Effectiveness of a Mobile Health System on Compliance with Two-Year Living Kidney Donor Follow-Up in the United States

Abstract

Background:

Organ Procurement and Transplantation Network (OPTN) policy requires two years of follow-up for living kidney donors (LKDs); however, many transplant hospitals struggle to meet this requirement. We developed and tested a mobile health (mHealth) system for LKD follow-up in a pilot randomized-controlled trial (RCT).

Methods:

LKDs were randomly assigned to either the intervention (mHealth + standard of care) or control arm (standard of care). We assessed OPTN policy-defined completeness and timeliness of 6-month, 1-year, and 2-year follow-ups. 400 LKDs were enrolled in the study (6/2018 – 2/2021).

Results:

At 6-month follow-up, a higher proportion of the intervention arm participants completed composite visits (97.5% vs. 91.5%, p=0.01). Both arms had similar compliance rates at 1 and 2-year follow-up (92.0% vs. 89.5%, p=0.49, and 66.5% vs. 65.0%, p=0.83). Intervention arm participants completed 6-month follow-up 11 days earlier than their counterparts (p=0.009).

Conclusion:

mHealth technologies improved 6-month follow-up, but did not impact 1- and 2-year LKD follow-up in this single-center RCT. Other strategies, such as providing services beyond data collection, may be necessary to improve donor engagement and support LDK long-term follow-up.

INTRODUCTION

Roughly 6,500 living kidney donors (LKDs) enter the United States (US) transplant system each year.¹ Living kidney donation (LKD) is the optimal treatment for end stage kidney disease (ESKD).^2–4 However, LKD is not without risk – kidney health, surgical, psychosocial, and financial risks have all been documented.^5–10 Robust characterization of donor risks is limited by the low availability of post-LKD data; this limits the information presented to LKD candidates during the informed consent process.^11,12 US Organ Procurement and Transplantation Network (OPTN) policy requires transplant hospitals to collect and report clinical and laboratory data at 6 months, 1 year, and 2 years post-LKD. LKDs can complete the OPTN follow-up form and submit laboratory data electronically.¹³ However, fewer than half of transplant hospitals meet the policy-defined thresholds for LKD follow-up.^7,14,15 Solutions are needed to facilitate timely and complete LKD follow-up data collection, help transplant centers maintain contact with and engage LKDs, mitigate inconvenience, and centralize LKD data.¹⁶

Mobile health (mHealth) technologies have been increasingly used to facilitate data collection and engage patients in chronic disease management.^17,18 A systematic review assessing mHealth tools’ impact on treatment adherence in a chronic disease management setting revealed that about half of mHealth tools improved adherence.¹⁷ For example, text messaging has been associated with improved appointment attendance and reduced cancellations.¹⁸ Orandi et al.’s survey of 110 LKDs found that >20% of LKDs who completed both their 6- and 12-month follow-up laboratory visits forgot to have their labs drawn until reminders were sent.¹⁹ In an exploratory study of LKD follow-up, we observed a high response rate to emails and/or text messages (72% response at 2-years post-donation).²⁰ Automated mHealth tools have the potential to lessen the LKD follow-up burden for donors and transplant hospital staff, and may be particularly appealing to transplant centers since there are no reimbursement mechanisms for policy-required donor follow-up care.²¹

We developed a mHealth tool to lessen the LKD follow-up burden and encourage LKD engagement. The mHealth was comprised of a patient-facing smartphone app and a provider-facing web portal. The smartphone app allowed LKDs to fill out the OPTN follow-up questionnaire, record lab values, and upload a photos of lab work. We conducted a pilot randomized controlled trial (RCT) to evaluate the impact of this mHealth tool on LKD follow-up rates and to inform the development and implementation of a centralized LKD registry (NCT03400085). This study was reviewed by the JHU Institutional Review Board (IRB00162212).

METHODS

Study Procedure

Trial design, including power calculations, and protocols are discussed in detail elsewhere. Based on our power calculations, we aimed to enroll 400 LKDs to detect a difference of 14% between the intervention and control arm 2-year compliance.²² We recruited patients who underwent LKD at Methodist Hospital Specialty and Transplant, San Antonio, TX, the US hospital with the highest volume of LKDs at the start of recruitment. We excluded LKDs who did not speak English or own a smartphone device. Recruitment occurred between June 2018 and February 2021. In this two-arm study, a blinded team member (AGT) used block randomization (block size ranging from 2–8), which improves the probability of having balanced arms throughout the study,²³ to assign participants to a study arm. This team member remained blinded until statistical analyses were complete. Another study team member used the numbered list with randomized assignments to create sealed envelopes. Transplant center staff then opened the next labeled envelope to allocate participants to either the control or intervention arms of the study. Each LKD had a 50% chance of being in the intervention arm. Participants in the intervention arm received the mHealth intervention in addition to standard follow-up procedures, while participants in the control arm received the standard follow-up.

Transplant hospital study personnel obtained a list of all patients undergoing donor nephrectomy and approached these patients for participation in the study. Written informed consent was obtained during the first post-operative clinic visit. After enrollment, study personnel helped participants assigned to the intervention arm download the application and explained its functions. Participants in the intervention arm were asked to use the mHealth tool to complete the OPTN questionnaire and submit laboratory values at 6 months, 1 year, and 2 years post-donation. Those in the control arm were instructed about required follow-up and told they would be contacted by phone, as is the standard of care for the transplant center (Figure 1).

Figure 1.

CONSORT Diagram

The primary outcome was the rate of complete (i.e., all components addressed) and timely (i.e., 60 days before or after the expected visit date) submission of required clinical (OPTN questionnaire) and laboratory follow-up data at each of the required follow-up visits. We collected UNOS ID, sex, and date of LKD to link participants to the Scientific Registry of Transplant Recipients (SRTR) for information on demographic and other characteristics.

Statistical Analysis

This study used data from the Scientific Registry of Transplant Recipients (SRTR). The SRTR data system includes data on all donor, wait-listed candidates, and transplant recipients in the US, submitted by the members of the Organ Procurement and Transplantation Network (OPTN). The Health Resources and Services Administration (HRSA), U.S. Department of Health and Human Services provides oversight to the activities of the OPTN and SRTR contractors. This dataset has previously been described elsewhere.²⁴

Using Fisher’s Exact test, we compared clinical (OPTN questionnaire), laboratory, and composite compliance at 6-months, 1-year, and 2-years for the control and intervention arms. Variation in average duration between visits among both arms was analyzed using Student T- and Mann-Whitney tests, depending on data normality. A p-value <0.05 was considered significant in primary analyses. In sensitivity analyses we used logistic regression (outcome: compliant follow-up) and adjusted randomization and potential confounders. Statistical analysis was conducted using Stata (version 15, College Station, TX) and RStudio (Boston, MA). Descriptive statistics were reported as counts (percentages) or medians (interquartile ranges [IQR]), as appropriate. Continuous values were compared using the Wilcoxon rank-sum test and Pearson’s chi-squared test for binary and categorical variables. Figures were created in RStudio using ggplot2, dplyr, tidyverse, hrbrthemes, viridis and scales packages.

RESULTS

Study Population

Four hundred LKDs were enrolled in this study. There were no statistical differences in baseline characteristics (Table 1). Most participants were female (71%), White (92%), Hispanic (55%), US citizens (98%), and employed (95%). Of the 200 mHealth intervention arm participants, 147 (75%) were female, 184 (92%) were White, and 115 (57%) were Hispanic. Median (IQR) age at enrollment was 39 (30–47) years. Less than half (88, 44%) were biologically related to their recipient. Most mHealth participants were U.S. citizens (98%) and employed (95%); 161 (81%) reported having health insurance at the time of enrollment into the study. Highest education level varied, with 19% having earned a high school diploma or GED, 36% having attended college or technical school, and 41% having obtained a college degree or higher.

Table 1.

Characteristics of the Study Population at Baseline.

Donor Characteristics	Control N=200	mHealth N=200	p
Age (Years), Median (IQR)	36 (29, 46)	39 (30, 47)	0.4
Female, N (%)	137 (69)	147 (74)	0.3
Race, N (%)
White	184 (92)	184 (92)	0.7
Asian	3 ( 2)	4 ( 2)
Black	8 ( 4)	7 ( 4)
Multiracial	2 ( 1)	0 ( 0)
Native American	1 ( 1)	2 ( 1)
Pacific Islander	2 ( 1)	3 ( 2)
Hispanic, N (%)	103 (52)	115 (57)	0.2
US Residency Status, N (%)
US Citizen	195 (98)	196 (98)	0.6
Non-US Citizen/US Resident	4 ( 2)	2 ( 1)
Non-US Citizen/Non-US Resident	1 ( 1)	1 ( 1)
Non-US Citizen/Non-US Resident, Traveled to US for Transplant	0 ( 0)	1 ( 1)
Functional Status, N (%)
100% - Normal, no complaints, no evidence of disease	194 (97)	194 (97)
90% - Able to carry on normal activity	1 ( 1)	3 ( 2)
Unknown	5 ( 3)	3 ( 2)	0.5
Living Donor(s) Relationship to Recipient, N (%)
Biological, blood related Parent	13 ( 7)	8 ( 4)	0.7
Biological, blood related Child	34 (17)	32 (16)
Biological, blood related Full Sibling	22 (11)	32 (16)
Biological, blood related Half Sibling	2 ( 1)	1 ( 1)
Biological, blood related Other Relative: Specify	11 ( 6)	15 ( 8)
Non-Biological, Spouse	16 ( 8)	14 ( 7)
Non-Biological, Unrelated: Paired Donation	60 (30)	63 (32)
Non-Biological, Other Unrelated Directed Donation	42 (21)	35 (18)
Highest Education Level, N (%)
None	1 ( 1)	2 ( 1)	0.4
Grade School (0–8)	2 ( 1)	0 ( 0)
High School (9–12) Or GED	43 (22)	38 (19)
Attended College/Technical School	70 (35)	71 (36)
Associate/Bachelor Degree	47 (24)	61 (31)
Post-College Graduate Degree	29 (14)	20 (10)
Unknown	8 ( 4)	8 ( 4)
Health Insurance Coverage (Yes), N (%)	158 (79)	161 (81)	0.5
Employed (Yes), N (%)	190 (95)	190 (95)	0.6

Visit Completeness and/or Compliance with Data Submission

At 6 months, a statistically significantly higher proportion of participants in the mHealth intervention arm completed visits as compared those in the control arm (97.5% vs. 91.5%, p=0.01) (Table 2). However, there was not a statistically significant difference in the visit compliance rates at 1-year (92.0% vs. 89.5%, p=0.49) and 2-years (66.5% vs. 65.0%, p=0.83). Similarly, for both clinical (OPTN questionnaire) and laboratory data, the mHealth intervention arm had higher compliance (complete and timely) rates at 6-months (99.5% and 97.5%, respectively) and 1-year (93.0% and 92.0%) as compared to the control arm (97.0% and 91.5% for clinical visits, and 94.5% and 89.5% for laboratory visits). At 2-years both the mHealth arm and the control arm had similarly low composite, clinical and laboratory compliance rates of 66.5% and 65.0%, 67% vs. 68.0%, and 66.5% vs. 65.0%, respectively (Table 2). In sensitivity analyses, older age (p<0.01) was associated with greater compliance.

Table 2:

Compliance in mHealth participants and controls.

	mHealth			Control
Compliance Factor	6-Month	1-Year	2-Year	6-Month	1-Year	2-Year
Composite Compliance	195 (97.5)	184 (92.0)	133 (66.5)	183 (91.5)	179 (89.5)	130 (65)
Clinical Compliance	199 (99.5)	186 (93.0)	134 (67.0)	194 (97.0)	189 (94.5)	137 (68.5)
Laboratory Compliance	195 (97.5)	184 (92.0)	133 (66.5)	183 (91.5)	179 (89.5)	130 (65.0)

Results are reported as N (%). Statistically significant results are highlighted in bold.

Timeliness of Required Follow-Up Contact

We examined variation in the number of days from donation to follow-up visit (6-month, 1-year and 2-year) across the two arms. Those in the mHealth intervention arm reached out for their first 6-month follow-up visit sooner than their counterparts in the control arm (95% CI: 155, 164 vs. 162, 171 days; p=0.03). These findings were confirmed by a Wilcoxon rank-sum test (z = 2.6, p=0.01), indicating that participants who received a message via the mHealth tool 6-months post-donation sought medical follow-up earlier than those who did not. Similarly, time to 1-year post-donation follow-up was shorter and statistically significant in the mHealth intervention arm. On average, participants in the mHealth intervention arm had their 1-year follow-up 6 days earlier than their counterparts in the control arm (95% CI: 346, 355 vs. 353, 361 days; p=0.02). In contrast, the number of days from donation to follow-up was greater for those in the intervention arm at 2-years post-donation, marking a shift from the 6-day lead in follow-up time observed at the 6-month and 1-year follow-up. Those in the mHealth intervention arm completed follow-up 719 days post-donation while those in the control arm completed follow-up 715 days post-donation for 2-year follow-up (95% CI: 714, 724 vs. 710, 718 days; p=0.9).

DISCUSSION

In this RCT of a mHealth tool to support LKD follow-up, the mHealth tool improved compliance with 6-month donor follow up; however, it did not improve OPTN policy compliance at 1 or 2 years after LKD. Importantly, this effect occurred in the context of very high compliance (>90% 6-month and 1-year compliance among all participants) relative to the average transplant center in a registry-based study.⁷ The effect of the mHealth intervention may be greater in centers with lower baseline levels of policy-compliant donor follow-up, and may be artificially inflated as the mHealth tool eases the burden of transplant center personnel, allowing them to use their time more effectively. Moreover, since the policy requires timely submission, improvement at 6 months has policy relevance. This may be important for centers who struggle to submit LKD donor follow-up within the 120-day window prescribed by policy. These findings have practical implications for both transplant hospital staff and living donors, specifically regarding engagement and communication with LKDs post-donation, addressing barriers to donor follow-up, streamlining the data collection process, and for ultimately understanding and characterizing long-term post-donation risks.

However, our results suggest that focusing on donor data collection may be insufficient for long-term LKD engagement. The low compliance rate among both arms observed at the 2-year mark warrants further study to identify the underlying cause(s) of this drop-off. Improved donor education on the importance of follow-up, incentives, complementary services, or alternative strategies are necessary to improve 2-year follow-up. Given the potential added burden of these strategies, avenues for reimbursement may be necessary.

Those in the mHealth intervention arm received their 6-month follow-up care earlier their counterparts, suggesting that mHealth tools, specifically automated SMS text and email messaging and push notifications, have the potential to increase follow-up compliance. Transplant hospitals, especially those performing large numbers of LKD transplants, should employ multiple tools to improve LKD engagement. Donor follow-up has been shown to help transplant hospital staff maintain relationships with LKDs.

Sensitivity analyses revealed that older age was associated with an increase in follow-up visit compliance, consistent with prior studies that identify younger age as a risk factor for donor non-adherence with laboratory follow-up visits.¹⁴ Because of this, transplant hospital staff should take each LKD’s age at the time of donation into account when speaking to LKDs about required follow-up visits. Future research should attempt to identify and overcome age-specific barriers to LKD follow-up to improve follow-up rates among younger LKDs.

This study is not without limitations. This was a single-center study which limits generalizability; however, outcomes in majority Hispanic donor populations are under reported LKD literature. Prior studies have highlighted the need to further study racial disparities in medical conditions among LKDs.²⁵ Future research should assess strategies to facilitate follow-up among a more racially diverse population of LKDs.

Complete and timely submission of clinical and laboratory LKD follow-up data is critical to understanding and monitoring the safety of LKDs. These findings can be used to refine current OPTN LKD follow-up practices and inform future use of mHealth technologies to promote follow-up and engagement. mHealth tools focused only on data collection for OPTN policy compliance may be insufficient to meet 2-year LKD follow-up goals. Future efforts should consider providing other services and improving donor education to improve donor engagement.

ABBREVIATIONS

ESKD

End-Stage Kidney Disease

HHS

Department of Health and Human Services

HRSA

Health Resources and Services Administration

LKD

Living Kidney Donor/Donation

mHealth

Mobile Health

OPTN

Organ Procurement and Transplantation Network

SRTR

Scientific Registry of Transplant Recipients

US

United States

Data statement:

The data that support the findings of this study are available from the corresponding author upon reasonable request.