The participants for this cross-sectional, exploratory study were recruited through consecutive sampling from both private and public sector health care facilities in Bangalore, Karnataka, India. The private facility was St John’s Medical College Hospital (SJMCH), Bangalore. This is a 1250 bedded, nonprofit, tertiary care teaching hospital that caters to patients largely from the South Indian states of Karnataka, Andhra Pradesh, and Tamil Nadu. The public health facilities involved were urban health centers in the vicinity of SJMCH in Bangalore. These public health facilities implement national health programs and provide health care at no cost to all those in need.

Patients suspected of clinically active TB are subjected to microbiological and/or radiological tests for confirmation. Newly diagnosed patients with TB are started on a Category I ATT comprising isoniazid (H), rifampicin (R), pyrazinamide, and ethambutol for 6 months. This includes an initial 2 months of an intensive phase of treatment with all the 4 drugs, followed by 4 months of a continuation phase of treatment with 2 drugs (H and R). Patients with relapse or drug resistance receive more complex ATT regimens involving injectable medications. ATT at the hospital is available either through (1) the DOT center, a public-private initiative that enables no-cost treatment through the RNTCP or (2) for a cost through the hospital’s pharmacy. Patients follow up with physicians every month routinely for health appraisals that include clinical examination, monitoring adverse effects, or prescription refills.

Diagnostic protocols at the public health care facilities are similar to those at SJMCH. However, all patients diagnosed with TB and treated at public health care facilities receive treatment only through the RNTCP. All patients within the RNTCP are expected to receive DOT.

Between February 2016 and December 2017, 185 patients with TB aged between 18 and 60 years, receiving treatment at the study sites, were enrolled in the study. Of the participants enrolled, 159/185 (85.9%) received ATT at SJMCH, whereas 26/185 (14.1%) received ATT at public health care facilities. Both newly diagnosed patients with TB and those already receiving ATT were included in the study. Patients who were seriously ill or those who did not understand the purpose of the study were excluded.

Subsequent to written informed consent, trained research assistants administered a questionnaire in the local language to the study participants. The questionnaire obtained basic sociodemographic information from the participants along with information regarding (1) the basic functionality of their mobile phones, (2) acceptability of delivering adherence support via mobile phones, and (3) the type of mobile phone intervention acceptable to them, such as SMS, voice call, interactive voice response system, or vDOT.

Data were analyzed using IBM-SPSS version 24. Frequencies, means, medians, SDs, and interquartile ranges were used to describe the data. The outcome variables studied were (1) the preference for voice calls compared with SMS reminders and (2) the preference for in-person DOT compared with vDOT. Some categorical variables that had multiple categories were converted into binary variables. Bivariate associations were assessed between demographics, clinical details, phone usage, and the preference-based outcome variables using the chi-square test and odds ratio (OR). Unadjusted logistic regression was used to derive OR for variables with more than 2 categories. Bivariate associations with a P value ≤.20 were included in an adjusted logistic regression model. Associations with P values <.05 were considered significant.

Ethical clearance for the study was obtained from the Institutional Ethics Committee, St John’s Medical College, Bangalore, India. Written informed consent was obtained from all participants after providing them with study-related information, either verbally or in writing, before administering the questionnaire.

A total of 185 patients with TB participated in the study. The mean age of the participants was 35.25 (SD 11.59) years. Of the participants, 114/185 (61.6%) were males, and 121/185 (65.4%) resided in an urban area. There were 44/185 (23.8%) participants on in-person DOT, 45/185 (24.3%) on self-administered treatment, and 96/185 (51.9%) participants for whom treatment was yet to be initiated. The demographic characteristics of the patients are presented in Table 1.

Clinically, 98/185 (53.0%) patients had pulmonary TB, and 159/185 (85.9%) were newly diagnosed patients with TB on category I ATT.

Of the 185 participants, 151/185 (81.6%) used a mobile phone. Among these 144/151 (95.4%) owned the phone, and 85/151 (56.3%) had used mobile phones for 6 years or more. Of those who owned mobile phones, 65/144 (45.1%) owned smartphones, and the rest (79/144, 54.9%) owned basic phones. The major reasons cited for not owning a phone included not needing a phone (19/34, 56%), inability to use a mobile phone (8/34, 24%), and financial constraints (3/34, 9%). Only 7/144 (4.9%) mobile phone users reported using a phone shared with other family members. Men were 4 times as likely as women to own mobile phones (unadjusted OR 3.816; 95% CI 1.747-8.338). Significant factors associated with phone ownership included education and a monthly income of 500 INR or more (Table 2).

Of the participants who used a mobile phone, 66/151 (43.7%) used the SMS feature, 72/151 (47.7%) used the alarm function, whereas 87/151 (57.6%) knew how to use the camera on their phone for photography and/or videography. Of those who used the alarm on the phone, only 2/73 (3%) used the alarm as a medication reminder.

Participants less than 32 years of age (adjusted OR 2.314, 95% CI 1.068-5.025) or those literate in English (adjusted OR 8.678, 95% CI 4.019-18.740) were more likely to use the SMS feature than their counterparts. In addition, those who were single (unadjusted OR 2.793, 95% CI 1.479-5.263), residing in an urban area (unadjusted OR 3.493, 95% CI 1.695-7.195), or formally educated (unadjusted OR 15.012, 95% CI 3.489-64.591) were more likely to use the SMS compared with those who were married, were from a rural area, or were not formally educated, respectively (Table 3).

Of the 185 participants, 182 (98.4%) agreed to receive health information on their mobile phones. Topics that the participants preferred included information on available medications, advances in TB management, and medication reminders. Participants also requested communication with health care provider, motivational health messages, specific diet, and prevention of TB as additional features (Figure 1).

In response to specific queries about the preferred mode of health communication, 137/185 (74.1%) of the participants chose voice calls over the SMS (40/185, 21.6%). In addition, 8/185 (4.3%) participants preferred either. Most of those (43/48, 90%) who preferred the SMS requested to receive them in English. On the contrary, most of those who preferred voice calls (127/137, 92.7%) requested for communication in regional languages. Of the 185 participants, 78 (42.2%) chose to receive reminders as often as they required to take their medication, whereas the rest preferred reminders either once a week or less frequently. Similarly, with regard to the timing of the reminders, 123/185 (66.5%) preferred reminders at specific times, whereas 22/185 (11.9%) preferred reminders just before they took their medication. The remaining 40/185 (21.6%) participants were willing to receive reminders anytime as they were free at home. Overall, 183/185 (98.9%) of the participants did not perceive the reminders as an intrusion of their privacy.

Preference for voice calls was significantly associated with age, marital status, literacy in English, type of TB, and the ability to use camera on their phones (Table 4).

Most of the study participants preferred vDOT over the conventional in-person DOT. This preference was associated with being male, residing in an urban area, and being formally educated. Other factors such as literacy in English, the ability to use SMS and phone camera, although associated with preference for vDOT in bivariate analyses, were not found associated with this preference in an adjusted logistic regression model. Clinical characteristics were not associated with the preference for the mode of communication or adherence monitoring strategy (vDOT or in-person DOT; Table 5).

Barriers such as stigma, medication side effects, and transport to the health care facility can affect adherence to ATT. In this light, exploiting the pervasiveness of mobile phone technology to overcome these barriers and support medication adherence is a promising solution. Motivational health messages and customized medication reminders via mobile phones are some interventions designed to support adherence to ATT [24]. Although several studies have explored the use of mobile phone interventions for adherence support in chronic infectious diseases such as HIV and TB, not all have shown favorable results [29,35]. Factors such as the complexity, personalization, and mode of communication could affect acceptability, uptake, and success of mobile phone interventions. We therefore chose to explore the acceptability of mobile phone adherence support interventions for ATT and identify the characteristics of such interventions, which the patients with TB would prefer, before developing such an intervention.

In this study, most participants were willing to receive adherence reminders via mobile phones and did not consider such interventions as an intrusion of their privacy. Randomized controlled trials have shown the effectiveness of SMS reminders in HIV infection and malaria in sub Saharan Africa [36]. However, the effectiveness of mobile phone reminders for antiretroviral treatment support in India was questionable [35]. Nevertheless, given that treatment for TB is for 6 months compared with HIV infection where the treatment is lifelong, mHealth interventions for TB are likely to face lesser intervention fatigue and are worth exploring. Although patients with TB from Salem in Tamil Nadu state, South India, considered communication via mobile phones useful, they preferred in-person contact with health care providers. The study, however, did not assess the preferred mode or type of ATT adherence support [37].

A study from Lesotho, Africa, reported a high uptake of SMS interventions in HIV/TB patients [28]. On the contrary and consistent with other studies from South India, most of our study participants preferred voice calls in comparison with the SMS [38]. The young, the employed, and the educated participants were more likely to use the SMS for communication. These participants probably preferred reading an SMS text as opposed to answering a phone call as it saved time and attracted lesser attention when received. Given the limited literacy in English, the roman script was used for both English and regional languages in SMS communication; it is not surprising that SMS communication was less popular than voice calls both in this study and in the literature in the Indian context [37,38]. However, with the availability of regional language options in mobile phones, developing interventions in regional languages is an option worth exploring. Furthermore, given that some of the participants were not literate but could use the basic functionality on mobile phones, interventions that use videos and pictures with limited requirements for literacy are an option.

Asynchronous vDOT is an accepted alternative suggested to circumvent the barriers to in-person DOT [39]. Studies have shown that vDOT is more confidential, easy to use [40], and allows health care providers to efficiently monitor a larger number of patients at a distance when compared with in-person DOT [31]. Barriers to vDOT include interruption of data connectivity [31], loss or theft of phone [40], and an inability to confirm that medicine was actually taken in certain settings [41]. Over three-fourths of the participants preferred vDOT to in-person DOT, despite no experience with the intervention, citing vDOT as an optimal solution for saving time and money or minimizing hospital visits.

Few participants expressed concern with sending their video to the health care provider. Fear of disclosure of their illness to family, unknown people watching their videos, fear of the videos getting published via social media, and discomfort with video-recording themselves were reasons expressed for the concern. Therefore, counseling the beneficiaries of the measures taken to safeguard their videos along with reinforcing the importance of adherence to ATT is essential to ensure the uptake of vDOT. Smartphone apps, although nonintrusive, were found beneficial in the management of HIV infection despite their limited functionality [42].

In this light, the abundant features for patient-centric mHealth interventions for TB can be explored. Most of the participants suggested newer apps incorporating disease-related information and behavior change communication, which could be incorporated into existing or newer apps. The concept of photovoice [30], where patients cured of TB shared their treatment experiences and replaced health care personnel, can be considered an option. Photovoice can also be incorporated into vDOT for health education and communication with patients.

Given that standard TB care in India is based on geographic location, most of the patients were from urban areas and therefore more representative of urban patients with TB. Also, as one-third of the patients received ATT through the RNTCP (at SJMCH or at public health care facilities), the study mirrors the public-private mix in TB care in India. Furthermore, as many of the study participants were newly diagnosed with limited treatment experience, their opinions are also likely to change with treatment. This, along with the limited experience of the participants with mobile phone interventions may mean that opinions may change with actual interventions. The limitations in estimating a sample size and the nonprobability sampling technique used may affect the generalizability of our findings. Nevertheless, the results of the study cannot be undermined as they inform patient-centric intervention design in contexts little exposed to mHealth interventions.

In addition, in terms of implementation reality, factors such as poverty, smartphone penetrance, internet access, and level of education necessary for using mobile phones will have to be addressed in an integrated manner to maximize the potential benefit of vDOT in ensuring treatment success in TB.

This study sought to assess whether communication via mobile phones could be an acceptable form of health care delivery in the context of patients with TB. We found that adherence reminders and information disseminating apps were acceptable in the management of TB. Contrary to the popularity of SMS-based reminders elsewhere globally, most of the study participants preferred voice calls. Efficacy of mHealth interventions could be improved when components that enable the inclusion of all demographic groups are incorporated along with enabling customizations to individual needs. Given the popularity of voice calls, interventions should include a voice component along with various language options in the Indian context. Although facing interceptable barriers such as privacy and stigma, vDOT as an alternative to DOT appears to hold promise in the Indian context. The effectiveness of mobile phone apps such as vDOT may therefore be worth exploring in the Indian context, while ensuring privacy and confidentiality of the end user.