Background:

Laboratory testing accounts for a significant percentage of the direct costs of hospitalization. Testing also imposes burdens on the patient due to phlebotomy, which not only causes pain, leads to anemia, and may necessitate placement of indwelling catheters after prolonged hospitalization. Efforts at limiting laboratory testing reduces the direct costs of hospitalization, decreases the workload of laboratory services and mitigates the burdens of phlebotomy.

Methods:

To better understand laboratory testing patterns among internal medicine residents on an inpatient general medicine ward, a retrospective review of all laboratory tests ordered during a 7–day period was performed. A pilot intervention of restricting test ordering to only a 24–h period was, thereafter, instituted for 2 weeks after an initial 2 week period of observation. A retrospective chart review was then performed. Residents were then surveyed about attitudes regarding laboratory test ordering and education by attending physicians.

Results:

During the initial observation period, internal medicine residents were noted to be ordering laboratory tests in a recurring fashion via computerized physician order entry (CPOE) for 3 up to 10 expected, i.e. future days of hospitalization. Most frequently ordered were daily basic metabolic panel (BMP), complete blood count (CBC), liver function tests (LFT), coagulation panel and type and screen. Despite stable results, tests continued to be performed. Ordering of certain lab studies also did not follow accepted diagnostic algorithms, particularly with the anemia work–up, thyroid function testing and autoimme work–up. During the two weeks prior to the intervention period, a culture of recurring laboratory testing via CPOE was similarly observed. Upon restriction of CPOE to only the next 24 h and encouragement to order only tests judged to be necessary, a significant decrease of two lab tests ordered per patient–day was observed, notably with significant decreases in the number of LFT’s, coagulation profiles and types and screens. Additionally, the number of patient–days when NO lab tests were ordered significantly increased. However, repeated testing of TSH, anemia work–up, etc. continued to occur despite availability of prior recent results. A survey of residents showed that discussion of laboratory test ordering with attending physicians rarely occurred and infrequent feedback regarding testing patterns was provided.

Conclusions:

Increased cost of hospitalization results from repeated, potentially unnecessary laboratory testing in the inpatient setting. Although recurring laboratory test ordering via CPOE may ease a resident’s daily burden of work, it results in excessive testing. Curbing recurring testing represents an area of intervention to reduce cost and decrease inherent burdens of phlebotomy. Additionally, this study revealed the need for greater resident education and feedback on laboratory test utilization.