The fall in rates of tuberculosis (TB) in many countries has been accompanied by the concentration of cases in the social risk groups including homeless persons.
Aim: Comparison of TB features in homeless persons and in non-homeless patients.
Methods: TB cases reported to National TB Register in Poland in whom information about the social status was available (the data about the social status were collected obligatorily in the years 2004-2013 only) were analysed. The results of DSTs were obtained from laboratory records and were available for the cases reported since 2010. Treatment outcome after 12 months was analysed for the cases registered between 2004-2012. The significance (Si) of the differences in proportions was assessed with chi-square test. P<0.05 was regarded as statistically significant. Test F was used to evaluate the significance of differences of the means of age. The multivariate logistic regression models were applied to find out the independently operating determinants of not achieving of success of treatment.
Results: 2,349 homeless persons (HP) and 72,989 other patients (OP) with TB were included. In the group of HP, there was a greater proportion of males in comparison with OP (90.5% vs. 66.3%) (Si). The mean age of HP was 49.8 years (SD±10.9); of OP-52.9 years (SD±17.5) (Si). 16.6% of HP and 10.4% of OP were previously treated for TB (Si). The previous treatment was adequate in 62.2% of HP and in 85.8% of OP (Si). Pulmonary TB was in 98.0%, extrapulmonary TB in 2.0% of HP and, respectively, in 92.5% and 7.5% of OP (Si). Pulmonary TB was confirmed by culture in 76.3% of HP and in 64.5% of OP (Si). Sputum smears were positive in 70.7% of HP and in 62.5% of OP (Si). Caseous pneumonia occurred in 2.7% of homeless subjects and in 1.1% of OP (Si); infiltrative TB in 95.5% of HP and in 97.5% of OP (Si). Resistance to isoniazid was observed in 2.9% of HP and in 3.1% of OP; to rifampicin in 0.0% of HP and in 0.2% of OP; to isoniazid and rifampicin in 0.4% of HP and in 0.8% of OP. These differences were not Si. Treatment success rate among HP was 44.1%; default rate 24.8%; 4.0% of HP died from tuberculosis; 3.2% died from other causes; 5.2% were transferred and their outcomes were unknown; 0.4% were still on treatment; 0.4% had treatment failure; in 17.9% of HP results of treatment were not available. Among other TB patients the rate of success was 66.8%; of default--8.8%; 1.9% died from TB; 3.1% died from other causes; 2.6% were transferred; 0.5% were still treated; 0.2% had treatment failure; in 16.1% the results were not reported. Differences between both groups were Si, except for the category "died from other causes", "still on treatment" and "treatment failure".
Conclusions: In the group of homeless persons with tuberculosis phenomena which indicate a delay in diagnosis of disease and in some cases the possibility of a recent infection and also unfavorable treatment outcomes occur in a greater proportion than among other patients. Targeted screening for TB and directly observed treatment could likely improve the epidemiological situation of tuberculosis in the population of homeless.