The role of Diabetes mellitus comorbidity on Tuberculosis treatment outcomes in Nepal: A prospective cohort study

Aim: The Objective of this study was to assess the effect of Diabetes Mellitus (DM) on treatment outcomes of tuberculosis (TB) patients in the Central Development Region of Nepal. Methods: A prospective cohort study was conducted in central Nepal. The study population of n=408 was consecutively recruited from treatment centers of all 19 districts of central Nepal. The TB cases (n=306) and TB with DM (n=102) cases were followed up for the estimation of blood glucose level, HbA1c level, and sputum examination on 2, 5, and 6 months after TB treatment started. The Generalized Estimating Equation (GEE) was performed to identify the risk ratio among TB and TB with DM cases on treatment outcome. Results: Our study identified that the magnitude of treatment failure among the tuberculosis cases was 19.7% (95% CI: 17.44-21.95). The GEE analysis observed that factors associated with the treatment failure had uncontrolled DM (HbA1C ≥7 % ) (adj.RR=5.24, 95% CI: 2.58-10.62, P value <0.001 ), aged ≥ 45 (adj.RR= 6.13, 95% CI: 2.55-14.76, P value <0.001), had inadequate financial status (adj.RR= 2.33, 95% CI: 1.07-5.06, P value 0.033) and had prior TB (adj.RR=2.33, 95% CI: 1.09-4.97, P value 0.028) respectively. Conclusion: The prevalence of worsening TB treatment among patients with TB and DM was significantly higher than those who had TB only. Poor glycaemic control, increasing age, inadequate financial status, and previous history of tuberculosis were strong predictors of worsening tuberculosis treatment outcomes.


Introduction
Nepal is passing through a phase of epidemiological transition from a higher prevalence of communicable diseases to noncommunicable diseases (NCDs).It is currently suffering from a double burden of diseases.Various small studies from different parts of the country on diverse populations have shown varying prevalence rates of type 2 diabetes mellitus ranging from 6.3 to 8.5%.However, a systematic review and metaanalysis from 2000 to 2014 illustrate that the prevalence of type 2 diabetes reached a minimum of 1.4% to a maximum of 19.0%.The pooled prevalence of type 2 diabetes was 8.4% (95% CI: 6.2-10.5%).In addition, prevalence of type 2 diabetes in urban and rural populations was 8.1% (95% CI: 7.3-8.9%)and 1.0% (95% CI: 0.7-1.3%),respectively (1).TB patients beginning TB treatment with Diabetes comorbidity experience tardy regain of body mass and haemoglobin (2,3), which are essential for the profound recovery from both diseases (4).In addition, previous studies have revealed that Diabetes may weaken sputum conversion (2,(5)(6)(7), cure and increase the risk of relapse (4,8,9), and raise the risk of anti-TB drug resistance as well (10,11).Furthermore, a recent study observed that TB with DM was associated with some critical sociodemographic factors, including age, unemployment, literacy, and polluted environment (12).A study from Nepal has also illustrated the prevalence of Diabetes among Tuberculosis patients, which was 9.1% among older age TB patients, tobacco users, people with high-income status, and a history of high blood pressure (8,13).Therefore, this present study aimed to identify the role of DM on the treatment response among TB patients in the Central Development Region of Nepal.

Methods
A prospective cohort study was conducted by administrating a structured questionnaire among the TB and TB with DM cases.
In addition, we examined their blood glucose level, HbA1c level, and sputum grade 2, 5, and 6 months after starting treatment of TB to identify the treatment outcome of TB.

Study population
A total sample of 408 patients was estimated to be required by taking reference of risk ratio 2.93 of non-cure rate (28.65%) among the TB DM cases from a previous study (5).408 TB cases were collected from the National Tuberculosis Centre and treatment centers of all 19 districts of the (Central Development Region) CDR, Nepal, and were examined for a blood glucose level.After that, 102 TB patients with Diabetes were considered cases, and 306 non-diabetes Tuberculosis patients were considered controls.Since six patients died and one got severe cancer during the study period, finally, 401 TB cases were followed up to identify treatment outcomes.Simultaneously, Body Mass Index (BMI) and blood glucose level were measured, and the sputum status was checked to determine treatment outcomes in two, five, and six months after starting treatment.The respondents who met the essential requirement for their family within the year of treatment were considered to have a good financial status.

Data Collection
The data was collected by using a structured questionnaire (Annex I).In addition, signs and symptoms of the tuberculosis cases were documented before the beginning of TB treatment, and additional history was obtained for the presence of DM or DM treatment, previous TB treatment, TB contacts, other comorbidities, and medication used.Similarly, the patients were followed monthly during the intensive phase and bi-monthly after that.History, physical examination, blood testing, and microscopic examination were repeated after the intensive phase (at two months), five months, and at the end of treatment (at six months).TB programspecific definitions were used to classify

Statistical analysis
All collected data were entered in Epi-Data (Version 3.1) and transferred to STATA (Version 13, Stata Corporation, College Station, TX USA) for analysis.The data collected after the respondents' follow-up in 2, 5, and 6 months were analysed using GEE to identify the risk ratio amongst the TB and TB with DM cases on treatment outcomes.

Results
Table 1 illustrates the characteristics of TB and TB with DM patients at 2, 5, and 6 months of the treatment period.The respondents (TB and TB with DM) aged ≥ 45 years old seemed to raise the non-curing rate from 43.30% at two months, 45.88% at five months, and 51.90% at six months of treatment.In addition, the tuberculosis patients living in rural areas were observed to fail sputum conversion at six months of treatment compared with two months of treatment, i.e., 12.50% to 11.49%, respectively.The increasing blood glucose levels among the TB with DM cases at 2, 5, and 6 months of the treatment period revealed a curing failure with 41.94%, 64.00%, and 62.96%, respectively.Similarly, an uncontrolled HbA1c level is also responsible for increasing the no-curing rate from 2 months (29.03%) to 5 months (56.00%).On the other hand, a raising BMI (Body Mass Index) level from low to normal was observed that enhanced the TB curing rate from 2 months (56.25%) to 6 months (58.82%) (Table 1).

Risk factors of the failure of treatment outcome: using the Generalized Estimating Equations model (GEE)
In this study, we analysed the risk factors for failure in treatment outcomes using the GEE model for repeated measures of the outcomes.

Discussion
The prevalence of DM with TB will continue to increase, given the projected global expansion of DM.However, to our knowledge, this is the first study on this region that has been performed to identify the treatment outcomes of tuberculosis cases associated with DM.The data presented in this prospective cohort study show that a total of 401 respondents from both TB and TB with DM cases were observed until the last month of the tuberculosis treatment period, of which 79 or 19.7% (95% CI: 15.79-23.61)were not cured.A study conducted in Taiwan observed similarly 17.0% of treatment failure (14).A study conducted in the urban setting of Indonesia revealed that 22.2% of the DM patients with TB had positive sputum smears after the treatment period (15).In Pakistan, nearly one-third (33.6%) of study participants who had a previous history of tuberculosis was not cured (16).In addition, more than two-thirds of the respondents were delayed in seeking treatment (≥ 7 days).In addition, most of the respondents who failed to cure visited more than two health facilities for their diagnosis.This might be due to some health providers being unable to diagnose TB as well as Diabetes in the same place.
In our setting, we determined the role of DM and other risk factors on TB treatment outcome 2, 5 & 6 months of comprehensive treatment of our tuberculosis cohort.The sputum conversion guides the duration of TB treatment and infectivity of the patient but delayed conversion is also associated with an increased risk of relapse.While most studies outside the Middle East ( 16) have shown no relationship between DM and conversion at the end of 2 months, we considered a more extended observation period of 6 months.Up to one-third of the world's population is infected with Mycobacterium tuberculosis; however, not all of those infected develop active TB because, usually, the immune system contains the germ.However, in some people, the bacteria remain dormant.They could become active, causing disease at later stages, especially those with risk factors such as old age, Diabetes, and other immunosuppressive treatments (7).So, after controlling the confounding factors, uncontrolled DM and five more risk factors showed an effect on the failure of TB treatment.The respondents who had uncontrolled DM with ≥7 % of HbA1c on two months of treatment were more than five times at risk of failing therapy.A systematic review found that uncontrolled DM (HbA1c ≥7) was a significant risk factor for positive sputum culture after two months (17).Another multicentre study conducted in South Korea revealed similar findings (18).Therefore, close monitoring of blood glucose and clinical conditions of TB patients with DM during the treatment period is crucial (19).Respondents aged ≥ 45 years had a greater risk of deteriorating TB treatment outcomes.A similar result has been observed by studies conducted in Indonesia (15), Taiwan ( 14 inadequate financial status was also associated with failure of treatment.However, a study conducted in Kuala Lumpur, Malaysia, revealed no significant difference in the economic situation between both groups (2).Furthermore, history of prior tuberculosis is doubling the effect of the non-curing rate of tuberculosis, supported by a study conducted in Malaysia: the authors observed that patients with a previous history of tuberculosis treatment were found to be three times more likely to have sputum smear nonconversion compared with those without prior exposure to tuberculosis (2).So, the reason might be a previous infection may induce initial cavitation and increase the extent of residual lesions of the lung (20).

Conclusion
This study outcome was a stepping-stone towards getting free of TB despite being diabetic.Our study observed that poorly controlled DM, increasing age, inadequate financial status, and previous history of Mahato RK, Laohasiriwong W, Koju R. The role of Diabetes mellitus comorbidity on Tuberculosis treatment outcomes in Nepal: A prospective cohort study.(Original research).SEEJPH 2022, posted: 20 March 2022.DOI: 10.11576/seejph-5329 P a g e 4 | 13 treatment response and outcome.TB registerswere cross-checked to ensure the quality of collected data.

Table 1 . Characteristics of TB patients at 2, 5 and 6 months of treatment (n=401) Characteristics 2 Months 5 months 6 months Cured Not cured Cured Not cured Cured Not cured Gender
Mahato RK, Laohasiriwong W, Koju R. The role of Diabetes mellitus comorbidity on Tuberculosis treatment outcomes in Nepal: A prospective cohort study.(Original research).SEEJPH 2022, posted: 20 March 2022.DOI: 10.11576/seejph-5329 P a g e 5 | 13

Table 2 . Risk Factors of Failure of Treatment Outcome among TB Patients Using the Generalized Estimating Equations Model
Mahato RK, Laohasiriwong W, Koju R. The role of Diabetes mellitus comorbidity on Tuberculosis treatment outcomes in Nepal: A prospective cohort study.(Original research).SEEJPH 2022, posted: 20 March 2022.DOI: 10.11576/seejph-5329 P a g e 6 | 13

Can choose more than one options)
Mahato RK, Laohasiriwong W, Koju R. The role of Diabetes mellitus comorbidity on Tuberculosis treatment outcomes in Nepal: A prospective cohort study.(Original research).SEEJPH 2022, posted: 20 March 2022.DOI: 10.11576/seejph-5329 P a g e 12 | 13 Part B: