Systematic Review of Survival Analysis in Leprosy Studies—Including the Following Outcomes: Relapse, Impairment of Nerve Function, Reactions and Physical Disability

Leprosy is a public health problem in South American, African and Oceanian countries. National programs need to be evaluated, and the survival analysis model can aid in the construction of new indicators. The aim of this study was to assess the period of time until the outcomes of interest for patients with or exposed to leprosy by means of survival analysis surveys. This review researched articles using the databases of PubMed, Science Direct, Scopus, Scielo and BVS published in English and Portuguese. Twenty-eight articles from Brazil, India, Bangladesh, the Philippines and Indonesia were included. The Kaplan–Meier method, which derives the log-rank test, and Cox’s proportional hazards regression, which obtains the hazard ratio, were applied. The mean follow-up until the following outcomes were: (I) leprosy (2.3 years) in the population who were exposed to it, (II) relapse (5.9 years), (III) clinical manifestations before, during and after treatment—nerve function impairment (5.2 years), leprosy reactions (4.9 years) and physical disability (8.3 years) in the population of patients with leprosy. Therefore, the use of survival analysis will enable the evaluation of national leprosy programs and assist in the decision-making process to face public health problems.


Introduction
Leprosy is considered a neglected and infectious disease [1,2]. The etiologic agent is Mycobacterium leprae, an obligate intracellular bacterium with an affinity for the peripheral nervous system [3]. In general, people affected by the disease suffer social and psychological repercussions due to deformities and physical disabilities caused by the disease's progression [4,5].
The global prevalence rate of leprosy has reached less than one case per 10,000 inhabitants, but it is considered a public health problem in countries located in South America, Africa and Oceania, which have not yet achieved the elimination of the disease [6]. In 2020, 127,558 new cases were detected worldwide, of which 74.0% were located in Brazil, India and Indonesia [6], and grade 2 of physical disability was diagnosed in 7198 (5.6%) individuals in 64 countries [6]. It is understood that there is a need to carry out studies to estimate the time between the date leprosy was diagnosed and the date of discharge, as well as to analyze the risk factors related to the outcome in search of its elimination. The survival analysis model can help in interpretations about the evolution of diagnoses and treatments, collaborating with the description of the behavior of leprosy and the prognostic factors related to it [7,8]. Survival analysis begins by counting the mean followup from the initial observation until the outcome of interest [9,10]. The epidemiological studies that can be used are experimental (clinical trials) and observational cohorts [11,12].
Given that leprosy continues to be a secular public health problem, and further development of the analysis is still a challenge, carrying out a systematic review was chosen as a form of contribution. This review sought to pursue the scientific production of survival analysis studies in patients with leprosy or those who have exposed to leprosy. Thus, the objective was to assess the period of time until the outcomes of interest, which may constitute indicators to evaluate the results of the disease control program in the future.

Registration Protocol
In this systematic review, we used the guidelines and checklist (Table S1) from the PRISMA 2009 preferred reporting items for system reviews and meta-analyzes [13] and the predefined protocol prospective international registry platform for systematic reviews [CRD42022296026] [14].

Data Sources and Research Strategy
The bibliographic survey of electronic databases was carried out in February 2022 without restrictions on the publication date and location. The databases used were PubMed (National Center for Biotechnology Information), Science Direct (Elsevier), Scopus (Elsevier), Scientific Electronic Library Online (SciELO) and the Virtual Health Library (BVS) for journal articles published in English and Portuguese, aiming to cover studies carried out in locations where the incidence of leprosy is higher. The keywords used for the searches were "Leprosy", "Mycobacterium leprae", "Survival", "Survival Analysis", "Survival Rate, Proportional Hazards Models" and "Kaplan-Meier Estimate". These keywords were researched in various forms of combinations. The search strategy performed in PubMed was as follows: ((("Leprosy" [Mesh]) AND (Lepros *)) OR ("Mycobacterium leprae" [Mesh])) OR ("Mycobacterium leprae")) AND ("Survival Analysis" [Mesh]). In addition, the references of eligible articles were consulted through manual searches (hand-searching).

Study Selection and Data Extraction
The articles were selected by two independent reviewers (CCB and ATX) initially by reading the titles and abstracts; some were ineligible for not meeting the inclusion criteria. After the removal of studies that did not meet the criteria and removing duplicates, the full texts were read in order to carry out a new eligibility certification. Based on the inclusion criteria of the study, data were extracted using a data extraction form and, in the case of any lack of consensus, an experienced third reviewer (CVB) would make her assessment. The inclusion criteria were: (I) survival studies on patients with or exposed to leprosy; (II) the mean follow-up until the occurrence of any type of outcome of interest, for example: cure, relapse, reactive episodes, degree of physical disability, illness in household and nonhousehold contacts; (III) application of survival curves using the Kaplan-Meier method (estimates the conditional probability between time and outcome) and/or the regression model (estimates the effect of predictor variables using the risk function). The outcomes were grouped and counted in order to address the five with the highest frequencies in the discussion. Articles were eliminated if they did not meet the inclusion criteria, were not in English or Portuguese, literature reviews, systematic articles with or without meta-analysis or book chapters, reviews, conference abstracts and letters to the editor.
The data were analyzed qualitatively and displayed in the form of tables and narrative text. The following items were extracted from each selected study: name(s) of author(s), year of publication, country, study design, duration of follow-up (years), study population, outcome of interest, number of patients/population and number of events, mean follow-up, incidence density, survival curve methods, comparison of survival curves, regression and measure of association. For the five main outcomes listed by high frequency, the following data were extracted for both paucibacillary and multibacillary: name(s) of author(s), year of publication, mean follow-up (years), incidence density/100 person-years and its 95% confidence interval (CI), Kaplan-Meier curve interpretations in years, hazard ratio and its 95% CI. This extraction was performed independently by two reviewers (CCB and ATX), and in the presence of any discrepancy, a third reviewer (CVB) verified the data.

Risk of Bias and Quality Assessment
The assessment of risks of bias and study quality was performed by two independent reviewers (CCB and ATX) using the Standard Quality Assessment Criteria for Evaluation of Primary Research Papers from a Variety of Fields [15]. The tool consists of 14 items to assess the methodological quality of each study applied in a systematic review. Each item received a score according to the response, such as "0" for negative responses, "1" for partial responses and "2" for positive responses. In this manner, the scores were performed as follows: (I) maximum points are based on 28-(2 × number of N/A), (II) total points are calculated by adding the total amount of points obtained from all items, and (III) the summary score is represented by the total score obtained divided by the maximum score presented in percentage. In the summary score, articles with a percentage above 75% were considered better quality, and articles with a percentage below 55% were considered low quality. Therefore, the higher the final percentage, the lower the risk of bias and the better the quality of the study. Any lack of consensus was discussed with a third reviewer (CVB).

Flow Diagram of Included Studies
The search generated 1601 results (PubMed-40, Science Direct-925, Scopus-10, Scielo-189, VHL-337), of which 1524 were removed after screening titles and abstracts. Of the 77 that remained, 44 duplicates were removed. Subsequently, 33 articles were evaluated based on their full text, with 12 being removed for not meeting the inclusion criteria. Finally, 21 studies were included in this review ( Figure 1).
In the manual search, 12 articles were found, one duplicate was removed and after reading the full text, four articles were removed because they did not meet the inclusion criteria. As a result of the manual search, seven articles were included. The total number of articles included was 28, as detailed in the PRISMA (preferred reporting items for system reviews and meta-analyses) flow diagram shown in Figure 1.

Study Description
All 28 studies included in this systematic review are described in Table 1. The studies were published between 1994 and 2022, which is equivalent to 28 years of scientific research addressing the topic of survival analysis and leprosy. The years that stood out were 2000, 2008 and 2012, with three articles each, and especially 2021, with four. Among the study designs, the cohort type predominated (19; 67.9%), being six (21.4%) prospective, five (17.9%) retrospective and eight (28.6%) unclassified. The follow-up duration of the studies averaged 9.5 years. Regarding the location of the studies, the highlighted countries were Brazil (11; 39.3%), India (6; 21.4%), Bangladesh (5; 17.9%), the Philippines (2; 7.1%) and Indonesia (2; 7.1%) ( Table 1).
Concerning the study population, the studies were divided into leprosy contacts (six; 21.4%). Of these, three studies were selected with intra-household and extra-household contacts. The other study population was leprosy patients (22; 78.6%), and in 10 studies, multibacillary leprosy patients were selected (Table 1).
Within the population of exposed patients and those with leprosy, 12 events of interest were found, and a single study may have up to four outcomes. In the population of leprosy contacts, the outcome was leprosy, and in the population of leprosy patients, the main events of interest were relapse, nerve function impairment, leprosy reactions and physical disabilities ( Table 1).
The total number of events was 3715 with a mean of 148.6 (maximum: 973 and minimum: 9) within a total population of 85,265. The mean follow-up of outcomes was found in 10 (35.7%) articles with an average of 5.6 years until the events occurred, and the mean follow-up in the following events was identified: leprosy (one article; average 2.3 years), relapse (four articles; average 5.9 years), nerve function impairment (one article; 5.2 years), leprosy reactions (two articles; average 4.9 years), physical disability (four articles; average 8.3 years) and other outcomes (two articles; average 2.8 years). The total number of articles exceeds 10, considering the articles contained more than one outcome or the mean result time contained more than one study group (Table 1).

Leprosy
The leprosy outcome was found in six articles (21.4%), of which the country locations were: Brazil (3; 50.0%), Indonesia (2; 16.7%) and India (1; 33.3%). The study population consisted of leprosy contacts and the contact population was specified in five articles, two (16.7%) in household and three (50.0%) in household and neighbor ( Table 1).
In household contacts, multibacillary patients have a higher incidence density of 1313/1000 person-years. The risk estimate (hazard ratio) of the disease occurring in the family contacts of multibacillary patients was 4.60 times higher (95% CI 1. 65-12.9) than in the non-contact group ( Table 2).
In paucibacillary patients, when there was an intervention with more drugs in the treatment, there was a decrease in the incidence density (C4: 1.9/100 person-years; C-ROM: 0.90/100 person-years). In multibacillary patients, those who received 24 doses of the regimen treated to the point of smear negative had an incidence density of 1.11/100 personyears. When comparing paucibacillary and multibacillary patients, the latter group had the highest incidence density, with 0.595/100 person-years. The FDT group multibacillary (24 doses up to the point of smear negative) had a higher probability of no relapse than the NRT (24 doses) multibacillary up to eight years, and the probability of relapse was higher in multi-bacillary patients up to 15 years of age with 12 treatment doses (90.19%) and 24 treatment doses (63.6%) ( Table 3).

Clinical Manifestations before, during and after Treatment
Outcomes encompassed in clinical manifestations before, during and after treatment that stood out were located in 14 articles (50.0%), distributed in leprosy reactions (5), nerve function impairment (7) and physical disabilities (5). The number exceeds 14, since the articles contained more than one outcome. The research locations were in Brazil (6; 42.9%), Bangladesh (4; 28.6%), India (2; 14.3%), Bangladesh and Nepal (1; 7.1%) and the Republic of Zaire (now the Republic of Congo) (1; 7.1%). The study population consisted of patients with leprosy, with multibacillary patients prevailing in six studies (42.9%) ( Table 1).
In leprosy reactions, a shorter mean follow-up was found in paucibacillary patients (six months) than in multibacillary patients (eight months). The probability up to six months of treatment of not occurring leprosy reactions in multibacillary patients with U-MDT (six doses) was higher at 64.14% than in relation to R-MDT (12 doses) at 62.23%. The risk estimate (hazard ratio) of leprosy reactions in the paucibacillary group was 1.244 times higher than in the multibacillary group ( Table 4).
The highest incidence density of nerve function impairment was in the multibacillary group, with 24.4/100 person-years. The probability of nerve function impairment occurring up to two years was higher in the multibacillary group (37.0%) than in the paucibacillary group (2.6%). The risk estimate (hazard ratio) of nerve function impairment occurring in the multibacillary group was 8.8 to 7.5 times greater than in the paucibacillary group ( Table 5).
The incidence density of physical disabilities in multibacillary patients ranged from 2.74 to 6.5/100 person-years. The multibacillary R-MDT group (12 doses) (33.8%) had a higher probability of physical disabilities than the U-MDT group (six doses) (30.06%) up to five years of treatment. The risk estimate (hazard ratio) of physical disabilities occurring in multibacillary after treatment was 2.80 times higher in those with grade 2 disabilities compared to those with grade 0 (Table 6).

Quality Assessment Criteria
To assess the quality of the studies, the Standard Quality Assessment Criteria for Evaluating Primary Research Papers from a Variety of Fields were used [15]. The quality percentages of studies ranged from 41 to 86%, with a mean of 63.2%. Five (17.9%) studies scored less than 55% on the summary score [8,16,19,26,37]. The studies that varied between 56% and 74% were 20 (71.4%) articles considered to have more complete data compared to the previous score. Whilst three (10.7%) articles had a score above 75%, having better quality information: Penna [22]. Five selected articles had scores below 55%, however, these remained because they were inclusive and comprehensive studies. Table 7 shows the breakdown of quality scores for each study.   NA-Not available. a The two single-dose treatment regimens (adult doses) were: C2: 40 mg/kg rifampicin (RMP) and 1200 mg clofazimine (CLO); C4: 40 mg/kg RMP, 100 mg CLO, 100 mg dapsone (DDS) and 500 mg ethionamide (ETH). b Evaluation of multidrug therapy (R-MDT) with three drugs-rifampicin, dapsone and clofazimine, in two patient groups MB, uniform treatment (U-MTD) for 6 months and R-MDT for 12 months according to the WHO recommendation. c Mean follow-up: 162 months for paucibacillary and 151 months for multibacillary. 0 if the response is "no"; 1 if the response is "partial"; 2 if the response is "yes"; followed by N/A if not applicable.

Discussion
This systematic review provides an opportune compilation of information on mean time estimates of outcomes of interest for leprosy that will be useful for national programs over the next decade. Considering that the topics discussed in this study seek to aid the strategies to achieve the proposed goal of ending neglected tropical disease epidemics by 2030 according to the third Sustainable Development Goal [41] and the main goal of the Global Leprosy Strategy 2021-2030, which is defined as the interruption of transmission [42].
In this review, 28 articles were analyzed that were published between the years 1994 and the month of January 2022. The study designs were cohorts and clinical trials. The countries with the highest number of studies were Brazil, India, Indonesia, Bangladesh and the Philippines, with the first three countries representing the highest incidence of leprosy in the world [6]. The main outcomes identified in the studies were leprosy, relapse, nerve function impairment, leprosy reactions and physical disabilities. The mean times and outcomes are key elements of discussion for the prevention and control of leprosy [4,39,40,43].

Leprosy
The six studies included in this review with leprosy outcomes were mostly from Brazil, with the study population being leprosy contacts [8] and household contacts [36,37]. This was followed by Indonesia [23,24] and India [25], with the study populations being household and neighbor contacts. According to Quilter et al. (2020) [44], individuals with the most contact with the sick would be the most susceptible to infection. While Romanholo et al. (2018) [45] emphasize that household contacts compose the group most likely to get sick. Corroborating with this systematic review, which identified the highest risk of becoming ill in household contacts [24,25] and specifically for contacts of multibacillary patients [24]. According to Teixeira et al. (2020) [46], contacts who lived with multibacillary patients, aged over 50 years and with schooling up to high school were more likely to develop the disease.
However, non-household spaces such as work and school provide intense inter-human contact [47]. Moreover, other forms of contact besides households can be important in the transmission of the disease [43,47]. A multi-centric study carried out in Brazil, India, Indonesia, Myanmar and Nepal, identified a percentage above 50% of patients with leprosy among social and neighborhood contacts [48]. In China, 69% of leprosy cases were nonhousehold contacts [43]. In this review, the highest incidence density was in household and neighbor contacts with 0.298/100 person-years [24] and the lowest was in the group of non-familial contacts with 0.046/100 person-years [25].
The statistical methods used in the six studies were Kaplan-Meier, identified in four articles which also used the log-rank test [23,25,36,37]. In three studies, Cox's proportional hazards regression was used [8,23,24], while in five other studies, the hazard ratio was applied [8,23,24,36,37]. The mean time until the development of leprosy was found in only one article, being 1.91 years for those who did not have the Bacille Calmette-Guerin (BCG) vaccine scar, 1.97 years with only one BCG vaccine scar and 3.00 years with two BCG vaccine scars [36]. In the research by Niitsuma et al. (2021) [49], BCG immunization was found to have a protective effect against illness (RR = 0.52; 95% CI 0.34-0.78).

Relapse
The studies included in this review with an outcome of relapse for leprosy are based on the concept of the World Health Organization (WHO): patients treated regularly using the standardized therapeutic regimen according to operational classification, discharged due to cure, and after a period of time, present once again with clinical signs and symptoms of the infectious disease [50].
In this review, three articles were from India with a study population featuring both multibacillary [19,39] and paucibacillary leprosy [30,39]. Two studies were from Brazil [35,40]. The remaining studies from Colombia [31], the Philippines [20] and the Republic of Zaire (now the Republic of Congo) also researched the same population [16].
In surveys carried out in specialized centers for leprosy, the results showed that the percentage of relapse cases in Northern India was 61% [51]; in Northern Italy in refugees/migrants, it was 35% [52] and in Midwest Brazil, it was 10 % [53] Studies carried out in two Brazilian states revealed that the relapse of the disease is more frequent in multibacillary patients [54,55]. Already under study in Colombia, the virchowian clinical form is specified, which is classified as multibacillary, with a four-times greater probability of the disease occurring compared to other clinical forms [31].
Other studies reaffirm that the greatest risk of developing relapse was in patients with multibacillary leprosy when compared to patients with paucibacillary leprosy, which is justifiable due to the high bacillary load [40,56], ratifying what was found in this systematic review [39,40]. However, in this review, the incidence density of relapse in multibacillary patients ranged from 0.28/100 person-years [20] to 6.74/100 person-years [31].
According to Rajkumar et al. (2021) [39], low relapse rates are indicative of treatment effectiveness. However, factors such as adequate and opportune treatment regimens for paucibacillary and multibacillary patients as well as drug resistance should be considered [39]. In this review, it was investigated when treatment was combined with other medications [16,30] or when a longer use of treatment with 24 doses was adopted [40], or in addition to prolonged treatment (24 doses) associated until the point of smear negativity [19], there is a lower risk of relapse.
The statistical methods used in all eight studies were Kaplan-Meier [16,19,20,30,31,35,39,40]. Of these, five articles used the log-rank test [16,30,31,35,40], three used regression models: Cox's proportional hazards regression [31], negative binomial regression [35] and the Andersen-Gill method [39], and two studies applied the hazard ratio [31,39]. The mean time to relapse was identified in four articles, the longest at 11.6 years [40] and the shortest at 1.6 years in paucibacillary patients undergoing ROM treatment [30]. There are differences among authors regarding the time to relapse, with some considering relapse at any time after the patient is discharged from treatment [31,57,58], and others emphasizing a minimum interval of three years [59], five years [40,60,61] and up to 16 years [55,62]. It is worth noting that WHO [50] considers the time until relapse to be three years for paucibacillary patients and five years for multibacillary patients. Gonçalves et al. (2019) [63] state that relapses greater than 15 years are commonly experienced by those who are continuously exposed to the bacillus. Chagas et al. (2021) [55], ratify this result, as 40% of relapses in their patients were due to household contact. Lobo et al. (1992) [64] advise about their finding of 24% of relapses detected in the first 18 months, as they may be late reversal reactions. In the work of Nascimento et al. (2022) [40], the importance of identified cases of relapses to immediately starting multidrug therapy is highlighted in order to avoid physical disabilities and the proliferation of the disease.

Clinical Manifestations before, during and after Treatment
In this review, in order to carry out the discussion, the outcomes with clinical manifestations that can occur before, during and after multidrug therapy were grouped [7,65]. These manifestations are derived from the bacillary load and the immune response to the etiological agent, Mycobacterium leprae, which is likely to affect between 30 and 50% of leprosy patients [66,67]. The grouping was carried out according to the outcomes: leprosy reactions and nerve function impairment that are interconnected and influence the final event, physical disabilities [7,68].
Among the findings, 14 articles presented one or more of these outcomes: five of leprosy reactions, seven nerve function impairments and five of physical disabilities. Most were from Brazil, the study population being patients with paucibacillary [26], multibacillary [33][34][35] leprosy, leprosy reaction during and after multidrug therapy [38] and cured [7]. Four articles in Bangladesh evaluated the general population with leprosy [17,18,21,28], while a study carried out in Bangladesh and Nepal was performed on multibacillary pa-tients [22]. This last study's population was the same investigated in India [29,32]. In the Republic of Zaire (now the Republic of Congo), a study was carried out with paucibacillary patients [16].
In studies carried out in specialized leprosy centers in Brazil, it was stated that leprosy reactions occur in approximately 10 to 50% of multibacillary patients [69,70]. On the other hand, this systematic review identified greater vulnerability to leprosy reactions in paucibacillary patients [38]. In a multi-centric research carried out at reference centers in India, Nepal, Bangladesh and Indonesia, the nerve function impairment occurred mainly in multibacillary patients with reactions [71]. In this review, it was found that multibacillary patients are at greater risk of having nerve function impairment [17,18,21,28] and physical disabilities are more favorable to multibacillary patients who have developed disability grade 2 [34]. Worldwide, the number of new cases of leprosy with grade 2 disability represents 7214; with a proportion of new cases with grade 2 disability of 5.66% [6].
Santos et al. (2020) [7] state that the risk of the progression of physical disabilities is associated with leprosy reactions, especially those reported during multidrug therapy. A Brazilian study identified that leprosy reactions are more frequent during multidrug therapy; however, approximately a third occurred after treatment [38]. It was also identified in other studies that leprosy reactions occur in about 15% to 23% of cases after multidrug therapy [72,73]. As for nerve function impairment, the risk of developing this clinical manifestation can reach up to 65% within two years after the start of treatment [74]. Other researchers observed that patients treated and cured after 10 and 15 years had a 30% and 35% probability of progressing to physical disability, respectively [7,75]. In this review, the highest incidence density was of multibacillary leprosy patients with physical disability outcome post multidrug therapy with 6.5/100 person-years [34].
The mean time until the onset of physical disabilities was shorter in multibacillary patients, averaging 4.28 years [32] and longer in paucibacillary patients, averaging 13.5 years [7]. This occurs because the bacillary load of multibacillary leprosy is higher [3]. However, in a cohort of multibacillary patients, there was a 40% progression of physical disability in patients after 10 years of completion of multidrug therapy [34]. Research in India [76], Nigeria [77] and Brazil [5,78] indicates the development of physical disabilities after the end of treatment.
Early detection and treatment are priority strategies for reducing physical disabilities [5,39,79], which can be irreversible and cause emotional, social and economic damage to patients [5,79]. Preventing these disabilities is one of the goals of the Global Leprosy Strategy 2021-2030 [42]. In this way, the implementation of surveillance of physical disabilities with the systematic follow-up of patients during and after completion of treatment is suggested [38,80].
In order to support the development of adequate surveillance in leprosy patients, monitoring and evaluating probable cases of relapses and physical disabilities is necessary, especially in multibacillary patients, since they have the highest load of Mycobacterium leprae. Another priority would be to investigate household and non-household contacts encouraging BCG vaccination.
On the other hand, for the evaluation and monitoring of national leprosy programs, it is recommended to include the mean time until the final outcomes in the epidemiological indicators. In addition, insert the incidence density, the use of Kaplan-Meier methods, log-rank test, Cox's proportional hazards regression and the hazard ratio [9,10,81].
In this way, these indicators will enable the evaluation of the leprosy control program in the following aspects: (I) leprosy in contacts: will evaluate the time of services to identify new cases between household and non-household contacts; (II) relapse: it will evaluate the duration of the therapeutic regimen and the effectiveness of the treatment; (III) clinical manifestations before, during and after treatment: it will evaluate the time until the appearance of physical deformities and consequently will assess the activities of opportune and/or early detection of cases in the health services.
These indicators described above aim to contribute to the Global Leprosy Strategy 2021-2030 by reducing the number of new cases and new cases detected with grade-2 disabilities [42]. The (I) leprosy in contacts can be an aggregate indicator in the second strategic pillar "scale up leprosy prevention alongside integrated active case detection", and the (III) clinical manifestations before, during and after treatment can be aggregated in the third strategic pillar "manage leprosy and its complications and prevent new disability". The (II) relapse is not addressed in the overall strategy but is a relevant indicator to evaluate drug therapy. The use of survival analysis resources has been shown to be a valuable tool for monitoring and controlling leprosy.
Regarding the quality of the analyzed articles, most of them (82.10%) are of high and medium quality. Thus, we can assume that the data extracted from them reflects the reality of the study areas. Other systematic reviews about leprosy have also included high-quality studies in their work [49,82,83]. Despite 17.9% of the studies included in our systematic review being evaluated as low-quality, it does not affect the overall standard of this research.

Limitations
Due to the heterogeneity of the study methods and the variation in mean time and incidence density, it was not possible to perform a meta-analysis of the results. As a limitation, articles in English and Portuguese were included, as the disease is found mostly in countries that mainly use these languages.

Conclusions
This is the first review to systematically investigate the mean time taken place until the outcomes: (I) leprosy, (II) relapse and (III) clinical manifestations before, during and after treatment. The mean time for the development of leprosy in those who were exposed was longer for those who had more than one dose of BCG. In the case of relapse, there were discrepancies between the authors regarding the mean time to this outcome; however, short periods must be given attention, as it can be mistaken with late reverse reactions. In clinical manifestations, the mean time was longer for paucibacillary patients compared to multibacillary patients. These last patients were highlighted in this review, and they need to be prioritized, given that the Global Leprosy Strategy 2021-2030 aims to interrupt the transmission of the disease. Therefore, the use of survival analysis will make it possible to evaluate national programs. In this way, it will assist in the decision-making process to face public health problems, affecting the quality of health services provided to patients affected by leprosy.