Current government thresholds for the "meaningful use" of computerized provider order entry in the hospital may not be high enough to actually save lives, according to a simulation conducted by researchers at the Rand Corporation.
The researchers estimated the impact on mortality in a sample of more than 2,500 hospitals used computerized provider order entry (CPOE) for 26%-50% of patients, and found that that level of use could reduce mortality for heart failure and acute myocardial infarction among hospitalized Medicare beneficiaries by 1.2% – not a statistically significant reduction.
But a second simulation found that mortality could be significantly reduced (2.1%) if CPOE were used for 51%-90% of patients hospitalized for those conditions. The findings were published Sept. 14 in the journal Health Affairs on Sept. 14 (doi: 10.1377/hlthaff.2011.0245).
The results could help influence policy makers as they set standards for the later stages of the electronic health records (EHR) incentive program, the authors noted. The program was authorized by Congress in the 2009 Health Information Technology for Economic and Clinical Health (HITECH) Act and will provide about $27 billion in payments to physicians, hospitals, and other providers by the end of 2016.
Under stage 1 of the program, hospitals are required to use CPOE on medication orders for at least 30% of their eligible patients; under initial regulations governing the program, the threshold could rise to 60% in stage 2 and 80% in stage 3 of the program.
However, even the initial requirements for CPOE have been criticized by health care providers as being too strict. And recently, the Health Information Technology Policy Committee convened by the Health and Human Services department to help implement the incentive program recommended delaying the implementation of stage 2 requirements by a year.
The Rand study results could be ammunition for policy makers seeking to make the program more robust.
"Our results do support the notion that the increased thresholds for computerized medication order entry proposed for later stages of the 'meaningful use' regulations have the potential to yield tangible benefits," Spencer S. Jones, an associate information scientist at the Rand Corp. in Boston and lead author of the study, said in an interview. "Our study should reassure policy makers at HHS and other stakeholders that high levels of use of computerized provider order entry and other health information technology have value and are likely to yield tangible health benefits for patients."
But Mr. Jones added that policy makers have a tough decision to make in this area because there is currently no research setting out the optimal pace for EHR adoption. While proceeding at a cautious pace may be an appropriate strategy, maintaining parallel paper and electronic systems could lead to unintended adverse consequences, he said.
The study, which relied on data from 2,543 privately owned general acute hospitals, also compared the impact of any use of CPOE to no use and found that hospitals that used CPOE even a small amount achieved lower mortality rates for heart failure, acute MI, and pneumonia among Medicare beneficiaries. However, when the researchers adjusted for potential confounding factors that might affect mortality, there was a statistically significant relationship between CPOE use and lower mortality only for acute MI and heart failure.
It’s unclear why pneumonia mortality was not significantly affected by CPOE use. It could be that CPOE is more helpful in reducing medication errors for such complex chronic conditions as heart failure than for acute conditions like pneumonia, the researchers wrote.
The researchers analyzed data from the 2007 American Hospital Association annual survey database and the AHA’s Information Technology Supplement from the same year. They obtained mortality data from the Centers for Medicare and Medicaid Services’ Hospital Compare database, which listed risk-standardized 30-day mortality rate for Medicare beneficiaries admitted to the hospital with acute MI, heart failure, or pneumonia.
The researchers disclosed no conflicts of interest. The study was funded by philanthropic contributions from members of the Rand Health Board of Advisors.