Sample description

The demographic and donor-specific characteristics of LLDs are shown in Table 1. The mean age of the participants was 37.9 ± 11.4 years, ranging from 19 to 63 years. Among the 124 participants, 56.5% were male; 72.6% had a bachelor’s degree or higher; and 62.9% had a monthly income of less than 3.5 million Korean won. Majority of the LLDs were children of the recipient (71%). The mean of length of hospital stay was 9.5 ± 3.2 days.

Table 1 Characteristics of the living liver donors

Sixty LLDs (48.3%) experienced complications that were categorized as follows by the Clavien–Dindo classification: grade I, which included fluid collection, pleural effusion, subcutaneous emphysema, atelectasis, dizziness and nausea, keloid and hypertrophic scars, wound dehiscence, hematuria, vaginal oozing, fatty liver, chest pain, shoulder pain, fever, and temporarily elevated aspartate transaminase or/and alanine transaminase levels over at least 1 year after the transplantation; grade II, which included dyspepsia, gaseous distention, chronic cough, urticarial rash, diarrhea, gastroenteritis, colitis (all requiring antibiotics and etc.), and portal vein stenosis (requiring aspirin); grade IIIa, which included common bile duct stenosis, biloma, and pulmonary thromboembolism; or grade IIIb, which included hematoma.

A small proportion of LDLT recipients (8.9%) was deceased at the time point of survey completion. The time since donation ranged from 1 month to 12 years; two-thirds of the donors (66.1%) underwent surgery for transplant within 3 years. The majority of donors (75.8%) were very satisfied with their decision to donate their liver.

Quality of life and unmet expectations about surgical outcomes

The mean donor SF-12 scores were 51.48 ± 7.44 (PCS) and 52.97 ± 8.47 (MCS) while the normative SF-12 scores for the general Korean population were 43.46 ± 3.05 (PCS) and 45.26 ± 4.35 (MCS) [19]. SF-12 component summary mean scores in LLD by time since donation and those of general Korean population are seen in Fig. 1. Among the study participants, 14 (11.3%) had poor PCS scores, and 16 (12.9%) had poor MCS scores. In terms of unmet expectations about surgical outcomes (Table 2), the percentages of the participants who reported worse-than-expected experiences for length of hospital stay, speed of recovery, pain, and complications were 9.7%, 22.6%, 34.7%, and 7.3%, respectively.

Fig. 1

Short Form-12 health survey (SF-12) component summary

The figure depicts the short Form-12 health survey (SF-12) component summary mean scores in LLD and general Korean population by time since donation

Table 2 Frequencies of the four items of postoperative experience in comparison with preoperative expectations Logistic regression models

The final logistic regression model for PCS is shown in Table 3, and that for MCS is shown in Table 4. In the univariable model for PCS (Table 3), time since donation and unmet expectations for surgical outcomes were significantly associated with poor PCS scores (unadjusted odds ratio [UOR] 0.53, 95% CI 0.32–0.88; UOR 6.93, 95% CI 1.67–28.74). In the univariable model for MCS (Table 4), education level, satisfaction with decision to donate, and unmet expectation were statistically related to poor MCS scores (UOR 0.32, 95% CI 0.11–0.93; UOR 0.37, 95% CI 0.17–0.83; UOR 5.67, 95% CI 1.40–22.97).

Table 3 Logistic regression model for PCS scores (n = 124) Table 4 Logistic regression model for MCS scores (n = 124)

For the multivariable logistic regression models for both PCS and MCS, age, sex, education level, monthly income, postoperative complications, recipient death, time since donation, and satisfaction with the decision to donate were controlled. The multivariable logistic regression model for PCS was significant (X2 (df = 9) = 19.313, p = .023) with acceptable goodness-of-fit statistics (Hosmer–Lemeshow: p = .177). In this adjusted model for PCS, unmet expectations for surgical outcomes predicted poor PCS scores (adjusted odds ratio [AOR] 7.46, 95% CI 1.38–40.49) after controlling for age, sex, education level, income, postoperative complications, recipient death, time since donation, and satisfaction with the decision to donate (Table 3). In other words, donors who reported three or four unmet expectations were more likely to have poor PCS scores than those who reported two or fewer unmet expectations. In this model, a shorter interval since donation was also associated with poor PCS scores (AOR 0.50, 95% CI 0.27–0.95).

The multivariable logistic regression model for MCS was significant (X2 (df = 9) = 18.638, p = .028) with acceptable goodness-of-fit statistics (Hosmer–Lemeshow: p = .266). In this adjusted model for MCS, unmet expectations for surgical outcomes were a predictor of poor MCS scores after controlling for other factors (AOR 7.15, 95% CI 1.35–37.97) (Table 4). The likelihood of poor MCS scores increased in donors who had three or more items of unmet expectations than in those who had two or fewer items of unmet expectations (AOR 7.15, 95% CI 1.35–37.97). In addition, LLDs having less than a bachelor’s degree and less satisfied with decision to donate were likely to experience poor MCS (AOR 0.24, 95% CI 0.06–0.96; AOR 0.30, 95% CI 0.12–0.77).