Perceived Support Among Families Deciding About Organ Donation for Their Loved Ones: Donor Vs Nondonor Next Of Kin

The aims of the survey were twofold: (1) to elicit family members’ accounts of various types of support received and their perceived quality of care for themselves and their loved ones during the time they made the donation decision and (2) to examine the relationship between those factors and the families’ decision about organ donation.

The survey was part of a study undertaken during a 6-year period that covered regions of 6 organ procurement organizations (OPOs) in the Northeast, Midwest, and Midsouth of the United States. The study included 326 families who constitute the population for the present survey. Owing to the large geographical study areas, telephone interviews were selected as the optimal method of data collection. A second reason for choosing this method was to maintain relative anonymity for the respondents. Six individuals with interviewing experience related to sensitive topics were recruited and trained during a 1-day workshop. Aside from using the survey to practice interviewing, training included presentations about the organ donation process, brain death criteria, grief and bereavement, and the protection of human subjects. One Spanish-speaking interviewer was available for participants wishing to be interviewed in Spanish. Interviews lasted an average of 45 minutes. Respondents were offered $50 for their participation. The study was approved by the institutional review boards at the academic institutions affiliated with the participating OPOs.

Potential survey participants received a letter from their OPO 8 to 10 months after the death of their relative informing them about the nature and purpose of the study and that they would be receiving a telephone call from a member of the research staff to ask if they were willing to take part in the study. Within 5 days, calls were made by the interviewers to review the study and its purposes, answer family members’ questions, obtain verbal consent, and if consent was given, to set up a convenient time for the interview. Interviewers asked for a family member who had played a significant role as a decision maker about donating their loved one’s organs. Respondents were told they could end the interview at any time. No one chose to terminate the interview before completion. At the end of the interview, participants were given information about bereavement services in their area.

Of the 326 letters mailed to potential respondents, 44 (13.5%) were returned as undeliverable. Of the remaining 282 persons who were called, 23 (8.2%) could not be located or were not reachable after 3 attempts. Sixty persons (21.3%) refused to participate when contacted. Reasons given by those who refused were similar for donor and nondonor groups and were primarily related to difficulties talking about their relative’s death. A total of 199 persons (70.6% of those who were called) were successfully interviewed. Respondents and nonrespondents were comparable in terms of their relationship to the deceased; most were mothers. Other sociodemographic factors were not available for nonrespondents. The proportion of nondonor next of kin who could not be located or who refused to participate was 36.4%, compared with 21.8% for donor family members. Of the final sample of 199 participants, 154 (77.4%) had consented to donation.

A 65-item structured survey questionnaire was developed by the researchers based on prior research and theoretical work, results from focus groups with donor and non-donor family members,¹¹  recommendations from donor families, and pilot testing. Survey items consisted of close-ended nominal-level questions and 3- and 4-point Likert scales. The survey was pilot tested with mothers of organ donors; the pilot test did not result in any substantive changes. A test-retest was undertaken with 8 persons (staff members of 1 of the participating OPOs) 1 week apart, producing a reliability of 98%.

The questionnaire addressed the following: background characteristics of the respondent and the deceased, perceived support, perceived quality of care for the family and the patient, and perceptions of the decision-making experience. Questions about support were grouped into 3 dimensions of social support: emotional, instrumental, and informational.³³  Emotional support was defined by behaviors such as listening, reassurance, demonstration of understanding, acceptance and nonabandonment, and physical touch; instrumental support was measured by meeting physical needs for comfort; and informational support addressed understandable information about organ donation and brain death, as well as having enough time to make an informed decision. A fourth dimension of social support referred to as “appraisal support” was not addressed in this study. Perceived quality of care was operationalized by perceived demonstration of compassion, respect, and dignity toward family members and toward the patient.

The social support items and their convergent and reliability properties are reported in the Results section.

Univariate analysis was first conducted to provide descriptive statistics for the donation decision as well as the core potential predictors of that decision. Multivariate logistic regression was then used to predict binary donation decisions by using the various predictors in a single equation, along with demographic and other covariates. Although many different hospitals were involved in the study, only 8 hospitals had 5 or more study participants. The analyses included adjustments for clustering due to hospital through the inclusion of relevant dummy variables. The SUDAAN statistical software package, version 10.0, was used for the analysis.³⁴ 

Slightly more than half (55.8 %, n = 111) of the respondents were from the Northeast, 28.6 % (n=57) from the Midsouth, and 15.6 % (n = 31) from the Midwest. As Table 1 shows, close to two-thirds of the participants were women and most were white. About one-third of respondents were either married or widowed. Slightly more than two-thirds had more than 12 years of education and a household income that exceeded $25 000. Respondents were primarily a parent of the deceased, followed by a spouse, an adult child, and a sibling.

Age of the deceased ranged from 1 month to 89 years, with a mean (SD) of 39.3 (19.4) years. Most were between 45 and 54 years of age. Slightly more than half (55.3%, n = 110) were males. Racial/ethnic background corresponded to that of the next of kin. Ninety-eight percent (n = 195) had been hospitalized in an intensive/critical care unit. Length of stay varied from less than 1 day to 28 days, with a mean (SD) of 3.4 (9.4) days. Age was the only characteristic that differed significantly with respect to the deceased’s donation status: donors were younger than nondonors (Table 2).

As noted earlier, a total of 77.4% (n = 154) of the respondents agreed to donate their relative’s organs. Whether consent was given or not, the primary decision maker was most often a parent (45.7%, n = 91), followed by a spouse (31.7%, n = 63), a child (12.6%, n = 25), and a sibling (10.1%, n = 20).

Demographically, donor and nondonor respondents differed only with respect to racial background. Among those who consented, 85.9% (n = 128) were white, 11.4% (n = 17) were African American, and 2.7% (n = 4) represented other racial/ethnic groups. (Data on racial/ethnic background were missing for 7 participants; Figure 1).

Donor and nondonor families differed significantly in their perceptions of how difficult their decision had been, with 60.4% (n = 93) of donors and 42.2% (n = 19) of nondonors stating that the decision had not been difficult at all (χ² = 9.7; P = .02). Conversely, 11.0% (n = 17) of those who consented to donation and 26.7% (n = 12) of those who declined described their decision as “very difficult.” Asked if they would make the same decision again, donor and nondonor next of kin also differed considerably in that 90.3% (n = 139) of donors compared with 57.8% (n = 26) of nondonors answered affirmatively (Figure 2). Nondonors were not only more likely to regret their decision but also expressed more uncertainty about the decision they had made.

Almost half of the respondents (48.2%, n = 96) reported that their loved one had talked about organ donation sometime during their lives. This variable did not affect the families’ decision. Of the entire sample, 14.6% (n = 29) were aware that their relatives had formally stated their donation wish by signing a donor card, their driver’s license, or signing up on a donor registry. Families were more likely to consent to donation when their relative had already formally expressed his or her wish. The same applied when families did not know if their loved ones had stated such wishes (Table 2). Wishes of the deceased were included as a variable in the logistic regression analysis.

Several questions were asked regarding respondents’ accounts of specific aspects of emotional support. Five items were of particular interest, with more than 50% of participants giving positive answers: (1) a staff member or volunteer was present who showed understanding of what they were going through (85.9%, n = 171); (2) a staff member or volunteer was there in case they needed them (83.9%, n = 167); (3) someone was present who listened to them (78.9%, n = 157); (4) a staff member or volunteer let them know that their feelings were normal (68.7%, n = 136); and (5) a volunteer or staff member gave them hope that they could go on (54.8%, n=109). A comparison of donors and nondonors on these support elements showed that significantly higher proportions of donor family members recounted having had someone present who showed understanding, who listened, who was there in case of need, and who gave them hope to go on (Table 3). Differences between donors and nondonors with respect to someone having let them know that their feelings were normal approached statistical significance (P = .05).

The 5 emotional support items were averaged to yield an overall index of emotional support for the logistic regression analyses. The items had an alpha coefficient indexing internal consistency of 0.70.

Instrumental support was assessed by means of the following 2 items: (1) physical necessities in terms of being provided with blankets, shower facilities, or toiletries and (2) food and/or drinks. Significant differences based on donor status were found in the area of physical necessities, with 62.3% (n = 96) of donors vs 40.0% (n = 18) of nondonors responding that they had been given such support (Table 3). No differences were found between groups with respect to being provided with nutritional items.

The physical comfort items showed a correlation of 0.30 (based on a phi coefficient) and were averaged to form an index of instrumental support.

At first, 2 general questions were asked of families regarding informational support: whether they were informed about the status of their loved ones and what to expect related to treatments and/or prognosis, and whether someone had been available to ask questions of hospital and/or OPO staff for them. In response to the former, 87.4% (n = 174) of the respondents gave an affirmative answer. Having had a friend or relative present who asked questions of staff was mentioned by 39.7% (n = 79). These 2 variables were not significantly associated with the families’ donation decision.

Two key and specific elements of informational support were then assessed: information about brain death and information about organ donation.

Information About Brain Death: Notable differences based on donor status were found regarding information on brain death, with 79.2% (n = 122) of donors and 57.8% (n = 26) of nondonors stating the information had been “quite understandable,” and 73.4% (n = 113) vs 48.9% (n = 22) asserting they had been given adequate time to understand brain death before considering organ donation. On a scale of 1 = “not at all understandable,” 2 = “somewhat understandable,” and 3 = “quite understandable,” means were 2.76 for donors vs 2.47 for nondonors for understandability of the information and 2.68 for donors vs 2.07 for nondonors for having enough time to understand befre making a decision (Table 3). Although only 6.5% (n = 10) of consenting next of kin felt that the amount of time for decision making was “not at all adequate,” the corresponding proportion among nondonors was 42.2% (n = 19; χ² = 35.4; P < .001).

The two informational support items had a correlation of 0.46 and were averaged to form a general index of provision of adequate information about brain death for the logistic regression analysis.

Information About Organ Donation: In response to the question regarding the adequacy of information about organ donation in order to make an informed decision and the extent to which this information was perceived as understandable, donors were significantly more likely than nondonors to report that the information had been adequate and understandable. Proportions were 90.0% (n = 137) for donors vs 42.2% (n = 19) for nondonors and 90.3% (n = 139) for donors vs 42.2% (n = 19) for nondonors, respectively. Mean scores (range = 1–3) were 2.91 for donors vs 2.22 for nondonors for understandability of the information and 2.85 for donors vs 2.07 for nondonors for having been given adequate information to make an informed decision (Table 3).

The 2 informational items showed a correlation of 0.77 and were averaged to form an index of provision of adequate information about organ donation for the logistic regression analyses.

Families’ perceptions of quality of care for their loved ones was assessed by the extent to which they felt that the patient was treated with compassion, dignity, and respect. Donor and nondonor family members differed significantly in their perceptions of compassionate care for their relatives. On a scale of 1 to 4 (with 4 being the most positive rating), mean scores were 3.81 for donor family members and 3.53 for nondonor family members (Table 3). Perceptions of the deceased having been treated with dignity and respect did not differ significantly according to the donation decision.

With regard to quality of care for themselves, family members who consented to donation more commonly felt that they had been treated with dignity and respect (means = 3.78 for donor family members vs 3.49 for nondonor family members) and that their religious/spiritual beliefs were respected (means = 3.72 for donor family members vs 3.40 for nondonor family members; Table 3). The 4 quality-of-care items were averaged to yield an overall index of perceived quality of care, yielding an alpha coefficient of 0.87.

Table 4 presents intercorrelations between the 3 dimensions of support and the donation decision. The correlations reported are point biserial correlations. All of the support dimensions were significantly correlated with the donation decision. Informational support about organ donation had the highest correlation (r = 0.55, P = .01), followed by informational support about brain death (r = 0.35, P = .01), emotional support (r = 0.29, P = .01) and instrumental support (r = 0.19, P = .01). When the variables were entered simultaneously into the logistic regression with the demographic (characteristics of respondents and the deceased), and hospital covariates, only informational support about organ donation yielded a statistically significant logistic coefficient (coefficient = 2.68, exponent of coefficient = 14.61, 95% confidence interval of exponent = 5.14–41.51).

To further elucidate this relationship, we calculated observed conditional probabilities of consent to donate as a function of informational support about organ donation (scale range = 1–3) by computing the proportion of respondents who consented to donate at a given score on the informational support scale. For those with a score of 1 (n = 9), the proportion donating was 0%; for those with a score of 1.5 (n = 11), the proportion was 45%; for those with a score of 2.0 (n=14), the proportion was 29%; for those with a score of 2.5 (n = 17), the proportion was 82%; and for those with a score of 3.0 (n = 148), the proportion was 89%. The data indicate a dramatic decrease in decisions to donate at scores less than 2.5. Perceived quality of care and the known prior wishes of the deceased to be a donor did not yield statistically significant logistic regression coefficients.

Families who are approached about organ donation for their relatives have to make their decision during profoundly stressful circumstances. Findings from this study advance our understanding of families’ experiences with and perceptions of particular elements of emotional, informational, and instrumental support and quality of care during these traumatic circumstances. Moreover, the analysis illuminates the association between the particular aspects of support and quality of care on the one hand and the families’ decision about donation on the other.

Meeting families’ support needs during the donation decision-making process is a challenge as well as an obligation for staff caring for eligible organ donors. Yet critical care practitioners often have inadequate communication skills and knowledge related to organ donation and brain death.^22,29,35,36  As noted earlier, they may also have ambivalent feelings about such issues. In combination, these factors have the potential to influence staff confidence and comfort and, in turn, prevent them from providing effective support to families and from facilitating the decision-making process.

Our multivariate analysis showed that the strongest predictor of consent to donation is informational support about organ donation. Previous research has shown families’ emphasis on their desire that such factual information be conveyed with compassion and sensitivity, suggesting a convergence of informational and emotional support.^11,13  The relationship between families’ consent, receiving understandable information about organ donation and brain death, and having been given sufficient time to make their decision mirrors the results of earlier research. The importance of family decision makers being provided with adequate informational support is underscored by longitudinal studies revealing that unanswered questions about aspects of the organ donation process complicated families’ grief.^13,17 

Regardless of their decision about donation, slightly more than one-third of participants reported having had a relative or friend act as a messenger between them and hospital or OPO staff. As Thelen³⁷  notes, nurses often serve as the intermediaries between families and physicians in end-of-life situations. The importance of this nursing role is accentuated when the issue of organ donation requires information be added to end-of-life conversations with next of kin. Yet various constraints limit nurses in fulfilling this role on a continuing basis, necessitating the involvement of informal caregivers.

The notable differences found between donors and nondonors in the domain of emotional support related to 4 particular elements: having had someone there to listen to them, having had someone there to show understanding, having had someone there to give them hope to go on, and having had someone there just in case of need. As noted earlier, previous research has addressed the emotional turmoil experienced by families during the donation decision process. However, only 1 empirical study³⁸  was found—and it was limited to parents of brain-dead children—that recommends particular staff behaviors that may address parents’ emotional needs. Such behaviors included giving parents the opportunity to be at their child’s bedside and to say goodbye. The authors’ other recommendations were stated in general terms related to providing emotional support and developing trusting relationships with patients’ families. Our findings move this area of inquiry forward by identifying more precise elements of emotional support and their association with consent to donation. Moreover, our findings bring into focus how hospital staff can respond optimally to families’ emotional needs and through that process facilitate the families’ decision making. Similarly, our results suggest that specific instrumental support in the form of physical comfort—also amenable to staff intervention—is associated with consent to donation.

The significant differences between donor and nondonor families’ perceptions of quality of care in terms of compassion, respect, and dignity for their loved ones and themselves warrant attention from the perspective of medical and nursing practice and education. How can these elements of quality of care translate into specific staff behaviors and be incorporated into skills training for relevant staff? We believe that principles of bereavement care—a central tenet of palliative care—are applicable in this context for staff training about providing support to families of brain-dead patients as well as support strategies for staff.³⁹ 

Evaluation of the aforementioned differences between donor and nondonor families sheds light on the nature of the decision-making experience itself. Nondonors were considerably more likely to describe their decision as “very difficult” and to state that they would make a different decision if in the same situation again. They also tended to express more uncertainty about their decision. The latter findings, which are consistent with those of Rodrigue et al,⁴⁰  highlight what may have been lost opportunities in obtaining consent from families and underscore the need for careful examination by OPO and hospital staff. Such examination should address strategies to reduce the potential for regrets and uncertainties among families by optimally meeting their needs for support.

Limitations of this study pertain to its retrospective nature, relying on respondents’ memory of past events. Respondents’ recall of events and facts may have been inaccurate. Another limitation is that nonrespondents could have differed from respondents on key background characteristics, thus weakening the generalizability of the study findings. Furthermore, the sample had limited racial and ethnic diversity. Another limitation pertains to the fact that some survey items were generated in a pilot study undertaken with mothers of donors only.

Our findings give us a deeper and more nuanced understanding of families’ experiences with support and quality of care and of the particular elements most predictive of consent to donation than was available from previous research. Gaining new insights into the “texture” of informational, emotional, and instrumental support and quality care helps us define modifiable elements of care that can be incorporated into professional educational programs. Innovative programs with simulated training and role playing constitute recent advances in this area.^41,42  Such educational interventions must comprise components about support of and strategies for critical care nurses, in particular, to cope with the stress they may experience when caring for prospective organ donors and their families. We argue that such educational programs not only will benefit bereaved families facing the donation decision and critical care staff caring for them and their loved ones but, ultimately, will improve the health of the public as a whole through increased rates of organ donation and transplantation.

We extend our thanks to the survey participants for their courage and generosity and thank the interviewers for their skillful interviewing. We gratefully acknowledge the cooperation of the participating OPOs and the assistance of all project staff and volunteers who made the study possible.