A Crystal Ball for Medicine

HamlinB_SON_DMG_2010_9143_150x200 (1)Twenty years is a long time to rely on one measurement approach. Imagine if in this technology-centric world we still relied on dial-up to connect to the internet. That’s basically where we are on quality assessment today. But we don’t need to be.

Predictive risk calculations allow doctors to look into the future. A risk score tells doctors how likely their patient is to develop heart disease or have a stroke. Working with their patients, doctors can discuss options for lowering this risk with the goal of preventing such events from happening.

With data from electronic health records, we should be able to create risk profiles for individual patients that actually take into account the different factors affecting their personal health—not just their age and gender, but their family history, whether or not they smoke, what medications and treatments they are receiving, and their own perspective on how they feel.

But right now, a 50 year old woman’s risk of developing heart disease is determined by a threshold set for the entire population of women aged 50-65 across the country. That’s a crude science. Everyone is not built the same. We should create risk profiles that change as patients change: as they reduce their risk by losing weight, quit smoking, or lowering their high blood pressure, thus reducing their chances of a heart attack or other adverse event.

That’s the vision of NCQA’s Global Cardiovascular Risk Score (GCVR). Leveraging the pioneering risk prediction work of Archimedes, it extracts data from electronic health records and uses a sophisticated algorithm to generate a highly sensitive, patient-centric risk profile for each clinician. It works like this: the higher the score the less likely a clinician’s patient will develop heart problems in the next five years.

The GCVR encourages doctors and patients to work together to set goals and track achievements, and helps them optimize the brief time they have together by focusing on what treatment adjustments or lifestyle changes would have the greatest impact on their health.

Those same scores which arguably give healers what they most want—timely information that will help them and their patients make precise decisions about care—could also revolutionize the way those doctors measure the quality of care. Physicians with higher predictive risk scores can be confident that they are being more effective and efficient in optimizing their patient’s health.

What a concept—quality measures that help with real time care.

Unlike current measurement strategies which suffer from a large time lag between when patient data is submitted and when quality results are generated, the GCVR approach looks to eventually present this information in real time at the clinician’s office. Because of NCQA’s highly standardized approach to EHR quality data collection, we can present clinicians with GCVR scores on a quarterly or monthly basis, compared to a traditional wait time of over a year for other measures. Our ultimate goal is to enable an automated quality measure process that allows providers to generate a GCVR report whenever they wish to view their progress.

We’ve tested the GCVR successfully with several healthcare organizations. We’ve still got a ways to go, but our preliminary findings suggest that we can get reliable data from EHRs to calculate GCVR scores.

And now we need your help to extend this pioneering work to all different types of community-based systems. With your active engagement, the GCVR can become the new gold standard to move us to a system that delivers the highest value care that is truly patient-centered.

Health professionals have made huge time and financial investments over the last several years in adopting electronic health records and measuring the quality of the care they provide. It’s way past time for all that investment to start benefitting those same professionals by providing tangible and timely quality results.

Ben Hamlin is the Director of the Department of Performance Measurement at NCQA

6 replies »

  1. You can say “it’s cheesesteak time.”

    Then a few weeks later “no cheesesteak no cry.”

    Repeat the cycle with cyclical variations in risk profile.

    Personalized medicine. Precision medicine. Patient-centred medicine. 101.

  2. “We should create risk profiles that change as patients change: as they reduce their risk by losing weight, quit smoking, or lowering their high blood pressure, thus reducing their chances of a heart attack or other adverse event”

    And this will have exactly what effect on how I treat the patient?

  3. hmm…let me see if I understand this. You say (correctly, in my opinon) that current risk scores are very crude measures and should be replaced…and make a very cogent argument for that, saying that current predictive scores basically don’t work.

    And yet the leader of your organization, who makes a ton of money certifying wellness companies, says exactly the opposite, that wellness programs save a lot of money because they can predict who is going to get sick, and prevent it.

    Since you’re saying the opposite things, only one of you is right. Which one of you is it — you or Margaret O’Kane?

    Hint: Her biggest argument in favor of her position and against yours is that she “doesn’t buy” my proofs that wellness is a waste of money.

  4. It is reports like this that confirm that many so called thought leaders do not get it, or are simply out if touch with reality.