A community based cross-sectional study was conducted. Study was conducted from April 1^st to June 30^th, 2019 in Prakasam District, Andhra Pradesh. Patient identifying information was removed prior to analysis. Research protocol was approved by Institutional Ethical Review Committee of Government Medical College, Ongole, Prakasam District (Approval number- IEC/G.M.C Ongole/2019/9). Drug resistant TB cases were confirmed by Cartridge Based Nucleic Acid Amplification Test (CBNAAT) to diagnose TB and Rifampicin resistance within 2 hours, RTPCR (Real time polymerase chain reaction) for rapid confirmation of diagnosis of Mycobacterium Tuberculosis (MTB) and MDR-TB and Line Probe Assay (LPA) to detect low level Isoniazid resistance. Most of the cases were detected by using CBNAAT technique, followed by RTPCR and LPA [7]. RTPCR and CBNAAT tests were performed by the technicians of the District TB Control unit under the control of District TB Control Officer, Prakasam district. For LPA, the sputum samples were sent to Intermediate Reference Laboratory, Visakhapatnam.

According to census 2011, population of Prakasam district was 33,92,764 [8]. It was standardised to the year 2019 as 36,85,898 by adding population based on population growth rate of Andhra Pradesh i.e., 1.08% per year.

The quarterly reporting system used in RNTCP enables analysis of cohorts of patients i.e., a group of patients who were registered for treatment in a specified area over a specified period of time. Under RNTCP, specified areas are TB unit, district, state and country whereas specified periods of time are four quarters of a year and one calendar year itself. A quarter is a three month period with the first quarter starting on 1^st January of the year and one year is divided into four quarters (1^st, 2^nd, 3^rd and 4^th). This information helps national, state and district levels to assess the performance and monitor the implementation of the programme.

Inclusion criteria: All the registered drug resistant TB cases in Prakasam district who gave consent during the study period.

Exclusion criteria: Drug resistant cases that did not give consent and were not available at the time of study.

The registered drug resistant TB patients were included in the study in Prakasam district in the second quarter in between April 1^st to June 30^th 2019. They were interviewed according to predesigned and pretested questionnaire. The questionnaire was developed by the principle investigator after consultation with experts in public health and district TB officer. All the patients were informed about the purpose of the study in Telugu language and consent was taken from them. Study tools were pretested and reliability and validity were checked. A pilot study was conducted on 20 participants to analyse the study tool and discussed with district level officials, to know validity of study tool. It revealed that the study tool is a valid schedule and it had content construct validity. Reliability of the tool was tested by measuring Cronbachs alpha value. Cronbachs alpha value of the tool was measured as 0.92, indicating good internal consistency in the data and measuring, what ought to be measured.

The investigator posed questions to the participant in Telugu and filled the answers himself. Questions were probed to the participants for eliminating memory bias or recall bias. Answers to the questions were verified by physical verification of the health records such as treatment cards and medical records. The questionnaire included various risk factors like demographic, socio-economic factors (Modified BG Prasad Socio-economic Classification) [9], literacy, income, overcrowding, personal hygiene, malnutrition, history of smoking, alcohol, place of diagnosis, clinical type, disease complications, place of TB treatment, family support, treatment with how many drugs, adherence to drug intake, drug toxicity, psychological factors, awareness about sputum disposal and MDR-TB.

Data was entered into an Excel spread sheet and double checked for errors. Data was analysed by using Epi Info and Excel software and depicted in proportions and cross tabulations. Chi-square test was applied to test the relationship of categorised independent, moreover dependent variables. A p-value of <0.05 was deemed statistically significant.

Prevalence of drug resistant TB in Prakasam district was calculated as 5.64 per one lac population. In the second quarter of the year 2019-2020, i.e., from April 1^st to June 30^th, 88 new cases were identified. Incidence of drug resistant TB in Prakasam district was calculated as 2.38 per one lac population. Out of 209 TB patients, registered 88 patients were having drug resistant TB in Prakasam district in the second quarter in between April 1^st to June 30^th 2019. Almost all the cases were pulmonary TB except one case. Majority had resistance to Rifampicin (79.55%) followed by Isonicotinic acid hydrazide or Isoniazid (INH) drug. Only one patient had drug resistance to both the drugs [Table/Fig-1].

[Table/Fig-2] depicts the risk factors associated with drug resistant cases. Among all the drug resistant cases, patients belonging to 20-40 years of age group (42%) and 40-60 (38.7%) were more in number. Less number of were from the age group below 20 years of age. Almost three fourths of cases (76.1%) were males. Most of the drug resistant cases belong to skilled workers group (71.6%) by occupation and had education up to school or no education. Most of the cases were from middle class (45.5%) according to modified BG Prasad classification, followed by upper middle class (25%). Overcrowding present in one-third of cases (36.4%). History of migration present in most of the cases (85.2%). Half of the patients had smoking habit and one-third had alcoholism. Drug abuse was observed in two cases.

Almost 70.5% cases had no previous history of TB and one-fifth of cases were relapses. Almost half of the cases had breathlessness and only one-tenth of the cases had haemoptysis. Eighty six per cent of them had history of retroviral disease and one-fourth had Diabetes mellitus. Half of the patients were ill nourished. One-fourth of patients had psychological problems. Most of them had awareness (76%) about MDR-TB complications. Adherence to treatment (93%) and family support (99%) was present in most of the cases. Drug toxicity was observed in 54.5% of the cases. After taking the treatment, two-thirds of cases (67%) became sputum negative and one-third of cases were still positive.

In [Table/Fig-3], association of history of different co-morbid conditions with type of drug resistance was tested by using chi-square test. Only two co-morbid conditions listed in [Table/Fig-3] i.e., history of drug toxicity and presence of psychological illness showed statistically significant association with type of drug resistance. History of drug toxicity and history of psychological illness were significantly high in Rifampicin resistant cases when compared to INH resistant cases.

In the current study, it was observed that males are at higher risk of developing drug resistant TB. Many are in the working age group i.e., above 20 years of age. It can be linked with migration history and occupation history. The people who went on migration from their native area for work purpose are at risk of developing drug resistant TB. Their low educational status is not enough to take preventive measures to get protected against drug resistant TB. In the current study, the problem was seen mostly in middle socio-economic class population according to modified BG Prasad socio-economic scale. History of smoking and alcohol are also contributing risk of drug resistance TB.

These findings are consistent with the study conducted by Workicho A et al., [10]. Their study also stated that young patients as seen in present study, living in a household with only one room, history of previous treatment and being HIV infected were found to be independent predictors of MDR-TB [10]. Raaji J et al., in their study also noticed that MDR-TB was more common in 26-45 year age group and males as noticed in present study, previously treated TB case, positive history of contact with MDR-TB patient, patients previously on non-DOTS treatment, patients with associated co-morbidities and in substance abuse [11].

Pradipta IS et al., in their systematic review and meta analysis stated that the adverse drug reactions, patients 40 years and older, HIV positive were the risk factors as seen in present study and also non-adherence, previous TB disease and treatment, unemployed, lacking health insurance were the risk factors [12].

Stosic M et al., in their study, identified six significant independent risk factors, monthly income of the family, defaulting from treatment, stigma associated with TB, subjective feeling of sadness, use of sedatives as seen in the present study (psychological illness) and chronic obstructive pulmonary disease [13].

The current study revealed that, most of the cases had no previous history. They got drug resistant TB cases directly. It is an alarming signal. But, it can be linked to the history of retroviral disease and immune compromised status. Because of the presence of retroviral disease, patients got more severe form of TB directly or they are getting relapses. Almost half of the patients were ill nourished indicating the vicious malnutrition and infection cycle.

The present study also revealed that awareness about the disease is good and almost all patients adhered to the treatment and had good family support. It indicated the changing scenario in the society, which will give a hope to reach the goal to end TB by the year 2035. But, these results are in contrary to the study findings of some other study of Pradipta IS et al., conducted in different settings where non-adherence to anti-TB (anti-TB) medication is a major risk factor for poor treatment outcomes in the study [14].

Johnson J et al., in his study said that adverse drug effect as seen in the current study and default, travel to a different place; symptom relief and cost of treatment and history of relapse are the factors [15]. Drobniewski FA and Balabanova YM, in their study revealed that the reasons are failure of the patient to adhere to therapy (patient non-compliance), failure to modify the drug regimen in light of drug sensitivities as seen in the present study and failure to deal with non-compliance are the factors [16].

The present study revealed that, one-fourth of patients had psychological problems. As TB is a chronic disease, patients are developing psychological problems over a period of time. Thus, psychiatric care should also to be provided to the patients for psychological rehabilitation of the drug resistant TB patients. Rumende CM in his study also said psychiatric illness as was seen in the present study and poor compliance with the treatment are important factors in the development of acquired drug resistance [17].

In the present study, the history of drug toxicity and presence of psychiatric illness had significant association with drug resistance. Rifampicin resistance was significantly high in the patients who had history of drug toxicity and who had psychological illness when compared to others. Jaber AAS and Baharudin I, in their study revealed that 26% of patients have drug toxicity as seen in this study also and were lost to follow-up because of drug toxicity [18]. Jaber AAS et al., in their study said that 32.7% had side effects associated with treatment as observed in the present study [19]. Goble M et al., has also shown drug toxicity as seen in the present study. Fifty-one (30%) of 171 patients had adverse reactions that led to the discontinuation of one or more anti-mycobacterial medications. Twenty of these patients had serious toxic reactions to one or more drugs during the first three months of therapy [20]. Prasad R, in his study stated that non-compliant patients due to monetary lack, lack of information, side effects of drugs as seen in present study and social myths and misconceptions, often do not adhere to treatment [21].

Chuchottaworn C et al., in their study among the 145 patients with pulmonary MDR-TB, the DST results showed drug resistance to INH and RMP in all patients as seen in present study, EMB-resistance in 30 (20.7%) patients, and SM resistance in 88 (60.7%) patients [22]. The prevalence of present study (5.6/1 lac population) was compared with other studies in Andhra Pradesh. The prevalence was 1.7 to 11% in the category 1 and 2 patients in the study by Suryakumari V et al., [23] and 15% in the study conducted by Armstrong E et al., the prevalence was 15% and 2-3% in the study of Chadha SS et al., risk factors were compared with present study [Table/Fig-4] [24,25].

Trend of drug resistance over 10 years is depicted in some studies. Kumar A et al., in their study stated that overall, MDR-TB cases gradually decreased from 2009 to 2015 in both new and previously treated cases in North India, but still, the prevalence rate was higher in North India than National prevalence rate of MDR-TB [26]. The National prevalence of MDR-TB is 3% in new cases and 12%-17% in retreatment cases [27]. Sharma N et al., showed a declining trend during the period of observation [28]. Mourya AK et al., in their study showed an increasing trend of MDR-TB in both new and previously treated cases in contrast to previous cases [29]. Suen SC et al., in their study stated that as transmission-generated MDR-TB becomes a larger driver of the MDR-TB epidemic in India, rapid and accurate MDR-TB diagnosis and treatment will become increasingly effective in reducing MDR-TB cases compared to non-MDR-TB treatment improvements [Table/Fig-5] [30].

The current study was a cross-sectional study. Incidence of drug resistant TB cases was calculated based on the secondary data collected from District TB centre. Drug resistant TB cases diagnosed in private clinics and unregistered TB cases might be missed while calculating incidence. Influence of risk factors can be assessed better in case control study. Further larger studies needed to confirm results.

Male persons having age above 20 years belonging to middle class (45.5%) and migrating for work are at the risk of developing drug resistant TB. Their low educational status is preventing them to take precautionary measures to get protection against TB. Most of the patients are developing drug resistant TB directly which is an alarming signal. Patients were also developing psychological complications, which needs to be addressed properly for better outcome of the patients. Further larger studies are needed to prove the association of risk factors with drug resistant TB.

Open data policy: The data was uploaded in github.com and made available in public domain after anonymising the collected data (https://github.com/DrSudhaKumari/MDRTB).