Living donor liver transplantation (LDLT) was developed to overcome the shortage of donor organs available for children1 and has been widely applied over the past 15 yr.2 Adult LDLT has also shown a remarkable growth, approximately 10% of all liver transplants in the United States are LDLTs.3 In Germany, 78 (8%) of 975 liver transplantations performed in 2005 were pediatric and adult living liver donations.4
Along with the problem of scarcity of deceased donations, the ethical discussion about the possible harm to a healthy donor still continues. Possible psychological stress on the donors that may arise in the family system has been discussed since the beginnings of living donations of kidneys.5 In the liver transplantation setting, urgent interventions are often on the agenda because alternative therapy methods for liver failure are lacking. However, some authors deem the setting of acute liver failure (ALF) to be a “strong relative contraindication” for adult LDLT2 because the ability of a donor to consent in this setting may be restricted by the circumstances associated with the impending death of the recipient. Otte6 reported that they do not take a living donation for a child in fulminant liver failure into consideration because they believe the psychological burden of the decision-making in this situation to be unacceptable. For Reding,7 living parental donor liver transplantation in fulminant hepatic failure is only to be considered ethical for cases in which access to an efficient emergency postmortem organ-sharing system is not possible.
In contrast to these authors, Malago et al.8 outlined that the basic medical ethical principles of autonomy and utility allow the extension of LDLT indications to patients not suitable for transplantation with deceased grafts. They maintained that the broadening of indications is appropriate only in centers with adequate surgical experience. After implementing statistical models, Durand et al.9 came to the conclusion that the sickest patients and those at high risk are those who benefit the most from LDLT. Their results show that the ethical principles of utilitarianism and beneficiency10 are preserved in the urgent LDLT setting.
Another entity that has been advocated as an indication for expeditious LDLT is hepatocellular carcinoma (HCC), because of the risk of tumor progression. Since the prevalence of hepatitis C and hepatitis C–related HCC has increased and waiting list mortality remains unchanged for patients with HCC, it has been discussed whether criteria for LDLT should be extended to include tumors that are beyond the Milan criteria.11, 12 Hiatt et al.11 dismissed extended LDLT in the context of HCC, because they see an additional psychological burden for the donor, which may lead to coercion.
In our centers, donors are evaluated for LDLT within days (in HCC) or hours (ALF); knowing that this urgent preparation may give rise to psychological distress, the expansion of the selection criteria may have a negative influence on outcome,13 and since the donor's psychological outcome depends on the well-being of the recipient,14 the psychological strain of the donor gains specific importance.
The aim of the present study was to investigate the psychological impact on the potential donors during the evaluation in urgent indications for LDLT and 3 months after transplantation. We explored the following hypotheses:
Comparing subgroups of donors with nonurgent, HCC, or ALF indications, a significant difference regarding anxiety, depression, and mental quality of life (QOL) at the time of donor selection (t0) is expected.
At 3 months after transplantation (t1), the psychological distress measured in anxiety, depression, and mental QOL is significantly reduced.
LDLT, living donor liver transplantation; ALF, acute liver failure; HCC, hepatocellular carcinoma; QOL, quality of life; SF-36, 36-Item Short-Form Health Survey; t, t-value; df, degrees of freedom; F, F-value; r, correlation coefficient; r2, coefficient of determination; ANOVA, analysis of variance.
PATIENTS AND METHODS
The Frame of the Donor Evaluation
For ALF all procedures are finished within 48 hours; in case of HCC, up to a maximum of 2 weeks. Patients were staged for listing according to the status code of medical urgency of the Organ Procurement of Transplantation Network Classification.4 Degrees of severity of liver disease were classified according to Child-Turcotte-Pugh scores. HCC patients did not have higher priority in listing.15
The psychological selection procedure is an integral part of donor evaluation and is based on 2 steps, the aim being to establish a frame for an autonomous decision-making process and protection of the donor.16, 17 The objectives of the first step are verification of informed consent and assessment of mental stability. In the second-step the dynamics of the relationship between donor and recipient are investigated and the transplantation procedures are anticipated with the recipient and the donor.18 In the framework of ALF, second-step interviews with the recipients were dropped, because they were not able to participate. In cases of ALF or HCC chances of recovery of the recipients were discussed in great detail with the donors, both in the surgeons' session of information disclosure and in the psychosomatic interview. All donors were informed about the possibility of declining donation at any step of the procedures by a medical excuse. Psychotherapeutic support was offered if demanded by donors or recipients.
After the procedures in our institutions all donors were presented to an external ethical board of the local Medical Association in an urgency meeting. This committee once more interviewed the donors and evaluated whether the decision to donate has been made without any external pressure.
Design of the Study
We have planned a prospective, longitudinal (5 time points), multicenter study at the Transplantation Centers of the University Hospitals Essen and Hamburg and the results of the first 123 donor candidates at 2 time points, i.e., during the donor selection (t0) and 3 months after transplantation (t1) are presented. All protocols of the survey were reviewed and approved by the Medical Faculty's Ethical Committee.
Hospital Anxiety and Depression Scale, German Version19
The Hospital Anxiety and Depression Scale20 is an internationally widely used reliable and valid instrument for the assessment of anxiety and depression among medically ill patients. This questionnaire was applied because it measures only mental and not physical symptoms of depression, an overlapping of bodily complaints could therefore be excluded. Each scale is made up of 7 items, which are summed up to a score ranging from 0 to 21. Scores lower than 7 are considered clinically not relevant.
Thirty-Six-Item Short-Form Health Survey, German Version21
The 36-Item Short-Form Health Survey (SF-36)22 is a self-assessment questionnaire measuring psychological, physical, and social aspects of QOL and is often used in the field of transplantation medicine. The instrument is considered reliable and valid. It contains 36 items, which are combined into 8 scales, out of which 2 summary scales can be computed: “physical composite score” and “mental composite score.” Each of the 8 scales results in a score ranging from 0 to 100 with a higher score indicating a better QOL. For the summary scales, T-scores with a mean of 50 and a standard deviation of 10 are computed.
Data were analyzed using the Statistical Package for Social Sciences, version 13.0 (SPSS, Chicago, IL). For descriptive analysis, data were expressed as mean values and standard deviation. A comparison with the normative sample for the SF-36 and with a healthy control group for the Hospital Anxiety and Depression Scale was conducted through 1-sample t-tests. Mean differences of continuous variables among subgroups were compared by analysis of variance (ANOVA procedure, followed by Scheffé-test) and t-tests according to group size, level of scales, and distribution parameters. Paired-samples t-tests and Wilcoxon-tests were conducted to compare the difference of the mean of 2 paired samples. Spearman correlations were calculated to test the linear relationship between variables. For all tests, a significance level of P < 0.05 was predetermined.
Sample at the Time Point of Donor Evaluation
From February 2004 to April 2006, 182 donors were evaluated in both centers for an adult or pediatric LDLT. Included were candidates who had a good command of the German language and a permanent residence status in Germany. A total of 123 potential donors (68%) for 100 recipients fulfilled the inclusion criteria and consented to participate in the study.
Sample Following Transplantation
All donors whose respective recipients had undergone transplantation (living or deceased donation) 3 months before were included at the second time point (April 2006). This was the case for 98 donors; 49 donors (actual donors) had donated, and for 49 donors (potential donors) the recipient was given a living donation from another family member or a deceased donation. The response rate was 59.2% (n = 29) for actual and 53.1% (n = 26) for potential donors at that point in time. Therefore, psychometric data were available for 55 donors (Fig. 1).
According to the urgency degree of indication, recipients were divided into 3 groups: nonurgent indications (n = 51); indications for HCC patients (n = 30); or indications for recipients in ALF (n = 19), including fulminant liver failures and cases of acute deterioration in chronic liver failure: “acute on chronic.” A total of 61, 40, and 22 donors for nonurgent, HCC, or ALF indications, respectively, were included in the study. Recipients' demographics and diagnoses are depicted in Table 1; donors' demographics and relationship to the recipient as well as subgroups concerning recipients' diagnoses are shown in Table 2.
Table 1. Recipient Demographics and Diagnoses
Recipients (n = 100)
Children (mean, range)
34 (2.71, 0-18)
Adults (mean, range)
66 (59.99, 24-68)
Nonurgent indications (n = 51)
Primary sclerosing cholangitis
Cirrhosis of unknown causes
Alcoholic liver disease
Primary biliary cirrhosis
Toxic liver damage
Multiple liver hemangiomas
Tumor indication (n = 30)
HCC within Milan criteria
HCC outside Milan criteria
HCC and posthepatitis cirrhosis
HCC normal liver
HCC and alcoholic liver disease
HCC and cirrhosis of unknown causes
HCC and hemochromatosis
HCC and primary sclerosing cholangitis
ALF (; n = 19)
Fulminant hepatic failure (n = 7)
“Acute on chronic” (n = 11)
Alcoholic liver disease
Alcoholic liver disease and Caroli syndrome
Primary nonfunction after liver transplantation (n = 1)
Table 2. Donor Characteristics
Donors (n = 123)
34.98 ± 8.43
Parent for child (%)
Child for parent (%)
Donor for other relatives (%)
Nonrelated donors (%)
Nonurgent indication (%)
HCC indication (%)
ALF indication (%)
As donor surgery for pediatric recipients is less extensive, we have investigated group differences for donors of pediatric and adult recipients. There were no group differences at the time point of donor evaluation regarding mental QOL (t = −0.938; degrees of freedom [df] = 122; P = 0.350), anxiety (t = 0.834; df = 122; P = 0.389), and depression (t = 0.764; df = 121; P = 0.457). Also after transplantation group differences remained without statistical significance for mental QOL (P = 0.865), anxiety (P = 0.756), and depression (P = 0.071). Therefore, we report on both groups as a whole.
Health-Related QOL and Psychological Well-Being of the Donors During Donor Evaluation
Regarding the whole sample, the overall score for physical QOL was significantly higher (t = 20.15; df = 122; P < 0.001), but mental QOL was lower than in the German normative sample (t = −2.79; df = 122; P = 0.006) (Table 3; Fig. 2)
Table 3. SF-36 and HADS Scores
Donors at time of evaluation (t0) n = 123
Donors 3 months after the transplantation (t1) n = 55
German normative sample
NOTE: ↓ In comparison with the German normative sample donors have a significant lower mental quality of life (P < 0.05); ↑ In comparison with the German normative sample donors have a significant higher physical quality of life (P < 0.05).
Abbreviation: HADS, hospital anxiety and depression scale, M, mean value; s, standard deviation.
In comparison with the German normative sample donors have significant lower anxiety scores (P < 0.05).
In comparison with the German normative sample donors have significant higher anxiety scores (P < 0.05).
In comparison with the German normative sample donors have significant lower depression scores (P < 0.05).
In comparison with the German normative sample donors have significant higher depression scores (P < 0.05).
Donors who were evaluated for nonurgent recipients showed the highest scores for mental QOL and donors for ALF recipients the lowest (Table 3; Fig. 2). There were significant group differences of the mean values for the SF-36 mental composite score (F = 8.36; df = 122; P < 0.001). Group differences were particularly explicit between the groups “nonurgent” and ALF (P = 0.001). The subgroups “nonurgent” and HCC (P = 0.069), as well as HCC and ALF (P = 0.171) did not differ from each other significantly.
In comparison with the German normative sample, the “nonurgent” indications group exhibited a mean value in the normal range (t = 0.63; df = 60; P = 0.529). The groups HCC (t = −2.73; df = 39; P = 0.010) and ALF (t = −2.86; df = 21; P = 0.009) showed significantly lower scores of mental QOL in comparison with the German normative sample. The low mental QOL for the whole sample (n = 123) was due to the scores of the subgroups HCC and ALF. There was a negative correlation between the degree of urgency and mental QOL (correlation = −0.35; P < 0.001).
Regarding the physical QOL, no significant group differences could be ascertained (F = 0.32; df = 122; P = 0.729).
Analysis of QOL on a Subscale Level
Significant differences between the indication groups could be established concerning the subscales “bodily pain” (F = 5.30; df = 122; P = 0.006), “vitality” (F = 4.97; df = 122; P = 0.008), “social functioning” (F = 5.65; df = 122; P = 0.005), and “emotional well-being” (F = 12.26; df = 122; P < 0.001). ALF patients showed the lowest QOL scores in all of those subscales (Fig. 2).
Depression and Anxiety
The mean values of the whole sample for anxiety and depression, measured with the Hospital Anxiety and Depression Scale-D, were within the range of a healthy control group (t = −1.48; df = 121; P = 0.141 vs. t = −0.14; df = 122; P = 0.889; Table 3, Fig. 3). Significant differences were found between the indication groups for anxiety (F = 11.92; df = 121; P < 0.001) as well as for depression (F = 15.73; df = 122; P < 0.001).
There were significant differences between “nonurgent” and HCC (P = 0.019) as well as “nonurgent” and ALF groups (P < 0.001) concerning anxiety. The HCC subgroup did not differ significantly from the ALF subgroup (P = 0.104).
There were significant group differences between the “nonurgent” and ALF subgroups (P < 0.001) as well as between the HCC and ALF group (P < 0.001) concerning depression. “Nonurgent” and HCC indications did not differ significantly (P = 0.349).
Compared to healthy controls, donors with “nonurgent” indications exhibited significantly lower scores for anxiety (t = −4.96; df = 60; P < 0.001) and depression (t = −3.31; df = 60; P = 0.00) and the HCC group stayed in the range of the healthy control group. Donors for ALF indications showed significantly higher scores for anxiety (t = 2.08; df = 21; P = 0.050) and depression (t = 3.30; df = 21; P = 0.003) than healthy controls. Anxiety and depression were significantly correlated with the degree of urgency (r = 0.40; P < 0.001). The ALF group showed the highest and the “nonurgent” group the lowest scores for anxiety and depression.
Changes in QOL and Psychological Well-Being 3 Months After LDLT
Donors' psychiatric complications following transplantation
A total of 9 patients contacted the transplant team because of psychological problems and fulfilled the clinical diagnostic criteria for depressive adjustment disorders or moderate episodes of depression and were given psychotherapeutic support, pharmacological treatment was necessary in only 1 case.
Changes in QOL
Compared to the time of evaluation, the physical composite score of QOL decreased significantly (t = −2.53; df = 53; P = 0.014). The decrease was due to donors who had hepatectomy (t = −2.68; df = 27; P = 0.013). Regarding the pre-post comparison of the whole group, the mental composite score of QOL did not change significantly (t = 0.622; df = 55; P = 0.537) but was within the range of the German normative sample at the second time point (t = −1.90; df = 55; P = 0.063). In the formerly burdened group with the ALF indications (n = 15), mental QOL improved (z = 0.079; P = 0.079) and was in the range of the normative sample (t = −1.23; df = 15; P = 0.236) (Fig. 4). The HCC group (n = 15) did not change significantly (z = −.79; P = 0.43), but showed scores in the norm range (t = −0.512; df = 14; P = 0.617).
Changes in Anxiety and Depression
The mean values for anxiety (t = 2.03; df = 54; P = 0.048) and depression (t = 2.91; df = 54; P = 0.005) decreased significantly for the 55 donors as a whole group. For the formerly burdened group of ALF donors, depression (z = −2.18; P = 0.029) and anxiety (z = −2.76; P = 0.006) scores also decreased significantly.
For the whole sample, symptom strain was comparable to a healthy control group for depression and lower than in healthy controls for anxiety (t = −2.81; df = 54; P = 0.007) at the second time point. For the ALF group, anxiety and depression were within the normal range. In a pre-post comparison of the subgroups, the donors with ALF indications demonstrated the strongest gain in psychological well-being (Fig. 4), followed by donors for HCC.
However, this psychometric development was not dependent on having experienced a living donation. No significant differences between potential and actual donors could be found in the mental composite score of the SF-36 (t = −0.508; df = 53; P = 0.614) or Hospital Anxiety and Depression Scale scores for anxiety (t = 1.043; df = 53; P = 0.302) and for depression (t = −0.532; df = 53; P = 0. 597).
Effect of the Recipient Outcome on Donors' Well-Being 3 Months After Transplantation
Recipients were subdivided into 4 groups according to the Clavien Classification (Clavien, 1994)23, grades 1a and 1b as well as 2a and 2b were united in single categories 1 or 2, respectively. Grade 1 (n = 50) includes non-life-threatening moderate problems. Grade 2 (n = 32) includes potentially life-threatening complications without residual disease or disability. None of the recipients experienced complications leading to long-term disability (grade 3). Grade 4a (n = 4) includes retransplantation and grade 4b (n = 14) death in the perioperative period.
At 3 months after the transplantation, the severity of recipients complications was correlated significantly to higher depression (r = .28; P = .038; n = 55) and anxiety scores (r = .36; P = .006; n = 55) and lower scores of mental quality of (r = −.36; P = .007; n = 55) in donors.
We present the largest consecutive group of pediatric and adult living liver donors reported to date. As donor surgery for pediatric recipients is less extensive, it might be expected that donors for pediatric recipients are prone to lower distress. Parents usually experience no ambivalence in their decision-making and the motivation to affect the child's suffering may be higher than in other constellations of living donation, which may also be factors protecting from distress. On the other hand, parents' concern for the well-being of their children might be high. Putatively, these factors counterbalanced one another and the perceived emotional distress therefore showed no significant difference for the pediatric and adult donors in our survey.
During evaluation prior to LDLT, mental QOL in this sample was lower in comparison with normal values. This result is not congruent with previous reports that showed scores for mental QOL (SF-36) higher24 or similar to the general population.25 However, it has to be considered that in our sample only one-half of the donors were in evaluation for nonurgent indications (61 donors, 49%), the other half (62 patients, 51%) were in evaluation for recipients with HCC or ALF indications. The comparatively low QOL is due to the high degrees of distress that HCC and ALF donors perceived. Furthermore, elevated mood and lack of physical complaints have been interpreted as a reaction to social expectations.25 We report on an advanced period of LDLT in our centers. After familiarization of professionals (physicians, transplant coordinators, etc.) with the procedure, which leads to a more open and competent handling of the donors, the self-report of the donor candidates was assumed to be no longer oriented toward social expectations.
However, even if no organ donation through a relative is planned, family members of a transplant patient are exposed to severe mental stress. It has been demonstrated within the context of heart transplantations that relatives will support the patient during the waiting period while at the same time putting their own needs aside.26, 27
At the time of donor evaluation, there is a correlation between the degree of urgency of recipients' indication and donors' perceived QOL. The psychological well-being of the donor depends on the health condition of the recipient: The more urgent the indication, the more serious the threat is for the recipient to die and the threat of losing him/her for the donor. Furthermore, the donor for an urgent recipient is under pressure to decide quickly and to consent to a transplantation with limited outcome chances. The correlation between higher degrees of urgency and reduced mental QOL is therefore comprehensible.
The lower mental QOL may be explicitly ascribed to the weighting of the subscales vitality, social functioning, and emotional well-being. In physical QOL, the ALF group showed significantly lower QOL scores in the subscale “bodily pain,” i.e., they experienced more pain. Knowing that liver donors are selected for excellent bodily functioning, a possible interpretation of this finding is that donors develop somatization as a psychological defense mechanism. Limb pain28 and abdominal pain as a somatoform disorder29, 30 have been reported after adult LDLT. Transplantation surgeons should be aware of the occurrence of somatoform pain disorders and feelings of reduced vitality after adult LDLT. This result suggests a close-meshed and comprehensive psychological follow-up for the donors of the groups at risk to allow an intervention when needed.
During evaluation, anxiety and depression are not elevated for the whole group but solely for donors in HCC and ALF indications. This is an expected result, as discussed above. According to their psychological distress, HCC donors score between nonurgent and ALF donors. Therefore, our first hypothesis that there are significant differences between the above-mentioned indication groups regarding anxiety, depression, and mental QOL at the time of donor selection could be confirmed.
At 3 months after transplantation, only 53.1% (potential) to 59.2% (actual) donors participated in our survey. In a previous study,17 we experienced that the response rates for survey questionnaires were 64% and 30% for actual and potential donors, respectively. We assume that some donors would like to keep the donation experience private. Nevertheless, we cannot exclude that some nonrespondent donors are dissatisfied with the results of LDLT or even have psychological problems because of it.
At 3 months after transplantation, in contrast to an own retrospective study,17 we could ascertain a significant impact of recipient outcome on donors' well-being following donation. But the mental QOL is again in the normal range for all subgroups. Furthermore, depression and anxiety are also comparable to healthy controls for the HCC and ALF donors. In previous studies, higher31–34 or similar scores of mental QOL have been established after LDLT.14, 34, 35 Satisfaction about having saved the life of the loved family member or at least having tried may intensify the perceived self-efficacy, relieve stress, and increase the mental QOL. Verbesey et al.24 observed that mental QOL stayed stable during evaluation and after donor surgery. As discussed above, our sample starts with mental distress above normal range that is normalized after LDLT. Therefore, the elevated distress is of temporary nature.
Returning to our introductory question, as to whether the above reported psychological burden on the donors is justifiable, some ethical arguments for LDLT should be recalled: Spital37 claimed that risks for the donor should be balanced with his benefits and not with those of the recipient. The most important benefit for the donor is to help save a loved family member's life. However, it is not only the recipient's well-being that is at the focus of attention but also the donor's genuine interests. A study on the expectations of kidney donors38 showed that mothers who had donated a kidney to their child did not expect any changes other than stabilizing their child's health. Wives as donors had further expectations concerning their marital relationship that were related to an improvement in everyday life, dialysis no longer being necessary. Both groups, therefore, had their own benefit from living donation and were satisfied with the results.
The most important result of this study is that mental QOL was lower for both HCC and ALF groups; anxiety and depression were higher for the donors for ALF patients during donor selection but 3 months after transplantation both groups were within the normal range regarding QOL, anxiety, and depression. Our second hypothesis could be confirmed for the whole group and for the ALF subgroup regarding anxiety and depression. Changes in mental QOL were not significant at that point in time. Surprisingly, the most burdened ALF donors showed the largest improvement after transplantation. Thus, the ethical criterion of the utility of LDLT for the donor himself claimed by Spital37 may be looked upon as fulfilled. On the other hand, the most relevant ethical consideration is the utility for the recipient himself; the outcome of LDLT in urgent settings must therefore also be compared to the outcome of deceased donation.
Our sample for the second time point is relatively small because some LDLTs were not carried out (Fig. 1) and the response rate was low. Furthermore, our findings need replication.
The psychological burden and benefit of LDLT are important criteria for its legitimacy. In employing standardized questionnaires and regular psychosomatic follow-ups, psychosocial outcome of LDLT may be compared in different conditions and centers. Our survey demonstrated that LDLT is not emotionally detrimental for donors in urgent conditions and therefore may be offered to patients in this setting as well.