a b s t r a c t

Study objective: Determine whether an expanded emergency medicine (EM) pharmacist Scope of practice re- duces the frequency of major delays in subsequent antibiotic administration in patients boarded in the emer- gency department (ED).

Methods: A pre-post, quasi-experimental study conducted from November 2019-March 2020 at a single-center tertiary academic medical center following the implementation of an expanded EM pharmacist scope of practice. Adult patients were included if they received an initial antibiotic dose in the ED and deemed to be high-risk. Sub- sequent antibiotic doses were reordered by EM pharmacists for up to 24-h after the initial order pending ED length of stay . The historical control group consisted of retrospective chart review of cases from the previ- ous year.

Results: The study identified that of the 181 participants enrolled, major delays in subsequent antibiotic admin- istration occurred in 13% of the intervention group and 48% of the control group (p < 0.01). When compared to the control group, the intervention group had a significant decrease in the number of delays among antibiotics dosed at 6-h (39% vs 13%) and 8-h (60% vs 8%) intervals. For antibiotics dosed at 12-h intervals, no statistically significant difference was observed between the control and intervention groups respectively (19% vs 5%). A sta- tistically significant lower incidence of in-hospital mortality was observed in the intervention group (3% vs 11%, p

= 0.02). In the intervention group, 97% of patients received subsequent antibiotic doses while boarded in the ED, compared to 65% in the control group (<0.01).

Conclusion: Expanding EM pharmacist scope of practice was associated with a significant reduction in the fre- quency of major delays in subsequent antibiotic administration as well as a decreased incidence of hospital mortality.

(C) 2021

1. Introduction

Emergency Department (ED) boarding is a growing public health and patient safety concern [1,2]. ED boarding occurs when an admitted patient is kept in the ED until an inpatient bed becomes available. ED crowding is often the result of input, throughput, and output factors [3]. Input factors are common issues that affect the influx of patients presenting to the ED, such as frequent-flyer patients, non-urgent visits,

* Corresponding author.

E-mail addresses: [email protected] (M. Payne-Cardona), [email protected] (V.A. San Luis), [email protected] (R. Aazami), [email protected] (M. Dermendjieva), [email protected] (M. Erin), [email protected] (J. Kirkwood), [email protected] (C. Tong), [email protected] (G. Marks), [email protected] (E.A. Smith), [email protected] (S.S. Torbati), [email protected] (J.F. Gilmore).

and influenza season. Throughput is reflected in the ED length of stay . Inadequate staffing and inefficient flow processes are common themes that decrease ED throughput. Hospital bed shortages, unit clo- sures, and inPatient boarding are output factors that also contribute to ED overcrowding [3]. Many adverse effects, such as increased hospital mortality, increased hospital LOS, and Medication errors and delays have been associated with increased ED boarding times [3-10]. Profes- sional organizations have also recognized the burden and peril ED boarding exposes to the health system. In 2017, the American College of Emergency Physicians (ACEP) released a policy statement acknowl- edging that boarding admitted patients in the ED contributes to lower quality of care, decreased patient safety, reduced timeliness of care, and reduced patient satisfaction [11]. Similarly, the Institute of Medicine suggests that ED boarding leads to poor patient outcomes, such as hav- ing an increased risk of Medical errors, delays in treatment, and overall decreased quality of care [12].

https://doi.org/10.1016/j.ajem.2021.07.039

0735-6757/(C) 2021

vulnerable patients, such as the critically ill, remain at risk for ad- verse outcomes associated with increased rates of medication delays and other time-sensitive therapies [10,13,14]. For instance, the impor- tance of early antimicrobial therapy administration is well known. The Surviving Sepsis Campaign recommends that intravenous (IV) antibi- otics be given within one-hour of the diagnosis of sepsis [15]. Although it is well established that prompt initiation of antibiotic therapy is asso- ciated with improved patient outcomes in sepsis, limited data exists on the impact of delayed subsequent antibiotic doses. A 2017 retrospective study evaluated the frequency of delays in second dose antibiotic ad- ministration among septic patients in the ED. [16] Of the 828 sepsis cases that were analyzed, 33% were found to have major delays in sec- ond dose antibiotic administration; defined as a first-to-second dose delay >=25% of the recommended dosing interval. The majority of delays were observed in antibiotics with 6-h dosing intervals (72%). Lastly, the authors observed major delays were associated with increased hospital mortality, hospital LOS, and mechanical ventilation requirement. Al- though ED boarding coupled with antibiotic delays may lead to an in- creased risk of Negative outcomes, a solution to combat delays has not been outlined.

At our institution, EM pharmacists have participated in the manage- ment of Medical emergencies to ensure safe and effective medication use since 2001. Clinical EM pharmacy services include participation in trauma resuscitations, code strokes, acute myocardial infarction activa- tions, and both cardiac and respiratory arrests. Additionally, EM phar- macists play an active role in implementing Medication Safety initiatives, answering drug information questions, and providing department-wide education. Pharmacy services also provide ED patient medication history review by pharmacists, transitions-of-care phar- macy technicians, pharmacy residents, and pharmacy students. During the time of the study, medication history pharmacy services were avail- able Monday through Friday from 0900 to 2330 pm. EM pharmacists also routinely review orders for boarded patients to ensure proper con- tinuity of care. Historically, one-time antibiotic orders entered by an ED physician would not be redosed until patients were transferred to the admitting units. However, subsequent doses were sometimes delayed or missed altogether due to the demands of a high census and a busy ED. EM pharmacist workflow for recommendation of subsequent anti- biotic doses include review of last administration, renal function, cul- tures, and provider documentation for assessment and plan. After reviewing appropriateness, the EM pharmacist would inform the ED provider that the subsequent dose was due and requested for the order to be placed. Subsequent antibiotic doses were sometimes de- layed or missed altogether if more critical tasks warranted the attention of the pharmacist or ED provider. An expandED pharmacist scope of practice was developed to allow EM pharmacists to schedule subse- quent antibiotics in selected high-risk patients to prevent major delays and to ensure continuity of care despite the unpredictable demands of the ED.

The primary objective of this study was to determine if expanding

EM pharmacist scope of practice to order subsequent antibiotic doses reduced the frequency of major delays. Major delay was defined as a first-to-second dose time greater than or equal to 25% of the recom- mended dosing interval. Secondary endpoints evaluated differences among antibiotic dosing intervals and the impact of major delays on pa- tient outcomes associated with hospital mortality, hospital LOS, inten- sive care unit (ICU) admissions, vasopressor usage, and mechanical ventilation requirements.

1. Methods
   1. Study design

This study was conducted at an 886-bed urban, academic hospital and level 1 trauma center that receives approximately 90,000 annual ED visits. This pre-post, quasi-experimental study targeted high-risk

adult patients that presented to the ED and were started on broad- spectrum antibiotic(s). Starting November 2019, the study site’s Phar- macy and Therapeutics Committee and the Antimicrobial stewardship Committee granted the expansion of EM pharmacist scope of practice to allow for subsequent antibiotic doses to be ordered by pharmacists for up to 24-h after the initial order for patients boarded in the ED.

• 1. Study population

Eligible patients for the intervention group were identified at the ini- tiation of antibiotics in the ED for adult patients (>=18 years old) and at a least one high-risk criteria: diagnosis of sepsis, diagnosis of pneumonia, temperature < 36?C or > 38?C, white blood cell count <4000/mm³ or > 12,000/mm³, systolic blood pressure < 100 mmHg, heart rate > 90 beats/min, respiratory rate > 20 breaths/min, lactate >2 mmol/L, or an abnormal chest x-ray. High-risk criteria characteristics were deter- mined a priori by a multidisciplinary workgroup that included physi- cians and pharmacists from both emergency medicine and infectious disease disciplines. Additionally, only antibiotics on 6-h, 8-h, and 12-h dosing intervals were included in this study to target intervals that have been associated with an increased risk of major delays. Patients were excluded if they did not receive the initial antibiotic(s) in the ED, antibiotic(s) selected were discontinued after the initial dose, patient expired prior to antibiotic redosing, patient refused care, or if the patient was admitted directly into hospice/palliative care. Antibiotic orders that were excluded from the protocol followed pre-intervention workflow, and the EM pharmacist would remind EM physicians of subsequent dose due times.

• 1. EM pharmacist scope of practice workflow

At the time of the study, the ED was staffed 20-h per day by EM phar- macists. Study efforts of patient enrollment and expanded pharmacist scope were only available during EM pharmacist coverage times. Pa- tients who met inclusion criteria for subsequent dose(s) during 4-h gap had antibiotics ordered and scheduled by EM pharmacist prior to end of shift.

Boarded ED patients were identified by Hospital overcrowding indi- cators: high hospital and ED census, increasED wait times, and increased overall time in the department. The expanded EM pharmacist scope targeted ED boarded patients that met inclusion criteria for subsequent dose(s) while waiting to be transferred to anInpatient unit. Subsequent antibiotic doses were only ordered if the patient was still in the ED at the time the following dose was needed. The pharmacist scope allowed for more than one subsequent dose to be ordered if appropriate, especially for more frequent antibiotic dosing intervals. (See Fig. 1.)

Antimicrobial stewardship initiatives at this institution require each provider to select an indication for each antibiotic ordered before it can be verified by a pharmacist. Each first dose antibiotic order was placed as a one-time dose in the ED. Upon verification of initial antibiotic, EM pharmacists reviewed lab results, prior cultures, vital signs, and prior provider documentation if available. Furthermore, pharmacists would recommend initial dose adjustments and alternative antibiotic selection if indicated. Renal function, pertinent labs, and diagnostic imaging were periodically not available at time of ordering initial antibiotics for high risk patients. Thus, to not delay administration of first dose antibiotics, subsequent antibiotics were not ordered during verification of the initial dose. EM pharmacists performed intermittent chart review to acquire additional information, such as updated lab results, vital signs, and pro- vider notes.

EM pharmacists reordered doses based on the initial physician- entered antibiotic, indication, and dosing-interval per hospital protocol. Subsequent doses were re-entered individually. The suggested intervals between doses were outlined in the institutional antimicrobial dosing policy, which take into consideration renal function, weight, and antibi- otic indication. Upon each antibiotic reordered, the pharmacists notified

Fig. 1. Flowchart of EM pharmacist scope of practice

the EM physicians either in person or over the institution-approved chat system to serve as a safety check that antibiotics were appropriate to continue. If provider plan documentation was not available when sub- sequent dose was due, EM pharmacists contacted EM provider to dis- cuss if appropriate to continue subsequent antibiotics. Additionally, EM pharmacists documented first dose administration time, recom- mended dosing interval, and if subsequent dose(s) were ordered per scope of practice in a pharmacist to pharmacist communication tool in the electronic medical record (EMR).

• 1. Data collection

Pre-intervention analysis consisted of retrospective chart review from data collected via an EMRreport that identified ED patients from November 2018 to March 2019 that met at least one high-risk criteria. The project lead reviewed each chart for accuracy and manually col- lected time stamps for antibiotic dose administrations in the ED, includ- ing the antibiotic frequency interval, and location of subsequent dose administration. Prospective intervention group data collection occurred between November 2019 to March 2020. Upon each antibiotic re- ordered under the EM pharmacy protocol, the patient profiles were manually stored in an EMR shared database for project lead to review. Data collection was done via a pre-approved and piloted standardized form. The primary outcome measure was to determine if antibiotics re-ordered under the expanded EM pharmacist scope of practice re- duced the frequency of major delays in second dose antibiotic adminis- trations. Secondary outcomes evaluated differences between individual antibiotic dosing intervals, location subsequent antibiotic dose was given, the impact of major delays on patient outcomes associated with

hospital mortality (due to infectious origin), hospital LOS, ICU admis- sions, vasopressor usage, and mechanical ventilation requirements.

• 1. Statistical analysis

Continuous variables were analyzed with the Mann-Whitney U Test and reported as medians (interquartile ranges [IQR]) to determine sta- tistical significance. Categorical variables were analyzed using a Chi- Square Test and reported as frequencies (percentages). For all data, a p-value of <0.05 was used to determine statistical significance.

• 1. Human subjects compliance

This study was performed as a quality improvement initiative and was exempt from institutional board review.

1. Results
   1. Characteristics of study subjects

A total of 181 patients were included in this study. Key clinical char- acteristics of ED Boarding patients who met inclusion criteria are shown in Table 1. No statistical differences were identified between interven- tion or control group demographics. The median age was 71 years in the intervention group and 69 years in the historical control group. Male patients consisted of 60% of the entire study population. Among patients in both arms, sepsis was diagnosed in 48% of patients, 24% were diagnosed with pneumonia, and 28% of diagnoses were classified as another infectious disease which included intra-abdominal infec- tions, urinary tract infections, cellulitis, and osteomyelitis. Regardless

Table 1

Baseline characteristics

Variable	Intervention	Control	P value
	(n = 117)	(n = 64)
Age, median - yr.	71	69	0.79
Interquartile range	54-82	56-83	-
Male sex - n (%)	67 (57)	32 (50)	0.35
Acute kidney injury - n (%)	22 (19)	11 (17)	0.79
Dialysis PTA - n (%)	4 (3)	3 (5)	0.67
Diagnosis - n (%)
Sepsis	54 (47)	32 (50)	0.62
Pneumonia	32 (27)	11 (17)	0.12
Other?	31 (26)	21 (33)	0.37
High-risk criteria met - n (%)	117 (100)	64 (100)	1
Systolic blood pressure <= 100 mmHg	15 (13)	10 (16)	0.6
Heart rate > 90 beats/min	74 (63)	40 (63)	0.92
Respiratory rate >= 22 breaths/min	35 (30)	15 (23)	0.35
Temperature < 36?C or > 38?C	28 (24)	11 (17)	0.29
Lactate >2 mmol/L	25 (21)	17 (27)	0.43
White blood cell <4 k/mL or > 12 k/mL	73 (62)	36 (56)	0.42
Altered mental status	16 (14)	9 (14)	0.94
Abnormal chest x-ray	48 (41)	16 (25)	0.05
ED LOS, median - hr.	16.8	10.5	<0.05

* Intra-abdominal, urinary tract, cellulitis, osteomyelitis.

of the suspected infectious etiology, all the patients included in both the intervention and control groups had the presence of at least one high- risk criteria. Table 1 displays the prevalence of each individual septic criteria at initial presentation to the ED.

When assessing ED LOS between groups, the intervention group had a significantly longer ED boarding time (median LOS 16.8 h), compared to the pre-intervention group (median LOS 10.5 h, p < 0.05). The num- ber of patients that received second dose antibiotics in the ED, also sig- nificantly varied. In the intervention group, EM pharmacists were able to use the expanded scope of practice on 117 boarded patients, of which had 140 second dose antibiotics administered during their ED LOS. The historical control group had 64 ED boarded patients, which consisted of 86 antibiotics being administered pre-intervention. The breakdown of antibiotics based on the interval between the first-to-sec- ond antibiotic doses are displayed in Table 2. Antibiotic intervals were determined by the pre-established institutional dosing guideline, there- fore the intervals determined were patient specific.

• 1. Frequency and magnitude of delays

Of second dose antibiotics administered in ED boarding patients, major delays occurred in 13% of the intervention group and 48% of the control group (p < 0.01) (Table 3). Evaluation of all delays within the study concluded that the median length of time from the first to second dose antibiotic administration for the intervention and historical control group were as follows: 6-h interval antibiotics (6.2 h vs 7.1 h; Fig. 2), 8-h interval antibiotics (8.6 h vs 10.6 h; Fig. 3), and 12-h interval antibiotics (12.1 vs 12.3 h; Fig. 4). A subset analysis determined that among the major delays, the intervention group displayed a significant decrease in the amount of 6-h (13% vs 39%, p < 0.01) and 8-h (8% vs 60%, p < 0.01) interval antibiotic delays, compared to the historical control group (Table 3).

Table 2

Amount of second dose antibiotics ordered in ED boarding patients

Variable	Intervention (n = 117)	Control (n = 64)
Antibiotics administered - n (%)	140	86
6-h interval	82 (58)	33 (39)
8-h interval	36 (25)	20 (24)
12-h interval	22 (16)	31 (36)

Table 3

Primary and secondary endpoints

Primary endpoint	Intervention	Control	P value
	(n = 117)	(n = 64)
Major delays - n (%)	15 (13)	31 (48)	<0.01
Secondary endpoints
Frequency of delays - n (%)
6-h interval	11/82 (13)	13/33 (39)	<0.01
8-h interval	3/36 (8)	12/20 (60)	<0.01
12-h interval	1/22 (5)	6/31 (19)	0.12
Subsequent dose given in ED? - n (%)	101/104 (97)	20/31 (65)	<0.01
Hospital LOS, median - hr.	139.5	131.7	0.69
Interquartile range	84-230	74-228	-
Total ICU admits - n (%)	14 (12)	10 (16)	0.49
Direct ICU admits - n (%)	8 (7)	4 (6)	0.88
ICU-free time, median - hr.	125.3	124.5	0.56
Interquartile range	81-215	70-215	-
Vasopressor usage
Vasopressor requirement - n (%)	8 (6)	4 (6)	0.88
Mechanical ventilation usage

Ventilation requirement - n (%)	6 (5)	4 (6)	0.75
Ventilator-free time, median - days	5.7	5.3	0.46
Interquartile range	4-10	3-9	-
In-hospital mortalitya - n (%)	3 (3)	7 (11)	0.02

* Subsequent antibiotic dose due and given during ED admission.

^a Death attributed to an infectious cause.

Magnitude of each major delay was also assessed. For 6-h dosing in- terval antibiotics, median first-to-second dose time was 8.3 h and 8.7 h in the intervention and control groups respectively. For 8-h dosing in- terval antibiotics, median first-to-second dose time for major delays was 11.4 h in the intervention group and 12.9 h in the control group. Lastly, for the 12-h dosing interval antibiotics, median first-to-second dose time was 15.5 h for major delays identified in the intervention group and 17.9 h in the historical control group. Location of where pa- tients received the second antibiotic dose was also assessed. The inter- vention group had 97% of patients receive the second antibiotic dose while in the ED, while the historical control group had 65% of second dose antibiotics administered in the ED (<0.01).

• 1. Impact of major delays on patient outcomes

Table 3 displays secondary outcomes among major delays identified. A significantly lower incidence of in-hospital mortality was observed in the intervention group (3% vs 11%, p = 0.02). In-hospital mortality was included if it was deemed to be of infectious origin, which was specified by the death summary in the EMR. No differences were identified in hospital LOS, total ICU admits, direct ICU admits, ICU-free time, vaso- pressor usage or mechanical ventilation requirement.

1. Discussion

Expanding EM pharmacist scope of practice at this institution led to a statistically significant reduction in the frequency of major de- lays in second dose antibiotic administration in ED boarded patients. Although not statistically significant, the magnitude of delays also decreased compared to the control group. This decrease was likely attributed to second antibiotic doses being administered prior to unit transfer. Unit transfer can be time consuming due to various workflow processes. This includes the need for the primary team to reconcile Medication orders, which subsequently need to be re- verified by a pharmacist before the nurse can administer. The signif- icant increase in second dose antibiotics administered in the ED prior to unit transfer supports how the expanded pharmacy scope allowed for more timely antibiotic administration. Subsequent antibiotic or- ders in the control group may have been delayed due to awaiting the patient to be transferred to the admitting unit. Dispensing loca- tion could also potentially contribute to delays of the second

Fig. 2. Magnitude of delays for 6-h dosing interval antibiotics.

Fig. 3. Magnitude of delays for 8-h dosing interval antibiotics.

antibiotic dose. For example, Time to administration may be short- ened if an antibiotic is available as floor stock compared to an antibi- otic that must be dispensed from the centralized pharmacy.

Since EM pharmacists at the study site had previously been involved in monitoring subsequent antibiotic doses in boarded patients and regu- larly assisted providers in scheduling these orders, their pre-intervention

involvement may have diluted the true impact of the expanded phar- macy scope. Additionally, a significant increase in ED boarding time in the intervention group was observed. It can be contended that since the intervention group faced longer ED Boarding times, there was greater risk for the patients to have delayed second antibiotic doses. However, this was not observed. The patients in the intervention group still had

Fig. 4. Magnitude of delays for 12-h dosing interval antibiotics.

improved times to subsequent dose antibiotic compared to the control group.

Leisman et al. observed an association between major second dose delay and increased mortality [16]. The authors explain that delayed second dose antibiotics may be a surrogate for patients who receive less attention and care overall. Our study demonstrates that expanding an EM pharmacist scope of practice to redose subsequent antibiotics may decrease hospital mortality. This observation suggests that patients treated under EM pharmacist scope of practice are receiving continuity of care more similar to an inpatient unit with routine scheduling of an- tibiotics. Continuity of care in the ED, expressed as more attention and care overall to boarded patients may explain the decreased hospital mortality association.

When comparing the results obtained in this study to the 2017 ret- rospective cohort study mentioned previously [16], similar trends were seen regarding identifying more major delays with the 6-h inter- val antibiotics and observing an association between major delays and increased hospital mortality. In contrast, this study did not find a signif- icant difference in hospital LOS or mechanical ventilation requirements. The small sample size of this study may have affected the ability to de- tect a difference.

The value of clinical pharmacists in the ED is well documented. Tra- ditional services by EM pharmacists include: responding to medical emergencies, providing consultation on medication issues, identifying and reducing medication errors [17]. These services have anecdotally increased patient safety by facilitating appropriate selection, prepara- tion, and medication administration [18]. This study contributes to the robust literature demonstrating the value of EM pharmacists on patient care.

Lastly, there are additional benefits to a pharmacist scope of practice to redose antibiotics. For example, the workflow for ED staff was im- proved. ED physicians’ workload was alleviated, as antibiotics were con- tinued under protocol on boarded patients and thus could focus on more critical emergency medicine tasks or treatments. ED nurses were also able to prioritize other tasks with the knowledge that subsequent doses were scheduled appropriately. EM pharmacist workflow also im- proved. The scope of practice provided an additional tool for pharma- cists to manage the critical demands of an overcrowded ED while ensuring continuity of care for boarded patients. For example, schedul- ing antibiotics can be cumbersome and complex. It requires medication profile review for timing of already administered medications, renal and Hepatic function as applicable, cultures if available, and previous shift notes and documents. If a high census was known, the ED pharmacist could schedule subsequent antibiotics for future administrations. This alleviated the burden of patient workup for a large volume of patients at one time. Overall, ED boarding adds additional strain to the health care system. Although more data is needed to assess the impact on sub- sequent antibiotic delays, an EM pharmacist scope of practice can be an additional tool to help care for ED boarded patients and possibly im- prove workflow for ED staff.

1. Limitations

Several limitations must be acknowledged. 1) The small population size may have missed significant risk factors that are associated with the negative clinical impact of major delays. Since current evidence varies in terms of patient outcome data and the association between hospital mortality, hospital LOS, and mechanical ventilation, more re- search is needed. 2) Historically, EM pharmacists at the study site have been involved in actively monitoring subsequent antibiotic doses expected due times and assisting providers in scheduling these orders prior to the implementation of the expanded scope. 3) Significantly higher ED boarding times of the intervention group relative to the his- torical control group reflect an increase in overall census of the study site. This may have skewed the results as more patients in the interven- tion arm may have been due for a re-dose during their ED stay. 4) The

study could have unintentionally excluded patients that are transferred to the ICU immediately after they are stabilized in the ED. Majority of these patients do not have subsequent antibiotics started within the short ED LOS and are not typically boarded in the ED. This could explain the low amount of ICU admits captured in the study population. 5) EM pharmacists manually enrolled study participants which could have po- tential for selection bias. 6) The difference in size between the interven- tion (n = 117) and historical control groups (n = 64) is another limitation of this study. This difference may be because the control group only included ED boarding patients captured in an EMR- generated report for ICD-10 code admissions specified for sepsis or pneumonia while the intervention group had a wider inclusion criteria.

7) Delay of subsequent antibiotics if due to delay of order placement or delay of administration was not investigated but would be an interest- ing observation for future studies. 8) Lastly, this was a single center study and the workflow processes implemented at other emergency de- partments may vary for the management of boarded patients.

1. Conclusion

In summary, this study identified that implementation of an ex- panded EM pharmacist scope of practice was associated with a reduc- tion in frequency of major delays of second antibiotic doses, as well as a reduction in overall hospital mortality of septic patients boarded in the ED. Implementation of an EM pharmacist scope of practice for sub- sequent antibiotic dosing may be a best practice and may provide an ad- ditional tool to help care for boarded patients when hospitals experience overcrowding. However, more research is needed to explore solutions to mitigate the impact of ED boarding and the consequences of delayed subsequent antibiotic administration on clinical patient outcomes.

Author contribution

JFG conceived the study. MPC, VSL, RA, MD, ME, JK, CT, JFG, GM, and EAS designed the study. VSL, RA, MD, ME, JK, CT, and MPC enrolled pa- tients and implemented the expanded emergency medicine pharmacist scope practice. MPC assumed data collection, management and analysis. EAS, GM, VSL, RA, MD, ME, JK, CT, and JFG provided statistical advice on study design and analyzed data. MPC drafted the manuscript, and all au- thors contributed substantially to its revision.

VSL takes responsibility for the paper as a whole.

Declaration of Competing Interest

All authors have no conflicts of interest to disclose.

References

1. Chalfin DB, Trezeciak S, Likourezos A, Baumann BM, Dellinger RP, DELAY-ED STUDY Group. Impact of Delayed transfer of critically ill patients from the emergency de- partment to the intensive care unit. Crit Care Med. 2007;35(6):1477-83.
2. Trezciak S, Rivers EP. Emergency department overcrowding in the United States: an emerging threat to patient safety and public health. Emerg Med J. 2003;20:402-5.
3. Hoot NR, Aronksy D. Systemic review of emergency department crowding: causes, effects, and solutions. Ann Emerg Med. 2008;52(2):126-36.
4. Singer AJ, Thode Jr HC, Viccellio P, Pines JM. The association between length of emer- gency department boarding and mortality. Acad Emerg Med. 2011;18(12):1324-9.
5. Salehi L, Phalpher P, Valani R, et al. Emergency department boarding: a descriptive analysis and measurement of impact on outcomes. CJEM. 2018;20(6):929-37.
6. Boulain T, Malet A, Maitre O. Association between long boarding time in the emer- gency department and hospital mortality, a single-center propensity score-based analysis. Intern Emerg Med. 2020;15(3):479-89.
7. Kulstab EB, Sikka R, Sweiss RT, et al. ED overcrowding is associated with an increased frequency of medication errors. Am J Emerg Med. 2010;28(3):304-9.
8. Backhsh HT, Perona SJ, Sheilds WA, et al. Medication errors in psychiatric patients boarded in the emergency department. Int J Risk Saf Med. 2014;26(4):191-8.
9. Patanwala AE, Warholak TL, Sanders AB, Erstad BL. A prospective observational study of medication erroris in a tertiary care emergency department. Ann Emerg Med. 2010;55(6):522-6.
10. Liu SW, Chang Y, Weissman JS, et al. An empirical assessment of boarding and qual- ity of care: delays in care among chest pain, pneumonia, and cellulitis patients. Acad Emerg Med. 2011;18:1339-48.
11. American College of Emergency Physicians. Boarding of Admitted and Intensive Care Patients in the Emergency Department; 2017.
12. Institute of Medicine: Agency for Health Care Research and Quality. https://www. ahrq.gov/research/findings/final-reports/ptflow/section1.html; 2018.
13. Sri-On J, Chang Y, Curley DP, et al. Boarding is associated with higher rates of med- ication delays and adverse effects but fewer lab related delays. Am J Emerg Med. 2014;32(9):1033-6.
14. Gaieski DF, Agarwal AK, Mikkelsen ME, et al. The impact of ED crowding on early in- terventions and mortality in patients with severe sepsis. Am J Emerg Med. 2017;35 (7):953-60.
15. Rhodes A, Evans LE, Alhazzani W, Levy MM, Antonelli M, Ferrer R, et al. Surviving sepsis campaign: international guidelines for management of sepsis and septic shock: 2016. Intensive Care Med. 2017;43(3):304-77.
16. Leisman D, Huang V, Zhou Q, Gribben J, Bianculli A, Bernshteyn M, et al. Delayed sec- ond dose antibiotics for patients admitted from the emergency department with sepsis: prevalence, risk factors, and outcomes. Crit Care Med. 2017;45(6):956-65.
17. Szczesiul JM, Fairbanks RJ, Hildebrand JM, et al. Survey of physicians regarding clin- ical pharmacy services in academic emergency departments. Am J Health Syst Pharm. 2009;66(6):576-9.
18. Jacknin G, Nakamura T, Smally AJ, Ratzan RM. Using pharmacists to optimize patient outcomes and costs in the ED. Am J Emerg Med. 2014;32(6):673-7.