Uncategorized

Next Generation Healthcare Analytics

Verisk_Deb_Bradley_PhotoMedical claims, pharmacy claims, lab values, HRAs, genetic markers, biometrics – the abundance of data  is having an immediate impact on how analytics shape healthcare.  Next generation analytics are bringing attention to health and wellness rather than disease-specific guidelines, and generating novel approaches to value-based medicine and care management.

Traditionally, analytics, such as predictive modeling, have been used to identify individuals for chronic care management and to set rates.  New predictive models, however, include financial and clinical algorithms, which allow healthcare organizations to implement advanced ways to identify, manage and measure risk across and within a population.

A few examples of these pioneering applications of advanced analytics are outlined below.

Genetic testing  More than 1500 genes can now be analyzed to find specific variations and mutations.  Approximately 10% are of clinical significance, with a growing number of gene tests being developed in clinical practice settings.  Genetic testing early on can lead to improved medical outcomes by optimizing therapeutic interventions and reducing morbidity, and lowering the cost of treatment.  Applying clinical algorithms specific to individuals who would benefit from genetic testing promotes better medical outcomes by reducing or eliminating complications, and ensuring correct application of medications.  An example is identification of individuals with a predisposition to Factor V Leiden.  Early identification avoids the potential complication of venous thromboembolic disease. 

Healthcare analytics applied to identify members for genetic testing offer the following benefits:

    •    Pharmacogenomics (‘Right Therapy’)    •    Ensures proper therapy and dosing    •    Disease Management (‘Right CarePath’)    •    Ensures proper care for individuals presenting disease symptoms    •    Pure Prevention (‘Right Future’)    •    Ensure proper preventive care for pre-symptomatic or at-risk individuals    •    Best Price (‘Right Price’)    •    Ensure diagnostic cost containment and best price for diagnosisUnderstanding uninsured populations

To better understand the healthcare needs of uninsured populations, and be better informed in contracting with health plans, the Commonwealth Health Insurance Connector Authority in Massachusetts is using a risk–adjusted capitation payment methodology for the FY 2010.  

Using available data and applying predictive models, individuals with the most costly manifestations of distinct diseases are identified.  A clinical profile of the expected costs expressed as a relative risk score is then generated.  Payment to the health plan is based on the risk and expected cost of caring for the individual.  This allows the Connector Authority to better align payment with risk by moving dollars from plans with better than average risk to those with worse than average risk.  

This is the first step in how the Connector Authority sees applicability of analytics across its program.  Future plans call for evaluating members for care management participation, and tracking utilization, quality and financial metrics across the plans.Realignment of Primary Care Physicians payment

As the number of primary care physicians (PCPs) decreases nationwide, a new approach to reimbursement is needed to encourage them to stay in practice.  Data analytics are used to promote a risk adjusted comprehensive payment plan to Primary Care practices to replace current encounter- based payments.  

Using predictive models, the ‘advanced medical home’ reimburses PCPs with a comprehensive monthly payment based on the relative risk of each patient that covers care costs and the cost for electronic health records.  In addition to risk adjusted payment calculations, analytics further support PCP practices by providing normative benchmarks and individual performance for financial, clinical and utilization outcomes.  

Analytics provide transparency of data to allow PCPs to review their practice patterns relative to other PCPs, aggregate lists of their patients with gaps in care, and comply with evidence-based guidelines.  The PCP is provided with the tools that focus on accountability and achieving health outcomes rather than focusing on the number of patient encounters.

Employee accountability

Analytics provide more than predictive modeling. Using financial and clinical algorithms, identification of new online health programs can lower the percentage of the medical dollar contribution by an individual.  Individual identification and stratification for engagement in condition and wellness programs allows employers to develop new benefit packages that offer incentives for case/disease management participation, and payment for participation in wellness activities as well as compliance with preventive health measures.  

Using analytics, employers can hone in on which intervention programs are needed, which members have care gaps, then adjust health coverage accordingly.  Let’s say the data shows 30% of employee medical claim costs are attributed to back pain treatment.  To respond, the employer may offer coverage for massage therapy or a reduced co-payment for enrolling in a back pain exercise program.  Individuals who ‘opt out’ of the program, or refuse to comply with recommended treatment, are assessed with larger out-of-pocket costs.

The future of analytics is now, as both employers and health plans combine disparate data sources to present one ‘truth’, and use that truth to improve overall health status by initiating health-focused programs, providing incentives, and tracking effectiveness of interventions.  The doors opened by healthcare analytics including predictive modeling, clinical algorithms and benchmarks are creating new opportunities to improve care.Deb Bradley is Vice President, Client Solutions at Verisk Health in Waltham, Massachusetts

Spread the love

Categories: Uncategorized

Tagged as:

9 replies »

  1. Thank you for any other wonderful article. The place else could anybody get that kind of info in such a perfect manner
    of writing? I’ve a presentation subsequent week, and I’m at the search for
    such info.

  2. One thing that is seriously overlooked in all of this is the social context relative to today’s environment. We really need to start doing analytics that provide insights into the impact that our new “social world” via social media has on all of this relative to the “subtle pieces of information” Mark mentions above. I am not a doc but as a former addict of the pseudo variety I know that social conditions whether addicted or not impact us all…and they impact the decision makers…and those who have back pain and even massage therapy is not going to mask their real core issues. I seriously doubt that genetic testing has found its way into the social media impacts in the work force as of yet? But…to say it (social media) is not playing a role is quite possibly a critical oversight into this entire issue…

  3. I look at this discussion from an Employer’s perspective. Well-run companies are engaged in a high-stakes balancing act of providing a competitive benefits package that will attract and retain good talent while not sinking the company under the weight of its benefits costs. Employers have both a vested interest and fiduciary responsibility to manage the performance of the benefits plans. An important part of this process is the use of sophisticated analytics — analytics that provide critical insights into how to design a competitive and effective plan.
    This type of analysis uses aggregated data — data collected at the person level, but rolled up to business unit or company level.
    Just as a manufacturing company would use six sigma methods to “control” its manufacturing processes, top performing benefits departments use aggregated data from all available sources to run its benefits plans. (Deb Bradley mentioned only a few of the potential data sources in her blog.) This concept has been around for many years, but many companies, pressed by the current economic situation, are just beginning to develop this capabilities. In a 2006 survey from McKinsey, 6 in 10 CEOs claimed they never considered measuring the performance of benefit plans against benefit objectives. I believe a similar survey taken today would reveal different results.

  4. Mark,
    In every patient encounter there are innumrable subtle pieces of information that are not gathered for the computer or database. A single case from the past can be the trigger to solve a new patient’s problem. Evidence lacking to support a teatment is not the same as eveidence of ineffectiveness. Don’t be too sucked in by evidence-based medicine, for it mostly is used as a tool to justify nonpayment.
    And us dinosaur docs have a work ethic unmatched by new docs hatched in the last decade. Computers are much slower than dinosaurs. Next time you are waiting in the ED for six hours, find out what kind of EMR they are using and short the stock for that company. Remember….to realy foul things up takes a computer.

  5. Odd to see the very negative response here to evidence based medicine. These docs seem to think that their random anecdotal experience is much better than statistically validated scientific deductions. These doctors are dinosaurs. The sooner that they retire, the better.

  6. Very well stated. It still confound me that most people do not understand that we have the scientific and technological advances TODAY that could provide insights and proactive solutions to a myriad of healthcare issues (both clinical and operational). This is not an issue of privacy or technology run amock. This is an issue of using for our benefit the technological advances already here. I imagine that some folks believe that if they cover their eyes they do not deal with the advancements in technology. As Galileo is supposed to say during the signing of his renunciation during the Inquisition…”[the earth] is still round.”

  7. As explicitly stated in any investment firm’s prospectus, experiential data has no ability to predict future performance, neither should health care analytics rely on patient care data or control group data to guarantee the reliability of analytic-based performance output for meeting customer expectations.
    Access to data for well-intentioned uses, such as optimal allocation of resources or for programmatic prevention or wellness programs, should be a feature of a well-developed system. As important, however, are the privacy of patient records and compliance with HIPAA. The author should have demonstrated how mandated privacy rules can be embedded in analytics or how analytics might have controls to not import data unless it is in compliance with privacy laws, policies and procedures.
    From a policy perspective, the potential range of abuses analytics offer employers and health plans is unacceptable without automated privacy controls. Otherwise, employers would be able to screen job applicants to disqualify those persons with too high potential for using sick days or other health-related criteria to reduce that applicant’s anticipated productivity. Existing risk management staffs have sufficient data now to reconcile health plan benefits with productivity needs of the firm.
    Like employers, insurance companies and other health plan providers could have individualized, prejudicial financial incentives to exclude applicants who are likely to consume relatively more provider resources than other members.
    For researchers, population data for analytics must meet the oversight requirements of their Human Subjects Panel to assure quality and appropriateness of patient experiential data, In the researcher’s proposals to funding sources, submissions should include specific analytics by certified software and procedural controls. Further, researchers’ methodology must assure stakeholders and those whose private data is used as the basis for any public health policy recommendations they may make.

  8. This sounds like the Borg collective in Star Trek.
    Health plans and employers define the “truth” and make all decisions for their drones, based on genetic tests, I assume, done at birth and mandatory.
    Patients and doctors “comply”…. or else…. As the Borg say “resistance is futile”…..

  9. This is scary stuff. Cookbook medicine is what this is, with no choices for the patient after they are profiled and no choices for the doctor. In fact, the doctor is completely unnecessary in this alternative reality.
    The employer will spend all his time managing his employee’s healthcare instead of running his business.
    The big brothr overtones of wellness programs imply mandatory compliance and success. Sounds like a wonderful life…to avoid like the plague. Total loss of freedom and portability.
    I’ve got an idea. The governemt needs to budget it’s resources for each year, and when they are gone, they are gone. Stop stealing from Social Security Ponzi scheme to pay for other things. Stop the unfunded Medicare Ponzi scheme, which is only for the benefit of the pols. People must remain free of government and mandated behavior. Otherwise sick people start being a lot like Soylent Green ingredients.
    After careful analytics of my own, I think you are a little too controlling for the Constitution.

Leave a Reply

Your email address will not be published.