Tuberculosis (TB) is a worldwide public health problem. Although the osteoarticular form represents from 3% to 5% of its manifestations, few data on the mortality caused by this type are available1. To expand the recently published study on its incidence and evolution over time2, we analysed mortality due to osteoarticular TB in Spain from a hospital perspective in the years from 1997 to 2018.
The mortality and lethality of osteoarticular TB (OA TB) over 22 years were estimated in an observational retrospective study based on the data gathered in the minimum basic set of hospital discharge data for patients with a main or secondary diagnosis (according to CIE-9 and CIE-10) of OA TB in Spain from 1997 to 2018. 336 deaths occurred in 5710 patients.
The average annual mortality of patients with OA TB was 0.35 cases per million inhabitants (CI 95% 0.31−0.38). The World Health Organization has detected a falling tendency in the estimated rates of mortality due to TB in all regions since 20003. The study found a significant difference between the annual average mortality per million inhabitants in the first period (1997–2007) with 0.12 deaths p.m., and the second period (2008–2018) with 0.07 deaths p.m. (P < .001).
It is striking that there was no mortality among the 31 babies and 133 children aged from 1 to 14 years with OA TB, when the average rate of lethality due to TB in children worldwide is about 24%, and that TB is one of the 10 main causes of death among children in the whole world4. Nevertheless, this may be explained by the fact that more than 96% of all deaths due to TB occur in children who received no treatment against TB5.
On the other hand, the incidence and mortality were higher among those aged above 75 years (0.25 deaths per million). The overall fatality rate was 5.9% (CI 95% 5.3–6.53). Mortality among the patients with OA TB as their main diagnosis at admission was 3.5 (CI 95% 2.85–4.20). A meta-analysis estimated that the combined percentage of patients with TB who died during treatment of the same was 18.8% in the patients infected with HIV and 3.5% in those not infected with HIV6.
The fatality rates for coinfection of OA TB with miliary TB or CNS TB were, respectively, 9.5% and 9.3%, while in two works in Spain that used methodology similar to ours, the fatality rates for miliary TB and CNS TB were 14% and 15.5%, respectively7,8. The fatality rate of the osteoarticular form is lower than that of the other forms of TB, and this may be due to the fact that it is a more localized infection that does not compromise vital organs.
Bivariate and multivariate analysis was performed on the different factors involved, to discover their relationship with death rates (Table 1). The patients with cardiovascular dysfunction were at 26.5 times higher risk of dying than the others, those who had malign neoplasia were at 6.2 times higher risk and those with respiratory failure were at 5.3 times higher risk of dying (Table 1).
Bivariate and multivariate analysis of risk factors and comorbidities for death.
Bivariate and multivariate analysis of the risk factors associated with mortality | |||||||
---|---|---|---|---|---|---|---|
Variables | Death in hospital with osteoarticular TB (n = 336) | ||||||
Yes | No | Bivariate analysis | Multivariate analysis | ||||
N (%) | N (%) | P value | OR (CI 95%) | P value | OR (CI 95%) | ||
Neurological deficit | |||||||
Yes | 23 (11.4%) | 179 (88.6%) | .001 | 2.133 (1.361−3.341) | .001 | 2.403 (1.446−3.994) | |
No | 313 (5.7%) | 5195 (94.3%) | Control group | Control group | |||
Cerebrovascular disease | |||||||
Yes | 28 (24.1%) | 88 (75.9%) | <.001 | 5.461 (3.515–8.483) | <.001 | 3.385 (2.044−5.608) | |
No | 308 (5.5%) | 5286 (94.5%) | Control group | Control group | |||
Diabetes mellitus | |||||||
Yes | 57 (9.2%) | 564 (90.8%) | <.001 | 1.742 (1.293−2.347) | .529 | 0.896 (0.637−1.261) | |
No | 279 (5.5%) | 4.810 (94.1%) | Control group | Control group | |||
Malign neoplasia | |||||||
Yes | 38 (24.8%) | 115 (75.2%) | <.001 | 5.831 (3.969−8.569) | <.001 | 6.229 (4.025−9.639) | |
No | 298 (5.4%) | 5.259 (94.6%) | Control group | Control group | |||
Chronic liver disease | |||||||
Yes | 22 (17.7%) | 102 (82.3%) | <.001 | 3.621 (2.253−5.821) | <.001 | 3.503 (1.923–6.381) | |
No | 314 (5.6%) | 5.272 (94.4%) | Control group | Control group | |||
Chronic kidney disease | |||||||
Yes | 43 (16.5%) | 217 (83.5%) | <.001 | 3.488 (2.463–4.938) | .027 | 1.626 (1.057−2.501) | |
No | 293 (5.4%) | 5.157 (94.6%) | Control group | Control group | |||
Malnutrition | |||||||
Yes | 11 (15.1%) | 62 (84.9%) | .001 | 2.9 (1.512−5.56) | .180 | 1.698 (0.782−3.687) | |
No | 325 (5.8%) | 5.312 (94.2%) | Control group | Control group | |||
Acute respiratory failure | |||||||
Yes | 85 (25.5%) | 248 (74.5%) | <.001 | 7 (5.306−9.234 | <.001 | 5.285 (3.861–7.234) | |
No | 251 (4.7%) | 5.126 (95.3%) | Control group | Control group | |||
Cardiovascular dysfunction | |||||||
Yes | 37 (68.5%) | 17 (31.5%) | <.001 | 38.994 (21.703–70.062) | <.001 | 26.514 (13.412−52.415) | |
No | 299 (5.3%) | 5357 (94.7%) | Control group | Control group | |||
Liver dysfunction | |||||||
Yes | 26 (25.2%) | 77 (74.8%) | <.001 | 5.770 (3.646−9.131) | <.001 | 3.963 (2.172–7.232) | |
No | 310 (5.5%) | 5,297 (94.5%) | Control group | Control group | |||
Kidney dysfunction | |||||||
Yes | 57 (31.1%) | 126 (68.9%) | <.001 | 8.509 (6.086−11.898) | <.001 | 3.641 (2.407–5.508) | |
No | 279 (5%) | 5,248 (95%) | Control group | Control group | |||
Nervous system TB | |||||||
Yes | 14 (9.3%) | 136 (90.7%) | .069 | 1.675 (.955−2.936) | .003 | 2.655 (1.403−5.024) | |
No | 322 (5.8%) | 5,238 (94.2%) | Control group | Control group | |||
Miliar TB | |||||||
Yes | 22 (9.5%) | 210 (90.5%) | .017 | 1.723 (1.094−2.712) | .079 | 1.634 (0.945–2.826) | |
No | 314 (5.7%) | 5,164 (94.3%) | Control group | Control group | |||
Transplantation | |||||||
Yes | 12 (12.4%) | 85 (87.6%) | .006 | 2.305 (1.246−4.262) | .005 | 2.678 (1.353−5.302) | |
No | 324 (5.8%) | 5,289 (94.2%) | Control group | Control group | |||
Sex | |||||||
Man | 137 (4.1%) | 3,180 (95.9%) | .060 | .788 (.615−1.011) | .726 | 1.046 (.812−1.348) | |
Woman | 124 (5.2%) | 2,269 (94.8%) | Control group | Control group | |||
Young adult | |||||||
Yes | 24 (1.4%) | 1,691 (98.6%) | <.001 | .168 (.11−0.255) | <.001 | .235 (.141−0.394) | |
No | 312 (7.8%) | 3,683 (92.2%) | Control group | Control group | |||
Adult | |||||||
Yes | 43 (3.4%) | 1,239 (96.6%) | <.001 | .490 (.353−.679) | <.001 | .429 (.278−0.663) | |
No | 293 (5.9%) | 4,135 (93.4%) | Control group | Control group | |||
Elderly | |||||||
Yes | 197 (13.3%) | 1,282 (86.7%) | <.001 | 4.524 (3.609−5.671) | <.001 | 1.977 (1.446−2.702) | |
No | 139 (3.3%) | 4,092 (96.7%) | Control group | Control group |
The average age of the patients who died in our study (73.19 years) was higher than that of the patients who did not die (55.24 years) (P < .001). On the other hand, those over the age of 74 years were twice as likely to die. On the contrary, those under the age of 45 years had almost 7 times higher probability of survival (Table 1).
To conclude, our studies show that the mortality due to OA TB in Spain is falling, and this may be due to the reduction in its incidence. The zero rate of mortality among children stands out, as does the fact that mortality is associated with age above 74 years and certain comorbidities, especially cardiovascular dysfunction, malign neoplasia and acute respiratory failure.
Conflict of interestsThe authors have no conflict of interests to declare.
This work received no grant or financial support.