Translation and adaptation of the Pediatric Cardiac Quality of Life Inventory to the Portuguese population

Background: Congenital heart diseases have a high prevalence in Portugal in the pediatric age. Understanding the quality of life of children with congenital heart diseases will allow to know its impact and the adaptive process involved. We lack, however, measurement instruments adapted to the Portuguese population. Objective: Translate and adapt the Pediatric Cardiac Quality of Life Inventory (PCQLI) for the Portuguese population. Methodology: The original questionnaire was translated according to international recommendations. The final ver- sion of the instrument was applied in a non-probability sample comprised 59 children, 80 adolescents and their parents, totaling 278 respondents. The facial validity analysis was performed with the pre-test and internal consistency analysis using Cronbach’s Alpha. Results: In the pre-test it was verified that the items are comprehensible, useful with a simple and colloquial language. The internal consistency for the total score varied for the different versions between 0.834 and 0.932. Conclusion: The Portuguese version of PCQLI allowed us to evaluate the quality of life of children with congenital heart disease in a satisfactory way. However, a larger sample is needed to develop the validation process.


Introduction
Congenital heart disease (CHD) is one of the most prevalent congenital diseases in Portugal, with an incidence of 6 to 8 per 1,000 live births (Kaku, 2018). Over the past decades, several factors have contributed to significantly reduce the morbidity and mortality rates of children with heart disease and, consequently, increase their life expectancy. Thus, a growing number of patients with CHD reach adolescence and adulthood. However, the literature on comorbidities associated with chronic diseases in children and adolescents is scarce. Quality of life has been difficult to define due to its complexity, comprehensiveness, and multidimensionality. According to the World Health Organization (WHO, 2014), quality of life is defined as an individual's perception of their position in life in relation to their goals, expectations, and concerns. Therefore, in addition to the complexity of defining this concept, it is also difficult to conduct studies which provide a comprehensive knowledge of the reality experienced by each child, adolescent or parent, in various contexts and stages of their life cycle. The perception of quality of life varies from person to person, across families. For this reason, it is important to assess the personal perspective of each child or adolescent and their parents (Marino, Cassedy, Drotar, & Wray, 2016;Niemitz et al., 2017). The analysis of the characteristics of the quality of life of children/adolescents with CHD is of undeniable interest because it will allow identifying the impact of the disease and/ or treatments and better understanding patients, their evolution, and adaptation to the disease. In addition, it will improve the communication between patients, parents, and health care providers, the definition of priorities based on patients' preferences, the monitoring of changes over time or in response to specific therapeutic strategies, the detection of other medical or psychosocial problems, the identification of risk factors for low quality of life, the implementation of a family-centered care philosophy, the promotion of adherence to treatment, and the satisfaction of the patient and the family (Marino et al., 2010;Marino et al. 2016). Given the lack of instruments in European Portuguese for assessing the quality of life of children and adolescents with CHD, a literature review was performed to identify scales in other languages that could be used in this research. After a detailed analysis of these scales, the Pediatric Cardiac Quality of Life Inventory (PCQLI) by Marino et al. (2008) was selected. Permission was obtained for its use. Thus, the objective of this study was to translate and adapt the PCQLI to the Portuguese culture.

Background
CHD is the most common birth defect. According to Muner-Hernando, Gil-Mira, and Zapardiel (2013), it represents 46% of neonatal deaths and is associated with moderate diseases in the long-term. Today, more than 90% of newborns with a diagnosis of CHD reach adult life (Daudt & Zielinsky, 2013). Survival rates vary by disease complexity: long-term survival (> 20 years) rates for children with simple CHD are estimated to be 95%, 90% for moderate CHD, and 80% for severe CHD (Marino et al., 2012). CHD is defined as a structural or functional abnormality of the heart or great vessels that occurs during the fetal period (Setty et al., 2017). CHD can be asymptomatic, where the child has a normal development and lifestyle, or symptomatic, conditioning the child's dayto-day life and causing delayed psychomotor development (Marino et al., 2016). The type of treatment depends on the type and severity of the heart disease, the presence and severity of symptoms, and the potential to determine reversible secondary changes. Thus, it can be a conservative or surgical treatment (Setty et al., 2017). In the last decades, several factors have contributed to significantly reduce the morbidity and mortality rates of children with heart disease. Simultaneously, there was an increase in the life expectancy of this population with significant disease-related physical, psychosocial, and neurodevelopmental comorbidities which have an impact on their quality of life (Marino et al., 2012;Niemitz et al., 2017). Chronic diseases are becoming increasingly common in child health. These conditions affect children's development, as well as their relationship with the environment and those surrounding them. According to Sousa and Cruz (2009), they are characterized by their duration, their long treatments and hospital readmissions, and their impact on children and their parents and on functions and emotions. However, some factors inherent to the pediatric age can limit the assessment of health-related quality of life, such as age range, devel- opmental stage, and diversity of anomalies, therapies (medical or surgical), and outcomes (Marino et al., 2012). Parents' perspective should be taken into account in the case of smaller children due to their age and lack of cognitive and linguistic skills to interpret data collection tools (Rajmil, López, López-Aguilà, & Alonso, 2013). In addition, Marino et al. (2016) share the opinion that it is essential to understand the perceptions of quality of life of children/adolescents and their parents. These views will contribute to developing a care plan directed at/focused on the needs of children and families, prioritizing problems, facilitating communication, detecting potential problems, identifying preferences, and monitoring the patient's evolution and response to treatment. In clinical practice, the integration of different points of view improves quality of life assessment; the perceptions of children/adolescents and their parents should complement each other (Niemitz et al., 2017). At an initial stage of this study, a literature review revealed the opinion of several authors about the key characteristics of a quality of life assessment tool. After a detailed analysis of the instruments used for assessing the quality of life of children and adolescents with heart disease, the PCQLI was selected. The PCQLI includes a questionnaire consisting of items rated on a Likert-type scale with versions for children aged 8 to 12 years, adolescents aged 13 to 18 years, and their parents. Its original version is written in English (American English), so it was translated into Portuguese according to international guidelines (DIS-ABKIDS, 2004).

Research question
Is the Portuguese version of the Pediatric Cardiac Quality of Life Inventory a reliable tool for assessing the perceived quality of life in a sample of children and adolescents with CHD and their parents living in a center region of Portugal?

Methodology
This study consisted in the process of translation and cultural adaptation of the PCQLI to the Portuguese population. The target population was composed of children aged 8 to 12 years (n = 59) and adolescents aged 13 to 18 years (n = 80) with CHD, as well as their parents (59 parents of children and 80 parents of adolescents). These children and adolescents attended the outpatient cardiology consultation of a pediatric hospital in the central region of Portugal, mastered the Portuguese language, and had no psychomotor problems. The following exclusion criteria were applied: children/adolescents with delayed psychomotor development; children/ adolescents and their parents who did not master the Portuguese language; children/ adolescents who had undergone surgery for less than 6 months; children aged less than 8 years. The minimum age for data collection was 8 years old because the original tool is validated for children above 8 years of age who are already capable of answering a written questionnaire. The maximum age was 18 years old because that is the age limit for receiving pediatric care and marks the transition to adulthood. The criterion of children and adolescents with congenital heart disease was established because CHD is one of the most prevalent congenital diseases. Therefore, this study aimed to analyze its impact on these children and adolescents throughout their life-cycle. Moreover, this study aimed to understand the parents' perceptions about their children's quality of life. The sample was composed of all individuals who met the inclusion criteria and showed interest in participating in the study, and who, after being explained the research objectives and project, agreed to participate on a voluntary basis by giving their written permission for data collection according to the planned methodology. The nonprobability consecutive sampling method was used. For reasons of accessibility and convenience, the sample was composed of children and adolescents with CHD who attended the cardiology consultation of a pediatric hospital in the central region on Mondays and Thursdays between February and July 2014. The questionnaire for children, adolescents, and parents included questions on the sociodemographic (age, gender, nationality) and academic characteristics (grade and academic performance addressed in the question on school retention) and the PCQLI. This instrument includes a general question and the questionnaire. First, it includes the following general health perception question, which is a distinct question regarding overall health status and general quality of life: "In general, would you say your health is…". It is rated The Disease impact and Psychosocial impact subscale scores are calculated individually based on the following formula:  (1), very good (2), good (3), fair (4), and poor (5).
The scores of this question are not included in the total score of the questionnaire (Marino, Hottinger, Drotar, Wray, & Ittenbach, 2014). The questionnaire is composed of a set of statements included in the PCQLI which are referred to as items. These items are grouped into dimensions, with each of them being composed of items rated on a 5-point Likerttype scale from strongly agree (1) to strongly disagree (5). Each subscale/dimension represents the patients' perceptions of the impact of the heart disease on distinct areas: Disease impact is associated with the physical func-tion and Psychosocial impact is associated with the psychological and social function (Marino et al., 2014). Thus, the PCQLI has four forms: child (aged 8 to 12 years), parent of child, adolescent (aged 13 to 18 years), and parent of adolescent. The child form consists of 23 items: 14 in the Disease impact subscale and nine in the Psychosocial impact subscale. The adolescent form is composed of 17 items in the Disease impact subscale and 12 items in the Psychosocial impact subscale. The PCQLI generates three scores: Disease impact subscale score, Psychosocial impact subscale score, and Total score ( Table 1). The Total score is calculated from the Disease Impact and Psychosocial Impact subscales using the following formula: Disease impact subscale score + Psychosocial impact subscale score As the sum of these two subscale scores, the total score has a maximum of 100 points (the higher the score, the higher the level of perceived quality of life). After evaluating the importance of the PC-QLI items, the associated concepts and their relevance to the target population through the literature review, the instrument was translated and culturally adapted to the Portuguese population according to the guidelines recommended by Vilelas (2009). The following steps were followed: (1) Initial translation -two bilingual translators performed an independent translation of the original version of the PCQLI from American English into Portuguese; (2) Synthesis of both translations -both translations were compared by the principal investigator, who identified some differences between them. These differences were analyzed and discussed by a group of experts, who selected words that were more simple and easier to understand, and the first consensus version of the PCQLI in Portuguese was obtained; (3) Translation -the consensus version in Portuguese was back-translated into American English by another translator. The obtained PCQLI versions and the original instruments were analyzed and compared. Given the simplicity of the instrument and the clarity of the statements, there was a high semantic, cultural, and idiomatic concordance. The few divergent questions were discussed and, after a consensus, the second Portuguese version of the PCQLI was obtained; (4) Revision by the group of experts -the Portuguese version of the PCQLI was revised by two specialist nurses in child health and pediatrics with over 7 years of profession-al experience, as well as by a different translator from those in the previous steps, who compared the items and analyzed their level of agreement with the original English versions. At the end of this meeting, this version (Tables 2 and 3) was considered to be well translated, accurate, easy-to-understand, and written in a clear language.  Subsequently, the questionnaire was applied in a pre-test and informal interviews were performed with 10 children/adolescents and their parents.
After the steps of translation and face validity analysis, the internal consistency of the Portuguese version of the PCQLI was assessed in a clinical setting. Duly completed questionnaires were considered valid for this study. Data were collected with the collaboration of the nursing team of the cardiology consultation. At this consultation, which occurred before the researcher's contact with the families, they were informed that they would be contacted to participate in the study. The questionnaires were applied by the researcher between the nursing consultation and the medical consultation. The respondents were supervised while completing the questionnaires to clarify any doubts and avoid contamination of results if children and parents shared ideas. It should be noted that there was a previous contact with the nurses working in that unit with the purpose of identifying the children who met the inclusion criteria. The questionnaire was delivered in a sealed envelope. Each envelope had another two envelopes: one with the version for children/ adolescents and another one with the version for parents, duly identified on the front of the envelope, with the informed consent form attached to each questionnaire. All questionnaires were coded s that each dyad could be identified later on while safeguarding privacy and anonymity.

Results
The process of translation of the PCQLI into European Portuguese showed satisfactory results. The consensus version preserved the meaning of the original version of the PCQLI. The original version of the PCQLI was translated into Portuguese by two independent translators. Both translations were compared by the principal investigator, revealing some differences in some items. Thus, a consensus version was obtained which was analyzed by a group of experts for clarity of language and equivalence of the meaning of the translated items. At the meeting with the expert group, after a comparison of the first translations, the final Portuguese version was obtained -the consensus version. However, the experts raised some doubts concerning some items. In the item "I miss social activities", they had doubts between "sinto falta de atividades sociais" and "falto a atividades sociais" and, after discussion, they decided to translate it as "sinto falta de atividades sociais". The criterion "I hold back when I am doing physical activities" raised some doubts between the expression "Eu tenho de me conter" and "Retraio-me" and they decided to use "tenho de me conter quando estou a fazer atividades físicas". In "other people tease me", doubts existed between "há pessoas que gozam comigo" and "os outros metem-se comigo", and they decided to use "as outras pessoas gozam comigo"." With regard to "I feel sluggish", they hesitated between "eu sinto-me sem energia" and "sinto-me lento", choosing the first possibility. In "I cannot wear what I want", the experts discussed between the option "não posso vestir o que quero" and "não posso usar o que quero", and decided to add more information to the item: "não posso vestir/usar o que quero (roupas, piercings, tatuagens,…)". Finally, in the item "I feel helpless", they had doubts between "sinto-me impotente" and "eu sinto-me desamparado(a)", but used the first option. The final result of this meeting was the consen-sus version of the scale in Portuguese, which was back-translated into English to check for any differences in meaning and/or content between the original and the translated versions, thus ensuring the quality of the instrument to be used (Vilelas, 2009). Then, the obtained PCQLI version and the original instrument were analyzed and compared to check for equivalence of meaning of the translated items. The differences related to the use of synonyms which did not alter the meaning of the sentences. The final version in Portuguese was analyzed by a group of experts who compared the items and identified their level of agreement (regarding clarity, colloquial language, translation, understandability, and reading style) with the original English versions. The experts found that this version was well translated, accurate, easy to understand, and written in a clear language. Thus, the pre-test version of the PCQLI in Portuguese was obtained.
In the pre-test phase, the researcher established a personal contact with all participants to explain the study objective and analyze face validity. This sample was composed of 10 parents of children and adolescents, as well as five children and five adolescents. Overall, they had no difficulty in completing the questionnaire and all of them considered that the questions were clear, easy-to-understand, and colloquial. Thus, the translation of this instrument was considered appropriate and the final versions of the PCQLI were obtained. A total of 280 questionnaires were gathered (two pairs were excluded because the quality assessment scale was not duly completed in the parent version). Thus, a sample of 59 pairs of questionnaires of the child version and 80 pairs of questionnaires of the adolescent version were obtained, in a total of 278 respondents among children, adolescents, and parents. The analysis of sample distribution by age showed that the mean age was 9.68 years in the group of children and 15.1 years in the group of adolescents. In relation to gender, there was a similar distribution, with a slight predominance of boys in the group of children (33; 55.93%).
As regards their academic characteristics, none of the children had a special education study plan. As for their academic performance, 13.56% of them had already failed at least one year. In this group of children attending 1 st to 7 th grade, most of them attended 5 th grade (22.04%). In the group of adolescents, only three of them (3.75%) had a special education study plan and 32.50% of them had failed at least one year. Most of them attended the 9 th grade (25%) and the 10 th grade (25%). In relation to the degree of kinship of children/ adolescents' legal representative who completed the questionnaire, in the group of children, 77.97% of the questionnaires were filled out by the mothers, 20.34% by the fathers, and 1.69% by a grandfather. In the group of adolescents, 70% of the questionnaires were filled out by the mothers, 26.25% by the fathers, and 3.75% by uncles and grandparents. The assessment of some of the psychometric properties of the Portuguese version of the scale showed a good internal consistency for the total score α = 0.834 and for the Disease impact subscale α = 0.824, as well as a reasonable consistency for the Psychosocial impact subscale α = 0.626 in the group of children (Table 4). In the original version, these two subscales had α = 0.88 and α = 0.78, respectively. Children's parents had higher internal consistency scores: total score α = 0.894; Disease impact subscale α = 0.861, and Psychosocial impact subscale α = 0.795 (in the original version, both subscales had a α = 0.89 and α = 0.83, respectively). In the group of adolescents, the total score showed a very good internal consistency (α = 0.932) and the Disease impact and Psychological impact subscales also had a high internal consistency (α = 0.892 and α = 0.851, respectively); in the original scale, the scores for both subscales were similar (α = 0.89 and α = 0.85, respectively). With regard to adolescents' parents, the results were identical, with α = 0.932 for the total score; α = 0.885 for the Disease impact subscale; and α = 0.882 for the Psychosocial impact subscale. In the original scale, the internal consistency of both subscales was α = 0.91 and α = 0.85, respectively. Similar results were found in the original version in American English.
SANDRA ILDA MORAIS LOPES et al.

Discussion
The terminology used in the PCQLI proved to be simple and the process of translation and adaptation into European Portuguese showed a high degree of agreement of translated items, despite minor differences between both translated versions, which, after the consensus version, did not correspond to major changes in meaning, but rather to the use of synonyms. The involvement of different translators and the panel of experts proved to be very important during this process. This study contributed to obtaining a relevant instrument for health research and clinical practice, which justifies the need for future validation studies. In this study, the authors used the same subscales and parameters as the authors of the original scale. In short, this study revealed high levels of quality of life in children and adolescents with CHD, which is consistent with the predominance of simple CHD in this sample. These results are in line with those of Ernst et al. (2018) and Niemitz et al. (2017) who found that patients with severe heart disease and greater need for medical care have worse quality of life. The 5-month data collection period proved insufficient to develop the validation process, as mentioned above. This was a limitation of this study, as well as the lack of analysis of concurrent validity and test-retest reliability, which should be analyzed in future studies. Considering the specificity of the sample, as well as the nonprobability sampling method, there is no guarantee of the representativeness of the population, thus the analysis, interpretation, and discussion of the results should be done with caution. For these reasons, future studies should develop strategies for controlling these factors, namely the use of probability sampling methods and larger samples. It should be noted that the limited number of studies on this topic in Portugal made it difficult to interpret the perceived quality of life of children and adolescents with CHD in Portugal.

Conclusion
This study resulted in the first Portuguese version of the PCQLI, which was translated and adapted to assess the quality of life of children with CHD. We hope that this can be the first step in Portuguese research in the area of CHD and a driving force for future studies. The results are interesting and promising regarding the use of the Portuguese version of the PCQLI in clinical practice. This questionnaire covers the main dimensions of quality of life and it is a short, easy to complete tool which uses a clear and simple language. The results found for this version confirm the potential of this instrument. Considering these properties and the size of the sample, the authors decided to use the same subscales and parameters of the original scale's authors.
Further studies should be conducted in Portugal on this topic, including studies aimed at validating the Portuguese version of the PCQLI.