The Long-Term Quality of Life of Living Kidney Donors: A Multicenter Cohort Study

Recruitment of participants

In brief, living kidney donor databases from nine centers in Canada, Scotland and Australia were reviewed, and donors were contacted by telephone for study participation. Across the centers, all donors were medically and psychosocially fit to donate according to accepted standard criteria for donation (11). Kidney donors were eligible for study participation if they were (1) at least 18 years of age at the time of donation, (2) donated a kidney between the years 1970 and 2007, and (3) spoke English. During the period of accrual no paired exchanges or nondirected donations were performed at any of the participating centers.

A comparison group of healthy nondonors was also assembled. Such individuals had to (1) be healthy at the time of the transplant (i.e. no renal disease, hypertension, diabetes, cardiovascular disease, pulmonary disease or active cancer by historical recall), (2) be at least 18 years of age at the time of the transplant and (3) they had to speak English. As a recruitment strategy, individuals for this healthy nondonor group were suggested by our donor participants. A database of potential nondonors was then generated. Donors could identify more than one potential nondonor. In some cases, donors were not able to identify any such an individual.

Participant assessment

All participants provided written, informed consent before taking part in our study. Participant assessments took place between April 2004 and June 2008. Donors and nondonors were assessed in person or at a distance (by mail or telephone), by research personnel not directly associated with the transplant team. As it has been utilized in prior living donor studies, we used the SF36 as one of our standardized quality of life scales. This validated tool measures quality of life based upon eight dimensions. Raw scores range from 0 to 100 with higher scores indicating a better quality of life. Physical and mental component summary (MCS) scores are also generated. The physical component summary (PCS) is an aggregate score of the physical functioning, role physical, bodily pain and general health domains of the SF36. The MCS is an aggregate of the energy/vitality, social functioning, role emotional and mental health domains (12). The average MCS and PCS scores of the general U.S. population are 50.

To add to our standardized quality of life assessment, we also used the 15D™ version 2.0, a 15-dimensional standardized, validated, questionnaire of health-related quality of life, based upon multiattribute utility theory (10). This measure is thought to be more comprehensive than other quality of life scales in that it assesses dimensions including vision and eating along with mobility, hearing, breathing, sleeping, speech, elimination, usual activities, mental function, discomfort and symptoms, depression, distress, vitality and sexual activity (10). Composite scores range from 0 (representing ‘dead’) to 1 (representing ‘ideal health’).

We also utilized the Feeling Thermometer, a visual analogue scale. Participants are able to choose from a continuum of values with scores ranging from 0 (the worst score, representing ‘worst imaginable health state’) to 100 (the best score, representing ‘best imaginable health state’). Furthermore, using 11 author-developed questions, demographic information was collected including participant marital status and their employment and income levels after the transplant. Participants were also asked to indicate if they had seen a psychologist, counselor or psychiatrist since the time of the transplant. We assessed their use of psychotropic medications by analyzing their reported medication list. Donation-related attitudes were investigated by asking donors about the benefit of donation to the recipient, if they felt adequately informed about the risks and benefits of donation around the time of informed consent, and if they would make the same decision to donate again. Questions were pilot tested by a group of experts in living donor issues. We report outcomes according to published STROBE guidelines (strengthening the reporting of observational studies in epidemiology guidelines) (13). Our study was approved by research ethics boards across participating centers.

Statistical methods

We compared donors and healthy nondonors using independent t-tests and Wilcoxon rank sum tests. Our method of identifying nondonors was a recruitment strategy only, and in many cases donors were unable to identify a nondonor participant. Thus, in our primary analysis, donors and controls were considered independent of one another. Based on our power calculations, we were able to detect (with 80% power) differences of 2.72 on the SF36 PCS, 2.80 on the SF36 MCS, 0.26 on the 15D and 4.35 on the Feeling Thermometer, if in truth such differences existed.

Appreciating the possibility of group interdependence, in an additional analysis we calculated the Pearson correlation coefficient of 98 donors and controls who were pairs (i.e. donors who were able to identify a nondonor to participate in the study). There was evidence of some correlation in the SF36 MCS score and some subscales (energy/vitality, social functioning, mental health) (correlation 0.3). Thus a paired dyad analysis was also employed in these instances to examine the consistency of the results with the unpaired primary analysis.

We used linear regression models to determine if SF36 scores and 15D scores differed between donors and nondonors after prespecified adjustment for race, gender, age, relationship to recipient, time since donation and marital status.

Finally, we carried out subgroup analyses using interaction terms to determine whether the effect of donor status (donor vs. control) on the MCS of the SF36 was modified by age at the time of donation (younger or older than the mean age of 43), year of donation (before or after the median year 2000) or relationship to recipient (genetically vs. not genetically related, spousal vs. nonspousal donors). We restricted these analyses to the MCS of the SF36, as it measures the psychological, social and emotional function of an individual, whereas the PCS score is thought to correlate more with one's physical health (12). Within donors, we used linear regression models to determine if SF36 MCS scores differed by surgery type (laparoscopic or open surgery) or recipient outcome (defined by the presence or absence of recipient death or graft failure at any time after donation).

We performed all statistical analyses using SAS version 9.1.3 (SAS Institute Inc., Cary, NC, USA). A p value of less than 0.05 for a test was considered statistically significant.

Description of participants

Across participating centers, there were 1140 donors identified from donor databases. A total of 421 donors were eligible for participation and contactable. A total of 235 donors provided informed consent for study participation.

These donors identified the names of 241 potential nondonor controls. In many cases, these were individuals who had come forward for donor evaluation and were deemed medically and psychosocially fit to donate by the transplant team, but for another reason, had not gone on to donate. In some instances another person with a more optimal HLA match to the recipient was chosen instead. Other reasons for nondonation included logistical issues such as work responsibilities and long travel distances. To our knowledge, none of our nondonors were excluded from donation because of psychosocial unfitness. Of those nondonor controls suggested, 172 were subsequently found to be eligible for the study and contactable. Of these, 114 consented to participate.

Questionnaires were completed by 203 donors and 104 nondonors (86% and 91% of individuals who provided informed consent, respectively) (Table 1). Participants who did not complete the study dropped out for personal reasons or we lost contact with them soon after they signed the initial consent form.

The baseline characteristics of donor and nondonor control groups are presented in Table 2. The characteristics of donors and nondonors were similar, although donors were slightly older than nondonors at the time of transplant and had different relationships to the recipient. Donors and nondonors were assessed an average of 7 (SD 6) and 8 years (SD 6) after the recipient's surgery, respectively. Eighteen percent (n = 36) of donors underwent a laparoscopic nephrectomy.

Recipient outcomes

Most recipient charts were available for review (n = 194). At the time of donor follow-up, 12 (6%) recipients had died and 20 (10%) had grafts which failed. Graft failure occurred 1 day to 14 years after transplant, and recipient death occurred 2 weeks to 13 years after transplant.

Quality of life outcomes

SF36, 15D and Feeling Thermometer scores: In the primary analysis there were no significant differences between donors and nondonors on any of the eight subscales or the physical and MCS scores of the SF36 (p values ranged from 0.33 to 0.98). Their scores were similar to the Canadian population norms (14) (Figure 1). When adjusted for age, gender, race, marital status at time of surgery, relationship to the recipient and year since donation, the groups remained similar (adjusted p values of 0.42–0.56). As we did note some correlations between donor and nondonor pairs in our dyad analysis, an additional analysis was completed, accounting for this correlation. Similar to the primary analysis we found no significant differences between donors and nondonors controls (p values ranged from 0.14 to 0.97).

The mean 15D quality of life score of donors was 0.93 (SD 0.09). Nondonors had a mean score of 0.94 (SD 0.06). There was no significant difference between groups (p = 0.55). When results were adjusted for race, gender, age, relationship to recipient, time since donation and marital status, the difference between groups (0.03, p = 0.55) was not statistically or clinically meaningful (15).

The mean MCS scores of various donor and nondonor subgroups are illustrated in Table 3. The association between donation and the SF36 MCS score was not modified by age, year of donation, or relationship to the recipient (interaction term p values ranged from 0.50 to 0.98). Donors who underwent laparoscopic nephrectomy did not experience a better MCS score on the SF36 than those who underwent open surgery (p = 0.32). Donors whose recipients experienced an adverse outcome did not score differently from those whose recipients faced no adverse outcome (p = 0.74).

Table 3. Mean mental component summary scores (± SD) of donor and nondonor subgroups

	Age ≥ 43	Age <43	Spousal relationship to recipient	Nonspousal relationship to recipient	Donation prior to year 2000	Donation in the year 200 or later	Laparoscopic nephrectomy	Open nephrectomy	Recipient with adverse outcome	Recipient without adverse outcome
Donors	54 ± 8	52 ± 9	54 ± 7	53 ± 9	53 ± 8	54 ± 8	52 ± 9	54 ± 8	53 ± 9	54 ± 8
Nondonors	56 ± 6	52 ± 9	54 ± 14	54 ± 7	N/A	N/A	N/A	N/A	N/A	N/A

• N/A = not applicable to nondonor population.

On the Feeling Thermometer, donors had an average score of 84 (SD 16), and nondonors, an average score of 85 (SD 10). There was no significant difference between groups (p = 0.46).

Donation-related attitudes: Donor thoughts on the overall benefit of donation to the recipient are illustrated in Figure 2. Ninety percent (n = 182) of donors felt that they were adequately informed about the risks and benefits of donation at the time of informed consent. Ninety-seven percent confirmed that they would make the same decision to donate again.

Main study findings

We conducted a multicentered study to compare the quality of life of living kidney donors to a healthy nondonor group using standardized scales and author-developed questions. Overall, our findings are reassuring. On quality of life scales donor scores were similar to nondonors. The marital status and income levels of donors were similar to nondonors in spite of their donation experience. About 1 in 10 donors visited a psychologist or counselor after donation with similar rates in nondonor controls. Our donors generally expressed positive attitudes about their experience.

Comparison with the previous literature

The results of our study are generally consistent with the previous literature. On standardized measures of quality of life including the SF36, past studies demonstrate that donor and nondonor scores tend to be comparable (1,4–8). In our study, we also noted similar SF36 and 15D scores in our donor and nondonor groups. Additionally, in past reports, donor relationships are often unchanged or positively impacted by the donation experience (16–19). We found that a similar number of donors and nondonors remained married after surgery. In the previous literature, on standardized psychological questionnaires including the Beck Depression Index, donors and nondonor scores tended to be comparable (20,21). A similar number of donors and controls in our study were using psychotropic medications and had visited mental health professionals. Also, donors have been noted to demonstrate positive donation related attitudes after their experience. Most would donate again if given the opportunity (5,6,8,17,22,23). This feeling was also expressed by donors in our study.

In previous subgroup analyses, more distant relatives had lower SF36 scores (4), laparoscopic donors demonstrated higher quality of life scores (24, 25) and donors whose recipient died or experienced graft failure had worse SF36 scores (26). In our study no subgroup analyses demonstrated differences that reached statistical significance. However, one must also appreciate that these analyses had adequate statistical power only for large effects.

Strengths and limitations

To our knowledge, our study is one of the first to compare donor outcomes with a healthy internal comparison group. Previous studies have compared donors with the general population, with three studies utilizing only ‘healthy’ controls that were identified from the general population (7,20,27). However, as donors tend to be inherently different from the public, it is possible that beneficial psychosocial outcomes were overestimated.

Our nondonor comparison group was more suitable as it consisted of individuals who were identified by donor participants and possibly had similar values, ideals and interpersonal qualities. Furthermore, they had personal ties with the kidney recipient and in many instances they had been assessed by a transplant professional to be medically and psychosocially fit to donate, but had not done so for another reason. Their demographic characteristics were also generally similar to those in our donor group.

The other strengths of our study include our sampling of a diverse group of participants from multiple countries and centers, as well as our comprehensive quality of life assessment.

There are limitations to our study. Despite our best efforts, recruitment was difficult. It was especially challenging to contact donors from our databases. Since donating, their contact information frequently changed, and in many cases we were not successful in speaking with anyone in their household. We cannot then exclude the possibility that nonparticipants experienced different measurable or unmeasurable outcomes from those who took part in our study, and that our study was biased toward participants who were functioning well.

We also used a questionnaire-based assessment which some individuals and cultures may find limiting. Some participants may have felt restricted in their responses and may have preferred an open-ended assessment where details could be fully described.

Further, the study was retrospective in nature, and some participants may have had a biased recall of past events. Having already gone through the experience, they may have found it difficult to express negative feelings. We also relied upon self-reporting and thus outcomes may have been either over- or underestimated. Additionally, as donors were not assessed prior to their surgery, we could not accurately measure changes in outcomes after donation.

Lastly, our participants were mainly Caucasian, were from a higher socio-economic class and had access to universal health care. They also came from an era when good access to life insurance has been documented (28,29) and when donor reimbursement strategies were not commonplace (30). The effect of these characteristics on our study findings is uncertain, but they may impact our study's applicability to other populations (31).

Practical implications

Our study has several practical implications for transplant centers and for future donors. First, most of our donors underwent a detailed psychosocial assessment with psychologists, psychiatrists or social workers prior to their donation. Their high function after donation supports the utility of these assessments (32,33).

In our study, only 90% of donors felt adequately informed about the donation process. It is difficult to know whether they felt inadequately informed about medical or quality of life outcomes. However, a survey of U.S. and non-U.S. transplant professionals noted that donors, especially with regard to psychosocial outcomes, are often presented variable information, with risks often omitted or inaccurate (2). Based on the results of our study, we can communicate clearer quality of life information to potential donors. We suggest that during the time of informed consent, potential donors be informed that based on structured questionnaires, donors have a similar quality of life as other healthy and motivated individuals. However, potential donors should also be made aware of the possible negative psychosocial outcomes reported in the previous literature, including strain in family relationships, impaired body image secondary to their surgical scar, depression and anxiety (3).

Finally, in our study, donors commented that they appreciated being reassessed by the transplant community. The literature suggests that this is a common desire of donors (17,22,34,35). Although we found no significant impact of recipient outcomes on quality of life, this seems to be especially important for those who have faced adverse recipient outcomes as early intervention and counseling may be facilitated (36).

Future research

In the past literature, some donors scored lower on measures of depression (37), and expressed more feelings of hostility in the early postoperative period (21). Additionally, the postdonation SF36 scores of some donors have been noted to decrease (25,35,38) with some (31%) developing incident episodes of diagnosable psychiatric disorders (35). Thus, future research should include long-term, multicentered, prospective studies, to clarify whether such observations are reproducible. Having participants complete assessments prior to the transplant, and at a variety of time points after the time of their surgery, will help identify any periods of vulnerability for donors.

Future research should also provide the opportunity to study the impact of paired exchange, nondirected donation and donor reimbursement programs on quality of life outcomes. Additionally, as the use of structured psychological scales is limited in the literature, their use in future studies may help to further elucidate any risks to the donor to aid in optimally maintaining their long-term health and well-being.