Utility of photography for trachoma surveys: A systematic review

As trachoma prevalence declines worldwide, performing field surveys with trained trachoma graders to validate its elimination becomes challenging. Graded conjunctival images could be used to train graders, validate field grades, and replace field grading. We review the use of photography to capture conjunctival images for the detection of trachomatous inflammation-follicular (TF) and -intense (TI), Trachomatous trichiasis (TT), and eyelid contour abnormalities. Literature searches were conducted between October 29, 2020 and February 15, 2021. We abstracted kappa scores for agreement between field and image graders, between field graders, and between image graders. The median kappa was calculated for studies discussing grading for TF and TI. Twenty-one studies were included - 19 described the use of imaging tools for grading TF and TI; 1 discussed TT; and 1 discussed eyelid contour abnormalities. For agreement between field and image graders, median kappa for TF was 0.71 (range: 0.57–0.92) and TI was 0.67 (0.37–0.74); between image graders, median kappa for TF was 0.65 (0.37–0.78) and TI was 0.76 (0.39–0.83); and between field graders, median kappa for TF was 0.86 (0.73–1) and TI was 0.82 (0.73–0.91). There is potential for the use of photography for trachoma prevalence surveys. Further research is warranted, particularly for TT and using newer smartphones.