Longer hospital stays do not mean better care

Sep 27 2005

Physicians at The Johns Hopkins Hospital (JHH) have disproved the notion that longer hospital stays mean better care. They have successfully cut back on wait times across one dozen hospital departments and, as a result, reduced to well below six the average number of days patients with congestive heart failure, a need for dialysis or surgery, and many other conditions must spend in the hospital.

From the outset, the Hopkins program was designed to improve patient care and cut down on wait times. According to statistics released today by the hospital, the program has led to a 4 percent decrease in the overall amount of time patients have to spend in any one of the hospital's 850 acute-care beds. On average, inpatients are spending a quarter of a day less waiting for their procedures, and the average hospital stay has been reduced from 5.93 days in 2002 to 5.7 days in 2005. This savings is equivalent to having added 26 new beds to the hospital.

And the denial rate for Hopkins by health insurers, who do not always agree to pay for days spent in the hospital that cannot be justified, has also declined. In 2001, Hopkins' denial rate was 2.08 days for every 100 patient days claimed in submissions to insurers. In 2003 and, again, in 2004, the rate fell to 1.85 days.

On its own, the hospital's Department of Medicine, which initiated the program, expects to decrease the waiting time for its 220 beds, representing 20 percent of all JHH beds, and thereby allowing it to see 14,000 patient admissions in 2005, 500 more than in 2004 and 2,200 more than in 2001.

Specifically, in the last year, the data-driven program has led to 78 percent of the department's admissions to acute-care beds doing better than nationally published industry standards for length of stay for a variety of specific diagnoses.

"Our results show that there is room for improvement in the way large teaching hospitals deliver high-quality care and that they can provide more of it without long wait times and to the better satisfaction of the patients we serve," says JHH physician in chief Myron Weisfeldt, M.D., who also is a professor and director of the Department of Medicine at The Johns Hopkins University School of Medicine. "Our next step is to strengthen this program as part of the management culture within Hopkins and to see if other hospitals can adopt our efforts to form the basis of a national program."

Weisfeldt noted that "spending four or five days in the hospital is not uncommon for many patients and for a variety of illnesses and conditions, but we need to question whether or not it is really necessary to spend so much time in an acute-care setting."

Indeed, he added, "most patients would rather not be in a hospital, preferring to return home as quickly as possible and, provided that it is safe to do so, to finish recuperating in the comfort of their own surroundings." The goal, Weisfeldt says, is to provide the best care, and "sometimes that means longer stays, sometimes not."

Historically, the dilemma for physicians and patients, he says, has always been to determine when it is best to leave the hospital, confident that the patient is on a road to recovery and that the patient's care has been delivered in the most effective and efficient manner possible.

Fiscal pressures compound the dilemma, with government, private payers and regulators eager to justify time spent and keep expensive hospital stays to a minimum. An average day of hospital care can cost $1,500.

"It is a tough judgment call that requires more than medical training; it requires constant questioning and review of how we manage a large hospital, practice medicine and treat our patients," says kidney specialist Paul Scheel Jr., M.D., an associate professor and vice chairman of medicine at Hopkins.

To put some evidence behind long-held suspicions that patients could be served better with a shorter length of stay, Hopkins began a more intensive look at its practices after a nearby community hospital closed in 2000. Hopkins physicians feared a spillover effect and influx of new patients that would strain its capacity. Any significant increases in demand for beds, they decided, could not be solved by expanding the number of beds and would have to be met by greater efficiencies from within the existing system. But they were not exactly sure where to find them.

Taking cues from best practices and management review systems already in place in the airline industry and at NASA, the Hopkins team set out to do a step-by-step analysis of how hospital services, such as tests and treatments, are managed as patients move among them.

A team of 10 physicians from the Department of Medicine first collected data from insurers and hospital information systems and identified specific procedures and physicians whose patients were spending longer-than-average times in the hospital.

All hospitals in the United States are required to collect and report information on how long they take to treat patients and for what diseases. The federal government then publishes this information as a list of procedures, grouped according to the severity of illness and with a number for each procedure. These are the so-called diagnostic-related group scores, or DRGs for short.

The Hopkins data was compared to those of other hospitals by using scores available from Millman and Robertson Inc., a well-known U.S. actuarial firm that has for decades calculated average lengths of stay for hundreds of hospital procedures and diagnoses based on their DRGs. The Department of Medicine team focused on procedures for which Hopkins had higher-than-average Millman scores.

The team not only interviewed physicians whose patients stayed longer than anticipated but also reviewed the patients' files in search of common sources of delay in delivery of care. Long waits for tests were found in all cases under review, and to the surprise of many, delays were frequently fixable. Over the four years of the program, the Hopkins team reviewed more than 200 different procedures and more than 5,000 patient files.

Among them were patients with congestive heart failure, for which patients were spending an average of six or seven days in early 2002, well above industry norms. Interviews with cardiologists revealed a nagging problem of one-day or two-day waits for ultrasound results, which are needed to update a patient's condition before discharge. In response, the Department of Radiology agreed to a policy of "same-day" discharge service for ultrasound procedures and revised its lab hours when necessary to accommodate the workload. As a result, length of stay fell to an average of 4.7 days by early 2003, below the Millman average of 4.8 days. The radiology department went even further, altering its consultation schedule to fit demand and alleviate backlogs and moving to a seven-day-per-week schedule for elective, or non-emergency, cases.

Perhaps the most dramatic example was applied to patients needing intravenous infusions because they were experiencing organ rejection after transplantation, or as part of their treatment for lupus. Before 2002, patients were admitted to Hopkins while they underwent once-daily infusions, which must be performed by a nurse. Treatment periods sometimes lasted as long as 14 days. Finding these extended wait times inconvenient for patients, the Hopkins team established an outpatient intravenous center, allowing patients to go home after one or two days before having to return to the hospital as outpatients for their remaining infusions.

"Same-day" or "48-hour" consultation service policies were introduced for other procedures, such as visits by social workers to patients newly diagnosed with HIV, and performance of colonoscopies and cardiac catheterizations.

Meanwhile, patient readmission rates stayed constant, with no increases, especially for the services and procedures in which average lengths of stay decreased, according to Scheel, who co-chaired the physician review team.

One of the biggest changes came in the hospital's treatment of patients with renal failure, many of whom lacked insurance and came to the Hopkins' emergency room for dialysis. According to Scheel, these patients were admitted to Hopkins for their routine dialysis, at a cost of tens of thousands of dollars, until their application for state Medicaid was approved and they became eligible for publicly funded dialysis in an outpatient setting. Before 2004, some patients waited in the hospital for as long as three months, and average stays lasted more than 20 days.

To bridge the gap between the emergency visit and receipt of state aid, Hopkins contracted with a community provider of outpatient dialysis services until the patients' Medicaid application was approved. At a cost of $120 per patient per treatment, or $360 per week, the community service was much cheaper than hospital care, and importantly, the move freed up beds for more acutely ill patients. Hopkins social workers experienced with Medicaid helped identify which patients were most likely to be accepted by Medicaid and which patients could be offered the community service. This augmented service immediately dropped Hopkins' Millman score for renal failure to 4.9 days, below the industry average of 5.4 days.

Within six months of the program's initial success, all other departments at JHH quickly followed suit by the end of 2003, establishing their own physician-review teams and with similarly dramatic results.

In Hopkins' surgical department, for example, better emphasis on discharge planning allowed it to achieve among the highest gains, with wait times and length of stay declining steadily from close to six days in 2002. Within the last two years, the average stay has decreased to 4.49 days, almost a half-day less than the average Millman industry score of 4.95 days. The department attributes some of this progress to revised job descriptions, which now include discharge planning, for more than a dozen nurse practitioners and physician assistants. This planning includes assistance and counseling for the family as they prepare for the patient's return home.

According to Hopkins' chief of surgery, Julie Freischlag, M.D., "Heading home from the hospital after a traumatic illness is a very emotional time for patients and their families. It is a major step in recovery for any patient to leave the hospital, a place where they have grown to feel safe and secure.

"But, it was always our suspicion that we could do better at helping our patients prepare to go home," she adds. "Up to 40 percent of our patients come from far away, traveling more than 100 miles for procedures only available at Hopkins. So, there are a lot of logistical details in planning for their discharge, such as who picks up the patient from the hospital and when, what steps make the home ready for the patient, some of whom may still require rehabilitation or outpatient therapy or have interim needs for special transportation. Unless there is a counselor readily at hand to provide this discharge service, people will naturally put off making decisions until very late in the process."

Equally impressive were improvements in neurosurgery, where changes in the last year alone for administrative activities, such as scheduling procedures and filing rehabilitation applications, shortened wait times and average length of stay by 6 percent, from 4.92 days in 2004 to 4.73 days in 2005.

Routine, postoperative MRI scans are now scheduled before the patient's surgery, avoiding back-ups in scheduling with one- or two-day wait times. The Department of Neurosurgery now preapproves up to half of its patients with their insurance providers for postoperative rehabilitation. Though less than 10 percent of patients will ever require therapy, the preapprovals avoid any lengthy delays in the processing of applications that can also hold up a patient's discharge.

"Linking quality of patient care to efficient lengths of stay was a learning process for everyone involved," says neurosurgeon Henry Brem, M.D., professor and director of neursosurgery at Hopkins. "But, it was not until we looked at the numbers that we realized we really could do more with less, especially delays. And, overall, our patients are more satisfied with their care when they understand that all of their care needs have been met and when there are as few roadblocks to their leaving the hospital as soon as possible."