Tech helps NHA save Rs 66.28 crore since 2018, 210 hospitals de-empanelled

Techniques like facial landmark detection or blur index are also being used to identify anomalies in the beneficiary images.A completely paperless claims management system and the use of advanced data analytics and artificial intelligence algorithms helped NHA in cracking healthcare fraud collusion among hospitals, the sources said. Social network analysis is being leveraged to identify instances of collusion between entities such as hospitals colluding together or PMAM, beneficiaries colluding for ghost billing as well as the creation of ecards for unauthorized beneficiaries.


PTI | New Delhi | Updated: 09-05-2022 15:26 IST | Created: 09-05-2022 15:21 IST
Tech helps NHA save Rs 66.28 crore since 2018, 210 hospitals de-empanelled
National Health Authority Image Credit: Twitter(@AyushmanNHA)
  • Country:
  • India

The use of technologies has helped the National Health Authority (NHA) deter impaneled healthcare providers from committing fraud while making a saving of Rs 66.28 crore since its inception in September 2018, official sources said on Monday. The Authority, the apex body implementing the Ayushman Bharat Pradhan Mantri Jan Arogya Yojana (AB-PMJAY), also de-impaneled 210 hospitals along with other punitive actions, the sources told PTI.

Right from the inception of the AB-PMJAY – the world's largest publicly funded health assurance scheme – the focus and philosophy of the NHA have been the use of technologies in scheme implementation.

''The NHA has a zero-tolerance policy towards fraud and abuse and has been taking steps for prevention, detection, and deterrence of different kinds of fraud that could occur in PMJAY at different stages of its implementation. Proactive action has been taken by NHA to implement a comprehensive framework of policy, and system. Artificial Intelligence/machine learning techniques/models currently being used includes, Fuzzy Logic, supervised/unsupervised learning, neural networks, and deep learning,'' the sources said.

These advanced techniques are helping identifying instances of impersonation or ineligible beneficiaries getting enrolled in the scheme against an eligible beneficiary. Image analytics algorithms are being used to identify duplicates or ghost billing, the sources explained.

Techniques like OCR/NLP are also being used to correlate medical documents with information entered in the Transaction Management System. Techniques like facial landmark detection or blur index are also being used to identify anomalies in the beneficiary images.

A completely paperless claims management system and the use of advanced data analytics and artificial intelligence algorithms helped NHA in cracking healthcare fraud collusion among hospitals, the sources said.

Social network analysis is being leveraged to identify instances of collusion between entities such as hospitals colluding together or PMAM, beneficiaries colluding for ghost billing as well as the creation of ecards for unauthorized beneficiaries. Giving an instance of fraud, the sources said, ''Arogyam Hospital and Research Centre in Jharkhand had used an image of a patient treated for cataract surgery in its claim in February 2019. The same image was also used by Gulab Hospital for Appendectomy in July 2019 for another patient. Sanjeevani Hospital and Research Center used the exact same image for the treatment of malaria in another patient in September 2020. This clearly establishes collusion among fraudulent entities''.

Real-time risk scoring of all individual transactions and entities such as hospitals, PMAM is being carried out based on machine learning algorithms to alert suspicious approvers.

The entire system is based on reinforcement learning models wherein the feedback being received from states is incorporated into the model to improve accuracy.

''Zero tolerance attitude to frauds and use of technologies helped NHA in savings of Rs 66.28 crore. A total of 210 hospitals have been de-impanelled along with other punitive actions including but not limited to recoveries, fines, and de-empanelment of healthcare providers,'' the sources stated.

(This story has not been edited by Devdiscourse staff and is auto-generated from a syndicated feed.)

Give Feedback