a b s t r a c t

Background & purpose: Alteplase is the standard of care for early pharmacologic thrombolysis after acute ischemic stroke . Alteplase is also considered a high-alert medication and is fraught with potential for error. We sought to describe the difference in medication error rates in in patients receiving alteplase for acute ischemic stroke from Regional hospitals compared to patients receiving alteplase at the comprehensive stroke center.

Methods: This was a retrospective cohort comparison of patients who were greater than 18 years old that received intravenous alteplase for the treatment of AIS from June 2015 to June 2018. Several institution specific databases were utilized to obtain pertinent data. A standardized taxonomy was utilized to classify Medication errors. Patients were excluded if they received any fibrinolytic other than alteplase or if alteplase was used for a non-stroke indication. Two cohorts (from regional hospitals or the Comprehensive Stroke Center (CSC)) were compared.

Results: A total of 676 patients received alteplase during the study period (34% from the CSC and 66% from regional hospitals). There were 133 (19.8%) errors identified. Ten errors (1.6%) occurred at the CSC and 123 (18.2%) errors occurred at regional hospitals. More patients who had an error with alteplase administration (12.7%) experienced a hemorrhagic conversion compared to those with no error in administration (7.2%, p = 0.04).

Conclusion: The error rate of alteplase infusion for ischemic stroke is high, particularly in patients from referring centers. Errors may be associated with adverse events. Further education and administration safeguards should be implemented to decrease the risk of medication errors.

(C) 2021

1. Background and purpose

Each year, approximately 795,000 people experience a new or recur- rent stroke. acute ischemic stroke is the most common type of stroke, accounting for 87% of all strokes [1]. Alteplase is the standard of care for early pharmacologic thrombolysis after AIS [2]. However, the dosing of alteplase for this indication is complex and has been asso- ciated with error [3,4]. Alteplase is considered a high-alert medication by the Institute of Safe Medication Practice (ISMP) [5]. There are several opportunities for error when using standard alteplase dosing for stroke including weight-based dosing, two-sequence drug administration pro- tocol (bolus and infusion), and the inability to completely administer the entire dose due to residual drug in the vial after dose administration. Previous trials have identified an increased rate of hemorrhagic conver- sions and a lower likelihood of a good functional outcome at discharge

* Corresponding author.

E-mail addresses: [email protected] (K.A. Dancsecs), [email protected] (M. Nestor), [email protected] (A. Bailey), [email protected] (E. Hess), [email protected] (E. Metts), [email protected] (A.M. Cook).

associated with errors in alteplase administration [3,4]. A multicenter survey comparing alteplase utilization in a Comprehensive Stroke Cen- ter (CSC) with regional hospitals identified that alteplase prescription and administration errors occurred in 64% of all patients with the most common error being incorrect dosage for body weight [6]. More errors occurred at the regional hospitals compared to the CSC.

No published studies utilize a standard taxonomy to classify and identify likely causative factors of medication errors related to alteplase use. The purpose of this study was to categorize alteplase errors using a standardized taxonomy and to identify the difference in medication error rates and outcomes in patients receiving alteplase for acute ische- mic stroke from regional hospitals compared to patients receiving alteplase at a CSC. The primary hypothesis was that the error rate be- tween regional hospitals and the CSC would not be different. Potential causative factors were also explored.

1. Methods

This was a retrospective cohort comparison of patients who were greater than 18 years old, received intravenous alteplase for the

https://doi.org/10.1016/j.ajem.2021.03.043

0735-6757/(C) 2021

treatment of AIS from June 1, 2015 to June 30, 2018, and received care at the CSC. The CSC, University of Kentucky Chandler Medical Center (UKCMC) is a 900-bed hospital that has approximately 1500 annual stroke visits. Patients were excluded if they received any fibrinolytic other than alteplase or if medical records were incomplete. The primary endpoint for the trial was to identify the difference in medication error rates in patients receiving alteplase from a regional hospital compared to patients receiving alteplase at the CSC. Secondary endpoints included categorizing the error type using a standardized taxonomy, and charac- terizing the incidence of hemorrhagic conversion, the Modified Rankin scale (mRS) at discharge, and ultimate discharge disposition. Data were obtained from several institution-specific databases along with the electronic medical record, including intervention databases (the on-call pharmacy resident intervention database and an emergency de- partment pharmacist intervention database). Get With The Guidelines(R) data and UKCMC’s internal safety reporting system were utilized to rec- oncile the data from the institution-specific databases. The level of stroke certification (if applicable) of the referring facilities were identi- fied through institution-specific searches and included Acute Stroke Ready (ASR) Hospital designated by Det Norske Veritas (DNV) and Pri- mary Stroke Centers (PSC) designated by the Joint Commission. Some regional hospitals were not stroke certified. modified Rankin Score (mRS) was dichotomized as good (0-2) and poor (3-6) neurologic out- come. Hemorrhagic conversion was classified according to the European Cooperative Active Stroke Study (ECASS) [7]. The National Co- ordinating Council for Medication Error Reporting and Prevention (NCC-MERP) standardized taxonomy was utilized to classify medication errors. (Table 1) [8]. The NCC-MERP taxonomy serves to provide a stan- dard language and structure of medication-error related data for use of developing databases analyzing medication error reports. Contributing factors developed internally by the UKCMC Medication Safety Officer were listed with each medication event. Contraindications were catego- rized based on the accepted warnings and contraindications for intrave- nous alteplase use [2].

At the CSC, clinical pharmacists are part of the stroke alert team which responds to patients that acutely present with stroke-like symp- toms or if they received alteplase at another institution. This response occurs 24 h a day, 7 days a week. Pharmacists are typically involved with verifying and documenting the alteplase dose and administration from the referring facility (if applicable), double checking the patient’s weight using a bed scale, and ensuring that alteplase is infusing at the correct infusion rate by verifying the rate displayed on the infusion pump. Documentation in the medical record regarding the dosing and details of alteplase administration is completed in this situation as

Table 1

Medication error definitions. Event type Definition

well. Neither the degree of pharmacist involvement or the extent of dos- ing verification at the regional hospitals could be assessed.

Data was managed using Research Electronic Data Capture (RED- Cap) electronic data capture tools hosted at the CSC [9]. Descriptive sta- tistics were used to categorize demographic and other data. All categorical variables were analyzed through Chi-squared analyses. Con- tinuous variables were analyzed through unpaired Student’s t-tests. The primary and secondary endpoints were calculated with a Chi-squared test of proportions (with a two-sided alpha level of 5%). Microsoft Excel Version 2016 was utilized to conduct statistical analyses. The study was approved by the University of Kentucky Institutional Review Board via expedited review and the need for patient consent was waived.

In this retrospective chart review, the primary author served as the abstractor and was not blinded to the study hypothesis. Charts and cal- culations were reviewed by the last author but no inter-abstractor reli- ability was completed. An abstraction guide was developed to standardized how data was extracted from the various databases. The data abstracted, outcomes, and cofounders were defined a priori.

1. Results

During the study period, a total of 676 patients were included the analysis. No patients were excluded. Two-hundred-twenty-seven (34%) patients received alteplase at the CSC and 448 (66%) patients received alteplase at regional hospitals. Of the patients receiving alteplase at a regional hospital, 58 patients (12.9%) were at an acute stroke ready facility (ASR) while 55 (12.3%) were at a primary stroke center (PSC). The remaining 335 patients (74.8%) received alteplase at a regional hospital with no stroke certification. There were a total of five regional centers that were ASR hospitals and three regional centers that had a PSC designation. No statistically significant differ- ences in baseline characteristics were identified between patients who received alteplase at a regional hospital or the CSC aside from the incidence of thrombectomy, which was more common in patients presenting to the CSC (Table 2).

A total of 133 (19.8%) patients had an error associated with alteplase administration. Ten errors (1.5%) occurred at the CSC and 123 (18.2%) errors occurred at regional hospitals. Of the errors related to alteplase administration that occurred at regional hospitals, 16 errors were at an ASR Hospital and fifteen errors at a PSC. The remainder of the errors occurred at undesignated centers (n = 92, 69%). The most common error identified was receiving an over-dosage of alteplase. Each of the 36 patients who received an over-dosage received alteplase at a re- gional hospital. Under-dosing and infusion errors were also very com- mon, as was administration of alteplase in patients with apparent contraindications (Fig. 1). The most common contributing factor leading to a medication error with alteplase was an incorrect calculation,

Wrong dose - overdose Wrong dose - under dose

Wrong rate - too fast

Wrong rate - too

slow

Patient received >110% of 0.9 mg/kg^a dosage of alteplase based upon their weight

Patient received <90% of 0.9 mg/kg^a dosage of alteplase based upon their weight

Alteplase was infused over <1 h based upon infusion pump settings

Alteplase was infused over >1 h based upon infusion pump

settings

Table 2

Demographic data.

CSC Regional hospital p-value (N = 228) (N = 448)

Contraindication Alteplase was infused outside of the 4.5 h window, specific

	risk factors for bleeding	Female, N (%)	109 (47.8%)	238 (53%)	0.17
Monitoring error -	Clinical lab value or monitoring parameter such as blood	Thrombectomy, N (%)	39 (17%)	31 (6.9%)	<0.001
clinical	glucose, or INRb	Mean NIHSS (St Dev)	9 (7)	9 (8)	0.57
Wrong technique	Incorrect and/or improper procedure used in the	Median weight, kg (St Dev)	85 (20)	83 (21)	0.97
	administration of a drug such as not flushing the intravenous	Median modified rankin score (IQR)	3 (3)	2 (3)	0.17

Median age, years (IQR) 62 (20) 63 (23) 0.24

line after alteplase infusion

Other Any medication error that does not fall into the defined categories

^a kg indicates kilograms.

^b Indicates Internationalized Normal Ratio.

N indicates Number.

IQR indicates 25%-75% interquartile range.

NIHSS indicates National Institutes of Health Stroke Scale. St Dev indicates Standard Deviation.

kg indicates kilogram.

Fig. 1. Comparison of medication error, event type.

accounting for 30 (23%) of the errors identified (Table 3). Other contrib- uting factors that led to an error associated with alteplase administra- tion included the infusion pump being programmed incorrectly (n = 27, 20%) and incorrect patient weight being used to calculate the dose (n = 21, 16%).

In total, fifty-six patients (8.3% of the entire study population) expe- rienced a hemorrhagic conversion after the administration of alteplase. Patients who had errors associated with alteplase administration more commonly experienced hemorrhagic conversion compared to those who did not have an error with administration (12.7% vs 7.1%,

Table 3

Contributing factors associated with medication errors.

Contributing factor	Regional hub center (13	Referring facility	Total (156
	errors)a	(143 errors)a	errors)
Calculation	0	30 (21%)	30
			(19.2%)
Infusion pump - programmed	0	27 (18.9%)	27

p = 0.04). Patients with hemorrhagic conversion also had a higher admission NIHSS (8.9 with no hemorrhagic conversion vs 16.3 with hemorrhagic conversion, p < 0.001). There was no difference in the hemorrhagic conversion rate of patients who received alteplase at the CSC compared to those who received alteplase from a regional hospital (p = 0.95).

Overall, during the three-year study period, 41 (5.9%) patients died during the index admission. No difference was seen in the number of patients who died based on where the alteplase administration was ini- tiated [2.3% the CSC patients vs 3.7% patients from regional hospitals (p = 0.45)]. The mortality rate was not associated with errors in alteplase administration (p = 0.63). The modified Rankin Score was only available in 388 patients. More recent versions of the Get with the Guidelines data collection prompted routine collection of mRS, so patients later in the cohort more frequently had an mRS available. The median mRS for all patients was 2 (range 0-6). Good functional out- come (mRS 0-2) was similar in patients who received alteplase at the CSC compared to referring facilities (46.3% at the CSC vs 53.6% from re- gional hospitals, p = 0.17). There was no difference in mRS of patients who had errors associated with alteplase administration compared to those who did not experience errors (Fig. 2, p= 0.13). Discharge dispo-

incorrectly Education	2 (1.5%)	20 (14%)	(17.3%) 22	sition was also similar between groups (Table 3). 376 (56%) patients
			(14.1%)	were discharged to home, accounting for the most common discharge
Patient weight	1 (0.8%)	19 (13.3%)	20 (12.8%)	disposition. A total of 123 (54%) patients who received alteplase at the CSC were discharged to home and 253 (56%) of patients who received

Drug information missing 2 (1.5%) 18 (12.6%) 20

(12.8%)

Preparation error by non-pharmacy	0	8 (5.6%)	8 (5.1%)	of a medication error was not associated with discharge disposition
Infusion pump - correct process not	0	5 (3.5%)	5 (3.2%)	(Table 4).
followed

alteplase at regional hospitals were discharged to home. The occurrence

1. Discussion

Communication	0	4 (2.8%)	4 (2.6%)
EHRb alert override	3	1 (0.7%)	4 (2.6%)
EHRb drug build	1 (0.8%)	2 (1.4%)	3 (1.9%)
Training	1 (0.8%)	2 (1.4%)	3 (1.9%)
Look alike sound alike	0	2 (1.4%)	2 (1.3%)
Delay	0	1 (0.7%)	1 (0.6%)
Drug monitoring	0	1 (0.7%)	1 (0.6%)
Inexperienced staff	0	1 (0.7%)	1 (0.6%)
Interruption/distraction	0	1 (0.7%)	1 (0.6%)
Medication history not completed	1 (0.8%)	0	1 (0.6%)
BCMAc not utilized	1 (0.8%)	0	1 (0.6%)
Atuomated dispensing cabinet (ADC)	1 (0.8%)	0	1 (0.6%)

override

Transfer 0 1 (0.7%) 1 (0.6%)

^a N indicates number. Data is presented as Number, percentage.

^b EHR indicates Electronic Health Record.

^c BCMA indicates Barcode Medication Administration.

The current study evaluated 676 patients who received alteplase at a CSC or regional hospitals. Errors were detected in 133 (19.8%) of alteplase administrations. Ten errors were identified with alteplase ad- ministration at the CSC and 123 errors were identified with alteplase ad- ministration at regional hospitals. Overall, the error rate of 19.8% was less than the error rate of other studies found within the literature but is still quite high given the risk of alteplase and the clinical situation [4]. The current study used the standard NCC-MERP taxonomy to clas- sify medication errors related to medication administration rather than just identifying contributing factors. Our data suggested that the most common medication error with alteplase administration is an overdosage, accounting for medication errors in 36 patients (5%).

Fig. 2. mRS at discharge in patients who experienced errors.

Table 4

Discharge disposition by facility.

Discharge disposition Regional hub center Referring facilities Total

ensuring the correct eligibility, dosing, admixture, and administration of alteplase. Consideration for tenecteplase may also be reasonable, though this is not currently supported as Standard care by the guide- lines. Regional hospitals who administer alteplase should have a phar-

	(Na=228)		(N = 448)		(N = 676)		macist involved in some capacity to handle the complexity of dosing
Home	123 (54%)		253 (56%)		376 (55.6%)		administering alteplase. Tenecteplase dosing (a one-time dose of 0.4
Acute rehab	55 (24%)		106 (24%)		161 (24%)		mg/kg) may reduce the rate of non-weight and non-calculation errors
Expired	16 (7%)		26 (6%)		42 (6%)		compared to alteplase. [11] The use of a standardized infusion pump li-

brary in regional hospitals should also be considered in order to help standardize the rate at which alteplase is infused.

Subacute rehab	2 (1%)	5(1%)	7(1%)	There are several limitations to this study. First, the study was retro-
Other facility	2 (1%)	10 (2%)	12 (2%)	spective in nature, limiting our ability to capture all data and all factors

Skilled nursing facility	12 (5%)	27 (7%)	39 (6%)
Hospice	13 (6%)	16 (4%)	29 (4%)
LTACb	6 (3%)	3 (1%)	9 (1%)

^a N indicates number. Data is presented as Number, percentage.

^b LTAC indicates Long-Term Acute Care Facility.

The most common contributing factor leading to a medication error in patients transferred from other facilities was a calculation error. This appears to be primarily due to the use of incorrect patient weights rather than faulty computations. The second most common contribut- ing factor leading to a medication error in patients transferred from other facilities was incorrectly programmed infusion pumps. This high- lights the need for continued education and awareness of the proper ad- ministration of alteplase as well as understanding the complex dosing scheme of alteplase for AIS. Hard-wiring programs for specific doses of alteplase into smart pump technology may be a helpful step for some in- stitutions to address this issue and could promote the standardization of smart pump libraries across regional networks.

Numerous factors are associated with hemorrhagic conversion rate, including the severity and location of stroke, blood pressure manage- ment early in care, the presence of relative contraindications or warn- ings related to alteplase use, as well as the timing and dosing of alteplase. Our findings confirmed the results of an increased hemor- rhagic conversion rate associated with alteplase administration errors as suggested by Messe and et al. [4] Hemorrhagic conversion was also associated with higher admission NIHSS, which also corroborates previ- ous reports [10]. We were unable to account for the other factors related to hemorrhagic conversion.

Several improvements to the process of alteplase administration could be considered to streamline care and reduce the potential for medication errors. At the CSC, pharmacists are present at all times to as- sist in the stroke alert response. Pharmacists can play a vital role of

related to errors. Many mRS scores (which were not routinely docu-

mented early in the cohort) were not available due to this reason. Sec- ondly, due to the nature of under-reporting of medication errors, we expect that a number of errors were not reported, thus causing our anal- ysis to under-estimate the actual incidence of errors in this population. The authors believe that the routine involvement of the pharmacist in stroke alert response in the CSC facilitated documentation of errors when they were evident, but it is still likely that some errors occurred both in the regional hospitals and in the CSC that were not detected. In addition, the authors only extracted data from the CSC. Therefore, the authors could not determine if the referring institution acknowl- edged the error occurred prior to arrival at the CSC. Lastly, it was not clear what, if any, safeguards or protocols were in place for alteplase ad- ministration at regional hospitals.

1. Conclusion

During the study period, a total of 133 mediation errors were re- ported in 19.8% of patients who received alteplase for the treatment of AIS. There were a variety of contributing factors that lead to errors with alteplase such as the use of an incorrect patient weight, calculation errors, and programming the infusion pump incorrectly. Patients who had an error associated with alteplase administration were more likely to experience hemorrhagic conversion, though this did not lead to a sig- nificant difference in neurologic outcome. The current study cannot link causation of hemorrhagic conversion and errant administration. How- ever, given the high error rate associated with alteplase administration, institutions should consider what additional safeguards or response personnel may be necessary to mitigate the potential for medication er- rors and Patient harm.

Credit author statement

All of the authors agree to the changes of the manuscript as listed below. The corresponding author and last author were responsible for manuscript creation. Each author contributed to the data collection and analyzing the data.

Sources of funding

This research did not receive any specific grant from funding agen- cies in the public, commercial, or not-for-profit sectors. REDCap is avail- able through the NIH CTSA grant (UL1TR000117) at the University of Kentucky.

Declaration of competing interest

The authors have no conflicts of interest to disclose.

References

1. Benjamin EJ, Virani SS, Callaway CW, Chamberlain AM, Chang AR, Cheng S, et al. Heart disease and stroke statistics-2018 update: a report from the American Heart Association. Circulation. 2018;137:e67-492.
2. Powers WJ, Rabinstein AA, Ackerson T, Adeoye OM, Bambakidis NC, Becker K, et al. 2018 guidelines for the early management of patients with acute ischemic stroke: a

guideline for healthcare professionals from the American Heart Association/ American Stroke Association. Stroke J Cereb Circ. 2018;49:e46-110.

1. Sahlas DJ, Gould L, Swartz RH, Mohammed N, McNicoll-Whiteman R, Naufal F, et al. Tissue plasminogen activator overdose in acute ischemic stroke patients linked to poorer functional outcomes. J Stroke Cerebrovasc Dis. 2014;23:155-9.
2. Messe SR, Tanne D, Demchuk AM, Cucchiara BL, Levine SR, Kasner SE, et al. Dosing errors may impact the risk of rt-PA for stroke: the multicenter rt-PA acute stroke survey. J Stroke Cerebrovasc Dis. 2004;13:35-40.
3. ISMP. List of high-alert medications. Institute of Medication Safety Practice; 2018 Accessed February 19, 2020, at https://www.ismp.org/sites/default/files/ attachments/2018-08/highAlert2018-Acute-Final.pdf.
4. Chung LS, Tkach A, Lingenfelter EM, Dehoney SB, Rollo J, de Havenon A, et al. Tissue plasminogen activator prescription and administration errors within a regional stroke system. J Stroke Cerebrovasc Dis. 2016;25:565-71.
5. Hacke W, Kaste M, Bluhmki E, Brozman M, Davalos A, Guidetti D, et al. Thrombolysis with alteplase 3 to 4.5 hours after acute ischemic stroke. N Engl J Med. 2008;359: 1317-29.
6. NCC MERP Taxonomy of Medication Errors. National coordinating council for med- ication error reporting and prevention (NCC MERP). Accessed February 19, 2020, at https://www.nccmerp.org/sites/default/files/taxonomy2001-07-31.pdf; 1998.
7. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap) - a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009; 42:377-81.
8. Al-Khaled M, Matthis C, Eggers J. Predictors of in-hospital mortality and the risk of symptomatic intracerebral hemorrhage after thrombolytic therapy with recombi- nant tissue plasminogen activator in acute ischemic stroke. J Stroke Cerebrovasc Dis. 2014;23:7-11.
9. Ronning OM, Logallo N, Thommessen B, Tobro H, Novotny V, Kvistad CE, et al. Tenecteplase versus alteplase between 3 and 4.5 hours in low National Institutes of Health Stroke Scale. Stroke: J Cereb Circ. 2019;50:498-500.