How effective is healthcare technology so far?
Two very exciting articles were published this month in the Journal of Health Affairs:
This study uses data mining to gain a sample size of 2.952 hospitals. The hospitals either had a Comprehensive EHR, a Basic EHR, or no EHR. The key measures were risk adjusted length of stay (ALOS), risk adjusted 30 day readmission rates, and risk adjusted inpatient costs. The staggering and sobering results show there was no difference between EHR and non EHR facilities on quality measures such as AMI, CHF, and pneumonia. In addition, there were no differences with ALOS for the most part (pneumonia showed a minor decrease with EHR - 0.5 days), no difference with 30 day readmission rates, and no difference with risk adjusted total costs. What are we spending billions on again?
This study looked at the 62 hospitals participating in The Leapfrog Group for Patient Safety CPOE analysis tool. It found that systems caught drug-allergies in most cases, but did poorly at drug-diagnosis contraindications such as pregnancy. The interesting measurement was in the prevention of fatal drug doses, where they were only caught in 47% of cases. In addition, drug-lab and drug-age alerts only flagged appropriately in 21% of cases. The investigators noted the vendor chosen by the hospital played a statistically significant part in the outcome of testing, suggesting some vendors may build to the leapfrog test or standards. Systems also seem to do a better job detecting adverse drug events that occur infrequently, as opposed to those that occur frequently.
These studies parallel some of the results we have seen with Barcoded Medication Administration. However, in contrast we should consider the stages of adoption in most hospitals. In the US, we are still trying to learn how to use technology in our daily routines. Healthcare has been a late player in the game, and the impact on productivity and safety has been minimal. This is what we should expect. We need to give healthcare workers time to learn to use the technology to do their daily tasks. They are still struggling to find where menus are or buttons. They don't easily find things they used to flip to in a paper chart. Once integrated into their workflows, we can start to look at obtaining some real benefits from clinical decision support and interoperability. Interoperability has a long way to go, so we certainly have time to refine our EHRs for the better. My final impressions are we need to train more, take ownership of our EHRs and work to improve patient care through their use.
Comments
Patience young Jedi, patience
Thanks for the information, pillguy. I think it's important to take a step back and realize that the true benefit from a completely centralized EHR will take years to realize. Is anyone really promoting them as a way to improve outcomes because if they are I think they're missing the big picture.
Centralized EHR will create a future way of handling the ever increasing mountain of data we have across the continuum of healthcare. I for one think of them as a repository to prevent duplication and redundancy which will save millions of dollars. In addition there's the added benefit of not having to hunt through 3 or 4 systems to find information that could be conveniently stored in a single location.
As far as CPOE is concerned I'm not convinced that's it's the panacea that everyone thinks it is. We haven't taken the time to build a proper infrastructure to support the model. Nor have we spent enough time with CDS and automatic order processing to optimize its use.
Of course these are just my opinions. Take 'em for what they're worth.
Who is promoting them? Well
Who is promoting them? Well our government. The entire basis for pushing adoption has been cost savings and increased patient safety. However, no one said they would be realized immediately after implementation :)