UpToDate is one of several online services that provide frequently updated monographs incorporating the latest, most relevant medical knowledge. Clinicians can access them directly through Web browsers or launch them from electronic health records (EHRs) using desktops, laptops, or mobile devices.
The study in the Journal of Hospital Medicine compared the performance of 1,017 hospitals that used UpToDate to that of 2,305 facilities that did not, employing publicly available administrative databases. In addition, performance of UpToDate users over time was compared to their own prior performance.
During the three years covered by the study, UpToDate hospitals had slightly shorter average lengths of stay than the control group did (5.6 vs. 5.7 days), and the same was true for each of the six conditions studied. The study group also had slightly lower risk-adjusted 30-day mortality rates, although the difference was significant for only three of the conditions. On all four conditions studied for quality scores--heart attack, heart failure, pneumonia, and surgical infection prevention--the UpToDate users did better than non-users.
[ For background on e-prescribing tools, see 6 E-Prescribing Vendors To Watch.]
In general, the study found, smaller community hospitals benefited much more from UpToDate than did larger hospitals and academic medical centers. For example, the small hospitals that used UpToDate did better on the quality scores than did non-UpToDate hospitals, "but almost no effect was found among large hospitals," the study said.
The quality metrics used in the study are based on very well-accepted clinical guidelines. But the strength of clinical knowledge support systems is that they can help clinicians answer difficult questions that come up when they're treating patients. If the quality measures in the study are so well-known, why would UpToDate have any impact on physicians' performance on these metrics?
Thomas Isaac, MD, the lead investigator, agreed that some of the studied measures are very basic, such as whether aspirin and beta blockers have been given to heart attack survivors, or whether antibiotics were provided to patients with pneumonia in a timely manner.
"These are things that you typically wouldn't need to look up using something like UpToDate to find out whether it's a good idea," Isaac said. "But some of these quality metrics may have an association with each other, so that hospitals that focus a bit more on quality and choose to use tools like UpToDate may put in place processes that are important for quality throughout the hospital."
That brings up the question of causality. "I don't think this paper proves causality at all," Isaac noted. However, he added, "The study controlled for teaching status, bed size, presence of an ICU, and other factors, and we still found that UpToDate was associated with these better metrics." The fact that hospitals' performance tended to improve over time when they used UpToDate also suggested--but did not prove--causality, he added.
The fact that the company that owns UpToDate sponsored the study raises the question of whether it influenced the results. But Isaac denied that it did. "We took all precautions to remain objective. All of the analysis, writing of the manuscript, discussion, and findings were done by the authors while we were conducting the analysis or writing the paper. We didn't share our work or discuss it with UptoDate."
The new InformationWeek Healthcare supplement on EHR Best Practices explains how the most astute healthcare providers are putting those billions of dollars in federal stimulus funds to productive use. Download the supplement now. (Free registration required.)