Objective Developing better tools to screen obstructive sleep apnea (OSA) in professional drivers is fundamental in occupational health services, given the serious consequences this condition can have on their safety.

Methods 1928 professional drivers were screened for OSA for three years using Berlin and Epworth questionnaires. 10.3% were suspected to be positive for OSA and send for polysomnography tests. We had a return of 39 data. We analysed the results with the MAP index because it has shown a high sensitivity and specificity in studies on professional drivers. The MAP index is easy to use and combines self-reported apnea symptoms (snoring and cessation of breathing) as well as objective data like age, sex, and body mass index (BMI). As the apnoea symptoms regarding snoring and observed apnoea are often not properly reported, we included the questions of the Berlin questionnaire about tiredness in order to improve the predictability and calculated a MAP2. To determine diagnostic test accuracy, we calculated both MAP indexes with apnea-hypopnea index (AHI) criteria already published in articles. Sensitivity, specificity, negative and positive predictive values (NPV, PPV) and area under curve (AUC) for receiver operating characteristic (ROC) were analysed. Predictive utility of both indexes was examined by characteristic variables, age ( < 50 or ≥ 50 years) and BMI ( < 30 or ≥ 30 kg/m²).

Results Employing the MAP_index1 for OSA, the sensitivity obtained was 63.3% and specificity was 41.2%. The PPV was 58.3% and NPV was 46.7%. The AUC was 0.524 (95% CI 0.339–0.709). The accuracy was higher in younger versus older drivers (AUC 0.701 versus 0.620). Sensitivity for Map index2 was 59.1% and specificity was 52.9%. The PPV was 61.9% and the NPV was 50.0%. The AUC was 0.560 (95%CI 0.376–0.744).

Conclusions The MAP 2 was had lower sensitivity but higher specificity. It cannot be considered as a better tool to predict obstructive sleep apnoea.