Shortage of Nurses Hurts Patient Care, Study Finds
May 30, 2002
By DENISE GRADY
In hospitals with low numbers of registered nurses,
patients are more likely to suffer complications like
urinary infections and pneumonia, to stay in the hospital
longer and to die from treatable conditions like shock or
gastrointestinal bleeding, researchers report today.
"I estimate that hundreds or, perhaps, thousands of deaths
each year are due to low staffing," said Dr. Jack
Needleman, an economist at the Harvard School of Public
Health and the lead author of a study on staffing published
today in The New England Journal of Medicine.
The nation has a serious nursing shortage, with 126,000
jobs unfilled, 12 percent of capacity, says the American
Hospital Association. The shortage is a result of hospital
mergers, layoffs and heavy workloads. Many hospital nurses
shifted to other work. The average salary, $46,000 a year,
has not increased much in a decade. Even though hospitals
are trying to hire again, nurses are no longer available.
The new study, paid mostly by the government, is one of the
largest to look at care in relation to staffing. The
findings were based on a computer analysis of the discharge
records of more than six million patients at 799 hospitals
in 11 states in 1997. Five million had medical problems.
One million had surgery.
The research sought to find out whether there was a
correlation between how patients fared and the levels of
registered nurses and two other types of workers with less
training, licensed practical nurses and aides.
Registered nurses receive several years of training and
usually hold associate's or bachelor's degrees. Practical
nurses usually have high school diplomas and a year of
training. Aides have less training.
The researchers found that for certain aspects of care just
nurses made a difference. In hospitals with higher nurse
staffing, stays were 3 to 5 percent shorter, and
complication rates 2 to 9 percent lower than in hospitals
with lower staffing. No relation was found between
patients' well-being or length of stay and the levels of
aides or practical nurses.
Registered nurses, Dr. Needleman said, are "the eyes and
ears of the hospital" for judging whether a patient is
"If something is going wrong," he said, "they can catch the
signs early, before the problem gets worse."
But problems cannot be spotted early, he added, if nurses
do not have time to observe their patients.
The president of the American Nurses Association, Mary E.
Foley, said, "We're very pleased that another large set of
research data has validated what the American Nurses
Association has been saying since the mid-90's, that there
is a direct effect on the outcome of patient care when you
have enough nurses."
As an example, Ms. Foley said, pneumonia and blood clots in
hospitalized patients often result from immobility, and
nurses help prevent such complications by turning and
repositioning bedridden patients, encouraging them to
exercise and cough and helping others to get up and walk
around. Even if nurses themselves do not do those things
for patients, they have the training and experience to know
when such help is needed and to ensure it is provided.
A senior vice president for policy at the American Hospital
Association, Carmela Coyle, said: "The study suggests that
registered nurses are key and critical to insuring good
patient care. This is something that hospitals and patients
have known for some time."
Ms. Coyle said she feared that the study would lead some
people to assume that health care problems could be solved
by laws that require specific staffing ratios, a move that
hospital trade groups have long resisted. She called such a
reaction overly simplistic.
"What you really need to look at," Ms. Coyle said, "is much
more than the number of nurses and the number of patients.
How many of the nurses have 20 years of experience or more,
and how many are right out of school? What technology is
there to support them? How sick are the patients?"
Dr. Needleman said his group made its study because the
government sought to fill a large research gap on the
connection between nurse staffing and quality of care. In
the early 1990's, nurses testified to Congress that
patients in some hospitals were endangered by managed-care
companies that were trying to save money by replacing
registered nurses with less skilled and lower paid workers.
Congress asked the Institute of Medicine to study the
issue. In a report in 1996, the institute said there was
not enough data to draw conclusions and called on the
government to finance studies.
Dr. Needleman's is one of those. It measured staffing two
ways, by the proportion of nursing hours provided by each
type of worker and by the number of hours a day the types
of workers devoted to each patient.
For registered nurses, the average number of hours a day
per patient worked out to eight. The hours included not
only time spent at the bedside, but also time spent
preparing medication, writing chart reports and consulting
colleagues or family members.
The averages include intensive care units, with one nurse
for every one or two patients around the clock.
"There were some hospitals, that if I were going to them as
a patient, I would be very concerned," Dr. Needleman said.
"The hospitals at the very low end of our sample had as
little as two hours, and some at the high end had 16
When the study compared the hospitals in the top quarter
with those in the bottom quarter on the proportion of
nursing time from registered nurses, it found that the
medical patients, as opposed to surgical patients, in the
bottom quarter had stays 3.5 percent longer, 9 percent more
urinary infections, 5.1 percent more gastrointestinal
bleeding, 6.4 percent more pneumonia and 9.4 percent more
shock or cardiac arrest.
In addition, the death rate was 2.5 percent higher for
"failure to rescue," meaning that the patients died from
conditions that might have been reversed if they had been
treated in time. Those conditions include pneumonia, shock
or cardiac arrest, upper gastrointestinal bleeding, sepsis
or a blood clot.