Patients between 12 and 21 years old with any chronic disease treated by specific pediatric subspecialties (Endocrinology, Rheumatology, Hematology, and Pneumology) of any gender were included. Patients included in the present study were receiving care at the outpatient consult at Hospital Universitario “José Eleuterio González” in Northeastern Mexico between 2020 and 2022. Patients with a neurological profile incompatible with the administration of the TRAQ questionnaire were excluded.

The patients included in each phase of the study (pilot testing, preliminary validation, and final validation) were independent groups, meaning no patient participated in more than one phase. The number of patients participating during each phase is specified in each section.

The TRAQ 5.0 questionnaire is an instrument that includes 20 items divided into five domains: (a) managing medications, (b) appointment keeping, (c) tracking health issues, (d) talking with providers, and (e) managing daily activities. It was specifically created for the evaluation of adolescent and young adult patients with chronic diseases or special healthcare needs.5,8 Each item is graded on a scale from 1 to 5, with 1 being “No, I do not know how,” and 5 being “Yes, I always do this when I need to.”

This process was undertaken in six phases.

Phase 1: Questionnaire translation. Two independent translations of the original English TRAQ were made into Mexican Spanish. Two non-physician-certified translators participated. One was familiar with the questionnaire terminology, while the other was not and had no prior medical knowledge.

Phase 2: Translation synthesis. The translators and a medical researcher met to compare both translations, looking for discrepancies or ambiguities. Finally, a synthesis of both versions was made.

Phase 3: Expert committee. The committee mentioned above was formed by 25 participants and included a PhD researcher with prior experience in transcultural adaptation and instrument validation, a medical team of the pediatric specialties mentioned above, a linguistics expert, three certified translators, and ten pediatric patients with chronic diseases that fulfill the inclusion criteria were randomly selected during routine visits to the clinic, as long as they agreed to participate in the discussions and matched the inclusion criteria of the present study.

The committee reviewed the translation synthesis, discussed the items, and modified the questionnaire to achieve semantic, idiomatic, and conceptual equivalence to the original TRAQ. In addition, changes based on the characteristics of the Mexican healthcare system were made to achieve transcultural adaptation. Decisions on modifications were made based on consensus from the committee participants.

A linguistics expert copyedited the committee's version, which resulted in the test version.

Phase 4: Reverse translation. A third certified translator (non-physician, native English speaker, without knowledge of the translation process) translated the final version back into English to ensure that the content was faithful to the original.

Phase 5: Pilot test. The test version was administered to 40 patients according to the inclusion and exclusion criteria. During the pilot test, common issues included overly literal translations and uncommon vocabulary. Medical staff flagged questions requiring clarification, which were revised for clarity and cultural relevance. Terms were adapted to better reflect Mexican healthcare terminology. All changes were approved by the expert committee. The final version showed improved comprehension, with no further changes needed after the reverse translation.

Phase 6: Validation. This phase included 141 patients who completed the final version of the TRAQ. Administrators were instructed to intervene as little as possible and, when needed, to only read the question as it was written without additional input. Patients feeling capable of answering the questionnaire alone could do so after briefly explaining the instrument and how to complete it. The purpose of having a healthcare professional administer the questionnaire was to identify potential issues related to the translation or comprehension of its questions.

The whole translation and transcultural adaptation process is summarized in Fig. 1.

Fig. 1.

Summary of the adaptation and validation process.

Descriptive statistics with central tendency (means) and dispersion measures (standard deviation (SD)) were obtained for numerical variables. Categorical variables were described as absolute numbers and percentages. Correlation matrices were obtained for the whole instrument and for each domain to evaluate questionnaire redundancies. A p-value of less than 0.05 was considered statistically significant. Variables with a correlation coefficient of more than 0.70 were reevaluated.

The reliability of the instrument was evaluated through homogeneity and internal consistency (Cronbach's alpha) for both the total score and for each domain. Cronbach's alphas equal or greater than 0.40 with a p-value of less than 0.05 were considered acceptable.9,10 These procedures followed the Ramada-Rodilla et al.11 guidelines and were similar to the ones used in the original validation of Wood and Sawicki.5

The statistical analysis was run on Stata version 16.1.

This study was approved by the internal Research Ethics Committee (Comité de Ética y de Investigación de la Facultad de Medicina and Hospital Universitario “José E. González” de la Universidad Autónoma de Nuevo León (number PE20-00008)). Informed consent was obtained from every patient or their legal guardian if needed, and informed assent was also obtained from patients under 18 years of age. Confidentiality was preserved throughout the study.

Phases 1–4 were carried out as specified in “Methods” section. Two additional meetings were held during these phases. During the first, the expert committee fine-tuned translation details and made minor syntax changes, such as rewriting the questions in the first person. At the second meeting, the expert committee and the participating patient group analyzed the translated instrument to identify difficult-to-understand questions flagged by the medical staff who applied the pilot test; syntax changes were made as needed.

Two questions, 9 and 10, were changed to better reflect the Mexican healthcare system. IMSS (Mexican Institute of Social Security) and ISSSTE (Institute for Social Security and Services for State Workers) are Mexico's main public healthcare institutions, providing medical services and social security benefits to private-sector employees and government workers, respectively. In this sense, question 9 was changed from “9. Would you apply for a new health insurance after losing your current coverage?” to “9. If you have any medical service or coverage (IMSS, ISSSTE, Major Medical Insurance, etc.), have you taken the necessary steps (or do you know what steps you must take) to maintain or renew it when you are of legal age?” Question 10 was changed from “10. Do you know what your health insurance covers?” to “10. In case you have any medical service or coverage (IMSS, ISSSTE, Major Medical Insurance, etc.), do you know what it covers or what you are entitled to (medicines, hospitalization, payments to doctors, reimbursements, etc.)?

Additionally, question 15 regarding financial support was considered inadequate since not all patients have access to assistance. Therefore, it was changed to include a “Not applicable” answer option. After the final revision of the instrument, a thorough syntax review was conducted to ensure clarity and accuracy.

During this phase, the translated TRAQ was administered to 40 patients. The participants’ age range was 12–21, and the mean age of the patients was 13.9 years (SD 1.75). Thirty two (80%) of the patients were female. Most (28, 70%) came from the endocrinology consult, and 12 (30%) from the rheumatology consult.

According to their scores, patients were generally well prepared for TC, especially regarding domains 1 (managing medications) and 5 (managing daily activities). Questions 9, 10, and 15 showed the lowest scores: 26 (65%) answered “No, I do not know how” in question 9; 25 (62.5%) in question 10; and 27 (67.5%) in question 15.

Minor changes in syntax were made after the pilot test to obtain the final translated version.

The final translated version of the TRAQ in Mexican Spanish and the reverse translation into English are shown in the Supplementary material (Supplementary Tables 1 and 2, respectively). The final translated version was administered to 141 patients. The mean age was 13.9 years (SD 1.75), 81 patients (57.45%) were female, and every single patient (100%) lived in urban areas. Regarding health insurance coverage, 43 (30.5%) patients had full coverage, 28 (19.8%) had partial coverage, and 70 (49.7%) had no coverage by either private or public health insurance (Table 1). The most common diagnoses among participating patients were acute lymphoblastic leukemia (28, 19.86%), asthma (13, 9.22%), idiopathic juvenile arthritis (15, 10.64%), type 1 diabetes mellitus (15, 10.64%), and hypothyroidism (8, 5.67%). At the time of the study, 109 (77.30%) patients were currently undergoing treatment.

The results of the administration of the final translated version (in English for the purposes of this manuscript) are shown in Table 2. Generally, the domains where the patients showed more preparation were 4 (talking with providers) and 5 (managing daily activities). In domain 1, question 3 (“Do you know how to take your medications correctly and without anyone's support (without help)?”) also showed a high score, with most patients responding “Yes, I always do this when it is required.” (65.96%) or “Yes, I have started to do it.” (19.86%). In domain 3, question 15 (“Do you receive any type of financial support for school or work?”) had the lowest score (74.4% answered “No, I don’t know how.”).

Hematology patients were more prepared to communicate with their physicians when detecting changes in their health status (question 8). Additionally, rheumatology patients are more frequently referred to as taking their medications correctly and without support (question 3). Pneumology patients had more frequent positive responses in domain 4. The Supplementary material (Tables 3–6) shows the full results divided by pediatric subspecialty.

Internal consistency analysis showed a global Cronbach's alpha of 0.76, while the results by domain oscillated between 0.47 and 0.60 (Table 3). Statistical analysis of the convergent and divergent validity through correlation matrices showed positive correlations with greater frequency (Supplementary material, Table 7). In domain 1, the most significant correlation was found between items 1 and 2 (0.4493, p=0.000). Similarly, items 5–6 in domain 2 showed a positive correlation greater than 0.40 (0.4899, p=0.0000). Domain 3 only had mild and moderate positive correlations. Domain 4 only had one moderate positive correlation (0.4112, p=0.0000). Finally, in domain 5, a positive and moderate correlation exists between items 18 and 19 (0.4252, p=0.0000). While domain 2 had negative correlations, these were all non-statistically significant.