The cross-sectional study was conducted in District Sonepat of Haryana among TB patients for a duration of one year from the reception of ethical approval (31 July 2021-22). Ethical approval for the study was obtained from the Institutional Ethics Committee (IEC) via letter no: BPSGMCW/RC634/IEC/21 dated 26/02/2021.

Inclusion criteria: Consenting TB patients domiciled in the district, notified on the Nikshay portal, and treated through the public sector between 1^st April 2020 and 30^th September 2020 (2^nd and 3^rd quarters of the year 2020) were included in the study.

Exclusion criteria: Those treated at private health facilities, those whose information was not reliable, and those not traceable after three attempts to contact them telephonically were excluded from the study.

Sample size and sampling technique: The number of patients diagnosed and treated for TB during the study period, the second and third quarters of 2020 under the public domain in the District of Sonepat under its seven Tuberculosis Units (TUs) was 1,225. The calculated sample size with 6% precision at 95% confidence for the finite population of 1,225, assuming a maximum prevalence of 50% delays, was 219, which was 17.9% of the study population. Hence, the sample size was rounded off to 20%. Thus, 20% of the total notified cases were taken as a sample size to get a sample representative of the population, which was calculated to be 245 study subjects.

TU-wise data pertaining to the study subjects were obtained through the Nikshay Portal for each quarter from the District Tuberculosis Office of Sonepat, thereby generating seven lists, each containing the following details: name, NIKSHAY ID, date of notification, contact numbers, residential area, treating facility, and TU of the treating facility. Twenty percent of the total participants were sampled from each of the lists at a class interval of five. In each case of exclusion, the subsequent patient in the list was selected. The data was collected by the principal investigator over the period of eight months from the date of ethical approval of the study.

The study subjects were contacted by telephone. After a brief self-introduction, the participants were briefed about the study by the principal investigator. Their willingness to participate in the study was confirmed, and appointments for in-person interviews were scheduled. The participants were then visited by the researcher at their residence or a location convenient to them. Written consent was obtained, followed by an interview. Data were collected using a pretested semistructured schedule designed by the researchers. The schedule was finalised after verification by the faculty of the Department of Community Medicine at the Rural Medical College, where the researchers are based. The variables analysed included demography, past medical history, symptoms leading to TB diagnosis, place of diagnosis, time between symptom onset and diagnosis, duration taken for initiating therapy, and reasons for delays. The variables in the semistructured schedule were finalised based on the coefficient of reliability, calculated using Cronbach’s Alpha. Scores of 0.80 were determined through pretesting, which was conducted on 20 subjects sampled by convenience and excluded from the main study. In each case, the participants were interviewed by the principal investigator.

Some of the key operational definitions used in the study are as follows:

The collected data was entered into a Microsoft Excel spreadsheet. Mean and Standard Deviation (SD) were calculated for quantitative data, while percentages and proportions were calculated for qualitative data. The Chi-square test/Fisher’s exact test was used to determine the association and significance of differences using R software. A p-value <0.05 was considered statistically significant.

Attempts were made to telephonically contact 400 patients, of which 140 could not be reached after three attempts. Ten were found to be non residents of Sonepat, and five refused consent. Excluding these 155 individuals, a total of 245 study subjects were interviewed for the study.

The mean±SD age of the study participants was 38.29±17.37 years. The majority (109 or 44.49%) belonged to the 19-40-year age group, were mostly males (147 or 60%), hailed from rural regions (153 or 62.45%), belonged to the upper middle class (95 or 38.78%), were married (181 or 73.88%), were literate (222 or 90.61%), were employed (139 or 56.73%), resided in nuclear families (186 or 75.91%) with 69 (28.16%) having addictions [Table/Fig-1]. A major proportion of study subjects consumed a non-vegetarian diet (110 or 44.9%), 14 (16.7%) were ovo-vegetarian, and the rest were vegetarian. Only 53 (21.6%) of the study subjects were under treatment for co-morbidities.

Extrapulmonary TB was diagnosed in one-fourth (60 or 24.49%) of the study subjects, and co-morbidities in 53 (21.6%) of them. Around 23 (9.39%) reported treatment interruption, 33 (13.47%) needed hospitalisation, and 19 (7.76%) of the patients were deceased at the time of contact [Table/Fig-2]. ‘Diagnostic Time delay’ was observed among 86 (35.10%) of the participants, with a 30-day median delay. Those who were tested on time mostly got tested on the 10^th day. ‘ATT initiation delays’ were observed in 51 (20.82%) of the patients, with a four-day median delay, and 21 (8.6%) were initiated beyond one week of diagnosis. Cough (148 or 60.4%) and fever (154 or 62.9%) were observed to be the most common presenting symptoms, followed by symptoms that cumulatively constitute ‘others’ (33.9%), namely, swelling in the neck, pain abdomen, haemoptysis, and chest pain.

Distribution study subjects=diagnosed clinically [Table/Fig-3]. Distribution of patients according to place of diagnosis is shown in [Table/Fig-4]. ‘Being diagnosed for other diseases’ 70 (28.6%) and ‘ignoring symptoms’ 28 (11.4%) constituted the major reasons for ‘Diagnostic time delay’. Non availability of drugs 11 (47.83%) and allergy to drugs/discomfort 07 (30.4%) formed the major reasons for treatment interruption [Table/Fig-5].

[Table/Fig-3]:

Distribution of TB patients according to their modality of diagnosis*.

(n=245) (*Multiple options)

[Table/Fig-4]:

Distribution of TB patients according to their place of diagnosis (n=245).

*Primary Health Centre, ^#Community Health Centre, ^District Hospital

The likelihood of diagnostic time delay increased with age (from 08 or 29.6% to 15 or 44.1%), was more prevalent among males (55 or 37.4%), rural populations (60 or 39.2%), the lower middle class (07 or 54%), married individuals (66 or 36.5%), illiterates (11 or 47.8%), those staying alone (03 or 60%), and individuals with addictions (29 or 42%) [Table/Fig-1]. Diagnostic delays were more common among those with extrapulmonary TB (24 or 40%), co-morbidities (28 or 52.8%), treatment interruptions (11 or 47.8%), individuals needing hospitalisation (17 or 51.5%), and those who were deceased (10 or 52.6%) [Table/Fig-2].

Therapeutic delays were observed more frequently among the age groups of 19-40 and 41-59 years (23 and 17 or ~21% each), males (35 or 23.8%), rural residents (38 or 24.8%), the lower middle class (04 or 30.8%), employed individuals (35 or 25.2%), illiterate individuals (08 or 34.8%), those staying alone (02 or 40%), and individuals with addictions (20 or 29%) [Table/Fig-1]. Directly Observed Treatment Short-course (DOTS) centres served as the most common source for 89.4% of the study subjects. Therapeutic delays were observed more among those with co-morbidities (12 or 22.6%), treatment interruptions (10 or 43.5%), individuals needing hospitalisation (11 or 33.3%), and among those who had discontinued therapy or were deceased (5 or 26.3%) [Table/Fig-2].

Approximately 15 (8.1%) of those with pulmonary TB (185 or 75.5%) did not follow-up with sputum examination post-therapy, of which 7 (46.67%) missed their follow-up as they were unaware or not informed about it [Table/Fig-6].

The proportion of Pulmonary TB in the current study was three fourths (185 or 75.5%) of the study subjects, which was similar to the observations of the Central TB Division of India in 2016 [12]. . This indicates that the trend of disease burden did not change much during the initial pandemic period. Currently, mass awareness and training campaigns [10] are being conducted in the public domain and health sector levels [11] to ensure that those developing symptoms such as cough, night sweats, fever, and loss of appetite for 14 days get tested for TB [8,9]. Considering the time for possibilities of unavoidable circumstances, the maximum acceptable time for suspecting TB was set at 15 days of symptoms. Thirty-five percent of the patients were tested beyond 15 days, with a mean of 30 days. This, when compared to a median of 50 cumulative days of diagnostic delay in 2014 [13] and 32.5 days in 2019 in Ballabhgarh [14], indicates that the efforts exerted towards the control of TB are starting to show favourable outcomes. However, the possibility of lead time bias cannot be ruled out, given the major focus of both TB and COVID-19 on the respiratory system during the time period. In this study, ‘ignorance of symptoms’ was the most common reason for patient-origin diagnostic delays, as found in studies conducted elsewhere in the country [15-17]. Delays related to diagnosis at healthcare facilities were observed to be much longer than patient-related reasons [14], even in the prepandemic period. The finding that secondary health centres were more frequently utilised for diagnosing reflects the habit of patients to delay seeking healthcare until they approach a higher centre for medical intervention.

As per the latest NTEP guidelines, ATT is to be initiated for the patient immediately upon receiving a positive report for TB [10,11]. Considering a generous margin of one day for addressing unavoidable situations, the time for initiating ATT was fixed at one day from day zero of diagnosis. The proportion of therapy initiation delays in present study was 20.8% with a median number of four days (SD 3.9 days). This was higher than that found in 2019 in Ballabgarh (2.5 days SD 33.6 days) [14] and in the nationwide systematic review in 2014 (2.5 days) [13]. This indicates a persistent delay in therapy initiation, which may have been further prolonged due to the impact of COVID-19-induced social restrictions. Approximately 8.6% of patients reported at least one episode of treatment interruption due to ‘non-availability of drugs’ (47.8%) and ‘drug allergy/discomfort’ (30.4%). Side-effects or allergies to drugs were also the major reasons in the study conducted by Jaiswal S et al., indicating that addressing the side-effects of drugs could ensure drug compliance [18], regardless of the pandemic background for TB therapy [10].

Although it is recommended that all PTB patients who complete their ATT are followed-up with sputum examination [19,20] at the end of therapy, the proportion that missed the follow-up in the current study was 15%. With no studies done so far in India to estimate this proportion, it becomes challenging to understand if the pandemic played a role in the observed compliance. The closest available parameter to this among a neighbouring region was the ‘loss to follow-up’ in Delhi in 2016, which stood at 12% [21].

Upon observing the demographic profile of study participants, the findings related to the male gender were consistent with previous studies conducted in India [7,14,15,16,22]. However, caution should be exercised when comparing trends between genders, as a systematic review [23] has warned of a high likelihood for women, especially older age groups, to delay seeking healthcare and therefore not receive proper notification. [Table/Fig-7] shows the comparison of present study with different studies [13-18,22,23]. Despite Haryana being a high-income state [24], in the public sector, the vulnerability to PTB was higher among individuals from relatively lower socio-economic statuses in both rural and urban populations. The financial crisis induced by the pandemic may have forced patients to prioritise earning over health to address their urgent financial needs.

This hypothesis was supported by the findings of a higher prevalence of therapeutic delays among employed adult study subjects of working age. Although the proportion of illiterate study subjects was 9.39%, a higher percentage of them experienced delays in diagnosis and therapy initiation. Illiteracy could be associated with poor reception and understanding of medical advice, leading to poor compliance. Adults who were separated, widowed, lived alone, had one or more addictive habits, co-morbidities, or a past history of TB were more likely to experience both diagnostic and therapeutic delays. These demographic factors have been identified as risk factors for depression and anxiety in various studies conducted among patients and the general public during the initial phases of the COVID-19 pandemic worldwide [25-27]. It is highly probable that these factors are linked to hopelessness in patients, influencing their poor health-seeking behaviours and making them more prone to delays. The observation that individuals experiencing delays were more likely to require hospitalisation and had a higher proportion of deceased members indicates that delays are associated with unfavourable outcomes.

The trends among people who availed TB services in the private sector could vary. The use of classical delay pathways for drafting operational definitions could have added more value to the study. Recall bias and truth bias cannot be ruled out. This was a cross-sectional study, and hence, cause-and-effect relationships cannot be established.

In the said study period, more than one-fourth of patients who availed ATT from the public sector experienced a delay in their TB diagnosis, with an average delay of 15 days. Additionally, 20% of the study population was initiated on ATT by day four. The most common reasons for diagnosis delays were ignorance of symptoms and misdiagnosis of a different disease. Drug intake interruption for at least one day was attributed to drug unavailability and allergy/discomfort. Individuals belonging to the working age, males, those residing in rural settings, individuals from lower socio-economic statuses, separated or widowed individuals, illiterate individuals, those with addictive habits and co-morbidities, were observed to have a higher proportion of diagnostic and therapeutic delays. Patients whose diagnosis and therapy were initiated late were proportionately more likely to require hospitalisation.