Dr Jeffrey Hammer has extensively worked on the economics of providing public health services in India. As part of the Accountability Initiative’s ‘The Cutting Edge’ series, he spoke to Ritwik Shukla prior to the Coronavirus pandemic in India, offering his insight on Ayushman Bharat, the Government of India’s new insurance programme, a critical missing link, and how this can be improved.  

About Ayushman Bharat

Ayushman Bharat, under the aegis of the Ministry of Health and Family Welfare (MoHFW), was launched by the Government of India (GoI) in 2018. The programme consists of two initiatives:

• The Pradhan Mantri Jan Arogya Yojana (PMJAY), a health insurance scheme aimed at providing access to quality inpatient secondary and tertiary care to poor and vulnerable families and reducing out-of-pocket expenditures arising out of catastrophic health episodes.
• Transforming 1.5 lakh Sub Health Centres (SHCs) and Primary Health Centres (PHCs) – the first point of contact in primary healthcare– to Health and Wellness Centres (HWCs) by 2022. 

In building a  comprehensive healthcare system – from HWCs to hospitals – this scheme faces some challenges. These include the lack of empaneled hospitals in some states and districts, some healthcare costs not being covered, difficulties in monitoring, and the poor status of existing primary healthcare facilities. The interview below delves into the conceptual nuances of these challenges. 

This piece has been edited for clarity. 

Q: What Ayushman Bharat is trying to do is build information systems. I’m going to quote an anecdote that I’ve heard. There is a mechanism where photos of the patient are taken– so if they’re getting a surgery done, then a photo of the scar or the mark of the surgery is taken.

Based on data validations measures such as this one, fraud has been reported, and the government has also de-empanelled hospitals. What would you say on such measures of catching fraud and misreporting? 

JH: Let’s take the example of a few other countries. So, for instance, in Canada, there are no insurance companies or intermediaries but the hospital bills the government directly. While Canada has way fewer packages, a simpler system (which is strange since the Canadian administrative capacity is pretty good), there have been estimates that there’s upcoding or providing more treatment than needed. The estimates are 10%, 15% of extra money going to hospitals and doctors due to upcoding. 

But what do you think would happen in India if doctors could just bill the insurance company or the government? Do you think it would be 10%, less or more? Also, how often will the government be able to send in the monitoring team?

Take the analogy of refrigerators delivered, or something that is easy to count and thus relatively easy for auditors like the Comptroller General of India to verify. Checking bills and whether treatment actually took place, in the same way as mentioned is much more complicated, especially given that doctors may make subtle distinctions even in the middle of the treatment. How do you check that? What’s the extent that you trust the doctor? 

So, in short- while I’m happy to see that the government did de-empanel hospitals and the errant service providers were caught, I’d also like to know what happened to the de-empanelled hospitals.

Q: I’m trying to find out. But what we do know is that there are penalties involved. At least, the guidelines are clear about it…

JH: Sorry, but I have to interrupt. That has always been the problem in India. The laws and guidelines of policy are always extremely well-written and comprehensive. It’s the implementation that is the problem. That there are guidelines is not even remotely sufficient.

Q: Point taken. I think the question is also more in the sense that the information asymmetry/imbalance around doctors is so strong because a doctor is a professional you can’t question. So, that’s a fundamental problem perhaps for every health system.

Is there anyone who’s actually been able to do this well, or even partly well? 

JH: The short answer is no. Or the short answer is yes, every rich country does it reasonably well. The fundamental problem is just as you said- you can’t really second guess if you pay someone for their services [doctors/hospitals]. If you pay doctors per-patient or per-case, you’re going to get more service provision than needed. If you pay doctors with a fixed wage, you’re going to get too little service provision. [1]

No matter how you do it, unless you have someone watching all the time, which isn’t possible, you’re going to either get too much or too little. Every system is trying to figure out how to strike a balance, and it’s very hard. 

For example, Finland has the formulae and people in place. They have a system that the Indian system was supposed to be like– Sub Centres, Primary Health Centres, everybody is in the public system. It’s 100% public and the doctors are civil servants, just like here. It works pretty well.

However, they figured that it was impossible for the national government to monitor these people, so the civil servants are hired by local governments. Local governments in Finland are like little villages; that means they thought that, even in a country like Finland with unbelievably high technical and educational standards and number one in Transparency International’s most honest governments list, etc. – the national government is too far away to be able to monitor effectively and thus required government employees at local levels.

These are levels of government even smaller than Gram Panchayats who can keep an eye on the doctors. Unfortunately, we don’t have such systems here. 

In contrast, Germany’s system is that there are insurance companies, but everybody has to be in one of these, and there’s regulation of the insurance companies. France also has something totally different– they have public hospitals but all regular doctors are private. In Sweden, it’s like Finland– they have local governments in charge. Again, even though it’s the third most honest country, in the city of Stockholm which is a large market, they decided it wasn’t working because of the incapability to monitor doctors in the city.

Let’s put this into perspective: the area of Stockholm is way less than Delhi and the population is 1/10th and they can’t do it. What they ended up doing was the payer-provider [2] splits. I have not kept up with what’s currently happening in Sweden – this was 20 years ago – and they might have something different now. But the overall point is that the implementation of public insurance is a difficult problem.

So basically, there’s no simple, single solution. Every single European country that is doing okay, is doing okay in entirely different ways along their own paths. They did one thing, tried another thing when it didn’t work and following this trial and error, ended up with completely different systems. In solving these problems, England ended up with National Health Services and Germany ended up with several insurance companies under one system.

Q: And what about India?

JH: The states are all different in what they can and cannot do. We can’t expect them to come up with the same solution, which is another thing that might make you suspicious of Centrally Sponsored Schemes as part of which everybody has to fall in line with practically the same guidelines.

So, we’re going to have to trust our doctors to perform without a lot of supervision for now, have information systems constantly in place to see how they’re doing and be willing to make adjustments along the way.

Until there is a well-functioning malpractice mechanism or an administrative system that can efficiently monitor and rectify each problem as reported by patients, accountability may be impossible to ensure.

Q: You mentioned Finland has a public health system with health centres at different levels. The second part of Ayushman Bharat has been trying to do exactly that– you build 1.5 lakh Health and Wellness Centres, and then you have hospitals at secondary and tertiary levels, and the idea is to integrate them through a system of referrals. Any thoughts on such a system?

JH: On referrals, you might like to look at the incentives which will be different for public and private doctors. You’re not going to be able to watch every single referral and, as designed, the system seems to be to trust that something works itself out.

But, personally, I’m skeptical. Public doctors are likely to refer too much– there is no incentive to hold on to a patient that requires a lot of thought or time. The private doctor on the other hand doesn’t want to lose the patient, and thus doesn’t want to refer the patient because it’s a guaranteed loss.

This difference between public and private doctors is always going to be a sticky point for the referral system. Private doctors might under-refer because of losses. Public doctors on the other hand, will get nothing for seeing more patients. And if we fix that – by tying public remuneration to cases seen –  there is a danger of turning public doctors into private doctors, which would mean we would lose the ‘publicness’ of the public health system.

Also, I have mentioned the role of monitoring in an insurance setting earlier on, and thus insurance providers may still have to second guess specific decisions of doctors. But, how will you figure out which doctors have kept the patient for too long? You can interview the patient, but no one takes the word of a poor person over the word of a doctor.

So, if the system is going to be integrating primary and secondary centres, they might have to start second guessing the primary health care centres for this, and this could lead to further complications. Again, information needs to be continuously collected and used to correct things if they start to go wrong – too many or too few referrals.

Q: What would a workable system look like? 

JH: So, I’ve been criticising the primary healthcare centres model for a long time because, basically, it’s complementary to the food line (where people have multiple options). Even in health systems, they have options to address their health needs, including private treatment and quacks. 

The path I would’ve chosen is as follows. Let’s start where doctors are willing to work and where we can watch them a little bit better. If they decide they’re seeing too many cases that are too simple to bother with at a hospital, then they may very well decide to establish small clinics someplace else, beholden to them and organised by the hospital, just to screen and save time.

This system is organised by the hospital, and it is solving a problem they face by helping them meet local needs. Otherwise, in a country so varied, you can’t conceivably anticipate what problem some of these hospitals will run into. We’re also underfunded, understaffed, under-everything! 

In 2001, doctors were asked what a satisfying career looked like for them? They said ‘being able to use what they learn in medical school, having the equipment necessary to do what they’ve learnt in medical school and having colleagues to discuss interesting cases’. But this is not going to happen if you’re the only doctor sitting in a PHC. It can happen where they really feel like they can do what they were trying to do in a hospital. And therefore, this might also address some of the absenteeism problem in the country’s public healthcare system.

Q: What about accountability within the system? 

JH: We’re back to better information systems and ways to use information collected to make policy changes. Not a strong point in Indian governance. The essential problem of managing millions of individual transactions that all require different solutions makes health a hard sector to regulate. Most of the time private markets do a good job in having people hold providers to account.

For food, people know exactly what they want and what it’s worth. In health it’s more complicated as I’ve explained till now. Patients are not in a position to second guess what doctors say. Of course this is a much worse problem in hospitals than in primary care. But again, second guessing all these transactions is prohibitively expensive.

So any system has to be watched carefully, information about visits recorded (symptoms, treatments, outcomes) and analysed and systematic problems fixed along the way. Until there is a well-functioning malpractice mechanism or an administrative system that can efficiently monitor and rectify each problem as reported by patients, accountability may be impossible to ensure. 

Ritwik Shukla is a Research Associate at Accountability Initiative. 

Editing by Avantika Shrivastava with inputs from Avani Kapur

Transcription by Cearet Sood 

To cite this blog, we suggest the following: Shukla, R. (2020) The  Cutting Edge Series: India’s Public Insurance Gamble, Interview with Dr Jeffrey Hammer. Accountability Initiative, Centre for Policy Research. Available at: www.accountabilityindia.in/blog/indias-public-insurance-gamble/.

[1] An example can be a doctor can upcode cases, push people for C-sections instead of regular childbirth if they are paid per case; admit more patients if they are paid per patient; and if they are paid a fixed amount irrespective of cases and patients, they may be slow or not admit anyone.

[2] Payers are beneficiaries of health insurance schemes while providers are those that provide the insurance cover. Payment is often split between the two to ensure that the risks are not borne entirely by one party.