by 
If you’ve followed the Smart Guide articles so far, your APM is ready for take-off. You’ve developed a data-sufficient technology infrastructure with both provider and claims data, carefully constructed your clinical network and engaged your clinicians, and implemented strategies for payers and consumers. Your APM may be operating already, with data ready to use.
Now you are at the crossroads, and your choice will determine whether your APM will really transform outcomes for your patients and achieve maximum Value, or whether your APM will be average. That choice pivots on interventions. Only through specific interventions with clinicians and patients can you change the trajectory of costs and outcomes. Let’s dig in.
Interventions as a Data-Driven Queue-Up for Patient Treatment Changes
The point of intervening is to change the course of events before those events happen—patient crises, for example. But many APM entities go about dealing with events in the reverse order: identify a patient crisis that culminated in an admission, and use population health outreach to bring the patient in for a visit or other intervention. This is appropriate to do. But it is better to do more with your data before the events occur, potentially averting a crisis. Your data should be able to identify patients at risk of failure before it happens.
We all know the saying about lightning not striking in the same place twice. If you believe that your patient who was called in after a crisis has improved because of outreach, that may be true. But relying on that strategy means that each measured improvement depends on having a patient in crisis, to begin with, necessitating an expensive event. A better strategy is to create a system that ensures your patients are improving without entering crisis mode, which means you need to engage clinicians in reviewing outcomes and intervention criteria before the crisis occurs.
What are intervention criteria? They are clinical and treatment criteria that point to an impending problem. Longitudinal outcomes that show persistently poor control in Type 2 diabetes are one such indication, because failure to deal with diabetes will result in its exacerbation and likely crises. But more important, combining this characteristic with medication evaluation, lack of appropriate specialist involvement, visit adherence, and potential social determinants of health will create the opportunity for specific improvements in patient treatment and avoid further deterioration of the patient’s health. Other conditions have similar constellations of outcomes and clinical symptoms that can be used predictively to stratify patient risk.
As a result of this data process, a patient may be queued into interventions for clinician review of medications (which could result in SDOH revelation and referral to financial resources), referral to a specialist, and prescribed nutritional services. Review of A1C levels may also result in continuous glucose monitoring and implementation of a self-management program.
Interventions to Refine Clinical Pathways and Clinician Engagement
You can also use interventions to help clinicians improve treatment plans. By creating episodes of care for conditions and procedures, you can plot variations in costs or outcomes across groups of patients. This allows you to differentially identify contributing factors to the episode costs or outcomes.
The point is to investigate variation, not judge, so that clinicians can collaboratively address refinements in pathways for both conditions and surgical procedures. Because longitudinal data and costs (especially patient-identified) have been so rarely provided to clinicians, a learning process is critical. By using episodes of care, all clinicians across disciplines can share in the investigation.
Five Intervention Strategies to Fuel Your APM Liftoff
1. Create a two-prong intervention strategy that empowers clinical interventions in (a) chronic disease and (b) specialty medical/surgical.
ACOs and other primary-care-focused APM entities frequently focus interventions on chronic disease patients, adding higher utilizers and other patients at high risk. Because these are the patients of primary care participating providers, this makes sense. However, it’s not enough.
With 40-60 percent of costs driven by specialists, you need to include specialty partnerships in your intervention strategy. You can achieve this with specialty partnerships that cement your referral network into a mutually beneficial collaboration based on data sharing, communication, and collaboration on referral criteria and treatments. This strategy enhances value to your patients and both primary and specialty physicians, providing avenues down the line for streamlining care as well as improvements.
Clinical interventions, which can be supported administratively through patient navigators and other clinician support means, are essential. The only way to reach ultimate value is to deliver better care that prevents advancement of disease as long as possible.
2. Use clinical episodes of select conditions and treatments as the basis for review of costs and outcomes.
Don’t limit your view of costs to category of service. Fee-for-service categories don’t make much sense in the APM world, where cost is calculated by all the care delivered to patients over a time period (per patient per year). For conditions, this time period is generally described as a year; for procedures or specialty incidents, for the period covered by the incident and any appropriate pre- or post- services such as imaging or physical therapy.
Clinical episodes enable you to create a unit of comparison. Consider these questions to get you started:
- What effect does differential use of anesthesia agent have on your costs or patient recovery?
- In how many of your patients may therapeutic inertia be a problem, and why?
- Which patients with both obesity and diabetes are on insulin-only?
- Why were some cholecystectomies performed as open procedures with higher costs versus laparoscopically?
Episodes of care are not magic, but allow your clinicians to evaluate the inputs of one case against the inputs of another and assess the results of each. And episodes of care also allow clinicians to examine costs in a way that makes sense to them—as a total of all services provided to a patient within a finite time frame.
Construct your episodes only for key areas of chronic disease and specialty services in your patient population—unless you are an academic or specialty-driven organization, where delving into episodes through the specialty side makes sense for specialty care models, research, and care pathways.
3. Start with clinical questions, and then activate data to identify potential interventions.
Interventions are the test of your data sufficiency, but only if you ask the right questions, first.
For conditions, the question of whether the patient is on track starts with clinical data, notably outcome values over time. If your data is not identifying these outcome values, there is more to do. Either your clinical data is not integrated with your claims data, or your clinical data is not being transported correctly into your repository. If your data is missing medications, then you will need to find prescribing or filled-benefit information from your EHR and/or claims data, most likely the former.
For procedures, claims data will provide a good supply of transactional information, but you may be missing diagnoses and outcome values coming from EHRs. This is one reason why specialty partnerships are important.
The less sufficient your data to fuel interventions, the harder you’re making it for your clinicians to review cases, which is both unfair and less effective.
4. Create a connection between interventions and health equity efforts.
You will immediately reveal health equity issues in reviewing patient episodes, especially for conditions. Cases where patients had no referrals, no treatment changes over years, and lower-level medications should all trigger reviews of patient circumstances that could point to lack of health equity. One solution is to use your patient navigators to interview (or visit) patients for a full understanding of their situations. Alternatively, a better approach may be to connect with community organizations who are already working with your patients.
Enriching your social determinants of health data is an important conduit for both improving your Value and for providing support for patients. You may also be able to include patient family and other support to your patient care team and magnify the effects of your interventions.
5. Test the effectiveness of your interventions.
Don’t make your interventions an intuitive program. Rather, back it up by verified data. You want to know that your clinical interventions are working, and in what patients, and how long it takes. You should be able to examine the critical points of failure in your process as well as in the data or the interventions themselves. The key to improvement is understanding what works, and what has not. Ensure that your system is capable of tracking every intervention event, the effect of that event, and the data results going forward. That creates the knowledge for changing the process or the people involved in it.
Interventions foster change. They have the power to transform the inevitable trajectory of a patient’s story into a better narrative, altogether. Well planned and executed interventions can completely transform the lives of many patients—and your organization. This is the nexus where all your organization’s efforts coalesce to create Value. Make the most of it!
Founded in 2002, Roji Health Intelligence guides health care systems, providers and patients on the path to better health through Solutions that help providers improve their value and succeed in Risk.
Image: Brian McGowan