Improving the “Performance” of Performance Measures

By Greg Seymann, MD, SFHM

As performance-based reimbursement strategies take center stage in healthcare payment system reform, the landscape for hospitalists remains stagnant, as we still lack a critical mass of relevant performance measures.

Under the Physician Value Based Purchasing (VBP) program, all hospitalist groups of ten or more eligible providers are subject to a pay cut from Medicare if they are deemed high cost and/or low quality. In January 2016, this cut will apply to all hospitalists, regardless of group size. The intricacies of how providers are evaluated on cost and quality are fleshed out here, but the majority of the quality barometer is based on self-reported performance measures via the Physician Quality Reporting System (PQRS). Currently, there are only 15 measures in the realm of hospital medicine that are reportable using inpatient billing codes, but several of them are something of a stretch. Even though they have links to inpatient billing codes, few hospitalists feel that reporting on diabetic eye exam or hypertension screening measures is up our alley.

One strategy to better position hospitalists for performance-based reimbursement is to develop new measures that are meaningful to hospitalists and the patients we serve. Transitions of care (TOC) is a niche that Hospital Medicine has uniquely championed, so measures aimed at safer transitions processes present a jumping off point.

Initiatives such as Project BOOST offer useful evidence-based best practices to make discharges safer. Increasingly hospitals and their quality teams are adopting BOOST elements to bolster their discharge processes; these elements involve screening for readmission risk factors, working with interdisciplinary teams, robust medication reconciliation processes, improving patient education and discharge instructions, timely follow up appointments, and post-discharge patient contact.

Accountability for disproportionately high readmission rates currently falls to hospitals through the Hospital Readmissions Reduction Program, and to primary care networks through Physician VBP, Accountable Care Organizations, and other alternative payment models. A focus on key outcomes like readmissions is here to stay, but as individual hospitalists or hospitalist groups, we seek measures that target processes that we impact more directly.

In the wake of SGR repeal, new opportunities to propose and develop specialty specific performance measures exist as part of the MACRA legislation have fueled discussion around the proverbial hospitalist water cooler about how we might best position ourselves in this regard. Targeting the BOOST best practices seems like a good place to start. Or does it?

Simple solutions don’t often work in complex systems like healthcare, so transforming BOOST recommendations into credible performance measures is not necessarily straightforward. Take medication reconciliation, for example. This is a process that hospitalists finesse daily as we navigate the care of medically complex inpatients. Many would say we perform it better than our colleagues in other specialties, so we should find a way to get credit for it as a next-generation performance measure.

How exactly would we distinguish ourselves as “high performers” on such a measure? After all, the key to good med rec is not simply doing it, but doing it RIGHT. It is pretty simple to document performance of the task: we could check a box when we complete it, submit a billing G code, or pull data from our EMR. But even if we relied on time consuming and costly manual chart audits, all we would learn is whether or not it was done. Auditors could not determine if it was skillful, sloppy, or somewhere in between.

What about ensuring all patients have a follow-up appointment at discharge? Hard to argue against the merits of such a step, but once again it is easy to audit, harder to master. What about patients with no PCP? Or patients from out of town? Weekend and holiday discharges? Underfunded patients? Those are obstacles that forward-thinking institutions can overcome, but the step that really matters is ensuring the patient shows up. Who is accountable for that?

We must not let the perfect be the enemy of the good, and for now we should embrace measures grounded in the principles of Project BOOST. However, we must continue to seek ways to refine and enhance the measurement process to incentivize meaningful quality. Because despite good intentions, sometimes tying rewards (or penalties) to such processes can inadvertently cheapen them. By transforming tasks we do in the best interest of patients into tasks done to earn a reward (or avoid a penalty), we may sap the intrinsic motivation to do them well (see here). Such an environment often spawns more workarounds than advances in patient care.

Measure me on how often I book discharge appointments? Sure, I’ll get it done, but whether it is with a doc my patient wants to see or in a convenient location or time slot will be less of a concern. Med rec every time? Sure I’ll check all the boxes, but if that is all I get credit for, how hard will I try to get it right? When we incentivize what is easy to measure, we may divert attention from the outcomes that really matter.

If you doubt the fact that well-meaning clinicians and hospitals would cut corners on good patient care just to satisfy a performance measure, hearken back to your premed years and come to terms with the fact that the “just tell us what we need to know for the test” crowd are now your colleagues. In the infancy of the pneumonia core measures, ED’s were giving out moxifloxacin in the triage bay to anyone who coughed in order to satisfy the mandate for rapid antibiotic administration. Smoking cessation counseling required for smokers with pneumonia? Sure, we’re leaving you this pamphlet and checking the “completed” box, good luck quitting.

Communities have wrestled with translating good ideas into actionable solutions since the dawn of civilization. No less a sage than Aesop shared an illustrative tale of a group of mice who grew increasingly frustrated with a vicious cat that lived nearby. The brood gathered to brainstorm solutions to this problem, and were compelled by a bold young mouse who declared the solution was obvious: simply placing a bell on the cat’s collar would serve to warn the mice when the predator was near, giving them time to safely disperse. The crowd’s initial enthusiasm quickly faded when another mouse spoke up: “Which one of you is going to place the bell on his collar?”

Hospitalists need meaningful measures, and for now we should embrace Project BOOST and solidify our reputation as outstanding stewards of care transitions. But we should not rest on those laurels; for the reasons above, we must continue to refine our measures, be vigilant for unintended consequences, and adopt an approach that is flexible enough to move us toward a measure set targeting meaningful quality. I’m hopeful that the next generation of hospitalist leaders and innovators will rally around this cause.

Dr. Greg Seymann is a hospitalist and Vice Chief for Academic Affairs at the University of California, San Diego Division of Hospital Medicine. He is a longstanding member and current chair of SHM’s Performance Measurement and Reporting Committee.