Better Hypertension and Diabetes Outcomes: From Basic Measurement to a Plan for Improvement

written by Dave Halpert | April 29, 2015

hypertension 2Are you caught in a squeeze between avoiding penalties in both PQRS and the Value-Based Payment Modifier (VBPM)? Medicare’s move to Pay for Performance has created a Catch-22 for many groups:  you may have enough data to report enough PQRS measures, but if you choose to report measures where your performance is below the CMS mean of your peers, you risk penalties under the VBPM.

As a CMS reporting registry that integrates VBPM Consultation Services, we commonly find at least one or two measures per client with scores that could negatively affect the VBPM if used in PQRS reporting—especially for large, multi-specialty groups. The most frequent measures at issue are for hypertension and diabetes outcomes. With increased requirements on the numbers and types of measures included in PQRS requirements, some groups have trouble coming up with substitutes for both PQRS and VBPM.

But selection of other measures is only a short-term technique to address the calculation of comparative performance under Value-Based Health Care. It’s time to put your performance data to use in developing an outcomes improvement program that will address the challenge in a substantive way for patients.

Unraveling the PQRS-VBPM Dilemma

The Value-Based Payment Modifier is calculated using your performance on measures you reported for PQRS, and some measures are more challenging than others. Each measure has at least one “performance met” response. The easy way out is not to report a measure where you cannot demonstrate performance; but often no other measures are available, either because there are no patients eligible, or data isn’t available, or another NQS Domain is needed to fulfill reporting requirements.

This last scenario is particularly risky, as your total quality score is based on a composite of your NQS Domain scores. In other words, your total quality score is not derived only from each of your nine measures, but also considers your three NQS Domains, regardless of how many measures are included in that Domain. If an NQS Domain includes only one measure, that measure’s performance rate is the sole determinant of that NQS Domain score—it will have a far greater impact on your quality composite than one measure in a Domain with six others.   

Being stuck with an undesirable measure as the lone representative in an NQS Domain can be profoundly detrimental to your VBPM Quality Tiering and, consequently, your Medicare revenue. This is particularly relevant in measures focused on control levels for patients with diabetes and/or hypertension—conditions seen and treated in a variety of practices.

Understanding the Difference Between Outcome and Process Measures

Why are certain measures related to hypertension and diabetes more difficult to achieve performance? Here we need to distinguish outcome measures from process measures, which simply explain that a service was provided. Many measures for hypertension and diabetes are outcome measures; what counts is the patient’s end result. Let’s say you have a patient with hypertension, whose a recent reading is below 140/90. Your PQRS measure result is based only on whether that patient is in control. But for VBPM, your results are compared with other groups. On the diabetes side, the same holds for the measure for most recent Hemoglobin A1c level (also a cross-cutting measure), and most recent LDL level. In short, the outcome means much more than it did in the Pay for Reporting days.

Certain outcome measures have a safety net. For example, a measure related to BMI enables providers to meet performance for a patient with a BMI outside the healthy range, if the record includes a plan of care. However, these qualifications are often strictly defined and frequently not captured in an EHR-Direct report, since the details are often noted in text fields or customized templates for practice use. Clearly, this measure (along with a similar screening measure related to blood pressure) is easier to meet when patients are healthy.

Make VBPM a Foundation for Population Health Projects: Hypertension and Diabetes

So, what steps can you take to make sure that your intermediate outcome measures related to hypertension and diabetes don’t sabotage your Value-Based Payment Modifier? Use the VBPM as a task list for your population health efforts. That’s the best answer to creating a systematic outcomes improvement program.

Step 1: Identify At-Risk Patients

This is harder than it sounds. Managing hypertension is challenging because there are often no symptoms, so patients are less motivated to seek treatment. This translates into less frequent visits. If a patient sees you only once a year, there isn’t an opportunity to improve.

There are similar risks with undiagnosed diabetes. Often the first diagnosis is made during a hospital admission. Diagnosing diabetes at an earlier stage could lead to more proactive management and patient education, with levels held steady. Bringing someone under control is far more challenging than maintaining control before the condition evolves.

Catching diabetes early has a separate, less obvious impact on your Value-Based Payment Modifier. When Medicare calculates VBPM, they look at cost and quality. In determining your cost composite, once your patients have been attributed, they will analyze all costs and utilization associated with each patient. Along with an “All Patients” composite, Medicare calculates both costs and hospitalizations for patients with at least one of these four chronic conditions:

  1. Diabetes
  2. Heart Failure
  3. Chronic Obstructive Pulmonary Disease (COPD)
  4. Coronary Artery Disease (CAD)

Early action to maintain a healthy diabetic population will influence those results. The costs incurred by a controlled patient will be less than for those with long-term, uncontrolled diabetes, partly because the controlled patient will spend less time in the hospital. Programs to improve the health status of these individuals are essential to their health—and to the well-being of your organization.

Step 2: Create a Plan Backed by Technology

Improving outcomes comes from a consistent, detailed plan of action for implementing interventions with patients. Interventions can be clinical, educational or administrative—and those are just categories. Organizations must decide how to measure the results of initiatives and interventions.

There is no magic solution for improving outcomes or population health. Hiring care coordinators, changing medications, providing educational material or sending letters to patients are all possible actions, but each is based on speculation about what is wrong and what will work to improve patient status. Evaluating the effectiveness of your efforts should be part of your plan, which should include these components:

  • Population Health technology with more than aggregate analytics that can
    • Segment populations and subsets by defining different criteria;
    • Establish a “control group” for populations that are part of effectiveness research;
    • Track specific interventions that are being evaluated across a population or subset;
    • Update patient information with new data from lab, visits and other sources;
    • Conduct multiple population health projects at any one time;
    • Enable physician practices to see their patient data and compare with others;
    • Display patient level detail;
    • Allow input of data as well as data import, useful in both practices and central offices.
  • Plan of Action with input from providers and stakeholders (e.g. your contracted health plan) of what projects you want to include and when. For primary care organizations, the Plan should include projects to address the four core chronic conditions included by CMS in the VBPM, plus readmissions and ambulatory-sensitive conditions. Additional projects should be evaluated based on the results of your VBPM “QRUR”, otherwise known as Quality Tiering reports, which will identify the areas where your organization is not favorably comparing to the CMS Mean.

Step 3: Undertake Projects with Interventions

You can see from these technology requirements that we are far beyond “Population Health Lite.” Once a Plan is established, the interventions themselves must also be scheduled and protocols established for each intervention. Let’s say that you decide to implement an intervention for a population with diabetes, to improve HgbA1C values of high-risk patients. You have gone through these steps:

  • You have already defined the population;
  • You have decided to implement a group education program for patients with the highest variation in results;
  • Your chosen intervention is a group education curriculum of three sessions.

In addition to the educational curriculum, you will need to:

  • Establish an outreach campaign to reach the patients;
  • Set a schedule of the educational sessions that has already been tested for acceptability;
  • Create the educational curriculum;
  • Create one or more mechanisms for patients, so they can easily reach you to participate;
  • Establish protocols for eliminating patients from the population because they are clinically inappropriate;
  • Involve and educate your providers and provider staff to refer and respond to patients;
  • Define your effectiveness criteria for evaluating the interventions and your post-intervention measurement;
  • Determine the timeline for adding patients into the project.

Be realistic about how effective your first efforts will be. Even if your actions produce great results, unless you are able to bring that patient in again before the end of the year, those results will have to wait until 2016 PQRS/2018 VBPM.

Step 4: Measure Your Results

An intervention without impact isn’t a failure—it’s a starting point. It’s important to know what works for your specific population and what does not. A good partner and capable technology should be able to help you modify projects with new targeted interventions, perhaps focusing on a smaller group, to see how those actions may apply to the whole group. Patients whose results have significantly changed in the last few months make good candidates. These are your “swing voters,” and determining what may have helped them (or what challenges they may be facing) can be revealing and create an opportunity to improve on a larger scale.

Start Now: Population Health Takes Time

Like most worthwhile endeavors, this will take time. There may be some patients who only need a slight tweak in medication, but they will be the exception, not the rule. Some patients are overdue for visits, and simply getting them in the office won’t happen overnight. Once you have a starting point (for example, your patients’ average Hemoglobin A1c level, the 25th-75th percentile A1c range, and your outliers), you’ll need to take actions, and leave enough time to see what outcomes your actions produce.

Diabetes and hypertension both reduce quality of life for patients, and both result in significant medical costs. In addition, Medicare is looking at outcomes stemming from these conditions to determine whether your practice’s payment should be adjusted, based on your results. For some, these outcomes will lead to financial penalties. For those who start now and track the efficacy of their actions over time, there is an opportunity for reward—both from the Value-Based Payment Modifier and, most importantly, for your patients’ improved health.

Download your free copy of the ICLOPS Insider’s Guide to PQRS 2015 Reporting: How to Succeed in the Value-Based Health Care Environment.

Founded in 2002, ICLOPS has pioneered data registry solutions for improving population health. Our industry experts provide comprehensive PQRS Reporting with VBPM Consultation, Population Health with Grand Rounds and ICLOPS Outcomes Research solutions that help you both report and improve your performance. ICLOPS is a CMS Qualified Clinical Data Registry.

Contact ICLOPS for a Discovery Session.

Image Credit: National Library of Medicine