Search results
Of the 751 results, 20 papers or abstracts on transition-readiness tools were found. Ten validation studies were published in peer-reviewed journals, and these were critically appraised. Ten conference abstracts were also found – 8 of which involved validation of a new tool, and 2 using or re-validating previously validated tools. Numerous checklists were found by searching with Google, including some listed as resources by the American consensus statement [
5]. The conference abstracts and general checklists contained inadequate information to allow formal evaluation and were subsequently excluded from analysis.
The tools published in peer-reviewed journals were variable in their design (listed in Table
1) and validation (Table
2). Most tools relied on patient self-report, and included questions on disease knowledge and self-management. Validation by measures of independence, knowledge, or self-management was most common. Seven of the 10 tools were disease-specific, with the majority directed at CF or solid organ transplantation. Most tools scored poorly according to the Terwee criteria (See Table
3) [
7].
Table 1
Design of transition readiness assessment tools
| Chronic diseases | 2 | 33 | Skills | 5-point Likert scale (Stages of change model) | Patient | 1 point each |
| Chronic diseases | 10 | 33 | Knowledge & self-management | Interview style (cross-referenced with medical records) | Patient | Each domain=1 Maximum 10 |
Self-Management Skills Assessment Guide [ 16] | Chronic diseases | 1 | 21 | Health-care awareness & decision-making | 5-point Likert scale | Patient & parent | 1-5 for each item, total score: 105 |
| CF | Multiple | 44 | Independent knowledge/behaviour | Yes/no | Parent | 1point for Yes, 0 for No. Maximum 44 |
Readiness Questionnaire [ 9] | CF | 2 | 24 | Knowledge & behaviour | Multiple choice or short answer | Patient | 1point each, Maximum 24 |
| Renal transplant | 3 | 22 | Involvement in behaviours & overall transition readiness | 4-point Likert scale (not/sometimes/often/always) | Patient & parent | 1-4 each question Maximum 48 |
| Liver transplant | 4 | Patient: 42. Parent: 36 | Skills, knowledge, behaviour | Likert scale & skill demonstration | Patient & parent | Items vary, Maximum 126 (parent: 108) |
| T1DM | 1 | 7 | Readiness to change responsibility | 5-point scale (Stages of change model) | Patient & parent | 1-5 each item, maximum 35 |
| Sickle cell disease | 5 | NR | NR | Variable, mostly 3-point Likert Scale (knowledge section worth 4 points) | Patient | NR (high score = more ready for transfer) |
| HIV | 6 | 21 | Knowledge of disease & transition process | NR | Patient | 1point each, 2 for disease knowledge |
Table 2
Cohort characteristics of transition readiness assessment tools
| Chronic diseases | 192 | 16-26 (mean: 19.7) | NR | 64% white, 56% female | USA |
UNC | Chronic diseases | NR | 12-22 | NR | NR | USA |
Self-management skills assessment guide [ 16] | Chronic diseases | 49 | 11-19 (mean: 15.6) | NR | 86% Caucasian | Canada |
| CF (with pancreatic insufficiency) | 76 patients, 70 parents | 4-17 (mean: 11.2) | NR | 94% white | USA |
Readiness questionnaire [ 9] | CF | 36 | NR | NR | NR | Canada |
| Renal transplant | 48 patients, 32 parents | 15-21 (mean: 18.6) | 5.73 y | 58% white, 29% black, 10% Hispanic | USA |
| Liver transplant | 71 patients, 58 parents | 11-20 (mean: 15.6) | 1-19 y (mean: 9.4 y) | 56% female | USA |
| T1DM | 69 | 12-17 (mean: 14.2) | >6 months (mean: 5.49) | 87% Caucasian | USA |
| Sickle cell disease | 69 | 14-21 | NR | 43% female | USA |
| HIV | 65 | 9-25 (mean: 15.8) | 1-17 years (mean: 10.5) | 44% female, 45% Caucasian | USA |
Table 3
Scoring of psychometric measures of transition-readiness tools by Terwee criteria
| + | +FA | +: total (0.93), domain 1 (0.92), domain 2 (0.82) | +: 100% (age, gender, disease type) | 0 | 0 | 0 | 0 | ? + mean/SD 0 MIC |
?sample size | |
| + | 0 | 0-? used PC | 0: inferred from development | + | ? Small cohort (n = 35) | 0 | 0 | 0 |
Self-management skills assessment guide [ 16] | – | 0 | +: 0.93 | +: 100% (correlation with independence) | 0 | ? inter-rater PC 0.56 Small cohort (n = 47) | 0 | + | 0 |
| - | 0 | +: 0.93 | ?: no hypotheses, correlates with age, years since diagnosis | ?: Small cohort (n = 35) | 0 | 0 | 0 | 0 |
Readiness auestionnaire [ 9] | - | 0 | 0 | +: 100% (caregiver ratings) | 0 | ? inter-rater PC 0.65 Small cohort (n = 36) | 0 | 0 | 0 |
| - | 0 | +: 0.79-0.94 | +: 86% (responsibility, medication knowledge, self-refilling, family relationship, decreased family involvement, adherence), no correlation with age | 0 | ? inter-rater PC 0.5-0.68 Small cohort (n = 48) | 0 | 0 | 0 |
| - | +FA ?sample size | -: 0.19-0.85 | -: 50% (self-management, age NOT adherence or health outcomes) | 0 | 0 | 0 | + | ? ? mean/SD (age) 0 MIC |
| - | 0 | +: 0.85-0.9 | +: 100% (responsibility, self-efficacy, decreased parenting stress) | 0 | ? inter-rater PC 0.33-0.58 | 0 | 0 | ? + mean/SD ? MIC |
| 0 | 0 | 0 | +: 100% (age, disease severity, gender) | 0 | 0 | 0 | 0 | ? + mean/SC ? MIC |
| - | 0 | 0 | +: 80% (anxiety, confidence in GP, decreased treatment length, improved with intervention) | 0 | 0 | 0 | 0 | 0 |
The Self-Care Independence Scale (SCIS) is a 44-item carer-report questionnaire assessing the child’s ability in and knowledge of, disease management. It was tested on 75 families who had 4-17 year olds suffering from CF with pancreatic insufficiency [
8]. The majority of children were Caucasian, of above average intelligence and above average socioeconomic status. This scale did not receive a positive rating for any measure in the Terwee criteria. A factor analysis of the scale wasn’t reported and thus its internal consistency is indeterminate despite an excellent Cronbach’s alpha (α = 0.93). The reproducibility of the scale appeared good due to a test-retest correlation of (r = 0.81), however the sample size of 35 did not meet the minimal criteria of 50 proposed by Terwee [
7]. The scale correlated with age (r = 0.67), years since diagnosis (r = 0.58), CF knowledge (r = 0.62), and general independence as determined by the 21-item validated Highland Dependency Questionnaire (r = 0.62), however the authors did not report their hypothesis and thus the construct validity is indeterminate. It also has a very specific cohort and consequently, the validity of the tool in other patient groups is uncertain [
8].
Cappelli
et al. developed another CF-specific questionnaire with 21 items testing disease knowledge and behaviour [
9]. It was validated in Canada by comparing the total readiness score with nominal caregiver ratings of either able or not able to cope with transfer. According to this measure, 77% of the adolescent respondents were correctly classified by summated questionnaire scores. Some limitations of this study are that content validation did not involve input from adolescents, and reliability was largely untested.
The Readiness for Transition Questionnaire (RTQ) is a 10-item tool for patients with kidney transplants. Notably, it uses multiple reporters, has an additional question about ‘overall transition readiness’, and also includes an assessment of non-adherence [
10]. Non-adherence is a significant barrier to successful transition as it is thought to be the cause of the high rates of kidney transplant failure in adolescents and young adults [
10]. The construct validity of the RTQ was therefore assessed using the Medical Adherence Measure. Medication adherence was found to contribute 33% of the variance of overall transition readiness scores, suggesting that adherence is a strong indicator of transition-readiness. However, despite transition readiness scores increasing with age, adherence actually decreased, raising the possibility that non-adherence occurs independently of other aspects of transition readiness. The RTQ also correlated with adolescent responsibility (r = 0.68), decreased parental involvement (r = -0.39), medication knowledge, self-refilling behaviour, and family relationship. There was a good Cronbach’s alpha, however internal consistency was rated as indeterminate due to a lack of factor analysis. Reliability was additionally scored as indeterminate because inter-rater reliability was tested using Pearson’s correlations (r = 0.5-0.68) without a weighted kappa.
Adherence, measured by blood levels of immunosupressants, was also used to validate the Transition Readiness Scale (TRS) for patients with liver transplants [
2]. The authors reported psychometrics for both an adolescent and parent version. A factor analysis was performed, however there was wide variability in Cronbach’s alphas. Most domains had good Cronbach’s alphas, but some scored <0.7. Construct validity was questionable as adherence and health outcomes did not correlate with total score. It should be noted that measuring non-adherence is a difficult and inexact science and although frequently used, both patient self-report and therapeutic drug monitoring may underestimate the extent of non-adherence [
11,
12].
Kaugars
et al. developed a 7-item questionnaire to assess the readiness of patients with Type 1 Diabetes, and their parents, to change the balance of responsibility of disease management [
13]. It was validated on 69 families in the US by correlation with self-efficacy scores (r = 0.90), decreased parental stress (r = 0.94), and responsibility. It has good Cronbach’s alpha (α = 0.85-0.9), but factor analysis and weighted kappa were not reported. There was poor inter-rater correlation between parent and patient (r = 0.58) and between mother and father (r = 0.33). Furthermore, it only assesses readiness to change responsibility and does not assess transition-readiness directly. Similar to the SCIS [
8], the study cohort was quite homogenous being 87% Caucasian, and 90% of parents having a college education.
Mcpherson
et al. employed a sickle cell disease-specific questionnaire with 5 domains: knowledge, thought, interest, difficulty, and importance of transition. Interest and knowledge domains correlated with age and disease severity in a group of 70 adolescents from a single centre. Reliability, reproducibility, and content validity were not reported, and there was a high non-response rate of 71% [
14].
The Transition Readiness Questionnaire (TRQ) is a 21-item HIV-specific questionnaire that was administered to patients before and after attending a transition program, an average of 6.8 months apart. The questionnaire assesses six variables including ability to arrange appointments, awareness of financial factors and knowledge of disease status and medications. Construct validity was adequate; transition-readiness scores were found to improve with time and were inversely related to state anxiety at baseline. However reliability and reproducibility testing were not reported [
15]. Although there is more focus on the improvement of scores with transition programs than the accuracy of the tool in predicting transition-readiness, this study highlights a purpose of the tool in identifying ‘problem areas’ which can then be targeted by transition programs.
The Self-management Skills Assessment guide is a 21-item youth and parent questionnaire. The scores increased with general independence as assessed by the Highland Questionnaire. Scores did not correlate with age, gender, ethnicity, or parent education. Internal consistency is indeterminate despite good Cronbach’s alphas (α = 0.89-0.93) as factor analysis was not reported [
16]. As with the SCIS, it measures self-management skills as a construct of transition-readiness, and although this relationship is theoretically assumed, evidence supporting it is minimal.
The Transition Readiness Assessment Questionnaire (TRAQ) involves 33 questions assessing skills/actions from 2 domains: self-management and self-advocacy, with 5 responses adapted from the ‘Stages of Change’ model. It was validated on 192 patients at 2 sites where the TRAQ score, as hypothesised, correlated with age, gender, and disease groups, but not race. It has excellent internal consistency with a Cronbach’s alpha of 0.93 and was one of two studies which conducted a factor analysis, although the sample size was arguably too small (6 patients/item, as opposed to 7 recommended by Terwee [
7]). It was also one of two tools whose development included a pilot on a group of adolescents, and thus received a positive rating for content validity. Its test-retest and inter-rater reliability were not reported [
4].
Ferris
et al. suggested that the TRAQ’s validity may be impaired as it relies on self-reporting, and instead offered the UNC TR[x]ANSITION, a tool with 33 questions across 10 domains which can be cross-referenced with medical records [
17]. It was the only other study that received a positive score for content validity, however construct validity wasn’t analysed and instead inferred from the development of the tool (interview of transition experts and 3 pilot tests on 185 adolescents in total). The authors did not report internal consistency by factor analysis or Cronbach’s alphas, and the inter-rater reliability (κ = 0.71) was performed on a relatively small cohort.
The only study of transition-readiness to originate outside of North America was a large exploratory study of factors that contribute to transition-readiness in Dutch adolescents [
18]. No transition-readiness assessment tool was developed, and instead, item specific scores were compared to each participant’s self-assessment of transition-readiness. Eleven variables significantly contributed to transition-readiness, including demographic factors (age, gender, ethnicity), attitude towards transition, impact of the disease, and health care self-efficacy. The main limitation of the study is the reliance on self-report for transition-readiness. The non-response rate was also notably high (64%).