Strategies to improve test utilisation
Some strategies may improve test utilization
· CPOE ordering templates
· CPOE alerts for redundant test orders
· CPOE ordering constraints to minimize recurring orders
· Display of relevant prior laboratory data
· Integration of practice guidelines into CPOE ordering screens
· Improving CPOE test search to permit improved test selection
· CPOE corollary test alerts
· Displaying test cost on CPOE ordering screen
· Unbundling CPOE test panels into their individual components
One of the most basic yet effective interventions to alter test utilization is to modify CPOE test ordering screens. A typical screen modification involves making commonly used and often appropriate tests more convenient to order than tests that are only occasionally indicated. Such modifications presumably reduce over-ordering of tests by reducing the “impulse-buy” mentality. Another strategy that may be employed is to “unbundle” testing panels, requiring physicians to select the individual tests instead of ordering the panel. Neilson et al., found that a combination of ordering frequency constraints and unbundling of metabolic panel tests decreased the use of the previously bundled tests by 51%.
Overutilization of inpatient laboratory tests may occur due to orders for recurring or “until discontinued” tests. CPOE systems that permit recurring orders may actually facilitate the overutilization of laboratory testing. With CPOE, clinicians may place recurring orders (e.g. CBC and electrolytes every morning until discontinued) but may fail to cancel them after the test is no longer clinically necessary. To curtail overutilization stemming from these types of orders, Vanderbilt University Hospital implemented CPOE pop-up boxes to alert clinicians of orders scheduled for longer than 72 h and offered clinicians the opportunity to voluntarily cancel these orders. A 24% reduction in orders for metabolic panel tests was attributed to this CPOE alert. CPOE systems provide a leverage point for restricting recurrent orders, although significant leadership may be required, as once recurrent orders are institutionalized on templates and in the culture of the hospital, they can be challenging to eliminate.
Ordering templates play an important role in standardizing care and encouraging adherence to clinical guidelines. Templates typically consist of an integrated order set including medication, laboratory and other orders appropriate for a particular clinical setting or diagnosis. For example, an admission template for myocardial infarction might include laboratory orders (e.g. serial troponin assays), medication orders (e.g. aspirin, beta blockers), specific dietary orders (e.g. low-salt diet), and nursing instructions. When placing orders for a specific patient, clinicians may start with the standard template and then make modifications to address the unique clinical circumstances of the patient. Another advantage of using CPOE templates is that when clinical guidelines change, templates can be readily updated and the change in guideline is thus made immediately apparent to all ordering providers.
CPOE provides a platform for informing clinicians of practice guidelines and facilitates the tracking of deviations from the guideline. The Massachusetts General Hospital Blood Transfusion service integrates evidence-based guidelines into CPOE blood product ordering screens.
For example, when ordering red blood cells (RBCs), clinicians must select an indication based on the patient’s age, hematocrit, and state of stress. Orders for RBCs not meeting guidelines are flagged by the computer system and transfusion service staff are electronically alerted to review flagged orders with the ordering clinician. In another example from transfusion medicine, Rana et al., demonstrated a significantly decreased rate of inappropriate transfusion upon integration of a transfusion algorithm into CPOE. In a cardiac intensive care unit setting, Wang et al., found that integration of practice guidelines into standard admissions order templates significantly decreased the use of laboratory tests without compromising care.
In several studies, display of cost information during order entry has been demonstrated to influence utilization. In a randomized controlled trial in an outpatient primary care setting, display of laboratory test cost on the CPOE ordering screen led to a 14% reduction in the number of tests ordered. Bates et al., studied the effects of displaying cost information at a teaching hospital and noted a non-statistically significant 4.5% reduction in ordered tests. In the US, there are financial incentives for reducing inpatient test utilization due to the costs of inpatient care being bundled into a single payment using the diagnosis-related group (DRG) payment schema. However, these incentives may be challenging to translate into reduced utilization by simply displaying costs, as providers may not be vested in cost reduction activities.
With several hundred to thousands of tests on a typical laboratory menu, search functionality is found in most laboratory CPOE implementations. With search, clinicians enter search terms and the system returns a list of corresponding tests. Clinicians then select the specific test they wish to order from the returned list. Search terms can include test names, corresponding synonyms and even corresponding disease states.
For example, a search for “celiac disease” might return “anti-transglutaminase IgA” and “endomysial IgA antibody,” two tests that may be ordered in the workup of celiac disease. By not requiring the clinician to know the exact names of the tests involved, the search function is made considerably more user-friendly. A user-friendly search function may help prevent free text ordering by providers. Free text orders cannot be electronically interfaced and must be manually translated into specific test names or codes, leading to inefficiencies, errors, and ambiguity. Monitoring free text orders may enable the CPOE team to improve the search process or contact ordering clinicians who routinely utilize free text inappropriately.
After a specimen has been processed and tested, clinicians may wish to order subsequent tests on the unused portion of the sample remaining in the laboratory. These tests are known as add-on tests and the management of add-ons often requires significant laboratory resources. Most CPOE systems do not support add-on testing. Thus, clinicians wishing to order add-on tests must call the laboratory with a verbal order, and laboratory staff must manually retrieve the sample and enter the orders for additional testing into the LIS. Such systems are often inefficient given the multiple manual steps involved, and the order may lack appropriate documentation. Extending CPOE systems to handle add-on testing can be logistically challenging as the add-on test process does not follow the same steps as a new test order.