How much is too much when collecting data for planning?

It is useful in delivering development outcomes, but technology has to be deployed carefully

September 11, 2020 12:53 am | Updated 10:53 am IST

Medical research report or contract vector, flat cartoon health or medical record paper or insurance document on work desk top view, concept of medicine check list, approved good test analyze

Medical research report or contract vector, flat cartoon health or medical record paper or insurance document on work desk top view, concept of medicine check list, approved good test analyze

On Independence Day, Prime Minister Narendra Modi announced the launch of the National Digital Health Mission (NDHM) under which “every Indian will get a Health ID card.” He said: “Every time you visit a doctor or a pharmacy, everything will be logged in this card.” The NDMH seeks to create an ecosystem under which health records will be digitised. The government has clarified that this would be voluntary, data will be stored locally, and only anonymised data will be shared upwards. In a conversation moderated by Varghese K. George , A.R. Nanda and Venkat Srinivasan discuss whether real-time capturing of demographic data and broad health indicators of the population will lead to benefits that outweigh the perceived and real risks of erosion of privacy. Edited excerpts:

Mr. Srinivasan, how can technology transform data collection and analysis?

Venkat Srinivasan: You can’t think of data without technology. This is digital data. And of course, it is important to recognise privacy issues. But the only way, I think, we need to think about this is [how it can] transform health outcomes, access and care, and not [in terms of] separate data alone. Use of this data with technology, I think, is the real intent of this policy. As you know, in the U.S., we’ve had this electronic health records initiative for a while. So, technology can, if used properly, dramatically transform healthcare at all levels. Technology can’t work without data. So, it can be at the population level, it can be at the individual health level. At the population level, it can be used effectively to control the outbreak of a pandemic. At the individual level, you have several levels of benefits. In India, millions of people don’t have access to quality healthcare. The silver lining is that the government has allowed for digital health, which allows a lot more people access. I’m sure that will happen over time.


Mr. Nanda, what problems do we face when it comes to demographic and health data? In terms of sampling, frequency, accuracy, range...

A.R. Nanda: We face problems in all these areas. Sampling size, sampling errors, non-sampling errors, how frequently we collect data. Cost is a big factor in addressing these questions. The use of technology is the cheapest and easiest way forward, in collection of data, monitoring the quality. The problems are daunting, but nothing that cannot be solved.

In your career as a demographer and a planner, did you ever wish the technology had been better?

A.R. Nanda: Yes. I have been associated with the Indian Census since the 1970s. It used to take years before Census data and family health survey data could be tabulated. By the 1990s technology came to our help, starting with early mainframe models. Now it is far smoother.

Mr. Srinivasan what can technology do for collecting data, keeping it safe, and using it for the public good, which was impossible a decade ago?

Venkat Srinivasan: You have nicely framed it in those three buckets. Technology is the only solution. It is an excellent solution on the collection side. We can expand the sample [size] at low cost. Computing devices that can use ‘store and forward’ type architecture [can overcome] connectivity challenges. But the quality of the data is where I think the biggest advances will be felt because you can control it through many methods such as pattern matching and looking for trends, all real-time. On the privacy and security side you hear daily reports of hacking. There are several ways of making technology more robust — create levels of anonymisation, make it much harder for somebody to figure out who you are. AI algorithms can protect data and prevent theft. But [that] has to be complemented with a legal framework that acts as a deterrent for anybody who is caught stealing or misusing personal health data. It must be independent of any political machinery. So, possibly, create an independent commission, like the Election Commission.

Comment | Mind the gaps in India’s health care digital push

Mr. Nanda, the Personal Data Protection Bill is currently before Parliament but it has been criticised for the sweeping powers it gives to state agencies. Are we still a long way from a reassuring legal framework?

A.R. Nanda: We have some legal framework for protecting Census and other survey data, but the problem is that organisations that handle these are not strictly independent regulatory authorities. For example, the Census Commissioner of India has the power under the Indian Census Act to say no — I had said no to governments —but I also think an independent commission is needed.

Do you fear vindictive action by state agencies or other private agencies who might be able to access data for political, economic or ideological reasons?

A.R. Nanda: That’s a risk. But at the same time, you have to see what you call the public good. You have a law, may be put it in the Constitution.

Parley | Can a digital ID aid India’s primary health ecosystem?

You’re suggesting that we go beyond a mere law and have a constitutionally empowered regulatory authority?

A.R. Nanda: Yes.

How is this conflict between privacy and public good handled in the U.S. with regard to data, Mr. Srinivasan?

Venkat Srinivasan: We have the Health Information Protection Act (HIPA), which is being enforced quite effectively. We have similar concerns across the spectrum. Recently, there was a very controversial idea of declaring your citizenship in the Census. The data are also encrypted, but nothing is hack-proof. But the regulatory framework is very important.

The Hindu Explains |  How will a centralised electronic medical records tool help to monitor trends about COVID-19?

How is private data used for analysis and policymaking?

Venkat Srinivasan: In policymaking, the use of data is increasing but it is not extensive. For instance, the response to the pandemic. We use electronic health records at the population level and individual level all of a sudden now. A much greater amount of information sharing has happened. All of a sudden, hospitals are talking to each other. And they’re all sharing information not only on individual patients but on incidents. With anonymised data, people have less of an issue but in some cases it may not be anonymised — for instance, in contact tracing. When you go beyond health, we have other interesting issues in this public health crisis, policy issues — pensions, loans that were given to businesses. The issue is these policy interventions are all based on data that are not real-time. There is a 10-year interval between two Censuses.

The decennial Census has been around for more than 130 years now. This time it is being disrupted by the pandemic. Can’t we have a real-time capture of demographic data?

A.R. Nanda: That is a distinct possibility. Some countries, such as Germany, have already gone into that. To get there, we have to improve the system of our birth and death and marriage registration and registration of other statistics. This has improved, but only in urban areas, and not to the level where you can say that you can make the data usable and also won’t have to go for the same data to the Census. But we must also remember that there are other types of data which are also collected in the Census regarding culture, language, economy, but yes, we don’t need that every 10 years. That could be collected at longer intervals.

The Hindu Explains |  What are the concerns around the Aarogya Setu app?

Mr. Srinivasan, is it possible and desirable to collect at least basic demographic data on a real-time basis rather than wait for 10 years?

Venkat Srinivasan: Totally. In fact, it is not going to be very expensive. Technology costs have come down so dramatically, and you have such a large penetration of mobile phones in India. Individuals will have to be willing to report and that’s really where I suspect India’s challenge might lie. For this, it is important that they see a personal benefit. You can introduce a lot of intelligent algorithms that detect the quality at the source. When you think about intelligent algorithms, I’m a total proponent of traceable technologies. And traceability here implies that you should be able to understand exactly what happened at the individual level in the use of the technology. If you can afford to do that, that will get the game more acceptability. It may seem trivial, but I think it’s critical.

We are discussing finding a balance between individual privacy and promoting innovation and efficiency. Now, at what level of specificity and granularity in data could that balance be optimally achieved? How much is too much and how much is too little when it comes to data?

A.R. Nanda: That’s a real problem. In a census as well as in the big demographic surveys, health surveys, we have this compromise. The smallest area level is the village. More sensitive data and what involves the question of qualityare shared only at the broader units such as the State or district level. Technology could make data capture and analysis possible in smaller units but at the same time, one has to be very judicious doing that. We need an independent agency outside of the government to take these decisions.

Also read | Draft health data management policy released for public feedback

Consent and anonymisation are the key words often used to reassure sceptics. Is it all good as long we adhere to these two principles?

Venkat Srinivasan: Both are very important, though, to me, they are elementary. You must have that, but you do need a stronger, autonomous regulatory framework. It is also potentially possible that you separate this into two categories. One is census. And your question is relevant on how granular you get, how you protect data, and again, you can use technology to separate different layers as necessary. There are things you need to be more sensitive about, more judicious, and so maybe you have layers, maybe three layers of access. You have one layer that many people can access; that’s completely at a population level. Another layer is only some people that you know, and so on. And their informed consent and anonymisation are important. The ability of an individual to have their health records to be stored digitally at a centralised location and access it across the country could be separate from this. But I think you can have a different set of protocols and permit. For that, all that is technologically possible, but I think you need a strong set of laws like this HIPA compliance.

Varghese K. George: We do have a consensus that technology can do a world of good in delivering health, education and other development outcomes, but one has to be very careful and judicious in its deployment, and there should be an independent and robust regulatory mechanism to oversee that process.

A.R. Nanda, is a demographer who served as Registrar General and Census Commissioner of India as well as Health Secretary; Venkat Srinivasan is a Boston-based technology entrepreneur who deploys AI in finance, accounting, health and education.

0 / 0
Sign in to unlock member-only benefits!
  • Access 10 free stories every month
  • Save stories to read later
  • Access to comment on every story
  • Sign-up/manage your newsletter subscriptions with a single click
  • Get notified by email for early access to discounts & offers on our products
Sign in


Comments have to be in English, and in full sentences. They cannot be abusive or personal. Please abide by our community guidelines for posting your comments.

We have migrated to a new commenting platform. If you are already a registered user of The Hindu and logged in, you may continue to engage with our articles. If you do not have an account please register and login to post comments. Users can access their older comments by logging into their accounts on Vuukle.