Whitepaper: Determine pre-existing well being circumstances from dental scientific notes

In April 2022, ExamOne and Sikka.ai introduced a collaboration to supply oral healthcare info by means of HealthPiQture. Sikka just lately revealed a whitepaper on the usage of Sikka Well being Indicators™ to establish well being dangers amongst insurance coverage candidates.

Sikka Well being Indicators whitepaper

Determine pre-existing well being circumstances from dental scientific notes

Sikka.ai is the main API Platform within the retail healthcare business that features opt-in dentistry, veterinary, audiology, optometry, chiropractic, orthodontics, oral surgical procedure, and different medical practices. Sikka.ai API Platform seamlessly connects to 96% of the retail healthcare market all through the US and is processing billions of transactions a day. Sikka.ai API Platform supplies the platform to assist practices optimize their enterprise, profitability, affected person communication, income cycle administration, affected person satisfaction, and affected person medical historical past evaluation, by enabling over 50 market functions on its platform.

Sikka has over 41,000 opt-in dental practices put in within the US and Canada by means of its market-leading API integration platform. Sikka leverages this wealthy dataset to find out if life insurance coverage candidates have 1 or extra of 10 pre-existing circumstances or habits with outsized impression on underwriting.  These indicators are based mostly on precise scientific notes from the licensed suppliers within the practices or affected person reported circumstances on well being historical past varieties.  Sikka’s oral healthcare indicators can assist decide the suitable threat class or adjustment based mostly on the applicant’s well being threat indicators. This can assist make sure that “much less dangerous” life insurance coverage policyholders don’t find yourself subsidizing “dangerous” ones, no matter whether or not or not circumstances are mischaracterized by accident or deliberately.

10 Sikka Well being Indicators

10 Sikka Health Indicators


Textual content Classification Mannequin

From Sikka’s huge database, sufferers with scientific notes that comprise particular key phrases are recognized for every well being indicator. These scientific notes had been preprocessed utilizing varied NLP preprocessing strategies. The Word2vec algorithm was used to generate a distributed illustration of phrases from scientific notes as numerical vectors, capturing the semantics and relationships between phrases.

The embedded phrase was fed into the Lengthy Brief Time period Reminiscence (LSTM) mannequin Determine 2, which is a sort of recurrent neural community able to studying order dependence in sequence prediction issues. The LSTM mannequin is efficient in memorizing necessary info and, in contrast to conventional classification algorithms, LSTM can use a a number of phrase string to search out out the category to which it belongs. The LSTM mannequin was educated on a 400,000 balanced dataset with an accuracy of 99%.

As a part of enhancing the textual content categorization, a guidelines engine was developed to include any incorrect classifications discovered within the retrospective research.

Fig 1. Text Classification Process


Fig 2. LSTM Architecture


Retrospective Research and Hit price

Sikka’s information has been validated in research performed by 3 main reinsurance firms, 3 main information suppliers, a number of carriers and MGAs in each the US and Canada, and a number one life settlement firm. These research vary from a choose 2,500 to an expansive 8,000,000 information and have match charges of as much as 54%. The tobacco indicator has recognized vital numbers of “smoking non-disclosers” that value carriers as a lot as $23,0001 per conventional time period policyholder in misplaced premiums as a result of misclassification, based mostly on the evaluation of 1 of the information suppliers3. Individually Sikka Indicators at the moment are in manufacturing with a number of carriers and have been useful at figuring out lacking underlying circumstances that affect underwriting.  Research of Sikka’s Tobacco indicators have been accomplished to establish gross protecting worth of just about 10x the associated fee.2 In 2019, Munich Re carried out a validation of Sikka’s Tobacco Rating utilizing insured information. Analysis confirms that details about dental well being may be informative about total well being. Munich Re recommends every service carry out a retrospective examine to finest assess the worth and software of the Sikka Tobacco Rating on its company-specific insured inhabitants.3

Current webinar

Be a part of Sikka and ExamOne for a captivating and well timed webinar dialogue with prime business panelists on how you can leverage oral healthcare information successfully for all times underwriting.

Watch Webinar 



1 https://www.verisk.com/siteassets/toprisks/how-audio-analytics-can-detect-undisclosed-tobacco-use-verisk-whitepaper.pdf

2 ExamOne Value Profit Evaluation, June, 2021, Brian Lanzrath

3 https://www.munichre.com/us-life/en/views/alternatives-for-stratifying-mortality-risk/oral-health-mortality-and-smoker-detection.html

Leave a Reply

Your email address will not be published. Required fields are marked *