The trusted source for
healthcare information and
Focused service, better than being stuck in hospital’
When Jane Jones’ neighbor drives her to the emergency department (ED) at Denver’s Exempla St. Joseph Hospital, the 80-year-old is dehydrated and too weak to heat a can of soup, but she’s not sick enough to admit — not yet. If she goes back home where she lives by herself, however, she’ll probably become a "bounce-back," an inpatient admission within 24 to 72 hours.
To prevent that, the director of the ED’s observation unit, David Magid, MD, MPH, keeps her overnight. "We can show that the unit improves clinical outcomes, as evidenced by the bounce-back rate," he says. With some patients, there’s the chance of underestimating the severity of a problem during a brief emergency visit. "The observation unit allows us to watch the trajectory of a condition," he notes. "Does it flatten or does it get worse?"
Since its creation two years ago, the unit has grown from six to 11 beds, while the bounce-back rate of 3.8%, out of the 50,000 annual ED visits, has dropped by more than half. Patient satisfaction runs high, and Magid estimates that the unit saves Exempla St. Joseph about $1 million a year. (See "Observation Units: National Highlights," p. 22.)
Admission depends more on the reason for observation than on specific problems. Criteria fall into three categories:
1. Diagnosis and treatment.
Some patients come to the ED with a clear diagnosis. They improve with treatment, but not enough to return safely to their homes. Jane Jones is one example. Many asthma sufferers fall into this group as well. When a particularly severe attack brings them to the ED, they typically receive steroids and breathing therapy. "While we might see some immediate improvement, it could take six hours for the steroids to kick in. So we keep them in observation, give them another breathing treatment, and send them home after they stabilize," Magid explains.
Sometimes, patients with kidney infections become so dehydrated that they vomit when they swallow antibiotic pills. In the observation unit, they might receive a liter of intravenous fluids for rehydration, which stops the vomiting. Most get better as soon as they can keep the meds down.
2. Diagnostic dilemma.
Chest pain is the most common problem in this category. For the initial evaluation of patients with questionable symptoms, observation is ideal. "Before we had the unit," notes Magid, "we found that we sent some patients home after an EKG and they would have a heart attack."
Twenty-four-hour physician staffing packs a high concentration of service into a 12- to 24-hour observation. For inpatients, the same processes would take longer, depending on the frequency of the attending physician’s rounds. The unit offers the additional benefit of treadmill tests followed immediately by interpretation of the results.
This enables patients to go home reassured or to receive rapid treatment if problems show up. By contrast, patients who leave the ED with orders to get treadmill tests often have to wait several weeks for an appointment. Louis Graff, MD, associate director of the ED at New Britain (CT) General Hospital, reports that patients like the efficiency of observation units: "They would rather go there [than to an inpatient unit] for a cardiac work-up because it’s focused, organized care. For them, it’s better than being stuck in the hospital for several days."
3. Safe haven during transition.
When a person is in a steady decline due to old age or chronic illness, a bout of flu or an infection can hasten an anticipated move from independent living to a higher level of care.
"In the past, a family might bring the patient, and we would admit [him or her]. It really wasn’t necessary, and it exposed people to infections," Magid says. An observation unit affords the relatives or case worker a respite of up to 23 hours in which to arrange the move.
Aside from an intermediate length of stay, observation care has other distinct features. It’s far from a catchall, Magid emphasizes. Prior to admission, patients are evaluated for risk of deterioration and intensity of care required. Protocols and checklists screen observation cases from inpatient cases.
Patients with identifiable heart problems go directly to an inpatient unit. "We prefer not to start an inpatient admission in the observation unit. If patients will require one-on-one nursing care, we prefer to admit them to the hospital," says Magid. Even so, one out of six observation patients ends up being admitted. And that’s as it should be, he notes. "If there were no admissions, that would tell us we were probably taking care of patients who should not have been in the unit anyway."
Within six months after Exempla St. Joseph opened the observation unit, the bounce-back rate in the ED went down 60% and has held steady ever since. "We didn’t know how things were going to work out when we designed the unit, but the hospital thought it would be good for patients so they approved it."