A Hospitalist‐Run Observation Service Safely Decreases Length of Stay for Low‐Risk Chest Pain Patients

1University of Colorado‐Denver, Denver, CO
2University of Colorado–Denver, Denver, CO
3University of Colorado–Denver, Denver, CO
4University of Colorado–Denver, Denver, CO

Meeting: Hospital Medicine 2009, May 14-17, Chicago, Ill.

Abstract number: 62


Our tertiary‐care academic hospital allocates low‐risk chest pain admissions without evidence of acute coronary syndrome (ACS) to either a nonteaching observation service staffed by hospitalists and nurse‐practitioners or a teaching cardiology service staffed by cardiology attendings, fellows, and house staff,


A retrospective cohort chart review was performed on chest pain admissions over a 3‐month period from July through September 2007. Modified Thrombolysis in Myocardial Infarction (mTIMI) scores were retrospectively calculated for all patients due to rare documentation of actual admission mTIMI scores in the charts. The mTIMI eliminates cardiac risk factors, aspirin use, and multiple chest pain episodes from the conventional TIMI score without losing diagnostic accuracy in patients presenting to the emergency department with undifferentiated chest pain. Low‐risk chest pain was defined as mTIMI ≤ 1 without ACS. Length of stay (LOS) was calculated from calendar dates of admission and discharge.


A total of 241 patients were admitted with chest pain during the study period. One hundred and seven of 108 observation patients (99%) and 91 of 133 cardiology admissions (68%) had an mTIMI ≤ 1. No observation and 8 cardiology patients were excluded because of an abnormal ECG on admission. Fifty‐eight of 107 observation patients (54%) and 28 of 68 cardiology patients (41%) underwent noninvasive stress testing (P = 0.12), and an additional 15 cardiology patients had a cardiac catheterization. Five observation patients had a positive stress test or a positive follow‐up troponin and were transferred to the cardiology service without adverse outcomes due to transfer of care. The length‐of‐stay analysis excluded transfers to cardiology and primary cardiac catheterizations, leaving 170 patients. The overall LOS was significantly shorter on the observation service than on the cardiology service (1.0 ± 0.8 vs. 1.5 ± 1.5 days, P = 0.03; Figure 1). Noninvasive stress testing doubled the LOS in both groups — from 0.7 ± 0.6 to 1.3 ± 0.8 days (P = 0.0001) on the observation service and from 1.1 ± 1.4 to 2.0 ± 1.4 davs (P = 0.006) on the cardiology service. Patients undergoing noninvasive stress testing on the observation service had a significantly shorter LOS than patients on the cardiology service (P = 0.01).

Figure 1. Length of stay on observation and cardiology services for patients with mTIMI ≤ 1 (*P < 0.05).


The hospitalist‐run observation service performed a similar amount of noninvasive stress testing with a significantly reduced LOS for low‐risk chest pain admissions (mTIMI ≤ 1) compared to the cardiology service. Inpatient throughput may be enhanced by maximizing the census of the observation service, minimizing unnecessary stress testing, and optimizing the stress testing process.

Author Disclosure:

D. Levin, none; A. Trosterman, none; J. Moloo, none; J. J. Glasheen, none.

To cite this abstract:

Levin D, Trosterman A, Moloo J, Glasheen J. A Hospitalist‐Run Observation Service Safely Decreases Length of Stay for Low‐Risk Chest Pain Patients. Abstract published at Hospital Medicine 2009, May 14-17, Chicago, Ill. Abstract 62. Journal of Hospital Medicine. 2009; 4 (suppl 1). https://www.shmabstracts.com/abstract/a-hospitalistrun-observation-service-safely-decreases-length-of-stay-for-lowrisk-chest-pain-patients/. Accessed May 26, 2019.

« Back to Hospital Medicine 2009, May 14-17, Chicago, Ill.