Given the good intentions behind the health insurance, what could go wrong?
We are all familiar with the options when somebody falls seriously ill. One, go to one of the public hospitals, some of which are very good - the Pakistan Institute of Medical Sciences (PIMS) and Jinnah Hospital come to mind - and charge nominal fees, especially from the poor. However, many are overcrowded, suffer from serious deficits in infrastructure and have doctors who are not motivated to provide high quality care. Two, go to a private hospital. Depending on your ability to pay, the quality and care there ranges all the way from top-of-the-line to downright dangerous, but generally, doctors there will try and do what they can - even do “too much”. The money they make depends on your experience.
Faced with a similar situation, in 2008, India asked a very simple question: if all that stops poor patients from accessing better quality care in the private sector is the expense, why doesn’t the government just pay for that care through a dedicated insurance scheme? On the surface, it seemed that this would solve the problem of improving care for the poor without having to improve public hospitals – which has proven extremely difficult. We still do not know, for instance, how to reduce absenteeism among public-sector doctors – it can run as high as 40 percent.
The scheme eventually implemented balanced incentives across multiple players in the system. Insurance companies are given a list of potential beneficiaries in a district and asked to submit a technical and a financial bid. They are then chosen based on the bids and proceed to the enrollment stage. Each enrolled family is given a smart card and the insurance company is transferred the premium amount in their bid. The families carrying the smart card can visit any empanelled hospital (public or private), and the hospital is reimbursed the package rate for the in-patient stay. These “cash-less” transactions were designed to reduce the burden on the poor.
Note the checks and balances — the more families the insurance company enrolls, the greater their margins, so the insurance company has the incentive to improve coverage. The hospitals would like more patients (and may try to attract them in multiple ways), but insurance companies want to keep the amount of care to a minimum. This, we thought, would prevent the system from doing too much or too little. As a part of the team that helped set up the insurance scheme called the RSBY (Rashtriya Swasthya Bima Yojana) or the National Health Insurance Plan, I felt that such a radical departure from our usual focus on the public sector was very welcome after 60 years of ineffectual health reforms. What could go wrong?
As it turns out, plenty.
First, researchers in 2017, namely Karan, Yip and Mahal showed that, at least by 2012, there was no effect on health expenditures. To quote: “The results suggest that the RSBY has been ineffective in reducing the burden of out-of-pocket spending on poor households.” So, the first premise of the scheme - that it would reduce expenditures for the poor - was not borne out by evidence.
Second, as researchers started digging into the data, they uncovered the potential reasons why. Using data from Rajasthan, Radhika Jain has shown that hospitals seldom adhered to the package rates mandated by the scheme and that under-the-table additional charges were common. The additional financial burden on patients came on top of continuing concerns around outright fraud and misappropriation of funds.
Further, hospitals respond immediately to any change in package rates by switching their claims towards better-reimbursed procedures. So, if the mandated package rate for a hysterectomy increased, the data showed, the number of hysterectomies suddenly increased. Some of this may be due to differences in coding, but more ominously, some of this could also be because hospitals genuinely increased the number of hysterectomies to increase their reimbursements.
On the flip side, Covid-19 comprehensively demonstrated that, under the right set of circumstances and investment, public hospitals can absolutely step up to the task of providing the care we need. Let me be unequivocal here: throughout the pandemic, and in both India and Pakistan, public sector functionaries have worked tirelessly around the clock to manage as best as they can. This is true not only for frontline workers who have gone beyond the call of duty on a daily basis, but also members of the administration who have worked non-stop for the last year under difficult circumstances.
The biggest problem is that it is impossible to price the services accurately and ensure that the government gets what it pays for.
As Pakistan now contemplates the Sehat Saluhat Card, with similar financing and public funding for private provision, it is worth stepping back and asking what we missed. The answer is three-fold.
The biggest problem is that it is impossible to price the services accurately and ensure that the government gets what it pays for. When the price is too low, a hospital will either choose not to enroll or will deny the services. When the price is too high, the hospital will make unfair profits, or worse, try to convince patients to receive the service even when it is not needed. Reports of unnecessary hysterectomies under the RSBY followed such erroneous pricing.
In the Indian scheme, a stent is reimbursed at Rs 40,000 but a heart bypass at Rs 120,000. Even if administrators can perfectly determine what operation the patient received, there is nothing stopping a hospital from choosing the operation that makes for higher profits. Why stop at a stent if the bypass nets additional profit? But worse, there is no way in the system currently to discover what the patient actually received. “Upcoding” (inserting a stent but reporting a bypass and perhaps just outright fraud) become more likely with greater deviation in prices from each hospital’s cost structure.
Getting the prices right is the central dilemma in any insurance programme and one that all countries struggle to solve. One thing that countries implementing large scale programs have in common is a large analytical and data centre that continuously examines procedures, procedure coding and charges from the insurance scheme. Prices have to be frequently updated based on the data, and this is a job for large specialised teams in each state.
Given that we will never get prices exactly right, hospitals may under-treat, over-charge or just deny services. The RSBY tried to deal with this problem through guidelines that allowed local NGOs and partners to set up health desks in hospitals to help patients navigate the scheme and the hospital. Unfortunately, there were few takers. How to allow, and staff, neutral third-party advocates is, therefore, a critical issue for such a scheme.
In tandem, the scheme will require dedicated teams with supporting legislation to control fraud. For the first time, the success of a health scheme is now tied not only to the functioning of the Health Department, but also the criminal justice and court systems. In the United States, insurance fraud is a federal crime that is vigorously prosecuted with prison time that can be as long as 15 to 20 years. It is worth stepping back and thinking about the legal framework around insurance fraud. Are the courts and legal system ready to take on what will inevitably be a massive challenge?
The key insight from the functioning of the scheme was that it was not simple at all. In fact, the scheme substituted our knowledge of how to provide health services through public institutions (which we have built up over the last 60 years) with our knowledge of how to regulate private provision of healthcare, a topic that we know very little about.
Some numbers on what this requires may help. In the United States, the single-purchaser Medicare scheme employs 6,000 people to cover 44 million beneficiaries. These are all highly trained professionals handling insurance audits, pricing and medical records; dealing with anti-trust cases and fraud and examining billing issues in each state. A similar staff to beneficiary ratio in the Indian state of Uttar Pradesh would require 10,000 highly trained regulators. The actual numbers? 42 even if we include a chief executive officer, nodal officer, contractual staff and medical officers. The story is very similar in every state in the country.
For a province like the Punjab to run such a scheme effectively, something like 5,000 to 7,000 highly trained regulators would be needed. Colleges and training institutions would have to be set up for this and staffed almost immediately if the scheme is to reach its potential.
Faced with the type of challenges, my plea is that we give up on the idea that this is a “small” change or an “easy” win. By all counts, it is not and in fact, moving to the kind of scheme that India has put in place will require massive state investments in an area that we have little experience with and do not understand very well. It will also require substantial changes to our legal framework and our ability to detect and prosecute fraud through the court system without political interference. Given the experience we now have on how our public hospitals can actually function given sufficient investment and trust in our doctors, it is worth thinking long and hard about the path we choose. Although it will take longer, a single-minded focus on improving our public institutions may, after all be a more desirable investment.
The writer is a professor at Georgetown University and a Senior Visiting Fellow at the Centre for Policy Research.