Home Cover Story Taking charge of observation units

Taking charge of observation units

With hospitalists at the helm, these units aren't just for chest pain any more

July 2007

Published in the July 2007 issue of Today’s Hospitalist.

In the weeks after the chest-pain observation unit opened at Cox Health’s Springfield, Mo., south campus in July 2003, Frank Romero Jr., MD, heard a lot of grousing from local internists and family practice physicians. The problem? Physicians were leery of relinquishing the care of their patients to hospitalists during such a potentially critical time.

Today, the unit has more business than it can sometimes handle, says Dr. Romero, who directs the hospitalist program and also rotates through the observation unit. For the first few months, the unit saw 40 or 50 patients a month “but that number quickly rose.

“Now I wonder if they are asking all patients if they have chest pain,” he quips. “Some months we might have 180 people.”

“There are a lot of politics, but the important thing about a decision unit is that it needs to be protocol-driven.”

~ Deanna Larson, MD
Methodist Hospital

The dramatic increase in patient volume is only part of the story. Before the eight-bed unit opened, the average length of stay for chest pain in the 534-bed facility was 72 hours. That’s now down to 18.

Revenue has also seen a drastic turnaround. In 2003, the hospital was losing about $600 on each of these patients because many didn’t meet Medicare guidelines for either observation or admission.

“In one year,” Dr. Romero now notes, “the hospital went from losing an estimated half-million dollars to $2 million in revenue on that unit alone.”

Similar trends are showing up in a new 18-bed unit at Philadelphia’s Temple University Hospital. The unit saw nearly 700 patients in the first three months of operation, about half for ruling out chest pain or acute MI. Early indications suggest that the unit, staffed 24 hours a day by two hospitalists working 12-hour shifts (and another on call during the busy afternoon hours), is helping patient flow from the ED onto the medicine units.

William Ford, MD, Temple’s section chief for hospital medicine, says there are no firm data yet, but anecdotal reports suggest that hospital capacity and throughput have improved. “From our length of stay “we’re shooting for 23 hours, but we’re at 36 now for the average patient “we know it has helped with ER volumes,” he says.

Removing bottlenecks
As EDs are increasingly stretched by capacity constraints and lack of specialist support, hospitals are turning to observation units, also known as clinical decision units, to fill those gaps without taking up inpatient beds. Patients admitted to observation units are expected to be evaluated and treated and rapidly improve within 24 hours. (The Centers for Medicare and Medicaid Services will pay for observation status for up to 48 hours or, in rare situations, even longer.)

A 2003 survey published in the American Journal of Emergency Medicine found that nearly 20% of U.S. hospitals had established observation units, with an additional 11% planning to open one within a year.

Since then, growth has only accelerated. “The big issue is the entrance point to the hospital,” maintains Bret Nicks, MD, secretary for the observation unit interest group for the American College of Emergency Physicians and director of a 15-bed unit at Wake Forest University Baptist Medical Center in Winston-Salem, N.C. “Hospital census is the underlying bottleneck now, and having an observation unit available helps minimize this problem.”

While you may think of observation units as primarily for conditions like uncomplicated chest pain, heart failure and abdominal pain, these services are also being created to handle diagnoses like asthma and exacerbations of COPD. The thinking is that extended evaluations better inform the “keep-or-discharge” determination.

One major goal of these units is to reduce potential liability by making sure that borderline or “gray area” patients don’t get sent home while they’re still at risk for a poor outcome. An even bigger motivation comes from payers, who want to avoid medically unnecessary admissions.

Putting hospitalists in charge
While such units have traditionally been run by EDs and directed by emergency physicians, there’s a major shift afoot, points out Stacy Goldsholl, MD, president of hospital medicine at TeamHealth, a national clinical outsourcing company based in Knoxville, Tenn. As more units are launched, a growing number are being led by hospitalists.

The thinking is that hospitalists, by virtue of their “insider” status and knowledge of hospital systems, are better positioned to coordinate care and get patients “re-dispositioned” “discharged home or admitted to a medicine unit “than emergency physicians.

“More hospitalists will be managing these units, and for continuity reasons that’s appropriate,” Dr. Goldsholl says. The majority of patients placed in observation will not end up being admitted to the hospital. But those “keepers,” she maintains, will benefit from having the same physician who cared for them during the observation period.

ED physicians now welcome that management change. “In some observation units now, the ER physician is there as an adjunct to allow hospitalists, like other subspecialists, to take care of the higher-risk patients and let the ED take care of the lower-risk ones,” says Roger Brooksbank, MD, an emergency physician and regional director for TeamHealth Midsouth in eastern Tennessee.

Sharing responsibility
Figuring out how to share responsibility between the ED and the hospitalist service is a big facet of keeping a unit up and running. At Methodist Hospital in Omaha, Neb., hospitalists organize and deliver care in the unit while the ED retains fiscal responsibility and overall medical direction. Hospitalists assumed primary responsibility for the three-year-old observation unit earlier this year because it “wasn’t being utilized to its full potential,” says Deanna Larson, MD, the hospitalist in charge of the switch-over.

The hospitalists first set a 90-minute goal for initial patient assessment. They then tackled the more challenging business of working with ED physicians and specialists to decide on criteria for patient inclusion and exclusion, and to figure out how to gets tests and labs done quickly.

“The whole goal of the decision unit is to expedite patients’ care, to get things ordered and get them out,” Dr. Larson says. “Having a primary doctor decide to leave their patients overnight and see them the next day on rounds, that doesn’t meet our goals.” While the unit is used primarily for chest pain now, there are plans to expand it to include patients with heart failure, syncope and transient ischemic attack, among other conditions.

Standardized protocols key
As hospitalists at the helm of these units have discovered, encouraging all parties involved “hospitalists, ED physicians and cardiologists “to agree on unit admission criteria is no small feat. In Methodist’s unit, patients must have an initial set of negative cardiac enzymes, be hemodynamically stable and have no history of serious heart disease.

At Temple Hospital, the clinical decision unit uses the acute coronary syndrome TIMI risk criteria as a starting point and insists that patients be “chest-pain free” on admission, Dr. Ford explains. Patients with a heart rate greater than 110 or a systolic blood pressure over 200 are excluded, as are those with abnormally high cardiac enzyme levels or patients deemed at risk for suicide or homicide.

Dr. Ford describes the inclusion/exclusion criteria as “a living document,” but he is unequivocal about its importance. “Whether the unit is ER- or hospitalist-run, you need strict admission exclusion and inclusion criteria,” he says. “When you leave it up to an individual doctor and don’t use evidence-based criteria, you’ll end up shooting yourself in the foot.”

You may also end up missing the window of opportunity for lowering costs and length of stay. According to 2006 statistics, hospitals should expect about 25% of observation unit patients to be admitted as inpatients. If that admission percentage runs much higher, the unit is likely accepting patients who should be admitted straight from the ED. If it’s less than 10%, on the other hand, the unit may be accepting patients who could be safely treated as outpatients.

Work in progress
The Methodist unit, which is run by a single hospitalist for admissions and one ED nurse, has since relaxed certain inclusion criteria. It now allows patients whose heart condition has been stable for several years and who are over age

The hospitalists at Cox Health’s Springfield unit also take a more broad-based approach than when they began. They admit primarily low to moderate risk patients. But the unit, managed by a single rotating hospitalist, also takes on a handful of high-risk patients, using the second of two distinct sets of inclusion criteria developed jointly by the hospitalists and cardiologists.

These patients, who have known cardiovascular disease and three or more cardiac risk factors, are accepted on a case-by-case basis as long as their cardiac enzymes are negative and their EKG is normal, Dr. Romero says. They are, however, treated differently.

“That population gets more aggressive therapy,” he says, with clopidogrel, anti-thrombotics and intravenous labetolol added to the mix.

As for gray-area patients, think of a 45-year-old male who had a 10-minute duration of chest pain but whose cardiac enzymes and EKG are normal. Because there’s a family history of heart disease, the patient is probably a good candidate for observation care, with repeat enzymes and EKGs, Dr. Brooksbank explains.

A same-age, same-presentation patient with diabetes, high blood pressure and hyperlipidemia, on the other hand, probably ought to be admitted as moderate-risk. “That’s because you believe the patient’s chance of coronary artery disease is much higher,” he says.

Standardized, expedited testing a must
Regardless of the admission protocols that units decide upon, hospitalists need to aim for general consensus with ED physicians and cardiologists.

“There are a lot of politics,” says Methodist’s Dr. Larson, “but the important thing about a decision unit is that it needs to be protocol-driven. Things have to be standardized so that everybody gets stress tests and enzymes done at the same time.” While some cardiologists do stress testing at four hours into observation, for example, others will wait 12 hours. Methodist has opted for a middle-of-the-road approach, with enzyme levels at three and six hours, and a stress test if the six-hour set remains negative.

Ensuring that tests are done when needed can be difficult in high-volume centers or during non-peak staffing times. For that reason, Drs. Ford and Larson urge hospitalists tapped to launch observation units to work out the logistics of lab and diagnostic scheduling well ahead of time. For stress tests, physicians should request designated spots or expedited status.

“If a patient who needs a stress test is sixth on the list, you’re not doing your unit a good service,” Dr. Larson points out. “You need to meet with the heads of lab and cardiology, and have them designate spots for you so your patients can get out quicker.” Pulling that off might require calling in some heavy hitters; in Methodist’s case, the chief medical officer got involved in the negotiations.

Addressing the costs associated with starting the unit may also require some fancy political footwork and the ability to “sell” the concept, cautions Dr. Ford. “There are a lot of upfront costs, but hospitals should take a long-term view, particularly in this pay-for-performance era,” he says. “In the long run, your length of stay and cost per case will go down.”

Bonnie Darves is a freelance writer specializing in health care. She is based in Lake Oswego, Ore.

Paying attention to getting paid

Whether hospitals have dedicated observation units or simply deliver observation care in a discrete setting, many are struggling to make sure those services meet reimbursement criteria from the Centers for Medicare and Medicaid Services (CMS).

Several years ago, the CMS reined in observation payment, notes Bret Nicks, MD, an emergency physician who directs the observation unit at Wake Forest University Baptist Medical Center in Winston-Salem, N.C. The agency also limited the number of reimbursable ambulatory payment classifications (APCs) that could be used in observation units, to avoid inappropriate use and charges.

In 2005, however, the CMS began reimbursing observation services for patients who stay a minimum of eight hours and a maximum of 48, for chest pain, asthma, heart failure and abdominal pain. Many private payers have followed suit.

But figuring out the complexities of billing and coding can be challenging. That’s because some fees for observation units are predetermined, while others “for breathing capacity and cardiac enzyme testing, for example “may be billed fee for service during the first 24 hours.

Many physicians, whether in the ED or the medicine units, aren’t well-versed in those basic criteria, Dr. Nicks contends. And if a patient admitted to the hospital should have been on observation status, that admission claim may be denied.

“At our hospital, we have initiated a program to correct the large number of one-day-stay admissions that should have been placed in observation,” Dr. Nicks points out. “There are major financial ramifications.”

Considering other diagnoses
Part of the solution should be a crackerjack case manager on the scene, one who understands the nuances of both patient and payer eligibility and CMS reimbursement. According to Steve Nahm, vice president of The Camden Group, a national practice consulting group in El Segundo, Calif., “the interaction of case management and the hospitalist is key regarding observation status” if hospitals “and physicians “hope to maximize reimbursement.

The good news is that the CMS is now considering expanding the number of APCs and the timeframe for reimbursable observation care, in certain circumstances.

And physicians heading up units are now considering conditions that might be good candidates for observation, regardless of APC status. Dr. Nicks cites pyelonephritis, cellulitis, mild traumatic brain injury, sickle cell crises and transient ischemic attacks as several examples. While the CMS may not reimburse the care of these patients under observation status, he says, the agency would likely cover physician services as well as clinically appropriate testing.

“There’s a push right now,” says Dr. Nicks, “to see if it makes more sense in terms of better outcomes and reduced re-admissions to manage some of these diagnoses in observation.”

Doing it right: protocols, staffing and support

Hospitalist programs considering launching an observation unit or taking over a fledgling one have their work cut out for them. Here are tips to ensure smoother management:

Work out nursing issues in advance. Observation-unit nurses should have requisite experience or training in the common diagnoses seen in the unit. Otherwise, they won’t be able to handle certain clinical changes.

“If the patient has persistent chest pain, instead of calling the doctor and asking what to do, the nurse should know to get an EKG, give the patient morphine and nitro, and then call the hospitalist with results,” advises William Ford, MD, section chief for hospital medicine at Temple University Hospital in Philadelphia.

Hospitalists also shouldn’t assume that ED nurses want to staff the unit, advises hospitalist Deanna Larson, MD, who runs the observation unit at Methodist Hospital in Omaha, Neb. “Many ED nurses don’t want routine admission work “and don’t want to be separated from their colleagues,” she says. Methodist solved that dilemma by rotating a core group of ED nurses through the unit.

Don’t lose sight of the “23-hour window.” Observation care is ideally concluded within 24 hours, when physicians must document either the patient’s need to remain in observation or to be admitted or discharged. Hitting that 24-hour window requires constant awareness of the “ticking clock,” says Bret Nicks, MD, an emergency physician who directs the observation unit at Wake Forest University Baptist Medical Center in Winston-Salem, N.C.

By the time a patient has been in the unit between 12 and 14 hours, he explains, the decision should basically be made whether the patient will go home or will require more time. “Don’t make that decision at 22 hours,” he says. Having strict, well-written protocols, he adds, makes that decision-making process go more smoothly.

Rotate hospitalists through the unit. Working the observation unit can lead to burnout, particularly when volumes are high. It’s best, sources say, to “share the pain.”

Many hospitalists who head up units discourage the idea of splitting hospitalists between medicine units and observation care on a single shift, for logistical and practical reasons. Instead, “you want to rotate all of your hospitalists through the unit, because it can be fast-paced,” Dr. Ford says, citing the 15 patients a day who might end up in observation at Temple and the half-dozen new admits over 12 hours.