Background: The Epworth Sleepiness Scale (ESS) is a tool widely used to assess excessive daytime sleepiness. Unfortunately, it is not reliable in low-income countries where situations such as reading a book, watching TV or theater or even driving a car are not common. The aim of this study was thus to assess the performance of a modified version of the Epworth scale in a low-income country. 

Method: Participants were recruited from both rural and urban area of Benin in the frame of the BeSAS study cohort. In the modified ESS (mESS), questions 1, 2, 3, and 8 were redesigned to reflect common living situations in Benin.  The dimensionality of the mESS have been evaluated through a screenplot and the internal coherence assessed using the Cronbach alpha coefficient.  The discriminatory ability of the scale was also assessed. 

Results: A total of 2909 participants were recruited, 1129 were male (38.9%) with a mean age (SD) 44.7 (14.5) years. Overall, 52.4% (1526) completed all the mESS questions while 453 (15.6%) completed the standard  ESS. The Cronbach alpha coefficient of the mES was 0.86 showing good internal coherence. Concerning the discriminatory ability, mean scores for mESS was 7.8 for participants with Insomnia Severity Index (ISI) < 8 vs 9.2 for participants with  ?8 (p<0.001), 7.8 for participants with Pittsburg Sleep Questionnaire (PSQI) <5 vs 8.3 for participants with PSQI?5 (p=0.03). 

Conclusion: Although the mES is not a perfect score, it appears more relevant in the Benin population than the original Epworth scale but needs further validation/improvement in other low-income countries.