Re-envisioning the "F" in SAFE: Facial cleanliness and hygiene metrics, personal hygiene practices, and trachoma risk: Operational research for enhanced evaluation in Tanzania

Background: Trachoma is spread from person to person through contact with infected ocular and likely nasal secretions. Evidence suggests that signs of an “unclean” face are associated with trachoma and facial “cleanliness” was protective against active trachoma. Facial cleanliness could help reduce transmission of trachoma by: (1) reducing the amount of infected eye secretions available for transmission to uninfected individuals; (2) reducing the attractiveness of faces to the putative fly vector. (3) reducing the volume of uninfected eye discharge that could act as a sticky trap and portal of entry for incoming infection. However, despite research definitions, within the trachoma community there exists no universally endorsed definition of facial cleanliness or recommendations for the standardized measurement of F in SAFE. This project proposed to test the reliability of a quantitative metric that measures dirt on the face (using wipes and standardized approach to clean the orbit area on the face). This was compared to standard observations of ocular and nasal discharge (unclean face). To test sensitivity to change over time, measurements were taken in children four hours apart and under a variety of washing face conditions. Methods: Part One: 7 face and wipe graders were trained and standardized against a master grader (HM and MW) before assessing reliability in 91 children seen twice in rapid succession. Grades on the wipes and grades on clean faces were assessed for intra-and inter-observer reliability using the kappa statistic, weighted in evaluation of the wipes. Part two: Population: Four communities were censused and 50 households with children ages 1-5 years were selected. Each were randomized to one of 4 groups: Washed with soap and water, washed with water, wiped with a towel, and control (where no cleaning was done). Each child enrolled was observed and wipes taken twice, 4 hours apart. After re-randomization, the study was repeated, with the addition of detailed instructions on how to wash/wipe faces. Results: The intra-observer reliability of grading the color on the face wipes ranged from weighted kappa= 0.40 to 0.75, with an average over all examinees of 0.57. The interobserver reliability ranged from 0.39 to 0.76, with overall reliability of 0.52. The intra-observer reliability of grading a clean face ranged from kappa=0.47-0.78, with an average of 0.64. The inter-observer reliability ranged from kappa=0.34 to 0.75, with an average of 0.51. All face washing interventions, regardless of type or degree of instruction, resulted in immediate improvement in the proportion of children with clean faces, and improvement in face wipe scores. Using a measure of clean face, the greatest immediate improvement was in the group with instructed washing with water and soap, who showed a difference of 44% in clean faces. This group, and the group with instructed washing with water, showed the most immediate improvement using the mean or median value of the face wipe score. The control group did not appreciably change over time. However, after 4 hours, the proportion of clean faces and the grade of facial wipes declined in all groups, although not to pre-treatment levels. Using measure of clean faces, the groups which declined the least and had the highest prevalence after four hours were the groups using usual or instructed wiping of face (without water or soap). Using the measure of mean score of facial wipes, the groups using instructed washing with water or water and soap had the highest mean scores after four hours. Conclusions/Next steps: These observational results need to be subjected to rigorous statistical testing. In general, both observation of clean faces and facial wipe scores have similar reliability, and detect changes resulting from facial cleaning and change over time. One metric, prevalence of clean faces, likely picks up change with wiping ocular and nasal secretions using a cloth, a strategy which may better than using a mother’s hand and water or soap and water. The other metric, facial wipe score, picks up change in dirt around the eye with the use of water, and soap and water; the latter maybe better strategies for removing dirt than wiping a face.