Delayed vs Early Admission to Intensive Care: A Retrospective Review of Outcomes and Clinical Indicators of Deterioration Following Trauma

Unplanned admission to an intensive care unit (ICU) is a trauma quality improvement indicator associated with increased morbidity, mortality, and economic burden on both the patient and hospital. ¹ Our study aimed to identify clinical factors associated with early and delayed ICU admission from a medical/surgical floor to reduce delayed ICU admission in the trauma population.

OPEN IN VIEWER

Table 1.

Clinical Factors Analyzed Between Early and Delayed ICU Admission.

	Delayed (>72 hrs), n = 48	Early (<72 hrs), n = 98	[DelAd/EAd] Odds Ratio (95% CI)	P-value
Mortality	18 (37.5%)	13 (13.3%)	3.923 (1.718-8.960)	<.001
Age, years	72.21 ± 15.17	65.31 ± 19.12		.043
BMI	25.03 ± 6.20	28.64 ± 8.06		.007
Obesity (BMI ≥ 35)	4 (8.3%)	14 (14.3%)		.424
Hospital LOS	14.83 ± 9.65	10.59 ± 6.47		.004
ISS	11.49 ± 5.72	12.85 ± 6.81		.301
Hematocrit on ED admission	35.69 ± 5.86	38.34 ± 6.04		.015
RR on ED admission	18.00 ± 2.49	18.82 ± 2.55		.021
Total time on floor (days)	6.47 ± 3.99	1.31 ± .82		<.001
Glucose 24 hrs prior to ICU (highest)	138.27 ± 48.61	171.14 ± 96.32		.01
Total MEQ in 24 hrs prior to ICU	17.13 ± 40.21	32.28 ± 81.76		.002
Prophylactic heparin on floor	13 (27.1%)	5 (5.1%)	6.909 (2.295-20.800)	<.001
Anticoagulation on floor	6 (12.5%)	1 (1.0%)	13.857 (1.618-118.693)	.005
Any form of anticoagulation prior to ICU	35 (72.9%)	41 (41.8%)	3.743 (1.764-7.944)	<.001
Gabapentin/Lyrica prior to ICU	16 (33.3%)	39 (39.8%)		.449
Benzodiazepine prior to ICU	10 (20.8%)	16 (16.3%)		.504
Abnormal EKG prior to ICU	40 (83.3%)	55 (56.1%)	3.909 (1.685-9.215)	.001
Surgery				.023
No surgery	28 (58.3%)	78 (79.6%)
Urgent surgery	17 (35.4%)	18 (18.4%)
Emergency surgery	3 (6.3%)	2 (2.0%)
Major abdominal surgery prior to ICU	4 (8.3)	1 (1.0%)	8.818 (.958-81.198)	.04
Orthopedic surgery prior to ICU	20 (41.7%)	22 (22.4%)	2.468 (1.172-5.196)	.016
Longest surgery (min) prior to ICU	69.75 ± 75.09	35.88 ± 69.92		.002
GCS day of ICU admission	12.36 ± 3.74	13.58 ± 2.42		.019
GCS decline from ED to ICU	28 (59.6%)	41 (41.8%)	2.049 (1.010-4.157)	.045
Chronic home aspirin use	9 (18.8%)	31 (31.6%)		.101
Chronic home anticoagulation use	14 (29.2%)	13 (13.3%)	2.692 (1.147-6.320)	.02
Chronic home antidepressant use	21 (43.8%)	24 (24.5%)	2.398 (1.152-4.991)	.018

An IRB-approved retrospective study was conducted at an ACS-verified level 1 trauma center from January 1, 2020, to March 31, 2023. Patients presenting following trauma and initially admitted to a medical/surgical floor before being transferred to the ICU were included. Patients were divided into two groups: early admission to ICU (EAd) vs delayed admission to ICU (DelAd), categorized as transfer to intensive care within or after 72 hours from initial presentation. Clinical variables including demographic information, injury type, ISS, vital signs, labs, tests, surgeries prior to ICU admission, medication administered in the emergency department (ED), home medications, chronic comorbidities, past medical history, and reason for ICU admission were compiled and analyzed for patients who were admitted to the ICU during their hospital stay. Continuous variables were reported as the mean ± standard deviation and were compared using independent-sample t test for normally distributed variables and Mann-Whitney U test for nonparametric variables. Statistical significance was defined as P < .05. Analysis was performed on SPSS version 28 (Armonk, NY: IBM Corp) (Table 1).

146 patients were admitted to ICU from a medical/surgical floor and met inclusion criteria—EAd consisted of 98 patients, while 48 were DelAd. No statistical difference was observed in injury mechanism, pattern, or ISS. EAd exhibited significantly lower in-hospital mortality of 13.3% vs 37.5% in DelAd patients (OR .255, 95% CI [.12-.58], P = .001). Patients in EAd were significantly younger (65.3 ± 19.2 vs 72.2 ± 15.17, P = .043). Upon admission to ICU, DelAd had a significant decrease in GCS between ED and ICU (−2.49 vs −1.03, P = .013). Patients in EAd were less likely to have been prescribed home anticoagulation (13.5% vs 29.2%, OR .371, 95% CI [.16-.87], P = .02) or antidepressant (25% vs 43.8%, OR .417, 95% CI [.2-.87], P = .018) therapy prior to admission. DelAd patients were more likely to have undergone orthopedic surgery prior to ICU admission (41.7% vs 22.3%, OR 2.468, 95% CI [1.172-5.196], P = .016). Lastly, EAd patients were given significantly higher amounts of opioids in the 24 hours preceding transfer to ICU, with an average morphine milligram equivalent dosage of 32.38 vs 17.13 in the delayed cohort (P = .002).

Unplanned admission to ICU is a metric that has gained international attention in recent years due to its significant associated mortality and enigmatic, multifactorial causes.^2,3 Many perspectives on the topic are to be considered. First, key outcome measures such as in-hospital mortality, length of stay, and health care resource usage are used to justify the impact of unplanned ICU admission on the outcome of the patient. ⁴ Building from this, analysis of inpatient indicators signaling necessitation for ICU admission has attempted to be identified but has not always been consistent. The topic of unplanned admission to the ICU has been studied across many specialties but holds a unique importance in the treatment of traumatically injured patients due to physician decision-making during the initial triage, stabilization, and admission process. To add to the literature regarding unplanned ICU admission and further describe the patient variables indicative of needing escalation of care, we performed analysis of trauma patients initially admitted to a medical/surgical floor and subsequently transferred to ICU either within or after 72 hours from initial hospital presentation. In doing so, we first observed findings consistent with the literature in terms of patient outcomes, with patients having delayed ICU admission exhibiting significantly higher in-hospital mortality and being significantly older, possibly representing the more indolent presentation of acute pathologies in elderly populations. Further, due to our access to more nuanced datapoints not available in large registries, an array of metrics with clinical significance not described in previous literature were observed to separate patients in the early and delayed admission cohorts. First, patients in the early admission cohort had significantly higher opioid burden in the 24-hour period preceding ICU transfer compared to the delayed cohort, possibly indicating that a more aggressive pursuit of adequate pain control necessitates earlier escalation of care. Interestingly, the significant drop in GCS observed in the delayed admission to ICU cohort may serve as a vital clinical indicator to providers for a patient’s need for transfer to ICU. Many more patients who experience unplanned ICU admission were prescribed at-home anticoagulation, similar to findings discussed in the CRASH study; however, we expand on this by our observation of a similar finding with antidepressant therapy. In addition, we observed that far more patients within the DelAd cohort had undergone orthopedic surgery prior to ICU admission when compared to EAd. While this study contributes observations on previously undescribed datapoints, limitations do exist. With relatively small sample size, this single-institution study is predisposed to type II error secondary to underpowering, leading us to believe that some datapoints exhibiting marginal insignificance may find significance in a larger cohort. In all, we add to the discussion regarding the clinical decision-making process for patient transfer to intensive care by describing clinical factors present in those potentially needing earlier transition to higher care. These findings indicate that home anticoagulant or antidepressant therapy, drop in GCS score, increasing opioid requirements, necessitation for orthopedic surgery, or abnormal EKG findings may warrant a more rapid escalation of care.