Sleep Breath (2014) 18:715–721 DOI 10.1007/s11325-013-0889-1
ORIGINAL ARTICLE
Evaluation and validation of four translated Chinese questionnaires for obstructive sleep apnea patients in Hong Kong Stanley Ching Nam Ha & Dennis Lip Yen Lee & Victor James Abdullah & Charles Andrew van Hasselt
Received: 5 May 2013 / Revised: 29 July 2013 / Accepted: 19 August 2013 / Published online: 3 September 2014 # Springer-Verlag Berlin Heidelberg 2013
Abstract Background The present study validates and evaluates the sensitivity and specificity of four internationally popular questionnaires, translated into Chinese, for assessing suspected obstructive sleep apnea (OSA) patients, namely, the Berlin questionnaire, the ASA checklist, the STOP questionnaire and the STOP–BANG questionnaire. Their predictive values in OSA risks in patients presenting with OSA symptoms are examined. Questionnaires may be helpful in prioritizing polysomnography (PSG) and in treatment for the more severe cases. Methods All patients attending our sleep laboratory for overnight PSG were recruited. They were asked to complete three questionnaires (Berlin, ASA checklist and STOP) 2 weeks before and on the same night as the PSG. STOP–BANG questionnaire, an extended STOP with demographic data, ‘B’-body mass index (BMI), ‘A’-age, ‘N’-neck circumference and ‘G’-gender was completed by our technologists using the patient’s completed STOP. Results A number of 141 patients were recruited. The sensitivities and specificities for STOP–BANG with cutoffs at PSG’s RDI=5, RDI=15 and RDI=30 were 81 % to 86 % and 34 % to 57 %, respectively. The high-risk group patients identified by STOP–BANG had significantly higher respiratory disturbance index and lower minimum oxygen saturation than the low-risk group patients.
S. C. N. Ha : D. L. Y. Lee : V. J. Abdullah Department of Otorhinolaryngology, Head and Neck Surgery, United Christian Hospital, Kowloon East Cluster, Hong Kong, China D. L. Y. Lee : V. J. Abdullah : C. A. van Hasselt (*) Department of Otorhinolaryngology, Head and Neck Surgery, The Chinese University of Hong Kong, Prince of Wales Hospital, 30-32 Ngan Shing Street, Shatin, N.T., Hong Kong, China e-mail: [email protected]
Conclusion Among the four questionnaires studied, STOP– BANG, with only eight questions and the highest sensitivity, is the best questionnaire of the four for OSA screening. This can potentially assist in prioritizing PSG and can be helpful in clinical or self-evaluation by the general public. Keywords Berlin . ASA checklist . STOP . Obstructive sleep apnea . Screening tool . Polysomnography
Introduction The prevalence of obstructive sleep apnea (OSA) in Hong Kong in middle-aged women and middle-aged men is 2 % and 4 %, respectively [1, 2]. Currently, overnight polysomnography (PSG) remains the gold standard in diagnosing OSA. It is, nonetheless, labor-intensive and resource-consuming. For most public institutions, a long waiting list is understandably inevitable [3]. This usually delays treatment for OSA patients. Questionnaire screening is an easy and simple-to-use tool. In the clinical setting, it could potentially help in the prioritization of cases for PSG. Early treatment might be relevant in moderate to severe OSA patients because of their related complications. The Berlin questionnaire (Appendix 1), developed at the Conference on Sleep in Primary Care in April 1996 in Berlin, Germany, is a commonly used questionnaire for the screening of OSA patients. It includes three main categories of questions with a total of 11 questions. Each question consists of two to five options. Patients can be classified as high-risk group or low-risk group, and the classification method is illustrated in Appendix 1. Previous studies have shown the sensitivities and specificities to be 62.5 % to 72.1 % and 53.8 % to 77 %, respectively, for identifying OSA [4–8]. The American Society of Anesthesiologists (ASA) Task Force on Perioperative Management of Patients with Obstructive Sleep Apnea had developed a checklist, ASA checklist
716
(Appendix 2), to assess the presence and severity of OSA patients [9]. It includes three main question categories with a total of 12 questions for adults and 14 questions for children. All questions are dichotomous. Patients can be classified as high-risk group or low-risk group, and the classification method is illustrated in Appendix 2. A previous study had shown the sensitivity and specificity to be 72 % and 38 %, respectively, for identifying OSA [4]. The STOP questionnaire (Appendix 3) is a relatively new screening questionnaire (S snoring, T tiredness during daytime, O observed apnea and P high blood pressure). It consists of four dichotomous questions. Patients can be classified as high-risk group or low-risk group and the classification method can be referenced to Appendix 3. The sensitivity and specificity had been shown to be 66 % and 60 %, respectively, for identifying OSA [4, 10]. The STOP–BANG questionnaire (Appendix 4), modified from STOP, includes B body mass index (BMI), A age N-neck circumference and G gender; hence, BANG has been reported to have sensitivities and specificities of 84 % to 85 % and 53 % to 56 %, respectively, for identifying OSA [10, 11]. The classification of high-risk group or low-risk group for STOP–BANG questionnaire is illustrated in Appendix 4. All four questionnaires are used internationally although there is, so far, only one study done in Asia for the STOP– BANG questionnaire [11]. Each questionnaire can potentially classify patients into either low-risk group or high-risk group of suffering from OSA. In the present study, we evaluated the sensitivity and specificity of the four translated Chinese questionnaires (Berlin, ASA checklist, STOP and STOP–BANG) in Hong Kong Chinese patients. Their values in predicting the presence of OSA and the correlation of low- and high-risk groups with the PSG respiratory disturbance index (RDI) results in 141 patients were also examined.
Study design A prospective study of suspected OSA patients referred for PSG at the Sleep Laboratory of the Department of Otorhinolaryngology, Head and Neck Surgery, United Christian Hospital.
Sleep Breath (2014) 18:715–721
weight, neck circumference and gender were recorded. They were asked to complete three questionnaires (Berlin questionnaire, ASA checklist and STOP questionnaire) at home 2 weeks before and in our ward on the same night as the PSG. The three questionnaires were translated into Chinese by the first author (SCH). The three questionnaires were then back-translated by other authors (DLY, VJA). The backtranslated questionnaires were compared with the original English version questionnaires to check that the questions were properly translated. The three Chinese questionnaires were distributed to 10 patients with snoring or daytime sleepiness. No modifications had to be implemented. Electroencephalogram, electrooculogram, electrocardiogram, leg and chin electromyograms, nasal/oral airflow by thermistor and nasal cannula, oxygen saturation, body position, thoracic and abdominal movements and snoring are recorded with synchronized on screen video recordings. Sleep technologists, who scored the PSGs according to the American Academy of Sleep Medicine Manual for the Scoring of Sleep and Associated Events [12], were blinded to the results of questionnaires. Statistical analysis was performed by SPSS version 20. The internal consistency of the translated questionnaires was assessed by Cronbach’s alpha. Since the four questions in the STOP questionnaire and the questions in ASA checklist reflected different parameters, internal consistency checking was not applicable. Cohen Kappa coefficients were obtained to assess the test–retest reliability of the translated questionnaires by completing the same questionnaires 2 weeks before and on the same night of the PSG. The construct validity was assessed by checking if the high-risk group patients were more severe than the low-risk group patients measuring by RDI. The results of the PSG and questionnaires were compared using sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) to evaluate the efficacy of individual questionnaire to screen and assess severity with a view to help prioritize OSA patients for PSG for Chinese patients in Hong Kong. The study is approved by the Cluster Research Ethics Committee (Kowloon Central/Kowloon East) (Ref: KC/KE10-0101/FR-3).
Results Subjects and methods Demographic characteristics of patients We recruited 141 patients aged 18 years old or above with habitual snoring, with or without daytime hypersomnolence referred for overnight PSG (SomnoMedics, SOMNOscreen) at the Department of Otorhinolaryngology, Head and Neck Surgery, United Christian Hospital in Hong Kong. Full informed consent was obtained, and the first part of the ASA checklist was completed in the clinic. Body height, body
A number of 141 patients [81.6 % males, average BMI= 25.83, average age=44.82, average apnea hypopnea index (AHI)=24.97, average RDI=29.70] were recruited. Table 1 shows the demographic data of the patients classified by different questionnaires into high- and low-risk groups, i.e., Berlin questionnaire (Appendix 1), STOP questionnaire
Sleep Breath (2014) 18:715–721
717
Table 1 Demographic data (comparison between low-risk group and high-risk group among four questionnaires, i.e., Berlin, STOP and STOP–BANG questionnaires and ASA checklist) All (n =141)
Berlin questionnaire
STOP questionnaire
STOP–BANG questionnaire
ASA checklist
Low riska (n =39)
High riska (n =101)
Low riska (n =42)
High riska (n =97)
Low riska (n =37)
High riska (n =102)
Low riska (n =56)
High riska (n =77)
No. of male/female Age BMI PSG data AHI RDI
115/26 45±11 26±4
29/10 41±11 24±5
85/16 46±11** 27±4**
36/6 42±12 25±4
77/20 46±11* 26±4**
24/13 38±11 23±4
89/13 ** 47±11** 27±4**
46/10 43±12 25±4
53/14 47±11 26±4
25±24 30±35
19±20 21±20
27±25** 33±40*
18±20 20±21
28±25** 34±40**
15±20 17±21
29±24** 35±38**
25±23 27±24
27±25 33±43
Min SaO2, %
80±9
82±8
80±10
84±8
79±10**
86±8
78±9**
81±9
79±10
STOP–BANG questionnaire has the best performance in identifying risk factors (gender, age and BMI) of OSA and more severe OSA patients. The data are presented in mean ± standard deviation *0.1
ORIGINAL ARTICLE
Evaluation and validation of four translated Chinese questionnaires for obstructive sleep apnea patients in Hong Kong Stanley Ching Nam Ha & Dennis Lip Yen Lee & Victor James Abdullah & Charles Andrew van Hasselt
Received: 5 May 2013 / Revised: 29 July 2013 / Accepted: 19 August 2013 / Published online: 3 September 2014 # Springer-Verlag Berlin Heidelberg 2013
Abstract Background The present study validates and evaluates the sensitivity and specificity of four internationally popular questionnaires, translated into Chinese, for assessing suspected obstructive sleep apnea (OSA) patients, namely, the Berlin questionnaire, the ASA checklist, the STOP questionnaire and the STOP–BANG questionnaire. Their predictive values in OSA risks in patients presenting with OSA symptoms are examined. Questionnaires may be helpful in prioritizing polysomnography (PSG) and in treatment for the more severe cases. Methods All patients attending our sleep laboratory for overnight PSG were recruited. They were asked to complete three questionnaires (Berlin, ASA checklist and STOP) 2 weeks before and on the same night as the PSG. STOP–BANG questionnaire, an extended STOP with demographic data, ‘B’-body mass index (BMI), ‘A’-age, ‘N’-neck circumference and ‘G’-gender was completed by our technologists using the patient’s completed STOP. Results A number of 141 patients were recruited. The sensitivities and specificities for STOP–BANG with cutoffs at PSG’s RDI=5, RDI=15 and RDI=30 were 81 % to 86 % and 34 % to 57 %, respectively. The high-risk group patients identified by STOP–BANG had significantly higher respiratory disturbance index and lower minimum oxygen saturation than the low-risk group patients.
S. C. N. Ha : D. L. Y. Lee : V. J. Abdullah Department of Otorhinolaryngology, Head and Neck Surgery, United Christian Hospital, Kowloon East Cluster, Hong Kong, China D. L. Y. Lee : V. J. Abdullah : C. A. van Hasselt (*) Department of Otorhinolaryngology, Head and Neck Surgery, The Chinese University of Hong Kong, Prince of Wales Hospital, 30-32 Ngan Shing Street, Shatin, N.T., Hong Kong, China e-mail: [email protected]
Conclusion Among the four questionnaires studied, STOP– BANG, with only eight questions and the highest sensitivity, is the best questionnaire of the four for OSA screening. This can potentially assist in prioritizing PSG and can be helpful in clinical or self-evaluation by the general public. Keywords Berlin . ASA checklist . STOP . Obstructive sleep apnea . Screening tool . Polysomnography
Introduction The prevalence of obstructive sleep apnea (OSA) in Hong Kong in middle-aged women and middle-aged men is 2 % and 4 %, respectively [1, 2]. Currently, overnight polysomnography (PSG) remains the gold standard in diagnosing OSA. It is, nonetheless, labor-intensive and resource-consuming. For most public institutions, a long waiting list is understandably inevitable [3]. This usually delays treatment for OSA patients. Questionnaire screening is an easy and simple-to-use tool. In the clinical setting, it could potentially help in the prioritization of cases for PSG. Early treatment might be relevant in moderate to severe OSA patients because of their related complications. The Berlin questionnaire (Appendix 1), developed at the Conference on Sleep in Primary Care in April 1996 in Berlin, Germany, is a commonly used questionnaire for the screening of OSA patients. It includes three main categories of questions with a total of 11 questions. Each question consists of two to five options. Patients can be classified as high-risk group or low-risk group, and the classification method is illustrated in Appendix 1. Previous studies have shown the sensitivities and specificities to be 62.5 % to 72.1 % and 53.8 % to 77 %, respectively, for identifying OSA [4–8]. The American Society of Anesthesiologists (ASA) Task Force on Perioperative Management of Patients with Obstructive Sleep Apnea had developed a checklist, ASA checklist
716
(Appendix 2), to assess the presence and severity of OSA patients [9]. It includes three main question categories with a total of 12 questions for adults and 14 questions for children. All questions are dichotomous. Patients can be classified as high-risk group or low-risk group, and the classification method is illustrated in Appendix 2. A previous study had shown the sensitivity and specificity to be 72 % and 38 %, respectively, for identifying OSA [4]. The STOP questionnaire (Appendix 3) is a relatively new screening questionnaire (S snoring, T tiredness during daytime, O observed apnea and P high blood pressure). It consists of four dichotomous questions. Patients can be classified as high-risk group or low-risk group and the classification method can be referenced to Appendix 3. The sensitivity and specificity had been shown to be 66 % and 60 %, respectively, for identifying OSA [4, 10]. The STOP–BANG questionnaire (Appendix 4), modified from STOP, includes B body mass index (BMI), A age N-neck circumference and G gender; hence, BANG has been reported to have sensitivities and specificities of 84 % to 85 % and 53 % to 56 %, respectively, for identifying OSA [10, 11]. The classification of high-risk group or low-risk group for STOP–BANG questionnaire is illustrated in Appendix 4. All four questionnaires are used internationally although there is, so far, only one study done in Asia for the STOP– BANG questionnaire [11]. Each questionnaire can potentially classify patients into either low-risk group or high-risk group of suffering from OSA. In the present study, we evaluated the sensitivity and specificity of the four translated Chinese questionnaires (Berlin, ASA checklist, STOP and STOP–BANG) in Hong Kong Chinese patients. Their values in predicting the presence of OSA and the correlation of low- and high-risk groups with the PSG respiratory disturbance index (RDI) results in 141 patients were also examined.
Study design A prospective study of suspected OSA patients referred for PSG at the Sleep Laboratory of the Department of Otorhinolaryngology, Head and Neck Surgery, United Christian Hospital.
Sleep Breath (2014) 18:715–721
weight, neck circumference and gender were recorded. They were asked to complete three questionnaires (Berlin questionnaire, ASA checklist and STOP questionnaire) at home 2 weeks before and in our ward on the same night as the PSG. The three questionnaires were translated into Chinese by the first author (SCH). The three questionnaires were then back-translated by other authors (DLY, VJA). The backtranslated questionnaires were compared with the original English version questionnaires to check that the questions were properly translated. The three Chinese questionnaires were distributed to 10 patients with snoring or daytime sleepiness. No modifications had to be implemented. Electroencephalogram, electrooculogram, electrocardiogram, leg and chin electromyograms, nasal/oral airflow by thermistor and nasal cannula, oxygen saturation, body position, thoracic and abdominal movements and snoring are recorded with synchronized on screen video recordings. Sleep technologists, who scored the PSGs according to the American Academy of Sleep Medicine Manual for the Scoring of Sleep and Associated Events [12], were blinded to the results of questionnaires. Statistical analysis was performed by SPSS version 20. The internal consistency of the translated questionnaires was assessed by Cronbach’s alpha. Since the four questions in the STOP questionnaire and the questions in ASA checklist reflected different parameters, internal consistency checking was not applicable. Cohen Kappa coefficients were obtained to assess the test–retest reliability of the translated questionnaires by completing the same questionnaires 2 weeks before and on the same night of the PSG. The construct validity was assessed by checking if the high-risk group patients were more severe than the low-risk group patients measuring by RDI. The results of the PSG and questionnaires were compared using sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) to evaluate the efficacy of individual questionnaire to screen and assess severity with a view to help prioritize OSA patients for PSG for Chinese patients in Hong Kong. The study is approved by the Cluster Research Ethics Committee (Kowloon Central/Kowloon East) (Ref: KC/KE10-0101/FR-3).
Results Subjects and methods Demographic characteristics of patients We recruited 141 patients aged 18 years old or above with habitual snoring, with or without daytime hypersomnolence referred for overnight PSG (SomnoMedics, SOMNOscreen) at the Department of Otorhinolaryngology, Head and Neck Surgery, United Christian Hospital in Hong Kong. Full informed consent was obtained, and the first part of the ASA checklist was completed in the clinic. Body height, body
A number of 141 patients [81.6 % males, average BMI= 25.83, average age=44.82, average apnea hypopnea index (AHI)=24.97, average RDI=29.70] were recruited. Table 1 shows the demographic data of the patients classified by different questionnaires into high- and low-risk groups, i.e., Berlin questionnaire (Appendix 1), STOP questionnaire
Sleep Breath (2014) 18:715–721
717
Table 1 Demographic data (comparison between low-risk group and high-risk group among four questionnaires, i.e., Berlin, STOP and STOP–BANG questionnaires and ASA checklist) All (n =141)
Berlin questionnaire
STOP questionnaire
STOP–BANG questionnaire
ASA checklist
Low riska (n =39)
High riska (n =101)
Low riska (n =42)
High riska (n =97)
Low riska (n =37)
High riska (n =102)
Low riska (n =56)
High riska (n =77)
No. of male/female Age BMI PSG data AHI RDI
115/26 45±11 26±4
29/10 41±11 24±5
85/16 46±11** 27±4**
36/6 42±12 25±4
77/20 46±11* 26±4**
24/13 38±11 23±4
89/13 ** 47±11** 27±4**
46/10 43±12 25±4
53/14 47±11 26±4
25±24 30±35
19±20 21±20
27±25** 33±40*
18±20 20±21
28±25** 34±40**
15±20 17±21
29±24** 35±38**
25±23 27±24
27±25 33±43
Min SaO2, %
80±9
82±8
80±10
84±8
79±10**
86±8
78±9**
81±9
79±10
STOP–BANG questionnaire has the best performance in identifying risk factors (gender, age and BMI) of OSA and more severe OSA patients. The data are presented in mean ± standard deviation *0.1
