Smart Cards for All: Digitalisation of Universal Health Coverage in India

Digitalised Public Health Services in India

To understand smart cards based universal health care it becomes pertinent to discuss the context of digitalised public health services in India. In recent years health care services have been subjected to large-scale use of biometrics based chip smart cards, considered as a key technological medium for health accessibility (Gold, 2013; Marohn, 2006). With the health policy of 1983, government of India for the first time outlined the role of technology by highlighting the importance of medical information systems for managing health care. ⁸ However, it took more than fifteen years to implement a prominent project on digital technology and health care, with a first telemedicine network for connecting three hospitals in Chandigarh, Delhi and Lucknow in the year 1998. The health policy of 2002 witnessed a subsequent shift from the policy of 1983 with emphasis on technology associated with an increasing role of the private sector in health care. This policy shift played a major role in extending the use of technology and digital technologies in particular with various projects on health information management systems in the first half of the decade of 2000 (Goldman, 2007; Negaard & Shaaban, 1989). Reflections on strong usage of biometrics and smart card technologies become more visible from the last years of the decade of 2000, as shown by the working group report on ‘Entitlement Reform for the Empowerment of Poor: The Integrated Smart Cards’ submitted on January 2007 as part of the eleventh five-year plan. ⁹ One of the most striking aspects of this report was the concept of ‘multi-application smart card’ to be used for various social sector schemes among them health care being a crucial one. The policy recommendations of this report became turning point in future construction of technological infrastructure of identification for health care, and also for other public services such as socio-economic welfare entitlements.

In this context of increased role of private players in health care and constant policy emphasis on using biometrics based smart cards, the RSBY scheme emerged in the year 2008 as a pioneer cashless and paperless scheme. One of the first aspects of RSBY implementation was the creation of a large digital database by enroling beneficiaries into this data driven health care program. This database became a large repository of personal information of people who enroled in the scheme, their photographs, fingerprint scans and address details aggregated for five members of a family. With the help of a software that is connected to this database, empanelled hospitals fetch beneficiaries information when they come for medical treatment. This practically takes place when the beneficiaries are enroled, have received the card and have taken it to an empanelled hospital at time of medical attention. However, from the very beginning this database is catered with incomplete and obsolete data, making it partial and leading to the exclusion of many people.

Partial Databases

The making of RSBY smart card and the enrolment process is based upon the selection of target BPL families and twelve other categories of beneficiaries. The BPL data list is provided by the district collector’s office to the labour department, which in turn forwards it to the State Nodal Agency (SNA) of RSBY. ¹⁰ The SNA locks the data and then forwards to the RSBY central agency, which generates a list of data having a 17 digit encrypted unique relationship number (URN) for each BPL family. The locked data list with URNs is then sent back to the insurance company. ¹¹ It contains an identification (ID) and password to access the data list from the central agency’s platform. After getting access, insurance company provides it to the technology provider (TPA). As per the data list, technology providers create a roadmap for enrolment of beneficiaries to RSBY. ¹² However, representatives of insurance companies, technology providers and government officials mentioned that this list was fraught with problems of redundancy as well as missing or incorrect data. ¹³ Indeed, the data relies on a thirteen years old census done in 2003–04 ¹⁴ . In thirteen years, kids of the census became adults, the elderly died and many families migrated and moved up or down the poverty line. Apart from BPL data, data for other categories were also difficult to access, as mentioned by a government official: ‘taking data was a major challenge, data categories like rickshaw pullers, street vendors, domestic workers, beedi workers, we did not have data on these categories’. ¹⁵ On a similar line, an insurance company representative said ‘some woman who came for enrolment, her age was 31 or 32, but the age during the census was 15 years or 16 years [..], we could not even update the age’ ¹⁶ .

During the enrolment process, biometrics of five family members were taken and matched with name, age and address details as per the pre-defined URN list. Frequent mismatching of data and biometrics was highlighted by the technology provider who looked over enrolments in several districts of Jharkhand: ‘there are problems with mismatch of biometrics, […] there is mismatch with URN data […] people in the villages work and their hands get rough and it becomes difficult to get their biometrics’. ¹⁷ Multiple data anomalies could explain the low enrolment of families as explained by the TPA representative ‘there is a lot of problem with data here; […] data has hampered the enrolment percentage, so enrolment is 40–50%’. ¹⁸ This low percentage of enrolment from the given BPL data list shows that more than 50% of the target population is not covered. According to several testimonies many households are missing from the BPL list used for enrolment, leading to their exclusion from the digital database and the RSBY scheme. This shows that right from the beginning of data acquisition and enrolment of beneficiaries the smart card is riddled with non-existing, fragmented and obsolete data.

At the beginning of the scheme in Jharkhand, people came to the enrolment camps to complain about the fact that they were not included in the enrolment list, or that their name or age was wrong in the file, but no mechanism was put in place to update the database and then improve the accuracy of data on site. Being outside the database, rightfully or not, lead to bear one’s own health care expenses. But people who were enroled in the database have also faced this situation. Getting the smart cards made and received did not mean that the beneficiaries accessed health services as per the card at time of medical requirement. The politics of (un)use of the smart cards brings into picture some further challenges faced by beneficiaries.

Smart Cards’ (Un)Use

Enrolment to RSBY was relatively poor in Jharkhand, with less than 50% of the targeted people actually subscribed to the scheme. For this scheme ‘Around 40 lakhs beneficiaries were identified, out of which 18 lakhs people got smart cards, so we had 45% enrolment rate in 2014–15’. ¹⁹ People in the villages got the feeling that if many people were enroled a few of them only used the scheme, as this community health worker explained: ‘In my knowledge only 2–3 people have used it, but enrolment was done by many people. There were around 70–80 families from my village who were enrolled in the scheme’. ²⁰ Low usage of the scheme was also confirmed by the state officials: ‘Out of 18 lakhs people enrolled, 74.7 thousand people used the cards to get treated in 2014–15’. ²¹ . The difference between enrolment data and usage data of RSBY is particularly striking, less than 1.8% of families targeted by the scheme actually used it in Jharkhand.

The first explanation provided by the insurers and state officials for the low usage of the smart card was lack of awareness of individuals who were supposed to benefit from RSBY: ‘It has been observed that beneficiaries got their smart cards but they are not availing the benefit of that because of lack of awareness’. ²² As explained by a representative from insurance company, some people had a very different understanding of the use of the RSBY smart card: ‘in some places the beneficiaries gave the smart card to the doctors in hospital and asked for ₹30,000 in cash thinking that it was like an ATM card and they can withdraw money’. ²³ For the majority of people who understood the purpose of the scheme, knowing how and where to use the cards was a challenge, as this patient explained: ‘I took it (RSBY smart card) to one hospital in Ranchi but they did not accept it. This hospital was not in the list. I did not read the details in the leaflet that was provided, but also it’s difficult to find the hospitals’. ²⁴ This testimony also highlights the low rate of hospitals empanelled in the scheme. Indeed, the number of empanelled hospitals and usage data are correlated; a second explanation for low usage of the card.

The low usage of the scheme is also directly associated with the one year validity period of the smart cards, which constitutes the third explanation for low usage of the scheme. Several patients and health workers explained to us that even when people knew about empanelled hospitals and travelled to these specific hospitals to use the smart card, they could not use it because the card was already expired. As told by this health worker: ‘we took the card to the hospital; it was R. hospital since it was in the list. So when we were admitting the patient they said yes the card will be accepted, but when the patient was getting discharged, so we asked whether they used the card and the money in that. They said no, the card cannot be used, the card is not valid, the card has lapsed’. ²⁵ Some hospitals told us that it was the case for the majority of patients that came to their hospitals: ‘This (expired smart cards) happens in 90% of the cases’. ²⁶

If the majority of RSBY smart cards were ‘unused’ in Jharkhand, the several cases of effective usage pointed towards several difficulties during the three important steps of hospitalisation process: registration, blocking and discharge. Registration, blocking and discharge procedures are done through several technologies like smart card readers, biometrics scanner and software where specific treatment packages are selected. ²⁷ When the smart card is accepted at the hospitals’ registration desk, it is punched into the smart card reader along with biometrics verification of the beneficiary family with the thumb readers. ²⁸ At this point, issues of biometrics mismatch can happen. Several interviews revealed that there were frequent biometric mismatch, for almost one person in ten. In that case, another member from the enroled family had to be called, otherwise the registration would not be completed and the smart card will not be used.

If registration is completed, the blocking of medical packages has to be done in the first 24 hours. The packages are selected from a list of six different categories: 1/non-surgical (medical) treatment in general ward; 2/admission in intensive care unit (ICU); 3/surgical procedure in general ward (not specified in packages list); 4/surgical procedure in general ward (specified in package); 5/maternity benefit package; 6/ in-patient care (IPD) and day care procedures with list of 1,090 medical conditions. ²⁹ However patients where only asked to provide the smart card and verify their biometrics for match, and did not know anything about packages that were booked. They were never told about the categories of packages and also about the amount taken on their card.

This non-communication about the blocking process and the amount retained on the card is further connected with possible manipulations of the card. It came into picture that many private hospitals blocked ‘non-required’ packages, as explained by this technology provider: ‘they booked an appendicitis but the patient was suffering from fever. For the fever the charge should be ₹500 but they blocked the package for ₹12,000’, ³⁰ and confirmed by several doctors we met. Another possible manipulation concerned the blockage of packages without any hospitalisation in reality or blockage of several days of hospitalisation, whereas the patient was discharged earlier. We encounter several cataract patients that spent less than 24 hours in private hospitals where a three days hospitalisation package was blocked indeed. This is done to block the maximum amount on the card, even when beneficiary did not require or use the service in reality. This could be correlated with testimonies provided by several stakeholders, on how in rural areas private hospitals flung rapidly after this scheme started to be implemented. ³¹ If people are not aware of the blocking process and the money retained on their card, those manipulations are easier to perform. These cases of manipulations also reveal the lack of monitoring of the smart cards usage. Talking about monitoring problems, one of the doctors of an empanelled private hospital mentioned, ‘in the case of private insurance when somebody is admitted, […], or has swiped the card then some representative of the company within 24 hours visits you to check whether it’s a genuine case or not, but in RSBY there is no monitoring’. ³² Further mentioning about the details, the doctor specified that ‘In the case of RSBY they come after one week or maybe sometime after the patient gets discharged, […]. Suppose I have done operation of hydrocele, but the booking is done for Hernia, […] you will come in 1 months or 15 days, the treatment is done and the patient is discharged, when someone comes to monitor, that person will be able to see paper not the manipulation’.

All the above-discussed issues highlight that this smart card based health care scheme has unfolded complex social realities, which help to better understand the dynamics of digital technologies. After discussing the social aspects of databases and smart cards, it is crucial to discuss the PPP through which this scheme is being implemented and administered.

The Double Winners

Insurance companies are seen as the main actor of RSBY but also the major ‘winner’ in this partnership: ‘the largest share of benefit is being received by CholaMandalam’. ³⁴ This perception that the partnership is a double win for private players comes from different pieces of evidence. In our study as well as in other research done on RSBY, two kinds of stakeholders were identified as benefiting from the scheme: insurance companies and private hospitals (Dasgupta et al., 2013; Shoree et al., 2014).

Indeed, if enrolment to RSBY in Jharkhand was low (around 45%), still 18 lakhs 65 thousands smart cards were made the last year of the scheme, which means that more than 18 lakhs premiums were paid to insurance companies. This constitutes a first obvious win for insurers that collected premiums for so many previously uncovered families. As explained by this district official, enrolment process was the most important part for insurers: ‘Insurance company guys used to be in high spirits when they were doing enrolment and making their cards, since more cards made more money for them’. ³⁵ The validity of the coverage will then play an important role in the ability to use the policy for enroled individuals. One insurance company explained that the card is valid for one year, but that in fact the patient can benefit from it for eight months because the enrolment period can take up to four months: ‘I will tell you this in short and brief, in insurance, a policy is for 12 months, the maximum upper limit is that the last beneficiary who has got the card, that beneficiary should get minimum 8 months policy, this is per IRDA guidelines’. ³⁶ According to both technology provider and district officers in charge of enrolment in the villages, the enrolment period was always more than six months: ‘The government directive is to close the enrolment in 3–4 months but it does not happen like this and the enrolment takes 6–7 months’ ³⁷ ; ‘The CholaMandalam company had a lot of benefit in this scheme, if we see that government is trying to implement it from April to next year 31st March, and also the money has been paid and also renewal and all has been done. They have become like the middlemen, they are taking money from the government and they are issuing the card in September, and the card has been delivered in October or November’. ³⁸ So according to our interviews in most of the cases, after receiving the card patients had only three to four months to use it before expiration.

The second win for insurance companies resides in the fact that empanelment of hospitals and claims settlements were also done by them. Firstly, it was not in their interest to empanel a lot of hospitals, as this state official explains: ‘Most of the time insurance companies are not showing their interests in empanelling the hospitals, because the more hospitals would be empanelled and accordingly they would have to settle more claims. They are discouraging the hospitals, not to create too much RSBY users’. ³⁹ So the less hospitals they would empanel, the less claims they would have to settle. In the next part of the article we will detail the processes of claim settlement that also favours insurers in many ways.

The second winners of RSBY scheme are private hospitals. In Jharkhand, 95% of reimbursement of packages (claims) were asked from private hospitals. In our study, as well as in another research conducted in Chhattisgarh, small private hospitals have been identified as central providers of care for RSBY beneficiaries (Dasgupta et al., 2013b) especially in rural districts, as this community worker explains: ‘That card was mainly used in the private hospitals and also more particularly in A. Hospital (a 10 bedded private hospital run by one MBBS doctor)’. ⁴⁰ When RSBY was launched, a hospital to be empanelled had to fill specific criteria—more than 10 bed, 24 hour MBBS doctors on site, lab facilities—but these criteria were considerably altered in states like Jharkhand and Chhattisgarh were few hospitals could meet them, as this state official highlights: ‘Major challenges were to have good hospitals, we didn’t have enough good hospitals to empanel, so we relaxed our criteria to empanel more hospitals, earlier we wanted only 10 bedded hospitals with all facilities. But we relaxed it to 5 bedded, and even to 2 bedded hospitals, so that people in those areas could benefit from the scheme also’. ⁴¹ Even in Ranchi the capital of Jharkhand we discovered that activity of small private nursing homes was considerably boosted by RSBY, as the head of this small hospital in Ranchi explains: ‘we use to have many patients from RSBY per month, if there were 30 patients per month in the hospital then 20–22 patients use to be from RSBY […] we also did this work for our advertisement’. ⁴²

Apparently small private hospitals benefited from the scheme, which brought a lot more customers to them, as this technology provider told us: ‘the private sector has a strong motivation for working for this scheme, they are aware that they can benefit from this scheme. Even to an extent many private hospitals have been opened to make benefits out of this scheme. […] People have opened hospitals in two rooms, they don’t even have MBBS but yet the registration and license is given to them that’s one of the biggest question mark’. ⁴³ However, all the private hospital managers we met argued that ‘the money never use to come on time’ ⁴⁴ and that many of their most recent claims are still pending after two years, because the insurer refused to clear them: ‘many hospitals have bills pending of up-to 2–3 months, even my hospital has pending bills, I have ₹21 Lakhs pending’ ⁴⁵ ; ‘99% of the hospitals have outstanding dues with CholaMandalam, some have 5 Lakhs, some have 2 Lakhs’. ⁴⁶ So even if private hospitals have benefited from the scheme at the beginning of it, the wind changed for them, several lost the ‘game of claims’ with insurers and some even shut down because of it.

The Game of Claims

The representatives of insurance companies gave different explanations to the non-settlement of hospitals claims. The late uploading of usage data by hospitals is among the first reason evoked by the different stakeholders. If data were not uploaded within the 24 hours of the discharge of a patient, the claim would be rejected by the insurer as explained by this state official: ‘Connectivity is not there, even in Ranchi district you have those problems. So hospitals couldn’t upload the data within the stipulated time frame of 24h. So very frequently insurers refused their claims with the comment “delayed uploading of data”; they said the hospitals are not respecting the delays’. ⁴⁷ The process of uploading the data in the servers is crucial for claims recovery. When the patient is discharged, a set of all three slips; registration, blocking and discharge along with all medical records is sent to the insurance company. This has to be matched and verified with upload of hospitalisation data in the server, on this the employee of a private hospital mentioned, ‘if the data is not uploaded in their sever, then there would be a mismatch since if there is blocking and the data of all the procedures is not uploaded then there would be a data mismatch, so the upload has to hit the server’. ⁴⁸ This was mentioned as a crucial part for claims to be processed, failure of which would lead to pending payment of claims. ⁴⁹

The second major reason for non-settlement of claims is related to misuse of packages by hospitals or even accusation of fraud. In those cases, the insurer decided that treatment that hospitals are claiming to be reimbursed for were not done properly or even not done at all and these hospitals could be de-empanelled for that reason: ‘One hospital in D. district, was de-empaneled for fraud issue. A patient was coming for diarrhoea and discharged after 3 days, but the hospitals blocked higher package in fact, different from the treatment, like hysterectomy for instance. That hospital was very frequently blocking high packages. That’s why it got de-empanelled’. ⁵⁰ The insurance company had two systems to identify misuse of the scheme by hospitals, through the data stored on their server or through random checks in suspicious hospitals as explained here: ‘We do checks in the software, if the hospital software is saying that today 10 patients have been admitted (blocking) then we check it. We can generate all that detail, we can take details of the beneficiary, we can take all the treatment documents, OPD paper, X-Ray paper. But we take the documents and then we try to assess using the documents. We can see common things, if the patient is admitted for severe fever, and the fever details is 98 or below 100 and if they have admitted in ICU then that creates a doubt element’. ⁵¹ But insurance staffs are not really qualified to be able to ‘separate the grain from the chaff’ online as well as offline, as they acknowledged themselves: ‘But then we are not doctors, so we cannot justify whether the patient needed ICU treatment or general ward treatment’. ⁵² However, claims that the insurance companies supposedly found had problems or was suspicious were not reimbursed.

The last reason for non-settlement of claims is because insurers did not get their full premiums paid the last year of the scheme, as stated here: ‘If you are giving me the premium then I will cover the risk, if not then we cannot provide the cover. We did not get money from the government so that’s why we could not pay to the hospitals, and this is because there were some political issues’. ⁵³ Central government that covers 75% of the amount paid its share but the state of Jharkhand that covers the rest refused to pay for the last month of the scheme because of pending vigilance enquiry regarding insurers practices for RSBY, as explained by this state official: ‘As per MOU, they (insurers) had to clear the claims of hospitals within one month. But they didn’t pay those claims; because they said there premiums were not paid by the state. They have not been paid their entire premiums because of the pending enquiry. Earlier they were paid some premium amount, but they said they paid to the hospitals because they had their premiums, but now they have no more money to set the claims’. ⁵⁴

If the majority of stakeholders we met had the impression that the scheme benefited to the insurance companies, they also clearly stated that this partnership was not a win for the government and public services: ‘They (insurers) took the money from the government, the money for the premium amount but they never paid back the claims amount which was suppose to be paid to the concerned hospitals’. ⁵⁵

The Losing Side

When RSBY started in Jharkhand, all the public hospitals from Community health centres to the state referral hospital were automatically empanelled. A strange idea knowing that in government facilities almost everything is free or at a very low cost for all the patients that are using those public facilities. Thus, an incentive mechanism was put in place so that public facilities could block the same packages as private hospitals and get reimbursed as well, as explained by this district official: ‘In the government hospital almost everything is free, so in that case there was some incentive for the government to put government hospitals so that they can get some reimbursements’. ⁵⁶ The majority of the reimbursed amount will go to improve infrastructures of the hospital and 25% would benefit directly to the hospital staff, which according to some officials was not enough to motivate the staff to use the card as explained by this official: ‘The public hospitals haven’t used the smart card. Only 25% of the claim amount will go to the hospital staff, individually they were not getting a big amount, the rest is going to the hospital account for the benefit of the hospital. So the staff there was not very eager to use the card’. ⁵⁷

In all the government hospitals we visited, managers and medical officers in charge (MOIC) said they used the scheme but did not received any money from RSBY. In one of the studied district, the accountants from the CHC and from the referral hospital with the highest number of claims of the district explained to us that they never ever received any fund from the insurer, even if they used the scheme and sent the data, a fact confirmed by the district accountant: ‘Even when all the paper work was ready and everything was fine, but still the money was not paid back. […] There is a society called HMS (Hospital Management Society) in which the money was supposed to come, but it never did. HMS is a society in all the CHCs in which government gives money for the betterment of the people/patients’. ⁵⁸

Even the technology provider working for the insurance company and supervising enrolment as well as claim data for RSBY in Jharkhand highlights the fact that public hospitals never got money out of the scheme: ‘Government focus was to promote and develop the public hospitals, but in this case all the funds were taken by the private hospitals, public hospitals didn’t get anything’. ⁵⁹ All those different actors highlighted the fact that even if the public hospitals used the scheme and claimed packages from the insurer they never received the claimed amounts for their staff or to improve their infrastructures as promised. Public hospitals did not win anything from the scheme and even lost patients for small private hospitals that were pointed out by several actors as low-performing and of low-quality in many ways.

For the government of India and the state of Jharkhand it was then a double loss. Firstly, because public infrastructures in Jharkhand did not benefited at all from the scheme and did not upgrade their facilities or staff thanks to the money injected through RSBY. Secondly, because insurers always kept the hand on data and adapted their premiums to the claim amounts they would have to disburse. Hospitalisation data were used by insurance companies to analyse the use of smart cards but also to evaluate districts performance and claim level. Thus, insurers knew where the treatments were happening and where packages were blocked and could fix premium rates according to this, as explained in detail by this technology provider: ‘using this data they change their premium rates, […] areas where the usage is high the premium rates would be more and where the usage is less premium would be a bit less, it is easy to know the usage data, the area and locations where it has been used more […], on this basis insurance company does bidding’. ⁶⁰ That way, insurers would bid according to the claim ratio of the districts, after carefully analysing the usability of the scheme both from the individuals’ side and the health infrastructures’ side, to ensure that their implication in the scheme will generate sufficient profits. Furthermore, in RSBY policy document, there has been no policy mechanism for stopping insurance companies and technology providers to do data dumping/offloading of RSBY usage data for future purposes. While discussing with private implementers about these issues, we realised that insurance companies can use the data for their own purposes^, because the agreement signed with the state has no mention about data regulation and privacy ⁶¹ .