Employment and Health Insurance in Long-Term Liver Transplant Recipients


*Corresponding author: W. Ray Kim, kim.woong@mayo.edu


This study was conducted to examine factors affecting health insurance and employment status in long-term liver transplant (OLT) recipients. All adult primary OLT recipients surviving at least 1 year were surveyed using existing questionnaires. Out of 217 eligible recipients, 186 (86%) responded. The median age of respondents was 55 years with a median survival after OLT of 3.4 years. The majority (98%) of respondents had health insurance coverage. Thirty-four (18%) reported having lost and/or having been denied health insurance since OLT, and 63 (34%) switched health insurance since OLT. Of the 179 that reported employment status, 98 (55%) were employed, including homemakers and students, while 39 (22%) were retired and 42 (24%) unemployed. The majority (76%) of those unemployed cited poor health as the reason for unemployment, followed by 5 (12%) who feared loss of disability or Medicaid benefits. Fourteen reported to have been denied or terminated from employment because of their transplant. In the regression analysis, employment prior to transplantation (odds ratio (OR) = 5.1), age less than 57 (OR = 5.1), physical function score >52.4 (OR = 3.6) and general health score >33.3 (OR = 7.6) were significantly associated with employment. These data may help identify high-risk pre-OLT patients for intervention measures such as work rehabilitation.


Liver transplantation (OLT) is now firmly established as an effective means to treat patients with end-stage liver disease. Outcomes of transplantation have improved steadily in the past 20 years and the most recent national average for 1-year survival for deceased donor is 86% and for 3-year survival 78% (1). In addition to prolonging survival, OLT improves recipient quality of life. Physical function and emotional, social well-being improve significantly following OLT (2–10).

Despite these achievements in overall patient outcomes, OLT has been criticized for its expense (11,12). The national expenditure for OLT was estimated to have reached $1.4 billion in mid 1990s (13). In an era of budgetary constraints in health care, consideration of cost-effectiveness of medical interventions has become critically important. There has been increasing emphasis on improving the economic efficiency with which liver transplants are performed (14–16).

Employment after OLT is not only a marker of clinically significant individual health recovery, but also, on a larger scale, an indirect means for society to recoup some of the resources that were expended in its support of transplantation activities. However, there have been several reports in the past that only a relatively small proportion (20–40%) of liver transplant recipients returns to work. These studies also explored the reasons for liver transplant recipients' inability to return to work, including depression (17), diminished physical health (18), older age (19) and retirement (19,20).

Health insurance coverage is an issue closely related to employment. The vast majority of private health insurance in the United States is obtained as a fringe benefit of employment; nevertheless, many full-time employees remain uninsured (21). Furthermore, means-testing of public health insurance programs (e.g. Medicaid), implies that individuals returning to employment may lose eligibility for public coverage. To the extent that private coverage is unavailable or less generous, the potential loss of public health insurance may be a barrier to employment. Indeed, in a small study by Hunt et al. (18), being on Medicaid pre-OLT was found to be a significant negative factor for post-OLT employment. Furthermore, employers may be discouraged from hiring someone with a significant health problem such as having a transplant, especially if they are self-insured or small employers unable to pass on the resulting group premium increases to employees (22).

This study surveyed long-term liver transplant recipients about their health insurance coverage and employment status in order to explore the relationship between health insurance and employment, controlling for observable health status. One of our questions was whether difficulty obtaining health insurance, loss of disability income or loss of Medicaid/Medicare coverage were associated with employment outcomes.


Study subjects and procedure

This study included all adult (age >18 years) liver transplant recipients at the Mayo Clinic who underwent primary OLT during a 3-year period between January 1996 and December 1998 and survived at least for 1 year. Eligibility included residency in the United States and understanding of written English.

A list of all eligible subjects was created from an institutional database, which contains a comprehensive array of data including demographic information, date of OLT, current vital status, pre-transplant characteristics, such as pre-transplant employment status and type and severity of liver disease. Liver disease was categorized into hepatitis C, alcoholic liver disease, cholestatic and other. Disease severity was measured by model for end-stage liver disease (MELD) prior to transplantation.

Questionnaires described in the following section were mailed to the eligible patients along with a letter explaining the study. In subjects who did not respond to the initial survey, a second, identical set of questionnaires was mailed. Subjects who did not respond to the second mailing were contacted by phone and if agreeable were interviewed over the phone. Patient participation was entirely voluntary and no remuneration was provided for participating in the study. This study was approved by the Institutional Review Board of Mayo Foundation.


Two self-administered questionnaires were used. A questionnaire inquiring about employment and health insurance consisted of 29 items: This questionnaire was adapted from two questionnaires that were previously used in other solid organ transplant recipients (23). The employment component consisted of 16 questions regarding employment status, ability to work and level of remuneration. The health insurance component (13 questions) addressed types of insurance coverage, any effects of transplant on health insurance and employment and impact of health insurance in obtaining employment. This questionnaire is available on request.

The second questionnaire was the Short Form-36 (SF-36), a widely used instrument to assess health status. It measures eight scales including physical and social functioning, physical and emotional role limitations, mental health, vitality, pain and general health perceptions.

Data analysis

Data from returned questionnaires were entered into a SAS (SAS Institute, Cary, NC) database. Employment status was categorized to full-time paid employment, part-time employment, temporary leave, homemaker/student, retired and unemployment. Employment status and other responses were tabulated separately for subjects under 65 years of age and those 65 and older.

Responses to the SF-36 instrument were scored for the eight sub-scales following the published algorithm (24). Physical and mental component summary scores were calculated. These scores were compared with age- and gender-matched norms as published, using the one-sample t-test. SF-36 has been scaled so that the average health-related quality of life of the reference population in each of the eight sub-scales is 50 (SD set to 10) with 0 being the worst to 100 being the best score.

In exploring sociodemographic factors and health status parameters that are associated with employment status, we used recursive partitioning (RP) analysis. RP is a nonparametric method used to classify observations on the basis of a large number of possible predictive variables. At the beginning, all potential predictor variables are considered and the one variable and cut-point value of that variable that most effectively divides observations into two sub-groups are identified. For the purpose of RP analysis, we sub-divided the employed group into employed and other that includes student, homemaker and temporary leave. Once the first binary partition is achieved, each new sub-group is then considered for further splits in a recursive manner. This iteration is continued until further partitioning would not improve accuracy of the model (25). We also used a parametric approach (multivariable logistic regression) to determine factors that are associated with employment (including full- and part-time jobs, student, homemaker and temporary leave) among respondents 65 years old or younger. Explanatory variables included age, pre-transplant employment, time since transplantation, original diagnosis of liver disease and pre-transplantation MELD, education level, marital status, type of insurance, loss or denial of insurance coverage and relevant health status parameters identified in the RP model. A full model was constructed first, incorporating all of these variables. A reduced model was then created, selecting only those that are significant using backward elimination using a p-value of 0.05 as exit criteria.

Our sample size calculation was based on our multivariable analysis being able to evaluate up to 10 predictors. We sought to recruit approximately 100 recipients with employment and roughly the same number without.


Of 249 patients who underwent OLT during the 3-year period, 217 who were alive and met the eligibility criteria were sent questionnaires. At the initial mailing, 125 subjects responded and 1 refused to participate. The subsequent mailing resulted in 32 additional responses and 2 additional refusals. Of the remaining 57 subjects, 29 responded to the telephone survey, leading to a total of 186 respondents, with the overall response rate of 86%. Phone interviewees did not differ from the rest of the respondents with regard to age, gender, employment and insurance status.

Demographic and clinical characteristics of the 186 respondents are summarized in Table 1. The median age at the time of survey was 55 years with a median interval between OLT and survey of 3.4 years. Slightly over half of the respondents (56%, n = 104) were men, and a substantial majority (82%, n = 152) were Caucasian. The vast majority of patients (92%, n = 169) completed at least high school, and 77% (n = 143) were married. Cholestatic liver disease (38%, n = 70) and viral hepatitis (19%, n = 36) were the two most common diagnoses.

Table 1.  Demographic characteristics of respondents
Age, median (range)
 At OLT52 (22–71)
 At survey55 (26–75)
Interval since OLT, median (range)3.4 (1.9–4.9)
 Male104 (56%)
Race (no response = 15)
 White152 (82%)
Education (no response = 2)
 College or higher74 (40%)
 High school or junior college95 (52%)
 No high school diploma15 (8%)
Marital Status (no response = 1)
 Married/cohabitating143 (77%)
 Separated/widowed/divorced26 (14%)
 Never married16 (9%)
 Alcoholic11 (6%)
 Viral hepatitis36 (19%)
 Cholestatic70 (38%)
 Malignant/fulminant19 (10%)
 Other/unknown50 (27%)
MELD prior to OLT, median (range)13 (6–49)

Responses to questions about health insurance coverage are summarized in Table 2. A large majority (98%, n = 183) of respondents had some type of health insurance coverage, including 18% who had their insurance through their spouse. Seventy-one (38%) reported having more than one source of health insurance coverage. The majority (55%, n = 102) of the respondents carried private insurance, while 56 (30%) had public insurance, including Medicare, Medicaid, Veteran's Administration (VA) or Native American programs.

Table 2.  Health insurance status among respondents
N = 186
Age <65
N = 155
Age 65+
N = 31
  1. *Five respondents with government programs reported not to have ‘health insurance.’

Currently have health insurance
 Yes183* (98%)152 (98%)31 (100%)
 No3 (2%)3 (2%)0
Has more than one insurance coverage
 Yes71 (38%)47 (30%)24 (77%)
 No114 (61%)108 (70%)6 (19%)
 No response1 (1%)0 (0%)1 (3%)
Type of primary coverage
 Fee for service58 (31%)52 (34%)6 (19%)
 Managed care44 (24%)42 (27%)2 (6%)
 Government56 (30%)36 (23%)20 (65%)
 Other/not sure16 (9%)16 (10%)0
 No response12 (7%)9 (6%)3 (10%)
 Positive125 (67%)103 (66%)22 (70%)
 Neutral29 (16%)26 (17%)3 (10%)
 Negative22 (12%)19 (12%)3 (10%)
 No response10 (5%)7 (5%)3 (10%)
Duration with current coverage
 <1 year13 (7%)13 (8%)0
 1–2 years26 (14%)22 (14%)4 (13%)
 2–5 years42 (23%)34 (22%)8 (26%)
 >5 years96 (52%)80 (52%)16 (51%)
 No response9 (4%)6 (4%)3 (10%)
Times health insurance was switched since OLT
 0123 (66%)103 (66%)20 (65%)
 1–258 (31%)47 (30%)11 (35%)
 3–55 (3%)5 (4%)0
Lost coverage since OLT20155
 Medical or OLT related4 (20%)4 (27%)0
 Job related5 (25%)4 (27%)1 (20%)
 Other11 (55%)7 (46%)4 (80%)
Denied coverage since OLT21201
 Medical or OLT related18 (86%)17 (85%)1 (100%)
 Nonmedical2 (10%)2 (10%)0
  No response1 (5%)1 (5%)0

Two-third of the respondents were satisfied with their insurance, and satisfaction with their insurance did not differ between the two age groups. More than one-half (n = 96) of the respondents reported having had the current coverage for more than 5 years. On the other hand, 63 respondents (34%) reported having switched health insurance since their transplantation, including 5 that switched their insurance three times or more. Moreover, 20 (11%) reported having lost insurance benefit since their transplantation, of whom 4 attributed it to medical issues. Twenty-one (11%) had been denied coverage following OLT, the majority of whom (86%, n = 18) attributed denial because of their transplant or other medical reasons.

Table 3 summarizes employment status of respondents. Seven of 186 respondents did not answer their employment status. Of the remaining 179 respondents, 98 (55%) were employed, including 62 paid full-time and 19 paid part-time employment; 17 were homemakers (n = 14), students (n = 2) or on temporary leave (n = 1). The employment rate was higher in respondents less than 65 years of age (61%, n = 90; overall employment, 50%, n = 74, paid employment), whereas only 26% (n = 8) of those 65 years or older were employed. Of those with paid employment, managerial or professional employment was the most common job category. With regard to income, 35% (n = 28) reported their current income to be greater compared to prior to transplantation. A similar number reported their income unchanged (37%, n = 30).

Table 3.  Employment status of respondents
N = 179
Age <65
N = 148
Age 65+
N = 31
  1. *Among those with paid employment.

  Full time62 (63%)57 (63%)5 (63%)
  Part time19 (19%)17 (19%)2 (25%)
  Homemaker/student/temporary leave17 (18%)16 (18%)1 (12%)
 Employment class*
  Managerial/professional30 (37%)29 (40%)1 (14%)
  Sales/admin/service26 (32%)21 (28%)5 (72%)
  Ag/product/construct/other21 (26%)21 (28%)0
  No response4 (5%)3 (4%)1 (14%)
 Current income comparison*
  Less than before OLT15 (18%)14 (18%)1 (14%)
  Same as before OLT30 (37%)27 (37%)3 (43%)
  More than before OLT28 (35%)27 (37%)1 (14%)
  No response8 (10%)6 (8%)2 (29%)
 Looking for work
  Yes14 (33%)14 (34%)0
  No28 (67%)27 (66%)1
 Wishes to return to work
  Yes21 (50%)21 (51%)0
  No17 (40%)16 (39%)1
  No response4 (10%)4 (10%)0
 Reason for unemployment
  Poor health (disabled)32 (76%)31 (76%)1
  Lose insurance benefits5 (12%)5 (12%)0
  Other4 (10%)4 (10%)0
  No response1 (2%)1 (2%)0

Thirty-nine (22%) respondents described themselves as retired, including 17 who were under 65 years of age. The median age of these 17 respondents was 62 (interquartile range: 58–63).

Overall, there were 42 respondents who were unemployed. One-half of the unemployed stated that they wished to return to work, whereas one-third of the unemployed were actually looking for work. Of the 21 who wished to return to work, 8 expressed interest in a work rehabilitation program. The majority (76%, n = 32) of those unemployed cited disability (poor health) as the reason for unemployment. Five individuals (12%) reported that fear of losing benefits because of increased income was the reason for unemployment.

Overall (n = 179), a small percentage reported to have been denied (6%, n = 10) or terminated (3%, n = 6) from employment because of their transplant. Of 42 respondents who were unemployed, 5 (12%) reported having been denied employment and 2 (5%) terminated for having a transplant.

One hundred and seventy-eight (96%) of the respondents filled out the SF-36 questionnaire. Figure 1 illustrates eight domain scores and two summary scores of SF-36. Responses in physical domains including physical function (PF: 45.4 ± 11.3), physical role limitation (RP: 43.5 ± 12.5), bodily pain (BP: 47.8 ± 11.5) and general health (GH: 44.1 ± 12.4), as well as the physical component summary score (43.7 ± 12.0) were uniformly lower than the norm obtained in general population. In the mental domain, vitality (VT: 45.6 ± 11.8) and social function (SF: 45.5 ± 12.8) scores were lower in our respondents than the norm. However, scores on emotional role limitation (RE: 48.2 ± 12.2) and mental health (MH: 50.1 ± 10.1) were comparable to the norm. Overall, the mental component summary score (49.5 ± 11.2) was not significantly different from that in the well population.

Figure 1.

SF-36 domain and summary scores in 178 respondents. Physical health is assessed by four domains including physical function (PF), physical role limitation (RP), bodily pain (BP) and general health, as well as the physical component summary score (PCS). Mental health is represented in four domains including vitality (VT), social function (SF), emotional role limitation (RE) and mental health (MH), which are summarized by the mental component summary score (MCS). p-values related to comparisons with population norms.

Results of the recursive partitioning (RP) analysis are shown in Figure 2. The criterion that most effectively classified the group into two was the employment status prior to transplantation. The majority (82%, n = 60) of the pre-transplant employed remained employed following their transplant. The subsequent hierarchical classification rule in those who were not employed pre-transplant is as follows: (i) Those who were 62 years or older were more likely to be retired (ii). Of those younger than 62 years, a score of less than 33.3 on the general health (GH) sub-scale of SF-36 was associated with unemployment (iii). Those who had better score on general health, but were aged 57 years or older, were less likely to be employed (iv). Of those who were younger than 57 years, a mental composite score greater than 56.4 was associated with employment, (v) Of those with a mental composite score less than 56.4, a physical function score less than 52.4 was associated with unemployment.

Figure 2.

Results of the recursive partitioning analysis. The factor that separated subjects into two groups most efficiently was pre-transplant employment status. If the answer was yes (white box), 60 of 73 (82%) were employed. The answer was no (black box) in 106 subjects, in whom the next question that separated subjects into two groups best was whether their age was greater than 61.5 years. If the answer was yes (white box), 31 of 37 (84%) were retired. Otherwise, further classification rule applied to the remaining subjects.

Table 4 summarizes the results of logistic regression analysis, modeling employment (including full- and part-time employment, student and homemaker) among subjects aged 65 years or younger. In the full model (all candidate variables considered), pre-transplant employment status, transplantation for the diagnosis of hepatic malignancy, being married, physical function score >52.4, general health score >33.3 and age 56 years or younger were significant variables associated with post-transplant employment status. In the final multivariable model, only four variables were significantly associated with post-transplant employment status. These were employment prior to transplantation (odds ratio (OR = 5.1, CI (1.8–14.0)), age less than 56.5 years (OR = 5.1, CI (1.8–14.3)), physical function score >52.4 (OR = 3.6, CI (1.1–11.2)) and general health score >33.3 (OR = 7.6, CI (2.4–24.4)), all of which were significantly associated with employment at the time of survey. Variables for which we found no significant effect include time since transplantation, original liver disease diagnosis, MELD score at the time of transplantation, education level, marital status, Medicare/Medicaid and having lost or been denied insurance since transplantation.

Table 4.  Factors that are associated with employment following liver transplantation among recipients 65 years or younger
 Full modelFinal model
OR95% CIOR95% CI
  1. OR = odds ratio.

  2. *p < 0.05.

  3. C(concordance)-statistic is equivalent to the area under receiver operating characteristic curve (36). Commonly used in evaluating a diagnostic test, this statistic may range from 0 to 1, with 1 corresponding to perfect discrimination and 0.5 to what is expected by chance alone. A c-statistic between 0.8 and 0.9 indicates excellent diagnostic accuracy and a c-statistic greater than 0.7 is generally considered as a useful test.

Pre-transplant employment7.5*2.3–24.55.1*1.8–14.0
Time since transplant (months)1.00.9–1.1
 Viral hepatitis0.40.1–2.2
 College or higher2.10.7–6.8
Health status
 Physical function >52.44.5*1.2–17.73.6*1.1–11.2
 General health >33.319.0*3.9–92.87.6*2.4–24.4
 Mental component >–1.5
Lost or denied insurance since OLT0.40.1–1.5


The goal of this study was to examine the relationship between health insurance and employment among liver transplant recipients who survived more than a year. Employment is a common vehicle to obtain health insurance coverage, whereas a transplant recipient may be deemed too ‘high risk’ for employment-based health insurance and be excluded from job opportunities. The latter is especially true with small businesses that lack a large pool of employees to disperse the burden of high costs associated with organ transplantation without increasing overall health insurance costs.

In light of the magnitude of expenses associated with the transplantation procedure and post-transplant care, procurement and maintenance of health insurance coverage is of critical importance not only for ensuring provision of adequate medical care but also for financial welfare of the recipients and their family. In this survey, the great majority of our patients (98.4%, n = 183) did have health insurance coverage, some from more than one insurer. Thus, health insurance coverage in our patients is obviously higher than in the general population. According to a recent statistic, 16% of Americans do not have health insurance (26). This however, should not be surprising, as high expenses associated with OLT and post-transplant care make it impossible for someone to undergo OLT without health insurance coverage. It was reassuring, however, to see that coverage commonly continued for more than 5 years and only a tiny fraction (4/186, 2%) lost insurance due to medical issues. Respondents were also mostly satisfied with their coverage.

Approximately, one-third of our respondents reported changing their insurance carrier following their transplant. This is a large fraction, even when compared to a group of relatively unhealthy individuals dissatisfied with their insurance (27). Although switching insurance is not typically regarded as the best marker of dissatisfaction, it is rather unusual for patients with chronic disease states to do so at such a high percentage (27). Typically, such patients would keep a less than desirable insurance in order to maintain a provider relationship (28). We also found that 29 recipients under 65 years of age (19%) had lost or were denied insurance coverage since their transplantation, although the majority (93%) of these individuals were currently under coverage. In the regression model, having lost or been denied insurance was not significantly associated with employment status. Thus, while denial of insurance coverage of transplant recipients is not uncommon, we did not find evidence that health insurance coverage has a significant impact on employment.

An implicit premise on which our society supports organ transplantation is that certain benefits, monetary or otherwise, will be returned from transplant recipients. Rehabilitation of patients to enable them to join the productive work force is both of social and economic value. However, previous studies have shown that there is ample room for improvement. In a survey conducted in Canada, only 40% of liver transplant recipients were engaged in full-time employment, including homemakers and students. Of the remaining 60%, 17% were employed part-time, while 43% were unemployed (20). Other studies conducted outside of the United States found employment rates of only 21–22% up to 6.4 years post-transplantation (19,29). With respect to reports from the United States, a retrospective study by Eid et al. found an employment rate of 57% (n = 46) excluding homemakers and 91% including homemakers (30). On the other hand, a cross-sectional study by Hunt et al., found a lower employment rate of 42% (n = 52) in liver transplant recipients (18).

Thus, our study's employment rate of 55% (n = 98) is one of the highest among studies reported so far, although small sample sizes and varying definitions of employment across previous studies make it difficult to compare directly. Such a comparison is further confounded by the fact that current health status is an important determinant of employment status, as shown in this and other studies. In those who were employed, the level of compensation was the same or higher than before transplantation. It is encouraging to note this improved productivity post-transplantation. As market wages are commonly used in economic evaluation of medical technology as an indicator for productivity gained due to successful treatment or lost due to poor health (31), this may be suggestive of improved health and quality of life. We did note that the majority of employed respondents in this study held jobs in categories that do not require hard physical labor. This and the fact that most of the unemployed were disabled may suggest that employment among liver transplant recipients may improve with education programs that help transition patients from physical labor to less physically demanding jobs.

Potential loss of health insurance (primarily Medicaid/Medicare) or disability benefits has been identified as a major disincentive for employment in other solid organ transplant recipients (32,33). A study by Raiz reported that although 72% of post-renal transplant recipients felt able to work, only 58% were employed. Regression analysis revealed that receipt of monthly disability check was a significant factor associated with unemployment (33). Similarly, Hunt et al. reported that liver transplant patients on Medicaid were 1.7 times more likely to remain unemployed (18). In our study, although poor health (disabled) was the most common reason for unemployment, fear of loss of benefits (disability and Medicaid) was an infrequent (n = 5, 12%) reason for unemployment. In the logistic regression analysis, Medicaid was not associated with unemployment, once other factors were accounted for.

Investigators have searched for health status indicators that influence employment status. Adams reported several health status indicators including ambulation, physical function and pain as significant predictors of employment (20). Another multivariate analysis showed that marital status and musculo-skeletal symptoms independently predicted employment status of liver transplant recipients (34). In the study by Hunt et al., although there was not a significant difference in objective measurements of overall health, unemployed patients scored lower in perceived physical function and physical role limitation compared to employed patients (18).

The relationship between health status and employment found in this study is informative. First, in overall patients, the scores in the physical domains were lower than the general population, whereas scores in the mental health domains were in line with the norm. This indicates that while our liver transplant recipients may still experience residual physical difficulties after transplantation, they enjoyed emotional well-being. The latter may be related to a response shift in that their experience prior to and through OLT gives them an appreciation of their health emotionally. Second, of the health status parameters, three were identified in the RP analysis to have significant association with employment, namely physical function and general health domains and mental component summary score. While the cut-off scores determined by the RP model for the physical function domain and mental component summary scores were slightly above the norm (50), those for the general health domain were almost two standard deviation lower at 33. This result indicates that liver transplant recipients were engaged in employment even when their perception of their health was quite low.

Other factors that were associated with employment identified in other studies included liver disease diagnosis (patients with alcoholic liver disease less likely to have employment) (35), marital status and educational level, none of which was replicated in this study. The only possible exception is the presence of malignancy prior to transplantation, which had a significant effect in the full model, but was excluded from the final model. There may be conceivable explanations for the association, both physical (e.g., recurrent tumor) and psychological (e.g. enjoying life cancer free), and our sample size may have prevented that from being recognized to be significant.

With regard to pre-transplant employment status, our data support the study by Adams et al., in which the duration of disability and unemployment pre-transplant as well as the age of the recipient were inversely correlated with likelihood of employment. While our results do not demonstrate causality between pre-transplant employment and post-transplant employment, one might surmise that transplanting patients before they become ill enough to become disabled and lose employment could potentially increase the probability of post-transplant employment. While this remains a conjecture, attempting to establish causality in this matter, for example, by a randomized trial, would not be ethical. However, the recent change in the organ allocation using the MELD scale favors patients with advanced liver disease and may have an adverse effect on the proportion of patients still employed by the time of transplantation. We look forward to further data in the future to track post-transplant employment rates after implementation of the MELD.

In summary, the most important factors associated with post-transplant employment were pre-transplant employment status and health status indicators in physical function, health perception and overall mental health. In our patient population, health insurance coverage was almost universal, although loss or denial of coverage due to transplant was relatively common. We found no evidence that health insurance coverage significantly impacted employment status. As the results of OLT continue to improve, rehabilitation of recipients is becoming increasingly important. It is important to keep in mind that rehabilitation encompasses not only recovery of physical functions but also achievement of psychological well-being and ultimately the ability to productively participate in society.


This work was supported by a grant from the National Institutes of Health (DK-34238) and a grant from the American College of Gastroenterology.