Background: The World Health Organization-recommended strategy for trachoma elimination as a public health problem is known by the acronym "SAFE", where "F" stands for facial cleanliness to reduce transmission of ocular Chlamydia trachomatis infection. Accurately and reliably measuring facial cleanliness is problematic. Various indicators for measuring an unclean face exist, however, the accuracy and reliability of these indicators is questionable and their relationship to face washing practices is poorly described.

Methods: Clean face indicator (ocular or nasal discharge, flies on the face, and dirt on the face), trachoma clinical sign, and ocular C. trachomatis infection data were collected for 1613 children aged 0-9 years in 12 Senegalese villages as part of a cross-sectional trachoma prevalence study. Time of examination was recorded to the nearest half hour. A risk factor questionnaire containing Water, Sanitation and Hygiene (WASH) questions was administered to heads of compounds (households that shared a common doorway) and households (those who shared a common cooking pot).

Results: WASH access and use were high, with 1457/1613 (90.3%) children living in households with access to a primary water source within 30 min. Despite it being reported that 1610/1613 (99.8%) children had their face washed at awakening, > 75% (37/47) of children had at least one unclean face indicator at the first examination time-slot of the day. The proportion of children with facial cleanliness indicators differed depending on the time the child was examined. Dirt on the face was more common, and ocular discharge less common, in children examined after 11:00 h than in children examined at 10:30 h and 11:00 h.

Conclusions: Given the high reported WASH access and use, the proportion of children with an unclean face indicator should have been low at the beginning of the day. This was not observed, explained either by: the facial indicators not being reliable measures of face washing; eye discharge, nose discharge or dirt rapidly re-accumulated after face washing in children in this population at the time of fieldwork; and/or responder bias to the risk factor questionnaire. A high proportion of children had unclean face indicators throughout the day, with certain indicators varying by time of day. A reliable, standardised, practical measure of face washing is needed, that reflects hygiene behaviour rather than environmental or cultural factors.

The authors declare that they have no competing interests.

• 078460/Z/05/Z/WT_/Wellcome Trust/United Kingdom