 19. Henry J Kaiser Family Foundation. Health coverage and uninsured. State health facts. Available at: http://kff.org/statecategory/health-coverage-uninsured/. Accessed October 28, 2014. 20. Bharat Thakkar KH, Henderson WG, Oprian C. A method to generate Kaplan-Meier and adjusted survival curves using SAS. Available at: http://www2.sas.com/proceedings/sugi23/Stats/ p226.pdf. Accessed April 4, 2016. 21. Henry J Kaiser Family Foundation. Status of state action on the Medicaid expansion decision. State health facts. Available at: http://kff.org/health-reform/state-indicator/state-activityaround-expanding-medicaid-under-the-affordable-care-act/. Accessed October 28, 2014. 22. US Department of Health and Human Services. Objective E: ensure access to quality, culturally competent care, including long-term services and supports, for vulnerable populations. Available at: http://www.hhs.gov/about/strategic-plan/strategicgoal-1/index.html#obj_e. Accessed December 17, 2013. 23. Singh GK, Hoyert DL. Social epidemiology of chronic liver disease and cirrhosis mortality in the United States, 1935e1997: trends and differentials by ethnicity, socioeconomic status, and alcohol consumption. Hum Biol 2000;72:801e820. 24. Foraker RE, Rose KM, Suchindran CM, et al. Socioeconomic status, Medicaid coverage, clinical comorbidity, and rehospitalization or death after an incident heart failure hospitalization: Atherosclerosis Risk in Communities cohort (1987 to 2004). Circ Heart Fail 2011;4:308e316. 25. Volkova N, McClellan W, Klein M, et al. Neighborhood poverty and racial differences in ESRD incidence. J Am Soc Nephrol 2008;19:356e364. 26. Wolfe RA, McCullough KP, Schaubel DE, et al. Calculating life years from transplant (LYFT): methods for kidney and kidneypancreas candidates. Am J Transplant 2008;8[Pt 2]:997e1011. 27. Mulligan D. Liver Committee Chair on status of liver distribution discussion. UNOS Transplant Pro. Available at: http:// transplantpro.org/liver-committee-chair-status-liver-distributiondiscussion/. Accessed July 9, 2015.
DuBay et al
Vol. 223, No. 4, October 2016
Medicaid Expansion and Transplantation Outcomes
620.e1
eTable 1. Baseline Characteristics of Heart Transplant Recipients (2002 to 2011) Stratified by Public Insurance, Medicare, or Medicaid Characteristic
Sample size, n Age, y, mean (SD) Female sex, n (%) Race/ethnicity, n (%) Caucasian African American Asian Hispanic Other BMI, kg/m2, mean (SD)* Education, n (%)y Less than high school High school Some college 4 years college Diabetes, n (%)* Hypertension, n (%)* Albumin, g/dL, mean (SD)y Creatinine, mg/dL, mean (SD)*
Total
Private
Medicare
Medicaid
p Value
17,239 52.0 (12.5) 4,233 (24.6)
9,925 51.3 (11.8) 2,406 (24.2)
5,157 57.3 (10.8) 1,113 (21.6)
2,157 43.2 (13.8) 714 (33.1)
<0.001 <0.001
12,406 3,033 411 1,242 146 26.9
(71.8) (17.6) (2.4) (7.2) (0.9) (8.4)
7,659 1,358 274 561 73 26.7
(77.2) (13.6) (2.8) (5.7) (0.7) (10.0)
3,653 1,008 83 367 45 27.2
(70.9) (19.5) (1.6) (7.1) (0.9) (4.8)
1,094 667 54 314 28 26.8
(50.7) (30.9) (2.5) (14.6) (1.3) (6.4)
<0.001
530 6,113 3,580 3,601 4,266 6,998 3.7 1.30
(3.8) (44.2) (25.9) (26.1) (24.8) (44.5) (0.7) (0.7)
197 3,065 2,127 2,534 2,280 3,823 3.7 1.30
(2.5) (36.7) (26.8) (32.0) (23.0) (42.4) (0.7) (0.6)
199 2,050 1,021 909 1,535 2,352 3.7 1.36
(4.8) (49.1) (24.3) (21.8) (29.8) (50.2) (0.7) (0.8)
134 998 432 158 451 823 3.6 1.20
(7.8) (58.0) (25.0) (9.2) (20.9) (40.8) (0.7) (0.6)
<0.001
<0.001
<0.001 <0.001 <0.001 <0.001
*Missing for <10%. y Missing for 10% to 20%.
eTable 2. Baseline Characteristics of Lung Transplant Recipients (2002 to 2011) Stratified by Public Insurance, Medicare, or Medicaid Characteristic
Sample size, n Age, y, mean (SD) Female sex, n (%) Race/ethnicity, n (%) Caucasian African American Asian Hispanic Other BMI, kg/m2, mean (SD)* Education, n (%)y Less than high school High school Some college 4 y college Diabetes, n (%)* Hypertension, n (%)y Albumin, g/dL, mean (SD) y Creatinine, mg/dL, mean (SD)* *Missing for <10%. y Missing for 10% to 20%.
Total
Private
Medicare
Medicaid
p Value
13,206 53.3 (13.1) 5,771 (43.7)
7,964 (60.3) 52.1 (12.4) 3,446 (43.3)
4,271 (32.3) 57.8 (11.9) 1,805 (42.3)
971 (7.4) 43.0 (15.6) 520 (53.6)
<0.001 <0.001
11,269 1,061 161 630 85 24.7
(85.3) (8.0) (1.2) (4.8) (0.6) (4.9)
6,853 600 115 349 47 24.8
(86.1) (7.5) (1.4) (4.4) (0.6) (5.1)
3,691 342 36 172 30 25.0
(86.4) (8.0) (0.8) (4.0) (0.6) (4.5)
725 119 10 109 8 23.3
(74.7) (12.3) (1.0) (11.2) (0.8) (5.0)
354 4,903 3,001 3,266 2,318 2,930 3.9 0.90
(3.1) (42.6) (26.0) (28.3) (17.6) (24.8) (0.6) (0.5)
144 2,591 1,876 2,318 1,354 1,655 3.9 0.90
(2.1) (37.3) (27.1) (33.5) (17.0) (23.3) (0.6) (0.5)
144 1,816 932 831 749 1,101 4.0 0.90
(3.9) (48.8) (25.0) (22.3) (17.6) (28.7) (0.6) (0.4)
66 496 193 117 215 174 3.8 0.88
(7.6) (56.9) (22.1) (13.4) (22.1) (19.8) (0.7) (0.8)
<0.001
<0.001 <0.001
<0.001 <0.001 <0.001 <0.001
620.e2
DuBay et al
J Am Coll Surg
Medicaid Expansion and Transplantation Outcomes
eTable 3. Baseline Characteristics of Liver Transplant Recipients (2002 to 2011) Stratified by Public Insurance, Medicare, or Medicaid Characteristic
Sample size, n Age, y, mean (SD) Female sex, n (%) Race/ethnicity, n (%) Caucasian African American Asian Hispanic Other BMI, kg/m2, mean (SD)* Education, n (%)y Less than high school High school Some college 4 years college Diabetes, n (%)* Hypertension, n (%)* Albumin, g/dL, mean (SD)z Creatinine, mg/dL, mean (SD)z
Total
Private
Medicare
Medicaid
48,201 53.1 (10.2) 15,922 (33.0)
30,941 52.4 (9.5) 9,524 (30.8)
10,158 58.1 (9.8) 3,698 (36.4)
7,102 48.9 (10.7) 2,700 (38.0)
35,026 4,451 2,190 6,043 491 28.2
(72.7) (9.2) (4.5) (12.5) (1.1) (6.2)
23,601 2,716 1,405 2,951 268 28.2
(76.3) (8.8) (4.5) (9.5) (0.9) (6.4)
7,277 847 394 1,535 105 28.4
(71.6) (8.3) (3.9) (15.1) (1.1) (5.8)
4,148 888 391 1,557 118 28.0
(58.4) (12.5) (5.5) (21.9) (1.7) (6.1)
2,030 18,315 9,200 8,651 11,706 8,853 3.0 0.90
(5.3) (48.0) (24.0) (22.7) (24.3) (21.5) (0.7) (0.5)
697 10,715 6,399 6,726 7,051 5,452 3.0 0.90
(2.8) (43.7) (26.1) (27.4) (22.8) (20.6) (0.7) (0.5)
642 4,209 1,757 1,443 3,073 2,219 3.0 0.90
(8.0) (52.3) (21.8) (17.9) (30.3) (25.4) (0.6) (0.4)
691 3,391 1,044 482 1,582 1,182 2.9 0.88
(12.3) (60.5) (18.6) (8.6) (22.3) (19.5) (0.7) (0.8)
p Value
<0.001 <0.001 <0.001
<0.001 <0.001
<0.001 <0.001 <0.001 <0.001
*Missing for <10%. y Missing for 10% to 20%. z Missing for >20%.
eTable 4. Baseline Characteristics of Renal Transplant Recipients (2002 to 2011) Stratified by Public Insurance, Medicare or Medicaid Characteristic
Sample size, n Age, y, mean (SD) Female sex, n (%) Race/ethnicity, n (%) Caucasian African American Asian Hispanic Other BMI, kg/m2, mean (SD) Education, n (%)y Less than high school High school Some college 4 y college Diabetes, n (%)* Hypertension, n (%)* Albumin, g/dL, mean (SD)z Creatinine, mg/dL, mean (SD)* *Missing for <10%. y Missing for 10% to 20%. z Missing for >20%.
Total
Private
Medicare
Medicaid
90,548 51.7 (13.2) 36,052 (39.8)
25,970 51.5 (11.8) 10,441 (40.2)
60,217 52.2 (13.5) 23,563 (39.1)
4,361 45.4 (14.1) 2,048 (47.0)
42,380 28,821 5,179 12,534 1,633 27.8
(46.8) (31.8) (5.7) (13.8) (1.9) (6.4)
15,190 6,289 1,635 2,506 350 27.8
(58.5) (24.2) (6.3) (9.7) (1.3) (6.5)
26,068 21,063 3,117 8,796 1,173 27.8
(43.3) (35.0) (5.2) (14.5) (2.0) (6.4)
1,122 1,469 427 1,232 110 26.6
(25.7) (33.7) (9.8) (28.3) (2.5) (5.7)
5,867 36,895 18,278 15,460 30,821 70,856 3.9 8.3
(7.7) (48.2) (23.9) (20.2) (34.1) (85.5) (0.6) (3.4)
801 8,921 5,755 6,328 8,422 19,880 3.9 7.9
(3.7) (40.9) (26.4) (29.0) (32.5) (84.5) (0.6) (3.4)
4,398 26,009 12,004 8,850 21,127 47,670 3.9 8.5
(8.6) (50.7) (23.4) (17.3) (35.2) (86.2) (0.6) (3.4)
668 1,965 519 282 1,272 3,306 3.9 8.5
(19.5) (57.2) (15.1) (8.2) (29.3) (82.2) (0.6) (3.4)
p Value
<0.001 <0.001 <0.001
<0.001
<0.001 <0.001 <0.001 <0.001
Vol. 223, No. 4, October 2016
DuBay et al
Medicaid Expansion and Transplantation Outcomes
620.e3
eTable 5. State-Level Population and Medicaid Data from 2010 to 2011 Used to Create Figure 2A and B
State
Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware District of Columbia Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma Oregon Pennsylvania Rhode Island South Carolina South Dakota Tennessee Texas Utah Vermont
Adult Medicaid population, n
Expected Adult population, n
Medicaid proportion, n
30,2000 42,100 609,900 179,000 3,557,200 316,000 291,600 82,600 99,700 1,137,700 446,400 116,600 73,600 854,900 470,400 228,700 128,100 339,800 348,800 196,600 323,500 911,100 817,900 393,100 249,800 380,500 445,00 80,000 114,700 40,400 502,000 170,600 2,624,800 611,200 24,200 810,200 213,600 286,000 970,000 90,800 356,500 42,600 528,100 1,224,200 119,400 94,800
2,992,600 374,700 3,894,800 1,692,300 21,581,500 3,072,700 2,383,600 604,100 464,200 11,169,700 5,256,000 891,000 901,100 7,798,700 3,924,300 1,971,200 1,684,600 2,625,500 2,540,700 912,000 3,719,600 4,811,600 6,252,900 3,543,000 1,711,400 3,725,400 583,900 1,144,600 1,489,300 859,000 5,394,700 1,154,900 12,747,300 5,575,200 446,900 7,211,500 2,141,400 2,424,300 8,556,600 678,000 2,811,000 496,700 3,913,900 1,313,1600 1,546,700 439,100
0.101 0.112 0.157 0.106 0.165 0.103 0.122 0.137 0.215 0.102 0.085 0.131 0.082 0.110 0.120 0.116 0.076 0.129 0.137 0.216 0.087 0.189 0.131 0.111 0.146 0.102 0.076 0.070 0.077 0.047 0.093 0.148 0.206 0.110 0.054 0.112 0.100 0.118 0.113 0.134 0.127 0.086 0.135 0.093 0.077 0.216
Observed adult transplant recipients Medicaid All Medicaid recipients, n recipients, n proportion, n
14 4 72 18 620 35 31 4 15 233 53 6 1 180 47 26 24 51 63 9 56 75 69 51 26 75 3 22 7 4 60 14 272 57 4 100 22 21 148 0 31 6 52 104 16 7
525 37 707 305 3,959 483 407 145 72 2,604 1,054 117 94 1,540 719 368 343 494 599 93 933 747 1,215 484 400 810 65 227 213 102 1,196 220 2,299 1,367 62 1,182 383 271 1,810 74 623 85 789 2,820 242 78
0.027 0.108 0.102 0.059 0.157 0.072 0.076 0.028 0.208 0.089 0.050 0.051 0.011 0.117 0.065 0.071 0.070 0.103 0.105 0.097 0.060 0.100 0.057 0.105 0.065 0.093 0.046 0.097 0.033 0.039 0.050 0.064 0.118 0.042 0.065 0.085 0.057 0.077 0.082 0 0.050 0.071 0.066 0.037 0.066 0.090
Observed to expected ratio
0.26 0.96 0.65 0.56 0.95 0.70 0.62 0.20 0.97 0.88 0.59 0.39 0.13 1.07 0.55 0.61 0.92 0.80 0.77 0.45 0.69 0.53 0.43 0.95 0.45 0.91 0.61 1.39 0.43 0.83 0.54 0.43 0.57 0.38 1.19 0.75 0.58 0.66 0.72 0 0.39 0.82 0.49 0.40 0.86 0.42 (Continued)
620.e4
DuBay et al
J Am Coll Surg
Medicaid Expansion and Transplantation Outcomes
eTable 5. Continued
State
Virginia Washington West Virginia Wisconsin Wyoming O/E, observed to expected.
Adult Medicaid population, n
360,500 450,100 158,000 484,400 30,400
Expected Adult population, n
Medicaid proportion, n
4,831,600 4,104,800 1,168,400 3,741,900 337,200
0.075 0.110 0.135 0.129 0.090
Observed adult transplant recipients Medicaid All Medicaid recipients, n recipients, n proportion, n
49 43 18 51 3
952 663 217 663 43
0.051 0.065 0.083 0.077 0.070
Observed to expected ratio
0.69 0.59 0.61 0.59 0.77