Background:
The Choosing Wisely Campaign encourages physicians to avoid over-utilizing routine labs in hospitalized patients with clinical and laboratory stability. A number of studies have shown adverse effects on hospitalized patients from lab overutilization. In previous research, we observed that internal medicine physicians at our center often order a daily complete blood count (CBC) and daily chemistry panel on admission and then do not discontinue these orders before discharge. This can lead to overutilization of these tests. In this study, we attempted to quantify the extent to which internal medicine residents and hospitalists at our center order redundant labs.
Methods:
This study was conducted at Northwestern Memorial Hospital (NMH), an 894-bed tertiary care teaching hospital in Chicago, and was approved by the Institutional Review Board (IRB). We used the Enterprise Data Warehouse (EDW) to quantify lab orders.
We looked at the extent to which hospitalized patients on a general medicine service with a length of stay (LOS) greater than or equal to four days had daily labs (CBC or basic chemistry panel) placed on admission and then not discontinued later in their admission despite repeated normal results. Moreover, we used broad exclusion criteria for CBCs and chemistry panels in order to exclude patients for whom daily labs despite lab stability may be justified (table 1). Patients who were in the ICU for any part of their hospitalization were also excluded.
Results:
Our quantification of redundant lab ordering varied significantly by type of routine lab (CBC vs. basic chemistry panel) and level of provider experience (hospitalist vs. resident) (table 2). However, the majority of patients who met inclusion criteria had a daily CBC and daily basic chemistry panel ordered within 24 hours of admission and also not discontinued before day 4. For patients with normal CBCs and basic chemistry panels for three consecutive days, the percentage of fourth day CBCs or basic chemistry panels with an abnormal result was below 20%. The percentage of hospitalist and medicine teaching service patients with a LOS greater than or equal to 4 days who had a CBC and/or chemistry panel drawn on the day of discharge, irrespective of whether prior labs were normal or abnormal, was 72.5% and 84.8%, respectively.
Conclusions:
Our study suggests that providers at our center have a high threshold to discontinue daily labs even in the face of stably normal labs. The ability to order daily recurring labs, though a matter of convenience, may play a role in lab overutilization. This may reflect the heuristic of default bias, and also an element of decision fatigue. It is certainly possible that clinical changes, which our study did not assess for, necessitated the additional labs.
Of note, our study does not evaluate the number of patients that continue to have daily labs despite stably abnormal labs, like stable anemia or CKD. Thus, it may underestimate the extent to which internal medicine physicians at our center overutilize routine labs. In summary our study suggests that the ability to order daily, recurring labs may lead to lab overuse, even in patients with a relatively long LOS and several days of normal results.
