Center Variation and Risk Factors for Failure to Complete Six-Month Post-Donation Follow-up among Obese Living Kidney Donors

Abstract

Background:

Living kidney donors in the United States who were obese at donation are at increased risk of end-stage renal disease (ESRD) and may benefit from intensive post-donation follow-up. However, they are less likely to have complete follow-up data. Center variation and risk factors for incomplete follow-up are unknown.

Methods:

Adult living kidney donors with obesity (body mass index ≥ 30 kg/m2) at donation reported to the Scientific Registry of Transplant Recipients from January 2005-July 2015 were included (n=13,831). Donor characteristics were compared by recorded serum creatinine at 6-months post-donation, and multilevel logistic regression models were used to estimate odds of 6-month creatinine.

Results:

After adjustment, older age, female sex, and donation following implementation of new center follow-up requirements were associated with higher odds of 6-month creatinine, with lower odds for obese donors with a history of smoking, biologically-related donors, and at centers with higher total living donor volume. 23% of variation in recorded 6-month serum creatinine among obese donors was attributed to center (intraclass correlation coefficient: 0.232, p < 0.001). The adjusted probability of 6-month creatinine by center ranged from 10% to 91.5%.

Conclusions:

Tremendous variation in recorded 6-month post-donation serum creatinine exists among obese living donors, with high volume centers having the lowest probability of follow-up. Moreover, individual-level characteristics such as age, sex, and relationship to recipient were associated with recorded 6-month creatinine. Given increased risk for ESRD among obese living donors, center-level efforts targeted specifically at increasing post-donation follow-up among obese donors should be developed and implemented.

INTRODUCTION

Living kidney donation offers the best option for long-term survival for patients with end-stage renal disease (ESRD) with shorter time to transplant and superior outcomes.¹ Low absolute risks to living donors compared to non-donors have been reported,^2–4 but follow-up care is essential to monitor progress and identify potential problems early post-donation. In fact, a 2015 consensus conference highlighted the importance of living donor follow-up, to promote data collection ensuring the safety of the procedure and to improve informed consent of potential donors.⁵ Current United States policies only require transplant centers to follow living kidney donors for two years, with recent policy changes now requiring minimum thresholds for follow-up of living donors.⁶

Despite minimum compliance standards, post-donation follow-up remains poor. Prior to implementation, complete follow-up for laboratory data was 51%, 40%, and 30%, and clinical data was 67%, 60%, and 50% at 6-months, 1-year, and 2-years post-donation respectively.⁷ While the proportion of timely and complete donor follow-up increased in the post-implementation period, a 2017 study reported that only 43% of transplant centers met all thresholds for living donor follow-up.⁸ This study also showed a wide range in compliance by transplant center that persisted over time.

Additionally, these and other studies have demonstrated that having a body mass index (BMI) of 30 kg/m² or greater at donation is a risk factor for incomplete post-donation follow-up. Schold and colleagues demonstrated that obese donors had 1.3-fold higher odds for missing laboratory data at one year post-donation when compared to normal weight donors.⁷ Henderson/Thomas et al. reported a 1.2-fold higher odds of incomplete or untimely follow-up among obese donors,⁸ and we showed that obese donors were 2% less likely to have a serum creatinine measurement at 6-months post-donation compared to their non-obese counterparts.⁹ This non-compliance is problematic, given that we also recently demonstrated that living kidney donors in the US who were obese at time of donation were at greater risk of ESRD than their non-obese counterparts.¹⁰ Thus, these donors may benefit from more intensive post-donation follow-up, particularly as the prevalence of obesity among living donors continues to mirror the increase seen in the general population.¹¹ While variation in compliance with post-donation follow-up has been observed across transplant centers among all living donors, it remains to be seen whether there are risk factors for lack of follow-up for obese donors and whether center variation exists in follow-up among this subset of donors. Understanding these risk factors is critical to improving living donor follow-up among obese donors.

METHODS

Data Sources

This study used data from the Scientific Registry of Transplant Recipients (SRTR). The SRTR data system includes data system includes data on all donors, waitlisted candidates, and transplant recipients in the US, submitted by members of the Organ Procurement and Transplantation Network (OPTN). The Health Resources and Services Administration (HRSA), US Department of Health and Human Services provides the oversight to the activities of the Organ Procurement and Transplantation Network (OPTN) and SRTR contractors. We also used data from the OPTN to identify the recovering center, as the facility recovering the living donor kidney is responsible for donor follow-up. The study was approved by the University of Alabama at Birmingham Institutional Review Board (161212003).

Study Population

We identified all adult, living kidney donors with a BMI of ≥ 30 kg/m² at donation between January 2005-July 2015, at transplant centers recovering at least 10 obese living donors during the study period, which excluded 36 centers. We excluded donors who died within 6 months of donation, resulting in a final cohort of 13,831 donors from 192 transplant centers.

Outcome Ascertainment

Our outcome was defined as presence of a serum creatinine measurement at 6-months post-donation in the living donor follow-up file.

Statistical Analyses

Exploratory Data Analyses

Demographic and clinical characteristics within the cohort were explored initially by frequency and medians. Characteristics were then compared by presence of 6-month creatinine using chi-square for categorical variables and Kruskal-Wallis for continuous variables. We examined the proportion of all donors recovered at a center that were obese, and Spearman’s correlation coefficient was used to explore the relationship between center volume of obese donors and the proportion of donors at the center with a recorded 6-month serum creatinine.

Multivariable Analyses

We utilized a multilevel logistic regression model to estimate the odds of having a recorded 6-month serum creatinine measurement with a random intercept for recovery center, adjusting for both patient and center-level characteristics found to be significant on univariate analyses at p < 0.10. We also calculated the intraclass correlation coefficient (ICC) to estimate the ratio of between-center variation to total variation explained by the model. To estimate an adjusted percentage of donors with 6-month serum creatinine at each center, we used empirical Bayes estimates of the center-level random intercepts for a reference donor (male, aged 40, less than college education, unmarried, no history of hypertension, non-smoker, uninsured, unrelated to the recipient, not working for income, pre-compliance era, center total living donor volume of 650 over the entire study period).

Sensitivity Analyses

Due to missing data for some of the covariates, such as education and history of smoking, a complete case analysis was performed and reported (n=9,676). As a sensitivity analysis, all missing levels of data were coded as such to allow for inclusion in modeling. We also explored inclusion of the proportion of obese donors of all living donors at a center as a covariate and produced a full model with all covariates of interest. Inferences were consistent with the presented analyses. All analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC).

RESULTS

There were 13,381 adult living kidney donors between January 1, 2005 and July 1, 2015 (Table 1). Median age was 41 years (interquartile range (IQR): 33–49), and nearly 60% were female. The majority of donors were Caucasian (81.2%), with approximately half of donors having less than college education. Four percent of donors had a history of hypertension, and 25.2% reported a history of cigarette use. Two-thirds of donors had health insurance, and most were working for income at the time of donation (81.4%). Approximately half of obese donors were biologically related to the recipient, and 22.9% donated following implementation of donor follow-up compliance standards. Overall, 69.6% of the cohort had a 6-month serum creatinine recorded.

Table 1.

Characteristics of obese living kidney donors (January 2005-July 2015), N=13,831

Age in years, median (IQR)	41.0 (33–49)
Female sex, N(%)	8,049 (58.2)
Race/ethnicity, N(%)
African American	2,213 (16.0)
Caucasian	11,231 (81.2)
Other	387 (2.8)
Less than college education, N(%)	7,797 (56.4)
Marital status, N(%)
Married/partnered	8,549 (61.8)
Single/widowed/divorced	4,999 (36.1)
Pre-operative systolic blood pressure, median (IQR)	123.0 (115–132)
Pre-operative diastolic blood pressure, median (IQR)	76.0 (70–81)
History of hypertension, N(%)	522 (3.8)
History of cigarette use, N(%)	3,480 (25.2)
Insured at donation, N(%)	9,366 (67.7)
Low GFR (< 90 mL/min/1.73m2), N(%)	4,746 (34.3)
Open procedure, N(%)	853 (6.2)
Surgical length of stay in days, median (IQR)	3.0 (2–3)
Biologically related to recipient, N(%)	7,854 (56.8)
Working for income, N(%)	11,257 (81.4)
Post-compliance era, N(%)	3,172 (22.9)
Total center LD volume, median (IQR)	531 (273–1,068)
Six month creatinine recorded, N(%)	9,628 (69.6)

Characteristics of obese living donors by 6-month serum creatinine

Characteristics were also compared by presence of 6-month serum creatinine (Table 2). Donors who had a creatinine measurement were older (median of 42.0 years vs. 40.0 years, p > 0.001), more often female (59.3% vs. 55.7%, p < 0.001), less often African American (15.3% vs. 17.7%) and more often Caucasian (82.0% vs. 79.5%, p=0.002). Donors with 6-month creatinine were more often married, had a history of hypertension, were insured at donation, more likely to have a college education, and donated in the post-compliance era but were less likely to be biologically related. Additionally, median center living donor volume was lower among donors with 6-month creatinine (442 (IQR: 264–1,037) vs. 654 (IQR: 321–1,072), p < 0.001).

Table 2.

Characteristics of obese living kidney donors by recorded 6-month serum creatinine

Characteristic	6-month creatinine (N=9,628)	No 6-month creatinine (N=4,203)	p
Age in years, median (IQR)	42.0 (33–50)	40.0 (32–47)	< 0.001
Female sex, N(%)	5,708 (59.3)	2,341 (55.7)	< 0.001
Race/ethnicity, N(%)
African American	1,471 (15.3)	742 (17.7)	0.002
Caucasian	7,890 (82.0)	3,341 (79.5)
Other	267 (2.8)	120 (2.9)
Less than college education, N(%)	5,423 (63.4)	2,374 (67.5)	< 0.001
Marital status, N(%)
Married/partnered	6,140 (64.8)	2,409 (59.3)	< 0.001
Single/widowed/divorced	3,317 (35.3)	1,642 (40.6)	< 0.001
Pre-operative systolic blood pressure, median (IQR)	123.0 (115–132)	123.0 (116–132)	0.77
Pre-operative diastolic blood pressure, median (IQR)	76.0 (70–81)	76.0 (70–81)	0.75
History of hypertension, N(%)	401 (4.2)	121 (2.9)	< 0.001
History of cigarette use, N(%)	2,425 (25.7)	1,055 (25.9)	0.84
Insured at donation, N(%)	6,780 (84.2)	2,586 (82.3)	0.01
Low GFR (< 90 mL/min/1.73m2), N(%)	3,327 (34.8)	1,419 (34.2)	0.49
Open procedure, N(%)	589 (6.1)	264 (6.3)	0.71
Surgical length of stay in days, median (IQR)	3.0 (2–3)	3.0 (2–3)	< 0.001
Biologically related to recipient, N(%)	5,316 (55.2)	2,538 (60.4)	< 0.001
Working for income, N(%)	7,936 (85.3)	3,321 (84.4)	0.18
Post-compliance era, N(%)	2,648 (27.5)	524 (12.5)	< 0.001
Total center LD volume, median (IQR)	442 (264–1,037)	654 (321–1,072)	< 0.001

Unadjusted donor follow-up by center

Over the study period, the total number of obese living donors per center ranged from 10 to 507 across 192 transplant centers, with the majority of centers recovering between 15 and 50 obese donors (Figure 1). The proportion of donors recovered from a center that were obese ranged from 9 to 44%, and the proportion of donors who had a 6-month serum creatinine per center ranged from 20% to 100% (Figure 2). As the total center volume of obese donors recovered increased, the proportion of donors at a center with a 6-month serum creatinine decreased, as evidenced by a Spearman’s rho of −0.23 (p=0.002).

Figure 1.

Frequency of obese living donors by recovery center

Figure 2.

Scatterplot of total obese donors recovered per center by percent of donors with a 6-month serum creatinine

Multilevel Logistic Regression Analyses

Patient-level

After adjustment, several patient-level factors remained significantly associated with presence of 6-month serum creatinine (Table 3). Older age was associated with a 2% increased odds (aOR=1.02, 95% CI: 1.01–1.02, p < 0.001), and female donors had a 14% increase in the odds of creatinine at 6-months (aOR=1.14, 95% CI: 1.03–1.26, p=0.01). However, history of cigarette use was associated with an 11% decreased odds of follow-up (aOR=0.89, 95% CI: 0.79–1.00, p=0.04), and biological relationship to the recipient was associated with a 12% decreased odds (aOR=0.88, 95% CI: 0.79–0.98, p=0.02).

Table 3.

Odds of having 6-month serum creatinine measurement, with random intercept for recovery center

	Unadjusted			Adjusted
Characteristic	OR	95% CI	p	aOR	95% CI	p
Age, per 1-year increase	1.02	1.01–1.02	< 0.001	1.02	1.01–1.02	< 0.001
Female sex	1.18	1.07–1.31	< 0.001	1.14	1.03–1.26	0.01
Race/ethnicity (ref=Caucasian)
African American	0.93	0.81–1.07	0.32
Other	0.99	0.73–1.33	0.93
Less than college education	0.85	0.77–0.94	0.002	0.93	0.83–1.03	0.18
Married/partnered (ref=single/widowed/divorced)	1.27	1.15–1.41	< 0.001	1.12	1.00–1.25	0.05
Pre-operative SBP, per 1-unit increase	1.00	0.99–1.00	0.34
Pre-operative DBP, per 1-unit increase	1.00	0.99–1.00	0.16
History of hypertension	1.38	1.07–1.79	0.02	1.15	0.88–1.50	0.32
History of cigarette use	0.91	0.81–1.01	0.08	0.89	0.79–1.00	0.04
Insured at donation	1.29	1.13–1.49	< 0.001	1.07	0.92–1.24	0.37
Low GFR (< 90 mL/min/1.73m2)	1.02	0.92–1.14	0.69
Open procedure (ref=Lap)	0.83	0.62–1.09	0.18
Surgical length of stay in days, per 1-unit increase	1.00	0.97–1.02	0.82
Biologically related to recipient	0.78	0.71–0.87	< 0.001	0.88	0.79–0.98	0.02
Working for income	1.12	0.98–1.28	0.098	1.12	0.97–1.29	0.12
Post-compliance era	2.87	2.54–3.25	< 0.001	2.85	2.52–3.23	< 0.001
Total center LD volume, per 25-unit increase	0.98	0.97–0.99	0.001	0.98	0.97–0.99	< 0.001

^*intraclass correlation coefficient=0.232, meaning that 23% of variability in recording of 6-month serum creatinine is accounted for by center (p < 0.001)

Center-level

We also identified center-level factors that were associated with presence of 6-month serum creatinine after adjustment. Individuals who donated post-implementation of the new compliance standards in February 2013 had a 2.9-fold increased odds of 6-month serum creatinine (aOR=2.85, 95% CI: 2.52–3.23, p < 0.001). Conversely, for every 25-unit increase in total center volume of living donors recovered, the odds of 6-month creatinine decreased by 2% (aOR=0.98, 95% CI: 0.97–0.99, p < 0.001). The ICC was 0.232, suggesting that nearly 25% of the variation in 6-month follow-up can be attributed to differences across transplant centers. When we examined the adjusted predicted probability of 6-month creatinine by center, we found that the proportion of obese donors with a 6-month creatinine ranged from 10% to 91.5%, with an average of 49.5% of obese donors across all centers having 6-month serum creatinine (Figure 3).

Figure 3.

Predicted probability of 6-month creatinine measurement by recovery center

DISCUSSION

In this study of obese living kidney donors, we found that obese donors who are older and female were more likely to have a 6-month serum creatinine measurement, while obese donors who reported a history of smoking or were biologically related to their recipient were less likely to have 6-month follow-up. Center volume was also associated with odds of 6-month follow-up, and we observed wide center-level variation in the predicted probability of follow-up among obese donors, even after adjusting for both patient and center-level factors.

These findings mirror what has been observed among non-obese living kidney donors. Prior studies have shown higher odds of incomplete or untimely follow-up among donors who were male, younger, had a history of smoking, and were non-spousal family members, irrespective of BMI class.^7,8,12 However, donors with these characteristics who were also obese may face even greater challenges to complying with follow-up recommendations. A 2013 study of 39 primary care physicians demonstrated that the physicians developed less emotional rapport (defined as counts of communication behaviors related to empathy, concern, and reassurance) with patients who were overweight or obese compared to normal weight patients.¹³ This finding is troublesome, given that emotional rapport and empathy among clinicians and patients has been linked with patient satisfaction and adherence.^14,15 Literature from the general population suggests that perceived stigma and previous negative experiences with providers among obese individuals are associated with poor patient satisfaction and delay and avoidance of health care.^16,17 In addition, higher BMI was shown to be associated with lower physician respect for the patient, which may adversely affect the quality of care and subsequent health care utilization.¹⁸ Therefore, it is not surprising that obese donors may be less inclined to remain engaged with the transplant center post-donation if they have not received what they consider to be proper emotional support. However, no studies to date have explored the opinions of obese living donors about their experience or care at transplant centers, and it is unknown how conversations about weight loss or maintenance at the time of evaluation or follow-up may further impact donor compliance with the requirement. Finally, the donation episode may be the only encounter with health care for some of these donors, as demonstrated by Alejo et al. in a cohort of prior living donors, ¹⁹ and may indicate a general lack of engagement with one’s health care and long-term health maintenance.

We found that both higher center volume of obese donors and total center volume of living donors were negatively associated with recording of 6-month serum creatinine. A 2009 survey of transplant centers demonstrated that the primary barriers to providing living donor follow-up were donor inconvenience and cost concerns.²⁰ These time and resource burdens increase as the number of individuals expected to meet compliance standards increases at a center. The United Network for Organ Sharing (UNOS) recently added a visual tool that permits transplant centers to track their compliance with donor follow-up, including due dates for upcoming forms and whether there are missed data fields.²¹ Use of this tool, in conjunction with center-initiated efforts (such as wristbands and reminder cards for donors) may serve to increase compliance rates across centers. Furthermore, while these results suggest the need for intervention at the center level, the intraclass correlation coefficient and the model results demonstrate that patient-level differences remain, even after accounting for the transplant center. As such, efforts targeted at subgroups of donors, including those with obesity who have been shown to be at even greater risk of loss to follow-up than their normal weight counterparts, are especially necessary as the utilization of obese donors continues to increase to meet the gap in supply and demand.

Center variation in donor follow-up has been observed by others.⁸ Variation by centers in acceptance of donor characteristics that impact likelihood of follow-up, such as insurance status, may contribute to this variation in the outcome. While the presence of center variation in the utilization of uninsured living donors has not been reported to our knowledge, a 2016 study of health insurance trends in living donors showed that those without insurance were slightly more often obese than non-obese (24% of all uninsured donors were obese vs. 22% of donors with insurance, p=0.004).²² Thus, it is not surprising that variation is seen when limited to the subset of donors with obesity. Interestingly, a survey of transplant centers conducted by Waterman et al. published in 2013 reported that only 68% of respondents from kidney transplant programs believed that follow-up with living kidney donors was a high priority.²³ While most (92% of respondents) reported that their center informed donors about the requirement for follow-up, only 67% reported having specific plans for achieving follow-up with donors. Per UNOS, currently there are no penalties for failing to meet compliance standards- rather these are recommendations for transplant programs’ voluntary use. If follow-up remains poor, it may become necessary to implement consequences for failing to comply with these standards, but this additional regulatory burden may unintentionally discourage the use of living donor transplantation.

Additionally, our finding that centers with a high volume of living donors were less likely to complete follow up of their obese living donors is important; these centers are among those most likely to accept this “higher risk” type of donor and this provides an opportunity to identify centers “at risk” for incomplete follow up. Given the concerns of donation-related risks¹⁰ and those associated with prolonged exposure to overweight and obesity, it may be beneficial for transplant centers to outline plans in conjunction with OPTN/UNOS for prioritizing follow-up of these and other medically complex donors. In addition to the use of existing resources provided by UNOS/OPTN, the transplant community may benefit from roundtable discussions or lunch and learn sessions at national meetings, in which high volume centers that are compliant with follow-up requirements share their experience and strategies for achieving compliance goals.

Our study is not without limitations. This was a retrospective analysis using registry data, therefore, we were restricted by the available information and cannot rule out unmeasured confounding. Moreover, we cannot account for whether the lack of medical follow-up was due to the failure of a center to contact the donor or the donor’s failure to respond. However, these data contribute novel information about obese donors who were lost to follow-up in early stages post-donation.

In conclusion, as we continue to utilize donors with a BMI of 30 kg/m² or greater, it is important to recognize not only their risk for loss to follow-up but also their need for long-term observation due to potential health outcomes. Development of center-specific efforts, in addition to accommodations from the regulatory agencies requiring follow-up, may help to improve compliance not only among obese donors but among all living donors nationwide.