Table 1.

The laboratory test utilization management toolbox.

Strength	Tool	Target	Strengths	Weaknesses	Example/References
Strong	Ban the test	Obsolete tests, “Quack” testing, Legitimate tests used in inappropriate circumstances	This is the “Nuclear Option”, as it ensures a complete cease to ordering	Only useful for tests with broad consensus as to lack of utility, which is unusual. Specific individuals may destroy consensus.	Bleeding time and other “Antiquated” tests (42).
Strong	Laboratory test formulary	All tests, especially those with utilization that is recognized, after analytics, to be above what is expected or justifiable.	A uniform policy across a system can be supported by a formulary, in the same way as a pharmacy formulary. Exceptions to formulary can be vetted by a committee or individual tasked with these decisions.	Requires authority and buy-in from multiple factions in a medical system, and likely participation by multiple specialties.	University of Michigan (43).
Strong	Combined intervention	Any test	By far the most effective, as the strengths of one intervention often complement the weaknesses of another.	Logistically complex, as many parties (the laboratory, clinicians, information services, payer systems, etc…) need to be involved.	Solomon meta-analysis (44), Massachusetts General Hospital Experience (45), hematology testing (46).
Strong	Stop paying for unnecessary testing	Any test	Similar to banning tests, this intervention is effective at nearly ceasing testing, depending on who decides to stop paying.	Depends on the payment system present in the medical system. Perceived as unfair, especially if the payer decides to stop paying for something without adequate justification. A physician may not and the cost could be transferred to know that a test will not be paid for, the patient.	Trends in reimbursement shown here (47), example of medical policy here (40).
Strong	Ban repetitive orders	Daily inpatient tests	Powerful method of reducing automatic ordering that providers often do not even know is occurring.	Worry amongst some physicians that they might “miss something”. Actual risk of missing something clinically important if a clinically indicated repetitive test is disallowed (i.e. coagulation tests in patients on anticoagulants).	Make repeated orders difficult through computerized order entry (48), ban standing orders (49), limit tests to 24-hour period (50).
Strong	Privilege ordering providers	Complex single tests, high unit cost and/or difficult to interpret.	Limiting testing to physicians who know how to use a test increases the prior probability in the tested patients, increasing cost effectiveness and diagnostic yield.	Multiple physicians may want privileges, even in the absence of evidence that they deserve them.	Neurogenetic testing diagnostic yield ∼30% for very rare diseases when expert providers order tests (51).
Strong	Require high level approval	Complex single tests, high unit cost and/or difficult to interpret.	Laboratory providers can have more insight into the utility of some tests than generalist providers.	Time consuming for laboratory staff or director, especially if there are no laboratory housestaff to take calls.	Large Genetics Sendout Testing Intervention (52).
Strong	Change computerized order entry options	Any test in a system with computerized ordering.	Computerized order changes can be made far more difficult to subvert than paper order form changes.	In the absence of a cultural change supporting modification of ordering practices, a complete stop to a specific order may increase provider abrasion. Unintended consequences can result if one is not careful in designing the intervention.	Reducing testing in coronary care unit (53). Change to routine testing menu (54).
Strong	Offer reflexive testing .	Any test where a cheaper screening test can be used before a more costly test.	Can work for computerized or paper ordering. Is a form of decision support that allows physician to follow correct testing algorithm with one order or click. Increases pre-test probability for more costly tests, making them more interpretable.	Requires an analyte for which a cheaper screening test exists. If using paper forms, one must realize that paper forms have a significant half-life in medical systems, and forms usually allow providers to “write in” tests that they cannot find on the form, thus allowing clinicians to subvert the intent of the reflexive panel.	Reflexive ionized calcium (32), coagulation panels (55).
Moderate/Strong	Utilization report cards	Routine outpatient panel testing, daily testing on inpatients.	Provides data on ordering to providers who may otherwise have no idea how they order tests, and thus may allow them to make informed decisions. reimbursement/financial Can be paired with penalties, or associated with peer feedback for added strength.	No one has to read the report card, especially if it is not associated with an incentive.	Outpatient report cards (56), intermittent feedback for physicians (57,58).
Moderate	Computerized reminders/decision support	Selected tests with moderate volume and high likelihood of being misordered.	Can provide support in real time to physicians to increase prior probabilities.	“Pop-up fatigue” occurs if too many reminders are implemented, leading to provider abrasion. Providers will also cease to continue to read pop-ups after some time.	Magnesium intervention (41), 1,25 dihydroxy Vitamin D email reminder (59) [cited example also uses privileging].
Weak	Post guidelines on paper order forms	Selected tests with moderate volume and high likelihood of being misordered, but no computerized ordering available.	Can provide support in real time to physicians to increase prior probabilities.	As opposed to pop-ups on computerized forms, written guidelines on a paper are likely easier to ignore.	Redesigning test requisitions and promulgation of factsheets (46,60).
Weak	Education alone/call for enhanced vigilance	Any test	Required as a component of nearly all successful utilization management efforts. Interventions lacking an educational component risk failure due to lack of buy-in from interested parties who do not understand the purpose of the change.	Almost never works alone, or when it does, the effect wears off over time or completely disappears if new staff takes over (i.e. in a teaching hospital).	Example showing effect wearing off after time (38), mixed effects of remindingphysicians of test costs (36–38).