Posttraumatic stress symptoms after solid-organ transplantation: preoperative risk factors and the impact on health-related quality of life and life satisfaction

Knowledge about possible predictors and risk factors of posttraumatic stress symptomatology after SOT is still very limited. Only few studies have aimed to explore preoperative risk factors of PTSS after transplantation. We therefore examined variables such as type of transplantation, history of retransplantation, preexisting psychiatric morbidity, preoperative chronic benzodiazepine consumption, preoperative alcohol abuse, and important sociodemographic characteristics. Furthermore, we investigated the association of PTSS, HRQOL and life satisfaction.

The trauma criterion was homogeneously defined as SOT followed by ICU treatment.

During the study period a total of 280 patients underwent SOT at the Transplantation Center of the Klinikum Grosshadern, Ludwig-Maximilians-University, Munich, Germany.

Before SOT, the transplantation candidates were evaluated by experienced consultation-liaison psychiatrists using the observer-rating Transplant Evaluation Rating Scale (TERS) [25]. 65 of these 280 patients died before study inclusion. 215 eligible SOT recipients were contacted in a letter and were asked to complete a research battery, which included an author-compiled clinical questionnaire, the PTSS-10 [26, 27], the Medical Outcome Study Form (SF-36) [28] and the life satisfaction questionnaire “Fragebogen zur Lebenszufriedenheit” (FLZ) [29]. 126 transplantation recipients returned the questionnaires, at a mean of 24.9 months (SD=11.9) after transplantation. 84 patients responded to our letter but refused to join the study and 5 patients were untraceable. Respondents with complete data were similar on the sociodemographic and treatment parameters to those lost due to refusal or untraceability.

Independent of type of transplantation the data analysis was performed for the whole sample according to the homogeneously definition of the trauma criterion as SOT followed by ICU treatment. The SF-36 data were compared with data from normative population whose individuals were drawn with respect to age and gender in a pairwise matching from a large data base (n=3000) used for the validation of the SF-36 in Germany.

The study was approved by the Institutional Review Board of the Ludwig-Maximilians-University of Munich. Data protection met the standards set by German law.

The psychiatric history is based on a psychiatric clinical interview before transplantation performed by experienced consultation-liaison psychiatrists (H.-B. R., T. K.).

Descriptive statistics were produced based on demographic, treatment-related and psychometric data (PTSS-10, SF-36, FLZ) and are presented as mean and standard deviation (SD). We applied the non-parametric χ² - test, the Fisher’s exact test and the Mann–Whitney-U test to test differences between patients with or without PTSS and the non-parametric Wilcoxon signed-rank test to compare solid-organ transplantation recipients with healthy controls matched by age and gender. All statistic tests were two-tailed, with significance set at p<0.05. In case of multiple comparisons (SF-36, FLZ), an alpha adjustment (Bonferroni) was used. By means of multivariate analyses in the form of multiple regression models, the level of influence by the identified risk factors and PTSS-associated manifestation in the SF-36 and FLZ dimensions in regard to a PTSS symptomatology was examined.

All statistical analyses were performed with IBM SPSS Statistics 20.0 for Windows.

All 126 participants (69.0% men, 31.0% women) were Caucasian, and the mean age was 52.4 (SD=11.6) years.

Preoperative psychiatric diagnoses based on the TERS and an examination by experienced consultation-liaison psychiatrists were recorded more frequently in patients with liver transplantation compared with patients after heart or lung transplantation (χ² =53.657; df=34; p=0.017). Main preoperative psychiatric diagnoses in the subgroup of patients with liver transplantation were substance abuse (n=9), personality disorder [narcissistic or histrionic] (n=6), and adjustment disorder (n=3).

Respondents with complete data were similar to those lost due to refusal and untracebility on the sociodemographic and treatment parameters.

Table 1. summarizes the pre-transplantation clinical characteristics of the whole sample, including the type of transplantation.

Table 2. shows the sociodemographic characteristics of the whole sample, including the type of transplantation.

Before SOT, the psychometric observer-rating scale TERS together with an examination by experienced consultation-liaison psychiatrists was performed on all SOT candidates, and in no case were preoperative posttraumatic stress disorder (PTSD) or PTSS diagnosed.

After the surgical intervention 19 (15.1%) SOT recipients had clinical significant PTSS as measured on the PTSS-10 scale. The mean PTSS-10 score for the SOT recipients with PTSS was 44.6 (SD=6.9) points and for those without PTSS 18.9 (SD=6.8) points (Mann–Whitney-U=0.0; p<0.001). The average length of time since transplantation at the investigation point was similar in SOT recipients with and without clinically significant PTSS (SOT-recipients with PTSS 28.58 months; SOT-recipients without PTSS: 24.22 months; Mann–Whitney-U: 801.5, p=0.142).

Seven (43.8%) SOT recipients with PTSS and 12 (10.9%) patients without PTSS had a preoperative psychiatric diagnosis before SOT (χ² =11. 765; df=1, p=0.001), suggesting an increased number of previous psychiatric illness in SOT recipients with PTSS. Frequent preoperative psychiatric diagnoses in the subgroup of SOT recipients with PTSS were adjustment disorder (n=5), alcohol dependence (n=2), histrionic personality disorder (n=2), dystymia (n=1), somatoform disorder (n=1)), obsessive compulsive disorder (n=1) and drug abuse (n=1).

SOT recipients with PTSS more frequently take benzodiazepines after the surgical intervention, but it needs to mentioned that even before the SOT, patients with postoperative PTSS had a higher level of benzodiazepines consumption (Fisher’s exact test, p=0.018). No differences between patients with PTSS and those without were found in the pre- and postoperative frequency of alcohol abuse (Fisher’s exact test, p=1.000).

Table 3. Pre- and postoperative frequency of benzodiazepine consumption and alcohol abuse.

Six liver transplant and one lung transplant recipients had a history of retransplantation and SOT recipients with a history of retransplantation suffered by trend more frequently from PTSS than those without (Fisher’s exact test, p=0.069).

Table 4. presents the history of retransplantations of SOT recipients with and without posttraumatic stress syndromes.

No significant differences between SOT recipients with PTSS and those without were found on the following sociodemographic characteristics: gender (χ² =0.363; df=1, p=0.594), marital status (Fisher’s exact test, p=0.555), years in education and/or vocational training (Mann–Whitney-U=988.5; p=0.846) and employment status (Fisher’s exact test, p=0.092).

SOT recipients with PTSS were younger than those without (mean age of transplantation recipients with PTSS 48.8 years [SD=10.1]; mean age of transplantation recipients without PTSS 53.0 years [SD=11.8]; Mann–Whitney-U=719.0; p=0.042).

Patients with liver transplantation displayed a higher level of PTSS after the surgical intervention than patients with heart or lung transplantation (Fisher’s exact test, p=0.003).

Table 5. illustrates PTSS according to type of transplantation.

The occurrence (χ² =1.178; df=2, p=0.585) and the number (Fisher’s exact test; p=0.933) of postoperative medical complication (e.g. bleeding, infections, cardio-vascular) did not significantly differ in the liver transplantation recipients in our study from that in heart and lung transplantation recipients with the exception that liver and lung transplantation recipients had significantly more frequently acute rejections (χ² =9.401; df=2, p=0.009). Furthermore, the incidence of posttraumatic stress symptoms after transplantation was independent of postoperative medical complications (χ² =0.676; df=1, p=0.462).

In comparison with those SOT recipients without PTSS, transplantation recipients suffering from PTSS displayed significant impairments in all HRQOL SF-36 domains: Physical Functioning, Role Physical, Pain, General Health, Vitality, Social Functioning, Role Emotional, and Mental Health.

Table 6. shows the HRQOL SF-36 domains according to PTSS.

As mentioned above, SOT recipients without PTSS have significant lower impairments in all HRQOL domains than those with PTSS. However, compared with healthy controls matched by age and gender, even SOT recipients without PTSS have impairments in some HRQOL SF-36 domains: Physical Functioning (Wilcoxon-Z=−7.434, p<0.001), Role Physical

(Wilcoxon-Z=−6.452, p<0.001), Pain (Wilcoxon-Z=−5.075, p<0.001), General Health (Wilcoxon-Z=−4.434, p<0.001), Vitality (Wilcoxon-Z=−4.112, p<0.001), Social Functioning (Wilcoxon-Z=−4.679, p<0.001), and Role Emotional (Wilcoxon-Z=−3.225, p=0.001).

In comparison with those SOT recipients without PTSS, transplantation recipients showing PTSS displayed significant impairments in the life satisfaction questionnaire (FLZ) global score and in the life satisfaction domains health and character/own skills.

Table 7. shows the life satisfaction domains (FLZ) according to PTSS.

In the univariat analyses described so far the following risk factors for the display of PTSS after SOT were identified: 1.) preexisting psychiatric morbidity, 2.) history of retransplantation (with the limitation that in our sample only patients with liver or lung transplantations had retransplantations in the past), 3.) chronic benzodiazepine consumption, 4.) age, and 5.) type of transplantation. In the second step a multivariate analysis in the form of a multiple-regression model was used to examine the levels of influence of the identified risk factors of PTSS. The following predictors were selected in the prediction of PTSS-10 values: preexisting psychiatric morbidity, history of retransplantation, benzodiazepine consumption before transplantation, age, and type of transplantation. Added to the list of predictors were medical complications, survival time since transplantation and gender. This model yields a R² (adj.) =0.145, F=3.125, p=0.002 and shows that significant predictors of PTSS post-SOT are preexisting psychiatric morbidity and history of retransplantation.

In a third step, an additional multiple regression model was calculated that included the PTSS-associated dimensions in HRQOL (SF-36) and life satisfaction (FLZ) identified in the univariat analyses as well as benzodiazepine consumption after SOT. This model yields R² (adj.) =0.562, F=37.229, p<0.001 and shows the statistically significant associative relationship between impaired HRQOL (SF-36 - physical health subscales: p<0.001; SF-36 mental health subscales p<0.001) and life satisfaction (FLZ - occupation, work: p=0.004) and PTSS-10 values.

Table 8. Multiple regression analyses.

Finally, there are several limitations to this study that deserve mention. The format of the study is a retrospective follow-up, and the absence of a prospective randomized trial makes it difficult to definitively conclude that the HRQOL associations observed are causal in nature.

The response rate is similar to that reached in other comparable studies (e.g. Rothenhäusler et al. [24]). As the eligible SOT recipients were asked in a letter, not in a face-to-face follow-up check, to participate in this study reasons for refusal could not be collected in all patients. A possible skewing of the study results due to the response rate of 59% cannot definitely be ruled out. The PTSS-10 is not sufficient for the diagnosis of posttraumatic stress disorder. Therefore elevated PTSS-10 scores were only classified as posttraumatic stress symptoms. Pre-transplant psychosocial assessments might be distorted due to the point, that SOT candidates are concerned that any information that they impart could influence the transplant team’s decisions about the patient’s candidacy. Furthermore, the study population consists of a mixed group of patients with different organ entities, in whom transplantation has been conducted and the small group size of the organ-subgroups is a limitation to differ between the different subgroups. However, in regard to posttraumatic stress symptoms common data interpretation is possible due to the homogeneously trauma criterion definition as SOT followed by ICU treatment. Finally, preoperative psychiatric diagnoses based on the TERS and an examination by consultation-liaison psychiatrists. Additional structured interviews have not been conducted to secure the diagnoses.