Health Status in Young Adults Two Decades after Pediatric Liver Transplantation

Abstract

We conducted a cross-sectional study of patients who underwent pediatric liver transplant between 1988 to1992 to evaluate long-term health status. Survivors completed socio-demographic, medical and Health Related Quality of Life (HRQOL) surveys by mail including the SF 36v2, PedsQL™4.0 Generic Core Scale, PedsQL™ Cognitive Functioning Scale and PedsQL™3.0 Transplant Module. SF 36 scores were converted to SF6D-based utilities and risk factors for lower outcomes were assessed.

Eighty-five of 171 patients had survived, Fifty-six were contacted with a response rate of 66%. Median age at LT was 0.86 yrs (IQR 0.58–3.0) and 64.3% had biliary atresia. Mean age at survey was 23.0±4.4 years. 62% attended college, 68% lived with parents and 80% of those over 23 were employed. Patient health utilities were lower than norms (0.75±0.12 versus 0.82±0.18, p<0.01) and correlated with unemployment (p<0.042), hospitalizations (p<0.005), and lower education level (p<0.016). Lower PedsQL™3.0 Transplant Module and PedsQL™ 4.0 Generic Core Scale scores correlated with unemployment (p=0.006, p=0.009) and hospitalizations (p=0.006, p=0.02). Pediatric transplant recipients who survive to adulthood have lower physical HRQOL, measureable transplant related disability and lower health utility. Transplantation is life saving however physical and psychological sequelae continue to affect health status up to two decades later.

Introduction

The durability of liver transplantation (LT) for the treatment of end stage liver disease in children is well established. Ten-year survival for pediatric LT recipients exceeds 80% and many long-term survivors have now reached young adulthood. Studies of long-term LT survivors are limited by small numbers and focus almost exclusively on graft function and patient survival. Few studies have addressed health status in long-term survivors.

Adolescents with chronic illnesses may struggle in their transition to young adulthood and a more independent lifestyle. Although they may be healthy, other factors such as activity restrictions, daily medications and fear of graft dysfunction may reduce their satisfaction with their overall health status. The framework of patients’ health status is composed of Health-related Quality of Life (HRQOL), and health values (also known as preferences or utilities). HRQOL can be conceptualized to include physical health, mental health, social functioning, role functioning, and general health perceptions.(1) Health utilities are values that represent the strength of an individual’s preferences for a specific health state. One year in perfect health is assigned a value of 1 and the state of death is assigned a value of zero. Health states between these two anchor points have a utility value between zero and one, and states considered worse than death have a negative value. Patients provide an assessment of their health utilities by estimating the desirability of a particular health state through their willingness to undergo risky treatments or sacrifice life expectancy in order to improve health (2)

Utility scores may be measured directly through techniques such as the Standard Gamble (SG) and Time Trade Off (TTO) or indirectly with a health state classification system that assigns values to discrete health states such as the SF-6D (Short Form 6 Dimensions) or the HUI (Health Utility Index). These scores are needed to calculate Quality Adjusted Life Years (QALYs) that are used for decision and cost effectiveness analysis, which in turn are methods used to inform individual or population level medical decision-making and health policy.

There is a growing body of literature on HRQOL after pediatric LT, including several reports that have followed survivors into young adulthood. (3–5)Young adults who have lived through a life threatening illness as children are a unique and growing population of patients. Growing up with a chronic medical condition, many may not know what it means to be ‘well’. For example, HRQOL in young adults who have had a heart or liver transplant, cancer survivors and those with hydrocephalus remain lower then the general population. (5–8) Although health utilities for adult populations both with and without chronic illness have been estimated, only a few studies have explored health preferences in adolescents with chronic diseases and none have reported on the health utilities of young adult survivors of pediatric solid organ transplantation.(9, 10) The present study was undertaken to evaluate the health statusina group of pediatric LT survivors followed into adulthood.

Materials and Methods

A cross-sectional study was performed to evaluate 18–22 year old survivors of pediatric LT, with emphasis on survival rate, employment, social function, HRQOL and health utilities. All pediatric LT recipients who received their first transplant at the University of Chicago Hospitals between July 1st, 1988 and Dec 31st 1992, who were able to complete the self-report questionnaires, and received follow up care at Children’s Memorial Hospital (CMH) or the University of Chicago Hospitals were eligible. Information regarding patient demographics, transplant type, and addresses were collected through the liver transplant administrative database at the University of Chicago. Current patient status was determined by reviewing the medical record, calling their last known home phone number, mailing certified letters to their last known address, calling their gastroenterologist/hepatologist of record and using the Social Security Death Index. (SSDI)(11) Attempts were also made to contact patients lost to follow up using the social networking site Facebook. The investigators created a Facebook account under the research groups’ name and contacted individuals who had the same birthdays and similar names as their subjects. A secure message was sent to these individuals, asking for their participation in a liver transplant research study and mentioning one investigator (EMA), who was a health care provider for the target participants at the time of their transplant. In cases where the profile page was in a language other than English, or there were more than ten results, and birthdays were not visible to assist with narrowing the search results, a message was not sent. The Institutional Review Boards of both institutions approved the study.

There were 178 primary LT performed between July 1, 1988 through December 31^st 1992; two other patients that were transplanted during that time period at other institutions but transferred their care to the University of Chicago Hospitals and Children’s Memorial Hospital were also included. Seven of these patients were foreign nationals living overseas and no attempt was made to contact them and five patients had a medical condition that caused sufficient incapacitation to preclude self-report. There were 86 deaths confirmed by the medical record and the SSDI.

A written questionnaire regarding socioeconomic and current medical status as well as generic and disease specific HRQOL surveys were sent by mail to the remaining 82 patients. We were successful in contacting 56 of these patients, measured by phone contact or acceptance of certified mailings, and twenty-six patients were lost to follow up. The overall participation rate among the patients who were contacted was 66%.

Survey Instruments

Questionnaire

After a review of the current literature on long-term survivors, a questionnaire was developed by the research team to evaluate demographic, social and medical characteristics of our population. Medical questions included: primary diagnosis, number of medications and side effects, hospitalizations and surgeries in the last three years and liver function test results. See Table 1 for sample questions and Supporting information section for the full battery.

Table 1.

Sample questions from General Demographic and Medical questionnaire.

Are you currently working?

Are you currently in school?

What is your highest level of education?

How do you pay for your medications?

What was your diagnosis?

What type of transplant did you have?

Has your doctor told you that you need another transplant?

Are you currently on the transplant waiting list?

Have you had a kidney transplant?

What medicines are you currently taking?

Have you had any side effects from these medications?

How often do you forget to take your medication?

In the past 3 years, how many admissions to the hospital have you had?

When was the last time you had a liver biopsy?

What were the liver biopsy results?

Have you had rejection in the past 3 years?

What were the results of your latest liver blood tests?

In the past three years, have you had a biliary stent placed?

In the past three years, have you had any surgery on your abdomen?

In the past three years, have you had any bloody vomiting?

Quality of Life Surveys

Two common approaches to measuring patient symptoms and other aspect of quality of life involve the use of generic or targeted scales. Generic scales are broad in focus, can be used across a number of patient populations, and as a result, can be used to compare scores across the general and patient populations. Targeted tools are developed to consider the unique aspects of disability faced by patients with a chronic illness. As such, disease specific tools are more sensitive to clinical changes. There is often reason to consider both levels of focus as part of a quality of life assessment strategy.

The Short Form-36 Health Survey

(SF-36v2) originated from the Medical Outcomes Study (MOS) which attempted to explain variations in patient outcomes by measuring changes in quality of life. It relies upon patient self-reporting and is one of the most commonly utilized generic health status measures. It is now widely utilized by managed care organizations and Medicare for assessment of care outcomes in adult patients as well as in general population health surveys conducted nationally. Use of this survey allows us to compare our results with multiple other studies in similar populations, compare with normal healthy controls and calculate health utilities.

The SF-36v2 provides a reliable and valid generic HRQOL health profile for use with individuals aged ≥ 18 years.(12) The survey includes 8 sub-scales: physical functioning, role physical, bodily pain, general health perceptions, vitality, social functioning, role emotional, and mental health. Physical component (PCS) and mental component (MCS) summaries are derived from these 8 sub-scales. PCS and MCS scores “reflect a combination of physical and mental function and well-being, the extent of social and role disability, and personal evaluation of health status.” Scoring of the SF-36v2 is standardized and norm-based for scale and summary measures, with a mean of 50 and standard deviation of 10 (range, 0–100). Norms are representative of the 1998 general United States population, with higher scale and summary scores corresponding to better HRQOL.(12)

Health Utilities were calculated by transforming the SF-36v2 score into a SF-6D equivalent. The SF-6D utilizes 11 questions from the SF-36v2 to create 6 domains and 249 unique health states. Valuations for the health states were obtained from an 836-person sample using the standard gamble method.(13, 14)

The PedsQL™

Measurement Model is a modular approach to measuring health-related quality of life (HRQOL) in children, adolescents and young adults. It has developmental and age appropriate generic core scales that may be used for healthy populations as well as patients with acute or chronic illnesses. There are also disease specific modules that complement the generic core scale, and are used to provide greater measurement sensitivity in the affected clinical populations. We have previous experience with multiple versions of the PedsQL™ and have also helped to validate the newly developed Transplant Module. The PedsQL™ Transplant Module was designed provide a more thorough understanding of the multidimensional nature of the child’s experience regarding the impact of solid organ transplantation. To supplement the transplant module, we also used the PedsQL™Cognitive Functioning Scale. Previous studies of HRQOL in liver transplant recipients have demonstrated lower school functioning scores on the generic PedsQL™. This is a domain that is not well captured on the generic core scale and is therefore of interest to measure using a more sensitive scale

The PedsQL™ 4.0 Generic Core Scales Young Adult Version

The 23-item PedsQL™ 4.0 Generic Core Scales designed for use in young adults aged 18–25 years encompasses: 1) Physical Functioning (8 items), 2) Emotional Functioning (5 items), 3) Social Functioning (5 items), and 4) School Functioning (5 items)To calculate the Psychosocial Health Summary Score, the mean is computed as the sum of the items divided by the number of items answered in the Emotional, Social, and School Functioning Scales.(15, 16)

PedsQL™ Cognitive Functioning Scale

This scale contains 6 items that address cognitive fatigue. The format, instructions, Likert response scale, and scoring method are identical to the PedsQL™ 4.0 Generic Core Scales, with higher scores indicating better HRQOL(17)

PedsQL™ 3.0 Transplaant Module

The 46-item PedsQL™ 3.0 Transplant Module encompasses 8 Scales: 1) About My Medicines I (9 items; barriers to medical regimen adherence), 2) About My Medicines II (8 items; medication side effects), 3) My Transplant and Others (8 items; social relationships and transplant), 4) Pain and Hurt (3 items; physical discomfort), 5) Worry (7 items; worries related to health status), 6) Treatment Anxiety (4 items; fears regarding medical procedures), 7) How I Look (3 items; impact of transplant on appearance), and 8) Communication (4 items; communication with medical personnel and others regarding transplant issues) The format, instructions, Likert response scale, and scoring method for the PedsQL™ 3.0 Transplant Module are identical to the PedsQL™ 4.0 Generic Core Scales, with higher scores indicating better HRQOL (see Table 2)(18)

Table 2.

Description of Survey Instruments Used to Measure Patient Symptoms And Quality of Life

Survey Instrument	Publisher	Description	Measures	Score Range
General Health Questionnaire	N/A
Short Form 36	Quality Metric Inc.	Includes 8 subscales which are then used to derive physical and Mental Summary Scores	Generic Health related quality of life	0–100
Short Form 6D		Health Utility using 6 domains from the SF 36	Calculated health utility	0–1
PedsQL™ 4.0 Generic Core Scale	Mapi Research Trust	Developed for use in patients regardless of health status. May be used to compare to healthy groups of children or across different diseases. Now validated for use up to age 25	Generic Health related quality of life	0–100
PedsQL™ Cognitive Functioning Scale	Mapi Research Trust	Developed as part of the Multidimensional Fatigue Scale.	Cognitive functioning	0–100
PedsQL™ Transplant Module Scale	Mapi Research Trust	Developed to include health domains unique to transplant recipients	Health related quality of life specific to solid organ transplant recipients	0–100

Statistical Analysis

All statistical tests were performed using PASW Statistics 12. (SPSS Chicago, IL) The SF 36v2 scores of LT survivors were compared with similar aged long term heart transplant survivors and 18–34 year olds in the 1998 US general population reference group.(6) The SF 6D, PedsQL™ Generic, and PedsQL™ Cognitive Functioning scores were compared to published healthy and chronic disease patient groups and the PedsQL™ Transplant module scores were compared to a cohort of pediatric solid organ transplant recipients. (15, 17–19)

The Students t test was used to compare the mean scores of the different HRQOL measures and health utility scores with their published norms. We also used Cohen’s d to estimate effect sizes for the PedsQL™ total scores and subscales. In cases when a statistical significant difference was not found we examined the minimal clinically important difference (MCID) of the PedsQL™ Generic and the SF 36v2.(12, 20)The MCID is defined as the smallest difference in a score that patients perceive to be beneficial and would mandate, in the absence of troublesome side effects, a change in the patients management. (21) Univariate analysis for risk factors that affect HRQOL and health utility was performed using nonparametric, 2-tailed tests with significance level of P ≤0.05. Multivariate analysis was not done due to our limited sample size

Results

Between July 1988 and December 1992, 178 patients received their primary orthotropic LT at The University of Chicago Hospitals. During the intervening years there were eighty-six deaths and the status of seven foreign patients is unknown. The twenty-year Kaplan-Meier survival of 171 patients who remained in the United States was 49.7%. (See Figure 1) The most common overall indication for transplantation was biliary atresia (64.3%). There were no significant differences between responders, non-responders and those lost to follow up with regards to gender, age at transplant, primary disease, type of transplant and insurance status. (See Figure 2)

Figure 1.

Kaplan Meier Curve of 178 patients transplanted at University of Chicago Hospitals between July 1988–Dec 31^st 1992.

Figure 2.

Characteristics of all eligible Long Term Liver Transplant Survivors n=82. There were no significant differences between the three groups

Clinical and demographic characteristics are shown in Table 3. The median age at LT was 0.86 yrs [IQR 0.58–3.0] and mean age at survey was 23.02±4.4 years. Sixty four percent were attending or had completed some college, and 80% of those over 23 were employed either full time or part time.

Table 3.

Clinical and Demographic characteristics of Long Term LT Survivors

Mean Age and SD (years)	23.02±4.44
Range	18.4–34.6
Over 21 years at time of survey	62%
Caucasian	86%
Female Gender	62%
Marital Status (Single)	78%
Education	College Grad or grad school 24%
	Some college= 38%
	GED= 22%
	9–12th grade= 16%
Full time student	43%
Employed Full or Part time	62%
Living Independently	32%
Government Insurance/Disability	19%
Followed by Hepatologist	57%
Primary diagnosis of Biliary Atresia	68%
Deceased Donor Transplant	62%
Retransplanted	24%
Have been told they may require another liver transplant	14%
Recipient of another solid organ transplant	1 (2.7%) (Kidney)
On the waiting list for another transplant	1 (2.7%) (Kidney)
Complain of medication side effects	54%
Forget to take medications once a week or more	19%
Off all immunosuppressive medication	22%
Self report Hypertension	24%
Currently on Blood Pressure medication	19%
Have had a Liver biopsy within the last 5 years	65%
Self report Autoimmune/Chronic Rejection on Biopsy	16%
Normal or near normal Liver Function Tests	86%
Have had Blood in stool within the last three years	14%
Currently Jaundiced	5%
Have had Hematemesis within the last three years	8%
Hospital admissions for transplant related issues in the past three years	46%

Eleven percent (4/37) of recipients self-reported biopsy proven chronic rejection, thirty percent (11/37) reported taking antihypertensive medication, and one patient had insulin dependent diabetes. One patient had a renal transplant (RT) and one was on the RT waiting list. Side effects attributed to immunosuppressive medications (IS) were reported by 54%, with intermittent abdominal pain being the most common complaint. Eight (22%) patients were no longer taking IS medications.

HRQOL

Results for the PedsQL™ Generic Core Scale, Cognitive Function Scale and Transplant Module are shown in Tables 4, and 5. LT survivors reported significantly lower physical health than healthy controls, but had scores that were equal to young adult cancer survivors and better than young adults with other chronic illnesses. Cognitive functioning was similar to healthy young adults, those with a chronic medical condition and cancer survivors. (Personal Communication Dr. Rhonda Robert, M.D. Anderson Cancer Center June 2, 2011)(15) When compared to a 5–18 yr old solid organ transplant survivor group, these LT survivors reported scores on the PedsQL™ 3.0 Transplant module that were significantly lower in all scales, with large effect sizes noted for the Total Score and How I look category. Table 6 (21) On the SF36v2, there were no statistically significant differences in the subscales, or summary scores compared to healthy controls aged 18–34. However, LT survivors scored one MCID above healthy controls in the subscale of Vitality and one MCID below in the subscale of general health. These results are shown in Figure 3 and compared with a group of young adults who survived ten-years after heart transplantation.(6)

Table 4.

Comparision of PedsQL™ 4.0 Generic Core Scale Results of Long Term LT Survivors with Young Adult Healthy Controls, Young Adults with a Chronic Illness and Young Adult Cancer Survivors

	LONG TERM SURVIVORS N=36	HEALTHY CONTROLS N=1171 (15)		CHRONIC ILLNESS N=102 (15)		ADULT CANCER SURVIVORS N=72
PedsQL™ 4.0 Generic Core Scale	MEAN± SD	MEAN± SD	EFFECT SIZE	MEAN± SD	EFFECT SIZE	MEAN± SD	EFFECT SIZE
Age in Years	23.02±4.44	19.7±1.65		19.7±1.65		23.82±3.79	0.2
Total Score	75.95± 16.50	78.18± 9.20	0.24	70.25±9.20	0.5*	81.80±14.72	0.39
Physical Health	82.29± 20.12	86.25± 10.63	0.36*	74.49±16.07	0.46*	81.33± 17.73	0.05
Psychosocial Health	73.84± 17.7	73.87± 10.53	0	67.99±11.85	0.43*	82.04±14.59	0.53
Emotional Functioning	69.03± 21.67	66.68± 15.00	0.15	60.02±17.30	0.49*	78.37±18.23	0.49
Social Functioning	82.22± 20.99	85.48± 11.90	0.27	82.21±13.10	0	88.50± 17.04	0.34
School Functioning	70.28± 22.96	69.47± 13.94	0.06	61.27±16.72	0.49*	79.24±15.85	0.49

Effect Size is compared to Long Term Survivors

^*p<0.05

Table 5.

Comparison of PedsQL™ 4.0 Cognitive Function Scale with Controls, Adults with a Chronic Illness and Adult Cancer Survivors

PedsQL™ Cognitive Functioning Scale	LONG TERM SURVIVORS N=31	HEALTHY CONTROLS N=1171(15)		CHRONIC ILLNESS N=102(15)		ADULT CANCER SURVIVORS N=72
	MEAN± SD	MEAN± SD	EFFECT SIZE	MEAN± SD	EFFECT SIZE	MEAN± SD	EFFECT SIZE
Age in Years	22.70±4.47	19.7±1.65		19.7±1.65		23.82±3.79
Total Score	69.01± 26.76	70.88± 18.15	0.1	62.89+20.26	0.28	75.93+21.67	0.3

Table 6.

Comparison of PedsQL™ 3.0 Transplant Module Scale with Solid Organ Recipients (Liver, Heart and Kidney)

PedsQL™ 3.0 Transplant Module	LONG TERM SURVIVORS N=32	SOLID ORGAN TRANSPLANT N= 270(21)	Effect Size
	MEAN ±SD	MEAN±SD
About my medicines I	76.63±17.99	83.08±14.98	0.42
About my medicines II	76.94±23.07	86.58±16.19	0.57
My Transplant and Others	65.92±23.71	74.03±19.77	0.40
Pain and Hurt	62.50±25.13	71.10±23.43	0.37
Worry	69.79±22.07	79.44±21.81	0.44
Treatment Anxiety	71.09±23.48	74.77±27.23	0.13
How I look	55.20±26.50	76.26±26.76	0.79
Communication	73.05±25.42	76.84±23.45	0.16
Total Score	69.03±15.72	79.03±14.36	0.69

The PedsQL™ Transplant module is a newly developed HRQOL measure for solid organ transplant recipients in which transplant specific symptoms correlated with lower generic HRQOL(18). Our cohort had lower scores than the published group however these were not age matched and the comparison group were much younger and had fewer years of post transplant follow up

Figure 3.

Scores of Subscales of the SF 36v2 compared with norms for the US General population (Age 18–34) and ten-year heart transplant survivors. The asterisks demonstrate scales where there is a minimal clinically important difference compared to the reference population

* Clinically important difference

BP, body pain; GH, general health; MCS, mental components; MH, mental health; PCS, physical components; PF, physical function; RE, role emotional; RP, role physical; SF, social functioning; VT, vitality; HT, Heart Transplant

Health Utility

Health utility values were comparable to a similarly aged group of young adult cancer survivors who were followed for a mean of 13.5 years (22) but significantly lower compared to healthy adults. (Table 7)

Table 7.

Measured Utility values for Long Term Survivors compared to healthy controls and other chronic disease states.

	Method	Mean Utility± SD	N	P Value	Ref
Long Term LT survivors	SF6D	0.75±0.12	36
• Patients on Immunosuppression		0.73±0.12	28	0.12
• Patients off Immunosuppression		0.81±0.10	8
Controls, healthy adults aged 35+	SF6D	0.80±0.01	311	0.0001	(46)
Young adult survivors of lower extremity bone tumors	HUI3	0.75±0.29	28	1	(41)
Medical Expenditure Panel Survey (MEPS: Age 20–29)	SF6D	0.82±0.18	3662	0.02	(19, 39)
Adolescents with Cystic Fibrosis	SG	0.92±0.15	65	0.0001	(9)
Adolescents with IBD	SG	0.97±0.07	67	0.0001	(10)
Chronic Kidney Disease	SF6D	0.67±0.13	185	0.0007	(47)
Chronic Liver Disease	SF6D	0.71±0.1	88	0.059	(48)
Cirrhosis.	SF6D	0.64±0.2	54	0.0039	(48)
HCV	SF6D	0.67±0.16	41	0.0165	(48)
HBV	SF6D	0.78±0.14	51	0.3	(48)
Cholestatic Liver Disease	SF6D	0.68±0.16	33	0.04	(48)
12 months Post Transplant	SF6D	0.62±0.05	183	0.0001	(49)

Patients off Immunosuppression

Eight patients were not taking any IS medication at the time of the survey. Their data is included in the supplementary material. There was a medium effect size (0.53) in the total score for the PedsQL™ 4.0 Generic Core Scale between the patients on and off IS and a large effect size (0.72) in the SF 36v2 Physical Component Summary Score. PedsQL™ 3.0 Transplant Module scores showed similar results with the largest effect sizes seen in the pain and worry subsections (2.44 and 2.52)

Univariate Analysis

In univariate analysis of patient self reported demographic and medical risk factors Table 8) LT survivors’ health utilities’ correlated with unemployment (p=0.04), hospitalizations (p=0.005), and lower education level (p=0.016). Lower PedsQL™3.0 Transplant Module and PedsQL™ 4.0 Generic Core Scale scores correlated with unemployment (p=0.007, p=0.005) and hospitalizations (p=0.02, p=0.006). Table 9 shows the correlation of Health Utility Scores with elements of the PedsQL™ Generic Core Scale and Transplant Module.

Table 8.

Correlation of Lower HRQOL and Health Utility scores with Patient characteristics

ANOVA	PedsQL™ Generic Total Score (p value)	PedsQL™ Transplant Module Total Score (p value)	Utility (p value)
Diagnosis of Chronic Rejection	NS	NS	NS
Unemployed	0.009	0.006	0.042
Hospital Admissions in the last three years	0.006	0.02	0.005
Diagnosis other than Biliary Atresia	NS	NS	NS
High School education only	0.042	0.06	0.016
Living with parents	NS	NS	NS
In a Relationship	NS	NS	0.047
Uninsured	NS	NS	NS
Over 21years old	NS	NS	NS
On CNI for immunosuppression	NS	NS	NS
Age at Transplant (less than 1)	NS	NS	NS
Deceased Donor Transplant	0.04	NS	NS

NS: Not significant

CNI: Calcineurin Inhibitor

Table 9.

Spearman Correlation of Health Utility Scores with PedsQL™ Scores

CORRELATION	PedsQL™ Generic Physical Score	PedsQL™ Generic Psychosocial Score	PedsQL™ Generic Total Score	PedsQL™ Transplant Total Score
SF 6D	0.55**	0.71*	0.73**	0.48*

^*<0.01

^**<0.001

Discussion

Improving long-term survival rates for pediatric LT has created a growing number of young adults with a unique set of physical, mental and social challenges. We describe a group of young adults who have survived almost two decades after undergoing LT as children. They have achieved a high level of education and employment despite suffering from the side effects of chronic immunosuppression and having lower HRQOL and health utility than their healthy peers. Previous reports have described lower HRQOL in LT survivors, ours is the first attempt to assess health utilities and to characterize clinical and health status aspects that may affect these scores.(3, 23, 24)

Our ten and twenty year survival rates were 57.5% and 49.7%, similar to a recent UCLA report, which included both adults and children.(5) A report of pediatric LT survivors from Pittsburgh using only whole cadaveric allografts demonstrated a 20 year survival of 77% in 166 children transplanted between 1989–1992.(25) The progressive decline in one-year mortality observed over the past twenty years and the declining rate of PTLD and life threatening infections in long-term follow-up should allow a greater number of LT recipients to reach adulthood with their original grafts. Data from The Studies of Pediatric Liver Transplantation (SPLIT) (26) estimates a ten-year graft survival of 80%; our data confirm that many grafts can function satisfactorily for 10years and of those 91%will survive to 18 years or longer. Chronic graft injury and/or fibrosis leading to graft loss may emerge as an important problem in the third and fourth decades after LT. Our data supports this concern with more than 10% of this cohort reporting a likely need for retransplantation.

Medical co morbidities may cause a decline in health status due to increased medication burden, more frequent physician visits and loss of the ability to participate in normal activities in the patient with an otherwise normally functioning graft.(27) Chronic Kidney Disease (CKD) one of the most common side effects of immunosuppressive medication is a growing a problem in the LT population. In the SPLIT database 2.5% of recipients at 5 years post transplant had stage 3 CKD and one underwent renal transplantation at ten years.(26, 28, 29).

Thirty one percent of our cohort were on anti hypertensive medication compared to 20% in a group of ten year survivors from the same era.(30) This is much higher than the 9–13% in five and ten year survivors in the SPLIT data set (26, 29) and suggests that renal disease may be a prevalent problem even though laboratory evidence of this was not collected. One patient received a kidney transplant and a second was on the waiting list. The Pittsburgh group has also reported three patients who required kidney transplants over twenty years of follow up which suggests these patients continue to be at risk for renal complications up to two decades after transplant. (25)

Eight patients (22%) were weaned off or had stopped taking immunosuppressive medication (IS) and did not have evidence of rejection. This is consistent with multiple studies that have reported successful weaning of IS in 15–30% of selected recipients.(31, 32) There was a trend toward higher HRQOL scores in patients who were no longer taking IS with effect sizes in the small to moderate range. Future studies specifically designed to examine the relationship between discontinuation of IS and both generic and disease specific HRQOL will be necessary to confirm this observation.

Long-term survivors reported a college attendance or completion rate of 63% which is greater than 32–50% in previous series (4, 5) and compares favorably with the general US high school population rate of 70%.(33) Given the concerns for academic and cognitive delay and poor school performance demonstrated by the SPLIT Functional Outcomes Group (FOG) collaborative (34, 35), this achievement indicates that cognitive function and learning may recover after LT. Alternatively, there may have been respondent bias toward patients with more successful academic outcomes.

Adult LT recipients report employment levels of 27–50% post transplant (36, 37) Typical predictors of employment post-LT in adults such as previous education and pre-transplant employment status have no bearing in pediatric patients. In our study 80% (9/11) of those over age 23 were employed; this age was used as it would allow college graduates to enter the workforce. This is similar to 82% employment in 20–33 yr olds and 88% employment in the 25–33 yr olds in a report from France.(4) Employment did predict higher HRQOL and Health Utility scores, as previously reported by Kousoulas et al (38)however this analysis was limited given our small sample size.

This is the first report to assess Health Utility in long-term LT survivors. Our results are consistent with young adult pediatric cancer survivors but lower than utilities reported by healthy young adults, adolescents with cystic fibrosis and inflammatory bowel disease.(9, 10, 22, 39) Mean health utility was greater in seven patients who were no longer on IS (0.82 vs. 0.73) Although this exceeds the minimal clinically important difference (MCID) of 0.03 for the SF 6D (40) it was not statistically significant, likely due to our small sample size. In contrast to pediatric bone cancer survivors, health utilities after LT did not differ by gender or age (41) This suggests that despite reporting good HRQOL, patients continue to find their current health state less than desirable and would be willing to undergo some risk, or sacrifice life expectancy for a chance to improve their health status.

Pediatric and adult LT recipients have lower HRQOL in some or all domains when compared to a reference population and similar or better scores than patients with chronic diseases.(3, 36, 42) Using the SF 36, Duffy et al reported significantly lower scores in role–physical, general health, social functioning, mental health and physical component summary in a group of adult and pediatric long-term LT survivors.(5)These findings are similar to our own cohort who demonstrated good HRQOL overall with some deficits in physical function both in the PedsQL™ and SF 36v2. In contrast to children following LT, the young adults in this cohort reported psychosocial outcomes that were similar to a healthy population (3). Predictors of higher HRQOL in our group included higher educational attainment, employment, and fewer hospital admissions. Type of transplant, whether living related or deceased donor, did not have a significant impact.

Limitations

Our study had several limitations. We collected limited clinical data and relied on patient self-reports to provide information regarding lab and biopsy results. Although participants did not differ from non-participants on measured medical and socioeconomic factors at the time of transplant, we cannot rule out the possibility that the groups differed on some other variable that may have impacted our interpretation of the results. Despite our best efforts using multiple approaches, we were unable to track 32.5%(26/80)of survivors. The difficulty we faced locating LT recipients highlights the complexity involved in the transition of care from adolescence to adulthood and from center to center. Where data were unavailable we attempted to make contact using social networking sites and Internet searches.

Our population was geographically diverse, and so we were unable to use the gold standard techniques for eliciting health utilities, which require face-to-face interviews. Instead, we administered multiple measures of HRQOL to augment our findings and to understand the factors associated with differences in patient reported health utilities. The variability in health utilities associated with using different measuring scales underscores the importance of using multiple methods to measure health status and developing transplant specific measures of health status.(43)

Our analysis has identified physical, social and psychological challenges faced by this growing group of transplanted young adults. More comprehensive, prospective follow up of all LT recipients will assist in identifying upcoming challenges and enable the transplant community to better care for future LT survivors.

The proliferation of transplant centers has increased the need for a national registry of long-term LT survivors to continue prospective follow-up and identify the operative, medical, and psychosocial factors that most strongly influence health status after LT.(5) There are numerous blogs, Facebook groups and web pages collecting patient information and helping transplant recipients communicate with each other. (44, 45) A web-based patient registry tool may provide an inexpensive, and secure method for tracking LT recipients and provide a platform to collect patient reported outcomes

Abbreviations

LT

Liver Transplant

SG

Standard Gamble

TTO

Time Trade Off

SF6D

Short Form 6 Dimensions

HUI

Health Utility Index

SF 36

Medical Outcomes Study 36-item short-form health survey

HRQOL

Health Related Quality Of Life

QALY

Quality Adjusted Life Year

CMH

Children’s Memorial Hospital

SSDI

Social Security Death Index

IS

Immunosuppresion