I’m often asked, what impact can a tool like ACR Select have in a typical Hospital inpatient imaging environment?
I think the truth is that we all know that there is a LOT of wasteful use of inpatient (and generally unmanaged) imaging services. I often chuckle when I recall many of my radiologist friends telling me that they can tell which exams are ‘normal’ just based on looking at their worklist and seeing the ordering clinician and reason for exam!
The easiest way to look at this is to simply compare this to the outpatient environment where a market goes to a pre-authorization or Clinical Decision Support (CDS) program. The typical reduction is about 10-12%.
I believe that at a minimum, CDS will replicate what happens in the outpatient environment. There are however, MANY variables that can affect this number.
I have a pretty ‘sobering’ belief as it relates to the impact we can create: 20-30% of all inpatient imaging should either probably have never been done, or was ordered incorrectly (i.e. wrong modality). Lets look at some the reasons why:
- Practice protocols and order sets; much of the inpatient imaging is done as a matter of ‘protocol’ or assumptions that are built into order sets that haven’t been reviewed or changed since inception. This is a big problem and I believe that a review of these practices and order sets is an important first step. Using ACR Select to review these is a big first step.
- There is very little interaction between the ‘expert’ (the rad) and the ordering clinician. This has to change. The radiologist needs to be incentivized to get involved in consults and the ordering clinician must know when a consult is necessary. Getting feedback from ACR Select plays a big role in identifying cases that are ‘borderline’ and suggest that a radiologist consultation would help to improve the decision.
- Ordering clinicians get no feedback on the clinical efficacy for the exams they are ordering. Most orderers order exams because “its what we always do”. Clearly, ACR Select solves this in an elegant way.
- Duplicate and repeat studies. This is a no brainer and has nothing to do with ACR Select. Having a proper ‘duplicate exam alert’ with report display and images in the CPOE is imperative to reduce over-utilization due to this.
- Self protection is a big issue too. Safe harbor for using guidelines/criteria will certainly impact what happens. We are long way from getting this, but this WILL happen. It has to!
The impact of ACR Select (and CDS in general) will be determined by how you implement. There are a few options. I’ll list in order of ‘boldness’:
- Disallow ALL orders that don’t reach a certain score threshold and require a consultation with Radiology. This will have the most impact.
- Display scores and ‘speed bump’ low scores by asking the orderer ‘why’ they are ordering a low scoring exam as well as offering a live consult with radiology. This is very effective and has been the most commonly used method.
- Display scores and offer no speed bumps. The doctors will see the low score and know data is being collected and this will affect behavior. Nobody ‘wants to be wrong’.
- Display no scores, but require that they select reason for exam based on the indications provided by ACR Select so at a minumum we have ‘structured reasons’ that can be analyzed.
In ALL of the above scenarios, there should be reporting tools used to analyze ordering patterns that identify the outliers so they can be educated. Keep in mind, our experience shows that inappropriate ordering is highly correlated to the ordering clinician, i.e. if you have a 25% low score rate, the data will show that they come primarily from the same group of doctors that represent just 50 or 60% of the doctors. And in the case, it will be usually for a specific set of exams. So, educating that smaller population will go a long way towards reducing waste through unnecessary utilization.