Published in the August 2004 issue of Today’s Hospitalist
When James P. Marcin, MD, MPH, began examining mortality rates in pediatric ICUs, he was pleasantly surprised to find that patients admitted during the weekend fared about the same as patients admitted during the week. Studies of adult populations, after all, have found significant differences between mortality rates for patients admitted during the weekend, and the preliminary data seemed to be good news for pediatrics.
That sense of optimism began to fade, however, when Dr. Marcin and his colleagues began comparing mortality rates for patients admitted during the night to patients admitted during the day.
In looking at more than 20,000 admissions to 15 pediatric ICUs between 1995 and 2001, mortality rates initially seemed similar. That changed when researchers adjusted admissions for severity of illness. They found that patients admitted between 5 p.m. and 7 a.m. faced a 28 percent higher risk of death within 48 hours than patients admitted during the daytime.
While unadjusted mortality rates between the two groups were similar, researchers discovered that patients admitted at night tended to be less sick. After risk-adjusting the data, investigators concluded that fewer of the patients admitted at night should have died because of their reduced severity of illness.
Further analysis found that patients in three categories ” children suffering from shock, congenital cardiovascular disease and cardiac arrest “faced the greatest risk of death when admitted at night compared to similar patients admitted during the daytime. For most other patients, mortality risk was the same whether children were admitted at night or during the day.
The importance of initial therapy
Dr. Marcin’s study, which was published in the June 2004 issue of Pediatrics, is only the latest study to examine the impact of admission time on mortality. As the article in Pediatrics points out, several studies have found that mortality rates for children born at night were higher than those born during the day. Studies of adult populations have similarly found that time of admission can affect mortality and other outcomes.
Many of the studies on time of admission, however, have tried to distinguish between weekend and weekday admissions, not nighttime vs. daytime admissions. That’s why Dr. Marcin and his colleagues were looking for differences in weekend admissions “and were surprised when they found relatively little variation.
Why did patients at night die more frequently than those admitted during the day? Dr. Marcin, assistant professor of pediatrics and critical care medicine at the University of California Davis Children’s Hospital in Sacramento and the senior author of the study, points to the challenges of treating the three subgroups.
“These patients are the sickest and are most likely to need the most immediate attention from nurses, physicians and other providers,” he explains. “For example, when you’re treating kids who come to the ICU with a congenital cardiac disease, you need an experienced pediatric critical care physician and/or a cardiologist immediately available. You may also need quick access to a catheterization lab, or you may need emergent surgery. With these patients, their initial therapy is critical to their outcome.”
Obviously, these patients’ health is not the only factor in their above-average mortality risk. Dr. Marcin acknowledges that factors like staffing levels, caregiver behavior, fatigue, or any combination of the two, could be behind the elevated mortality rates.
He speculates that another factor might be found in patient loads. In some studies, research has found, for example, that more patients tend to be admitted during nighttime hours.
“Granted, these patients were a little less sick in our study than those admitted during daytime hours,” he explains. “Still, admission density could play some role.”
Staffing issues
None of those factors alone, however, would explain why children admitted at night would fare worse than those admitted during the day. Their severity of illness was on average lower than patients admitted during daytime hours, yet their mortality was still higher.
The obvious question is what part staffing levels play into the nighttime effect Dr. Marcin and his colleagues found. “I think most people would admit that hospital staffing is diminished during the night hours, both in numbers and perhaps in experience and quality,” he says.
He points to nurse staffing as one example. “In most pediatric ICUs,” Dr. Marcin explains, “when a day shift position opens up, it goes to nurses with seniority who have been working the night shift. New grads don’t generally go on the day shift. They usually start at night.”
Even seasoned physicians and nurses, he adds, are more prone to make errors at night than during the day, largely because of fatigue. “While people on night shift sleep during the day, they’re still more likely to get by on less sleep than people who work during the days,” Dr. Marcin explains. “That’s been studied and demonstrated over the years.”
Other possible factors could include the availability of services like diagnostic imaging. MRIs, angiography and X-rays are typically a little more difficult to come by at night, at least in part because of staffing issues.
The weekend effect
While night shifts are unpopular, so is weekend work, yet there were no differences in mortality rates between weekday and weekend admits. In addition, several studies have found a “weekend effect” in mortality rates in adult settings.
While Dr. Marcin admits that he has little in the way of hard data to answer that question, he does have a hunch: Pediatric ICUs typically have somewhat better attending coverage on weekends than adult ICUs. Also, he points out, while all pediatric ICUs are staffed with pediatric critical care physicians, many adult ICUs are not staffed with critical care physicians.
“I think there’s probably more of an attending presence in the pediatric ICU,” he explains. “In our children’s hospital, there are two or three pediatric attendings in the house at all times. In addition to the pediatric critical care physician, there are also pediatric anesthesiologists, and pediatricians from the ER or pediatric ward, but we generally have attendings present throughout the weekend. Compared to ICUs in the adult world, pediatric ICUs are much more regionalized, and therefore you’re more likely to find an attending in the pediatric ICU than in the adult world.”
As proof, he points to a study in JAMA that recently found there was no difference in the care provided on weekends in neonatal ICUs.
Dr. Marcin hopes to be able to delve into some of these factors in the near future. The federal database he used to compile data for his most recent study is already collecting some information about the experience level of physicians and nurses. He hopes that in the next few years, more in-depth data from pediatric ICUs will allow him to correlate outcomes with the experience levels of health care providers.
“It’s further evidence that there needs to be a call for health care providers, health care administrators and health policy makers to look at differences in quality of care, differences in personnel, differences in staffing and everything else that happens differently at night than during the day,” he says.
