Data fidelity: Black and white vs. color
Think of using claims data alone as seeing only in black and white. There are still some black and white cinematic masterpieces out there, and black and white photos certainly have character. But seeing movies and photos in color has become popular for a reason. What’s in front of you becomes richer and clearer. You get an enhanced picture.
It’s a similar dynamic when you move from analyzing claims data to integrated data. You start to see the deeper, richer, more “colorful” nuances of the patient journey that you may have missed when the picture was only in black and white. You glean market insights data that wouldn't be apparent without combining cost, clinical outcomes, adherence data and health care utilization information.
Layering additional clinical detail onto claims data through compliant and secure data linkage will help your analyses capture a more holistic view of the patient experience. And with this expanded view comes the potential to identify more of your target population and achieve more of your business and research goals.
Painting the picture: multiple sclerosis case study
To illustrate just how enhanced the picture can be with integrated data, consider the following real-world example* from Optum Life Sciences researchers.
Optum researchers conducted a study to investigate the early symptoms that precede the onset of multiple sclerosis (MS), a chronic neurological condition, to determine their predictive value in identifying individuals at high risk for the disease.
The literature review showed studies which used information from de-identified clinical notes to improve prediction models. To demonstrate the added insight gained from layering claims with clinical data, the researchers conducted their analyses twice, using 2 different RWD sources. First, they used claims data alone, then second, used claims data linked with notes derived from EHRs.
Researchers employed a machine learning (ML) approach to predict incidence of MS based on:
- Claims data (documented MS diagnosis)
- Combined claims data (documented MS diagnosis) and signs and symptomology (SDS) derived from EHR notes
The study findings suggest that symptoms could appear long before the diagnosis of the disease. The addition of EHR data improved the odds ratio of each predictor variable. Overall, analyzing SDS data along with claims significantly improved the accuracy of risk prediction.
The benefits of utilizing SDS terms from EHR systems in addition to claims data highlight the potential of ML approaches for improving MS diagnosis — and possible applications for other diseases.
This example sets the stage for generating the strongest, most impactful evidence you can have — something that is ever more critical as the rules of competition continue to intensify and evolve.