METHODS: All patients seen at Lanakila Chest Clinic with HbA1c ≥ 6.5% at the start of TB treatment were offered a TB clinic-based diabetes management program. This novel program included frequent glucose testing, HbA1c testing every three months during treatment, diabetes clinic referrals, and individualized diabetes education provided in TB clinic. A total of fifty patients with diabetes who started TB treatment between January 2013 and December 2014 and with at least two recorded HbA1c tests were evaluated. Paired t-tests were used to compare initial and final HbA1c readings, independent two-sample t-tests and one-way ANOVA were used to determine if changes in HbA1c were significantly different based on initial HbA1c reading. Chi-square and univariate linear regression was used to determine the association between change in HbA1c and diabetes education received.
RESULTS: Initial mean HbA1c was 8.7% (95%CI: 7.0, 10.1); 42% of the cohort had an initial HbA1c of 6.5-7.5%, 26% had initial HbA1c of 7.5-9.0%, and 32% (19.1, 44.9) had initial HbA1c ≥9.0%. A significant change in mean HbA1c of -0.5 %(p=.039) was observed from initial to final reading for entire cohort. Individuals with initial HbA1c ≥ 9.0% had a greater decrease in mean HbA1c (-2.0%) than individuals with initial HbA1c 7.5-9.0% (0.0%; p=.002) and initial HbA1c 6.5-7.5% (0.3%; p ≤.001). TB clinic-based diabetes education effort did not correlate with changes in mean HbA1c during treatment.
CONCLUSIONS: These results demonstrate that, for a majority of TB-DM cases, diabetes control can be significantly improved during TB treatment. Individuals who begin TB treatment with poor glucose control demonstrated the most benefit during TB treatment. Multiple factors likely contribute to overall improved glucose control for patients with TB and diabetes. Although the initial evaluation of our TB clinic-based education program did not correlate to improved glucose control, this finding is likely affected by inconsistent implementation of the diabetes education sessions, incomplete charting of diabetes education efforts in TB clinic, and small sample size. Further research with improved adherence to a more structured diabetes education program is necessary to properly evaluate the potential impact of diabetes education in TB clinics.