Psychosocial Characteristics of Patients Evaluated for Kidney Transplant and Associations with Functional and Frailty Metrics at a Veterans Affairs Hospital

Abstract

Background:

The effect of psychosocial problems on listing outcomes and potential interactions with functional metrics is not well-characterized among Veteran transplant candidates.

Methods:

The results from psychosocial evaluations, frailty metrics, and biochemical markers were collected on 375 consecutive Veteran kidney transplant candidates. Psychosocial diagnoses were compared between patients listed or denied for transplant. Functional abilities were compared among patients with or without psychosocial diagnoses and then evaluated based on reason for denial.

Results:

Eighty-four percent of patients had a psychosocial diagnosis. Common issues included substance or alcohol abuse (62%), psychiatric diagnoses (50%), and poor adherence (25%). Patients with psychiatric diagnoses, cognitive impairments, and poor adherence were more likely to be denied for transplant (p<0.05). Patients with depression, PTSD and anxiety did not have worse functional ability, but experienced more exhaustion than patients without these problems. Patients denied for medical but not purely psychosocial reasons had worse troponin and functional metrics compared with listed patients.

Conclusion:

Over 80% of patients with a psychosocial diagnosis were listed; however, poor adherence was a particularly important reason for denial for purely psychosocial reasons. Patients with psychosocial diagnoses generally were not more functionally limited than their counterparts without psychosocial diagnoses or those listed for transplant.

1. INTRODUCTION

Candidate evaluation is an important initial step in determining transplant outcomes. Mental health and psychosocial issues are common among renal failure and kidney transplant candidates and are associated with short and long-term outcomes.^1-14 Therefore, these issues are considered in the transplant evaluation process.^15,16 While pre-transplant mental health evaluation methods are variable across the US^13,14,17, to accommodate the high rate of mental health issues among Veterans^18-20, the VA kidney transplant evaluation process was developed to include both a social work and psychological evaluation for each candidate prior to an onsite evaluation^21,22.

Frailty, also common among transplant patients²³, is known to be associated with transplant outcomes including length of stay²⁴, readmission²⁵, delayed graft function and mortality^24,26-28 and has been increasingly incorporated into the candidate selection process.^29,30 Frailty, mental health, and psychosocial issues commonly intersect^31-33 and can synergize to produce worse outcomes in transplant patients.⁹ Experience is evolving that considers the effects of frailty and mental health issues in the post-transplant period; however, less is known about the impact of these issues when identified at the time of candidate evaluation.

The Iowa City VA transplant program collects frailty, pedometer, treadmill and cardiac biochemical markers on all candidates offered an on-site evaluation. B-type natriuretic peptide (BNP) and troponin are known to be associated with survival among dialysis and pre and post-transplant patients and are also collected at our center as part of the transplant evaluation.^34-40 Prior retrospective analysis of our patient population showed a close correlation between frailty, functional and biochemical markers and transplant evaluation outcomes.⁴¹ The associations were made with our population as a whole where the majority of candidate denials were based on medical reasons. However, we are aware that some candidates are turned down for purely psychosocial reasons and therefore were interested in studying potential interactions of functional and biochemical metrics among patients in this subset. Given past reports of interactions of frailty with psychosocial diagnoses, we hypothesized that frailty would be associated with candidate denial for psychosocial reasons.

Our study had three goals: 1) to characterize our evaluation and listing process with respect to our patients’ psychosocial diagnoses, 2) to evaluate potential associations between psychosocial diagnoses and functional, frailty and biochemical markers, and 3) to compare functional, frailty and biochemical markers between patients listed and those denied for either medical or purely psychosocial reasons.

2. METHODS

2.1. Study population

Following Institutional Review Board approval through the Iowa City VA Health Care System and University of Iowa, we performed a retrospective analysis of patients undergoing a first-time kidney transplant evaluation at the Iowa City VA Health Care System between January 1, 2015 and December 31, 2018. The VA kidney transplant referral and evaluation process has been previously described.^21,22 Briefly, in the VA system pre-transplant evaluation testing is assembled and screened prior to on-site evaluation. The pre-transplant packet evaluation includes a social work and mental health assessment for all patients. A third psychosocial evaluation is performed by the Iowa City transplant social worker at the time of the on-site evaluation. Electronic and paper chart records of these three assessments were available for all patients and served as the source of mental health and psychosocial information as described below. Patients denied at the ‘packet’ level account for approximately 20% of total referrals and were not included in this analysis.

Basic demographic factors that were abstracted from the medical records included patient age, sex, race (as self-reported to and categorized by the VA), body mass index (BMI), history of hypertension and diabetes, length of time on dialysis at the time of the evaluation, and current and past smoking history (yes/no).

Biochemical tests and frailty assessments were collected on all patients at the time of the on-site evaluation as part of our routine evaluation as described below.

Following onsite evaluation, patients were divided according to their initial transplant evaluation outcome as falling into one of three groups: listed, denied, or deferred. Transplant evaluation outcomes were recorded in the medical record for each patient. For each patient deferred or denied, the reasons were enumerated in the medical record and extracted for this study. The reasons were subsequently divided into medical or psychosocial categories. The evaluation flow chart is presented in Table 1.

Table 1.

Evaluation flow

Following evaluation, 375 patients were denied, deferred, or listed for transplant during the initial listing conference. Deferred patients were further evaluated and denied or listed after obtaining additional information. Sixty-two patients were denied after deferral and added to the 95 initially denied, for 157 total denials. Similarly, 110 patients were listed after deferral and added to 98 initially listed patients, for 208 total listed patients. Ten patients remained in deferral at the time of the analysis.

2.2. Outcome Assessment

The primary outcome of interest was the final decision to list or deny a patient for kidney transplant. Final listing disposition was chosen over initial evaluation outcomes as it was felt to be more clinically important. Furthermore, it nearly doubled the sample for our comparison between listed and denied patients (due to the initially deferred observations). Additional key study outcomes include psychosocial diagnoses, biochemical and frailty metrics among patients based on whether they were denied for medical or psychosocial reasons.

2.3. Psychosocial Information Acquisition

Psychosocial information was gathered from retrospective review of templated notes, including 1) Mental Health Assessment for Transplant, 2) Social Work Assessment for Transplant, and 3) the onsite Social Work transplant evaluation, which were completed for each patient. The elements captured by the Mental Health and Social Work Assessment for Transplant notes completed following in-person assessments at a patient’s local VA and submitted as part of the packet prior to the onsite evaluation are presented in Table 2. Psychosocial information that was collected included all psychiatric diagnoses, current and past smoking history, current and past alcohol use and abuse history, illicit substance use history, presence of cognitive impairment if supported by objective testing, and history of poor adherence. Of note, information related to mental health diagnoses, illicit substance abuse, tobacco, alcohol, adherence, and cognition are solicited in both assessments.

Table 2.

Comparison of the components of the Mental Health and Social Work Assessments for Transplant

The VA transplant referral process requires an assessment by both a mental health and social work provider. Providers perform a template guided semi-structured evaluation specific to the VA kidney transplant program. Elements required for each templated evaluation are shown. In general, the social work assessment is more specific in describing required elements. Mental health assessments are performed by either a licensed clinical psychologist (PhD) or psychiatrist (MD). The social work assessment is done by a social worker with a Master of Social Work degree with a clinical or independent license.

Component	Mental Health Assessment	Social Work Assessment	Comments
History of present medical illness	R	NR
Past medical history	R	R	Social work assessment titles this category “background clinical information”
Education history	NR	R
Employment history	NR	R
Current medications	R	NR
Family history	R	R	Social work assessment also requires comments on marital history
Adherence/compliance	R	R
Substance abuse history	R Alcohol Drugs Tobacco Other	R Alcohol Amphetamines Barbiturates Benzodiazepines Cannabinoid Cocaine Methadone Opiates Propoxyphene Tobacco Other	Social work assessment is more prescriptive by providing checklist of substances to be considered
Mental health history	R Appearance Behavior Speech Mood Affect Thought process Thought content Attitude Ability to relate to interviewer Brief cognitive assessment using either MOCA, Mini-Cog, SLUMS, or BOMC	R Past and present treatment/hospitalizations Emotional status (anger, anxiety, depression, suicidal ideation, gender identity) Past abuse Mental status (orientation, memory, intelligence, mood)	Mental health assessment template indicates use of a brief cognitive assessment tool
Legal history	R	R Driving under the influence of alcohol Jail time Other legal (child custody, pending or recent divorce, immigration, pending lawsuits, parole or probation)	Social work assessment is more prescriptive by providing checklist
Social history	R	R
Financial resources	NR	R
Support person	NR	R
Mental status exam	R	R Orientation Memory Intelligence Mood	Validated instruments are usually used, but the system does not require specific tests for either mental health or social work assessments

R, required; NR, not required; MOCA, Montreal Cognitive Assessment; SLUMS, St. Louis University Mental Status Exam; BOMC, Blessed Orientation Memory Concentration Test

A patient was considered to have poor adherence if any compliance or adherence issues were noted in either the mental health provider or social worker packet notes. In addition, adherence was screened by the onsite social worker by review of dialysis appointment attendance records for the past three years.

Cognitive impairment was operationally defined as a validated test result consistent with at least mild cognitive impairment. The Social Work and Mental Health Assessments templates indicate that a cognitive assessment is part of both evaluations. A preference for the Montreal Cognitive Assessment© (MoCA), Mini-Cog©, Saint Louis Mental Status Exam (SLUMS), or Blessed Orientation-Memory-Concentration (BOMC) test is indicated in the Mental Health Assessment template. A variety of other tests were also used by centers according to local protocol. These included, Mini-Mental Status Examination© (MMSE), Repeated Battery for Neuropsychological Status© (RBANS), the Minnesota Multiphasic Personality Inventory© (MMPI), Wechsler Adult Intelligence Scale© (WAIS), and the Brief Neuropsychological Cognitive Examination (BNCE^™).

2.4. Biochemical and Frailty Factors Assessed at the Iowa City VA at the Time of On-Site Transplant Evaluation

Biochemical markers

Blood for biochemical tests was drawn before functional tests. Cardiac troponin T and N-terminal pro-brain natriuretic peptide (BNP) drawn at the time of evaluation were immunoassayed in our clinical lab using electrochemiluminescence on a Roche Cobas^® 6000 analyzer (Roche Diagnostics, Indianapolis, IN). The troponin T reference range was 0.00 to 0.03 ng/ml, with a critical level defined as > 0.09ng/ml. BNP reference ranges were age (by decade) and sex adjusted. For males and females respectively, the 95^th percentile for BNP levels in pg/ml were as follows: ages 45-54, 138 and 192; ages 55-64, 177 and 226; ages 65-74, 229 and 353; and ages ≥75, 852 and 624.

Frailty metrics

Four functional metrics were collected including handgrip, 30 second chair sit-stand, chair sit-reach, and timed up-and-go. Each metric was measured as previously described.^42-44 Briefly, handgrip was measured in kilograms on a calibrated Jamar^® Dynamometer (JLW Instruments, Chicago, IL) in the right and left hands and a mean value was calculated.⁴⁵ The number of times that a patient rose to the full standing position within a 30 second period was recorded in the sit-stand test. For the sit and reach test, the gap in centimeters between the middle fingertip and toes of an extended leg was measured with the patient reaching with both arms extended, in the seated position, at the edge of the chair. Ability to touch the toes or go beyond was recorded as zero. In the up-and-go test the patient was timed while proceeding as quickly as possible from the seated position around a cone positioned 8 feet from the chair and returning to the seated position.

Two self-reported answers to questions from the Center of Epidemiological Studies Depression Scale (CES-D) measuring level of exhaustion were recorded. These included in how many days in the past week (0-7 days) a) was everything you did an effort, and b) could you not get going?⁴⁶ Our previous work⁴¹ showed that question b was more closely related with listing outcome and was chosen for use as the exhaustion marker in the current work.

Treadmill and Pedometer testing

Treadmill and other functional frailty measurements were made on non-dialysis days (in the case of hemodialysis) or prior to dialysis if conducted on a dialysis day.

Symptom limited treadmill testing was performed on a GE Health care T-2100^® treadmill (GE Medical Systems Milwaukee, WI) according to the modified Bruce or Naughton protocols as previously described.⁴⁷ Early in the data collection phase, the modified Bruce or Naughton was used, but with experience we evolved to exclusively use the Naughton protocol because we found that its more graded acceleration into stress gave the renal failure patients a better opportunity to achieve their maximal plateau. Treadmill data that was recorded included maximal METS level achieved, total time on treadmill, and reason for stopping the test. For this analysis only METlevel was analyzed.

An EKHO Two pedometer (EKHO^®; Dallas, TX) was calibrated for each patient per manufacturer instructions. Patients were instructed, to wear the pedometer all the time while awake. Pedometer “on” and “off” times were recorded. This data was measured as steps taken and calculated steps per time.

At the time of the multidisciplinary listing conference the team was blinded to pedometer results, handgrip, 30 second chair sit-stand, chair sit-reach, and timed up-and-go. Treadmill, troponin T, BNP, and information from the medical and psychosocial patient evaluations were presented at listing conference.

2.5. Statistical analysis

Comparisons were made between patients based on their final listing disposition, splitting the sample into two cohorts: final list and final denial for transplant listing (10 patients still in deferral status were excluded from this part of the analysis). Summary statistics for demographic factors and psychosocial diagnoses were generated for the cohorts, displaying means (SDs) or medians (IQRs) for continuous measures, depending on the distribution, and counts (percentages) for categorical measures. Univariate generalized linear models (GLMs) were utilized to test for differences in summarized measures between the cohorts. The GLM framework accommodates an array of outcome distributions, including normal, Bernoulli, multinomial, and gamma (right skewed), while providing a more powerful approach than common non-parametric analogues. A similar set of summary statistics and comparisons of the demographic and psychosocial diagnosis measures once again employed the GLM framework, now stratifying on reason for final denial: medical only, combined medical and psychosocial, and psychosocial only. These assessments report only the overall cohort effect on the demographic and psychosocial measures. Pairwise comparisons between the three cohorts were not reported for this part of the analysis.

We further employed GLMs of order two and three to evaluate factors associated with denial for purely psychosocial reasons. Model fits were compared using the Akaike information criterion (AIC)^48,49 where smaller values indicate a more appropriate predictor set specification. Interaction effects between predictors in the top models (according to AIC) were also assessed for inclusion in the models. Additional models involving depression and adherence (with and without interaction effects)predicting rejection for purely psychosocial reasons allowed us to assess suspected multicollinearity.

One final stratification on the full data set (excluding the 10 patients still in deferral status) was used to further compare frailty measures among patients who were: listed, denied due to medical reasons only, denied due to a combination of medical and psychosocial reasons, and denied for purely psychosocial reasons. GLMs were used to test for an overall difference in our outcome measures between the four groups, as well as make pairwise comparisons among the three deny groups, and between the listed patients and those denied for purely psychosocial reasons. Comparisons with p-values < 0.05 were considered statistically significant. All analyses were conducted in SAS 9.4.

3. RESULTS

3.1. Demographics

Among the 375 total evaluated patients, 97% were male, 57% White, and 32% Black/African American, with other races constituting the remainder. History of diabetes mellitus and hypertension were present in 75% and 95% of patients, respectively. Causes of renal failure included diabetes (51%), hypertension (18%), focal and segmental glomerulosclerosis (8%), polycystic kidney disease (6%), primary urologic (3%), glomerulonephritis (3%), IgA nephropathy (2%), and other (9%). Average length of time on dialysis among all patients was 854 days at the time of evaluation. Approximately 7% were current smokers and 70% were former smokers.

3.2. Evaluation outcomes

Twenty-six percent of the candidates were immediately listed, 25% were immediately denied, and 48.5% deferred. Among the patients immediately denied, 51% were denied for purely medical reasons, while 36% were denied for a combination of medical and psychosocial reasons, and 14% for purely psychosocial reasons (Table 1).

Among 182 deferred patients, the deferral reason was medical in 55%, combined medical and psychosocial in 37%, and purely psychosocial in 8%. Among the deferred patients, 60% were eventually listed and 34% were denied. The remainder were still in deferral at the time of data analysis. Sixty-three percent of the candidates where medical reasons contributed to the deferral and 57% of those deferred with contributing psychosocial reasons were eventually listed. Only four of the 14 patients deferred for purely psychosocial reasons were eventually listed.

Among the 157 patients whose final status was denial, 51% were denied for medical reasons, 35% for combined medical and psychosocial reasons, and 14% for purely psychosocial reasons. Of these purely psychosocial denials, poor adherence was a factor in 62%, psychiatric disease in 50%, low cognition in 33%, poor social support in 25%, substance abuse in 12.5%, and financial constraints in 0.04%.

3.3. Demographic and psychosocial factors among patients listed or denied for transplant

Listed patients had lower BMI and were less likely to have diabetes mellitus. There was no significant difference in race, length of time on dialysis, or current smoking status comparing listed and denied patients. Prior smoking history trended toward favoring denial (p=0.07). Fifty percent of candidates had a psychiatric diagnosis and only 15.5% had no psychiatric, substance abuse history (alcohol or illicit drug), cognitive impairment or adherence problem history. Depression, PTSD, and anxiety were the most common psychiatric diagnoses. Other psychiatric diagnoses were present in 13% of patients. Patients with psychiatric diagnoses, particularly depression, cognitive impairment, and histories of poor adherence were more likely to be denied for transplant. (Table 3)

Table 3.

Demographics and psychosocial factors among patients based on final listing decision

Demographic	Final List	Final Denial	p-value
N	208	157
Age (mean (sd))	60.0 (9.8)	61.8 (9.5)	0.08
Male	199 (95.7%)	153 (97.5%)	0.36
Race/ethnicity
White	124 (59.6%)	86 (54.8%)
Black/African American	60 (28.9%)	56 (35.7%)	0.33
Hispanic	20 (9.6%)	10 (6.4%)
Other	4 (1.9%)	5 (3.2%)
BMI (mean (sd))	29.5 (4.2)	30.6 (5.1)	0.0146
HTN	200 (96.2%)	155 (98.7%)	0.20
DM	122 (58.9%)	110 (70.1%)	0.0288
LoT dialysis (median (IQR))	587 (227-1297)	500 (249-1185)	0.57
Current Smoking	13 (6.3%)	12 (7.6%)	0.60
Previous Smoking	138 (66.4%)	118 (75.2%)	0.07
Psychosocial Factors
No psychiatric diagnosis	114 (54.8%)	68 (43.3%)	0.0118
All psychiatric diagnoses†	94 (45.2%)	89 (56.7%)	0.0297
All psychiatric diagnoses on psychoactive meds	47 (22.6%)	45 (28.7%)	0.19
All anxiety spectrum‡	31 (14.9%)	28 (17.8%)	0.45
Depression	54 (26.0%)	61 (38.9%)	0.0087
PTSD	36 (17.3%)	37 (23.6%)	0.14
Schizophrenia/schizoaffective	2 (1.0%)	2 (1.3%)	> .99
Bipolar	7 (3.4%)	1 (0.6%)	0.14
Personality disorder spectrum*	6 (2.9%)	8 (5.1%)	0.28
Other psychiatric diagnoses	12 (5.8%)	11 (7.0%)	0.79
Alcohol abuse history	87 (41.8%)	63 (40.1%)	0.74
Illicit drug	76 (36.5%)	70 (44.6%)	0.12
Cognitive impairment	27 (13.0%)	47 (29.9%)	0.0001
Adherence issues	27 (13.0%)	63 (40.1%)	<0.0001

BMI, body mass index; HTN, hypertension; DM, diabetes mellitus; LoT dialysis, length of time on dialysis prior to date of transplant evaluation; PTSD, posttraumatic stress disorder

^†All patients with psychiatric diagnoses

^‡Includes patients with anxiety, panic, and obsessive-compulsive disorders

^*Includes patients with anger management, borderline, antisocial and unclassified personality disorders

Fifty percent of patients with psychiatric diagnoses (or 25% of all patients) were being prescribed psychiatric medications at the time of transplant evaluation. There was no difference in listing outcome comparing patients with psychiatric diagnoses being treated with psychiatric medications and those not taking psychiatric medications. The most common classes of psychiatric medications among our cohort included selective serotonin reuptake inhibitors in 50 patients, norepinephrine-dopamine reuptake inhibitors in 9 patients, serotonin-norepinephrine reuptake inhibitors in 7 patients, and benzodiazepines in 19 patients. Among the evaluated patients, 17.6% were taking an antidepressant. An additional 12 patients were taking other medications related to psychiatric diagnoses.

3.4. Concurrent diagnoses

There were 296 separate psychiatric diagnoses among 183 patients showing that concurrent psychiatric problems were common. Concurrent psychosocial diagnoses are displayed in Figure 1. Cross-tabulation of the various diagnoses shows that among patients with depression, 46%, 38%, 35% and 30% also had histories of illicit substance use, alcohol abuse, anxiety, and PTSD, respectively. Among patients with an alcohol abuse history 29% had depression, 20% PTSD and 17% anxiety. Among patients with cognitive impairments, three conditions, namely depression, illicit substance use history, and alcohol abuse history were relatively common (53%, 54%, and 32% respectively). Similarly, among patients with nonadherence, 42% had depression, 50% had an illicit substance use history, and 42% had an alcohol abuse history.

Figure 1.

Bubble plot with proportional representation of co-existing psychosocial diagnoses. Percentages within bubbles represent the percent of patients with the coordinate x and y axis psychosocial diagnoses relative to the total number of patients with each diagnosis along the x-axis. For example, among the 115 patients with depression, 41 (35%) also had a diagnosis of anxiety. Among the 59 patients with an anxiety disorder, 41 (68%) also had depression.

Cog. Deficit, cognitive deficit; Illicit subst., history of illicit substance use; alcohol, alcohol abuse history; PTSD, post-traumatic stress disorder history.

3.5. Comparison of patients denied for medical versus psychosocial reasons

The demographics and psychosocial diagnoses stratified according to reason for denial (medical only, combination of medical and psychosocial, and psychosocial only) were compared among patients whose final listing status was “denied for transplant” (Table 4). Patients denied for purely psychosocial reasons trended toward lower BMI and were more likely to be younger, have shorter time on dialysis, and have diagnoses of depression, personality disorder, illicit drug use, cognitive impairment, and adherence issues. There was no difference in likelihood of denial for purely psychosocial reasons based on race or ethnicity.

Table 4.

Demographics and psychosocial factors stratified according to final denial reason

Demographic	Medical Only	Med/Psych Combined	Psychosocial Only	p-value
N	80	55	22
Age (mean (sd))	63.6 (8.2)	60.1 (9.5)	59.4 (12.9)	0.0480
Male	80 (100%)	52 (94.6%)	21 (95.5%)	0.06
Race/ethnicity
White	47 (58.8%)	28 (50.9%)	11 (50.0%)
Black/African American	28 (35.0%)	21 (38.2%)	7 (31.8%)	0.23
Hispanic	4 (5.0%)	5 (9.1%)	1 (4.6%)
Other	1 (1.3%)	1 (1.8%)	3 (13.6%)
BMI (mean (sd))	31.0 (5.1)	31.0 (5.4)	28.2 (3.5)	0.0524
HTN	80 (100%)	54 (98.2%)	21 (95.5%)	0.12
DM	54 (67.5%)	41 (74.6%)	15 (68.2%)	0.66
LoT dialysis (median (IQR))	481 (258-876)	904 (293-1720)	260 (141-488)	0.0474
Current Smoking	8 (10.0%)	3 (5.5%)	1 (4.6%)	0.49
Previous Smoking	63 (78.8%)	40 (72.7%)	15 (68.2%)	0.49
Psychosocial Factors
No psychiatric diagnosis	39 (48.8%)	21 (38.2%)	8 (36.4%)	0.37
All psychiatric diagnoses†	41 (51.3%)	34 (61.8%)	14 (63.6%)	0.37
All psychiatric diagnoses on psychoactive medication	18 (22.5%)	20 (36.4%)	7 (31.8%)	0.20
All anxiety spectrum‡	12 (15.0%)	12 (21.8%)	4 (18.2%)	0.62
Depression	28 (35.0%)	22 (40.0%)	11 (50.0%)	0.0001
PTSD	16 (20.0%)	16 (29.1%)	5 (22.7%)	0.45
Schizophrenia/schizoaffective	1 (1.3%)	0 (0%)	1 (4.5%)	0.38
Bipolar	0 (0%)	0 (0%)	1 (4.5%)	0.14
Personality disorder spectrum*	1 (1.3%)	4 (7.3%)	3 (13.6%)	0.0327
Other psychiatric diagnoses	3 (3.8%)	6 (10.9%)	2 (9.1%)	0.20
Alcohol current	23 (28.8%)	14 (25.5%)	4 (18.2%)	0.63
Alcohol abuse history	34 (42.5%)	21 (38.2%)	8 (36.4%)	0.83
Illicit drug	24 (30.0%)	34 (61.8%)	12 (54.5%)	0.0008
Cognitive impairment	13 (16.3%)	23 (41.8%)	11 (50.0%)	0.0004
Adherence issues	16 (20.0%)	31 (56.4%)	16 (72.7%)	<.0001

BMI, body mass index; HTN, hypertension; DM, diabetes mellitus; LoT dialysis, length of time on dialysis prior to date of transplant evaluation; PTSD, posttraumatic stress disorder

^†All patients with psychiatric diagnoses

^‡Includes patients with anxiety, panic, and obsessive-compulsive disorders

^*Includes patients with anger management, borderline, antisocial and unclassified personality disorders

Multivariable modeling results for denial for purely psychosocial reasons were fit using demographic and psychosocial measures as predictors. Top multivariable models included: 1) “all psychiatric diagnoses” and PTSD, 2) age and cognitive impairment, and 3) age, PTSD, and cognitive impairment (Table 5).

Table 5.

Multivariable modeling of factors associated with denial for purely psychosocial reasons

Model Variables	Predictor*	Odds Ratio (OR)	OR 95% Conf Int Lower	OR 95% Conf Int Upper	p- value	AIC
Best two Order 2 Models
All psychiatric diagnoses PTSD	All psych	5.52	1.37	22.26	0.018	76.57
	PTSD	0.10	0.01	0.84	0.036
Age† Cognitive impairment	Age	0.94	0.88	0.99	0.028	77.24
	Cog. Imp.	4.29	1.21	15.15	0.026
Best two Order 3 Models
Age PTSD Cognitive impairment	Age	0.92	0.86	0.98	0.014	73.67
	PTSD	0.09	0.01	1.12	0.065
	Cog. Imp.	6.84	1.70	27.55	0.008
All psychiatric diagnoses PTSD Adherence issues	All psych	5.08	1.22	21.15	0.028	74.15
	PTSD	0.10	0.01	0.88	0.040
	Adherence	3.83	1.04	14.09	0.046

AIC, Akaike information criteria; PTSD, posttraumatic stress disorder

^*All non-age variables are dichotomous (yes/no) and are comparisons between statuses. For example, PTSD vs. No PTSD.

^†For age, the odds ratio estimates are based on a one year increase.

Following significant univariate assessments of depression and adherence on rejection for purely psychosocial reasons, we fit bivariate models both with and without an interaction effect. These additional comparisons showed that adherence was significantly associated with higher denial rates for purely psychosocial reasons (p<0.01). Depression corresponded to higher denial rates overall, but this relationship was not statistically significant. There was no significant interaction between depression and adherence issues (p=0.22).

3.6. Frailty and functional metrics among patients with or without psychosocial diagnoses

Frailty and functional metrics were compared between patients with and without psychiatric diagnoses, history of illicit substance abuse, smoking, alcohol abuse, cognitive impairment, and adherence issues and are summarized in the Supplemental Table. Troponin and BNP were higher in patients with poor adherence. Troponin, but not BNP, was lower among patient taking medications versus those with psychiatric diagnoses not on medications. BNP was also higher in patients with history of PTSD. Patients with depression walked fewer steps per day (2832 v. 3507, p=0.04). Hand grip was weaker in patients with cognitive impairment (28.5 v.1.1 kg, p=0.03). Patients with anxiety and personality disorders performed better, and smokers performed worse on the sit-stand test. Patients with depression, anxiety, PTSD, cognitive impairment, and poor adherence answered that they “could not get going” (CES-D exhaustion question) more days per week than patients without those diagnoses. No difference was measured in exhaustion in patients with personality disorder, smoking, and alcohol abuse history. No differences in the up-and-go test, treadmill (METS) and sit-reach tests were measured among patients with and without psychosocial diagnoses.

3.7. Frailty and functional metrics in patients denied for medical or psychosocial reasons

Overall tests for differences in the list, deny purely medical, deny med/psych, and deny pure psych groups were significant for all examined frailty metrics (Table 6). Except for the CES-D exhaustion question, patients denied for psychosocial reasons did not perform worse than listed patients on the metrics. The group denied for medical reasons had higher BNP (3017.5 v. 1350.5) and performed significantly worse than the group denied for purely psychosocial reasons on the up-and-go (6.73 sec v. 4.71 sec) test. Both medical and psychosocial denial groups had more days of exhaustion (CES-D exhaustion question) per week than listed patients, with psychosocial denial patients trending toward more days of exhaustion than patients denied for medical reasons.

Table 6.

Comparison of frailty and functional metrics among patients listed or denied for transplant

Frailty Measure N	List 208	Deny Medical Only 80	Deny Med/Psychosocial Combined 55	Deny Psychosocial Only 22	p (all 4 groups)	p (3 deny groups)	p (list vs. deny psychosocial only)
Troponin (ng/ml) median (IQR)	0.03 (0.01–0.07)	0.06 (0.02-0.12)	0.05 (0.03-0.11)	0.02 (0.01-0.12)	0.0027	>.99	0.14
BNP median (IQR)	1457 (485.5-2999)	3017.5 (1073.5-6178.5)	2987 (1208-12886)	1350.5 (482-1916)	<.0001	<.0001	0.96
Mean handgrip (Kg) mean (sd)	32.37 (9.91)	28.81 (8.32)	27.31 (8.59)	29.25 (7.38)	0.0054	>.99	0.72
Chair sit-stand (reps/30 sec) mean (sd)	15.22 (5.98)	11.66 (4.58)	12.10 (6.46)	15.18 (5.15)	<.0001	0.20	>.99
Up-and-Go (sec) median (IQR)	5.15 (4.30-6.30)	6.73 (5.35-8.28)	6.17 (5.20-7.90)	4.71 (4.10-5.20)	<.0001	<.0001	0.58
Sit-Reach (cm) median (IQR)	0.00 (0.00-6.50)	5.00 (0.00-10.00)	1.00 (0.00-10.00)	0.00 (0.00-6.00)	<.0001	<.0001	0.42
Days Exhausted/week * median (IQR) mean (SD)	0.00 (0.00-0.00) 0.63 (1.52)	0.00 (0.00-2.00) 1.19 (1.94)	0.00 (0.00-2.00) 1.53 (2.24)	0.00 (0.00-3.00) 1.73 (2.64)	<.0001	0.61	<.0001
Treadmill (METS) mean (sd)	5.96 (2.16)	3.70 (1.82)	4.42 (2.37)	5.14 (1.77)	<.0001	0.0102	0.56
Pedometer (steps/day) median (IQR)	3778.20 (2362.93-5488.80)	2427.97 (1342.12-3615.51)	2924.24 (1615.20-4289.97)	3443.63 (1958.40-4526.48)	0.0054	0.98	>.99

^*CESD Exhaustion question 2: How many days per week could you not get going?

4. DISCUSSION

Eighty-four percent of patients presenting for transplant evaluation at the Iowa City VA had psychosocial issues, which were a factor in 44% of the deferral decisions and 49% of the decisions to immediately deny patients for transplant. A quarter of immediate denials were for purely psychosocial reasons and patients deferred for purely psychosocial reasons were unlikely to progress to listing. Similar to a prior study⁵⁰ considering the effect of demographic, medical and psychosocial factors on access to transplant among Veteranidney transplant candidates, we did not find an interaction with race and listing outcome and our denied patients trended toward older age and more comorbidities. However, unlike Freeman et al., our study did show an association between depression and listing outcome. In addition to depression, cognitive impairment and poor adherence were the most significant psychosocial factors that were associated with overall transplant denial and denial for purely psychosocial reasons.

The 6.7% rate of current smoking among our candidates is in line with the 6.1% incidence reported among all US end-stage renal disease patients who completed Medical Evidence Forms between 2012 and 2014⁵¹ and is similar to the 5.5% and 5.8% rates of active smoking among candidate¹⁷and listed¹³ renal transplant patients. In contrast to Sandhu et al. who reported a limiting effect on transplant access related to active smoking, our results showed a trend toward transpant denial for smoking history, but not active smoking. This is perhaps both a reflection of our local practice of not counting smoking as an absolute contraindication to transplant and the known association of smoking with medical comorbidities which were frequent reasons for transplant denial. Our 68% rate of prior smoking history is consistent with the 77% rate of smoking history among Veterans captured by McKinney⁵² using the 1987 National Medical Expenditure Survey. However, it is known that smoking among Veterans has declined over time^53,54 and this trend coupled with coaching and cessation efforts to prepare patients for a successful evaluation perhaps contributes to the low rates of active smoking seen in our candidates.

The association of poor adherence and impaired cognition¹⁴ with listing denial is not unexpected as these issues, when not addressed, are considered contraindications to transplant due to reported associations with worse transplant outcomes.^55-59 The rates of nonadherence among our entire cohort and listed patients (24% and 13%, respectively) are similar to the 11% previously reported for candidates⁶⁰, but lower than those reported among general dialysis⁶¹ and post-kidney transplant recipients^59,62,63. Rates of low cognition among our evaluated (20%) and listed patients (13%) were lower than the 42% previously reported for listed patients¹⁴, and lower than the up to 70% rates⁶⁴ reported among dialysis patients which perhaps reflects bias in the referral process and differences in screening methods.

The association of depression with listing outcome is more unexpected as this is not a specific selection criterion. Other authors have previously reported the association of non-adherence and cognitive impairment with other mental health diagnoses including depression^10,60,65 and in our study 53% of patients with cognitive impairment and 42% of patients with poor adherence also had depression. However, our results suggest that poor adherence and not depression was the primary driver of listing denial. Of note, patients with psychiatric diagnoses taking medication, which we used as a potential marker for disease severity, were no more likely to be denied than unmedicated patients. Symptoms of social withdrawal associated with PTSD, depression, and anxiety disorder can impact patients’ ability to create and sustain supportive relationships.⁶⁶ This loss of social connectiveness can also exacerbate issues of adherence and may help explain the high rate of psychiatric diagnoses among nonadherent patients.

Fifty percent of the patients in this study had psychiatric diagnoses which is similar to the 47% reported previously by Trivedi et al. in their survey of mental illness among Veterans receiving primary care within the VA system¹⁸ and the 56% rate among Veteran kidney transplant recipients captured by Evans¹⁹. Also 25% of our patients were taking a medication for a psychiatric diagnosis which is comparable to the 24% rate of serious mental illness (SMI) previously reported among Veteran kidney transplant recipients¹⁹. PTSD, which has not been as well studied among kidney transplant candidates, affected 19% of our candidates and is on the upper end of the spectrum compared to prior reports of PTSD in VA populations^18,67,68 and is higher than the 4% rate of PTSD among post-kidney transplant Veterans reported by Evans¹⁹. Thirty-one percent of our candidates overall and about 26% of the listed patients were diagnosed with depression which is close to previously reported ranges for prevalent depression among dialysis (39%)² , waitlisted (33%-52%)^5,6, and post-kidney transplant (2.8% -31%)^5-10 patients. However, the rates of depression and anxiety (16%) in our cohort are higher than the moderate depression (2.8%) and anxiety (1.5%) assessed using the Brief Symptom Inventory (BSI) by Freeman and Myaskovsky in contemporaneous VA kidney transplant candidates.^5,50 The rates reported by Freeman and Myaskovsky are also low compared with the 25% estimated rate of depression in the general VA population and seem low considering that almost 18% of our patients were taking an antidepressant. Questionnaires are known to estimate higher depression rates⁶⁹, thus the low rates obtained via the BSI are surprising, but could perhaps be partially explained by undercounting patients receiving effective treatment⁶⁰. Additional studies may be necessary to clarify rates of mental illness among Veteran transplant candidates.

In contrast to work by Konel et al. showing a monotonic rise in depressive symptoms with increasing frailty among post-transplant patients we did not find an association between functional or frailty metrics and depression in our cohort of patients studied at the time of evaluation. With the exception of steps per day measured by pedometer, patients with depression in our study had similar performance on tests of strength and speed as non-depressed patients. The strong association of depression with feelings of exhaustion in our patients perhaps explains why patients with depression may have been less motivated to walk while still having the capacity to do so. Unlike the work by Konel, we did not grade the degree of depression and cannot say for sure if the group of patients denied for medical reasons who had worse functional performance may have also had worse depression. This seems unlikely though because patients denied for medical reasons, had fewer days of exhaustion (CES-D question) than patients denied for purely psychosocial reasons. Similar to the association shown by Konel and Dobbels between depression and younger age, patients in this study denied for psychosocial reasons where depression was more prevalent, were younger than in the group denied for medical reasons.⁸

Interestingly, patients with a history of poor adherence had higher troponin and BNP, and had more days they could not get going, indicating a potential interaction between medical and psychosocial factors contributing to patient adherence. Prior studies have shown higher cardiac mortality among nonadherent patients⁷⁰ and synergistic effects with depression and adherence on cardiac outcomes^71,72. As nonadherence is often multidimensional, nonadherence among renal failure patients also may involve missed dialysis sessions or larger interdialytic weight gains, which could exacerbate cardiac stress.^73,74 Given past reports of associations between cognitive abilities and cardiovascular health in renal failure patients^75-77, our results showing no association between BNP and troponin and cognitive function are perhaps unexpected, but could be influenced by the fact that the majority of our patients had only mild cognitive impairments, and should be considered in light of accumulating evidence that uremic neurotoxicity is also important in explaining cognitive decline in renal failure patients⁷⁸.

Our prior work⁴¹ showed that interactions predictive of listing outcomes for several functional, frailty and biochemical markers and exhaustion (as measured by CES-D question response) factored in all of our best multivariable models predicting listing decision outcome. However, our prior study did not parse our denial groups according to denial reason. As Konel’s and others’ work showed associations between frailty and depression we hypothesized that functional and frailty metrics would be associated with denial for purely psychosocial reasons. Performance in the majority of the metrics for patients denied for purely psychosocial reasons was not significantly different from listed patients and better than that of patients denied for medical reasons despite the association between depression diagnosis and listing denial. Patients denied for psychosocial reasons were exhausted more days per week than listed patients and those denied for medical reasons which may partly explain why the CES-D question was predictive in several of our previous multivariable listing models.⁴¹

Limitations of this study include those inherent in retrospective analyses. Since our results come from a single center study that included almost all male patients, our results may not be able to be extrapolated or be reflective of psychiatric diagnoses and psychosocial risk factors among all transplant patients. Also, our study did not systematically use specific screening tools to capture psychiatric diagnoses as done in other studies and instead captured information from three separate psychosocial providers’ assessments.⁹ However, this may not have led to mis-estimation of psychiatric diagnoses as psychiatric questionnaires are known to overestimate diagnoses when compared to interview techniques.⁶⁹ The true incidence of cognitive dysfunction may not have been detected as it is possible that not all patients were tested using an objective test. If centers screened patients with objective cognitive testing, then there is also the possibility of variability in screening methods. Also, as there is heterogeneity in objective cognitive tests used, there is likely inter-center variation in the definition and thresholds for the detection of cognitive dysfunction.

In conclusion, psychosocial issues were prevalent in this cohort of Veterans referred for kidney transplant affecting more than three quarters of our patient population. However, over 80% of the patients with a psychosocial diagnosis were eventually listed for transplant highlighting the importance of psychosocial support services and the psychosocial portion of the VA transplant evaluation. Transplant candidates with psychiatric diagnoses generally were not more functionally limited than those without psychiatric diagnoses or those denied for medical reasons. However, there was an association between poor adherence, cognitive impairment, and feelings of exhaustion. Also, patients denied for purely psychosocial reasons reported more exhaustion than patients denied for medical reasons. The transplant evaluation process could be improved by more thoroughly considering how concurrent problems interact. In consideration of the potential effects of exhaustion on adherence, an approach that more frequently leads to a deferral for additional mental health counseling and evaluation of the reasons for exhaustion, rather than denial, should be part of the process. As this work suggested that the exhaustion that patients denied for psychosocial reasons experience may be independent from depression, the VA system might benefit from future work that clarifies the way transplant candidate mental health diagnoses interact with factors such as exhaustion and adherence.

Data Availability Statement:

The data that support the findings of this study are available from the corresponding author (Daniel Katz, MD daniel-katz@uiowa.edu) upon reasonable request.