Promoting Medication Adherence from the Perspective of Adolescent and Young Adult Kidney Transplant Recipients, Parents, and Health Care Professionals: A TAKE-IT TOO study

Abstract

Background.

Medication nonadherence is an important factor limiting allograft survival after kidney transplantation in adolescents and young adults. Some interventions, including the Teen Adherence in Kidney Transplant Effectiveness of Intervention Trial (TAKE-IT), showed some success in promoting adherence but the potential for scalability and use in routine clinical practice is limited. We applied user-centered design to gather the perspectives of recipients, parents, and health professionals concerning their needs, challenges, and potential intervention strategies to design an optimal, multi-component medication adherence intervention.

Methods.

The qualitative study was conducted at 4 Canadian and 3 American kidney transplant programs. Separate focus groups for recipients, parents, and health professionals were convened to explore these stakeholders’ perspectives. Directed content analysis was employed to identify themes that were shared vs. distinct across stakeholders.

Results.

All stakeholder groups reported challenges related to taking medications on time in the midst of their busy schedules and the demands of transitioning toward independence during adolescence. The stakeholders also made suggestions for the multi-component behavioral intervention, including an expanded electronic pillbox and companion website, education materials, and customized digitized features to support shared responsibility and communication among recipients, parents, and health professionals. Several suggestions regarding the functionality and features of the potential intervention reported in this early stage will be explored in more depth as the iterative process unfolds.

Conclusions.

Our approach to actively involve all stakeholders in the process increases the likelihood of designing an adherence intervention that is truly user-informed and fit for the clinical setting.

Introduction

Robust gains in patient and graft survival have been achieved after pediatric kidney transplantation.^1,2 However, graft failure rates begin to rise around 11 years of age and peak in young adulthood, regardless of the recipient’s age at transplant.³ A systematic review focused on adherence to the immunosuppression regimen⁴ and its impact on outcomes after kidney transplant found that adolescents and young adults (AYAs) are at particularly high risk for nonadherence.⁴ Furthermore, 44% of graft losses and 23% of late acute rejection episodes among AYAs were associated with medication nonadherence. Thus, nonadherence is considered the most important modifiable factor contributing to poor graft survival among AYAs.

A qualitative study of adolescent transplant patients’ and parents’ perceptions of barriers to medication adherence⁵ led to the development of measures to assess such barriers, and confirmed that barriers are important correlates of adolescents’ clinical outcomes. ⁶ Subsequently, Tong and colleagues ⁷ conducted a systematic review of qualitative studies to better understand the experiences of adolescent recipients after organ transplantation. Results revealed that medication nonadherence remains a challenge for AYAs. Most recently, a systematic review of quantitative studies found inconsistent evidence of the relationships between psychosocial factors and medication nonadherence risk in pediatric transplantation, with the exceptions that nonadherence was indeed associated with adolescence, racial/ethnic minority status, and presence of mental health issues.⁸

Patients, families, and healthcare providers (HCPs) share responsibility for the promotion of medication adherence to achieve optimal health outcomes after kidney transplantation; thus a comprehensive understanding of the needs and perspectives of all of these stakeholders is critical. However, barriers to adherence are rarely systematically assessed in the clinical setting and patients’ and parents’ perspectives may not be actively sought during time-limited clinical encounters.⁹ Thus, despite advances in understanding the barriers to adherence, there remains a need for effective, practical interventions for use in real-world clinical settings.

In fact, a variety of adherence-promoting interventions have been developed and tested. They have been evaluated almost exclusively in adults, or general chronic conditions in adolescents, and range from coaching to monitoring using electronic devices (e.g., pill caps) and pharmaceutical refills.^10–12 Some interventions incorporated education, counseling, and financial support.^11,12 In general, interventions employing a single strategy (e.g., education) are not effective.^13–15 Multicomponent interventions have a greater likelihood of impact,^12,16,17 However, given their complexity, it is unlikely that they could be disseminated for long-term use by patients and transplant programs.¹⁸

In contrast, few immunosuppressant adherence-promoting interventions have been tested in AYAs, and show limited evidence of efficacy.¹⁹ The Teen Adherence in Kidney Transplant, Effectiveness of Intervention Trial (TAKE-IT) was the first randomized trial of a multicomponent intervention in this area. Participants assigned to the TAKE-IT intervention used an electronic pillbox (ePillbox) and worked with a coach to facilitate medication adherence and minimize adherence barriers. Compared with usual care, participants receiving TAKE-IT were over 66% more likely to take medications and take them on time.²⁰ However, the ePillbox was not portable, the coaching was labor-intensive, and neither the ePillbox nor the coaching were integrated into transplant programs’ plan of care. Thus, the intervention’s usability, scalability, and dissemination into routine clinical practice were unlikely. The Teen Adherence in KidnEy transplant Improving Tracking To Optimize Outcomes (TAKE-IT TOO) study seeks to make the promising intervention more scalable and the present research reports on this effort.

Qualitative research allows stakeholders to share their perspectives, thereby generating relevant, in-depth and often compelling insights to inform care and adherence interventions.²¹ Ultimately, this approach may help improve the outcomes they regard as critical for patient care through integration of their perspectives into adherence intervention design. User-centered design is an approach that involves stakeholders throughout the development process so that technologies support their tasks and are easy to use and adopt. ^22,23 The more accurate the assessment of user requirements, the higher the level of user acceptance and adoption.²⁴ User-centered design has successfully promoted the usability of health technologies for transplant populations. ^25,26 To adapt and develop new components of the TAKE-IT intervention to meet stakeholders’ needs and facilitate dissemination, it was critical to involve AYAs, their parents, and HCPs.

In sum, while past work has explored barriers and interventions related to adherence, our study is unique in that we sought the views of all key stakeholders in the posttransplant care process to better understand their needs, and we undertook the work so that we could refine the Take-IT technology-based adherence intervention to make it more usable, scalable and feasible for use in routine clinical practice. We thus conducted stakeholder focus groups to: a) identify the challenges and strategies related to medication adherence in AYAs, b) determine stakeholders’ needs related to strategies to promote adherence, and c) explore stakeholders’ views on how to modify possible ePillboxes and coaching efforts (such as those from the TAKE-IT intervention) to address these challenges and needs.

Materials and Methods

Study Design and Setting

We conducted a qualitative study in which each participant attended a single focus group meeting. Separate focus groups were convened with AYAs, parents, and HCPs between January- August 2017 at four Canadian programs (Montreal Children’s Hospital of the McGill University Health Centre; Centre Hospitalier Universitaire Ste. Justine; The Hospital for Sick Children; British Columbia Children’s Hospital) and three U.S. programs (University of Pittsburgh Medical Center (UPMC)/UPMC Children’s Hospital of Pittsburgh; St. Louis Children’s Hospital; and Seattle Children’s Hospital).

Sampling

AYAs were eligible if they were beyond three months after kidney transplantation and aged 12–24 years. Patients with neurocognitive disabilities severe enough to preclude participation, those unable to communicate in French or English, or with multiorgan transplants were excluded. One parent per AYA (12–17 years) was eligible to participate. AYAs and parents could participate and was not contingent upon both AYA and parent participation. Focus groups of HCPs included pediatric and adult transplant team members representing a range of disciplines.

We convened separate groups for AYAs aged 12–14, 15–17, and 18–24, and parents of AYAs aged 12–14 and 15–17. Our rationale was to maximize homogeneity in developmental stage within the groups in order to promote engagement and interaction with peers. Purposive sampling ²⁷ was used to ensure that we included representation from all AYA age groups and their parents, and HCPs across a range of professional disciplines and years of experience.

Once eligible participants were identified, convenience sampling was used to recruit participants until we were able to schedule each focus group. Across all sites, of 57 AYAs approached for participation, 11 (19%) refused (citing lack of interest or time). Of 46 willing to participate, 32 (70%) were enrolled (including 19 aged 12–17 and 13 aged 18–24); the remaining 14 could not attend due to scheduling issues. Of 26 parents who were approached, 1 (4%) refused (citing lack of time); 19 (76%) were enrolled; scheduling issues precluded the remaining 6 parents from participating. Data for one parent focus group was lost due to technical difficulties leaving a final sample of 15 parents for inclusion in the analysis. Finally, across all sites, of 65 HCPs approached, 3 (4%) refused (citing lack of time); of the remaining HCPs, 46 (74%) were enrolled; the remainder could not participate due to scheduling or other issues.

Procedures

Each site’s institutional review board approved the study. Transplant program personnel approached eligible potential study participants. If they were interested, their names were given to the researchers, who contacted them to describe the study. AYAs aged ≥18 provided written informed consent. Parents of AYAs < 18 provided written consent for themselves and their child, and the younger AYAs provided written assent. Each study participant received $25, refreshments, and parking reimbursement.

To encourage the participants to speak openly, focus group facilitators were not involved in the AYAs’ clinical care. An additional study team member observed group sessions to ensure that all topics in the script were addressed and to assist with logistical issues. To ensure consistency across study sites, facilitators and observers underwent web-based training that provided a study overview, focus group goals, their roles and responsibilities, a review of the semi-structured scripts and completed a training competency checklist at the conclusion of training.

Measures

Before their focus group began, AYAs and parents completed a self-report questionnaire about sociodemographic and transplant-related characteristics. HCPs completed a questionnaire about sociodemographics, professional discipline, and years of experience.

In consultation with patient and parent co-investigators (AD, AG), we designed stakeholder-specific, semi-structured focus group scripts (see supplemental information). Topics included: 1) challenges and strategies for medication adherence; 2) views about possible interventions to improve adherence including the use of a personal coach and the development of action plans; and 3) opinions about using an ePillbox and a website to track medication-taking and to integrate medication-taking data into routine transplant program clinical care.

Analysis

Analyses were descriptive. For questionnaire data, summary statistics (means ± SD, medians with IQR, proportions) were used to describe sample characteristics.

The 90 to 120-minute focus group sessions were audio recorded and uploaded via a secure server to the User-Centered Design and Qualitative Analysis Center (UCD-QAC) established for this study at the University of Pittsburgh where analyses were performed by co-investigators and staff with qualitative analysis expertise. Professional transcriptionists transcribed audiotapes. Focus groups conducted in French were transcribed and translated into English. Transcripts, with participant names redacted, were reviewed for accuracy. Observer notes were reviewed to add context to transcripts. The transcript documents were transferred to Atlas.ti (Scientific Software Development GmbH) to organize, sort, link, and code the text.

We used the directed approach to qualitative content analysis ²⁸ to systematically code and interpret the text data using an a priori coding scheme based on the major topics that were addressed in the focus group scripts. ²⁹ We then segmented the focus group text into units of analysis (quotes) and tagged them with meaningful codes. Initial coding was performed separately by two members of the UCD-QAC team. Any discrepancies were discussed until consensus was reached. Code labels were sorted and clustered under themes based on linkages within the overall categories. ³⁰ In addition to these coding procedures, coded transcripts were randomly selected for an additional level of review by two other co-investigators. Any discrepancies in coding were discussed and reconciled. The process was iterative until descriptive redundancy was reached, (i.e., consensus that the coding scheme captured the data and no new descriptive codes emerged). ³¹

Results

Participant Characteristics

A total of 32 AYAs participated in 9 focus groups, and 15 parents participated in 4 groups. Their demographic and transplant-related characteristics are in Table 1. AYAs ranged from 13–24 years of age, almost half were female, and a majority was non-Hispanic white. Six of the 32 AYAs had also participated in TAKE-IT. None of the 15 parents had a child who participated in TAKE-IT. Characteristics of the 46 HCPs who participated in 7 groups are in Table 2.

Table 1.

Characteristics of adolescent and young adult (AYA) kidney transplant recipients and parents

	AYA all ages (n=32)	AYA aged 18–24 (n=13)	AYA aged 12–17 (n=19)	Parents of AYA aged 12–17 (n=15)
Number of focus groups	9	5	4	4
SOCIODEMOGRAPHIC CHARACTERISTICS
Age, M (SD), range	17.6 (3.6), 13.0–24.8	21.3 (2.2), 18.4–24.8	15.1 (1.7), 13.0–17.7	43.9 (8.8), 31.0–65.0
Age of recipient at transplant, M (SD), range	12.7 (4.8), 1.7–13.0	15.8 (3.8), 10.0–23.0	10.7 (4.3), 1.7–16.3	10.5 (4.4), 1.7–16.3
Sex, female, % (n)	43.8 (14)	38.5 (5)	47.4 (9)	93.3 (14)
Race/ethnicity, % (n)
NonHispanic White/European	81.3 (26)	84.6 (11)	78.9 (15)	73.3 (11)
Black/African	3.1 (1)	7.7 (1)	0.0 (0)	0.0 (0)
Hispanic/White	3.1 (1)	0.0 (0)	5.3 (1)	0.0 (0)
Asian or Native Hawaiian/Pacific Islander	3.1 (1)	0.0 (0)	5.3 (1)	6.7 (1)
Mixed	9.4 (3)	7.7 (1)	10.5 (2)	13.3 (2)
Missing	0.0 (0)	0.0 (0)	0.0 (0)	6.7 (1)
Level of education, % (n)
9th grade or lower	40.6 (13)	0.0 (0)	68.4 (13)	6.7 (1)
10th to 12th grade	37.5 (12)	46.1 (6)	31.6 (6)	6.7 (1)
> high school	21.9 (7)	53.8 (7)	0.0 (0)	53.3 (8)
Missing	0.0 (0)	0.0 (0)	0.0 (0)	33.3 (5)
Parent’s relationship to recipient, % (n)	---	---	---
Mother/stepmother				86.7 (13)
Father/stepfather				13.3 (2)
TRANSPLANT-RELATED CHARACTERISTICS
Years since patient’s most recent transplant, Median (IQR)	3.6 (1.7–8.0)	4.6 (1.7–8.5)	3.6 (2.1–5.4)	3.8 (1.1–5.4)
Lifetime number of kidney transplants received, % (n)
1	84.4 (27)	76.9 (10)	89.5 (17)	93.3 (14)
2	15.6 (5)	23.1 (3)	10.5 (2)	6.7 (1)
Indication for transplant, % (n)
Congenital anomalies of the kidney and urinary tract	34.4 (11)	23.1 (3)	42.1 (8)	33.3 (5)
Nephronophthisis	18.8 (6)	15.4 (2)	21.0 (4)	26.7 (4)
Acquired disorders (focal segmental glomerulosclerosis, glomerulonephritis)	15.6 (5)	30.8 (4)	5.3 (1)	6.7 (1)
Other	18.7 (6)	23.1 (3)	15.8 (3)	33.3 (5)
Unknown	12.5 (4)	7.7 (1)	15.8 (3)	0.0 (0)
Type of most recent transplant, living donor, % (n)	65.6 (21)	53.8 (7)	73.7 (14)	80.0 (12)
Study participant parent was a donor for child, % (n)	---	---	---	46.7 (7)
Required dialysis before transplant, % (n)	56.3 (18)	53.8 (7)	57.9 (11)	53.3 (8)

Table 2.

Characteristics of HCPs

Characteristic	Focus group members (n=46)
Number of focus groups	7
SOCIODEMOGRAPHIC CHARACTERISTICS
Age, M (SD); 1 case was missing	42.0 (9.9)
Sex, % female (n)	84.8 (39)
Race/ethnicity, % (n)
White/European	80.4 (37)
Black/African	4.3 (2)
Hispanic/White	0.0 (0)
Asian	8.7 (4)
Mixed	6.5 (3)
Professional discipline, % (n)
Nurse	37.0 (17)
Transplant physician	23.9 (11)
Social worker	10.9 (5)
Psychologist	6.5 (3)
Pharmacist	6.5 (3)
Other	6.5 (3)
Number of years in clinical practice, Median (IQR)	16.0 (7.0–22.0)
Number of years in transplantation, Median (IQR)	8.0 (3.0–17.0)

Challenges of Medication Adherence

AYAs, parents, and HCPs described perceived challenges in medication adherence, and Table 3 provides illustrative quotes. AYAs reported their top challenges were (a) remembering to take medications on time and (b) learning the steps required for medication management. Strict routines were difficult to maintain, specifically timely morning medication administration, managing medications when away from home, and when daily routines were more relaxed. The more frequent the medication dosing, the less confident AYAs felt about successful administration. They reported little experience in independent medication administration. While some AYAs were eager to master the necessary skills for taking medications independently, others were less interested and let parents take charge of medication management.

Table 3.

Challenges to Medication Adherence by Stakeholder Group

Challenges Expressed by AYAs
Themes	Illustrative quotes
Difficulty remembering to take medications on time	“It’s easier in the school year because you are getting up early… but during the summer, it’s kind of more relaxed…and it can be a little bit more difficult.” −22 year old “I have to take my medications four times a day. When I am on my own I have a hard time remembering, so my mom reminds me to do it.” −14 year old Once the doses go down it does not really seem to change a whole lot (unless the manufacturer changes the pill color or the pill size or shape) but it does not really affect me because I know what pills are routine and plus I am on like the Prograf and Cellcept – 22 year old “If I don’t get up and take them right away, then I’ll say to myself that I will remember to take them when I get up, but then I forget and I do other things.” −15 year old -I know which ones they always look like and it’s not hard for me to figure out which I am taking and it’s just become normal – 22 year old “I think as you get older, you kind of get more busy with different school things so I think it can be a problem to take them at night.”. −24 year old
Learning the steps required for medication management	“To begin with, [my coordinator] helped me learn my pills by color, by name, by dose and when they are due. I can tell I’m getting better at remembering these details which makes me feel good.” −12 year old “I’m going to want my parents there to help with everything because I have problems that prevent me… I have issues remembering or just saying stuff normally….I have a hard time being social because I have anxiety and stuff like autism… and stuff that makes it hard for me to be able to do it on my own now.” −22 year old “Obviously, I know my dad is not going to let me forget to take my meds, so sometimes I just wait for him to remind me.” −17 year old “You have to decide to either take your meds or you don’t … you need to know it’s your responsibility for life.” −24 year old
Challenges Expressed by Parents
Themes	Illustrative quotes
Fear of the consequences of non-adherence	“…they’ve been given a gift ….if that means I’m getting up at 6:30 on Saturday when I would love to sleep in, I’m going to do it.” -mother of 12 year old “[My son] never goes on trips without me except for [kidney camp]…they really were taking care of him so I allowed that, but he doesn’t go on any other trips with anybody else. -mother of 13 year old “The possibility of rejection is real…. because everything changes if she doesn’t take those meds” -mother of 15 year old
Knowing whether their children took their medications	“It’s just the extra reassurance that if I don’t receive a text then I know she didn’t even think about it so I can get on her and get her to take them.” -mother of 15 year old “You think it will be fine but then when you say to them, ‘did you take them’ – ‘yes, I took them” and you come into the room and they are still there, so… I kind of put it in his mouth almost.”-mother of 13 year old “For me, it’s more if he is somewhere else and I’m not with him…how do I know that he actually took his medication? To me it’s not so much his scheduling, its whether he took them or not.” -mother of 15 year old
Supporting children’s transition toward independence	“… there’s a lot that we’ve had to go through to get our kids to this point and none of us are willing to sacrifice their health… and we are sacrificing ourselves… until they honestly can show us that they can do it for themselves consistently.” -mother of 17 year old “Of course, there is always that resistance, especially with a teen. We’ve had alarm watches, alarms on the phone…sometimes because he has the ear plugs in or he’s online he doesn’t hear them.” -father of 16 year old “It’s like there’s so much outside of their control, but you know that’s the one thing they can control. They say, ‘I’m not going to do it’.” -mother of 17 year old
Challenges Expressed by Health Care Professionals
Themes	Illustrative quotes
Lack of reliable measures of medication adherence	“….some sort of tools need to be developed to help determine non-adherence early to prevent bad outcomes.” -physician’s assistant adult transplant “Because there is no way to know for sure whether medications are being taken as they should be, I had to assume that no adolescents were achieving adherence….” –pediatric transplant psychologist “Patients won’t tell clinicians the truth about med-taking to avoid disappointing them. Teens tell us what they think we want to hear….” -pediatric transplant nephrologist
Limited capacity to support adherence	“We have limited time to spend on adherence during a follow-up visit and geographic distance from the center makes it difficult for families to plan additional visits.” -transplant psychologist “Ideally, the entire team would be mobilized to help solve adherence problems… Adherence is only one of many issues addressed in team meetings. Finding the time for the team to develop individualized solutions regarding adherence is difficult.” -pediatric transplant nephrologist “We all feel the need for training about how to assess educational needs and how to deliver material appropriately to kids and parents with variable needs.” -pediatric transplant coordinator
Maintaining a trusting relationship with children and parents	“Sometimes it’s behaviors of the parents that undermine adherence…For instance, lack of financial responsibility, or making sure that kids get their prescriptions refilled, or come to clinic appointments. It is difficult to make understand that it is not ‘us against them’. …we work with them for the good of their child and their family.” -transplant social worker “Some parents consider non-adherence as a reflection on them so they make kids afraid to report non-adherence; this is another example where the trusting relationship is threatened when parents prevent kids from openly discussing their struggles with adherence.” –transplant coordinator “Considering the role parents play in promoting adherence, it is important to recognize that they [parents] run the spectrum of independence levels, education, cognitive ability, degree of involvement, capacity, social support and resources. They need to be able to trust that we are willing to help meet their needs.” –transplant social worker
Supporting children and parents toward transition to independence	” After the first year, adolescents become less vigilant and early habits are broken. The adolescents tend to become less vigilant about taking medications during periods of relative stability. This is an important factor for determining readiness for greater independence.” -pediatric transplant nephrologist “Older teenagers often struggle the most because their lives become more complicated. Teens tend to believe they are invincible so they take risks… and gamble with their health. These behaviors are often most pronounced in the late teenage years when they should be transitioning to more independence.” -transplant psychologist “It seems like there is sometimes a disconnect between the parents thinking the kid is ready to become more independent and the kid not wanting to disclose that they are forgetting to take their medications. Or sometimes, the adolescent is ready to take responsibility and the parent is unwilling to let go.” -transplant coordinator

Parents reported challenges dealing with (a) fear of the consequences of non-adherence, (b) knowing whether their children took their medications, and (c) supporting children’s transition toward independence. Parents’ fear of organ rejection led to high degrees of vigilance over their children’s medication-taking, particularly early post-transplant when the medication regimen was new and more complex than later post-transplant. They expressed uncertainty regarding whether their children took medications unless witnessed. Parents had difficulty relying on their child to remember to take their medications, even when reminders were sent and the children sent texts to confirm medications were taken. The repeated attempts to confirm medication administration often strained parent-AYA relationships. Parents struggled knowing when and how to guide their children to assume more independent roles. Their children’s natural urge to separate from them as they got older by pushing back against established rules, roles, and relationships were often at odds with parents’ desire for AYAs to demonstrate that they could reliably manage their medications.

HCPs’ challenges for supporting AYA medication adherence included (a) a lack of reliable measures of adherence, (b) limited capacity to support adherence, (c) issues related to maintaining a trusting relationship with children and parents, and supporting transitions to AYA independence. Lack of definitive information on adherence complicated prescribing decisions, given the risks associated with both under- and over-dosing immunosuppression. Some HCPs reported that their default position was often to assume that the AYAs were nonadherent, although they acknowledged that asking AYAs about medication taking was vital. When adherence was less than ideal, they often questioned whether they had provided the necessary support or had established sufficient trust to allow adherence difficulties to be openly discussed. HCPs acknowledged that AYAs’ vigilance in following their regimen declined over time post-transplant. Although promoting adherence was a high priority, HCPs commented that transplant programs have insufficient clinical resources to provide practical and psychological support because adherence issues are complex. They also acknowledged their own training deficits and wanted to learn better, more practical ways to support children and parents with medication management and transition to independence.

Strategies for Medication Adherence

AYAs reported that their main strategy involved linking medication administration to daily routines, such as getting ready for school/work in the morning and getting ready for dinner in the evening. They also tried to fill their pill containers on the same day weekly. They used cell phone alarms as medication reminders. AYAs kept backup medications with them or with parents in case they were away when medications were due. They relied on others, including family and friends, to “double check” that they took their medications. If they missed a dose, they knew what to do (e.g., how late they could take it or if the dose should be skipped).

Parents noted the importance of having consistent, repeated conversations with their children about the gravity of their condition and need to take medications as directed. They reported that they wished their children had more opportunities to connect with others who have undergone transplantation to share experiences and learn from each other. Parents, like AYAs, emphasized maintaining structured routines, anticipating activities that disrupt those routines, and having backup doses of medications available. Parents relied on a select group of other trusted adults within the family to help monitor medication taking. They relied on reminders on their own their cell phones to indicate when medications were due. They expected their child to inform them when they had taken medication doses. Parents took opportunities to gradually teach AYAs the necessary steps to assume more responsibility for medication management.

HCPs relied on triangulation of data sources as proxy measures of adherence, including therapeutic drug levels, rejection history, missing laboratory studies or clinic visits, AYAs’ knowledge of medications and engagement in their care, and the opinions of other HCPs. They felt AYAs and parents should both be included in medication-taking from the beginning with the hope of gradually increasing AYAs’ level of accountability and transition to independence. They noted that conversations with AYAs and parents about medication-taking gave the opportunity to build trust, discuss barriers to adherence, reinforce the importance of adherence, and review consequences of nonadherence. They addressed adherence as a routine part of each encounter in order to be proactive rather than reactive. They recommended holding team meetings to discuss adherence; some wished to include the AYA and parent in such meetings. Some transplant programs included a pharmacist who could meet with individual AYAs and HCPs to help simplify medication regimens or reduce side effects.

Views about Personal Coaching and Action Plans

The views of each stakeholder group regarding the use of personal coaches and action plans are summarized in Table 4. AYAs were reluctant to believe a coach would provide benefit beyond what they received from their parents and healthcare team. As one 17 year old remarked, “I don’t think a coach would have any better ideas than the transplant social worker or my nurses.” When pushed to imagine an ideal support person, AYAs were split about whether they would want to meet someone during or outside of the clinical visit, with concerns about such meetings extending the already long stays for clinic appointments vs. concerns about having more of their lives occupied with medication-related issues. Some felt that having a coach would be acceptable, but only if it was optional.

Table 4:

Suggestions for Enhancements of the Adherence Intervention

	AYAs	Parents	HCPs
Coach	Could be a peer coach or support person Could work to support group session Should be someone they can personally identify with that understands transplant life and can help master medication management skills	Should explain what parents should expect at each post-transplant phase, especially in the first 3 months post-transplant Advise on how developmental stages influence how a parent might approach medication management and adherence Could be a peer	Would need knowledge and skills in communication and behavior motivation Could serve as a liaison between the AYA, parents and transplant team
Action Plans	Useful only when a problem with adherence is identified or in special circumstances	Should outline how to help their children, including when and how to execute plans	Universally viewed positively, but unclear how to integrate plans into clinic flow
ePillbox Design	Need to detect when the pills were removed from the compartments Should send immediate confirmatory text notifying parent that correct compartment was opened and pills were taken at appropriate time Need reminders when pills are due Compartments should snap together and pull apart like a pill organizer for portability	Need sensor that detects opening and closure of the compartments Need a warning alert after the medication should have been taken if no response Pill compartments should separate to take a few doses away at a time Should provide alert when ePillbox is empty/needs to be refilled	Size is important for portability and discretion Build in individualized rewards and motivators Include ability to connect with social media (e.g., facebook, IMessaging) Avoid potential for ‘tricking the system’ Should allow for synchronization to AYA’s phone (e.g., for alerts about medication times, appointments)
Website Design	Provide summary report listing how many medications were taken on time out of total number possible Include educational material to help them learn their medications Send an alert when ePillbox is empty/needs to be refilled Need a way to post a note to HCP about why a dose was late or missed	Provide summary report listing how many medications were taken on time out of total number possible Include personalized quizzes related to medications Send immediate confirmatory text notifying parent that correct compartment opened and closed at appropriate time Send warning alert after the medication should have been taken if no response Provide decision support about what to do when AYA is late in taking medications Have ability for multiple people to receive alerts	Merge sources of data about adherence (ePillbox, electronic medical record) Send feedback about adherence to clinicians via web or electronic health record Allow alerts, timing, and mode of alerts to be customized based on AYA preference AYA and family should decide who should have access to data on website Allow HCPs to view patterns and trends in medication taking Expressed concerns about amount of data that would be generated and liability associated with monitoring

Parents and HCPs both expressed mixed feelings about personal coaches, including concerns about adding another provider to a large clinical team. Parents suggested coaches could undermine parental authority and preferred the AYAs received their guidance from them as they knew their children better. They commented that a coach might be helpful early post-transplant when regimens are new and seem complex. HCPs felt routinely meeting with a personal coach (at the clinic or patients’ homes) could reduce the emphasis on medication-taking problems in clinic and allow HCPs to discuss other important issues. The coach could participate in HCP team meetings and provide an outside perspective. On the other hand, some HCPs felt that a transplant team member, particularly transplant coordinators, were in the best position to offer coaching. They were concerned a coach would prolong visits and, if coaches did not have a long-standing relationship with the families, they might not provide an individualized approach to care.

Although action plans were viewed as potentially helpful, AYAs felt such plans would be most relevant for peers who were having difficulty remembering their medications. None of the AYAs felt they personally would benefit from an action plan. Parents thought action plans might help if they were very specific. HCPs noted that action planning was useful in other patient groups, but were not sure how they could integrate plan development into their workflow because of time constraints. They noted that action plans would only be helpful if AYAs and parents were involved in the development of individualized plans.

Views about Design of an ePillbox and a Related Website for Medication Monitoring

Stakeholders’ views are summarized in Table 4. AYAs and parents both felt an ePillbox needed to be able to detect use, send confirmatory messages to parents and HCPs when compartments were opened, and be portable. HCPs suggested social media compatibility and a reward system might be useful to AYAs. For the website design, parents and AYAs were interested in access to education materials including, for example, personalized quizzes. All stakeholder groups wanted to be able to view summary reports related to medication taking. AYAs and parents wanted to be able to submit comments to explain why doses were taken late or missed. HCPs also thought it would be important to have the possibility to receive messages confirming medication-taking but they did not want AYAs and parents to rely on web-based communication to report health-related concerns. HCPs also wanted decisions about alerts, timing, and mode of alerts intended for AYAs and parents to be customized based on their preferences. HCPs liked the potential to view patterns and trends in medication taking, but expressed concerns about the volume of data that would be generated by the ePillbox and liability associated with monitoring the website. HCPs also hoped that data about medication adherence from the ePillbox and website could be merged with other data in electronic medical records such as labs and clinic notes.

Discussion

This study allowed us to gain insights directly from key stakeholders (AYAs, parents and HCPs) about issues related to medication adherence that were not considered in Take-IT. Although the Take-IT trial demonstrated that participants randomized to the TAKE-IT intervention showed greater medication adherence,²⁰ the feasibility and acceptability of the TAKE-IT intervention for use in routine clinical care were not clear. Moreover, the intervention was designed by pediatric nephrologists without any other stakeholder input. The current TAKE-IT-TOO project is designed to further refine and augment the successful intervention to promote scalability. Furthermore, we are taking a comprehensive approach to adapt and optimize the TAKE-IT intervention by applying principles of user-centered design as the intervention is modified so that we can gain input from key stakeholders on multiple levels. Those levels include not only the collection of focus group data but the inclusion of AYA and parent stakeholders as co-investigators on the study team. This approach allows us to gain insights directly from the individuals who share direct responsibility for promoting medication adherence.

All stakeholder groups reported dealing with challenges related to taking medications on time but from different perspectives. The AYAs reported that one of their primary challenges was remembering to take medications. The parent groups expressed fear of the consequences of non-adherence, and the need to know that their children did not forget to take their medications. The HCPs reported the challenge of determining reliably whether AYAs were taking their medications. Our findings about the challenges that AYAs, parents and HCPs face with medication adherence were consistent with those reported in other studies. For instance, forgetting to take medications was a common barrier to adherence in the AYA age group.^7,8 A study of 110 health care providers specializing in pediatric transplantation found that providers reported challenges similar to those reported by our HCP stakeholder groups. ³² However, in contrast to past research, our study participants were also asked to consider possible design solutions to overcome barriers to taking medications as prescribed.

The views expressed by stakeholders in our study also indicate that the challenges to medication adherence in this age group are complex and situated in the context of transition toward independence. The AYAs reported challenges such as learning the steps required for medication management. Parents reported challenges supporting children’s transition toward independence, and HCPs found it difficult to know how to best support AYAs and parents to make transitions to independence. This is consistent with findings from qualitative reports that found that managing the multiple demands of transplantation along with the normal developmental tasks of adolescence posed many challenges for young transplant recipients and their families, including managing medical demands, such as adherence. ⁷ Areas of particular importance included the AYA’s desire for independence and need to assume more of the responsibility for medication taking. Killian’s review ⁸ also identified adherence to be an issue for the family as a whole, particularly when transitioning responsibility for medication taking from the parent to the patient. Reviews have suggested that better communication and interaction between the parent and AYA facilitate the transition.^7,8 Interestingly, systematic review of perceived barriers to medication adherence among 2501 adolescents who had at least one of 14 chronic diseases, ³³ found that despite disease-specific differences in requirements, barriers were similar across conditions and aligned with the barriers reported by our focus groups and other studies of AYA kidney recipients.

However, our findings go beyond expanding our understanding of the challenges of medication adherence. Our stakeholders also suggested ways to adapt and add new components to the TAKE-IT intervention to overcome issues and problems that AYAs face adhering to the complicated medication regimen while in the midst of their busy schedules and demands to transition toward independence. Some of the suggestions included an expanded ePillbox and companion website, education materials, and adding features to individualize and customize digitized alerts, notifications and reminders to support sharing responsibility and communication between AYAs, parents and HCPs based on the stakeholders preferences and AYA’s unique routines and schedules. In fact, the use of ePillbox technologies is not a new adherence intervention strategy; higher medication adherence was observed with the use of ePillboxes in other chronic disease states.^34,35 The suggestions voiced by our stakeholders regarding an ePillbox and the addition of the companion website will inform refinements of these designs to promote ease of use, tracking and displaying of objective data, and sending feedback about the AYA’s medication adherence patterns to parents and HCPs.

eHealth interventions show promise for promoting medication adherence among adolescents with a variety of chronic diseases.^4,36,37 Mobile technologies are being designed with increasing frequency and may be efficacious in improving medication adherence, but to date have been used primarily in adult transplant populations.³⁸ Some of these technologies are relatively passive (e.g., ingestible sensors) or simple (e.g., phones with texting capabilities), while others use more complex interactive technologies, such as smartphones to support patient-provider communication, and blue-tooth enabled medical devices for remote monitoring of physiological parameters or other medication adherence targets. However, it is rare for stakeholders to be involved in the design of eHealth technologies and for the technologies to be intended for clinical use, targeted to the AYA age group, be multi-component in nature, or support the work of the multiple stakeholders involved in medication adherence of AYA kidney recipients all at once. Stakeholders in this study offered suggestions for an eHealth behavioral intervention that included multiple components and highlighted the importance of education, communication, and social support to improve medication adherence.

Importantly, HCP stakeholders expressed specific concerns about use of both the ePillbox and website for adherence monitoring, based on the volume of data to be collected and the amount of information they would have to deal with. They were also concerned about liability issues related to their uncertainty regarding their responsibility to respond to indicators of non-adherence on a day-to-day basis. Summaries of the critical issues in the success of eHealth interventions ^39,40 revealed that concerns related to clinician workflow are not new, and ambiguity regarding clinician accountability is common. These findings underscore the importance of seeking HCP perspectives about how to establish workable routines in order to design an intervention that will be useful and adopted in the real-world clinical setting.

Perspectives on personal coaching and action planning were mixed. Most AYAs and parents were ambivalent about including an additional person (a coach) on the care team. HCPs felt that coaches might be useful for AYAs who were having difficulties following the regimen. When personal coaches were considered, all stakeholder groups agreed that to be effective, coaches would need skills in communication and behavioral motivation. A coach would also be expected to work with the team to avoid undermining the relationships between AYAs, their parents and HCPs, or prolonging clinic visits. The stakeholders also saw value in concrete, individualized action plans with input from AYA and parents, such as those used in the TAKE-IT intervention. It is possible that our participants did not mention other factors that might have influenced their views about adding a coach or action plan. For example, HCPs’ own caseloads may have affected their perceptions about the value of adding various intervention strategies (i.e., the strategies could either facilitate HCPs’ work or interfere with it). AYAs’ own adherence levels may have affected the views they and their parents expressed. The AYAs and parents in our study generally reported that they were adherent to requirements; this may explain why they did not see value in adding a coach.

Indeed, one limitation of our study is that our sample may not include full representation of the range of adherence in the AYA transplant recipient population. Studies focused on medication adherence often have an inherent selection bias toward recruiting individuals who are adherent.⁴¹ Alternatively, our AYAs and parents may have felt uncomfortable acknowledging nonadherence with other families. However, this did not seem to preclude the participants from offering many and very extensive comments on barriers to adherence suggesting the personal acknowledgement of problems may not have been essential.

Another study limitation is the fact that our sample was predominantly white with little representation of minorities. This lack of racial diversity is a concern particularly since the aim of involving a wide range of users in the design is to increase the likelihood that the intervention will be adopted by AYAs with different needs. To this end, efforts to recruit ethnic and racial minorities in the next phases of the study will be heightened. However, our sample included AYAs, parents and HCPs who spanned the age range and disciplines of interest, who varied considerably in time since transplant, transplant program volume and geographic regions.

The findings of our study have implications for clinicians and researchers involved in the promotion of medication adherence among AYA kidney recipients. While AYAs and parents across transplant programs may share common challenges to adherence, our findings indicate the need to consider adherence fully in the context of recipients’ unique situations, as defined by their own routines, their age and degree of independence, time since transplant, and regimen complexity. Therefore, adherence must be assessed individually and often to reflect these variations. In general, incorporating medication-taking into daily routines is known to promote adherence, but this is particularly important for adolescents during times when their routines are altered, such as weekends, vacations and breaks from school, and for young adults as they begin to take on jobs and other commitments outside the sphere of parental oversight. The transition toward independence appears to be a particularly critical issue to be considered by all stakeholders, yet is fraught with uncertainties. AYAs may seek more freedom but may not have the action plans needed to maintain adherence during this move. Parents and HCPs often find it difficult to judge the recipient’s readiness to assume sole responsibility and may not know optimal strategies to promote such behavior.

Our findings may benefit researchers who may be considering developing adherence interventions that leverage technology. Thus, not only is it important to apply the principles of user-centered design to future work, but such work might benefit from consideration of our AYA and parent stakeholder perspectives with regard to the technology features that they thought would be helpful, as well as the concerns expressed by HCPs about the time commitment and liability of using medication-tracking websites.

Considering that the development and refinement of technology-based interventions is an iterative process, several suggestions regarding the functionality and features of the potential intervention identified at this early stage will be explored in more depth. Our approach to actively involve all stakeholders in the process, and plan to assemble as representative a sample as possible, increases the likelihood that the goal of designing an adherence intervention that is truly user-informed and fit for the clinical setting is more likely to be realized.⁴²

Abbreviations:

AYAs

adolescents and young adults

HCPs

healthcare providers

TAKE-IT TOO

Teen Adherence in KidnEy transplant Improving Tracking To Optimize Outcomes

UPMC

University of Pittsburgh Medical Center

UCD-QAC

User-Centered Design and Qualitative Analysis Center