Cross-cultural adaptation and validation evidence of the Hand-Foot Syndrome 14 (HFS-14) instrument

Abstract

Objective  To cross-culturally adapt the Brazilian version of the Hand-Foot Syndrome 14 (HFS-14) instrument to Brazilian Portuguese and obtain evidence of construct validity.

Methods  Methodological, quantitative study divided into two stages. Stage 1: cross-cultural adaptation, which consisted of translating the instrument from English into Portuguese, and back into English, and cross-cultural validation of the instrument in Portuguese, through content validity analysis and assessment by the target population. Stage 2: obtaining evidence of construct validity by applying the questionnaire to a sample of 110 participants undergoing therapy with drugs that potentially cause hand-foot syndrome. Evidence of validity was investigated using exploratory and confirmatory factor analysis techniques, item response theory parameterization, and relationships with the Skindex-16 instrument.

Results  The Brazilian Portuguese version of the HFS-14 showed excellent cross-cultural validity with an overall Content Validity Index of 0.92 and 100% acceptability by the target population. The instrument presented a unifactorial structure with factor loadings ≥ 0.516 for all items in both analyzes and good internal consistency and model fit indices. Correlation coefficients with the Skindex-16 ranging from 0.477 to 0.589 showed an association between the constructs measured by each instrument.

Conclusion  This study provides evidence of validity for the Brazilian version of the HFS-14 as an instrument capable of assisting in the assessment of the quality of life of patients with hand-foot syndrome.

Keywords:
Hand-foot syndrome; Drug-related side effects and adverse reactions; Quality of life; Validation study; Translating; Cross-cultural comparison; Surveys and questionnaires

Resumo

Objetivo  Adaptar transculturalmente para a língua portuguesa brasileira e obter evidências de validade de construto para a versão brasileira do instrumento Hand-Foot Syndrome 14 (HFS-14).

Métodos  Estudo metodológico, quantitativo, dividido em duas etapas. Etapa 1: adaptação transcultural, que consistiu na tradução do inglês para a língua portuguesa e para o inglês novamente, e na validação transcultural do instrumento na língua portuguesa, através da análise da validade de conteúdo e avaliação pela população-alvo. Etapa 2: obtenção de evidências de validade de construto, com a aplicação do questionário em uma amostra de 110 participantes em terapia com fármacos potenciais causadores de síndrome mão-pé. As evidências de validade foram investigadas por meio de técnicas de análise fatorial exploratória e confirmatória, parametrização pela teoria de resposta ao item e relações com o instrumento Skindex-16.

Resultados  A versão traduzida para o português brasileiro do HFS-14 apresentou ótima validade transcultural, com Índice de Validade de Conteúdo geral de 0,92 e 100% de aceitabilidade pela população-alvo. O instrumento apresentou estrutura unifatorial, com cargas fatoriais ≥ 0,516 para todos os itens nas duas análises e bons índices de consistência interna e de ajuste do modelo. Coeficientes de correlação com o Skindex-16 variando entre 0,477 e 0,589 evidenciaram associação entre os construtos medidos por cada instrumento.

Conclusão  O presente estudo traz evidências de validade para a versão brasileira do HFS-14 como uma ferramenta capaz de auxiliar na avaliação da qualidade de vida de pacientes com síndrome mão-pé.

Descritores:
Síndrome mão-pé; Reações adversas relacionados à medicamentos; Qualidade de vida; Estudo de validação; Tradução; Comparação transcultural; Inquéritos e questionários

Resumen

Objetivo  Adaptar transculturalmente al idioma portugués de Brasil y obtener evidencias de validez de constructo de la versión brasileña del instrumento Hand-Foot Syndrome 14 (HFS-14).

Métodos  Estudio metodológico, cuantitativo, dividido en dos etapas. Etapa 1: adaptación transcultural, que consistió en la traducción del inglés al portugués y al inglés nuevamente, y en la validación transcultural del instrumento en el idioma portugués, mediante el análisis de la validez de contenido y la evaluación por el público destinatario. Etapa 2: obtención de evidencias de validez de constructo, con la aplicación del cuestionario en una muestra de 110 participantes en terapia con fármacos potenciales causadores del síndrome mano-pie. Las evidencias de validez fueron investigadas mediante técnicas de análisis factorial exploratorio y confirmatorio, parametrización por la teoría de respuesta al ítem y relaciones con el instrumento Skindex-16.

Resultados  La versión traducida del HFS-14 al portugués brasileño presentó una excelente validez transcultural, con un Índice de Validez de Contenido general de 0,92 y 100 % de aceptabilidad por parte del público destinatario. El instrumento presentó estructura unifactorial, con cargas factoriales ≥ 0,516 en todos los ítems en los dos análisis y buenos índices de consistencia interna y de adaptación del modelo. Los coeficientes de correlación con Skindex-16, que variaron entre 0,477 y 0,589, evidenciaron relación entre los constructos medidos por cada instrumento.

Conclusión  El presente estudio ofrece evidencias de validez de la versión brasileña del HFS-14 como una herramienta que puede servir de ayuda en la evaluación de la calidad de vida de pacientes con el síndrome mano-pie.

Descriptores:
Síndrome mano-pie; Efectos colaterales y reacciones adversas relacionados con medicamentos; Calidad de vida; Estudio de validación; Traducción; Comparación transcultural; Encuestas y cuestionarios

Introduction

New oncological therapies offering more targeted and effective treatments and better survival results for many patients have been introduced recent decades.(¹) However, these new classes of drugs have brought challenges related to the onset of dermatological adverse effects that may negatively impact patients’ quality of life (QoL).(²,³)

Dermatological toxicities can lead to reduced therapeutic adherence, negatively impacting patients’ prognosis and survival.(²,³) Therefore, it is clinically relevant that health professionals are aware of them and know how to manage them. Cutaneous manifestations related to antineoplastic drugs have distinct classifications with emphasis on acral reactions, such as hand-foot syndrome (HFS), or palmar-plantar erythrodysesthesia and its variants, such as PATEO syndrome (periarticular thenar erythema and onycholysis) and hand-foot reaction.(⁴-⁷)

The HFS is a reaction associated with the high sensitivity of skin tissues to the action of some antineoplastic drugs and the class of tyrosine kinase inhibitors, characterized by painful, erythematous and symmetrical lesions on the digital and distal tips of fingers and toes, heels and, more frequently and severely, on soles of the feet, especially in areas of pressure and friction. Symptoms include dysesthesia, paresthesia, edema and, in severe cases, blisters, scaling and ulcers that impair mobility, limit activities of daily living and compromise the QoL of patients.(²,⁷)

In healthcare practice, the assessment of dermatological toxicities by health professionals is not always satisfactory in relation to the condition of the skin and the severity of the lesions. There is also a risk that the assessment does not include the suffering experienced by patients and the impact on their QoL. The effective monitoring with multidimensional instruments leads to early diagnosis and the adoption of interventions capable of alleviating symptoms and reducing their duration, leading to a better QoL.(³,⁸)

The Skindex-16, already translated and validated in Brazil, and the Hand-Foot Syndrome-14 (HFS-14), not translated or validated in Brazil until the present investigation, are among the specific instruments for the assessment of dermatological toxicities.

The HSF-14 was developed in English in 2011 by Vincent Sibaud, a researcher affiliated with the Toulouse Cancer Research Center, France, with the aim of assessing the QoL of patients with HFS induced by chemotherapy or targeted therapies, through 16 items that include specific signs and symptoms of HFS.(⁹,¹⁰) The instrument has also been translated into Japanese and Turkish, providing evidence of validity for monitoring the impact of HFS in different population contexts.(¹¹,¹²) It is noteworthy that both the original study and subsequent validations included patients with similar clinical and therapeutic profiles.

The justification for this study was based on the need for accurate assessment of HFS in people undergoing cancer treatment at a higher risk of dermatological changes. Making the HFS-14 available to the national community of oncology professionals may help them measure the impact of this toxicity on patients’ QoL, favoring early interventions and individualized care.

To cross-culturally adapt it to Brazilian Portuguese and obtain evidence of validity for the Brazilian version of the Hand-Foot Syndrome-14 (HFS-14) instrument were the objectives of the present study.

Methods

This is a methodological, quantitative study divided into two stages: cross-cultural adaptation and search for evidence of construct validity. The organization of data in this article was guided by the Standards for Quality Improvement Reporting Excellence, SQUIRE 2.0 checklist.(¹³-¹⁵)

The investigation of evidence of validity of the HFS-14 was guided by the guidelines proposed by the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN).(¹⁶)

The data collection period was from December 2021 to June 2022 at the chemotherapy outpatient clinic of a large hospital in São Paulo.

The HFS-14 is a self-administered questionnaire with 14 items, each of them is scored on a three-point Likert scale according to which: (0) no, never; (3) once in a while; (6) always. There is also a question to score limb involvement and another that assesses the presence of pain.(¹⁰) The sum of the 16 items generates a raw score (RS) ranging from 2 to 90 points and, as per the original study’s guidelines, this is converted to a adjusted score (AS) ranging from 2 to 100 points using the following equation: AS = (1.1136 × RS) – 0.2273. The reading of the result should be inversely proportional, i.e., higher scores indicate worse QoL related to HFS.

The adaptation process to obtain the final version consisted of the following phases: translation, back translation, expert committee and pre-testing of the instrument.(¹³,¹⁴)

The original HFS-14 instrument was initially translated from English to Portuguese by two independent translators, native Brazilians with command of the English language, and only the first was informed of the study’s objective. After obtaining the two translations with minimal discrepancies, a summary was prepared and submitted to the back translation process, which resulted in Version 2.

The Delphi technique was used in the process of searching for evidence of content validity in relation to semantic, idiomatic, cultural and conceptual equivalences of Version 2, based on the evaluation by a committee of experts formed by seven professionals (four nurses, two oncologists and one dermatologist) with three to 17 years of experience in the field of oncology (average 10.6 years). The model proposed by Fehring, widely used in the literature, in which criteria such as qualifications, specialization, scientific production, knowledge and time of experience with the topic under discussion are considered, was used in the classification of professionals.(¹⁷) An average score of 10, ranging from 5 to 14 points, was obtained emphasizing that the cutoff point was 5, considering at least, practical experience or some degree of academic qualification in oncology. To quantitatively measure the degree of agreement regarding content validity, we chose to calculate the modified kappa, an improved version of the content validity index (CVI) to control the agreement bias by chance based on a 3-point Likert scale, in which kappa values above 0.60 were considered indicative of satisfactory content validity.(¹⁸-²¹)

Next, a pilot study was conducted on a sample of 15 participants using the preliminary version of the HFS-14 with the main intentions of assessing the comprehensibility of the instrument by the target audience, estimating the time required to collect data, as well as any other problems that might prevent its proper use.

This sample consisted of participants undergoing therapy with drugs that could potentially cause HFS and its variants, with or without symptoms, diagnosed with any type of cancer, over 18 years of age, literate, undergoing any line of oncological therapy, regardless of when it had started. Exclusion criteria comprised patients in the final phase of life or unable to respond to the items on the instrument. Participants answered the preliminary version of the HFS-14 and an additional assessment document containing open-ended questions about the patient’s understanding of the instrument with the ultimate goal of making the effective adjustments suggested in order to improve the understanding of the instrument and obtain the final version of the questionnaire (Appendix 1).

During the pilot study phase, participants did not report any difficulties in understanding the instrument, the estimated collection time was 15 minutes and no problems that could impact data collection were observed.

Subsequently, data collection was conducted using the final version of the HSF-14, which was completed by 110 volunteers who met the same inclusion and exclusion criteria and exhibited characteristics similar to those of the pilot study sample.

Evidence of structural validity was assessed by investigating the internal structure using the exploratory factor analysis (EFA) technique, which consisted of factor extraction procedures, analysis of proximity to unidimensionality and model fit indices. This assessment was further enriched by examining the internal consistency, stability and factor replicability. Next, the model proposed by EFA was validated using confirmatory factor analysis (CFA). To perform both analyzes, the sample was randomly organized into two groups of equal size (n = 55). Studies involving simulations and real data have demonstrated that robust instruments produce stable factor solutions with at least 50 participants.(²²-²⁴) Thus, considering the evidence of validity of the HFS-14,(¹⁰-¹²) it was assumed that the two samples presented an appropriate size for the proposed analyzes.

Evidence of validity based on response processes was investigated using item response theory (IRT). For the validation of health instruments, IRT offers significant advantages. Specifically, IRT models enable the characterization of the relationship between individual items and the underlying latent trait (θ) targeted by the questionnaire, along with assessing each item’s discriminative power to differentiate individuals across the continuum of θ.(²⁵,²⁶) Thus, from the factor analyzes, IRT parameters regarding the discriminative power for each question of the HSF-14 were extracted.

To investigate evidence of construct validity through the relationship with external measures, the sample also answered the Skindex-16 questionnaire. As explained, the Skindex-16 also aims to assess the impact of dermatological toxicities on QoL, and has been applied in all previous studies to seek evidence of the validity of the HFS-14.(¹⁰-¹²)

The Skindex-16 has 16 multiple-choice questions divided into three dimensions: physical, emotional, and functional. The possible answers are distributed on a Likert scale from zero to six, and the total score for each dimension is obtained from the sum of the answers. The higher the value found, the worse the QoL.(⁹)

The IBM SPSS Statistics software, version 28.0 (IBM Corp., Armonk, NY, United States) and FACTOR, version 12.06.07, were used. Sociodemographic, clinical, and education-related data on the management of HFS were analyzed descriptively using absolute numbers and percentages. In the EFA, the factorability of the correlation matrix was assessed using the Bartlett test and the Kaiser-Meyer-Olkin (KMO) sampling adequacy measure. Factor extraction was performed using the Diagonally Weighted Least Squares (DWLS) technique, considered ideal for ordinal data with different types of distribution and retention of factors using the parallel analysis technique with random permutation of the observed data.(²⁵-²⁷) The approximation to unidimensionality was assessed by calculating the Unidimensional Congruence (UniCo), Explained Common Variance (ECV) and Item Residual Absolute Loadings (IREAL) indexes, and as data suggested a unifactorial structure for the HFS-14, factor rotation techniques were not applied. The criterion for maintaining the items was a factor loading value ≥ 0.5 for both EFA and CFA.(²⁶) Internal consistency was assessed by calculating the ω coefficient and composite reliability, while the stability and replicability of the factors was assessed by calculating the Latent and Observed variables.(²⁵,²⁶,²⁸) In EFA and CFA, the indices to assess the quality of the model fit included the χ² statistic and its division by the degrees of freedom, Root Mean Square Error of Approximation (RMSEA), Comparative Fit Index (CFI), and Tucker-Lewis Index (TLI). The selection and interpretation of these parameters were determined in accordance with recent literature on the subject.(²⁵,²⁶) Item discrimination (a) and difficulty (b) parameters derived from the IRT were obtained by adjusting a graded response model.(²⁹) When 95% confidence intervals (95%CI) were calculated, the bias-corrected and accelerated bootstrap resampling technique with 500 samples was utilized. Correlations between the total HFS-14 score and the total and factor scores of the Skindex-16 were measured using Pearson’s correlation tests with calculation of the p-value and 95%CI performed using 2,000 bootstrap resamplings. A significance level of 5% (p ≤ 0.05) was considered in all analyzes.

The HFS validation process respected ethical aspects and had the consent of the author and copyright holder prior to the start of the study and after its completion. The project was approved by the Research Ethics Committees (CEP) of the hospital and the Universidade Federal de São Paulo (Unifesp), opinion number 4.260.317 (Certificate of Presentation of Ethical Appreciation: 34390820.1.0000.5505), and data collection started after obtaining the informed consent form by all those involved.

Results

Evidence of cross-cultural validity

After the first round with the experts, kappa values ≥ 0.77 were found for all items, in addition to a kappa value = 0.91 for the general questionnaire, demonstrating excellent agreement by the committee regarding the content validity of the instrument. The second round of evaluation consisted of only one question regarding question 8. In the consensual translation, the item was described as follows: “Tenho dificuldade para calçar os sapatos por causa da minha síndrome mão-pé” (I have difficulty putting on shoes because of my hand-foot syndrome), but one of the judges suggested “Tenho dificuldade para colocar calçados fechados por causa da minha síndrome mão-pé” (I have difficulty putting on closed shoes because of my hand-foot syndrome). The suggestion to change question 8 was unanimously accepted by the experts, consolidating the final version for the pre-test, applied to 15 patients.

Instrument evaluation by the target population

Regarding understanding and ease of answering the HFS-14 instrument, 100% of participants reported that questions were sufficiently clear for understanding the items and their respective answers. At the end of the stage, the Brazilian version of the HFS-14 instrument was defined for the evaluation of psychometric qualities. In relation to the sample that answered the questionnaire for evaluation of psychometric properties (n = 110), their mean age was 57.5 ± 14.9 years (median 59), with a minimum age of 25 and a maximum of 86 years, 78 were male (70.9%), 83 (75.5%) had complete higher education and 22 (20.0%) had complete high school or incomplete higher education. Regarding comorbidities, 57 (51.8%) did not have any associated disease and 53 (48.2%) reported some comorbidity, among which the most frequent were hypertension (23%) and diabetes (7.3%). Regarding smoking habits, 30% were former smokers and 66.4% had never smoked.

Evidence of structural validity

Excellent factorability indices were observed (Bartlett’s test: χ² (91) = 558.4, p < 0.001, KMO = 0.839), supporting the implementation of the EFA. The parallel analysis recommended the retention of only one factor, which was able to explain 57.33% of data variance, and this result was corroborated by the indices of approximation to unidimensionality (UniCo = 0.990, ECV = 0.914, IREAL = 0.197). Thus, the results obtained demonstrated that the HFS-14 can be considered a unidimensional instrument, that is, formed by only one factor. In the EFA, all items had factor loadings ≥ 0.516 with a lower limit of the 95% CI higher than or very close to 0.3, indicating a satisfactory relationship between the questions and the construct (Table 1). In CFA, factor loadings ≥ 0.680 were observed and statistically similar to the values obtained in EFA. The adjustment indices demonstrated satisfactory appropriateness of the model proposed by EFA to data (χ² (77) = 5.316, p > 0.999; RMSEA = 0.000, CFI = 0.999, TLI = 1.153). The adequacy of this solution was confirmed by the CFA indices (χ² (77) = 84.069, p = 0.272; RMSEA = 0.041, CFI = 0.996, TLI = 0.996). The reliability and replicability analysis obtained the following values: Composite reliability = 0.939, ω = 0.946, Latent variable = 0.954, Observed variable = 0.950. These results demonstrated satisfactory reliability and replicability rates of the factor structure.

Table 1
Factor structure of the Brazilian version of the HFS-14

HFS-14	EFA		CFA
	Factor loading	95%CI	Factor loading	95%CI
1	0.579	0.304–0.952	0.788	0.395–0.990
2	0.814	0.394–1.012	0.776	0.451–0.967
3	0.722	0.431–1.091	0.883	0.456–1.033
4	0.639	0.315–0.988	0.853	0.440–0.992
5	0.730	0.359–1.085	0.769	0.396–0.960
6	0.727	0.381–0.947	0.816	0.398–0.969
7	0.855	0.436–1.055	0.788	0.385–0.949
8	0.658	0.380–0.968	0.876	0.446–0.995
9	0.898	0.449–1.070	0.876	0.452–1.009
10	0.516	0.261–0.887	0.746	0.386–0.961
11	0.613	0.322–0.943	0.805	0.420–1.002
12	0.596	0.291–0.917	0.775	0.455–0.963
13	0.844	0.434–1.024	0.778	0.402–0.975
14	0.854	0.437–1.044	0.680	0.334–1.003

• EFA: Exploratory Factor Analysis; CFA: Confirmatory Factor Analysis; 95%CI: 95% confidence interval

Evidence of validity based on response processes

Regarding the IRT parameters, questions 3 (“Tenho dificuldade para realizar as atividades do dia a dia por causa da minha síndrome mão-pé” / “I have difficulty performing daily activities because of my hand-foot syndrome”) and 9 (“Permanecer em pé é difícil para mim por causa da síndrome mão-pé” / “Standing up is difficult for me because of my hand-foot syndrome”) were among those with the greatest discriminative power in the models proposed by CFA and EFA, respectively. The items with the least discriminative power included questions 14 (“Meu relacionamento com outras pessoas está prejudicado pelos sintomas da síndrome mão-pé” / “My relationships with other people are impaired by the symptoms of my hand-foot syndrome”) and 10 (“Tenho dificuldade para andar, mesmo curtas distâncias, por causa da minha síndrome mão-pé” / “I have difficulty walking, even short distances, because of my hand-foot syndrome”). The IRT parameter b (difficulty) demonstrated satisfactory behavior for all items in both solutions, with all second thresholds presenting higher θ values compared to the first thresholds, determining that individuals with a higher level of the latent trait (impaired QoL due to HFS) were more likely to present responses indicating greater impairment of QoL compared to individuals with a lower level of the latent trait (Table 2).

Table 2
Item response theory parameterization of the Brazilian version of the HFS-14

HFS-14	IRT parameters
	EFA			CFA
	a	b1	b2	A	b1	b2
1	0.711	-2.766	-0.860	1.279	-2.034	-0.838
2	1.401	-1.789	-0.489	1.232	-1.360	-0.512
3	1.045	-2.218	0.095	1.880	-1.395	-0.077
4	0.831	-3.274	-1.317	1.632	-1.566	-0.585
5	1.069	-2.194	-1.244	1.205	-2.082	-0.786
6	1.059	-22.469	-0.758	1.414	-1.783	-0.368
7	1.646	-1.562	-0.524	1.278	-1.564	-0.382
8	0.873	-1.872	-0.035	1.819	-1.405	-0.398
9	2.043	-1.621	0.127	1.813	-1.663	-0.454
10	0.602	-2.388	-0.133	1.120	-1.651	-0.403
11	0.776	-2.614	-0.730	1.358	-1.658	-0.373
12	0.741	-3.013	-1.109	1.227	-2.315	-0.780
13	1.575	-1.725	-0.997	1.238	-1.872	-0.777
14	1.643	-2.101	-0.985	0.928	-2.638	-1.675

• EFA: Exploratory Factor Analysis; CFA: Confirmatory Factor Analysis; IRT: Item Response Theory; a: discriminative parameter; b1 and b2: difficulty parameter, thresholds 1 and 2

Evidence of validity through relationships with external measures

Given the unidimensional structure of the HFS-14 evidenced by EFA, only the calculation of the total score was performed. The mean scores obtained for the Skindex-16 domains ranged from 17 to 27 points (Table 3).

Table 3
Characterization of the study sample in relation to the scores on the Skindex-16 and Hand-Foot Syndrome-14 questionnaires

Variable	n	Mean	SD	Median	Min.	Max.
Skindex-16 – Symptoms	110	27.39	27.23	20.83	0.00	100.00
Skindex-16 – Emotions	110	25.84	29.36	14.29	0.00	100.00
Skindex-16 – Functioning	110	17.12	24.68	3.33	0.00	100.00
Skindex-16 – Total	110	23.50	25.08	13.54	0.00	96.88
HFS-14 – Total	110	22.96	22.10	15.92	2.00	100.00

• SD: standard deviation; Min: minimum; Max: maximum; HFS-14: Hand-Foot Syndrome-14

Moderate to strong and statistically significant positive correlations were obtained between the two instruments, suggesting convergence between the constructs evaluated (Table 4).

Table 4
Correlation analysis between scores of the Skindex-16 and HFS-14 instruments

Skindex-16
		Symptoms	Emotions	Functioning	Total
HFS-14 – Total	Coef. [95%CI]	0.477 [0.271–0.654]	0.478 [0.286–0.657]	0.589 [0.398–0.746]	0.555 [0.371–0.721]
	p	< 0.001*	< 0.001*	< 0.001*	< 0.001*

• Pearson correlation test; HFS-14: Hand-Foot Syndrome-14; Coef: correlation coefficient; 95%CI: 95% confidence interval; *: statistically significant value at the 5% level (p ≤ 0.05)

Discussion

The HFS-14 presented good evidence of validity regarding its structure. In the EFA, the high factor loadings, reliability measures and parameterization according to the IRT demonstrated the relevance of all items of the instrument for assessing the impact of the HFS on the QoL of people undergoing cancer treatment.(²⁵-²⁹) These results and the unidimensional structure of the questionnaire were corroborated by CFA, which presented excellent fit indices for the model proposed by EFA. Unlike what was proposed in the original study and its cross-cultural adaptation for the Turkish population, and in agreement with the results of the cross-cultural adaptation study for Japan, the Brazilian version of the HFS-14 presented strong evidence of having a single-factor structure.(¹⁰-¹²)

Some justifications can be raised for these discrepancies and similarities with previous studies. In the original study, the HFS-14 factors were proposed based solely on the theoretical perspective and without being tested by factor analysis, whereas in the study by Okten and Ozer,(¹²) factor extraction and rotation techniques considered inadequate in the current literature were used, such as the “eigenvalue > 1” criterion and orthogonal rotation.(²⁵,²⁶) In the latter study, this point was reinforced by the goodness-of-fit indices presented by the authors, which were lower and would not be considered satisfactory according to criteria used in the present investigation. On the other hand, in the Japanese cross-cultural adaptation of the instrument,(¹¹) although the “eigenvalue > 1” criterion was used, an oblique rotation was chosen, which is more in line with best practices in factor analysis. Thus, these results demonstrate the importance of the structural validation stage through the use of appropriate analysis techniques during the validation process of an instrument.(²⁵,²⁶)

The translation process of the HFS-14 into Portuguese was harmonious and minimal adjustments were necessary to ensure the accuracy of the synthesized version. The cross-cultural validation was successful, each item showed satisfactory kappa values, and the overall kappa was classified as excellent in the literature.(¹⁶-²¹)

These results demonstrated that the Brazilian version of the instrument adequately covered the factors associated with the impact of HFS on the QoL of patients undergoing cancer treatment, in addition to being adapted to the linguistic and sociocultural characteristics of this group.

Most study participants showed little impairment in carrying out their activities, which leads to the assessment that the HFS condition was restricted to mild to moderate degrees, a fact confirmed by the responses to the Skindex-16.

The relationship between the Skindex-16 and the HFS-14 demonstrated coherence between the constructs assessed by the two instruments, that is, the QoL in Brazilian patients with HFS and its variants. In this sense, the Skindex-16 is a widely used questionnaire and recognized by several national studies as a valid measure, which reinforces the evidence of construct validity of the HFS-14. A statistically significant and positive correlation was observed between the Brazilian HFS-14 scores and the Skindex-16 scores. The correlation coefficients were satisfactory, indicating that the HFS-14 provides a valid score.

However, these coefficients had a moderate magnitude, indicating a more modest correlation strength compared to the results reported in other studies.(¹⁰-¹²) In fact, after converting to the coefficient of determination (R²), the shared variance between Skindex-16 and HFS-14 ranged from 23 to 31%. These results may be related to the fact that each item of the HFS-14 was developed to be specific to HFS and suggest that the HFS-14 may be a better scale for identifying differences in QoL impairment compared to other specific dermatological QoL scales. Furthermore, the HFS-14 may be very useful for monitoring patients with HFS and its variants.(¹⁰) In addition, although the estimated correlations for the present study were lower compared to data in the literature, this comparison is limited by the lack of information regarding the 95%CI for the other studies. However, considering the upper limits presented in Table 4, it can be inferred that the correlations reported here were similar to those of other HFS-14 validation studies.

Recognizing and appropriately grading the signs and symptoms of HFS and its variants, as well as assessing QoL can help determine the most effective management strategies in these patients. If appropriate management is not implemented early, this toxicity can worsen, and the impairment of QoL can lead to changes in medication dosage or even discontinuation of treatment, impacting the prognosis of these patients.(⁸,⁹)

Another relevant issue to be considered is the importance of patients’ self-reporting during their journey, since they often present symptoms originating from the disease and/or treatment-related adverse events. Despite objective toxicities, subjective experiences, such as symptoms, are much better reported by patients themselves.(³,⁹)

Due to the low HFS-14 scores of most respondents, it was not possible to evaluate the results in a population with greater clinical impairment of HFS. Although the sample size was favorable for the psychometric tests performed, it proved to be limiting for evaluations of correlations between the results of the use of the HFS-14 instrument and the independent variables, as well as for possible analyzes of invariance of the factorial structure obtained. Still regarding the sample size, although evidence suggests that factorial solutions with excellent fit indices (such as those of the present study) are stable even for small samples, the small sample size attributes a significant degree of uncertainty to the estimates of the models, which is illustrated, for example, by the considerable width of the confidence intervals for the factor loadings in the solutions of the factor analyzes. In this sense, a sample calculation analysis with the recently proposed SENECA technique(³⁰) suggested n = 300 for confirmation of the factorial structure, and this is a recommended objective in future studies.

Another important limitation of the present study was the failure to perform one of the reliability tests, such as the test-retest method, although the internal consistency through composite reliability and the replicability of the factor structure reached appropriate values, which points in favor of the stability of the instrument. Furthermore, it is suggested that future longitudinal studies evaluate the responsiveness of the instrument, that is, its ability to reflect changes in the QoL of patients.

Conclusion

The version adapted for the Brazilian context of the HFS-14 proved to be easy to apply and understand in a sample of people undergoing cancer treatment with HFS, and presented equivalence to the original version in terms of content. The evidence of construct validity indicated that the questionnaire appropriately measured the impact of the HFS on QoL among participants.

Data availability

The authors did not make the data from this article available in repositories prior to submission.

Acknowledgements

We thank Henrique S. Souza for his assistance in deriving the formula for converting the HFS-14 raw score into a weighted score.

Appendix 1

. Final version of the Hand and Foot Syndrome (HFS-14) instrument

Síndrome Mão-Pé (SMP)-14
Por favor, responda as seguintes afirmações da forma mais verdadeira possível. Não há resposta certa ou errada, apenas experiências ou situações que você vive diariamente. Especifique a área afetada pela sua síndrome mão-pé: ( ) Mãos ( ) Pés ( ) Mãos e pés Você diria que a sua síndrome mão-pé tende a ser: ( ) Muito dolorosa ( ) Moderadamente dolorosa ( ) Sem dor
1. Acho difícil girar a chave na fechadura por causa da minha síndrome mão-pé.
( ) Sim, sempre ( ) Sim, às vezes ( ) Não, nunca
2. Acho difícil preparar refeições por causa da minha síndrome mão-pé.
( ) Sim, sempre ( ) Sim, às vezes ( ) Não, nunca
3. Tenho dificuldade para realizar as atividades do dia-a-dia por causa da minha síndrome mão-pé.
( ) Sim, sempre ( ) Sim, às vezes ( ) Não, nunca
4. Tenho dificuldades para me lavar, maquiar ou barbear por causa da minha síndrome mão-pé.
( ) Sim, sempre ( ) Sim, às vezes ( ) Não, nunca
5. Acho difícil dirigir qualquer veículo por causa da minha síndrome mão-pé.
( ) Sim, sempre ( ) Sim, às vezes ( ) Não, nunca ( ) Não tem importância para mim
6. Acho difícil vestir meias ou meia-calça por causa da minha síndrome mão-pé.
( ) Sim, sempre ( ) Sim, às vezes ( ) Não, nunca
7. Demoro mais do que o normal para me vestir por causa da minha síndrome mão-pé.
( ) Sim, sempre ( ) Sim, às vezes ( ) Não, nunca
8. Tenho dificuldade para colocar calçados fechados por causa da minha síndrome mão-pé.
( ) Sim, sempre ( ) Sim, às vezes ( ) Não, nunca
9. Permanecer em pé é difícil para mim por causa da síndrome mão-pé.
( ) Sim, sempre ( ) Sim, às vezes ( ) Não, nunca
10. Tenho dificuldade para andar, mesmo curtas distâncias, por causa da minha síndrome mão-pé.
( ) Sim, sempre ( ) Sim, às vezes ( ) Não, nunca
11. Geralmente fico sentado ou deitado por causa da minha síndrome mão-pé.
( ) Sim, sempre ( ) Sim, às vezes ( ) Não, nunca
12. Tenho dificuldade para pegar no sono por causa da minha síndrome mão-pé.
( ) Sim, sempre ( ) Sim, às vezes ( ) Não, nunca
13. Meu trabalho está sendo prejudicado por causa da minha síndrome mão-pé.
( ) Sim, sempre ( ) Sim, às vezes ( ) Não, nunca ( ) Não tem importância para mim
14. Meu relacionamento com outras pessoas está prejudicado pelos sintomas da síndrome mão-pé.
( ) Sim, sempre ( ) Sim, às vezes ( ) Não, nunca
Assinatura do paciente: ___________________________ Data:* _________________ Hora:* _________________ (*Campo obrigatório)
Assinatura: ___________________________ Cargo:* ___________________________ Data:* _________________ Hora:* _________________ ( ) Consulta inicial ( ) Retorno (*Campo obrigatório)

References

• ¹ Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021;71(3):209-49.
• ² Cury-Martins J, Eris AP, Abdalla CM, Silva GB, Moura VP, Sanches JA. Management of dermatologic adverse events from cancer therapies: recommendations of an expert panel. An Bras Dermatol. 2020;95(2):221-37.
• ³ Williams L, Ginex P, Ebanks Jr. G, Ganstwig K, Ciccolini K, Kwong B, et al. ONS Guidelines ™ for Cancer Treatment-Related Skin Toxicity. Oncol Nurs Forum. 2020;47(5):539-56.
• ⁴ Ding F, Liu B, Wang Y. Risk of hand-foot skin reaction associated with vascular endothelial growth factor-tyrosine kinase inhibitors: A meta-analysis of 57 randomized controlled trials involving 24,956 patients. J Am Acad Dermatol. 2020;83(3):788-796.
• ⁵ Deutsch A, Leboeuf NR, Lacouture ME, McLellan BN. Dermatologic adverse events of systemic anticancer therapies: cytotoxic chemotherapy, targeted therapy, and immunotherapy. Am Soc Clin Oncol Educ Book. 2020;40:485-500.
• ⁶ Ceglio WQ, Rebeis MM, Santana MF, Miyashiro D, Cury-Martins J, Sanches JA. Cutaneous adverse events to systemic antineoplastic therapies: a retrospective study in a public oncologic hospital. An Bras Dermatol. 2022;97(1):14-21.
• ⁷ Canille RM, Pinto MH, Galisteu KJ, Czorny RC, Bertolazzi LG, Faria TV. Adverse dermatoneurological events and impacts on daily activities of patients with gastrointestinal neoplasms undergoing chemotherapy. Rev Bras Enferm. 2023;76(1):e20220161.
• ⁸ Almeida V, Pires D, Silva M, Teixeira M, Teixeira RJ, Louro A, et al. Dermatological side effects of cancer treatment: psychosocial implications-a systematic review of the literature. Healthcare (Basel). 2023;11(19):2621.
• ⁹ Cárcano CB, de Oliveira CZ, Paiva BS, Paiva CE. The Brazilian version of Skindex-16 is a valid and reliable instrument to assess the health-related quality of life of patients with skin diseases. PLoS One. 2018;13(3):e0194492
• ¹⁰ Sibaud V, Dalenc F, Chevreau C, Roché H, Delord J, Mourey L, et al. HFS-14, a specific quality of life scale developed for patients suffering from hand-foot syndrome. Oncologist. 2011;16(10):1469-78.
• ¹¹ Mikoshiba N, Yamamoto-Mitani N, Sato K, Asaoka Y, Ohki T, Ohata M, Miyashita M. Validation of the Japanese version of HFS-14, a disease-specific quality of life scale for patients suffering from hand-foot syndrome. Support Care Cancer. 2015;23(9):2739-45.
• ¹² Okten C, Ozer Z. Validity and reliability study of the Turkish Version of the hand-foot syndrome quality of life scale. J Educ Res Nurs. 2023;20(1):7-13.
• ¹³ Guillemin F, Bombardier C, Beaton D. Cross-cultural adaptation of health-related quality of life measures: literature review and proposed guidelines. J Clin Epidemiol. 1993;16(12):1417-32.
• ¹⁴ Beaton DE, Bombardier C, Guillemin F, Ferraz MB. Guidelines for the process of cross-cultural adaptation of self-report measures. Spine (Phila Pa 1976). 2000;25(24):3186-91.
• ¹⁵ Ogrinc G, Davies L, Goodman D, Batalden P, Davidoff F, Stevens D. SQUIRE 2.0 (Standards for QUality Improvement Reporting Excellence): revised publication guidelines from a detailed consensus process. BMJ Qual Saf. 2016;25(12):986-92.
• ¹⁶ Mokkink LB, Terwee CB, Patrick DL, Alonso J, Stratford PW, Knol DL, et al. The COSMIN study reached international consensus on taxonomy, terminology, and definitions of measurement properties for health-related patient-reported outcomes. J Clin Epidemiol. 2010;63(7):737-45.
• ¹⁷ Fehring RJ. Methods to validate nursing diagnoses. Heart Lung. 1987;16(6 Pt 1):625-9.
• ¹⁸ Echevarria-Guanilo ME, Gonçalves N, Romaniski PJ. Psychometric properties of measurement instruments: conceptual bases and evaluation methods - part I. Texto Contexto Enferm. 2017;26(4):e1600017.
• ¹⁹ Taylor E. We agree, don't we? The Delphi method for health environments research. HERD. 2020;13(1):11-23.
• ²⁰ Niederberger M, Spranger J. Delphi technique in health sciences: a map. Front Public Health. 2020;8:457.
• ²¹ Polit DF, Beck CT, Owen SV. Is the CVI an acceptable indicator of content validity? Appraisal and recommendations. Res Nurs Health 2007;30:459-67.
• ²² Barrett PT, Kline P. The Observation to Variable Ratio in Factor Analysis. Personal Study Group Behav. 1981;1(1):23-33.
• ²³ Kyriazos TA. Applied psychometrics: sample size and sample power considerations in Factor Analysis (EFA, CFA) and SEM in general. Psychol. 2018;9(08):2207-30.
• ²⁴ Winter JC, Dodou D, Wieringa PA. Exploratory factor analysis with small sample sizes. Multivar Behav Res. 2009;44(2):147-81.
• ²⁵ Ferrando PJ, Lorenzo-Seva U, Hernández-Dorado A, Muñiz J. [Decalogue for the factor analysis of test items]. Psicothema. 2022; 34(1):7-17. Spanish.
• ²⁶ Rogers P. Best Practices for your exploratory factor analysis: a factor tutorial. Rev Adm Contemp. 2022;26(6):e-210085.
• ²⁷ DiStefano C, Morgan GB. A comparison of diagonal weighted least squares robust estimation techniques for ordinal data. Struct Equ Modeling. 2014;21(3):425-38.
• ²⁸ Hayes AF, Coutts JJ. Use omega rather than cronbach's alpha for estimating reliability. But…. Commun Methods Meas. 2020;(14):1-24. doi:
• ²⁹ Sousa LA, Braga AE. [Test theory and item response theory in educational assessment]. Rev Impa. 2020;1(1):e020002. Portuguese.
• ³⁰ Lorenzo-Seva U, Ferrando PJ. Determining sample size requirements in EFA solutions: a simple empirical proposal. Multivariate Behav Res. 2024;59:899-912.

Edited by

• Associate Editor:
  Alexandre Pazetto Balsanelli (https://orcid.org/0000-0003-3757-1061) Escola Paulista de Enfermagem, Universidade Federal de São Paulo, São Paulo, SP, Brazil.

Publication Dates

• Publication in this collection
  09 Jan 2026
• Date of issue
  2025

History

• Received
  10 Oct 2024
• Accepted
  6 May 2025