Wednesday, September 15, 2010

Appropriate Metrics of Counting Beans

Good evidence demonstrates that once an Emergency Room doctor determines a patient is better served by a specialized care unit, such as the Medical Intensive Care Unit, it's best that deliverance arrive quickly. This notion strikes us as common sense, yet still, they went to the trouble of studying the question.

The results are striking. Of course, they confirm our suspicion: patients do better when they are promptly taken to the ICU than they do hanging out in the ER. The results also quantify the difference in outcome produced by reducing delays. The difference in mortality between deliverance and waiting compares well to another procedure-coronary artery stenting.

Given this data, we can expect ER physicians to measure themselves by how quickly and precisely they direct critical patients to appropriate care units. Discretion is required; care units have limited space. ER doctors make the decision to transfer largely on the unique capabilities of a unit's nursing staff and equipment. While a cardiologist is easily called down for a consult, the staff of the coronary care unit is tethered to their facility. Expertly managing patient flow is an appropriate metric to grade ER doctors.

Two positive outcomes are possible when transfer is not delayed. First, mortality improves. Towards higher quality is the first direction to lead our medical care system. Second, reducing transfer time by an hour and a half has been shown to reduce patient stays by a day and a half. And, this is the second direction needed in our medical care system: cost down.

But does the hospital's corner office wish to grade ER doctors by this metric? On first glance, turning beds over more rapidly lowers revenue--certainly that is the corollary of lowering patient costs. Hotels don't operate on rushing their guests out the door, so how can hospitals survive by doing the equivalent?


So then, how can I suggest that bed turnover, not bed utilization, is the most appropriate metric for hospital performance? Hospitals do outsource bed management to companies such as Sodexo that have their roots in hotel management. However, because hospital beds are highly utilized with low seasonality, the revenue drivers of these two industries are quite different. Utilization rates are a secondary metric to turnover.

For Medicare patients, bills aren't reduced when they are sent home more quickly. DRGs have already bundled payments for Medicare admissions. Once the patient is diagnosed, only co-morbidities will increase reimbursement. Additional days in the hospital procure no additional revenue, only additional expense.

Therefore the goal of a hospital administrator is more similar to a restaurateur on a Saturday night: turn the tables over. In this way, doctors and hospital administrators have aligned interests, just as the wait staff and management share the same goals at a restaurant.

So a Medicare patient who does not heal rapidly is akin to that table who lingers in conversation after finishing their meal. Or, to place it in terms of pharmacology--the beds are saturated and hospitals are operating on 0-order kinetics. Either more beds, or a faster turnover of beds must be achieved to yield more profit--the enzymatic product of a hospital bed from the POV of a hospital CEO.

Clearly, freeing bed resources doesn't reduce revenue, it provides opportunity to pursue more business (heal, in physician lingo). Under DRGs, faster bed turnover, and associated cost reduction entirely benefits the hospital. Bed turnover is the more appropriate metric than bed utilization to grade resource productivity.

The risk emerges for hospital administrators will push physicians to release patients earlier than is clinically indicated. This is a well-worn and legitimate argument, which also supports bed turnover as a primer driver of hospital accounting. We can't both insist hospital accountants won't support ER patient flow improvement--because QI may reduce revenue, and bemoan the perverse incentives of hospital administration to dismiss patients before they're fully well.

Bed turnover dictates sales velocity, one of the most important metrics for generating revenue in a business. Bed turnover also influences capital budgeting. Consider an industry rule: each new bed, with all associated equipment, costs $1 million. If management is presented two methods to generate equal revenue: 1.) improve bed turnover, 2.) expand the hopsital, improve bed turnover should win or the hospital board needs to scout new talent.

Improving quality undoubtedly benefits patients. Does decreasing costs only benefit providers? In the long run, no. Improving bed turnover will increase hospital profit margins only as long as the market remains static. So, not long. The ability to produce more revenue from a single bed will stave off the need to embark upon hospital expansions, it provides payers with bargaining power to reduce reimbursements. While the overhead of a new bed will likely remain around $1 million, as clearing that hurdle becomes easier, hospitals become relatively less resource intensive (assuming they remain demand driven).

In the long run, the nimbleness of ER doctors, expertly managing patient flow, can reverberate to produce a more agile medical care system. Projects can achieve both quality improvement and cost reduction. Flow projects align management and physician interests and leverage a key business driver: bed turnover. ER physicians should recognize that bed turnover influences both revenue and capital budgeting. By appreciating the value their QI projects deliver to the business, physicians can insist that the patient-doctor relationship benefits from the improvement just as much as the management-investor relationship.