Validation of the St George’s respiratory questionnaire Vietnamese Version in new pulmonary tuberculosis patients

Background: According to the Global Tuberculosis Report 2019, Vietnam is one of the 20 countries with the highest TB prevalence in the world. Pulmonary tuberculosis has a signicant effect on lung functions, causing many obstacles in daily activities and affects the quality of patient’s lives. Methods: The case-series study conducted on 43 newly-diagnosed pulmonary tuberculosis patients at the Department of Tuberculosis - Military Hospital 103 within 4 months. The aims of the current study were to evaluate the validity of the Vietnamese version of the St. George's Respiratory Questionnaire and to investigate the relationship between SGRQ scores and the clinical and subclinical symptoms in new pulmonary tuberculosis patients. Results: The results indicate that Vietnamese version of the SGRQ has high reliability with Cronbach's alpha of Total score was 0.9451, Cronbach's alpha of all domains was above 0.6, of which the Symptom domain was 0.6635, the Impact domain was 0.9069, the Activity domain was 0.9121. The study also showed that SGRQ score was proportional to the aggregate size of all cavities on chest X-ray (r = 0.3772) and inversely proportional to BMI (r = -0.2843), MGIT days to positivity (r = -0.1635). Conclusions: The Vietnamese version of the SGRQ is a highly reliable and valuable questionnaire in assessing symptoms and life effects in new PTB patients. We recommend it as symptom measurement and quality of life evaluation in patients with new PTB in future studies. Trial registration: The and All participants had provided informed


Background
Tuberculosis in general and pulmonary tuberculosis, in particular, remains the leading cause of disease burden worldwide. According to the Global Tuberculosis Report 2019, an estimated 10.0 million people fell ill with TB and approximately 1.3 million TB deaths in 2018 [1]. Almost all PTB patients experience lung function impairment which leads to limited mobility and decreased quality of their life [2,3].
St. George's Respiratory Questionnaire (SGRQ) is a detailed questionnaire developed to measure health status in patients with lung diseases. It is mainly used to evaluate patients with COPD, asthma, and bronchiectasis and has been translated into many languages around the world [4][5][6][7]. Many studies are using SGRQ to evaluate and prognosis patients with latent tuberculosis and post pulmonary tuberculosis worldwide [8][9][10]. In Vietnam, the Vietnamese version of the SGRQ was created to be used as a measure of quality of life in patients with COPD, asthma and bronchiectasis.
However, the use of SGRQ in new pulmonary tuberculosis patients has not been studied much throughout the world. Besides, at present, there is no single scale to assess symptoms and impact of life in new PTB patients. In this paper, the authors offer to evaluate SGRQ value in new PTB patients and examine the relationship between the SGRQ and clinical and subclinical symptoms in new PTB patients. We believe that the ndings presented in our paper will provide signi cant evidence of the validity of the SGRQ as a reliable scale for pulmonary tuberculosis patients. In doing so, we hope that our research contributes to the global ght against tuberculosis.

Methods
The aims of the current study: The purposes of the current study were to determine the validity of the Vietnamese version of the St.
George's Respiratory Questionnaire and to evaluate the association between SGRQ scores and the clinical and subclinical symptoms in new pulmonary tuberculosis patients.

Study design:
This was a case-series study Study participants: All patients diagnosed with new PTB, over 18 years old were recruited into this study. The main inclusion criteria were: abnormalities on CXR compatible with PTB and a positive result on the MGIT culture (sputum or BAL); principal exclusion criteria were previous treatment for active TB disease.

Study location and time:
The current study was conducted at the Department of Tuberculosis -Military Hospital 103 from February 2020 to June 2020.
Sample size and sampling: 43 eligible patients with new pulmonary tuberculosis were included in the study.

Measurements
The dependent variable was the SGRQ score. The SGRQ is a self-assessment scoreboard exclusively for respiratory disease including 3 sub-sections: symptom section (including 8 questions), activity section (including 16 questions) and impact section (including 26 questions). Translations of the SGRQ questionnaire are available from the St George's library. To verify the translated version, a native Vietnamese speaker with a good English pro ciency was selected and transitionally translated from the SGRQ English version into Vietnamese. The translated version has been critically reviewed to con rm equivalence in semantics, idioms, experiences, and concepts.
The independent variables were BMI (kg/m 2 ), MGIT days to positivity (day), Aggregate size of all cavities on chest X-ray (cm). BMI was computed as = weight (in kilogram)/ (height in meter) 2 . BMI (body mass index) under the cut-off point of 18.5 kg/m 2 was considered underweight [11]. The aggregate size of all cavities was calculated by adding up the widest diameters of all ones. To measure the widest diameter of each cavity present on CXR, we used the standard radio-opaque ruler visible on the lm. We used the MGIT BACTEC 960 system to perform liquid cultures. Time to MGIT positivity was calculated by the number of days from sample inoculation to detection of MTB growth.
Data collection Data collection tools: Eligible participants were asked to complete the Vietnamese version of the St George's respiratory questionnaires.
Data collectors: Studying doctors were responsible for data collection.
Data collection procedures: All new PTB patients were invited to the study. A consent form was given to the participants before administering the research. It took about 15 minutes for each participant to complete the questionnaire. Personal information (eg: name, phone number…) was anonymized before the analysis. We also collected age, gender, BMI, CXR results, sputum smear microscopy results, time to MGIT positivity (sputum or BAL) and random blood glucose.
Data quality assurance: The data quality was monitored by a researcher in the studying team.
Data analysis and statistical methods SGRQ score was calculated using an algorithm designed by PW Jones, St George Medical University Hospital, London, UK, available online from http://www.healthstatus.sgul.ac.uk/sgrq-app. The score ranges from 0-100 points, the higher the score, the greater the corresponding respiratory disease.
Continuous variables data were presented as means ± SD [standard deviation] and categorical data were presented as numbers and percentages. We assessed internal consistency reliability using Cronbach's α coe cient. As adapted from Taber, K.S (2018) [12], internal consistency reliability for each scale is considered as excellent if Cronbach's α is ≥ 0.9, strong if Cronbach's α is ≥ 0.8, acceptable if Cronbach's α is ≥ 0.7 and reasonable if Cronbach's α is ≥ 0.6. Correlations between SGRQ score and other factors were determined using the Pearson correlation coe cient. The known-group validity was evaluated based on different diagnoses using the student's t-test. The signi cance level was set at a p-value < 0.05. The analysis was performed using STATA version 14 (College Station, Texas 77845 USA).

Participants' characteristics
The study included 43 new PTB patients from February 2020 to June 2020. The characteristics of the whole sample (N = 43) are presented as follows in Table 1. Data are reported as n (%) or mean ± SD (standard deviation).    Table 3 show all P-value > 0.05. Therefore, there is no signi cant difference between the sexes in the indicators of the SGRQ scale.  As shown in Table 4, SGRQ showed correlation with the BMI, aggregate size of all cavities on CXR, MGIT days to positivity. Accordingly, SGRQ score correlated inversely with BMI and MGIT days to positivity (r < 0).
As a result of Table 4 The relationship between MGIT days to positivity and SGRQ score entries was inversely proportional with r in all domains < 0, meaning that the longer MGIT time to positivity, the lower the SGRQ score, but the sig were > 0.05, so it was not statistically signi cant.

Discussion
This study investigated the validity and reliability of the SGRQ Vietnamese version in 43 patients with new PTB. Also, the current study was to explore the relationship between SGRQ score and BMI and subclinical symptoms. As stated in the WHO report (2019), the TB incidence in men/women was 2/1, in our study on 43 new PTB patients, the rate was 30/13 ≈ 2.3, equivalent to the WHO report. Although the WHO Global Tuberculosis Report 2019 also shows that tuberculosis can be acquired at any age, the highest incidence is in adult men (> 15 years old). Our study data was consistent with the WHO report [1].
BMI less than 18.5 kg/m2 is considered underweight [11].  [14]. These results provided an additional fact that high proportions of TB patients were malnourished.
As reported by Rachel W. Kubiak et al, the mean RBG among TB patients overall was 10.04 ± 5.74 mmol/l and 49% TB patients had a RBG > 7.8 mmol/l, which are higher than our study [14]. This can be explained by the differences in the diets between the two countries.
In our study, sputum AFB smears had a sensitivity of 58.33% as compared with the result of 67.5% from the research of Philip Mathew et al. were analyzed at two university-a liated on 267 sputum samples [15].  [17], which are higher than our study.
In agreement with previous studies on the reliability of SGRQ scale, the study by Adnan et al on PTB patients in Indonesia showed that the Cronbach's Alpha scores of all subscales (symptom, activity and impact) were above 0.7 [17], the results were similar to that of Zeina Akiki's study in patients with COPD and asthma in Lebanon with Cronbach's alpha score of 0.80 [4]. Research by M. Ferrer conducted in Europe on COPD patients with Cronbach's alpha results of symptom domain was > 0.7 and impact, activity domains were > 0. 9 [6]. The Cronbach's α coe cient for Japanese version in Mariko Morishita-Katsu's study on COPD patients was reported as 0.933 [5], comparable to Anees Ur Rehman's study in Malaysia on COPD patients where the Cronbach alpha report for SGQR was 0.87 [7].
However, the data on SGRQ study results in new pulmonary tuberculosis patients worldwide is limited. In our study on new PTB patients, the SGRQ scale in Vietnamese version is a highly reliable scale, with Cronbach's alpha score of the total score of 0.9451, Cronbach's alpha of all other domains were above 0.6, in which the symptom domain was 0.6937, the impact domain was 0.9069, the activity domain was 0.9121. Therefore, SGRQ can be used to assess symptom levels and life effects in patients with pulmonary tuberculosis. Our results demonstrated that there is no gender difference in the SGRQ scale.
However, in the study of Adnan et al. the symptom and impact domain did not differ in gender, except for the activity domain [17]. In our opinion, perhaps due to religious and cultural differences in the two countries, Vietnamese women are not limited to participating in social activities and equally as men.
To our knowledge, our study is the rst to demonstrate an association between SGRQ score and BMI and subclinical symptoms in new PTB patients. We found that SGRQ score was proportional to the aggregate size of all cavities on chest X-ray (r = 0.3772) and inversely proportional to BMI (r = -0.2843), MGIT days to positivity (r = -0.1635). This means that the lower the BMI, the higher the SGRQ score (the more respiratory symptoms and life effects) and the longer MGIT time to positivity, the lower the SGRQ score. However, not all relationships were statistically signi cant. The results of the correlation with sig > 0.05 can be explained by our study on a small sample size of 43 PTB patients. Therefore, it is necessary to conduct a larger sample size research to evaluate the relationship between the SGRQ score with the clinical and subclinical indicators.
There were some limitations in the present study. Firstly, the number of PTB patients in this study was small (n = 43) and this might be a cause of the results that were not statistically signi cant. Secondly, we could not follow up the participants to assess the SGRQ score after TB treatment. Future studies are suggested to conduct in a larger sample and in multiple centers to con rm the ndings.

Conclusion
The application of the SGRQ scale in Vietnamese version on 43 new PTB patients at the Department of Tuberculosis -Military Hospital 103 from February 2020 to June 2020 shows that the SGRQ scale is valuable and reliable. In addition, SGRQ scores have a close associate with BMI and other subclinical symptoms. Although SGRQ can be applied in clinical practice to evaluate symptoms and affect life in patients with PTB, it needs further adjustment to be completely understandable and more suitable to Vietnamese culture.

List Of Abbreviations
PTB-Pulmonary tuberculosis SGRQ-St. George's Respiratory Questionnaire data was completed by HNT, KMX and CTN. HNT, KMX prepared a draft of the manuscript that was modi ed by CTN, TTN and QD. The nal version was read and approved by all authors.