Performance of the Health System Network in Formosa, Argentina, in the Diagnosis of Leprosy

ABSTRACT.

Leprosy is a chronic, neglected tropical infectious disease, currently endemic in Formosa, a province in northwestern Argentina. To analyze the performance, distribution, and effectiveness of the health system in leprosy diagnosis in Formosa, we estimated the trend of the number of new cases of leprosy diagnosed between 2002 and 2019 and estimated a forecast for 2022 at the primary health care centers (PHCCs) of at the first level of care (1stLC), at district hospitals (DHs) of the second level of care (2ndLC), high-complexity hospitals at the third level of care (3rdLC), and in rural and urban areas. The general trend was calculated based on the new cases detection rate (NCDR) using the autoregressive–moving average model (ARMA). The 1stLC, 2ndLC, and 3rdLC and the rural/urban variables were assessed using a proportional Bayesian trend ARMA (TrARMA) model. A predictive model was used for estimated forecasts. Markov-Monte Carlo chains were applied with A Metropolis-Hastings’s algorithm. The highest median proportion (Mp) of new cases of leprosy was diagnosed at the 2ndLC (Mp, 0.67; 97.5% credibility interval [CI] [0.56–0.77]), at the 3rdLC (Mp, 0.11; 97.5% CI [0.08–0.15]), and in urban areas (urban median proportion (uMp), 0.86; 97.5% CI [0.83– 0.88]), whereas the lowest proportion of new cases was diagnosed at the 1stLC (Mp, 0.082; 97.5% CI [0.061–0.108]) and in rural areas (rural median proportion (rMp), 0.13; 97.5% CI [0.11–0.16]). Our model predicts for 2022 that a median number of new cases of leprosy of 19.70 will be diagnosed in urban areas (97.5% CI [15.94–23.80]), and will continue to be diagnosed at the 2ndLC (median number of cases, 15.33; 97.5% CI [12.40–10.52]) and 3rdLC (median number of cases, 2.43; 97.5% CI [1.97–2.94]).

INTRODUCTION

Leprosy is a chronic infectious disease caused by Mycobacterium leprae that produces pain, physical and sensory disability, and social stigma commonly associated with vulnerable contexts.¹ The global strategy to reduce the burden of leprosy in the world is based on the organization of an integrated health-care service that provides early diagnosis, timely treatment, prevention and care of physical disabilities, monitoring of contacts in the primary health-care level,² and support of specialized professionals in the secondary and tertiary levels of care, based on the disabling potential of the disease.³

In 2016, a total of 27,356 new cases of leprosy were reported in the Americas, of which 295 were from Argentina,⁴ a country where the national prevalence of leprosy was 0.18/10,000 inhabitants and had reached its elimination goals in 2012.⁵ In particular, the province of Formosa, located in the subtropical region of northeastern Argentina, currently reports alarming epidemiological indicators that need to be investigated and clarified.⁶

Since 1957, Formosa has recognized health as a fundamental right for its population, including indigenous communities with their ancestral cultural patterns. The elimination of leprosy is a current challenge for the public health system of the province. Accordingly, Formosa’s Health System Network is organized by care levels of increasing complexity: assuming the strategy of primary and/or first level of care (1stLC)⁷ is represented by primary health-care centers (PHCCs). PHCCs distributed throughout the entire province are the gateway for patients to the health system and are responsible for managing the most frequent and relevant low-risk pathologies.⁷ The second level of care (2ndLC) is represented by Sanitary Districts (SDs), which are the health territorial divisions of the province in 12 areas that perfectly match the radii and census segments, with common sanitary needs.⁸ The importance of SDs lies in the fact that each one has at least one district hospital (DH) that attends to medium-risk pathologies, and they are the referral centers of PHCCs.^8,9 Hospital Central, Hospital de la Madre y el Niño, and Hospital de Alta Complejidad belong to the tertiary level of care (3rdLC); they are referral hospitals for complex pathologies. The Program for the Control of Leprosy in Formosa (PCL-F)¹⁰ was integrated into Formosa’s Health System Network in 1999 for taking direct action to control leprosy. However, in 2016, patients with grade 2 disabilities and the occurrence of leprosy in children younger than 15 years of age were reported, revealing a clear delay in the diagnosis of the disease.⁴

Previous studies from our research group have already confirmed that leprosy is still expanding throughout Formosa and will not be eliminated in the short term.¹¹ It is possible that the urbanization process that took place in the province during the past decade and that caused the depopulation of rural areas in favor of urban ones greatly influenced the use of health-care services in general and of primary health-care services in particular.¹² On the other hand, it has also been noted that some patients seek care in major hospitals far from where they reside.¹³

In this study, we hypothesized that the low detection rate of new cases of leprosy in PHCCs at the 1stLC may be the result of the bypassing PHCCs and towards DHs (2ndLC) as a result of an internal migration pattern. Given the fact that there is a clear need to gain ground on new and more effective ways to eliminate leprosy in Formosa, we aim use our results to provide recommendations and innovative strategies for moving forward.

MATERIALS AND METHODS

Study designs and data source.

This is a descriptive observational study with a predictive modeling component. Data from 811 patients diagnosed with leprosy between January 1, 2002 and December 31, 2019 were obtained from the medical records, which are centralized in the PCL-F. Informed verbal consent was obtained from adult patients and parents or legal guardians of child participants. The Research Ethics Committee of the Centro Nacional de Genética Médica (CENAGEM) de la Administración Nacional de Laboratorios e Institutos de Salud (ANLIS), Dr. “Carlos G. Malbrán” dependent on Ministerio de Salud de Argentina, certified that patient data were stored safely and securely to ensure patient confidentiality and that the study raised no ethical concerns (Comité de Ética en Investigación [CEI]-Arnaiz, September 19, 2017).

First, we explored our database to identify the number of new cases of leprosy diagnosed in PHCCs, sub-classified according to the care level of increasing complexity as PHCC-1: basic health center; PHCC-2: outpatient care only; PHCC-A3R: births and temporary hospitalization with no auxiliary diagnostic or treatment services; and PHCC-4: outpatient and hospitalization (Figure 1).

Figure 1.

Map showing the location of primary health-care centers (PHCCs) in the province of Formosa corresponding to the first level of care (1stLC). The location coordinates of each PHCC were obtained by geocoding the PHCC addresses listed on the official database from the Ministry of Health and from the Google Maps platform, when unavailable from official sources. Some PHCCs may not show an accurate location, but an approximate one, because of the lack of updated reliable information available from official sources. This map and its data is available at: www.google.com/maps/d/u/0/viewer?mid=1mmICTwklAydDTETuv1mqbgT3w89KXEB6&ll=-24.681567699457414%2C-59.95390449999999&z=7.

Second, we identified the number of new cases of leprosy diagnosed in the DHs of each of the 12 SDs (I–XII) into which the province is divided: SD I (Engineer Juárez), with 34 PHCCs and two DHs; SD II (Las Lomitas), with 43 PHCCs and three DHs; SD III (Ibarreta), with 19 PHCCs and five DHs; SD IV (Laguna Blanca), with 29 PHCCs and seven DHs; SD V (El Colorado), with 13 PHCCs and three DHs; SD VI (Pirané), with 17 PHCCs and two DHs; SD VII (Misión Laishi), with 20 PHCCs and one DH; SD VIII (Formosa), with nine PHCCs and one DH; SD IX (Formosa), with seven PHCCs and one DH; SD X (Formosa), with seven PHCCs; and SD XII (Clorinda), with nine PHCCs and one DH. SD XI (Formosa) has three high-complexity hospitals and provides the 3rdLC¹⁴ (Figure 2).

Figure 2.

Map showing the territorial health division of the province of Formosa in Sanitary Districts (SDs) at the second level of care (2ndLC), adapted from Formosa’s Ministry of Human Development (2011). This map was created using QGIS software. The original map was digitalized and overlapped the 2010 National Census radio polygon shape to accurately delimit the nine SDs: SD I, Engineer Juárez; SD II, Las Lomitas; SD III, Ibarreta; SD IV, Laguna Blanca; SD V, El Colorado; SD VI, Pirané; SD VII, Misión Laishi; and SD XII, Clorinda. SDs VIII, IX, X, and XI correspond to the capital city of Formosa. This map and its data is available at: www.google.com/maps/d/u/0/viewer?mid=1mmICTwklAydDTETuv1mqbgT3w89KXEB6&ll=-24.681567699457414%2C-59.95390449999999&z=7.

Overall, Formosa’s health system has 207 PHCCs providing the 1stLC, 23 DHs distributed in 11 SDs providing the 2ndLC, and three high-complexity hospitals in SD XI providing the 3rdLC.

Third, we identified the number of new cases of leprosy that were diagnosed in the PHCCs of the rural areas and in the DHs of urban areas. A rural or urban area was defined based on the results of the latest national census (2010) and official estimates (https://www.citypopulatio.de/php/argentina-formosa_s.php).¹⁵

Statistical analysis.

The collected variables were analyzed via classic time-series methods¹⁶ using a fully Bayesian approach.¹⁷ Because we have a matrix containing a time series with a trend and autocorrelated data, instead of calculating the descriptive statistics directly from raw data, we estimated the basic statistical parameters by using the inference made from the statistical models, free of trend and autocorrelation effects.

Trend analysis and forecasts.

To detect the general trend of the number of new cases of leprosy diagnosed, we studied the total number of new cases in the province, and then we carried out the same analysis considering the diagnosis of new cases in PHCCs at the 1stLC and in DHs of the 2ndLC to determine whether it was a general pattern or was restricted to one of the categories mentioned earlier. The trend of the number of new cases of leprosy, both in general and for each of the previously mentioned variables, was studied using a Bayesian approach.

General trend model.

As a general trend model, we propose an exponential decay trend, where the average number of new cases of leprosy decreases exponentially to zero with a λ rate, so the number of new cases of leprosy is

$$X_t=X_{t_0}{\lambda }^{t-t_0}$$ (1)

where X_t is the estimated number of new cases of leprosy as a function of time and t₀ = 2002, in other words, the starting point of this analysis. To correct for the lack of independence between successive data, given that it is a time series, an autoregressive–moving average (ARMA) model (pq)¹⁶ was added to the basic model. Because the data series is very short, orders of the autoregressive (p) and moving average (q) greater than one were not used. Last, the fit of the model to the data was evaluated with a Poisson likelihood function.

Therefore, the final model used to estimate the number of new cases of leprosy by year was

$$X_t=\text{λ}{X}_{t-1}+\varphi \left(N_{t-1}-X_{t-1}\right)+\theta {\epsilon }_{t-1}+{\epsilon }_t,$$ (2)

where X is the number of new cases of leprosy estimated per year t, λ is the rate of exponential decay, φ is the coefficient of autoregression (AR), N is the number of new cases of leprosy observed, θ is the coefficient of the moving average (MA), and ε is the error, which follows a Poisson distribution $\text{ε}\sim \text{Poisson}\left(X_t\right).$ The last three terms make up the ARMA model, from now on.

The final model is defined as follows:

$$X_t=X_{t_0}{\lambda }^{t-t_0}+\text{ARMA}\left(p,q\right)+\epsilon \sim \text{Poisson}\left(X_t\right).$$ (3)

In its complete form, the model consists of four parameters to be estimated by series: origin (t₀ = 2002), rate (λ), AR term (φ), and MA term (θ). This model is called trend ARMA (TrARMA) from now on.

Trend model by category.

To test whether the number of new cases of leprosy showed different types of trends according to the different types of health-care categorizations, the following models were proposed. The first was a no-trend model only with the ARMA term of Equation (3). Here, the variation of cases is assumed to be a purely stochastic autocorrelated process.

The second model proposed was one in which the trend was the same general trend as estimated previously, but with a proportional distribution of cases that was constant over time:

$$X_{ti}=p_i{X}_t+\text{ARMA}\left(p,q\right)+\epsilon \sim \text{Poisson}\left(X_{ti}\right),$$ (4)

where X_ti is the number of new cases of leprosy estimated for category i at time t, p_i is the proportion of the total number of new cases of leprosy categorized as i over the total number of cases recorded in our study, and X_t is the total number of new cases of leprosy estimated according to Equation (3). In this model, the trend and the autocorrelation are shared across the categories and the only variation is the error term. This model is called proportion/trARMA (PrTrARMA) from now on.

Third, the TrARMA model was extended so that each parameter could be global (shared across the categories) or category specific:

$$X_{ti}=X_{t_{0}i}{\lambda }_i{}^{t-t_0}+\text{ARMA}\left(p_i,q_i\right)+\epsilon \sim \text{Poisson}\left(X_{ti}\right),$$ (5)

where subscript i defines the parameter as being specific to a category similar to that of Equation (3) proposed for each variable. Here, each variable had its own trend and autocorrelation structure.

Bayesian process of model selection.

In all cases, we began by proposing a null model (only with the mean, without trend or autocorrelation) and then added the parameters one by one until the explanatory value began to increase or reached the complete ARMA model, as stated in Equation (3). The explanatory value was measured as the Bayesian information criterion,^17,18 which is an index that considers the goodness of fit, measured as the logarithm of the likelihood function, the number of parameters, and the amount of data. This procedure was applied, as described here, for a series with the sum of all cases per year, to obtain a global trend.

To categorize patients at PHCCs, DHs, and rural/urban areas, we first proposed the PrTrARMA model of Equation (4). Therefore, the model selection procedure started with the PrTrARMA model, and then proceeded to the model of Equation (5), adding the parameters one by one, as in the global trend model. This procedure was repeated for each of the listed categories.

After selecting and fitting, the model was validated by comparing the forecasted values for 2017, 2018, and 2019 with the observed values of new cases of leprosy that resulted within the credibility interval of the prediction. The data from 2017 (for urban areas) was outside the predicted credibility interval, being much higher than in previous years and as high as the data from 13 years before, so it is possibly an outlier; therefore, we considered that the model was valid for estimating forecasts.

For all the models, Markov-Monte Carlo chains were used with the Metropolis-Hastings algorithm¹⁷ to calculate the subsequent distributions of parameters. For this, 100,000 iterations of the algorithm were needed, of which the first half were discarded, resulting in a total of 50,000 iterations containing these distributions a posteriori. Noninformative a priori distributions were used in all the parameters. These calculations were performed using the 2.4 version pymc module of Monte Carlo methods for the python programming language.¹⁹

RESULTS

Trend of new cases of leprosy diagnosed from 2002 to 2019 by the Health System Network.

The model selection procedure selected a global trend model with an MA model (ARMA [0.1] for autocorrelation structure) (Supplemental Table S1). The global trend analyses of 811 new cases of leprosy diagnosed from 2002 to 2019 evidenced a decreasing trend with an origin of 66.701 (62.17–71.34) and a slope of 0.948 (0.937– 0.959), which is an annual decrease of 5.2%, and an autocorrelation MA of –0.2302 (–1.3308 to 0.2176).

The general trend evidenced that the greatest proportion of new cases of leprosy was diagnosed in SDs of the 2ndLC (median proportion (Mp), 0.67; 97.5% CI [0.56–0.77]) compared to the PHCCs at the 1stLC (Mp, 0.082; 97.5% CI [0.061–0.108]) (Table 1).

Table 1.

Proportion of new cases of leprosy diagnosed at the 1stLC and 2ndLC in Formosa from 2002 to 2019

Level of care	CI 2.5%	Median estimate	CI 97.5%
1stLC
PHCC-1	0.0223	0.0347	0.0515
PHCC-2	0.0437	0.0616	0.0843
PHCC-A3R	0.0588	0.0799	0.1062
PHCC-4	0.1204	0.1532	0.1920
2ndLC, SDs I–XII	0.5641	0.6705	0.7793

1stLC = first level of care; 2ndLC = second level of care; PHCC = primary health-care center; SDs = sanitary districts. The table shows the estimated median and the credibility interval (CI) of the proportion of new cases of leprosy diagnosed at the 1stLC and 2ndLCs of care. The 1stLC treats low-risk pathologies in PHCCs, sub-classified according to the complexity level of attention as: PHCC-1: basic health center (nurse and health agent); PHCC-2: outpatient care only (general practitioners, gynecologists, dentists, pediatricians, health agents, kinesiologists, and social workers); PHCC-A3R: childbirth and hospitalization services are added, (no auxiliary diagnostic or treatment services); PHCC-4: outpatient and hospitalization (low-complexity surgery, low-risk hospitalization, psychology, and social assistance are added). The 2ndLC corresponds to the organization of SDs, which treat middle-risk pathologies (outpatient and inpatient).

Considering SDs in particular, the greatest proportion of new cases of leprosy were diagnosed in a small district in the southeastern region of the province: SD V (El Colorado; Mp: 0.15: 97.5% CI: [0.12–0.19]), followed by SD VIII (Mp: 0.11; 97,5% CI: [0.08–0.14]) of the 2ndLC and SD XI (Mp: 0.11; 97,5% CI: [0.08–0.15]) of the 3rdLC, the two last located in the capital city of Formosa (Table 2, Figure 2).

Table 2.

Proportion of new cases of leprosy diagnosed at the 2ndLC in Formosa from 2002 to 2019

SDs I–XII	SD name	CI 2.5%	Median estimate	CI 97.5%
I	Engineer Juárez	0.0656	0.0899	0.1208
II	Las Lomitas	0.0559	0.0778	0.1058
III	Ibarreta	0.0557	0.0777	0.1061
IV	Laguna Blanca	0.0272	0.0416	0.0612
V	El Colorado	0.1207	0.1566	0.1999
VI	Pirane	0.0523	0.0734	0.1001
VII	Misión Laishi	0.1006	0.1320	0.1705
VIII	Formosa	0.0860	0.1139	0.1497
IX	Formosa	0.0647	0.0884	0.1184
X	Formosa	0.0180	0.0296	0.0463
XI	Formosa	0.0868	0.1155	0.1508*
XII	Clorinda	0.0180	0.0297	0.0463

2ndLC = second level of care; 3rdLC = third level of care; SDs = sanitary districts. The table shows the median and the credibility interval (CI) of the proportion of new cases of leprosy diagnosed from 2002 to 2019 at the district hospitals of SDs l–Xll. This territorial division matches perfectly the radial census segments, so the disaggregated population of the censuses enables us to assign the corresponding population under the responsibility of each area.

^*SD XI belongs to the capital city of Formosa but corresponds to the 3rdLC because it has three high-complexity hospitals.

Regarding the 1stLC, most new cases of leprosy were diagnosed in PHCC-4 (Mp, 0.15; 97.5% CI [0.12–0.19]), which are the most complex primary care centers, whereas the lowest proportion of new cases was diagnosed in PHCC-1 (Mp, 0.03; 97.5% CI [0.02–0.05]), the least complex primary care centers (Table 1).

Trend of new cases of leprosy diagnosed from 2002 to 2019 in the entire province, including rural/urban areas.

For the rural/urban categorization, a TrARMA proportional model showed a decreasing trend in the proportion of new cases of leprosy diagnosed in both: the rural area (rural median proportion [rMp], 0.1357; 97.5% CI [0.1116–0.1625]) (Figure 3) and the urban area (urban median proportion [uMp], 0.8643; 97,5% CI [0.8375–0.8884]) (Figure 4).

Figure 3.

Number of new cases of leprosy in rural areas as a function of time. The circles indicate the number of new cases of leprosy observed, the solid line is the median trend calculated using the trend ARMA (TrARMA) model, and the dashed lines indicate the 97.5% credibility interval (CI). The values to the right of the solid vertical line are the forecasted values for 2022.

Figure 4.

Number of new cases of leprosy in urban areas as a function of time. The circles indicate the number of new cases of leprosy observed, the solid line is the median trend calculated using the trend ARMA (TrARMA) model, and the dashed lines indicate the 97.5% credibility interval (CI). The values to the right of the solid vertical line are the forecasted values for 2022.

Number of new cases of leprosy forecasted for 2022.

We forecasted the number of new cases of leprosy for 2022 at two levels: 1) the Health System Network and the 2) entire province, including rural/urban areas. The estimated forecasts in the Health System Network indicate that in the year 2022, leprosy will be diagnosed mainly in the DHs of SDs that correspond to the 2ndLC, with 15.33 new cases of leprosy (97.5% CI [12.40–18.52]), and in PHCC-4, with 3.49 new cases of leprosy (97.5% CI [2.83–4.22]) of the 1stLC (Table 3).

Table 3.

Predicted values of the number of new cases of leprosy that will be diagnosed at the 1stLC and 2ndLC in 2022

Level of care	CI 2.5%	Median estimate	CI 97.5%
1stLC
PHCC-1	0.44	0.55	0.66
PHCC-2	1.13	1.39	1.68
PHCC-A3R	1.47	1.81	2.19
PHCC-4	2.83	3.49	4.22
2ndLC, SDs I–XII	12.40	15.33	10.52

1stLC = first level of care; 2ndLC = second level of care; PHCC = primary health-care center; SDs = sanitary districts. The table shows the median and credibility interval (CI) of the estimated values of the number of new cases of leprosy that will be diagnosed in 2022 at different levels of complexity of the 1stLC and 2ndLC from Formosa’s health system.

Figure 5 show the population imbalance of the SDs in Formosa and is correlated to the forecasts that indicate that in SDs V (El Colorado) and VII (Misión Laishi), which correspond to the 2ndLC, and SD XI, which corresponds to the 3rdLC, the greatest number of new cases of leprosy will be diagnosed (3.30, 97.5% CI [2.67–3.99]; 2.79 97.5% CI [2.25–3.37]; and 2.43 97.5% CI [1.97–2.94]), respectively (Table 4).

Figure 5.

Map showing a range of the number of inhabitants per Sanitary District (SDs) in Formosa. This map was created using QGIS software. The population data were obtained from the 2010 National Census at a census radio polygon scale. These data are used only in this map to portray how the migration process has had a direct influence on the population size of each SD and to show how the population is distributed with respect to the health system, which also indicates that SDs with more inhabitants present the greatest number of newly diagnosed cases of leprosy. This map and all of its data is available at: www.google.com/maps/d/u/0/viewer?mid=1mmICTwklAydDTETuv1mqbgT3w89KXEB6&ll=-24.681567699457414%2C-59.95390449999999&z=7. PHCC = primary health-care centers.

Table 4.

Predicted values of the number of new cases of leprosy that will be diagnosed at the 2ndLC in 2022

SDs I–XII	SD name	CI 2.5%	Median estimate	CI 97.5%
I	Engineer Júarez	1.53	1.89	2.28
II	Las Lomitas	1.32	1.63	1.97
III	Ibarreta	1.32	1.63	1.97
IV	Laguna Blanca	0.70	0.86	1.04
V	El Colorado	2.67	3.30	3.99
VI	Pirané	1.24	1.54	1.86
VII	Misión Laishi	2.25	2.79	3.37
VIII	Formosa	1.94	2.40	2.90
IX	Formosa	1.50	1.86	2.24
X	Formosa	0.49	0.61	0.70
XI	Formosa	1.97	2.43	2.94
XII	Clorinda	0.49	0.61	0.74

2ndLC = second level of care; SDs = sanitary districts. The table shows the median and credibility interval (CI) of the estimated values of the number of new cases of leprosy that will be diagnosed in each one of the SDs from the province of Formosa in 2022.

Finally, the forecast estimates that, in 2022, 2.58 new cases of leprosy (97.5% CI [18.27–27.28]) will be diagnosed in Formosa, with a net predominance in urban areas (19.70 new cases; 97.5% CI [15.94–23.80]) with respect to rural areas (2.88 new cases; 97.5% CI [2.33–3.48]) (data not shown).

DISCUSSION

The primary result of this study proves that, although leprosy in Formosa evidences a sustained decreasing trend in leprosy diagnosis, the disease will not be eliminated by 2022 and it will be diagnosed mainly at the 2ndLC and 3rdLC instead of the 1stLC PHCCs of the Health System Network.

Several studies affirm that to control leprosy better and prevent its complications, the diagnosis of new cases should be carried out at the 1stLC.^2,3 Formosa’s Health System Network is structured according to levels of increasing complexity,⁸ with PHCCs providing the 1stLC. Although the PCL-F was integrated into the general health system and works systematically in actions to control leprosy, the results are not expected.

Several aspects inherent to PHCCs at the 1stLC can be considered when explaining our findings. The effectiveness of the actions to control leprosy requires that PHCCs at the 1stLC provide accessibility (first-contact access), integrality (services suitable for the health needs of the population), family and community orientation, and cultural competence, among other aspects.^20,21

Although in Formosa accessibility to PHCCs at the 1stLC can sometimes be affected by inclement weather (considering the sub-tropical climate with a rainy season), the PCL-F plans monthly epidemiological surveillance actions to search actively for new cases of leprosy and to follow up with contacts throughout the province.

On the other hand, the lack of professional training in the recognition of the signs and symptoms of leprosy at the most basic care levels^22,23 has proved to have a direct impact on the early diagnosis of new cases and on the integrality of the examination offered. Our results show that the diagnosis of new cases of leprosy increases with the level of complexity of the PHCCs. This is evidenced by the fact that PHCC-1 offers only a nurse and a health agent whereas PHCC-4 has a general practitioner who provides continuous and comprehensive medical care of the individual and family, low-complexity surgery, and low-risk hospitalization.⁸ This solidifies our argument that to improve the performance of the 1stLC in the diagnosis of new cases of leprosy, the PCL-F should prioritize training campaigns for PHCC staff in clinical, laboratory, and epidemiology of leprosy detection sustainably over time to produce a lasting improvement in an early diagnosis rate.²⁴

Another important recommendation is to make available family and community orientation at the PHCCs, so that information on how to carry out self-diagnosis, detect the disease, and arrange for a consultation is accessible at all times.²³ It is also suggested that education and awareness campaigns on leprosy be carried out using standard materials (brochures with PHCC addresses, days and hours of attention), radio broadcasts, and innovative social networks (@Instagram and Facebook).²

We wonder why most new cases of leprosy were and will be diagnosed mainly in the 2ndLC and 3rdLC. The first explanation could be related to the fact that patients’ perceptions of the quality of medical services in DHs and high-complexity hospitals that correspond to the 2ndLC and 3rdLC are more reliable than the services offered by PHCCs at the 1stLC.¹² The DHs have basic services that include dermatology services and complementary diagnostic services for care of middle-risk pathologies and are the referral of the PHCCs, whereas hospitals at the 3rdLC have high-quality medical services with specialized equipment, attend high-risk pathologies, and are referral hospitals of the DHs. The second explanation could be the urbanization process of Formosa that began in 1947 and continues to this day, during which there was an imbalance in the population of the eastern SDs with respect to the central and western regions, bypassing the use of rural primary health-care services^12,25 and consolidating the capital city of Formosa as a services provider. Evidence of this is that the best performance of the Health System Network in the diagnosis of new cases of leprosy was and will continue to be in urban areas in general and in SDs of the eastern region of the province in particular. SD XI provides medical coverage to 70% of the population and corresponds to the 3rdLC, indicating that the complexity level of the health system directly influences the diagnosis of new cases of leprosy.

Our study suffers from the limitations imposed by the use of a secondary database not designed by us. It would have been useful to have detailed additional information (e.g., exact address of the patients, the PHCCs and the district hospitals where the diagnosis was made, date and directions of social mobility [migration], whether it was temporary or permanent, and social variables) to include in our analyses. Moreover, case underreporting may result from the difficult access to health-care facilities in rural areas.

Our results show the health system must redirect the early diagnosis of leprosy to the PHCCs, because human and economic resources are used at the 2ndLC and 3rdLC to diagnose a condition that should have been detected at the 1stLC. Second, considering the social and economic differences between rural and urban areas, strategies consistent with local priorities must be implemented.²⁶

We hope our work contributes to the elimination of leprosy in the province of Formosa and worldwide.

Note: Supplemental table appears at www.ajtmh.org.