Assessment of integration of the Leprosy Program into Primary health care in Aracaju , State of Sergipe , Brazil

Introduction: The aim of this study was to assess the epidemiological and operational characteristics of the Leprosy Program before and after its integration into the Primary healthcare Services of the municipality of Aracaju-Sergipe, Brazil. Methods: Data were drawn from the national database. The study periods were divided into preintegration (1996-2000) and postintegration (2001-2007). Annual rates of epidemiological detection were calculated. Frequency data on clinico-epidemiological variables of cases detected and treated for the two periods were compared using the Chi-squared (χ2) test adopting a 5% level of significance. Results: Rates of detection overall, and in subjects younger than 15 years, were greater for the postintegration period and were higher than rates recorded for Brazil as a whole during the same periods. A total of 780 and 1,469 cases were registered during the preintegration and postintegration periods, respectively. Observations for the postintegration period were as follows: I) a higher proportion of cases with disability grade assessed at diagnosis, with increase of 60.9% to 78.8% (p < 0.001), and at end of treatment, from 41.4% to 44.4% (p < 0.023); II) an increase in proportion of cases detected by contact examination, from 2.1% to 4.1% (p < 0.001); and III) a lower level of treatment default with a decrease from 5.64 to 3.35 (p < 0.008). Only 34% of cases registered from 2001 to 2007 were examined. Conclusions: The shift observed in rates of detection overall, and in subjects younger than 15 years, during the postintegration period indicate an increased level of health care access. The fall in number of patients abandoning treatment indicates greater adherence to treatment. however, previous shortcomings in key actions, pivotal to attaining the outcomes and impact envisaged for the program, persisted in the postintegration period.

In Brazil, actions for diagnosing and treating leprosy have historically been conducted at specialized state-run units.Since 1994, Brazil's Ministry of health has devised and promoted the Family health Strategy (FhS) as a priority model for organizing primary health care services.In response to initiatives by the World health Organization (WhO) recommending integration of leprosy control into primary health care (PhC), Brazil has established the integration of leprosy control actions (LCA) into primary health care as a key directive toward achieving the goal of eliminating leprosy 1 .Thus, the main strategy proposed for implementing leprosy actions was training family health care teams (FhCT), through training courses and distribution of technical guides.
The FhS was implemented in Acaraju in 1998 2 .Integration of the LCA began only in 2000 following training of FhCT 3 .In 2007, the municipality had 128 FhCT, distributed among 43 family health units, providing coverage for 86.7% of the population 2 .
Traditionally, appraisals of the performance of the Leprosy Control Program (LCP) have been based on a set of indicators defined by the National LCP, compiled from information contained in notification/investigation records held on the Brazilian Disease Notification System (Sistema de Informação de Agravos de Notificação -SINAN), a system developed in the 1990s as a tool for collecting, processing, storing, and analyzing data on mandatory notifiable diseases in Brazil 4 .These indicators are categorized into two groups denoted by the Ministry of health as epidemiologic and operational 5 .however, the use of these indicators for assessment purposes has several drawbacks.One of these weaknesses is the lack of robustness of the information sourced from data collected at local and municipal levels.Many municipalities lack adequate systems to oversee data entry into the system or to monitor the information held.Nevertheless, for states and municipalities in which

RESULTS
Raposo MT et al -Integration of the Leprosy Program into primary health care the system is well managed, the assessment of information from the system enables valuable analyses, which help managers assess and monitor local actions 6 .
In this regard, the use of data from SINAN for individual municipalities and locales is potentially positive.Reports have described highly heterogeneous situations among different municipalities from within the same area or even across services offered by the same municipality.Greater proximity of the action agents, namely, health care managers and professionals, with these data increases the likelihood of the information leading to improvements.
The aim of the present study was to analyze the epidemiological indicators and the clinico-epidemiological characteristics of leprosy during the preintegration and postintegration periods of LCA at family health units in the municipality of Aracaju.This descriptive study involved analysis spanning 12 years (Jan 1996 to Dec 2007) and was restricted to diagnosed cases of leprosy in patients residing in the municipality of Aracaju in Sergipe, Brazil.Data were sourced from the SINAN through the Aracaju Municipal Secretariat for health and the State Secretariat for health of Sergipe.In a bid to increase the reliability of the data and reduce the risk of under-registration, the study used data collected in 2009 on the years 1996 to 2007.
Aracaju, the capital of State of Sergipe, covers an area of 174km 2 and had an estimated population of 544,035 inhabitants in 2009.The city accounts for the highest number of leprosy cases in the state and had an overall detection rate of 49.22 cases for every 100,000 inhabitants in 2007 7 .The integration of the LCP into family health units commenced in 2000 3 , and therefore, the analysis was split into two study periods: the 1996 to 2000 period corresponded to the phase prior to the integration of the LCA and was denoted the preintegration period, and the period spanning from 2001 to 2007 represented the postintegration period.
The database was organized to ensure exclusion of duplicate and inconsistent records.Individual compulsory notification records and/or clinical charts were analyzed in the remaining cases of inconsistencies.To standardize the records for current classification, cases with a disability grade 3, registered up to 2001, were converted to grade 2 disability, in accordance with the instrument standardized by the WhO, revised in 1998, and adopted in Brazil.
Given that leprosy detection rates fluctuate over longer periods, to assess the endemic picture for the municipality, a descriptive analysis of the indicators recommended for monitoring and assessing the epidemic, compiled on an annual basis encompassing aspects of morbidity, magnitude, and epidemiological profile, was carried out 5 .
The following indicators were included: annual rate of detection of new cases per 100,000 inhabitants, annual rate of detection of new cases in the population aged between 0 and 14 years per 100,000 inhabitants 5 , and proportion of disability grade 2 in new cases per 1 million inhabitants at time of diagnosis 8 .Because of the inconsistency of some of the variables held on the database, the following indicators were excluded: proportion of leprosy cases with grade 2 physical disability on diagnosis among new cases detected and assessed on an annual basis and proportion of leprosy cases with grade 2 physical disability among cases assessed at conclusion of multidrug therapy (MDT) on an annual basis.It was decided not to use the rate of annual prevalence of leprosy per 10,000 inhabitants because of the change in the criteria used to calculate this figure during the study period and also owing to the intrinsic weakness in the prevalence indicator currently in use.
From the year of initial diagnosis, the clinico-epidemiological variables were taken (gender, age-group, operational classification, disability grading (DG) rating at diagnosis, DG rating at end of treatment (MDT), detection method, contacts examined, and case outcome) as a proxy representing the quality of the service.The evolution of these variables over time was analyzed for the two time frames studied.The first period was the preintegration stage, characterized by a health care service centralized at a single health unit, whereas the second encompassed the postintegration period, during which, the LCA were integrated into PhC.
Data were stored on the EpiInfo™ software program, version 3.3, from the Centers for Disease Control and Prevention (CDC), October 2004.Frequency distributions were compared using the Chi-squared (χ 2 ) test adopting a 5% level of significance (p < 0.05).

Ethical considerations
The project was previously approved by the Research Ethics Committee of the Clinics hospital, University of São Paulo Medical School (n o 1032/08) and by the Research Ethics Committee of Tiradentes University (n o 030108).

Detection rates
From 1996 to 2007, a total of 2,249 cases of leprosy were detected in Aracaju, the municipality with the highest number of cases of the disease in Sergipe.Table 1 shows data obtained by this study set against official figures on the epidemic in Brazil 9 .Shifts over the period point to an increase in overall detection during the years subsequent to integration of the LCP in PhC in 2000.A peak in detection was found in 2003, reaching a rate of over 50/100,000 inhabitants, a level maintained for the two proceeding years.
The raw rates of detection in individuals between 0 and 14 years of age showed some variations, predominantly during the period following integration, peaking both in 2002 and 2005 when the detection rate reached its highest point of more than double the rate for the previous year.The trend in the detection rate in children revealed a pattern of a slight rise, where rates for Sergipe consistently exceeded those of Brazil as a whole for almost all years studied 9 .
Taken together, the overall detection rate between 2003 and 2006, plus the detection rate among individuals younger than 15 years, classify the municipality as hyper-endemic.During the other years studied, the pattern of the epidemic was classified as very high 5 .The shift in the epidemiological picture for the disease coincided with the rising number of FhCTs incorporated into PhC and with the intensifying of the process of implementing the FhS in the municipality.
The proportion of new cases rated as DG2 at diagnosis (per million inhabitants) was calculated only for years in which the DG percentage assessed at diagnosis exceeded 75%.For the years 2001,  2002, 2003, 2005, and 2006, the proportion of cases (per million inhabitants) showed a stable pattern with rates of 14.95, 12.66, 27.1, 16.04, and 11.87, respectively.

Clinico-epidemiological characteristics of cases
With regard to gender for the study period as a whole, the cases were predominantly women, but the gender distribution for the two study periods did not differ significantly (p = 0.164).
Of the 2,249 cases included, 2,077 (92.4%) were older than fifteen years when first diagnosed.Comparing the two study periods, the proportion of leprosy patients aged between 0 and 14 years was 8% and 7.32% for preintegration and postintegration periods, respectively, a difference not reaching statistical significance (p = 0.696).A total of 351/780 (45%) cases in the age bracket of 20 to 44 years were diagnosed during the preintegration period versus 707/1,469 (48.3%) during the postintegration period.
The majority of cases detected were of the paucibacillary (PB) type for both study periods.Although an increase in multi-bacillary (MB) cases was seen (from 36.2% to 40%), this difference was not statistically significant (p = 0.073).
Table 2 shows the epidemiological variables by period of LCA integration in the municipality.
The proportion of new cases with DG rated at diagnosis was 60.9% and 78.1% for preintegration and postintegration periods, respectively.A decrease in the percentage of cases in which DG was not rated was found for the postintegration period (p < 0.001).
Analysis of DG assessment at end of treatment revealed that DG was not rated in 58.6% and 55.7% of cases during preintegration an d posti nte g rati on p er i o d s , respectively, where the reduction in cases without registered disability reached statistical significance for the postintegration period (p < 0.023).
In terms of detection method, a significant association (p < 0.001) was found with probable increased detection by contact examination, following the integration of LCA to PhC.Entry to the system as relapsing was included in this study under the category of other methods of detection, and from 1996 to 2000, only 3/780 (0.4%) of cases were entered as relapsing , whereas from 2001 to 2007, 59/1,469 (4%) of cases were registered under this mode of entry (data not shown in Table 2).
With regard to the number of patients undergoing treatment during the years assessed, 106/780 (13.6%) and 1,014/1,469 (69%) of patients had their contacts registered during preintegration and postintegration periods, respectively.The proportion of contact examinations per new DISCUSSION leprosy case was 568/780 (0.73) during the preintegration period versus 2,100/1,469 (1.43) for the postintegration period, showing improved compliance with the contact examination action.
Concerning the type of outcome registered, a significant reduction in the percentage of defaulters was found for the postintegration period (p < 0.008).
Comparing raw rates of detection overall, from 1996 to 2007, the epidemic was consistently worse in Aracaju than in Brazil as a whole for virtually all years studied 9 .These results point to an overall increase in detection after the integration of the LCP into PhC.
Worldwide, the trend in the rates of detection of new cases showed no decline between 1995 and 2000 10 , but a downward shift was evident from 2002 to 2008 11 .In India, the country with the highest number of leprosy cases, the rate of detection steadily increased between 1982 and 2002 12 .A similar steady upward trend also has been seen in Latin America 10 .Brazil accounts for over 90% of cases found in South America, and detection rates 1980 and 2004 showed a steady rise 13,14 with levels set to fall from 2010 13 .For the North-East region of Brazil in particular, the largest increase in detection rates occurred between 1980 and 2004, although this area also had a higher regression rate than any other region in the country 14 .
The upward shift in detection rates seen in Aracaju is similar to the trend reported in other Brazilian studies [15][16][17][18][19] , but a steady decline is evident from 2003 onward.These results contrast with the pattern reported for states in the process of eliminating the disease as a public health problem, such as Paraná 20 and São Paulo 21 , which exhibited a pattern of steady decline in detection rates over the same time frame.
The relatively stable pattern of detection rates in individuals between 0 and 14 years of age, with a marked rise during the postintegration period, is consistent with a study conducted in Duque de Caxias where twin peaks of detection were identified in this age group.This double-peak pattern may be explained by the initial phase of integration or strategic actions carried out the previous year 16 .
Other studies in Brazil involving individuals younger than 15 years have revealed a pattern of endemicity varying from average to very high 18,22,23 .Rates for the Vale do Jequitinhonha were found to exceed those State of Minas Gerais, where parameters for the region were responsible for sustaining the epidemic and contributed to its classification of high to very high 24 .In the State of Espírito Santo, analysis of the trend of detection in individuals under 15 years of age showed an upward pattern between 1980 and 2003, with rates stabilizing between 1996 and 2003 19 .Although other studies have shown similar slight increases in detection in children 25 or have pointed to a steady pattern over time 26 , a study in a historic series performed in Bahia is noteworthy.The cited study found a substantial increase in the rate of detection among minors over the past two decades 15 , demonstrating the recent surge in transmission of the epidemic 5 .
Frequency of new cases with DG 2 at diagnosis is a recent measure currently calculated per 1 million inhabitants 8 , which is employed to assess the magnitude of deformity caused by leprosy in the general population, to compare it against other diseases, and to indicate the efficacy of early case detection 5 .The measure currently lacks parameters defined by WhO or the Brazilian Ministry of health to aid its interpretation.Nevertheless, the world target is to reduce this figure by at least 35% between 2011 and 2015, whereas the target for Brazil is a 13% reduction from 2008 to 2015 5 .Despite the accuracy issues associated with this indicator because of the characteristics of the epidemic and to the limited knowledge of the operational reality of actions in each country, the measure can contribute toward improving the quality of leprosy services 27 .The limitation of this study in interpreting this indicator may be attributable to the lack of data or to the under-reporting of physical disabilities at diagnosis, particularly from 1996 to 2000, during which time, the mean percentage of DG assessment was less than 75%.however, inconsistencies in data regarding disability registration, particularly in the first few years of the series analyzed, must be taken into account.
Biological and sociocultural factors have been implicated as determinants of the greater incidence of leprosy found in men 28 .Other studies carried out in Brazil have confirmed the disease's predominance in male subjects 19,[29][30][31] .however, the greater number of women found in the years assessed in the present study is in line with the findings of other investigations 16,[32][33][34][35] .One possible explanation for this phenomenon is that traditionally, women have a tendency to seek PhC assistance more readily than men 28,36 .This difference is due to cultural gender issues (men delay longer in seeking medical help when ailing), as well as to the greater number of programs and activities available to women in the primary health care sphere.
The predominance of young adults observed in the present cases is consistent with other reports 17,29,30,32,33,35,37,38 .This population group is more vulnerable to the risk of exclusion from the production cycle as a result of the disease and to the sequelae and stigma associated with leprosy 39 .The percentage of individuals aged between 0 and 14 years afflicted by the disease may be a sign of active and recent transmission of the endemic, with under 15s being continuously exposed to the bacillus in the environment where they live 4 .This result also exposes the passive stance of local health services and limited efforts to actively identify cases 24 or implement other means of health surveillance.Some Brazilian studies have confirmed the predominance of the clinical PB form 29,35,37 , whereas others have reported a predominance of MB 19,32,33,38,40 .In the present study, the PB clinical form was most prevalent in both study periods.Surprisingly, no increase in the proportion of PB cases was identified between the two periods, where an increase can usually be expected when a large increase in the number of cases detected occurs.
Disability grading at diagnosis is extensively used as an epidemiological indicator and as a tool for clinical assessment.The increase in the percentage of cases with DG ratings at diagnosis points to the probable involvement of health services fulfilling and performing this control action more effectively.The value obtained for this parameter is regarded as regular as a rating of health care quality and for monitoring results of the LCA 5 .
The high percentage of cases lacking disability rating records, especially during the preintegration period, attests to the fact that physical disability ratings is an LCA, which presents problems in quality and registration of assessments 27 .Although a slight increase in the proportion of cases with DG rated at diagnosis was seen in the postintegration period in this investigation, other Brazilian studies have confirmed a rise, followed by a fall, in the proportion of DG 2 ratings during the period of postintegration of LCA as a result of the measures put in place 16,26,30 .