Money & Banking

PM-JAY adopts machine learning to tackle fraud

Maitri Porecha New Delhi | Updated on December 13, 2019

Indu Bhushan, Chief Executive Officer, Pradhan Mantri Jan Arogya Yojana   -  V_V_KRISHNAN

Provides State and district authorities with field app to improve fraud-control mechanisms

To better tackle fraud, the National Health Authority (NHA), which implements the government-run cashless health insurance scheme Pradhan Mantri Jan Arogya Yojana (PM-JAY), has devised a mobile-based field application for State and district authorities, to upload evidences from the site of investigation on real-time basis.

Up to 338 hospitals are under NHA scanner for alleged frauds and 111 have been de-empanelled, which means that they have been removed from the scheme. Penalties worth over ₹3.5 crore have been recovered. In Uttarakhand alone, penalties worth ₹2.5 crore have been recovered from 15 hospitals. In Uttar Pradesh on the contrary, ₹2.7 lakh have yet been recovered from up to 30 hospitals. In Chhattisgarh, action has been taken against 76 hospitals, and recovery made is ₹4.13 lakh.

PM-JAY CEO Indu Bhushan said, “We have seen up till now that 1 per cent of all cases accepted under the scheme have been fraudulent.”

Unlike its predecessor Rashtriya Swasthya Bima Yojana (RSBY), which did not have an efficient fraud control mechanism, PM-JAY is dabbling in machine learning and artificial intelligence to identify malpractice triggers. Five companies — SAS, MFX, LexisNexis, Optum and Greenojo — were taken on board for six months, starting July earlier this year, to conduct fraud analytics. According to an official, “There was no investment on the scheme’s part during this exercise. We will now float a Request for Proposal to get a company on board.”

WhatsApp-like group

“In a bid to improve our fraud control mechanisms, we are equipping the State and district authorities with the field app. We have formed a WhatsApp-like group at district levels, in which we red-flag suspicious claims raised or pre-authorisations sought,” an NHA official told BusinessLine.

The teams at the field level swoop in on the suspect hospital and investigate the medical documents. During the on-going inquiry, they upload the documents onto the central server on real-time basis, and these papers are not retained after their upload at the local level,” the official said.

The official further explained, “This helps reduce the time of investigation as we get only 15 days to check the claim, as it has to be paid within two weeks of pre-authorisation.”

Every pre-authorisation and claim submitted will be vetted through the system in a bid to identify fraud. “For example, if a male claims money for a delivery, which is a female procedure, there will be a system-level trigger pop-up. If a patient is being shown as signing up for a cataract surgery, third time around there will be a red-flag, as one cannot have three eyes, or for that matter, if a woman is shown to be undergoing a hysterectomy (uterus removal) twice, as a woman has only one uterus,” the official further said.

NHA currently does not have the legal sanctity to prosecute hospitals for fraud. “This is left to the discretion of States. We are feeling the need for a National Health Insurance Act, like a law, to be put in place. Who will draft it, is the question. Will it be the NHA or the Insurance Regulatory and Development Authority?” said the official.

“It is too early for us to say anything about an impending Act at the moment,” Bhushan said.

Published on December 13, 2019

Follow us on Telegram, Facebook, Twitter, Instagram, YouTube and Linkedin. You can also download our Android App or IOS App.

This article is closed for comments.
Please Email the Editor

You May Also Like