Standardized Mortality Analysis of Military Fireghters from 2000-2016 in Rio de Janeiro, Brazil

Backgrounds


Conclusion
The re ghters' group had the lowest overall mortality detected in both the general and the main chapters of the 10th Revision of the International Statistical Classi cation of Diseases.

Background
The tasks of re ghters are extremely exhausting and often present risks to their own health. Their workplace, characterized by tense and dangerous environments, is considerably harmful to their own physical and mental integrity in the course of their lives. 1,2 Fire ghters are constantly exposed to strong emotional charges as they often experience dangerous situations, deal with death, and witness tragic scenes. Moreover, after attending to these circumstances, they return to their barracks to work without receiving any support to help them overcome these traumatic situations that is part of their daily lives. 3 Besides that, they have a heavy workload, are often under tension, deal with human lives, and are exposed to dangerous agents. However, since these professionals need to be ready for new calls, they try to neutralize the stressful experiences brought about by these situations. 3 There are inherent dangers in the re ghters' profession. They are exposed to physical risks because of the noise and the extreme temperatures, biological risks because of their exposure to microorganisms that are in contact with blood and organic uids, and chemical risks because of their exposure to chemical substances. 4 During their work, they face multiple challenging situations such as re ghting, rescuing people and animals, and landslides. They are also constantly exposed to combustion remains, smoke, carcinogenic substances that are volatile in the re and pyrolysis or debris. These exposures have raised concerns on the increase in potential cancer cases among re ghters and stimulated epidemiological investigations to assess them. 5 Fire ghters are surprised by emergency calls that must be immediately answered during their shift; therefore, they need to have good physical conditioning. These professionals react to these calls with a signi cant increase in heart rate. 6 During re suppression, they work with maximum load and their heart rate reaches high values. 7 According to studies, cardiac overload is aggravated by the use of heavy clothing for extended periods. 8 Moreover, heat stress and consequential uid loss can result in decreased cardiac output despite the initial tachycardia. 8 Many deaths during service are precipitated by the inherent aid stress that occur to re ghters with underlying cardiovascular diseases. [6][7][8] Furthermore, the mental health of these military personnel deserves full attention in order to recognize their health problems and its consequences early on. Studies on the subject indicate that these professionals have a high incidence to certain types of cancer [9][10][11] and deaths in service due to cardiovascular diseases. [6][7][8] To carry out this study, we identi ed the causes of death of these deceased re ghters, then compared it with the mortality rate of the general population in the state of Rio de Janeiro.
This study aimed to characterize the re ghters' mortality from 2000 to 2016 by using the 10th Revision of the International Statistical Classi cation of Diseases and Health Related Problems (ICD-10) and demographic variables; and comparing the observed distribution of deaths to those expected according to the mortality rate of the general population in the state of Rio de Janeiro.

Study design
This study is about the mortality of active and inactive re ghters of Rio de Janeiro present in the payrolls of Rio de Janeiro State Military Fire ghters (CBMERJ -Corpo de Bombeiros Militar do Estado do Rio de Janeiro) from 2000 to 2016. The population of this study consists of a cohort of male CBMERJ re ghters from January 2000 to December 2016, excluding those who worked for less than 12 months in the corporation. A database was also organized containing the data category of the deaths that occurred in the same period.

Data source
To establish the person-time of the cohort, we obtained the payrolls from 2000 to 2016 from the General Directorate of Finance (DGF).
Moreover, we collected the data of deceased re ghters through their death certi cates, which is required for their pension concessions to dependents, from the Directorate of Personal Management's (DGP -Diretoria Geral Pessoal) occupational history and the Directorate of Inactive and Pensioners (DIP -Diretoria de Inativos e Pensionistas) of CBMERJ. With these death certi cates, it was possible to access the underlying causes of death of 383 fatalities by directly searching the public and unrestricted database of the Mortality Information System (SIM -Sistema de Informação sobre Mortalidade).
Moreover, when the death certi cates or the number of the deceased re ghters were not found in the DGP or DIP, we searched the Vital Data services of the Health Departments of the Municipality (SMS -Secretaria Municipal de Saúde) and the state (SES -Secretaria Estadual de Saúde) of Rio de Janeiro for their names and dates of death.
For information on the underlying causes of these deaths, we created a linkage between the data of the deceased re ghters and the database of the Municipal Health Department of Rio de Janeiro, whose methodology pursued the following steps: Cleaning and standardization of the variables of the Fire ghters on Excel, lled with the data collected from CBMERJ and SIM from 2000 to 2016, using Stata 11.0. A unique identi er was created at the re ghters base to further merge them, building a single re ghters bank with the found and not found deaths; Probabilistic relationship (linkage) of databases using the OpenRecLink (ORL) software; Probabilistic linkage (ORL): Relationship variables: name, date of birth, and gender (for visual inspection). Blockage: Step 1: FBLOCK ( rst name) + LBLOCK (last name) + DB (date of birth) + gender Parameters of the variables for the probabilistic relationship in the ORL: Name: "Approximate" Type, "92" Correct, "1" Incorrect and "85" Threshold, according to the III Reclink Manual.
Step 2: FBLOCK + DB Parameters of the variables for the probabilistic relationship in the ORL: Name: "Approximate" Type, "92" Correct, "1" Incorrect and "85" Threshold, according to the III Reclink Manual.
Construction of the nal base where the found cases are contained after the manual non-pairs search.
In this initial linkage carried out with the SMS database of Rio de Janeiro, we found 1,131 pairs out of the 1,816 deaths recorded in the studied cohort of re ghters.
The deceased who were not found through this strategy were searched in the Database of the State Department of Health; identifying 256 more individuals.
However, the underlying cause of death of 46 re ghters was not obtained even though the records from CBMERJ showed that they were deceased. Thus, we classi ed them under "indeterminate cause of death" (R99.0).

Data analysis
To obtain the Standardized Mortality Ratio (SMR), 12 everyone's participation time in years in the corporation was calculated, and the sum of the data set the person-time of this cohort. As a reference, we used the population of men over 19 years old in the state of Rio de Janeiro in 2010 according to the Demographic Census, which brought more accurate information, in the intermediate period of the cohort.
The SMR was calculated by dividing the number of actual observed deaths in the cohort of military re ghters by the number of expected deaths, that is, those that would occur if the cohort was submitted to the same mortality rates per age group of the male population of the state of Rio de Janeiro in 2010.
As indicated by the literature, speci c SMRs were calculated for the most frequent pathological groups in the career of re ghters such as circulatory system disease, neoplasia, and death by external causes (accidents and by violence). [6][7][8][9][10] Lastly, the statistical analysis was performed using the Stata 20 program.

Results
Fire ghters' characteristics  For deaths by external causes, an excess was detected for those by rearm projectile (CID-10: X93; X94 and X95), presenting 189% SMR which is increasing according to age group from 30 to 69 years. There was no death due to this cause for the age group over 69 years (Table 4).

Discussion
This study was the rst to evaluate mortality of military re ghters from the state of Rio de Janeiro, which had immense value for the entire corporation, being of great relevance to the issue and for future projections in relation to the health of these professionals. The study population was the cohort of male military re ghters from 2000 to 2016.
This is a young cohort since more than 40% of all people-years were under 40 years old. However, the increase of people-year over 69 years old, especially from 2010 onwards, reveals that the cohort is aging.
The observed causes of death in the cohort were concentrated in Chapters II, IX, and XX of the ICD-10, with percentage values of 16.6%, 28.0%, and 22.6% respectively. It is worth mentioning that this distribution differs from that presented in the population of the state of Rio de Janeiro in 2010, with 15.2%, 29.3%, and 13.8% respectively. 13 The SMR for all causes of death in re ghters was 64.6% (95%CI: 61.6-67.6%). This favorable result to re ghters can be explained by the effect of healthy workers, rigorous selection on admission, periodic examinations in military environments with earlier detection of morbidities, and the practice of regular physical activity. 13 Other studies about re ghters corroborate this nding, with SMR for all causes of death ranging from 74% to 99%, which the authors also attribute to the effect of healthy workers, careful entry into the military units, and performance of physical capacity tests according to age in addition to admission exams. 10,15-19. Brice et al. (2015) studied mortality in re ghters from France and used SMR to make a comparison of the French male population between 1979 and 2008. 19 From 89 French administrative departments, 10,829 employed re ghters were included in the cohort in 1979. A total of 1,642 deaths were observed with SMR for all causes at 81% (95%CI: 77-85%), which is increasing with age but not exceeding 100%. They did not detect high SMR for any speci c cause and this nding was related to the effect of healthy workers in the cohort. 19 Between 1970 and 2014, Petersen et al. (2018) conducted a study on mortality in a cohort of Danish re ghters. 10 A total of 11,775 re ghters representing 235,526 people-year were monitored. The authors used two reference populations to calculate the SMR: a sample of male workers' population (n= 262,168) and of military personnel (n= 396,739). They observed 1,017 deaths in the cohort, with low SMR for all causes of death when comparing both workers (SMR=74%; 95%CI: 69-78%) and military (SMR= 88%; 95%CI: 83-93%). Death from stomach cancer (SMR = 196%; 95%CI: 122-316%) was signi cantly higher in re ghters who served full-time, while the authors observed a signi cant increase in prostate cancer in the volunteers (SMR= 189%; 95%CI: 122-293%) as compared with the reference populations. The authors justi ed the SMR under 100 for all causes of death by citing the strict selection system upon entering the corporation, and the early diagnosis of certain pathologies that were referred to for treatment and for necessary follow-up. 10 In this study, the chapter that concentrated the highest proportion of deaths in men over 20 years old in 2010 was circulatory system diseases (28.0%), a similar proportion to that of the general population of the state of Rio de Janeiro (29.4%) and Brazil (28.4%). 13 A systematic literature review showed that studies in military re ghters highlighted a high incidence of heart disease, often related to a sedentary lifestyle and being overweight. Acute and chronic occupational factors have also been related to the risk of heart disease among them. From the chronic factors, the long periods of sedentary lifestyle, exposure to smoke, noisy environments, shiftwork that impairs sleep quality, inadequate eating in military environments, occupational stress, and posttraumatic stress disorder (PTSD) are highlighted. Among the acute ones, the main factors are environments with high temperatures culminating in dehydration, response to alarms, exposure to gases and particles in the re scenarios, and the execution of the activity itself. 7 A study in Denmark on the incidence of cardiovascular diseases in re ghters showed an increased proportion in this occupational group, especially angina pectoris, acute myocardial infarction, and heart ischemic disease. 20 In this study, the mean age of death occurrences from cardiovascular diseases was 45.4 years for active military re ghters and 62.4 years for the inactive ones. The SMR for cardiovascular diseases was 64.4% (95%CI: 58.8-68.9%), with acute myocardial infarction (I21 in ICD-10) being the most frequent basic cause of death in this chapter with SMR of 72.4%. This SMR below 100 can also be attributed to the effect of healthy workers and periodic tests that contribute to the detection of risk factors and early treatment. 14 Regarding neoplasms, the SMR for all cancer types was 72.8% (95%CI: 64.5-81.0%). Anent primary location, prostate cancer presented 94.0% SMR, surpassing 100 in the age groups of 50-59 and 60-69 years. These values may be related to the performance of periodic tests because they allow greater diagnosis possibility than the general population.
Cancer mortality studies in re ghters report that there is a relationship between their profession and neoplasms, although the main locations vary from one study to another.  9 assessed mortality and incidence of cancer in volunteer re ghters in Australia. When compared to the male population of the country, they observed an incidence decrease for almost all types of cancer, with incidence increase found only for prostate cancer. This was the only re ghters' mortality study that related the causes of death to the accident types that the re ghters were called for (all accidents types, all re types, res in structures, res in landscapes and res in vehicles). An SMR of 183% was observed (95%CI: 110-286%) in deaths due to smoke, re, and ames exposure.
Deaths by external causes in this study represented the second most frequent cause, presenting an SMR of 88.5% (95%CI: 79.9-97.0%), highlighting deaths caused by rearm projectile (SMR=119.4%). It is noteworthy that studies in the literature on this underlying cause of death of re ghters were not found. A low SMR for rearm death was observed in the age group of 20 to 29 years, over 100 in age groups above 30 years, and even higher in the age group of 60-69 years when the military is no longer active in the corporation. The low SMR value observed in people under 30 years old may be related to the high mortality ratio of the general population of the state of Rio de Janeiro in this age group for the same cause, most often poor, black, and culturally and economically less favored. 22 Other hypotheses can be raised in this context because when the re ghters start their military lives, many begin to carry out other activities, seeking to maintain a better standard of living. Thus, it is possible that in the state of Rio de Janeiro, this nancial complement is carried out in risky activities such as security. These may be the causes of the highest number of fatalities due to rearm death in this age group, considering the population of this state with the same gender and age.
It should be noted that Brazil has very high numbers of death by violence. Upon analyzing the country's male mortality rates from 2000 to 2016, it was observed that the external causes chapter always presented a magnitude higher than 115 deaths per 100,000 men, surpassed only by the circulatory system diseases chapter. Moreover, in the state of Rio de Janeiro, external causes had even higher rates, occupying the second place between 186 deaths per 100,000 men in 2002 and 126 deaths in 2015. Among deaths by external causes in this state, the main cause of male fatalities was aggression by rearm, although with a decreasing tendency from 45.8% in 2000 to 33.9% in 2016. 13 However, this study presents some limitations. First, we did a retrospective design when the better approach would be a prospective study with risk factors quanti cation for occupational-related diseases and exposures, accurate diseases documentation, and life habits reports. Nevertheless, the results presented are unprecedented and offer a realistic overview of the mortality of the corporation. Second, the cohort follow-up period (from 2000 to 2016) was shorter than desired due to the unavailability of previous data. However, this restriction can be considered bene cial since it allowed analysis within the same ICD-10 at a time when the mortality system offers very good quality data. 23 Third, we collected data from death certi cates, speci cally the underlying cause of death, instead of data selected by a competent agency using its own application, which data are more trusted than those found on death certi cates and corporation records.
This study allowed mortality evaluation, which is previously unknown, in this occupational group that performs services of excellence to the population of Rio de Janeiro. Its results may support the implementation of intervention programs in order to prevent the most frequent mortality causes in the future.

Conclusions
This study showed high clinical and social relevance since the causes of death in this occupational group had never been addressed, especially with the discovery of many unknown results. Given the importance of the work of military re ghters and the dangerous situations they experience in their daily lives, knowing their main causes of death deserves full attention from researchers who are interested in the subject. The data we discovered should be disclosed to all active and inactive re ghters of the Corporation.
The SMR was used to evaluate deaths in this occupational group. There is no report in national and international literature on the death excess due to rearm, but this mortality cause stood out in this study. Another relevant nding was the SMR for prostate cancer, which is above 100 in the age group of 50 to 69 years, although without statistical signi cance.