Validation of the EORTC-QLQ-HN35 Questionnaire in Turkish Head and Neck Cancer Patients

Introduction

Methods

Study Population
Eighty HNC patients who were referred to the Radiation Oncology Department of Hacettepe University Medical School for either definitive or adjuvant RT with/without concurrent CT were invited to participate in this study and recruited following their provision of written informed consents. The QLQ-C30 and QLQ-HN35 scales were completed by patients at three-time points as the beginning, middle, and end of the treatment period. After each patient completed answering the questions in both modules, they were also asked about whether any questions in the QLQHN35 module were confusing, upsetting, or difficult to understand. The data collection continued between January 2014 and September 2018. The study protocol was approved by the Hacettepe University Ethics Committee for Non-Invasive Clinical Research.

EORTC QLQ-C30 and HN35 Scales
EORTC QLQ-C30 measures the general QoL and can be used alone or together with other questionnaires developed for specific anatomic locations. EORTC QLQ-C30 provides a general health status score, also provides scores for symptom and functional domains. On the other hand, the QLQ-HN35 is specific to HNC and provides scores for various symptom domains. The QLQ-HN35 includes 35 questions: 11 single item subscales relating to teeth, opening the mouth, dry mouth, sticky saliva, coughing, feeling ill, pain killers, nutritional supplements, feeding tube, weight gain and weight loss. QLQ-HN35 also includes 24 items grouped into seven subscales as follows: pain (4 items), swallowing (4 items), senses problems (2 items), speech problems (3 items), trouble with social eating (4 items), trouble with social contact (5 items), and less sexuality (2 items). The response format was a four-point Likert scale in both QLQ-C30 and QLQ-HN35. Responses to the questionnaires were transformed into a 0-100 scale using EORTC guidelines.[] The decrease in scores of general health status and functional scales in EORTC QLQ-C30 imply deterioration of these scales, whereas the increase in scores of symptom scales in both questionnaires implies deterioration of symptoms.

Statistical Analysis
Descriptive statistics were presented using either mean±standard deviation or median [interquartile range] for numerical variables, and frequencies and percent for categorical variables. The comparisons of numerical data between independent groups were performed using the Mann-Whitney U test for two groups, and the Kruskal-Wallis test for more than two groups. The comparisons of numerical data between dependent groups were performed with the Friedman test for more than two groups. Internal consistency was assessed by Cronbach alpha, and test-retest reliability was assessed by intraclass correlation coefficients (ICCs). Construct validity was evaluated by correlation matrices between subdomains of the scale. Discriminant validity was analyzed by comparison of QoL scores between the most common three diagnostic groups. For clinical validity, an absolute change of 10 points on a 0-100-point score was suggested to be clinically important for the QoL assessments (1-23). Thus, a difference between the preand post-treatment domain scores of QLQ-HN35 was calculated and compared with a reference value of 10 using a one-sample T-test. All statistical analyses were performed with SPSS 25® (IBM Corp., Armonk, NY, USA) software, with a two-tailed design and a type-I error level of 5%.

Results

Table 1: General demographic and clinical characteristics of the patients

The attrition rates were 1.3% (n=1) and 18.8% (n=15) at treatment onset and end-of-treatment assessments, respectively. The primary reason for high attrition at the end of the treatment was a refusal to complete the questionnaire due to the completion of the RT.

The analyses for treatment-related changes in clinical parameters revealed that the grade of mucositis (p<0.001) and pain score (p<0.001) significantly increased, and mean body weight significantly decreased (p<0.001) (Table 2, Fig. 1). For the health-related QoL assessments, dyspnea and diarrhea scores in the QLQ-C30 scale, and the teeth, pain killer, nutritional supplements, and feeding tube scores in the QLQ-HN35 scale did not change significantly during the treatment period. The remaining symptom scores in both scales generally increased during the treatment period. On the other hand, functional scales of QLQ-C30 decreased during the treatment period (Table 3).

Table 2: Changes in the treatment-related clinical parameters during the treatment

Fig 1: Changes in the treatment-related clinical parameters during the treatment.

Table 3: Changes in the QoL assess nebts during the study period

Overall internal consistency of QLQ-HN35 was excellent at each assessment period (treatment onset ?=0.926, mid-treatment ?=0.937, end of treatment ?=0.944). Overall test-retest reliability was satisfactory and ICCs ranged between 0.77 and 0.84. Correlations between corresponding domains of QLQ-C30 and QLQ-HN35 showed satisfactory convergent validity (r=0.61 to r=0.73). Comparisons of the QLQ-HN35 scores between the most common three diagnostic groups concerning divergent validity showed that the differences among the symptom scores were compatible between the diagnosis and clinical behavior. The clinical validity was assessed by comparing the absolute changes between treatment onset and end-of-treatment with a reference value of 10, and all domains showed statistically significant clinical changes (Table 4).

Table 4: Clinically significant changes in the QLQ-HN35 scale scores during the treatment

Discussion

The EORTC QLQ-HN35 was developed by Bjordal et al. to measure the QoL in patients with HNC and the pre-testing was performed in patients from Norway, Sweden, Denmark, United Kingdom and Belgium.[] In 1999, a preliminary reliability and validity study was performed in 500 patients from Norway, Sweden and the Netherlands during and after treatment which resulted in proposal of additional questions based on the feedback of the patients.[] They reported the Cronbach's alpha coefficient ≥0.78 in all samples and the compliance rate as 83%. In 2000, a cross-validation study approved the reliability of the QLQ-HN35 module in patients with advanced HNC.[] In the same year, Bjordal et al. published the final clinical validity of scales and single items and results on psychometric properties of the QLQ-H&N35 in patients from 12 countries speaking nine languages.[] This study included 622 patients under treatment for the first time or for the recurrent disease that underwent surgery and/or RT and/or CT, and also disease-free patients 1-3.5 years after treatment. The rate of unanswered questions was <3% in this study, and they reported the Cronbach"s alpha coefficient >0.70 for all scales but the senses scale in patients with pharyngeal cancer, which was found 0.68.

In the following years, validation of the QLQ-HN35 module in several more languages has been performed in patients with HNC of different sites undergoing different treatment modalities or under follow-up. In 2013, Singer at al. reviewed 136 studies in 19 different languages in 27 countries.[] They reported that the scales of sexuality and speech were the scales with the highest percentages of missing values. They found that the median Cronbach"s alpha was between 0.61 and 0.93. The authors also published their results in patients that received targeted or multimodal therapy for HNC and detected some deficiency in the QLQ-HN35 module regarding the toxicity of these therapeutic options, such as rash, nail changes, pulmonary symptoms and impairment of fertility, and proposed an update of the module.[] This study led the way to the development of the QLQ-HN43 module. Our next aim is to validate this updated version in Turkish HNC patients.

Limitations
A higher number of patients in the present study could have resulted in a more accurate statistical analysis. In addition, although the attrition rate was satisfactory at the beginning of the treatment, the attrition rate increased to 18.8% at the end. A lower rate of attrition at the end could have also yielded different statistical results.

Conclusion

Peer-review: Externally peer-reviewed.

Conflict of Interest: The authors declare that they have no conflict of interest.

Ethics Committee Approval: The study protocol was approved by the Hacettepe University Ethics Committee for Non-Invasive Clinical Research.

Financial Support: None declared.

Authorship contributions: Concept - G.Y., M.C., G.Ö.; Design - G.Y., S.Y.S., C.S.B.E.; Supervision - G.Y., M.C., G.Ö.; Funding - None; Materials - G.Y., Ç.S.B.E.; Data collection and/or processing - S.Y.S., C.S.B.E.; Data analysis and/or interpretation - G.Y., Ç.S.B.E., S.Y.S., D.Y.; Literature search - S.Y.S., G.Y.; Writing - S.Y.S., Ç.S.B.E., D.Y., G.Y., M.C., G.Ö.; Critical review - G.Y., M.C., G.Ö.

References