I spend a lot of time promoting the value of electronic health records over traditional paper records because I believe in the efficiency gains and cost savings they bring to underwriting budgets. Because so many people are using AI tools to answer all types of inquiries, I prompted ChatGPT with the question – “Is electronic health record data inferior to traditional paper medical records?” Here is a synopsis of the response, “The idea that electronic health records are ’inferior’ to traditional paper records is more perception than reality, and it depends on what you are comparing. A better comparison of the two types of medical records is that one is not “better” than the other, they are simply different.”

Nobody will deny that in the early stages of EHR adoption, underwriters will often experience transition fatigue and a bit of cognitive overload as they navigate the differences between how patient data is presented within an EHR record vs. the familiar APS. Inexperienced users are also likely to work at a slower pace during this learning curve period. However, a slight delay during this time of adjustment is to be expected - and the temporary impact on productivity will be outweighed by the significant cost savings and efficiency gains reflected in cycle time metrics and expenses in the long run. And when we consider these gains in combination with improvements observed in data quality, the positive impact is undeniable.

For the last four years, I have performed an annual review of content depth on 100+ cases from MIB’s individual data sources for EHR. For consistency, I review the same number of cases each year (114) and use a consistent number of USCDI categories (United States Core Data for Interoperability) as the measurement tool. USCDI is a standardized set of health data classes and elements for nationwide, interoperable health information exchange. The classes & elements included in the review are selected based on their relevance to mortality and morbidity assessment and include:

Results from 2022- 2025:

In most categories, the content depth is performing consistently or better than prior years, at or above 80% in key areas such as medications, vitals, and overall health encounters and problems. Even tobacco/smoking status is showing almost 80% of the time.

Other notable observations:

• The HIEs (Health Information Exchanges) performed similarly to the EHRs in terms of robustness of content and useful directional value for accelerated programs, such as:
  • Medication compliance, long term obesity, mental capacity issues, substance and tobacco use history, and non-medical information.
• Some records are normally considered as lacking in protective value (examples: eye exams or orthopedic consultations), were found to include critical medical and non-medical high-risk information.
  • Our review revealed a record for a young male who was last seen for knee pain, was in the U.S. for less than a year, in a new job, and had a significant medical history.
• Records for young applicants may provide high value / high-risk information.
  • Our review revealed a record for a homeless teen - which makes you wonder who was applying for the coverage?
• Records with a limited number of pages or older information (more than 24 months old), often provide high- risk information.
  • Our review revealed a four-page record for a middle-aged man in need of aa psychiatric evaluation.

Given these findings, why does the perception that electronic health record data is “inferior” continue to linger? Could it be that EHR data is held to a higher standard than other types of electronic medical data (such as claims, prescription, and clinical lab data) because it is most often used as a replacement for the traditional APS and, therefore, expected to provide exactly the same information?

We need to start reframing the conversation about “value” to one of cost/benefit and challenge underwriters to recognize when there is ‘enough’ medical information to assess mortality and morbidity risk. Until we are willing to challenge the risk assessment process, put aside the comfort of traditional methods, recognize that EHR data often does include ‘enough’ information to assess risk, and that an APS is not “better” but just “different” - we will be stuck in this misperception of “inferior” value.

We, as underwriters, can do better.