The Nepal country roadmap to zero leprosy recommends high levels of expertise to facilitate the early detection and reduction in transmission of leprosy (9,21). Contrary to that, the median score of knowledge and skills of LFPs in Dhanusa and Mahottari was just 0.53. The low documentation score of 0.37 in the study districts is a challenging issue, because high quality documentation is required to assess the progress and validate the elimination of leprosy (22). The exact situation of leprosy elimination in Dhanusa and Mahottari was therefore likely uncertain, as the median case management and median overall performance was just 0.54 and 0.48 respectively.

Fifty percent of LFPs had knowledge and skills less than 53%, which is low and is in contrast Mohite et al’s study in Bangladesh, where 88% of health workers had good knowledge of leprosy (20). The highest scores were just 63%. Similarly, most peripheral health facilities had poor performance scores posing a serious concern for elimination of transmission of leprosy by the year 2030.

Ninety six percent of the LFPs knew at least 1 cardinal sign, a more satisfactory result compared to a study in Bangladesh (72%) (23). However, only 19% of LFPs could palpate at least one peripheral nerve, revealing gaps in leprosy diagnosis in Nepal which need addressing. Only 19% (6 out 31) health facilities received a 50% aggregate performance regarding leprosy management, crucial for achieving the goal of elimination of transmission by 2030. The finding that >80% of the LFPs could not palpate any peripheral nerves is comparable to the study conducted by Roy et al that showed 83% health workers in parts of India were poorly skilled (13).

Those workers who have taken any of the leprosy training (BLT or CLT) had significantly higher levels of knowledge and skill (OR = 8.1, 95% CI, 1.4-66.3, p-value = 0.01), highlighting the importance of training since leprosy is either not included or not prioritized in the academic course of paramedical staff. Other variables including education level and duration of experience were not significant markers of increased knowledge and skills. However, in multivariate regression of the raw scores, Mahottari had significantly higher knowledge and skills than Dhanusa (β= 0.63, standard error (SE) = 0.21, p-value = 0.002; Figure 2), but not training. This needs to be explored further.

Less than 5% (4.8%) of cases were diagnosed through active case detection and the majority of cases (78%) were diagnosed through the Lalgadh Leprosy Hospital, which drew the governments’ attention to hidden cases in the community. The cases are diagnosed at the specialty hospitals only after the leprosy cases reach the advanced stage. This means the proportion of the cases diagnosed at the peripheral health facilities should be greater, as people in the early stages of leprosy visit these facilities. In this study, few leprosy cases were diagnosed in the peripheral health facilities. So, early cases could be hidden in community due to the inefficient performance of health worker, unavailability of diagnosis and/or the stigma associated with the disease.

Both disability score and EHF score were not significantly improved after full treatment and compliance with MB MDT treatment, which was surprising as comparable studies conducted by Kumar et al show as the disability prevalence in non-compliant cases was significantly greater than fully treated patients (15). This may, however, be due to the very small sample size and time frame and requires further investigation. That all cases have completed treatment in time and were released after treatment indicates a well performing health system for those recruited. Moreover, counselling of patients by a LFP had significantly changed the belief of the patient that self-care can improve disability (OR = 7.9, 95% CI, 1.3-89.1, p-value = 0.01). The result was comparable with the studies conducted by Lusli et al and Devkota et al (24,25), where counselling has a significant role in changing the belief and attitude towards health behaviors. Similarly, counseling significantly improved self-care practice (OR = 46, 95% CI, 2.5-867%, p-value = <0.001).

Among the cases monitored, only 8% developed lepra reactions, of which 2.7% (5 patients) and 5.4% (10 patients) were type 1 and type 2 reactions respectively. Other studies found 17.9% of leprosy cases developed type 1 reactions and between 1.2% to 15.4% of the leprosy cases developed type 2 reactions (26,27). These are similar trends but likely limited by sample sizes.

Regarding the satisfaction from serviced provided from health facilities, 95.2% leprosy cases were satisfied with the services, which is comparable to a study in Brazil, where between 92-100% of patients were satisfied at some level (28).

Weaknesses of our study are the convenience sampling of the districts, though this was deliberate due to disease burden, and patient enrolment by the LFPs of respective health facilities. This might have biased the sample and future prospective and randomized control studies might improve this. Moreover, the sample size is low due to low case numbers, potentially preventing more detailed analyses. For example, we used univariate count data analyses where continuous data and multivariate analyses would be beneficial, but we had too many variables for the small sample size. Similarly, we had a large, zero-inflated sample, with many patients having zero disability or EHF score. Zero-inflated beta regression tools exist and could be used, but failed to fit to our data, even using a Bayesian framework (not shown), likely due to too many zeros and prior distribution specification sensitivity. These factors limit some of the conclusions that can be made highlights the need for more comprehensive study.