Lynda Taina Abicher
Saint Damien Hospital
Title: Risk factors for bad evolution in children hospitalized at Saint Damien hospital for pulmonary tuberculosis, January 2015 to December 2017.
Biography
Biography: Lynda Taina Abicher
Abstract
Background
In 2017, tuberculosis was still a major public health concern in some countries, as Haiti. In 2016, Haiti counted 20 000 cases of tuberculosis among which 2 600 pediatric cases. In children the diagnosis is very challenging and relies on a careful history of exposure, clinical examination and relevant investigations. Unfortunately, bacteriological confirmation is not always conclusive, then the diagnosis stays presumptive: based on the clinical evaluation and/or exposure’s history, creating a delay of treatment. Now begs the question: What is the influence of the diagnosis methods on the prognosis? This study objective is to analyze the association of presumptive diagnosis and other risk factors with a poor evolution of pediatric inpatients admitted for tuberculosis at Saint Damien hospital.
Design
A case-control study was conducted from January 2015 to December 2017. Cases were any child less than 14-year-old treated for tuberculosis, dead or without any clinic improvement after the intensive phase of the treatment. Control represented any child with a satisfying clinical improvement. ODD ratio with a confidence interval of 95% were calculated and integrated into a multivariate analysis. Confirmed diagnosis was defined as a positive bacteriological exam or a positive Gene-Xpert for tuberculosis. Statistical tests chi2 and Fisher were realized for categorical variables.
Results
The study included 167 children: 85 cases and 31 controls. Mean age was 7-year-old. 38 children had a confirmed diagnosis: 31 (26.72%) with a positive bacteriologic exam and 7(6.03%) with a positive gene-expert. For children with presumptive diagnosis, the exposition source was identified in 48.26% cases, while the tuberculin skin test was conclusive for 62.93%. A presumptive diagnosis with negative bacteriological exam did not show any relationship with the prognosis. Thus, absence of BCG immunization (OR= 4.31, (IC95%= 1.34-13.8) and Clinical severity at admission [OR=3.56, (IC95%= 1.35-9.34) were the risk factors associated with bad outcome.
Conclusions
The study results do not show any relation between the mode of diagnosis and patients improvement. Nonetheless, some measures of presumptive diagnosis such as tuberculin skin test should be used more efficiently in order to optimize not only the diagnosis in itself, but also the evolution by reducing the delay to be treated. The use of the Xpert MTB test should also be increased. BCG vaccination should be optimize in the country as a measure helping to decrease tuberculosis infection and bad outcomes. Clinical status at arrival should also be considered to decrease mortality rate associated with tuberculosis.