HbA1c admission levels do not predict outcomes in pulmonary tuberculosis
HbA1c levels on admission do not predict therapeutic outcomes in in-hospital, smear-positive, culture-proven, HIV-negative tuberculosis, a new study shows.
Medical and clinical records of 239 patients (62 percent male; median age 73 years) diagnosed with pulmonary tuberculosis were used for this study. Confirmation of tuberculosis was done through cultures and participants were consistently monitored for HIV. Those who had culture-negative tuberculosis, were HIV positive, had multidrug-resistant tuberculosis and were taking antituberculosis treatment were excluded.
After baseline blood, sputum, HbA1c and X-ray assessments, participants received isoniazid, rifampicin, pyranizamide and ethambutol (HRZE) or isoniacid, rifampicin and ethambutol for 2 months followed by isoniazid and rifampicin. In-hospital death, becoming noninfective and time to live discharge were the main outcomes of the study.
The median HbA1c and blood glucose levels were 5.9 percent (5.6 to 6.6) and 109 mg/dL (96 to 133), respectively. Of all participants, 81 percent (n=193) were discharged alive and 86 percent (n=206) became noninfective before death.
Admission HbA1c levels was above 7.0 percent in 44 patients. Over the course of 3 months, there were clear and significant downward trends in HbA1c levels 1, 2 and 3 months after admission(p<0.001 for months 1, 2 and 3).
There was no apparent correlation between the time to discharge alive and the levels of HbA1c on admission according to the Spearman’s rank correlation test (r=0.17), Kaplan-Meier curves (p=0.431) and multivariate Cox analysis (hazard ratio [HR], 1.03; 95 percent CI, 0.89 to 1.20; p=0.659).
Similarly, there were no significant correlations between the time to becoming noninfective and levels of HbA1c on admission (r=0.17; p=0.113; HR, 0.93; 0.80 to 1.06; p=0.277 for every 1 percent increase in HbA1c).
Finally, HbA1c levels were not significantly associated with in-hospital mortality (p=0.631) and time to in-hospital death (p=0.427).