Design

We adopted an exploratory, interview-based methodology. Approval for this research was obtained from the Ethics Committee of IBAB and the research was carried out in compliance with the Helsinki Declaration. Prior to conducting the interviews, a detailed questionnaire was constructed based on a review of the peer-reviewed and grey literature about telemedicine in India including publications of the World Health Organization, the mHealth Alliance (http://www.mhealthalliance.org/) and The Center for Health Market Innovations (http://healthmarketinnovations.org/). The questionnaire was validated during a few initial interviews and the final version considered details of (i) the onset of a programme, (ii) its implementation and scale-up, including issues related to cost, financing, business model, human resources, infrastructure, technology and evaluation and (iii) the enablers and challenges in running it.

Sampling

SJ sent email invitations to all 48 programmes in India listed at http://www.mhealthalliance.org/ at the time. Early interviews with those who responded led to other contacts. Only persons who had worked in organizations directly employing eHealth technologies were selected. Potential interviewees were invited to be part of the study that was outlined to them, and a positive response to the email invitation was taken as informed consent to participate. In all, we conducted 30 interviews, representing 28 organizations, with a few follow up clarificatory emails. Further attempts, over several weeks, to recruit interviewees did not succeed.

Data collection

The interviews were conducted by SJ between March and September 2012, inclusive. Interviews were semi-structured around the validated questionnaire (Additional file 1). Notes were taken during each interview or immediately thereafter. In order to address the confidentiality concerns of the interviewees, the interviews were not recorded. The interviews were 25 to 90 minutes long, with an average of 55 minutes. Eight interviews were carried out face-to-face, 21 by phone and one by e-mail. In each case, complete annonymity and confidentiality were assured. Data collection was stopped when all those who had agreed had been interviewed.

Data analysis

Descriptive analysis of the interview notes was done as follows: the replies of each interviewee were collated under the corresponding questions of the questionnaire and analyzed together in thematic clusters as in the questionnaire. Categorization of mHealth programs in India by the respondents turned out to be largely in accordance with that proposed by the mHealth Alliance [24] and this framework was adopted in this study. Other themes are presented as a direct compilation of the replies without an attempt to create a framework. Overall, an inductive analytical approach was adopted. This approach allowed us to reflect the rich diversity of experiences of the various mHealth stakeholders.

Nature of work

Technology

Challenges in eHealth

There are several challenges in this field of work, some pertaining to one or two programmes and others to several. These are listed below.

Donors: A major challenge concerns the tendency for donors to like the concept of mHealth and to require it of the NGOs they fund. The NGO hastily proceeds, without fully understanding the requirements on the ground. The programmes are therefore donor-driven rather than needs-driven. Donors may also prefer to give money for a specific disease, although patients require overall healthcare. Furthermore, donors (or clients or technology providers) sometimes prefer the latest technology, which can be a challenge to implement once users are comfortable with an earlier one.

Government: The GoI’s Planning Commission strongly recommends telemedicine. Nevertheless, there may be a lack of suitable policies, or lacunae in their implementation, that can prevent more effective telemedicine.

In a government hospital, for instance, neither is telemedicine mandatory, nor does a doctor earn more for this work. Therefore these doctors are often unmotivated to use this modality. In private hospitals, in contrast, a certain amount of van duty or other telemedicine may be made a mandatory part of the doctor’s job. In the largest government (satellite-enabled) programme which involves 425 set ups donated by ISRO, 85% are not in use due to bad management, insufficient incentives, or lack of technical staff or electricity. In this programme, although private hospitals constitute only 30% of the participating hospitals, they account for 70% of the tele-consultations, that are delivered on a non-profit basis.

Although there are examples of NGOs or private healthcare providers working with the government and doing a good job of creating telemedicine facilities and programmes, there can be several challenges in working with a government (at the state level or national). If a government agency is required to clear a programme or requires permission to be part of it, there can be large delays. More seriously, a change in the government may lead to a programme being axed. Also, by government guidelines, the number of people covered by a programme often serves as a surrogate for impact, which can be unsatisfactory. In terms of funding, there may be a good budget, but no assurance of how it will be spent. There needs to be a better matching of the needs and the financing, especially because government funding usually comes with the requirement that the money has to be spent within the year or be forfeited. Sometimes due to fear of misutilization much of the funds may remain unspent at the end of the year. Furthermore, government funding often targets secondary and tertiary care that requires hospitalization, instead of primary or preventive care, an exception being the ASHA programme.

In terms of health education, the multi-levelled administration can be inefficient in implementing complex programmes. Thus, a proposed 12-step process may be reduced to three steps by the time it is implemented, losing much of the meaning of the intervention. Yet another issue concerns corruption. By a minister’s own admission, preventive care projects are not lucrative, and therefore such projects tend not to be supported. Sometimes a state government issues a tender just for infrastructure, since this is where people can make illegal money, and does not plan for other elements of a programme.

Telemedicine Society of India (TSI): In 2003–5 TSI’s telemedicine hype lead to the creation of infrastructure and to ISRO donating bandwidth and equipment. Unfortunately, due to the inefficiencies of the public sector described above, the infrastructure has remained largely unused, leading to huge disillusionment in government.

Records and data: Many telemedicine projects are based on the electronic collection and storage of data. However the government either mandates a paper-based system or it fears that the people dealing with paper records will become redundant with a switch. Also, even if some data is collected electronically, the remainder has to be done by paper, lowering the incentive to shift to electronic records. Yet another challenge is that it is not trivial to collate data in a format in which it can be analyzed, and also map the data to the geographic area of coverage. Furthermore, as NGOs or governments use mobiles, they may suddenly have large amounts of data that they are not in a position to handle. They may then need to invest in human resources for data analysis.

Technology: There are several technological challenges in telemedicine. First, those who design the mobile applications may do so without much understanding of the ground reality, making them sub-optimal. Second, as mentioned above, there is a trend to constantly upgrade software while the existing one is still functional. Instead of implementing a project with the current software, money is spent on developing and piloting a new version. Low end mobiles are widespread, but cheap smartphones such as the open-source Android devices would be better suited for people with poor literacy, and for applications that employ a global positioning system (GPS) tracking function. However if one upgrades to a smartphone, one needs an application that may no longer run on a simpler phone. Thus there may be rapid obsolescence of the currently used hardware and software. Third, the novelty of the technology has required extended learning and optimization periods for both manufacturers and implementers, thereby slowing or decreasing its spread, especially in the early years. Fourth, the electric supply and network coverage can be poor, which is sometimes circumvented by the store-and-forward modality. If the government subsidised it, a satellite connection could be used where internet connectivity is absent, but no new satellite-based connections are being established. Fifth, companies that were pioneers in this area were forced to come up with their own technical solutions. Thus, hospitals may use proprietary software that is not Dicom compliant, and therefore not transferable across hospitals with different patient flows, employment structures and so on. As and when modalities for scale up across organizations are worked out, standardized software will become essential. Thus, there are several technical challenges, and yet human resources are a far bigger issue, as discussed further below.

Fragmentation of programmes: The most successful eHealth programmes are driven by organizations outside the government, whether for-profit or non-profit. However even if a programme covers most of a state, it will be offered by a single organization, possibly in partnership with the government. An exception is an ophthalmology programme, developed by a private hospital in one state, that the government of another state is copying. It is being rolled out as a public-private partnership, with the original hospital acting as a consultant to the private partner in the latter programme.

Thus, an over-arching issue concerning telemedicine, whether amongst those who run programmes, or those who develop the software or hardware, is that each initiative is running on its own, without the sharing of experiences. Also, there isn’t much research in this area. This wastes resources due to the duplication of effort, and also impedes scale up.

Communication: There are different kinds of challenges related to communication. First, since IT and medical professionals don’t really understand each other, developing appropriate software is not straightforward. Second, for the educational projects, it is important for beneficiaries to be engaged so that there is acceptance of the information. Not everyone believes the messages concerning washing hands and sleeping properly for instance. Thus, health has to be made an interesting topic if such educational programmes are to succeed.

Logistics: There are logistical challenges, especially when working in remote areas. (i) It can take three months to plan a van trip. The planning includes identifying NGOs and donors, educating patients about which conditions need medical attention and motivating them to attend the camp. The follow-up of patients is also non-trivial. A specific challenge is that the van returns to the city every evening. It would take separate incentives to get doctors to stay on the road overnight to enable them to reach the most distant patients. (ii) Patients may come to the health centre only in the morning, although the specialist is available for tele-consultation only in the late afternoons. (iii) After the tele-consultation, there can be the challenge of patients’ access to drugs. To deliver drugs to a rural area, one organization tied up with public buses, and logistics were worked out according to the bus schedules. Although the same organization also tried to work with unbranded generics, they found that the better distribution channels of branded generics was worth the extra cost.

Manpower: The shortage of skills is evident in a variety of settings, as follows. (i) Although supposedly educated up to middle school, some ASHAs are illiterate and can barely answer a call. Therefore learning how to use the phone can be a challenge. Their professional training may also be poor, leading to low quality counselling of beneficiaries. (ii) Technicians who work in the telemedicine centres are not skilled in computer use and need to be trained. This can lead to high training expenditure and staff shortages. (iii) A specific challenge in MDS is the narrative part of the autopsy where health workers need to make accurate notes based on what family members say. This requires a special skill. (iv) The NGOs and research groups in health and population studies, and in the government, may find technology challenging. There may be reluctance to learn something new, and even if provided a ready product they will not necessarily make full use of it. (v) A doctor needs special skill to interact with rural patients, being empathetic and humble, and discussing other things before health issues. A change in doctor can lead to a big drop off in the attendance at a rural facility if the new person lacks these skills. (vi) The doctor must also be flexible, and be able to function with the few facilities available in a rural setting. (vii) People shifting to other jobs is another persistent challenge. After some experience, a fresh medical officer will soon get a higher paying job elsewhere. Women birth attendants may get married, or pregnant, and leave. To tackle this issue, one organization only employs retired doctors, who move less often. (viii) Some of the challenges relate to working with technicians at a distance. Illustratively, one ophthalmology hospital does not allow a technician to dilate the pupil since he would be unable to deal with any side effects, and this limits the care that can be offered. (ix) Even staff in rural areas who have been trained may need re-training, sometimes weekly.

Consulting the independent specialist: A challenge specific to tele-consultations relates to the delay and hassle involved in operating the computer and other technology without assistance, waiting for the connection and so on. Further, billing and transferring money electronically means that doctors have to pay taxes on this income. Unless there are 30–50 consultations a day it is usually not worth the specialist’s time, and this may not have happened yet.

Attitude: Aside from skills, the attitude of everyone involved in telemedicine plays a very important role in its success. (i) Implementers need to have contagious enthusiasm for their projects. If they take the health workers into confidence at every stage of the project, the latter develop a much needed sense of ownership of the project. (ii) An ASHA is often more preoccupied with her own household than with her official work. (iii) A common problem is that a rural doctor may be too ‘proud’ to consult an urban specialist, despite patients’ requests. (iv) Users’ acceptance of both technology and mode of healthcare delivery is important, and convincing them may take considerable effort. (v) Services related to family planning, contraception and abortion cannot be implemented easily in South India, which is more conservative. Also there are more qualified allopathic medical practitioners available in the south, so a project cannot be run with the less qualified Registered Medical Practitioners who will be accepted in the north.

Finances and sustainability

One of the biggest challenges facing most of the programmes concerns funding and sustainability. The sources of funds (or other crucial support) for each programme are listed in Additional file 2. These are primarily philanthropic donors, government (that usually provides non-cash support), cross-subsidy or revenue, and rare cases of venture capital. The financial sustainability of each programme is also discussed. In terms of government-run or supported schemes (including those involving ASHAs), these will be sustainable so long as the government continues its support. In several cases, sustainability is dependent on patient volume. Thus there are geographic limitations on where such programmes can run. Low-cost efforts such as health information, where corporates see value in large-scale visibility, are sustainable due to the latters’ support. For-profit endeavours that depend on venture capital investment are yet to prove their sustainability, even in urban areas.

The costs for some aspects of the programmes are listed in Additional file 3. These include capital expenditure such as that for a van (Rs. 25–30,00,000), a stationary clinic (Rs. 1–2,00,000 for a basic clinic or Rs. 5–6,00,000 for an eye clinic) or a mobile phone (Rs. 4,500 for a mid-range model), and running costs such as doctors’ salaries (Rs. 6000 for a fresh graduate in a rural area), incentives to ASHA workers (Rs. 100 a month to bring patients to a tele-medicine centre) or mobile phone connection charges (Rs. 199 for a month’s rental for biosurveillance work). Since most of the programmes are run by non-proft organizations and are geared to poor patients, running costs tend to be a small fraction of what would be charged by a for-profit entity in an urban environment, for example.

A specific mention should be made of how doctors were paid for their involvement in eHealth. In the most successful programmes, run by large non profit hospitals, doctors were not paid beyond their regular salaries. In government hospitals, too, there was no separate fee to a doctor. In the less successful website-based consultation programs doctors were paid via bank transfers by the website.