Objectives: Pediatric patients are at higher risk of nonadherence to immunosuppressive medication after kidney transplant and the resulting adverse outcomes. Factors associated with nonadherence vary, which follow an epidemiological framework and according to health system patterns. The Brazilian public health system covers all costs of kidney transplant, including immunosuppressive medications. We aimed to assess the prevalence and correlates of nonadherence to immunosuppressive medications in a pediatric kidney transplant population who received free access to immunosuppressive medications within the health care system.
Materials and Methods: In this single-center cross-sectional study, we studied a convenience sample of 156 outpatients (< 18 years old) who were a minimum of 4 weeks posttransplant. Implementation nonadherence to immunosuppressive medications was measured by the 4 questions of the Basel Assessment of Adherence to Immunosuppressive Medications Scale. Multilevel correlates to nonadherence (patient, micro, and macro levels) were assessed.
Results: In our patient population, 61% were males, mean age was 13.6 ± 3.1 years, 77% were adolescents, and 84% received organs from deceased donors. We found that 33% were nonadherent to immunosuppressive medications, mainly in timing (25%) and taking (10.9%) dimensions. Being an adolescent (odds ratio: 2.66; CI, 1.02-6.96), religion other than Catholic or Protestant (odds ratio: 4.33; CI, 1.13-16.67), and family income higher than 4 reference wages (odds ratio: 3.50; CI, 1.14-10.75) were factors associated with nonadherence.
Conclusions: In our patient population of mostly adolescents, one-third displayed nonadherence to immunosuppressants. Unexpectedly, a higher economic profile, potentially representing better previous access to health care, was independently associated with nonadherence. This result highlights the need for identifying specific correlates to nonadherence before designing interventions.
Key words : Children, Clinical research/practice, Immunosuppression, Kidney transplant, Medication nonadherence
Similar to that shown in adult patients, kidney transplant (KT) in children increases patient survival compared with other renal replacement therapies, with lower cost for the health care system.1,2 Nonadherence to treatment is a significant risk factor for graft loss after KT.3-7 In children, this behavior has the potential to be more devastating since the challenges of controlling immune responses in this population are higher than in adults, with resulting consequences of therapeutic failure being more likely.3,4,8-11 With the consideration that KT in pediatric patients provides improvements in growth and development and reductions in morbidity and mortality, feasible measures to reduce nonadherence are valuable strategies to improve outcomes in pediatric KT recipients.5,11
Adherence to treatment involves not only the prescribed drugs but also dietary recommendations and lifestyle modifications.12 However, a consensus conference on transplant proposed that nonadherence should be defined as the failure of the medication regimen that would be enough to undermine results.13 Adherence to medication is currently considered to be a 3-stage process, subjected to measurement, consisting of initiation, implementation, and persistence. In KT recipients, the challenges are how patients take their immunosuppressive drugs (implementation) and the continuity (persistence).14 With regard to the implementation phase, variances in the prescribed regimen by greater than 5%, measured by electronic monitoring, have been shown to result in adverse outcomes.15,16
Studies in nonadherence after pediatric KT have reported an incidence of approximately 30%, with a range of 5% to 72%.3,4,9-11 Nonadherence is reported to increase up to 43% during adolescence, particularly in the period of transition from pediatric to adult care, during the time when adolescents progress in their abilities to be more self-sufficient and independent.3,4 As a result, 23.2% of acute rejection episodes are associated with nonadherence, and there is a 5-fold increase in the risk of graft loss by nonadherent patients.3,9,11
According to the World Health Organization, adherence is a multidimensional phenomenon determined by the interaction of 5 factors: (1) social/economic factors, (2) patient behavioral factors, (3) condition-related factors, (4) therapy-related factors, and (5) health care team and system-related factors.12 The ecological model proposes that individual behavior is influenced by these factors at the levels of health professionals (micro), health services (meso), and health care system and policies (macro).17 The identification of these factors in each population is an essential tool for the design of strategies aimed at the reduction of nonadherence.3,5,7,18
Brazil ranks second internationally in the total number of KTs performed annually.19 All of these procedures are performed in the largest public health system in the world, which includes the free dispensation of immunosuppressive medications.20 The number of KTs performed in Brazil has increased by 66.5% over the past 10 years.21 Pediatric KT, comprising 2% to 13% of the total number of KTs, has also experienced an improved trend but with lower annual growth rates compared with the adult population. However, on average, 80% of pediatric patients on wait list are effectively transplanted per year. Transplant services are concentrated in the more developed regions of the country, and there is a shortage of data on the evolution of these grafts.21 Unfortunately, even with an estimated population of approximately 2000 pediatric KT recipients, there are no studies assessing nonadherence in this population in Brazil.
This study aimed to assess the prevalence of nonadherence to immunosuppressive medications in a pediatric KT population with free access to these medications and to identify the factors associated with this behavior, based on the ecological model.
Materials and Methods
This was a single-center observational and cross-sectional study, conducted at Hospital do Rim of the Oswaldo Ramos Foundation and Federal University of São Paulo (UNIFESP), São Paulo, Brazil. The project was submitted to the Ethics and Research Committee of the Federal University of São Paulo (approval number 471.626, 11/29/2017) and conducted in accordance with the ethical standards of the institutional research committee and with the 1975 Helsinki Declaration and its later amendments or comparable ethical standards.
The sample was defined by convenience. We invited all pediatric patients followed in the period after KT to participate who met the inclusion criteria; all who agreed to participate in the research were included in the study. During the 2-year period (2014 and 2015) of the study, of 240 total patients who received follow-up, 212 came with their families for routine visits. Of these, 56 were not included (51 did not meet the inclusion criteria, 4 refused to participate, and 1 was not taking immunosuppressive medication). Inclusion criteria were as follows: age < 18 years and at least 4 weeks post-KT. The time post-KT was arbitrarily chosen as a minimum period required to adjust to a new treatment regimen, allowing time to follow the recommendations of the selected scale to evaluate nonadherence (Basel Assessment of Adherence to Immunosuppressive Medications Scale [BAASIS©]).
Sociodemographic and clinical characteristic variables
The following characteristics were collected through interview with parents and from medical records: age (adolescents were those 12-18 years old and younger age were those < 12 years old), sex, race (white, black, mixed-race, Asian), family income (1-4 times the Brazilian reference wage [approximately US $232.06]), religion (Catholic, Protestant, Spiritualist, no religion), education (elementary, high school), actively participating in school (currently or unable to attend due to disease), type of transport to KT service (car, plane, public transportation, or provided by health care system), use of more than 1 type of transport, and presence of a caregiver. We also collected the following clinical variables: cause of chronic kidney disease (CKD), type of pre-KT treatment (predialysis, hemodialysis, peritoneal dialysis), type of donor (living or deceased donor), time post-KT (months), and scheme and dosage of immunosuppressive medications (type of drug, number of pills).17
Adherence to immunosuppressive medication
To assess adherence to immunosuppressive medications, we administered the BAASIS© by interviewing the patient’s parents or guardians.22,23 The BAASIS© is a self-report questionnaire, which has been validated in Brazilian Portuguese for use in transplant patients.22,23 This instrument assesses the implementation phase of medication adherence.14,22,23 It assesses the 4 most frequent dimensions of nonadherence during the period of the previous 4 weeks. These dimensions are dose omissions (taking dimension), a delay greater than 2 hours (timing dimension), an omission of successive doses (drug-holidays), and dose reduction. Any positive answer classifies the patient as nonadherent.22
Factors associated with nonadherence based on the ecological model
Factors associated with nonadherence were based on the ecological model. We evaluated at the patient level: (1) sociodemographics, including age (determined as adolescents [12-18 years old] or younger [< 12 years old]), sex, race (white, black, Asian, or mixed-race), family income nominal monthly per capita (1-4 times the Brazilian reference wage [US $232.06]), religion (Catholic, Protestant, Spiritualist, no religion), education (elementary or high school); (2) related to disease, including CKD etiology, type of pre-KT treatment (predialysis, hemodialysis, peritoneal dialysis); and (3) related to the treatment, including type of donor (living, deceased donor), time post-KT (months), and scheme and dosage of immunosuppressive medications (drugs, number of pills). At the micro level, we evaluated social support based on the presence of a caregiver. At the macro level, we assessed the type of transport to access to the KT service (type and use of more than 1 type of transport). The meso level, related to the health care team and service structure, was not addressed.17,24
Data collection occurred over a 2-year period during routine clinic appointments. Parents/guardians of pediatric KT recipients were invited to participate in the study if they met inclusion criteria. After agreeing to participate, they signed the Terms of Free and Informed Consent Form. Data collection was conducted by the principal researcher, who was not part of the KT team, by interviewing parents or guardians and patients. It was decided to interview the parents because, as caregivers, they follow the outpatient appointments. In addition, this method avoided collection bias, since the sample included children and adolescents. Correlates to nonadherence were collected by interview or from medical records.
Categorical variables were presented as frequencies, and numerical variables were presented as means ± standard deviation. Associations between adherent and nonadherent patients were evaluated using the chi-square test/Fisher test or t test/Mann-Whitney test, based on the perspective of the ecological model. Normality was checked by the Kolmogorov-Smirnov and Shapiro-Wilk tests. Logistic regression to nonadherence included variables whose associations in the univariate analysis were meaningful to 10% (adolescents or younger, religion, family income nominal monthly per capita, type of transportation). The adequacy of the final model was evaluated by the Hosmer-Lemeshow test. A level of significance was set at 5% for all statistical tests. We used SPSS version 20.0 (IBM Corp.) for all analyses.
Characteristics of the sample
Of the 156 patients assessed, 23.1% were children and 76.9% were adolescents, with a mean age of 13.6 ± 3.1 years, ranging from 3 to 17 years. Most study patients were male (61%), and most were of mixed race (51.3%). Family income was low; 55.4% reported earning less than 3 times the Brazilian reference wage (approximately US $696.18), and most received social benefits because of the disease (56.4%). Catholic religion was the prevalent religion, and most (64.3%) had completed elementary school. Regarding access to KT follow-up, transportation provided by the health care system (37.2%) and bus (30.8%) were the most frequent; however, 22.4% required air transportation and 20.5% needed more than 1 type of transplantation (Table 1).
When we analyzed clinical characteristics, concerning the initial diagnosis of CKD, the most frequent causes of end-stage renal disease were hereditary/genetic, obstructive/chronic pyelonephritis, and chronic glomerulonephritis. Hemodialysis was the most common pre-KT treatment (60.3%), and 84% received a graft from a deceased donor. Living donors were all related (92% parents, 4% grandparents, 4% uncles). The mean follow-up time was 45.1 ± 35.8 months. Most KT recipients (93.6%) were on a protocol of triple immunosuppressive therapy, with the most common therapy being tacrolimus + azathioprine + prednisone (61.5%) (Table 1).
Nonadherence to immunosuppressive medications
According to the BAASIS©, the most frequent dimension of implementation nonadherence was the timing dimension (25%), followed by the taking dimension (10.9%). In the overall assessment, among individuals who reported at least 1 dimension of nonadherence, one-third of the study patients were nonadherent (Table 2).
In our comparison of adherent and nonadherent groups by univariate analysis, we found differences only at the patient level. Age (adolescents, P = .044) and patients without religion or who were Spiritualists (P = .012) were the only variables associated with nonadherence (Table 3). In the multivariate logistic regression analyses, adjusting those with a P value < .10 as predicting variables, adolescence (2.6 times), having a religion other than Catholic or Protestant (4.3 times), and having a higher family income (> 4 times the reference wages) (3.5 times) were independently associated with nonadherence (Table 4).
On the basis of our study that used a self-report instrument, one-third of pediatric KT patients who received free access to health care were nonadherent to the implementation of the immunosuppressive therapy regimen. Our study population has unique aspects because, in Brazil, post-KT patients have free access to immunosuppressive medications and better access to health care, especially those who depend only on public health care before their KT.24,25 Our study is the first to evaluate multilevel correlates of nonadherence in KT pediatric patients and is one of the few to assess Brazilian KT patients with a broad evaluation.24 Our multivariate analysis showed that adolescents, as reported by others,4 and those having a better economic profile had a higher chance of being nonadherent. Given the significant social disparities seen in Brazil, this is an interesting epidemiological profile.25,26 A possible explanation for this finding may be that patients of a lower socioeconomic profile, who receive better and more comprehensive access to health care after KT (including access to immunosuppressive medications, laboratory examinations, and hospitalizations), may value their improved health and consequently take better care of themselves and their transplanted kidney. Similar results were already described in KT adult patients.24
In our study, most patients were adolescents (> 12 years of age), representing the group of pediatric patients at a higher risk of being nonadherent.4,10 The demographics of our patient population, including self-declared mixed race, history of hemodialysis, receiving a deceased donor kidney, and causes of CKD, were similar to those reported in other Brazilian KT population studies.27-29 There were also similarities to international studies on nonadherence of pediatric KT patients regarding sex, age, and cause of CKD.3,4,8,30
One-third of patients were considered to be nonadherent. Although similar to other studies, this is a worrying finding, especially because the Brazilian health system provides free access to all immunosuppressive medications. The prevalence of nonadherence varies according to the method applied to diagnosis. In studies that used self-report questionnaires, such as the BAASIS© that we used here, a mean of 30% (range, 10%-69%) has been reported in cross-sectional studies.4,11 In other health systems that provide free access to immunosuppressive medication, nonadherence varies from 21% to 29%, even when measured by different methods.31,32 Although there are no studies in the Brazilian KT pediatric population for comparison, a single-center study that used the BAASIS© reported 34% of adult Brazilian KT patients were nonadherent.21 Otherwise, it is interesting to point out that nonadherence to the implementation phase was mostly due to the timing dimension (25%), meaning a delay greater than 2 hours. This finding has also been reported in adult patients.24,33,34 Patients must be made aware of the importance of timing when taking immunosuppressive medications. Even a small deviation (5%) from the prescribed regimen can be associated with suboptimal outcomes.15,16
From the 16 possible variables, driven by the ecological model, only patient-level factors were correlated with nonadherence. Adolescents are classically more nonadherent than younger children.3,4,10 The adolescent experiences a dynamic and complex process of maturation, such as body changes, new intellectual abilities, development of identity, social and affective relationships, dreams, a constant need for independence, and a new social role. These factors cause daily routine changes and a decrease in parental control. These changes usually increase the chance of the adolescent neglecting or deciding to reject the medication treatment.35,36 Specific strategies should be designed for adolescents because previous studies have suggested a higher risk of graft loss during the transition to adult care.37,38
A higher socioeconomic level, defined as a better family income in terms of reference wages, was a correlate of nonadherence. We suppose that people with a higher socioeconomic profile could believe they have easy access to transplant as a treatment and might then behave recklessly after KT. The same finding has been reported in adults.24 Patients with a religion other than Catholic or Protestant were more nonadherent; 77% of this subgroup reported having no religion. In CKD, in more advanced stages, patients who reported having a religion showed better outcomes. They showed a better doctor-patient relationship, quality of life, and coping with the disease, contributing to a better adjustment to treatment and better outcomes.39 A mature type of religiosity, the intrinsic religiosity, which refers to the feeling of believing without clearly identifiable external motives, was more frequent in adherent adult KT patients.40 This finding was never described in pediatric patients and obviously needs further exploration.
There are some limitations of our study. First, we evaluated a convenience sample of a single center. However, the demographic profile is similar to that reported in other studies, and there are few studies about nonadherence in pediatric KT patients involving more than 100 patients.4 Second, it is a cross-sectional study, allowing us to make only associations and limiting causal inference. Third, the diagnosis of nonadherence was based on a self-report instrument (BAASIS©) answered by parents, and we did not assess the adolescent’s independence. This type of measurement tool frequently underestimates nonadherence, compared with the more expensive electronic monitoring. However, BAASIS© has good psychometric characteristics, is inexpensive, and is easy to administer; therefore, it has the potential to be included in clinical practice.5,16 Fourth, the multilevel analysis of potential correlates to nonadherence, including 3 of the 4 levels of the Ecological Model, was not broad. Some additional aspects should have been addressed, such as the interpersonal relationship between the patient and health care provider (micro level) and aspects of the health care system profile (clinical practice patterns, skills of the team).10,11 We understand a further evaluation of all of these aspects will enrich the much needed explanation of the complex behavior of nonadherence, for example, in a multicenter design. Nevertheless, our results bring relevant epidemiological data because there are no studies in Brazilian patients and because a multilevel analysis of correlates of nonadherence has never been previously explored in pediatric KT patients.
With the use of a specific self-report instrument, we studied nonadherence to immunosuppressive medications and its correlates in a sample of 156 pediatric KT recipients under a health care system with free access to transplant, which was based on their parents’ perspectives. The prevalence of nonadherence was 33% and was associated with adolescence, the absence of a religion, and a better economic profile. The correlation between nonadherence and a higher socioeconomic profile, while demonstrating better adherence in lower-income individuals, is surprising in a developing country such as Brazil. These results highlight the need for identifying specific correlates to nonadherence before designing interventions. Specific aspects of the target populations, such as the proportion of adolescents and the availability of transitional services, should guide coordinated actions. These coordinated strategies may also be modulated by the involvement of health professionals with abilities to identify psychosocial aspects and behavior patterns.
Volume : 18
Issue : 5
Pages : 577
DOI : 10.6002/ect.2020.0101
From the 1Nursing Post-Graduation Program, Paulista School of Nursing, Federal
University of São Paulo, São Paulo, Brazil; the 2Kidney Transplantation Unit,
Division of Nephrology, Federal University of Juiz de Fora, and Center of
Interdisciplinary Studies and Researches in Nephrology, Juiz de Fora, Minas
Gerais, Brazil; and the 3Division of Nephrology, Hospital do Rim e Hipertensão,
Paulista School of Medicine, the 4Department of Women’s Health Nursing, and the
5Department of Clinical and Surgical Nursing, Paulista School of Nursing,
Federal University of São Paulo, São Paulo, Brazil
Acknowledgements: The authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no potential declarations of interest.Ana Carolina Maximo Silva was funded through scholarships by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) (financial code 001). The authors declare that there are no conflicts of interest regarding the publication of this paper. Author ORCID numbers are provided. Ana Carolina Maximo Silva: https://orcid.org/0000-0001-9673-3441; Helady Sanders-Pinheiro: https://orcid.org/0000-0001-8603-1331; Renata Fabiana Leite: https://orcid.org/0000-0003-0017-6770; Marina Pontello Cristelli Joseph: https://orcid.org/0000-0002-2813-0400; Jose Osmar Medina Pestana: https://orcid.org/0000-0002-0750-7360; Janine Schirmer: https://orcid.org/0000-0003-0783-2961; Bartira de Aguiar Roza: https://orcid.org/ 0000-0002-6445-6846.
Corresponding author: Bartira de Aguiar Roza, Paulista School of Nursing, Federal University of São Paulo, Napoleão de Barros St, 754, 04024-002, São Paulo, Brazil
Phone: +55 11 99109 8843
Table 1. Sociodemographic and Clinical Characteristics of Pediatric Renal Transplant Recipients
Table 2. Implementation Nonadherence to Immunosuppressive Medications According to BAASIS©
Table 3. Variables Studied at Patient, Micro, and Macro Levels in Pediatric Renal Transplant Recipients According to Implementation Nonadherence to Immunosuppressive Status
Table 4. Multivariate Logistic Regression of Correlates of Implementation Nonadherence