Mortality from tetanus between 1990 and 2015: findings from the global burden of disease study 2015

Background Although preventable, tetanus still claims tens of thousands of deaths each year. The patterns and distribution of mortality from tetanus have not been well characterized. We identified the global, regional, and national levels and trends of mortality from neonatal and non-neonatal tetanus based on the results from the Global Burden of Disease Study 2015. Methods Data from vital registration, verbal autopsy studies and mortality surveillance data covering 12,534 site-years from 1980 to 2014 were used. Mortality from tetanus was estimated using the Cause of Death Ensemble modeling strategy. Results There were 56,743 (95% uncertainty interval (UI): 48,199 to 80,042) deaths due to tetanus in 2015; 19,937 (UI: 17,021 to 23,467) deaths occurred in neonates; and 36,806 (UI: 29,452 to 61,481) deaths occurred in older children and adults. Of the 19,937 neonatal tetanus deaths, 45% of deaths occurred in South Asia, and 44% in Sub-Saharan Africa. Of the 36,806 deaths after the neonatal period, 47% of deaths occurred in South Asia, 36% in sub-Saharan Africa, and 12% in Southeast Asia. Between 1990 and 2015, the global mortality rate due to neonatal tetanus dropped by 90% and that due to non-neonatal tetanus dropped by 81%. However, tetanus mortality rates were still high in a number of countries in 2015. The highest rates of neonatal tetanus mortality (more than 1,000 deaths per 100,000 population) were observed in Somalia, South Sudan, Afghanistan, and Kenya. The highest rates of mortality from tetanus after the neonatal period (more than 5 deaths per 100,000 population) were observed in Somalia, South Sudan, and Kenya. Conclusions Though there have been tremendous strides globally in reducing the burden of tetanus, tens of thousands of unnecessary deaths from tetanus could be prevented each year by an already available inexpensive and effective vaccine. Availability of more high quality data could help narrow the uncertainty of tetanus mortality estimates. Electronic supplementary material The online version of this article (doi:10.1186/s12889-017-4111-4) contains supplementary material, which is available to authorized users.


Background
Tetanus, commonly referred to as "lockjaw", is a serious infection caused by Clostridium tetani. The bacterium is commonly found in the environment (usually in soil, dust, and animal waste). Tetanus spores can enter the body through cuts or abrasions. Newborns can become infected through contaminated instruments used to cut the umbilical cord or by improper handling of the umbilical stump [1]. Neonatal tetanus is more likely to occur in low and middle income countries especially in places such as urban slums and rural areas; in those places unhygienic deliveries at home are common, and coverage of antenatal care services and maternal tetanus toxoid immunization are usually inadequate [2][3][4].
During the past two decades, there has been a dramatic decline in tetanus cases and deaths due to the scale up of immunization programs [5,6]. Despite the availability of an inexpensive and effective tetanus vaccine, many people in low and middle income countries continue to die from tetanus. In developed countries, tetanus is rare but occasional cases and deaths continue to occur in unvaccinated individuals. The current patterns and distribution of tetanus mortality have not been well documented. In this study, we identify the global, regional and national levels and trends of neonatal and non-neonatal tetanus mortality between 1990 and 2015, based on the findings from the Global Burden of Disease Study 2015.

Methods
Data from vital registration, verbal autopsy, and mortality surveillance data covering 12,534 site-years from 1980 to 2014 were used for this study [7]. The International Classification of Diseases (ICD) codes for neonatal tetanus include ICD-10 codes (A33-A35.0) and ICD-9 codes (037-037. 9, 771.3). Further details about data sources are provided in the Web Appendix. We used the Cause of Death Ensemble model (CODEm) strategy [7][8][9][10], which has been widely used for generating global estimates of cause-specific mortality. The CODEm strategy evaluates potential models that apply different functional forms (mixed effects models and space-time Gaussian Process Regression models) to mortality rates or cause fractions with varying combinations of predictive covariates [7], including DTP3 coverage proportion, educational attainment, health system access, in-facility delivery proportion, lagged distributed income, skilled birth attendance proportion, and tetanus toxoid coverage proportion. An ensemble of models that performs best on out-of-sample predictive validity tests was then selected as the best model. A complete time series of the parameters for each covariate for each location was estimated using data from household surveys, censuses, official reports, administrative data, and systematic reviews. The sources and imputation methods used to generate time series for the covariates have been published elsewhere [11].

Results
There were 56,743 (95% uncertainty interval (UI): 48,199 to 80,042) deaths due to tetanus in 2015: 19,937 (UI: 17,021 to 23,467) deaths occurred in neonates and 36,806 (UI: 29,452 to 61,481) deaths occurred after the neonatal period (Table 1). Of all neonatal tetanus deaths, 45% of deaths occurred in South Asia. Sub-Saharan Africa accounted for additional 44% of deaths; 67% of these deaths occurred in eastern sub-Saharan Africa, 27% in western sub-Saharan Africa, and 6% in central sub-Saharan Africa. Of tetanus deaths after the neonatal period, 47% of deaths occurred in South Asia, 36% in sub-Saharan Africa, and 12% in Southeast Asia. Figure 1 shows the global age-sex distribution of tetanus mortality in 2015. Tetanus deaths were concentrated in neonates when they were compared with deaths in each of the other age categories (Fig. 1). More deaths occurred in males than females in most age groups (Fig. 1). Agestandardized tetanus mortality rate (per 100,000 people) among males (0.93, UI: 0.72 to 1.44) was also higher than that among females (0.63, UI: 0.50 to 0.90) (data not shown).
Between 1990 and 2015, the global mortality rate due to neonatal tetanus dropped by 90% and that due to non-neonatal tetanus dropped by 81% (Table 1). At the country level, the decline in neonatal tetanus mortality rate varied from −47% in Somalia to −95% in Angola in sub-Saharan Africa. The decline in tetanus mortality rate after the neonatal period varied from −0.12% in South Sudan to −92% in Mauritania in sub-Saharan Africa (Table 1).
There were also substantial between-country variations in tetanus mortality rates (Figs. 2 and 3). For example, neonatal tetanus mortality rates per 100,000 people varied from 3,376.4 (1,731.6 to 6,447.9) in Somalia to 1.0 (0.4 to 2.0) in Zimbabwe in sub-Saharan Africa in 2015 (Table 1). Tetanus mortality per 100,000 people after the neonatal period varied from 10.3 (3.6 to 23.7) in Somalia to 0.04 (0.03 to 0.06) in South Africa in the same year (Table 1).
Although both neonatal and non-neonatal tetanus deaths were concentrated in low and middle countries, a small number of deaths from non-neonatal tetanus continued to occur in high-income countries. We estimated 36 (UI: 28 to 51) deaths in Western Europe, 13 (UI: [11][12][13][14][15][16] deaths in high-income Asia Pacific, and 9 (UI: 8 to 11) deaths in high-income North America due to tetanus in 2015 ( Table 1): most of these deaths occurred in adults, especially among elderly people. More detailed results showing the location-year-age-sex specific distributions of tetanus mortality from 1990 to 2015 in 5-year interval are viewable in an interactive online visualization tool at http://vizhub.healthdata.org/gbd-compare.

Discussion
Exceptional progress has been made over the past two decades in reducing mortality from tetanus worldwide. Nevertheless, mortality from tetanus was still unnecessarily high in a number of low and middle income countries in 2015. The scale-up of immunization coverage to prevent maternal and neonatal tetanus represents a huge success of a collective effort. However, the scale-up has not been universal, with low vaccination coverage being documented in several countries [6,12,13]. Constraints related to financial and human resources and difficulty vaccinating people in hard-to-reach rural areas were among the factors influencing the tetanus toxoid vaccine coverage [12].                Tetanus mortality rates were the highest among neonates in low and middle income countries, indicating failures of health systems to provide immunization, antenatal care, and clean deliveries for all births. Mortality rates from tetanus after the neonatal period were much higher in low and middle income countries compared with high income countries, but a small number of deaths continued to occur in high income countries due to low vaccination coverage in adults [14,15]. Our findings showed that agestandardized mortality from tetanus was higher among males than females globally. Previous studies have also reported male sex as a risk factor for both neonatal and nonneonatal tetanus [16,17]. Although the exact reason is not clear, possible explanation for the increased risk of tetanus among newborn boys include medical-care seeking for boys, differential cord care, maternal recall, and circumcision practices [13,16]. Among adults, occupational exposure and relatively lower vaccination coverage in men were among the reasons for the increased risk [17].
A main limitation of this study concerns the poor availability of data in many sub-Saharan African countries where tetanus mortality is most common. For countries without reliable vital registration systems, our analysis relies on verbal autopsy data. Variations in analytical methods and the instrument used for collection of verbal autopsy data may also introduce measurement bias and reduce the comparability of tetanus cause-of-death data across countries. Estimating tetanus mortality for every geography over time is challenging especially for those with sparse or no data. We applied sophisticated modeling methods, borrowing strength across geography and covariates to help predict for locations and years with limited data. Accordingly, the estimates for a geography with sparse data are reflected by wider uncertainty intervals (Detailed information on data availability, model estimates and uncertainty intervals for each region and country are available online at http://vizhub.healthdata.org/cod/). New data for countries, especially in the sub-Saharan African region would narrow the uncertainty in the tetanus mortality estimates for countries in the region.

Conclusions
Up-to-date information on the levels and trends of tetanus mortality is critical to guide prevention and intervention efforts. Despite the availability of a safe, inexpensive, and effective vaccine, our findings on tetanus mortality suggest that the vaccine is not fully utilized. Despite the general decline in tetanus mortality, tens of thousands of lives could still be saved by scaling up interventions.

Additional file
Additional file 1: Tetanus data sources and citations. (XLS 114 kb)

Funding
The Global Burden of Disease Study 2015 was funded by the Bill & Melinda Gates Foundation. The funding body has no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.

Availability of data and material
The data sources that support the findings of this study are available as Additional file 1. The datasets generated during the current study are available through the GBD Results Tool (http://ghdx.healthdata.org/gbdresults-tool). Additionally, metadata for all sources of raw data analysed in the current study are available in the GBD Data Input Sources Tool (http:// ghdx.healthdata.org/gbd-2015/data-input-sources), which includes information about the data provider where interested parties can inquire about data access. Some restrictions apply to the availability of unpublished data, which were used under license for the current study, and so are not