Consider what we know and don't know about how capacity affects patient safety. Admission during off-hours for stroke and heart attack yields a five percent increase in mortality. (JAMA 2008;299:785; Am J Med 2007;120:422.) One of 25 deaths in patients with ST segment elevation heart attack can be attributed to ambulance diversion, according to a New York City study. (Inquiry 2010 Spring;47:81.) Two Australian studies that looked at short-term mortality rates found that admission into the hospital system during times of crowding is associated with 20 percent to 30 percent increased mortality. (MJA 2006;184:213; MJA 2006;184:208.)
We are supposed to be fixing “systems” problems. So what's holding us back?
Consider the following. In the 1960s, most admissions to the hospital were elective. Patient acuity was modest, hospital capacity was generous, and there was little concern for length of stay. Physicians had little reason to round more than once a day. Perhaps appropriately, hospitals were a 9-to-5, Monday through Friday operation, with a skeleton crew on evenings, nights, and weekends. Fast forward to our current environment. Most admissions are urgent or emergent, unscheduled, and occur with almost equal frequency seven days a week. Unfortunately, the hospital industry, in large part, continues to have a five-day-a-week response to a seven-day-a-week problem. The resulting traffic jams, more than any other single thing, have created our capacity problems. It also seems to be the problem we are most eager to ignore.
Over the past decade, there have been tremendous and successful efforts to eliminate queues in the emergency department, hastening the triage process, moving the registration process to the bedside, and moving patients directly into the clinical area rather than having them sit in the waiting room. (All that's needed now is an EMR capable of keeping pace with the speed of the environment.) Inadequate attention, however, has been paid to similar problems with the inpatient services.
Everywhere one looks at the inpatient services, one can find queues and batching by design. Elective admissions enter in far greater numbers at the beginning of the week, and there are fewer discharges on weekends than weekdays. Orders for labs, x-rays, and consults are batched each morning. In academic centers, morning rounds osmose into noontime conferences, leaving discharge orders to be written in the afternoon. Patients leave in the late afternoon, finally allowing for some of the boarded patients in the emergency department to move to the inpatient services. The timing of this places an extraordinary burden on the evening nursing staff, which first has to batch the discharges and then the admissions.
We know from the great work done by a group that included Eugene Litvak, PhD; Donald Berwick, MD; and Michael McManus, MD, that “smoothing” elective admissions (reducing their peaks and valleys) dramatically decreases capacity problems and directly benefits more than just the service that underwent smoothing. In one hospital, in the year prior to smoothing elective surgical cases, there were more than 700 cases cancelled because of urgent or emergent cases. In the subsequent year, with a slightly higher volume of case, seven cases were cancelled. (Yes, seven.) This practice also relieved ICU congestion, and sharply reduced ED crowding and ambulance diversion. Some very important work by Dr. Gabe Kelen's group demonstrated that close to 50 percent of hospital inpatients could be discharged if needed. (Disaster Med Public Health Preparedness 2009;3[Suppl 1]:S1.) On Sept. 11, 2001, 20 percent of all inpatients in the greater New York City area were discharged by noon that day. Hospital occupancy took months, not days, to rebound.
All of this begs the question: Should we be having capacity issues? Smoothing flow would mathematically require fewer resources and yield better capacity (which has been proven to be true in real life). It becomes all the more striking how resistant we are to any change, however modest or obvious.
Compare the following two styles. An emergency physician sees a patient who requires an urgent neurology consult (who, for the sake of this story, actually exists and is available). What happens? The neurologist sees the patient, and perhaps recommends an MRI. The MRI is obtained, interpreted by the radiologist, and the consult is then completed by the neurologist. This entire process may take hours. Compare how this may happen on the inpatient service. The patient is admitted. The attending sees the patient the next morning (day 2), and orders a neurology consult. The neurologist sees the patient the next morning (day 3), and recommends an MRI. The MRI is ordered that afternoon, and obtained the next morning (day 4). The MRI is interpreted by the radiologist that afternoon. The neurologist returns the next morning and completes the consult (day 5).
The core difference in these two tales is not a difference in skill but a difference in approach and culture. In the emergency department, we want to consult now, test now, results now, and an action plan now. This is, in large part, simply not the culture of the inpatient services. If one rounded on an inpatient unit, and asked which patients were still there only because they were waiting for a test, one might find that a distressing percentage of the patients really are not sick enough to require continued hospitalization. Although the average length of stay at my institution is a bit over five days (longer if the stay involves a weekend), the median length of stay is about three days. This is a very short time, when rounding once-a-day, to solve the patient's problems.
So here's a great research question. What if the “do it now” approach of the emergency physician was practiced on the inpatient units, where labs, x-rays, and consults were not ordered for tomorrow morning but for now? Do the test now. Get the results now. Act on the results now. Don't change anything else. Just change this behavior. What would the results be? How much capacity would be created? Would we stop endangering our patients who bear the consequences of our inaction?
Comments about this article? Write to EMN at firstname.lastname@example.org.