Clinical and microbiological predictors of healing in Buruli ulcer disease

Abstract

IntroductionWound measurements are relevant in monitoring the rate of healing (RoH) and may predict time to healing. Predicting the time to healing can help improve the management of Buruli ulcer. We examine three methods for the determination of RoH and their use as predictors of time to healing. MethodsLesion measurements of Buruli ulcer patients treated from 2007 to 2022 were obtained with acetate sheet tracings (2D) or Aranz software (3D) fortnightly. RoH was determined using the absolute area, percentage area reduction and linear methods at 4 weeks post onset of antibiotic treatment. Predicted time to healing was compared to the actual healing time. Baseline characteristics were assessed for associations with healing. ResultsAll three methods for calculating the RoH significantly distinguished between fast and slow healers (p < 0.0001). The predicted healing time using the linear method was comparable to the actual healing time for fast healers (p = 0.34). The RoH was influenced by the form of lesion, with plaques [OR 2.19 5 %CI (1.2–3.6), p = 0.009], and oedemas [OR 8.5; 95 %CI (1.9––36.9), p = 0.004] being associated with delayed healing. The proportion of patients with paradoxical reactions 16 % vs 3 %, p < 0.0001), higher baseline bacterial load (75/104;72 % vs 21/47;45 %, p = 0.001) and delayed clearance of viable organisms (71/104;68 % vs 9/47;19 %, p < 0.0001) was higher in the slow healers than the fast healers. ConclusionPredicted healing rates were comparatively lower for slow healers than fast healers. Baseline characteristics associated with healing can be explored for an improved disease management plan to reduce patient and caregiver anxiety.