Simplified Berlin Questionnaire for Screening of High Risk for Obstructive Sleep Apnea Among Thai Male Healthcare Workers

Abstract

: Obstructive Sleep Apnea (OSA) is a common disease associated with major cardiovascular diseases. Male subjects are more at higher risk for OSA than female subjects. The Berlin questionnaire is a beneficial screening tool for OSA and has 14 items. The Berlin questionnaire may need some adjustment for Thai or Asian populations. We aimed to find items that should be asked in the Berlin questionnaire to identify high risk for obstructive sleep apnea among Thai male healthcare workers. This study was performed in Thai male healthcare workers over the age of 35 and currently working at the Faculty of Medicine, Khon Kaen University. The Thai version of the Berlin questionnaire was randomly distributed. A study population of 273 subjects was required to provide a confidence value of 95%. An item analysis of the Berlin questionnaire was evaluated as independent factors for being high risk of OSA by using a multivariate logistic regression analysis. Of the 273 distributed questionnaires, 135 subjects returned then (49.5% response rate). Of those, 41 (30.4%) were identified as being at high risk of OSA. Only three items of the Berlin questionnaire, including frequent snoring, high body mass index and hypertension, were independently associated with being at high risk for OSA. In conclusion, the Berlin questionnaire can be shortened to identify high risk for OSA by itself; not polysomnography.

A New Berlin Questionnaire Simplified by Machine Learning Techniques in a Population of Italian Healthcare Workers to Highlight the Suspicion of Obstructive Sleep Apnea

“…The subset of three variables was reasonably robust and did not depend too much on the particular dataset split; after about 60 runs, the subset was found to contain the variables computed from B10 (selected at every run), B1 (present in 73% of the “best” feature subsets), B6 (presence in 38%), B7 (37%), B3 (25%), B4 (2%). We remark that hypertension B10 is always among the most useful features [which was already known from the single-variable calculations; this result confirms what was found in ( 18 )]. Considering now the remaining five features, three concern snoring (B1, the most voted after B10; then B3 and B4) while two concern feeling tired in daytime, either at wake-up or along the day (B6 and B7), with similar presence in the subsets.…”

“…As to the two variables with relatively objective measurement, i.e., having high blood pressure, B10, and the body mass index (computed from the subject physical data), the former had high predictivity (AUROC = 0.74, F = 0.80) while the latter showed lower discriminant power (AUROC = 0.64, F = 0.03). This result was quite surprising if compared with the one reported in ( 18 ) where BMI is found to be quite a strong predictor.…”

“…This reduced questionnaire shows some similarity with the one proposed in Arunsurat et al ( 18 ) with the important difference that (as already remarked) BMI is not preserved in the reduced set. The discrepancy might partly come from the different group considered, i.e., the high percentage of young and prevalently female respondents in our sample compared to the all-male healthcare workers investigated in Arunsurat et al ( 18 ).…”

Machine Learning (ML) Models to Enhance the Berlin Questionnaire (BQ) Detection of Obstructive Sleep Apnea (OSA) at-Risk Patients

“…In the quest for a straightforward questionnaire to ascertain OSA risk, clinicians demand enhancements to existing tools. Arunsurat et al [34] posited that with certain modifications, the BQ could serve effectively as an OSA screening instrument. Furthermore, Stelmach-Mardas et al [35] add to the growing body of evidence indicating the BQ inadequacy in distinguishing between high and low-risk patients, suggesting the need for the development of alternative protocols to heighten diagnostic precision for such individuals.…”