Hospitals that do lots of orthopedic procedures have a substantial advantage in measurable outcomes over those that do relatively few procedures, a new study has found.
Teaching hospitals have better outcomes than nonteaching hospitals, and teaching hospitals with orthopedic residency programs have the strongest results.
HCIA, a Baltimore-based healthcare information company, conducted the study. Since 1993 HCIA has chosen the 100 top hospitals in the U.S. and elevated them as benchmarks for the rest of the industry. The criteria include financial, clinical and operational measures.
"Practice makes perfect" is the theme of the new report. Those hospitals that perform more orthopedic procedures see better results as measured by the number of deaths and complications.
"The tables very clearly show that there is a huge difference in outcomes between the very highest-volume, the benchmark hospitals, and all the rest," says Jean Chenoweth, vice president of HCIA. "We haven't had that with the Top 100 hospitals up to now."
Yet one independent medical researcher questioned HCIA's methodology and said more research must be done to verify the results.
"The data available in the Medicare database are crude and don't permit risk stratification of patients," says Michael Rie, an anesthesiologist at University of Kentucky Medical Center in Lexington.
The HCIA report shows that high-volume programs with a low or moderate growth rate have the fewest deaths and complications, while programs that are growing quickly tend to have more complications.
And, remarkably, if all hip and knee replacements were performed by only the most experienced hospitals, patient lengths of stay would decrease by about one day per patient.
Hospitals that performed the fewest procedures-between 100 and 160-in 1998 had a mortality rate of 1.32 per 100 patients. Those that performed the most procedures-more than 433-had the lowest mortality rate, .87 per 100 patients.
Hospitals that completed the most orthopedic procedures were better able to prevent complications, too. The complication rate for the lowest-volume hospitals was 3.36 per 100, and for the highest-volume hospitals, 3.02 per 100.
"Where they do lots of these programs, they get superior results," Chenoweth says. "Over the years, complications have gone down, and severity of illness has gone up."
At teaching hospitals with orthopedic residency programs, benchmark hospitals had a median mortality index of 0.78. The index for the nonbenchmark teaching hospitals with residency programs was 0.95. An index of 1.0 is the national norm.
At teaching hospitals without orthopedic residency programs, the mortality index for the benchmark hospitals was 0.70, compared with 1.02 for the nonbenchmark teaching hospitals. And at nonteaching hospitals, the mortality index for benchmark hospitals was 0.54, almost half the 1.0 rate for the nonbenchmark hospitals in that category.
Beth Waibel, an analyst with HCIA, said even though the mortality index is adjusted for risk, that doesn't completely make up for the fact that teaching hospitals with residency programs have the highest case mix. Risk adjusting doesn't completely even the playing field.
In each group of hospitals, the complications index followed a similar pattern. Furthermore, costs per case were substantially lower at the benchmark hospitals, and lengths of stay were shorter.
Orthopedic care is one of the most common service lines in U.S. hospitals. Almost every hospital can provide some orthopedic care, and almost 4,600 hospitals offer major orthopedic services.
HCIA chose the Top 100 orthopedic hospitals from 73 teaching hospitals with orthopedic residency programs, 780 teaching hospitals without residencies and 1,313 nonteaching hospitals. The orthopedic study is a first for the company.
HCIA says that analyzing the data and determining the benchmark organizations will lead to an increase in the overall quality of orthopedic services. It also hopes to narrow the gap between the benchmark providers and the universe of orthopedic providers.
Rie says HCIA's proprietary methodologies for calculating results should be made public so that outsiders can evaluate the validity of the study.