Study Finds Little "Weekend Effect" Related To Intensive Care Admissions

Not so, according to investigators at the University of Iowa and the Veterans Affairs Medical Center in Iowa City who analyzed the relationship between outcomes and day of the week of admission to intensive care units (ICUs). The study, which was published in the June issue of Medical Care, tracked 156,136 patients admitted to 38 ICUs in 28 northeast Ohio area hospitals from 1991 to 1997.

The researchers found that patients admitted to ICUs on the weekend had nearly the same odds of death as those admitted on Friday or Monday. Overall, the increased risk of death for Friday through Monday admissions was only about 9 percent higher than the mortality odds for ICU patients admitted Tuesday through Thursday. In addition, people admitted Friday through Monday stayed only about 4 percent longer than those admitted mid-week.

The differences in outcomes likely have more to do with patients' conditions than with staffing during the six-year time period studied, said Gary Rosenthal, M.D., UI professor of internal medicine and director of the UI Division of General Internal Medicine, and one of the paper's authors.

The investigators said they were somewhat surprised by the findings, given that hospitals often have different patterns of staffing on the weekends compared to the weekdays.

"It turned out that most of the differences were relatively small with respect to mortality and length of stay. We believe those differences were due to the acuity, or severity, of patients' illnesses, not to staffing," said Rosenthal, who also is a staff physician and researcher with the Veterans Affairs Medical Center (VAMC) in Iowa City, where he directs the Program for Interdisciplinary Research in Health Care Organization.

The UI results challenge the conclusions of a paper on weekend and weekday hospital mortality, based on Canadian data for nearly the same period (1988 to 1997), that was published by other investigators in the New England Journal of Medicine in August 2001.

That study found a higher incidence of hospital deaths for three life-threatening conditions -- rupture of the major artery to the stomach, serious inflammation of the epiglottis (the cartilage that stops food from entering the windpipe), pulmonary embolism (a blocked blood vessel in the lung) and 23 other of the most common 100 hospital deaths. At the same time, that study found that for the remaining majority of those 100 conditions, including serious ones such as heart attacks, brain hemorrhages and acute hip fractures, no increased risk of death was associated with weekend admissions.

However, those Canadian findings, as reported in some media stories last year, made it sound as if it is significantly riskier to be admitted to hospital on a weekend, said Mitch Barnett, health services research specialist with the Iowa City VAMC and another of the paper's authors.

"There may be a small adverse affect associated with receiving care on the weekends. However, it's extremely small and not nearly as grave as what the other authors found," said Barnett, who also is a doctoral student in the UI College of Pharmacy. "Our study did a better job of adjusting for severity of illness. People who show up on the weekend tend to be much sicker than those who seek care during the week."

Rosenthal also stressed the importance of adjustments in order to get a true picture of risk.

"The data used in the prior New England Journal of Medicine study to adjust for how sick patients were at the time of admission were relatively crude. We utilized much more sophisticated methods of adjusting for severity of illness," he said.

The UI study involved clinical data and a sophisticated method called APACHE to adjust for severity of illness, unlike the Canadian study, which used less informative insurance claims.

However, Rosenthal and Barnett urged caution in extrapolating their findings.

First, their study did find a small weekend--admission disadvantage for certain types of surgical procedures such as procedures to treat lung and brain cancer or to relieve narrowing of the carotid or peripheral arteries, which made up a relatively small part of the team's overall study population. In addition, because the UI findings were based on data from 1991 to 1997, it is possible that new studies might show a slightly different picture.

"Since 1997 hospitals have been under increasing financial pressure, and nursing shortages have affected staffing patterns, so it is relevant to re-visit this question of weekend staffing," Rosenthal said. "The challenge in doing another study like this is getting access to data. We were fortunate in our study to have had access to a large amount of very detailed clinical data that allowed us to make the adjustments for severity of illness."

The database of patient information used by the team originated at Case Western Reserve University, where Rosenthal previously served on the faculty.

In addition to Rosenthal and Barnett, the study team included Peter Kaboli, M.D., a UI associate professor in internal medicine and an Iowa City VAMC staff physician, and Carl Sirio, M.D., assistant professor of anesthesiology, critical care medicine and internal medicine at the University of Pittsburgh School of Medicine.

The study was funded by a federal Department of Veterans Affairs Grant. In addition, Kaboli was supported in part by a VA Research Associate Award and a Quality Scholar's Fellowship.