Introduction
The discontinuation of follow-up among individuals with CHD predominantly occurs during the transition from paediatric to adult care, Reference Gilboa, Devine and Kucik1 and studies have reported that approximately 26.1% of individuals with CHD at the transitional age discontinue their medical follow-up. Reference Moons, Skogby, Bratt, Zühlke, Marelli and Goossens2 Patients who discontinue their follow-up often do not seek medical attention until their heart failure symptoms have significantly worsened, to the point of being life-threatening in some cases. Reference Müller, Norozi and Caroline3,Reference Wray, Frigiola and Bull4 Therefore, it is of vital importance for individuals with CHD to acquire the necessary knowledge and skills for managing their disease in adulthood at the appropriate time, enabling them to independently manage their condition. Additionally, they should transition from paediatric to adult care to ensure they receive lifelong follow-up care. Reference John, Jackson and Moons5,Reference Moons, Bratt and De Backer6
The process through which individuals with CHD attain independent adult disease management, while simultaneously accomplishing medical, psychosocial, educational, and vocational developmental tasks, is referred to as the transition. Reference Moons, Bratt and De Backer6,Reference Carrizosa, An, Appleton, Camfield and Von Moers7 Those who undergo a suitable transition process gain an understanding of their disease, make informed decisions regarding necessary disease management and a healthy lifestyle (which includes physical activity, diet, career choices, sexual health, and contraception), and are able to utilise the social resources required for adult disease management. Reference John, Jackson and Moons5,Reference Moons, Bratt and De Backer6 Consequently, heart associations across the globe recommend providing transition programmes for adolescents and young adults with CHD. Reference Moons, Bratt and De Backer6
However, the actual impacts and effectiveness of transition programmes provided to individuals with CHD are not well-established. Despite growing interest in these programmes and related research, their practical implementation for CHD patients is not yet commonplace, Reference Campbell, Biggs and Aldiss8,Reference Wakimizu, Sasaki, Yoshimoto, Miyazaki and Saito9 and there is a clear need to establish evidence supporting the use of CHD transition programmes. Consequently, this study aimed to systematically review the literature on the effectiveness of transition programmes for adolescents and young adults with CHD, comparing their outcomes with control groups. Furthermore, through a meta-analysis of various effects, it is hoped that this study will contribute to the evidence base for transition programmes for CHD patients.
Methods
Search for studies
The literature search was conducted in accordance with the COre, Standard, Ideal search model. Reference Bidwell10 On February 26, 2023, keyword searches were executed across seven academic databases: CENTRAL, Embase, PubMed, Web of Science, CINAHL, KISS, and Kmbase. From February 26, 2023, to March 15, 2023, additional searches were carried out in clinical trial databases (ClinicalTrials.gov, ICTRP), on relevant academic journal websites, and through grey literature sources (ERIC, ProQuest, RISS, OPENGREY.EU). The search keywords comprised combinations of the following terms: “adolescent/puberty/young adult/pediatrics/child,” “heart defects, congenital/congenital heart malformation,” and “patient transfer/continuity of patient care/health transition/transition to adult care/self-management/patient transport/patient care/population dynamics/transition to adult care/self-care.”
Inclusion and exclusion criteria
The studies included adolescents (aged 10–17 years) and young adults (aged 18–29 years) who had simple, moderate complexity, or severe complexity CHD. Studies that involved non-structural cardiac conditions or individuals outside the age range of 10–29 years were excluded.
The experimental group (or exposure group) participated in a transition programme that incorporated five key components: an introduction to transition, medical knowledge, living with CHD, self-management, and self-advocacy. The goal of this programme was to foster independent adult health management, following the guidelines set forth by the American Heart Association (Table 1). Reference John, Jackson and Moons5,Reference Moons, Bratt and De Backer6 The intervention (or exposure) was deemed to have been provided if one or more topics within each component were addressed. The control group (or non-exposure group) either received no intervention or usual care. This control group comprised individuals who had not participated in a structured transition programme or received any other form of intervention, as defined in the literature.
The inclusion criteria encompassed randomised controlled trials, non-randomised controlled trials, cohort studies, and case-control studies, provided they included a control group. Cohort studies, single-group pre-test-posttest studies, cross-sectional studies, and descriptive studies that did not feature a control group were not considered.
Table 1. Transition programme components for adolescents and young adults with CHD
CHD: congenital heart disease.
Data extraction
Two researchers, BRL and SL, independently carried out literature selection. Following this, one researcher (BRL) extracted the data using a predefined data coding form, while two other researchers independently reviewed the data that had been extracted. The findings from the study selection and data extraction were shared among the researchers, with any discrepancies being resolved through discussion.
Quality evaluation
Quality evaluation was conducted using the Revised Tool to Assess the Risk of Bias in Randomized Trials (RoB 2.0) Reference Sterne, Savović and Page11 and the Risk of Bias Assessment tool for Non-randomized Study. Reference Kim, Park and Lee12 This evaluation was performed by two researchers, BRL and SL. The interrater reliability of the quality evaluation among the researchers was assessed using Cohen’s kappa statistics. In the event of any discrepancies in the assessments between the researchers, these were resolved through discussion and consensus. The certainty of evidence pertaining to each major outcome was analysed using the Grading of Recommendations Assessment, Development, and Evaluation approach.
Statistical analysis
Effect sizes were analysed using the meta package in R software version 4.2.1. Due to variation in participant age, as well as the composition and methods of interventions across the selected literature, a random-effects model was used to calculate the pooled effect sizes. Reference Higgins, Thomas and Chandler13 When three or more studies reported the same outcomes, the intervention effect estimates were combined. Reference Shenhav, Heller and Benjamini14 The effect size for continuous data was calculated using Hedges’ g, while odds ratios were used for dichotomous data. Additionally, 95% confidence intervals for effect sizes were computed. Reference Borenstein, Hedges, Higgins and Rothestein15 Hedges’s g is interpreted based on effect size as follows: small effect size (0.34 or lower), moderate effect size (0.35–0.74), and large effect size (0.75 or higher). Reference Cohen16 Heterogeneity was assessed using Cochran’s Q test, Reference Cohen16 and the extent of heterogeneity was determined using Higgins I2 statistic. Reference Higgins, Thomas and Chandler13 Subgroup analyses were performed to investigate whether the timing of effect measurement contributed to heterogeneity and to examine potential differences in loss to follow-up observations based on parental involvement. Publication bias was assessed using funnel plots, the Egger test, and trim-and-fill analysis. Reference Higgins, Thomas and Chandler13,Reference Duval and Tweedie17,Reference Harrer, Cuijpers, Furukawa and Ebert18
Results
Study selection results
We conducted a keyword search and reviewed the titles and abstracts of 13,429 papers, excluding any duplicates. From these, we selected 20 studies. However, four of these studies were abstracts for which the full text was not available, two clinical trials did not report their results, and two clinical trials were still ongoing, thus providing insufficient information about the characteristics or outcomes of transition programmes. Additionally, two papers were not related to transition programmes, and two were descriptive studies, leading to their exclusion from the analysis. This left eight studies in the initial selection. We also performed a manual search, which yielded an additional 11,351 papers. After reviewing the titles and abstracts of these papers, we selected 13 studies. Among these, six had already been chosen in the initial selection, two did not have full text available, two were not related to transition programmes, and one was a descriptive study. This resulted in the inclusion of two additional studies. In total, we selected and analysed 10 studies (Fig. 1).
Figure 1. PRISMA flow diagram of the study selection process. PRISMA = Preferred reporting items for systematic reviews and meta-analyses.
Intervention outcomes
In the 10 studies, outcomes were reported for disease-related knowledge, Reference Mackie, Islam and Magill-Evans19,Reference Goossens, Fieuws and Van Deyk20,Reference Mackie, Rempel and Kovacs23,Reference Lee and Jung24,Reference Mackie, Rankin and Yaskina28 loss to follow-up, Reference Hergenroeder, Moodie and Penny22,Reference Mackie, Rempel and Kovacs23,Reference Gaydos, Chowdhury, Judd and McHugh25,Reference Bushee, Ginde, Earing, Buelow, Reinhardt and Cohen26 self-management, Reference Mackie, Islam and Magill-Evans19,Reference Mackie, Rempel and Kovacs23,Reference Lee and Jung24,Reference Mackie, Rankin and Yaskina28 quality of life, Reference Lee, Lee and Choi21,Reference Lee and Jung24,Reference Hwang27 excess time between paediatric and adult CHD care, Reference Hergenroeder, Moodie and Penny22,Reference Mackie, Rempel and Kovacs23 self-advocacy, Reference Mackie, Islam and Magill-Evans19,Reference Mackie, Rempel and Kovacs23 health and health risk behaviour, Reference Goossens, Fieuws and Van Deyk20,Reference Hwang27 transition to adult CHD care, Reference Bushee, Ginde, Earing, Buelow, Reinhardt and Cohen26 unplanned cardiac hospitalisations, Reference Bushee, Ginde, Earing, Buelow, Reinhardt and Cohen26 and deterioration of heart failure status Reference Hergenroeder, Moodie and Penny22 (Table 2). Among these, disease-related knowledge was the most frequently measured outcome, assessed in five studies. Conversely, outcomes related to the impact on participants’ health, such as unplanned cardiac hospitalisations and deterioration of heart failure status, were each evaluated in only one study. Detailed outcome statistics for each specific study can be found in Table S1.
Table 2. General characteristics of the included studies
ACHD = adult congenital heart disease; CHD = congenital heart disease; Cont. = control group, Exp. = experimental or exposure group; HBS-CHD = Health Behavior Scale-Congenital Heart Disease, KYHBS = Korean Youth Health Behavior Survey, LKQ-CHD = Leuven Knowledge Questionnaire for Congenital Heart Disease; NRCT = non-randomized controlled trial; NYHAFS = The New York Heart Association Functional Classification of Heart failure; PedsQL = Pediatric Quality of Life; PCQL = Pediatric Cardiac Quality of Life Inventory; QoL = quality of life; RCT = randomised controlled trial; SRAHP = Health Self-Efficacy Measure; TRAQ = Transition Readiness Assessment Questionnaire; USA = United States of America.
* Upper age limit not determined.
Reporting statistical significance: ↑* Increase and a statistically significant result; ↑ Increase and not a significant result; ↓ Decrease and not a significant result; ↓* Decrease and a statistically significant result; (n) Not significant.
Intervention characteristics
The characteristics of interventions are summarised and presented in Table 3. Among the included transition programmes, six were not based on any specific model or theory. One study utilised guidelines and tools from GotTransition© (https://www.gottransition.org/), Reference Gaydos, Chowdhury, Judd and McHugh25 two were based on Bandura’s Reference Bandura29 self-efficacy theory, Reference Lee and Jung24,Reference Hwang27 and one applied the five psychosocial factors of resilience. Reference Lee, Lee and Choi21
The frequency of interventions varied, with four studies offering interventions once, Reference Mackie, Islam and Magill-Evans19–Reference Lee, Lee and Choi21,Reference Mackie, Rankin and Yaskina28 one study providing two interventions, Reference Mackie, Rempel and Kovacs23 one study conducting interventions weekly for 4 weeks, Reference Hwang27 and one study offering interventions weekly for 6 weeks. Reference Lee and Jung24 Two studies had varying intervention frequencies, with one providing interventions annually until the transfer to adult CHD care was complete, Reference Bushee, Ginde, Earing, Buelow, Reinhardt and Cohen26 and one providing interventions at least every 6 months until completion of the transfer. Reference Gaydos, Chowdhury, Judd and McHugh25 One study did not report the exact frequency of interventions. Reference Hergenroeder, Moodie and Penny22
Seven studies provided individualised transition programmes tailored to each patient’s condition and needs, Reference Mackie, Islam and Magill-Evans19,Reference Goossens, Fieuws and Van Deyk20,Reference Hergenroeder, Moodie and Penny22,Reference Mackie, Rempel and Kovacs23,Reference Gaydos, Chowdhury, Judd and McHugh25,Reference Bushee, Ginde, Earing, Buelow, Reinhardt and Cohen26,Reference Mackie, Rankin and Yaskina28 while three provided group transition programmes that involved group activities such as sharing experiences and exercising together among individuals with CHD. Reference Lee, Lee and Choi21,Reference Lee and Jung24,Reference Hwang27 Among the group transition programmes, two involved creating self-management plans and providing one-on-one coaching from nurses via phone or email during the self-management process. Reference Lee and Jung24,Reference Hwang27
The topics covered by almost all (9 or 10) of the transition programmes comprised anatomical-haemodynamic considerations (n = 10), managing medications (n = 10), medical appointments (n = 9), and communication with the healthcare provider (n = 9). In contrast, lifestyle topics, such as educational/vocational considerations (n = 3), anxiety and depression (n = 3), insurance (n = 2), and parent support (n = 1), were covered in three or fewer transition programmes.
Nurses were the primary providers of the interventions in seven studies, Reference Mackie, Islam and Magill-Evans19,Reference Goossens, Fieuws and Van Deyk20,Reference Hergenroeder, Moodie and Penny22–Reference Lee and Jung24,Reference Hwang27,Reference Mackie, Rankin and Yaskina28 while in two studies, the intervention was delivered by both nurses and physicians. Reference Lee, Lee and Choi21,Reference Gaydos, Chowdhury, Judd and McHugh25 Additionally, one study involved a multidisciplinary team, Reference Bushee, Ginde, Earing, Buelow, Reinhardt and Cohen26 and in all cases, nurses were part of the intervention delivery. Two studies included parents in the transition programme, considering their opinions when determining the timing of transferring to an adult care environment or providing brochures on transferring health management responsibilities to their children. Reference Gaydos, Chowdhury, Judd and McHugh25,Reference Bushee, Ginde, Earing, Buelow, Reinhardt and Cohen26
Three studies used electronic tools to efficiently manage participants during the transition process. Gaydos et al. Reference Gaydos, Chowdhury, Judd and McHugh25 created an electronic registry to manage the progress of the transition programme and follow-up. Goossens et al. Reference Goossens, Fieuws and Van Deyk20 used a computerised checklist during education to differentiate information that had been discussed with participants or that participants were already aware of. Furthermore, Hergenroeder et al. Reference Hergenroeder, Moodie and Penny22 developed an electronic medical record-based transition planning tool to track and assess participants, which served as the basis for providing the transition programme.
Table 3. Characteristics of the interventions of the included studies
ACHD = congenital heart disease; APN = advanced practice nursing; CHD = congenital heart disease; GP = group programme; IP = individual programme; NP = nurse practitioner; NR = not reported; USA = United States of America.
a Description of the transition.
b Need for lifelong care.
c Need to transfer to adult health care setting.
d Anatomical-haemodynamic considerations.
e Complications.
f Catheter and surgical interventions.
g Educational/vocational considerations.
h Physical activity.
i Pregnancy and birth control.
j Diet and weight management.
k Smoking/alcohol/drugs.
l Anxiety and depression.
m Connecting with others who have CHD.
n Parent support.
o Insurance.
p Medical appointments.
q Managing medications.
r Communication with the healthcare provider.
s Use of community resources.
Risk of bias assessment
Out of the 10 selected studies, four randomised controlled trials Reference Mackie, Islam and Magill-Evans19,Reference Mackie, Rempel and Kovacs23,Reference Hwang27,Reference Mackie, Rankin and Yaskina28 were evaluated using RoB 2.0 (5 domains), while the remaining six studies Reference Goossens, Fieuws and Van Deyk20–Reference Hergenroeder, Moodie and Penny22,Reference Lee and Jung24–Reference Bushee, Ginde, Earing, Buelow, Reinhardt and Cohen26 were assessed using Risk of Bias Assessment tool for Non-randomized Study (6 domains). The Cohen’s kappa for quality assessment across a total of 56 domains was 0.75 (95% CI = 0.61–0.85), indicating moderate agreement. Reference McHugh30
Randomised controlled trials
Mackie et al. Reference Mackie, Islam and Magill-Evans19 employed random allocation, clustering by the week of attendance in the cardiology clinic. However, a detailed description of the allocation sequence was not provided, resulting in a rating of “some concerns.” In all four randomised controlled trials, Reference Mackie, Islam and Magill-Evans19,Reference Mackie, Rempel and Kovacs23,Reference Hwang27,Reference Mackie, Rankin and Yaskina28 there were no deviations from the intended interventions, and the risk of bias in terms of missing data and outcome measurement was low. Two studies by Mackie et al. Reference Mackie, Rempel and Kovacs23,Reference Mackie, Rankin and Yaskina28 had follow-up durations ranging from 6 to 18 months, and some mean and standard deviation values were not reported. This raised concerns about potential bias. The overall bias assessment indicated that one study Reference Hwang27 had a low risk of bias, while three studies Reference Mackie, Islam and Magill-Evans19,Reference Mackie, Rempel and Kovacs23,Reference Mackie, Rankin and Yaskina28 were rated as having some concerns regarding the risk of bias (Fig. 2).
Figure 2. RoB 2.0 for randomised controlled trials. ROB = risk of bias assessment using a revised tool to assess the risk of bias in randomisedtrials.
Non-randomised controlled trials, Cohort studies, and case-control studies
The cohort studies conducted by Goossens et al. Reference Goossens, Fieuws and Van Deyk20 and Hergenroeder et al. Reference Hergenroeder, Moodie and Penny22 did not report baseline differences in outcome variables or utilised a historical control group, resulting in a high risk of bias in the selection of participants. Bias due to confounding variables was low in all six studies Reference Goossens, Fieuws and Van Deyk20–Reference Hergenroeder, Moodie and Penny22,Reference Lee and Jung24–Reference Bushee, Ginde, Earing, Buelow, Reinhardt and Cohen26 . Two studies Reference Lee, Lee and Choi21,Reference Lee and Jung24 collected all outcomes using self-reported tools, which raised concerns about the risk of bias. Lee and Jung’s study Reference Lee and Jung24 did not provide details about blinding of outcome assessors, leading to an unclear risk of bias. The study of Goossens et al. Reference Goossens, Fieuws and Van Deyk20 reported detailed results only for some participants who participated in all outcome measurements within the control group, resulting in an unclear risk of bias. The study of Goossens et al. Reference Goossens, Fieuws and Van Deyk20 did not report the mean and standard deviation for secondary outcomes, leading to a high risk of bias (Fig. 3).
Figure 3. Risk of bias assessment using RoBANS. RoBANS = Risk of Bias Assessment tool for Non-randomized Study.
Results of the meta-analysis
Disease-related knowledge
In the five studies reporting disease-related knowledge, Hedge’s g was 0.89 (95% CI = 0.29–1.48), indicating a large effect size that was statistically significant. Heterogeneity testing revealed a Cochran’s Q value of 32.7 (p < .001, df = 4), indicating heterogeneity in effect sizes across studies, and the degree of heterogeneity was substantial (Higgins I2 = 87.8%; Fig. 4). Since the timing of outcome measurements varied among the included studies, a subgroup analysis was conducted by dividing studies into those reporting immediate post-intervention results and those reporting results from follow-up assessments (Fig. 5). Disease-related knowledge showed a significant increase in the experimental group compared to the control group when measured immediately after the intervention (Hedge’s g = 2.26, 95% CI = 1.57–2.95). Furthermore, in follow-up assessments of 6 months or more, the experimental group exhibited a significant increase in disease-related knowledge compared to the control group (Hedge’s g = 0.55, 95% CI = 0.18-0.93), and this difference in effect sizes between the two groups was also significant (F2 = 18.07, df = 1, p < .001).
Figure 4. Forest plot of the effects of transition programmes for adolescents and young adults. TE = estimate of treatment effect; seTE = standard error of treatment estimate; SMD = standardised mean difference; CI = confidence interval; QoL = indicates quality of life.
Figure 5. Forest plot of the disease-related knowledge effect of transition programmes for adolescents and young adults according to the timing of measurement. TE = estimate of treatment effect; seTE = standard error of treatment estimate; SMD = standardised mean difference; CI = confidence interval.
Self-management
In the four studies reporting self-management, Hedge’s g was 0.67 (95% CI = 0.38–0.95), indicating a moderate effect size that was statistically significant. Heterogeneity testing showed a Cochran’s Q value of 3.97 (p = .265, df = 3), suggesting that the effect sizes across studies were not heterogeneous, and the degree of heterogeneity was low (Higgins I2 = 24.3%; Fig. 4).
Disease-related quality of life
In the three studies reporting disease-related quality of life, Hedge’s g was 0.60 (95% CI = −0.24 to 1.44), indicating a moderate effect size that was not statistically significant. Heterogeneity testing showed a Cochran’s Q value of 11.22 (p = .004, df = 2), indicating heterogeneity in the effect sizes across studies, and the degree of heterogeneity was substantial (Higgins I2 = 82.2%; Fig. 4). However, since the number of studies included in the effect size synthesis was less than four, a moderation analysis could not be conducted. Reference Shenhav, Heller and Benjamini14
Loss to follow-up
The pooled effect size for loss to follow-up, reported in four studies, showed that the experimental group receiving the transition programme had a 0.41-fold (95% CI = 0.22−0.77) lower rate of loss to follow-up compared to the control group, and this difference was statistically significant. The converted value for this effect size, represented by Hedge’s g, was 0.49 (95% CI = 0.15−0.83), indicating a moderate effect size. Heterogeneity testing revealed a Cochran’s Q value of 4.88 (p = .181, df = 3), indicating no heterogeneity in effect sizes across studies, and the degree of heterogeneity was moderate (Higgins I2 = 38.5%; Fig. 6). In line with the recommendation to include parents in transition programmes to facilitate the transfer of self-management responsibilities, Reference Moons, Bratt and De Backer6,Reference Sable, Foster and Uzark31 we analysed differences in the effect of loss to follow-up according to parental involvement. The results showed that when parents participated in the transition programme, the experimental group had a 0.36-fold (95% CI = 0.22−0.59) reduction in the loss to follow-up rate compared to the control group, and this was statistically significant. However, when parents did not participate, the OR for loss to follow-up was not statistically significant, and the difference in effect size between the two groups was also not significant (χ2 = 0.01, df = 1, p = .936; Fig 7).
Figure 6. Forest plot of the effect of transition programmes for adolescents and young adults on loss to follow-up. CI = confidence interval; OR = odds ratio.
Figure 7. Forest plot of the effect of transition programmes for adolescents and young adults on loss to follow-up according to parents’ participation. CI = confidence interval; OR = odds ratio.
Publication bias
Publication bias was assessed using the effect size of the primary outcome for each study. A right-skewed funnel plot was observed, and the Egger test indicated that the regression equation was not statistically significant (t = 2.20, p = .059). The Hedge’s g, adjusted using the trim-and-fill method, decreased compared to the unadjusted Hedge’s g. However, both effect sizes remained statistically significant. It was concluded that even though publication bias was present, it did not have a significant impact on the results of the studies (Figure S1; Figure S2).
Quality of evidence
The quality of evidence for the outcome variables included in the meta-analysis—namely, disease-related knowledge, loss to follow-up, self-management, and disease-related quality of life—was assessed. Disease-related knowledge and loss to follow-up were evaluated as having low-quality evidence, indicating limited reliability of effect estimates. Self-management and disease-related quality of life were assessed as having very low-quality evidence, suggesting that there is almost no certainty regarding the effect estimates. Detailed results are provided in Table S2.
Discussion
Research characteristics
Transition programmes for CHD patients have been implemented in the USA, Canada, South Korea, and Belgium since 2014. These programmes primarily target adolescents and young adults, with an average age range of 13−25 years. They encompass a broad age spectrum, from adolescents to young adults (n = 6), with the majority focusing on those with moderate to severe complexity CHD (n = 7).
Among the 10 transition programmes reviewed, over nine studies incorporated topics related to the anatomical-haemodynamic considerations of CHD (n = 10), medication management (n = 10), medical appointments (n = 9), and communication with healthcare providers (n = 9). The anatomical-haemodynamic considerations of CHD are part of the medical knowledge component of the transition programme, while medication management and medical appointments fall under the self-management component. Communication with healthcare providers, in contrast, is part of the self-advocacy component. Reference John, Jackson and Moons5 In other words, transition programmes for adolescents and young adults with CHD focus on enhancing their independent self-management and self-advocacy skills, grounded in a thorough understanding of the disease. Experts in the field of transition concur that acquiring skills related to medical knowledge, self-management, and self-advocacy is vital among the various topics covered in transition programmes. Reference Sable, Foster and Uzark31–Reference Mackie, Fournier, Swan, Marelli and Kovacs33 Self-management and self-advocacy skills can bolster a CHD patient’s medical independence, decision-making capacity in healthcare, and control over their psychosocial environment. Reference Sable, Foster and Uzark31 A lack of medical knowledge can hinder patients from developing self-management and self-advocacy skills. Reference Ladouceur, Iserin, Cohen, Legendre, Boudjemline and Bonnet34 Therefore, it is crucial for transition programmes to incorporate education and role supplementation concerning medical knowledge to understand the condition, self-management skills for managing the condition, and self-advocacy skills to enhance decision-making and communication.
Interventions related to educational/vocational considerations (n = 3), anxiety and depression (n = 3), and parent support (n = 1) were relatively underrepresented in transition programmes for lifestyle management. Education for patients with CHD is closely linked to adult employment and income. Reference John, Jackson and Moons5 Adolescents and young adults with CHD often encounter educational delays and face difficulties in making career choices. Reference Sable, Foster and Uzark31,Reference Kovacs, Brouillette and Ibeziako35 They are less likely to pursue higher education or vocational training compared to their peers without CHD. Reference Cocomello, Dimagli, Biglino, Cornish, Caputo and Lawlor36 Therefore, transition programmes for adolescents and young adults with CHD should consider incorporating additional educational and vocational interventions. Moreover, individuals with CHD have a higher lifetime prevalence of anxiety and depression compared to the general population, Reference Westhoff-bleck, Briest and Fraccarollo37 especially during the adolescent and young adult years when the transition process involves shifting disease management responsibilities from parents to the individual. This shift can intensify psychological distress in these individuals. Reference Kovacs, Brouillette and Ibeziako35 Parent support is also recognised as a crucial factor in the psychosocial well-being of CHD patients. Reference Luychx, Goossens, Rassart, Apers, Vanhalst and Moons38 Therefore, when developing transition programmes, it is important to include psychological interventions that are based on assessments of anxiety and depression. Additionally, considering parental support during the process of empowering patients to manage their own healthcare decisions can prove beneficial. Reference John, Jackson and Moons5,Reference Heery, Sheehan, While and Coyne39
Among the 10 transition programmes analysed, seven were individual programmes, outnumbering the group programmes. This contrasts with the findings of a systematic review of transition programmes for paediatric patients with chronic conditions, where individual and group programmes were reported to be equally represented. Reference Wakimizu, Sasaki, Yoshimoto, Miyazaki and Saito9 This discrepancy may be due to the diverse nature of cardiac diseases within CHD, which vary in severity, characteristics, and management strategies based on individual patients. Reference Flocco, Lillo, Dellafiore and Goossens40 The studies included in this analysis targeted various CHDs, ranging from simple to severe complexity or moderate to severe complexity. Consequently, a more personalised approach is often favoured. Notably, all three transition programmes reported in South Korea were group programmes. This could be because, in South Korea, adult CHD care is not distinct from paediatric care. Reference Eun41 Therefore, these programmes tend to focus more on enhancing the general disease management of adult CHD in a community-based group setting rather than individual, hospital-based approaches primarily aimed at transitioning to adult CHD care.
Among the transition programme studies analysed, four were observational studies, while seven demonstrated a risk of bias. The primary sources of bias included the omission of some outcome values when measurements were taken multiple times, the use of historical control groups, the failure to clearly report baseline differences in outcome variables, and self-reporting. Therefore, it is crucial to consider these biases when examining the effects of transition programmes in future research. Additionally, some studies did not provide detailed descriptions of programme aspects such as the exact age of participants, programme frequency, and duration. Consequently, future studies should adhere to reporting guidelines to ensure that transition programmes can be replicated.
Effects of transition interventions
After the transition programmes, the participants showed a substantial increase in disease-related knowledge, with a large effect size. In this context, disease-related knowledge refers to understanding medical aspects related to CHD, including its complications, medical interventions, and disease management strategies. Reference Moons, De Volder and Budts42,Reference Sawicki, Lukens-Bull and Yin43 Youth with CHD often have limited knowledge about topics such as physical activity, diet, and sexual health related to disease management, Reference De Lima Campos, Perin, Assmann, Lucchese and Pellanda44–Reference Van Deyk, Pelgrims and Troost46 and they frequently struggle to accurately locate their specific cardiac abnormalities and comprehend the complexities of cardiac defects and complications. Reference De Lima Campos, Perin, Assmann, Lucchese and Pellanda44,Reference Van Deyk, Pelgrims and Troost46 During the transition programme, it is recommended to employ strategies such as creating portable electronic files like the MyHealth passport to provide information about heart conditions and using diagrams to educate patients about anatomical defects. Reference John, Jackson and Moons5 In the transition programmes included in this study, similar approaches were likely employed. These programmes likely utilised visual materials, such as MyHealth passports and anatomical diagrams, to educate participants about medical knowledge, contributing to the improvement in disease-related knowledge. Notably, the study conducted by Lee and Jung Reference Lee and Jung24 , which reported the largest effect size, was based on self-efficacy theory. In addition to written educational materials, they produced successful disease management videos to provide audio-visual materials. Unlike other studies that delivered the intervention 1–2 times, they implemented the programme over 6 weeks and augmented knowledge comprehension by offering weekly phone and email support to address participants’ queries.
Disease-related knowledge showed a larger effect size immediately after the intervention than in follow-up assessments conducted after the intervention. According to the Forgetting Curve Theory Reference Ebbinghaus47 , knowledge tends to be forgotten at a rate of approximately 80% within 1 month. Therefore, it is plausible that the effect size diminished in assessments conducted after a period of 6 months or more. Given that the effectiveness of intervention on disease-related knowledge may wane over time, regular re-evaluations and supplementary education are essential. However, in this meta-analysis, the follow-up periods across the studies varied from 6 to 26 months, highlighting inconsistencies between studies. Further research is required to ascertain the optimal timing for reassessment.
Self-management among participants saw a moderate increase following the transition programme. This concept of self-management includes behavioural aspects such as obtaining prescribed medications when necessary, arranging follow-up appointments, and attending medical consultations. Reference Riegel, Lee, Dickson and Carlson48 To encourage changes in health behaviour, interventions that not only impart knowledge but also promote self-monitoring, risk communication, and social support have proven effective. As such, interventions should incorporate behaviour change strategies that combine knowledge, awareness, and facilitation. Reference Van Achterberg, Huisman-de Waal, Ketelaar, Oostendorp, Jacobs and Wollershein49 The transition programmes analysed in this study utilised various role supplementation techniques for awareness and facilitation. These included discussions, encouragement, sharing and providing experiences, and offering relevant resources. Of particular note is the study by Lee and Jung, Reference Lee and Jung24 which reported a large effect size. This study provided participants with a self-management diary to encourage self-monitoring of medication adherence and symptoms. They also offered one-on-one telephone counselling to promote self-management and mailed necessary resources to participants. Additionally, the programme facilitated discussions among participants, allowing them to share their self-management experiences.
Loss to follow-up decreased after the transition programmes, with a moderate effect size. In a previous meta-analysis by Moons et al., Reference Gilboa, Devine and Kucik1 the effect size for discontinuation of follow-up in CHD patients after transition programmes was not statistically significant. However, in this study, the addition of one more study to the effect size pooling conducted by Moons et al. Reference Gilboa, Devine and Kucik1 yielded a significant effect on loss to follow-up. Furthermore, a meta-analysis by García-Rodríguez et al., Reference García-Rodríguez, Raygoza-Cortez and Moreno-Hernandez50 which included various chronic disease populations, also found a significant effect size for clinic drop-out. By combining the results of the previous meta-analysis with the findings of this study, it can be deduced that transition programmes may help reduce loss of follow-up. A recent systematic review identified structured transition programmes led by nurses and formal handovers to adult CHD care as strategies to decrease treatment discontinuation in CHD patients. Reference Bassareo, Chessa, Di Salvo, Walsh and Mcmahon51 The transition programmes evaluated for loss to follow-up in this study employed structured transition programmes, which included guidelines from Got Transition©, electronic medical record-based transition planning tools, and shared patient information among healthcare providers. These programmes were primarily nurse-led or included nurses as part of the transition team. Notably, the study by Mackie et al., Reference Mackie, Rempel and Kovacs23 which reported a significant effect size on the reduction in loss to follow-up, involved paediatric cardiac nurses visiting an adult CHD outpatient clinic with patients and directly transitioning them, thereby reducing the rate of loss to follow-up.
The analysis found no statistically significant differences in loss to follow-up based on parental involvement. However, it’s important to note that the number of studies included in this meta-analysis was limited, and the specific roles of parents and the details of their interventions in transition programmes were not thoroughly described. The reason is that when parents become overly involved in the transition process and hinder their child’s acquisition of independence, parental involvement can actually diminish the effectiveness of transition programmes. Ideally, parents can play a significant role in the follow-up process during transitions by engaging in discussions with their children about the follow-up before visiting the outpatient clinic, preparing a list of questions for healthcare providers, and encouraging their children to communicate directly with these providers and ask questions. Reference Moons, Bratt and De Backer6 Therefore, a structured transition programme that involves parents from the planning stage of transitions, providing them with relevant information, is necessary. This will enable parents to participate appropriately and play a positive role in the transition process. Subsequently, an analysis can be attempted to evaluate the effectiveness of parental involvement.
The effect of the transition programmes on disease-related quality of life was not statistically significant. The disease-related quality of life in adolescents and adults with CHD is influenced by the complexity of the CHD, clinical symptoms, and complications, with only limited associations with demographic and socio-economic factors. Reference Bertoletti, Marx, Hattge and Pellanda52–Reference Liu, Chaou, Lo, Chung and Hwang54 Therefore, to enhance disease-related quality of life in individuals with CHD, the focus should be on alleviating clinical symptoms and complications associated with CHD. A prior meta-analysis indicated an improvement in disease-related quality of life when individuals with CHD participated in an exercise programme lasting from 10 weeks to 3 months, which improved cardiac function and reduced the need for medication. Reference Xu, Su and Ma55 In this study, the transition programmes that evaluated disease-related quality of life also incorporated topics aimed at improving cardiac health through lifestyle factors such as physical activity, diet, and weight management. These programmes were intended to motivate participants to actively participate in exercise and make dietary modifications. However, these programmes were relatively brief, lasting only 1 day or 6 weeks. To effectively improve disease-related quality of life, future programmes should include topics related to lifestyle management, focusing on enhancing cardiac function and clinical symptoms, and should be implemented over longer periods, such as 10 weeks or more, to evaluate their effectiveness.
This study marks the first attempt to conduct a meta-analysis of the varied effects of transition programmes for individuals with CHD, by identifying five key components of CHD-targeted transition programmes and analysing appropriate programmes based on these components. By defining transition programmes clearly and conducting an exhaustive literature search with heightened sensitivity, this research sought to strengthen the validity of effect assessment. This was achieved by selecting studies that included control groups and analysing their outcomes to yield meaningful conclusions. Consequently, this study carries significant importance as it provides a current overview of transition programmes for adolescents and young adults with CHD. This serves as an evidence base for researchers and practitioners interested in developing and implementing transition programmes for CHD patients in the future.
Limitations
This study has several limitations: First, the specific topics included within the transition programme components and the structures of the programmes themselves varied, resulting in significant heterogeneity among the studies. Alternatively, most studies had similar characteristics, such as the classification of CHD and providers. Therefore, due to a lack of studies for each outcome, we could not perform subgroup analysis on characteristics with substantial diversity and identical characteristics. As a result, there were limitations in understanding which factors had a positive impact on specific outcomes. Second, there was a dearth of studies reporting actual health-related outcomes, such as emergency room visits, hospitalisations, and disease or complication status, which precluded an analysis of their effects. Third, methodological weaknesses in the included studies (risk of bias), high heterogeneity between studies, and small sample sizes resulted in low or very low levels of certainty in the evidence produced by the meta-analysis. Therefore, these limitations should be taken into account when interpreting the results of the analysis.
Conclusions
This study conducted a systematic review of transition programmes designed to aid adolescents and young adults with CHD in managing their condition as they transition into adulthood. To evaluate the effectiveness of these programmes, a comprehensive search of research literature, clinical trials, and grey literature was undertaken. Ten studies that met the selection criteria were ultimately analysed. The transition programmes for adolescents and young adults with CHD focused on providing education and role supplementation to enhance their medical knowledge to understand the disease, their self-management skills, and their self-advocacy skills to improve decision-making and communication. The impact of these transition programmes was evident in the improvement of disease-related knowledge, self-management, and a reduction in loss to follow-up. However, there was no significant effect on improving disease-related quality of life. It is important to acknowledge that the studies included in this review had methodological weaknesses (risk of bias), a high degree of heterogeneity, and limited sample sizes. As a result, the evidence supporting the effect estimates was of low or very low certainty. Therefore, caution should be exercised when interpreting these results, and further replication studies are recommended for the future.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S104795112400026X.
Acknowledgement
None.
Financial support
None.
Competing interests
This article was adapted from a thesis by Bo Ryeong Lee in partial fulfilment of the requirements for the dissertation at Daegu Catholic University of Korea.
