Trachoma-associated morbidity and mortality in Brazil: an ecological study focusing on hospitalization and mortality data, 2000−2022

ABSTRACT Background: Trachoma is the leading infectious cause of blindness worldwide. It is a neglected tropical disease caused by Chlamydia trachomatis. The objective of this study was to analyze the trachoma-associated morbidity and mortality in Brazil from 2000 to 2022. This ecological time-series study was based on secondary data on trachoma obtained from hospital admissions (trachoma as the primary or secondary cause) and death certificates (trachoma as the underlying or associated cause). Methods: We calculated the sex- and age-standardized rates of hospital admissions and trachoma-specific mortality according to sociodemographic variables and analyzed the spatial distribution. Results: We identified 141/263,292,807 hospital admissions (primary cause: 83.0%) and 126/27,596,830 death certificates (associated cause: 91.3%) related to trachoma. Trachoma-related sequelae were reported in 8.5% of hospital admissions and 6.3% of death certificates. Trachoma was more common in males (hospital admissions and death certificates), people aged ≥70 years (hospital admissions and death certificates), those with brown skin (hospital admissions and death certificates), and those living in the North (hospital admissions) and Northeast (death certificates) regions of Brazil. Conclusions: Despite the relatively low rates of trachoma morbidity in Brazil, the associated mortality rates are of concern. The heterogeneous patterns of occurrence in the country in terms of population and territory reinforce the need to evaluate and monitor the available data, despite the low prevalence, in order to achieve and maintain the elimination targets in Brazil in the future.


INTRODUCTION
Trachoma is the primary infectious cause of blindness globally 1 .It is a neglected tropical disease (NTD) 2 caused by Chlamydia trachomatis 1 .Trachoma has strong social determinants related to poverty and other vulnerable conditions 3 .Accounting for 1.4% of global blindness cases, an estimated 115.7 million individuals resided in the endemic regions of trachoma in 2023, with 1.5 million experiencing sequelae from the disease across 40 countries 1 .
To evaluate the endemicity of the disease and declare it a public health concern, the World Health Organization (WHO) relies on prevalence indicators such as trachomatous inflammation− follicular in children aged 1-9 years, prevalence of trachomatous trichiasis (TT) in individuals aged ≥15 years who are "unknown to the health system," and evidence of the health system's capacity to identify and manage incident TT cases 1 .
During the initial phase of the national trachoma survey in non-indigenous areas from 2018 to 2019 in Brazil, the country's overall prevalence was highlighted at a technical level to eliminate the disease 4 .
Despite implementing a national survey in Brazil that aimed to assess the endemicity of trachoma at the population level, the Trachoma Surveillance and Control Program (in Portuguese: Programa de Vigilância e Controle do Tracoma [PVCT]) in the country is structured around evidence gathered from surveys of school children.This involves finding active cases, treatment and treatment monitoring, health promotion, prevention, control, and disease surveillance 5 .The effectiveness of the PVCT's actions is evaluated by monitoring the percentage of the population receiving treatment, the eligible population undergoing TT surgery, and the positivity rate 5 .The positivity rate, indicating the proportion of individuals testing positive among those screened, is commonly used to assess survey results, with thresholds set as follows: <5%, low positivity; 5-10%, medium positivity; and ≥10%, high positivity 5 .
Despite the issuance of the latest Decree No. 217 on March 1, 2023 6 , trachoma remains under national surveillance in Brazil, with elimination efforts centered on reporting aggregated data via the "Trachoma Survey Bulletin" in the Brazil Information System for Notifiable Diseases (in Portuguese: Sistema de Informação de Agravos de Notificação [SINAN]) 5 .However, routine analyses currently do not include the hospital morbidity and mortality data 7,8 .
The limitations of data collection in low-endemicity contexts and characterization of only positive trachoma cases in the SINAN without recording the demographic and clinical data of those examined for the disease and their contacts hinder a comprehensive analysis of the epidemiological situation 9 .Furthermore, the low sensitivity of the healthcare and surveillance network of the Unified Health System (in Portuguese: Sistema Único de Saúde [SUS]), along with limited prioritization, is evident from the scarcity of scientific publications on trachoma in Brazil 10 .
Although trachoma is not directly linked to mortality, analyzing the hospital and general morbidity and mortality data highlights the vulnerabilities associated with the disease, particularly regarding access to SUS care.This study examined the trachoma-related morbidity and mortality in Brazil from 2000 to 2022 using data from the Hospital Information System (in Portuguese: Sistema de Informações Hospitalares [SIH]) and the Mortality Information System (in Portuguese: Sistema de Informação de Mortalidade [SIM]) of the SUS.This study provides a comprehensive understanding of the attention, surveillance, and control measures required for trachoma.

METHODS
This was an ecological time-series analysis based on secondary data of hospital admissions (HA) and death certificates (DC) associated with trachoma in the different regions and states of Brazil from 2000 to 2022.
Geographically, Brazil is divided into five administrative regions (North, Northeast, Southeast, South, and Central-West), comprising 26 states and 1 federal district, with 5,570 municipalities serving as the analysis units (Figure 1).The country spans a territorial area of 8,510,417.771km² and has approximately 203,080,756 inhabitants 11 .Data regarding trachoma-associated HA (trachoma being the primary or secondary cause) and DC (trachoma being the underlying or associated cause) were extracted from the SIH and SIM databases, respectively, provided by the Department of Information Technology of the Unified Health System (in Portuguese: Departamento de Informática do SUS), Ministry of Health.The following International Classification of Diseases codes (in Portuguese: Classificação Internacional de Doenças e Problemas Relacionados à Saúde [CID10]) were used to identify patients with trachoma-associated HA and DC: trachoma (A71), early phase of trachoma (A71.0),active phase of trachoma (A71.1),unspecified trachoma (A71.9), and trachoma sequelae (B94.0).
For the spatial distribution analysis, the average rates for the 2000-2004, 2005-2009, 2010-2014, 2015-2019, and 2020-2022   periods were calculated and standardized using the direct method based on the age structure by sex from the 2010 census (per 10 6 inhabitants).The natural break method of the Jenks classification algorithm (natural breaks) categorized the spatial classes of the adjusted rates.The areas of residence for HA and DC were utilized as the units of analysis (26 states and 1 federal district), excluding patients from an unknown state of residence.

• Ethical considerations
This study was approved by the Research Ethics Committee of the Hospital São José of Infectious Diseases of the Health Department of the State of Ceará (Approval number: 5.132.182).

RESULTS
During 2000-2022 in Brazil, 141/263,292,807 HA related to trachoma were identified, with an adjusted mean rate of 0.031 per 10 6 inhabitants.Additionally, 126/27,596,830 trachoma-related DC were recorded.Of the HA cases, 83.0% were primarily associated with trachoma, of which 12.1% progressed to death.Among the recorded DC, 8.7% cited trachoma as the underlying cause, and 83.3% of deaths occurred in hospital settings.Trachoma sequelae (CID10 code B94.0) were noted in 8.5% of HA and 6.3% of DC (Table 1, Table 2).The age-and sex-adjusted rates (per 10 6 inhabitants) for HA and DC did not exhibit a discernible temporal pattern across Brazil and its regions throughout the study period (Figures 2A-2B, Table 1, Table 2).
Regarding the municipality classification variables, trachomarelated HA were more frequent among residents of inland municipalities (78.0%), large size municipalities (56.0%), and areas with a very high IBP (24.8%).The highest adjusted rates per 10 6 inhabitants were observed in residents of inland municipalities, small size II municipalities, and areas with a low IBP.Trachomarelated DC were more frequently recorded for residents of inland municipalities (75.4%), municipalities with >100,000 inhabitants (45.2%), and areas with a very high IBP (27.8%).The highest adjusted rates per 10 6 inhabitants were observed in residents of capital cities, small municipalities with a population of 20,001−50,000 inhabitants, and areas with a very high IBP (Table 1).
Among the regions, trachoma-related HA were the most common in the Southeast (41.1%) and high adjusted average rates observed in the North region (Table 1).Trachoma-related DC and the highest adjusted average rates were observed in the Northeast region (Table 2).
The spatial distribution of trachoma-related HA and DC rates exhibits heterogeneity over time across the analyzed periods, with various states recording high rates (HA: ≥0.111 per 10 6 inhabitants; DC: ≥0.054 per 10 6 inhabitants) (Figure 3).This spatiotemporal  pattern, varying across regions over the years (average adjusted rates every 5 years), is also evident in the annual time-series analysis of the country's regions (Figures 2 and 3).

DISCUSSION
Despite having identified the limitations, trachoma-related morbidity and mortality persist in Brazil.The spatial distribution of trachoma-related HA and deaths across the country indicates occurrences in all states, particularly in large municipalities and areas characterized by greater social inequality and vulnerability.The Northeast region of Brazil had the highest proportion of trachoma-related deaths, suggesting more than just an operational issue with registration.Notably, one state situated in the North region of Brazil, exhibited the highest trachoma-related mortality rate.Moreover, the North and Northeast regions had the lowest proportions of dwellings with access to a general water supply network (58.8% and 80.0%, respectively) and garbage collection (72.4% and 70.8%, respectively) in Brazil 11 .Hence, it is imperative to analyze these occurrences in areas where a significant portion of the Brazilian population faces limited access to healthcare services and resides in precarious living conditions 12 .
This study underscores the necessity of enhanced evaluation and monitoring of data within the country's Health Information Systems (In Portuguese: Sistemas de Informação em Saúde [SIS]) to ensure higher quality of analysis.Although death is not a direct clinical outcome of trachoma, HA could be associated with the need for surgical correction of eyelid sequelae of the disease 5 .These occurrences may indicate more severe disease states and potential limitations in accessing timely care with greater technical complexity within the SUS 5 .A review of the registration of trachoma-related deaths and HA in a long-term national historical series further underscores these perspectives 9 .
Trachoma as a cause of death suggests a likely inconsistency in the registry, as it is directly associated with disability but not with death.In addition to this inference not correlating with the clinical status of the disease, inadequate coding of records with the clinical forms of trachoma, fragmentation and/or duplication, operational limitations and a lack of interoperability highlight the limited reliability of the records.This could lead to misinterpretation, making them unrepresentative of the population and/or the health-disease process in question 13 , and thus influencing the correct decision making for SUS management 14 .

Maciel AMS et al. • Trachoma morbidity and mortality in Brazil
A better understanding of the possible operational factors that may have influenced these results is warranted to bring positive changes in the process of management and analysis of health data from the SIH-SUS and SIM, especially considering the low endemicity of the disease in the country 7,8,9 .
The difficulties experienced by the municipal management and local health professionals in conducting care and surveillance actions indicate the need for operational and implementation research related to the SIS to improve its use and enhance transparency in the database analysis strategies within the SUS 15 .Therefore, a detailed evaluation of the quality analysis reports of these systems is recommended to detect causes incompatible with the occurrence of these events, thereby improving the data adequacy and accuracy.This is crucial for planning public health policies aimed at eliminating trachoma.
The lack of qualified data for evidence-based decision-making 16 and operational difficulties in managing and recording TT cases 17,18 in information systems make it difficult to understand the trachoma morbidity and mortality patterns and eliminate the disease.
Underreporting and absence and/or inconsistency of information can lead to the under-or overestimation of health indicators.Therefore, the data collected by the services should be evaluated, and the country's health professionals should be trained to enhance the epidemiological quality of actions that closely align with the real situation 10 .
Nevertheless, the morbidity of the disease is considered relevant in terms of public health, with cases recorded in more than 9% of Brazilian municipalities (508) and associated with leprosy, leishmaniasis, and schistosomiasis in almost all (96.6%) cases detected as NTD in Brazil in 2015 10 .It is also worth considering the limited understanding of the record of hospitalizations due to the disease in the country, an understudied aspect, and the possible indirect impact of the disease on mortality 10 .
Global NTD programs acknowledge the strategic importance of progress in national health systems for more effective and planned responses to achieve the elimination targets set by the WHO 13 .Furthermore, healthcare and surveillance interventions can be enhanced by using better quality data 14 .Therefore, monitoring the completeness of DC, as recommended by the WHO, is considered a strategic and essential approach for the SUS to obtain consistent information on mortality for conducting assertive interventions 19 .
Improvement in the quality of SIM records was particularly evident after 2006, with a reduction in records of deaths due to undetermined causes.The most critical aspects of the SIH-SUS are related to coverage and completeness due to imprecision in the definition of the cause of hospitalization 20 .
Therefore, the SIS used to characterize morbidity and mortality must provide a specific functional perspective, with individualized analysis through its own critical reports, in addition to the ability to interoperate with different databases 21 that include trachoma.By utilizing probabilistic or deterministic resources and tools, data linkages expand the scope for qualifying the study of these diseases and provide space to ensure the care of people with these conditions 22 .Furthermore, the systematic use of integrated data at the local, regional, and national management levels is essential for achieving the Sustainable Development Goals (SDGs) of the 2030 Agenda 23,24 .Furthermore, it is crucial to assess the composition and systematically monitor health information systems to qualify information and support the verification of elimination in populations at risk of the disease 2 .
In Brazil, trachoma remains among the diseases targeted for elimination 4,25 ; however, it was not included among the NTDs presented in the National Agenda of Priorities for Health Research, and therefore lacks funding, as defined in that proposal 24,26 .Despite presenting a considerable global burden of disease (measured in disability-adjusted life years), particularly because it is associated with a low and very low "Human Development Index" and lower "expected years of schooling" dimension, there is a probable decrease in detection due to better socioeconomic and educational development conditions in the population 27 .
In 2023, trachoma was included in the scope of the prioritized socially determined diseases by the Interministerial Committee for the Elimination of Tuberculosis and Other Socially Determined Diseases (In Portuguese: Comitê Interministerial para a Eliminação da Tuberculose e de Outras Doenças Determinadas Socialmente).As progress has been made, the unfolding of the State Policy -Healthy Brazil Program and expansion of inter-sectoral actions have been aligned with the 2030 Agenda to achieve the SDGs.This focus aims to eliminate and/or reduce public health problems such as trachoma that affect populations facing social inequality 24,28 .
However, challenges associated with improving these intersectoral actions include ensuring completeness and consistency of records, monitoring them, and evaluating actions and strategic analyses by integrating information systems.Nonetheless, there are prospects for expanding the planning and decision-making capacity to control the disease prevalence in the country 23 .
The limitations of this study include the use of secondary data from the SIH-SUS and SIM, which may result in incomplete recording of variables.The results indicate the presence of probable operational inconsistencies in the databases and the need to validate the information for the true characterization of this NTD.Despite these limitations, the use of large databases from different SISs combined with specific critical analyses tailored to these systems and spatial and temporal distributions provides new perspectives for efficient and reliable situational analysis of trachoma morbidity and mortality in the country.
Trachoma continues to impose high morbidity and mortality burdens on the country.However, there is an increased need for the evaluation, monitoring, and systematic critical operational analysis of the SIS data to ensure the completeness and consistency of HA and DC records.Expanding research to understand the factors influencing these outcomes better implies changes in the SIH and SIM management processes.Appropriation of this information will provide knowledge for health management and planning, with a view toward more qualified and integrated interventions for healthcare and surveillance, particularly in primary healthcare, which is essential for controlling trachoma and NTDs in general.
Despite the relatively low rates of trachoma morbidity in Brazil, the associated mortality rates are of concern.The heterogeneous patterns of occurrence in the country in terms of population and territory reinforce the need to evaluate and monitor the available data, despite the low prevalence, in order to achieve and maintain the elimination targets in Brazil in the future.Therefore, there is a clear need to qualify disease surveillance, care, and control interventions.

FIGURE 3 :
FIGURE 3: Spatial distribution by state of the trachoma-related (A) hospital admission rate and (B) mortality rate in Brazil (2000-2022).

Trachoma morbidity and mortality in Brazil FIGURE 2:
Maciel AMS et al. • Trachoma-related (A) hospital admission rate and (B) mortality rate adjusted by sex and age in Brazil overall and each region (2000-2022).