The title of this post is a terrible pick-up line, but it reflects my recent musings on how we identify patients in primary care who may benefit from behavioral health services. Health screening is standard practice in primary care. We screen for physical health, mental health, substance use, health behaviors, family functioning, social determinants of health, and more. We assume that the screening tools are accurate and use the data to inform our treatment decisions. We may even assume that patients do not completely understand their own health and that screening can be the first step toward an accurate diagnosis. Screening allows care delivery systems to process high numbers of patients, categorize patient health problems, calculate risk levels, and assign appropriate treatment.
Screening is a ubiquitous practice in healthcare. I have worked in five different primary care systems and each one had a protocol for identifying patients with depression. Other industries use screening as well. Employers screen new employees for drug use and citizenship status; athletic teams screen players for physical prowess and potential ability; banks screen customers for credit worthiness. There seems to be this underlying belief that certain discrete points of data can predict something important about employees, athletes, customers, and patients. Although there is strong evidence that health screening works, it is certainly an imperfect science. For example, many patients with chronic pain symptoms will flag on the PHQ-9 (a depression screening measure) even if they don’t have depression symptoms. Most behavioral health measures rely on self-report and are at risk for false positives (and false negatives too!). Screening data should inform clinical intuition and experience; it is a starting point, not an end point.
It may help to take a step back and think about the functional purpose and rationale of screening. By taking a step back and choosing the target of your screening efforts, you can create a comprehensive strategy that meets multiple needs while efficiently using your resources. In my career, all of my clinical experience has involved universal screening. The purpose of universal screening is to identify as many candidates as possible; in other words, the wider your net, the more fish you catch. The rationale is that universal screening can identify anything on the spectrum: potential, developing, active, and even urgent health problems. For example, SBIRT (screening, brief intervention, and referral to treatment) is an effective universal screening approach for identifying substance use and assumes that, because there are more risky substance users than those with full blown dependency, a greater number of problems exists with the former sub-population. Another example of universal screening is the recommendation from the United States Preventative Services Task Force for depression screening. They recommend all adults be screened, although they do not specify frequency. The collection of physical vitals data (blood pressure, heart rate, weight) is a universal screening approach too. Imagine seeing a medical provider and not having your blood pressure recorded.
Universal screening is pretty, well, universal in healthcare. There is another approach that has gained some recent attention that you may not know about. Although there is no official term for it, let’s call it focused case identification (FCI), a population health management strategy. This approach includes tools like patient registries, risk calculation, and non-clinical data. The purpose of FCI is to identify patients who would benefit most from a service based on an algorithm or risk score. The rationale is that, in a healthcare system of limited resources, our sickest patients are most likely to be high utilizers of care and have the worst outcomes.1,2 By determining risk level and focusing our resources on this segment of the patient population, we can get the biggest bang for our buck. There are four recommendations to follow when taking this approach: first, focus your case-finding efforts on patients with chronic medical conditions and high health care costs; second, deploy treatment resources in a fully integrated fashion; third, use only highly trained clinicians and evidence-based treatments; fourth, escalate care when appropriate using care management and coordination strategies.3 The assumption is that a focused case identification approach would reach the Triple Aim faster compared to universal screening.
So, what should your clinic do? Well, first let me say that I may be creating a false dichotomy here, a straw man so to speak. This comparison of two case identification approaches is just meant to highlight the strengths and limitations of each model. On the one hand, universal screening is a fairly straightforward and low technology-based strategy (you could do it all with paper surveys) that can create false positives and negatives. On the other hand, a FCI model can identify your sickest patients and efficiently use resources, but it requires significant expertise and sophisticated health information technology for maximum benefit (e.g., care management software, risk calculation tools). I can see hospitals using the FCI model because they have access to the knowledge and technology it requires, but a small clinic in rural New York? Not so much.
Here are some thoughts about moving forward. First, combine both case identification approaches into a comprehensive strategy. Clinics can’t focus on patients with chronic health problems and comorbidities unless they first screen for those issues. That usually requires a universal approach. Perhaps the solution is to only universally screen all patients with chronic health conditions like diabetes and congestive health failure, but not the entire patient panel. Second, clinics need access to knowledge and technology that make FIC possible. Payers, both private and public, can create knowledge networks that clinics can easily join to learn more. Some clinics may need practice facilitators to accelerate their transformation process. Finally, policymakers should create rules requiring EMR vendors to develop easy to use and affordable population health management tools, making FIC more likely to happen. So, perhaps instead of using “What is your philosophy of screening” as a pick-up line, maybe a better line is “What is your value-based case identification model?”