A judicious approach to ordering lab tests
Following these guidelines to order fewer tests can improve health care quality and patient experience, while reducing wasteful costs.
PRACTICE RECOMMENDATIONS
› Follow US Preventive Services Task Force and professional society recommendations for laboratory testing, including choice and frequency of tests. A
› Consider the pretest probability of your patient having a disease, and order the most sensitive and specific test to diagnose a new condition. Employ a 2-step approach with a second laboratory test when the first is outside the reference range. B
› Refrain from ordering routine preoperative testing for patients undergoing low-risk surgeries; these data do not improve postoperative outcomes, can lead to costly testing cascades, and may delay necessary surgical care for patients. A
Strength of recommendation (SOR)
A Good-quality patient-oriented evidence
B Inconsistent or limited-quality patient-oriented evidence
C Consensus, usual practice, opinion, disease-oriented evidence, case series
Narrowly out-of-range tests. Test results that fall just outside the “normal” range may be of questionable significance. When an asymptomatic patient has mildly elevated liver enzymes, should additional tests be ordered to avoid missing a treatable disorder? In these scenarios, a history, including possible contributing factors such as alcohol or substance misuse, must be paired with the clinical presentation to assess pre-test probability of a particular condition.14 Repeating a narrowly out-of-range test is an option in patients when follow-up is possible. Alternatively, you could pursue watchful waiting and monitor a minor abnormality over time while being vigilant for clinical changes. This whole-patient approach will guide the decision of whether to order additional testing.
Habitually paired tests. Reflexively ordering tests together often contributes to unnecessary testing. Examples of commonly paired tests are serum lipase with amylase, C-reactive protein (CRP) with erythrocyte sedimentation rate (ESR), and TSH with free T4 to monitor patients with treated hypothyroidism. These tests add minimal value together and can be decoupled.15-17 Evidence supports ordering serum lipase alone, CRP instead of ESR, and TSH alone for monitoring thyroid status.
Some commonly paired tests may not even be necessary for diagnosis. The well-established Rotterdam Criteria for diagnosing polycystic ovary syndrome specify clinical features and ovarian ultrasound for diagnosis.18 They do not require measurement of commonly ordered follicle-stimulating hormone and luteinizing hormone for diagnosis.
Serial rather than parallel testing, a “2-step approach,” is a strategy made easier with the advent of the electronic medical record (EMR) and computerized lab systems.8 These records and lab systems allow providers to order reflex tests, and to add on additional tests, if necessary, to an existing blood specimen.
Repetitive laboratory testing. Repetitive inpatient laboratory testing in patients who are clinically stable is wasteful and potentially harmful. Interventions involving physician education alone show mixed results, but combining education with clinician audit and feedback, along with EMR-enabled restrictive ordering, have resulted in significant and sustained reductions in repetitive laboratory testing.19
Continue to: Ongoing management of chronic conditions
