Article, Emergency Medicine

Rate of patients at elevated risk of opioid overdose visiting the emergency department

a b s t r a c t

Objective: To determine the rate of patients visiting the emergency department who are at risk of opioid overdose. Methods: The electronic records of a 412 bed tertiary care county teaching hospital were searched for emergency department (ED) visits from January 1, 2013 to October 31, 2016 to find patients on at least 100 mg Morphine equivalents (MME) of oral opioid therapy, or an opioid in combination with a benzodiazepine. Records were also searched for patients with a positive urinalysis for opioids when no opioid was present on their home med- ication list. Medication reconciliations were searched for patients at risk of opioid overdose who were subse- quently discharged on naloxone.

Results: An analysis of 2521 patients visiting the ED was performed, and the overall rate of risk of opioid overdose increased from 25.84% to 47.41% (p b 0.0001) in patients meeting inclusion criteria from 2013 to 2016. For patients on opioids, the rate of patients on 100 MME daily or greater increased from 9.72% to 28.24% (p b 0.0001) from 2013 to 2016. The rate of patients on opioid therapy in combination with benzodiazepine therapy did not change signif- icantly from 2013 to 2016. When comparing patients at risk of opioid overdose to total emergency department visits, we found the rate of at risk patients increased significantly from 0.12% to 0.56% (p b 0.0001) from 2013 to 2016. Conclusions: The rate of patients visiting the emergency department at risk of opioid overdose increased significantly from 2013 to 2016. Naloxone was not routinely prescribed to this patient cohort.

(C) 2018

Introduction

Background

The rate of opioid mortality has increased from 1.5 deaths per 100,000 people in the year 2000 to 5.9 deaths per 100,000 people in 2013. In 2013 drug overdose became the leading cause of Accidental death in the United States with a total of 44,000 deaths. Sixteen thou- sand deaths were attributable to prescription opioid painkillers [1]. In 2015 there were 52,404 drug overdose deaths with 20,101 deaths oc- curring from prescription pain relievers and 12,990 from heroin [2].

In 2014, the World Health Organization (WHO) released data de- scribing demographics of patients who are at risk of opioid overdose. Those who are more likely to experience an overdose include: those who are opioid dependent, Injection drug users, people who have un- dergone imprisonment, those who have been in detoxification/rehabil- itation programs or other periods of abstinence, and those who use Prescription opioids. This includes patients who have been appropri- ately prescribed opioids with on-going monitoring. The WHO also

* Corresponding author at: School of Pharmacy, Texas Tech University Health Sciences Center, 3601 4th Street, STOP 8162, Lubbock, TX 79460, United States.

E-mail address: [email protected] (C.F. Seifert).

included a study which found that patients taking 100 morphine milli- gram equivalents (MME) or greater by mouth daily are at increased risk of overdose [4]. Additionally, patients with multiple prescriptions, particularly for benzodiazepines, are at increased risk of overdose [5].

In 2016, the Centers for Disease Control and Prevention (CDC) re- leased guidelines for opioid use in chronic pain patients. Recommenda- tions include provision of a co-prescription for naloxone when the following risk factors are present; history of overdose, history of sub- stance use disorder, use of 50 MME daily or greater, and concurrent use of a benzodiazepine [6].

Naloxone is the most common antidote for opioid overdose with several drug formulations and delivery devices available for lay person administration. It is effective in reversing the effects of opioid agonists for up to 30-120 min [7,8]. Naloxone has demonstrated a favorable safety profile with Opioid withdrawal being the most significant adverse reaction experienced [7,8].

Demand for access to naloxone products has increased substantially in the US alongside increased mortality rates from opioid overdose. The 2014 WHO “Community Management of Opioid Overdose” guidelines recommend that people who may witness an opioid overdose should be provided the opportunity for training in the use of, and have access to, naloxone products [3]. Although naloxone access has increased, data measuring the impact of a Naloxone distribution and education

https://doi.org/10.1016/j.ajem.2018.03.055

0735-6757/(C) 2018

program are sparse [9]. Available data have shown a reduction in opioid induced mortality in communities with naloxone distribution and edu- cation [10].

Importance

To make meaningful interventions, there is a need for simple and ef- fective methods of identifying patients at risk of opioid overdose who visit the emergency department. Brady et al. demonstrated patients vis- iting the emergency department are more likely to experience a pre- scription drug overdose death, with increased frequency of visits being associated with a higher mortality rate [11]. By looking at the rate of patients at risk of opioid overdose, we can determine if the emer- gency department would be an impactful location to dispense naloxone products. Kestler et al. demonstrated when at risk patients are offered naloxone in an emergency department setting, 68% accept the drug as a Rescue medication [12]. Although similar methods were used in deter- mining patients at risk of overdose, this study was conducted in Canada, and several of the means to identify at risk patients are not readily avail- able in the United States, such as the provincial pharmacy database used to screen for at risk patients. If patients at risk of opioid overdose can be identified upon visiting the emergency department, using readily avail- able information, then interventions can be made with the intent to re- duce mortality.

Goals of this investigation

This study sought to determine the rate of patients at risk of opioid overdose compared to all patients who use opioids by identifying pa- tients using high dose opioids, have a co-prescription for a benzodiaze- pine and an opioid, and have a positive urinalysis for opioids upon visiting the emergency department.

Methods

Study design and setting

This study was a retrospective chart review of a public, non-profit, 412 bed county teaching hospital serving a county and surrounding area population of around 300,000. The authors obtained institutional review board approval from the university and hospital (IRB approval number L17-057).

Selection of participants

The electronic medical records were searched for patients reporting to the hospital’s emergency department from January 1, 2013 to Octo- ber 31, 2016 who had an opioid medication on their emergency depart- ment (ED) medication reconciliation, a benzodiazepine and a co- prescription for an opioid on their ED medication reconciliation, and pa- tients with a positive urine drug screen for opioids upon admission.

Table 2

Doses used to qualify 100 MME daily use or greater.

Opioid High Risk Dose

Codeine >=660 mg/daya

Fentanyl transdermal patch (ug/h) >=50 ug/hb

Hydrocodone >=100 mg/day

Hydromorphone >=25 mg/day

Methadone >=20 mg/dayc

Morphine >=100 mg/day

Oxycodone >=66 mg/day

Oxymorphone >=33 mg/day

Tapentadol >=250 mg/day

a The maximum recommended daily dose of codeine is 360 mg/day [13].

b Equianalgesic dosing for fentanyl is 41.6 ug/h., but patches come in 25 ug/h. and 50

ug/h. concentration. Of note, the dosing for fentanyl is in ug/h, not mg/day.

c Due to the varying pharmacokinetic properties of methadone, patients receiving 20 mg of methadone or more daily will be considered high risk even though this does not directly equate to 100 MME.

Exclusion criteria were as follows: age b18 and over 89, patients who are pregnant, and patients who are prisoners.

Interventions, methods, and measurements

A data request was issued to the Report Developer in the IT Analytics department of the hospital specifying parameters for data collection. Pa- tients were sorted by medical record number (MRN) as well as risk fac- tor for opioid overdose; total daily opioid use of 100 mg morphine equivalents (MME) daily, co-prescription for an opioid and a benzodiaz- epine, and positive urine drug screen for opioids with no opioid pre- scription being present on the patient’s ED medication reconciliation. Each MRN was only used once, so no patients were counted as repeat visits. Data was analyzed by manual review of information using Microsoft Excel. Analysis was confirmed independently by two reviewers.

Table 1 illustrates the criteria used to search for the presence of opi- oids on the patient’s ED medication reconciliation. All patients reporting to the emergency department with an opioid on their medication rec- onciliation were reviewed for the presence of 100 MME daily use or greater. One-time orders, duplicate orders, and orders without a clear total daily dosage were excluded from analysis. Table 2 indicate MME criteria.

Patients were also assessed for a co-prescription for a benzodiaze- pine. To maintain separation of risk factors for overdose, patients who were already receiving 100 MME or greater daily were removed from the total count. One-time orders, duplicate orders, and rescue doses were removed from analysis. Table 3 was used to determine if a patient had a benzodiazepine on their ED medication reconciliation.

Patients with a urine drug screen with a presumptive positive or positive result for opioid drugs were compared with ED medication rec- onciliations to determine if a prescription for an opioid was present. Du- plicates and results other than “presumptive positive” or “positive” were removed from analysis. The intent of using urine drug screen was to include illicit use as a risk factor for opioid overdose.

Table 1

Search terms used to identify opioids on ED medication reconciliation.

Generic Name Brand Name

Codeine Tylenol #3, Tylenol #4

Table 3

Search terms used to identify benzodiazepines on ED medication reconciliation.

Generic Name Brand Name

Fentanyl transdermal patch (mcg/h)

Duragesic

Alprazolam Xanax

Hydrocodone Hysingla ER, Zohydro ER, Vicodin, Norco, Lortab, Xodol Hydromorphone Dilaudid, Exalgo

Methadone Dolophine

Morphine MS Contin, Kadian

Oxycodone Oxycontin, Roxicodone, Xtampza ER, Oxaydo, Endocet, Percocet, Xartemis XR

Oxymorphone Opana, Opana ER

Chlordiazepoxide Librium

Clobazam Onfi

Clonazepam Klonopin

Diazepam Valium

Lorazepam Ativan

Temazepam Restoril

Triazolam Halcion

Table 4

Patient demographics.

Patient Demographic

Total Population n = 2521

100 MME or Greater n = 88

Opioid & Benzodiazepine n = 165

Male (%)

55.97

47.72

46.06

Median (IQR) Age (Years)

47.63 (IQR = 23)

51.8 (IQR = 21.25)

53.15 (IQR = 18)

Race (%)

Asian

0.08

0.00

0.00

Black/African American

13.8

10.23

13.3

Hispanic

9.44

6.81

5.45

Native American/Alaskan

0.08

0

0

Pacific Islander/Native Hawaiian

0.08

0

0

White

71.9

82.95

78.78

Other

4.56

0

2.42

Patients who met the criteria for being at risk of opioid overdose then had their discharge medication reconciliation reviewed to deter- mine if a naloxone prescription had been written for the patient.

Outcomes

The primary outcome was to determine the rate of patients visiting the emergency department who were at risk of overdosing on opioid drugs based on the previous criteria.

Analysis

Data were analyzed using the add-on Excel statistical package Ana- lyze-it 3.90.7 (copyright Leeds, England 1997-2017). Continuous data were reviewed for normality using the Shapiro-Wilk test. All continuous data were found to be non-parametric and central tendencies are re- ported as medians with interquartile ranges (IQR). Descriptive statistics were used to compare patient demographics, odds ratios and relative risk ratios. Nominal data were compared with the Pearson Chi-Square test. An alpha level of significance was defined a priori at a p-value b0.05.

Results

Main results

Table 4 shows the patient demographics of the groups. While there was a greater proportion of males, it was noted that more women were in the at-risk groups. Racial breakdown was comparable for this geographic area, however, much fewer Hispanic patients were at risk of opioid overdose than the racial makeup of the area. A total of 2521 pa- tients visiting the emergency department were assessed for the use of a high daily dose of opioid medications. Of those, 540 patients had an opi- oid medication on their ED medication reconciliation with 88 patients taking 100 MME or greater daily. From 2013 to 2016 the rate of patients using high dose opioids increased significantly from 14/144 medication reconciliations to 37/131 medication reconciliations (p b 0.0001, Table 5). Significance is lost when measuring year to year with 2013 to 2014 showing an increase from 14/144 to 18/161 (p = 0.678), 2014 to 2015 showing an increase from 18/161 to 19/104 (p = 0.104), and 2015 to 2016 showing and increase from 19/104 to 37/131 (p =

0.075). Of note, the data for 2016 is limited to those visits occurring on or before October 31, 2016.

Of the 540 patients with an opioid medication on their ED medica- tion reconciliation, 208 patients also had an order for a benzodiazepine drug. After excluding patients taking 100 MME or greater, 165 remained (Table 6). No significant changes were recorded in the rate of patients being co-prescribed benzodiazepines and opioid medications. From 2013 to 2014 the number of patients with both classes of medications on their ED reconciliations increased from 43/118 to 46/81 (p = 0.08), from 2014 to 2015 the rate of patients with both medications on their profiles decreased from 46/81 patients to 36/100 (p = 0.09), from 2015 to 2016 the rate decreased from 36/100 to 40/135 (p = 0.46), and the overall change from 2013 to 2016 was 43/118 to 40/135 (p = 0.41).

A total of 2022 opioid urinalyses were assessed. After excluding du- plicates, 1547 urinalyses that returned a positive or presumptive posi- tive result were included. Positive or presumptive positive results were identified for 806 patients with no history of opioid medications on their ED medication reconciliation (Table 7). From 2013 to 2014 there was no significant change in the rate of positive urine drug screens for opioids in patients without an opioid on their ED medication recon- ciliation (p = 0.23). From 2014 to 2015 there was a significant increase in positive urine drug screens with 180/412 being positive in 2014 and 285/485 being positive in 2015 (p = 0.01). There was no significant change from 2015 to 2016 (p = 0.74). Overall, the rate of positive uri- nalysis for opioids in patients without an opioid on their ED medication reconciliation was significant (p = 0.008).

When all three factors for assessing risk of opioid overdose are com-

bined and compared to overall emergency department visits, we find the percentage of patients visiting the emergency department who are at risk of opioid overdose has increased significantly over the study pe- riod as shown in Table 8. In 2013, we found 100 patients at risk of opioid overdose from 80,375 total emergency department visits indicating 0.12% of total patients visiting the emergency department are at risk of opioid overdose. This percentage of at risk patients increased to 0.29% in 2014, 0.42% in 2015, and 0.56% in the first 10 months of 2016. Each of these year to year increases were significant with a p b 0.0001. Additionally, the overall increase in percentage of patients visiting the emergency department demonstrated a significant increase (p b 0.0001). As the rate of at risk patients visiting the emergency depart- ment increased, so did the county mortality rate due to opioid overdose

Table 5

Rate of patients on 100 MME or greater daily opioid therapy.

Table 6

Patients with an opioid and benzodiazepine medication on ED medication reconciliation.

Year

# of Patients on

# of Patients w/Opioid on

% of Patients

p-Value

Year

Patients w/Benzo &

Total Benzo Rx

% of Patients

p-Value

>=100 MME N = 88

Admit Med Rec N = 540

w/ >=100 MME

Opioid N = 165

N = 434

w/Benzo & Opioid

2013

14

144

9.7

2013

43

118

36.4

2014

18

161

11.2

0.678

2014

46

81

56.8

0.08

2015

19

104

18.3

0.104

2015

36

100

36.0

0.09

2016a

37

131

28.2

0.075

2016a

40

135

29.6

0.46

Overall

b0.0001

Overall

0.41

a 2016 Data only through 10/31/2016. a 2016 Data only through 10/31/2016.

Table 7

Opioid urinalysis results.

Year

Positive Urinalysis without Opioid on

Total Positive Urinalysis

% Positive Without Opioid

p-Value

Profile N = 806

N = 1547

on Profile

2013

43

125

34.3

2014

180

412

43.7

0.23

2015

285

485

58.8

0.01

2016a

298

525

56.8

0.74

Overall

0.008

a 2016 Data only through 10/31/2016.

0.60% 8

7

% of ED Pts. at High Risk of opioid OD

Lubbock County Opioid Mortality

0.50%

6

0.40%

5

0.30% 4

3

0.20%

2

0.10%

1

as shown in Fig. 1. Several more years of data will need to be collected

0.00%

0

2014 2015 2016

before a true correlation can be determined.

A total of 17 orders for naloxone were found on discharge medica- tion reconciliations. Of those, only one order was for a patient who met the criteria for being at risk of overdose in this study. Once again, using the discharge medication reconciliation is prone to error due to inaccurate reporting on the list.

Limitations

Several limitations are present including the retrospective nature of the study. A major limitation is the use of medication reconciliations to gather medication data where human error could bias information lead- ing to over or under reporting of opioid dosage and use. Errors in med- ication reconciliations could also result in inaccurate reporting of naloxone orders written for high risk patients upon discharge from the emergency department. It is important to note the total number of urine drug screens increased from 2013 to 2015, and this may contrib- ute to artificially increasing the rate of positive urinalysis. Utilizing urine toxicology screens has multiple limitations due to potential con- founders including legitimate opioid administration prior to arriving in the ED, bias from indications for ordering urine toxicology screening, and no confirmatory testing, among others. Despite these limitations, the ability to identify patients at risk of opioid overdose is very pertinent in the emergency department setting.

Discussion

As opioid mortality continues to increase in the United States, it is important for healthcare professionals to take a front-line role in mini- mizing the adverse impact opioids have within our communities. Suc- cessfully identifying patients at risk of overdose is key to initiating necessary interventions to, hopefully, reduce mortality. Through a ret- rospective study we successfully identified patients at risk of opioid overdose using information that is readily available on the patient’s electronic medical record. Fig. 1 shows that the overall risk of opioid overdose in the ED of a county facility mirrors the actual mortality rate from opioids in that same county. If this relationship holds, this could potentially be used as a measure to determine if a reduction in mortality exists when patients identified as at risk of opioid overdose are provided with counseling, education, and take-home naloxone.

Table 8

Total patients at risk of opioid overdose.

Year

Total # of Patients at Risk

Total # of Patients

% of Patients at Risk

p-Value

Total # of ED Visits

% of Total ED Patients at Risk

p-Value

2013

100

387

25.84

80,375

0.12

2014

244

654

37.31

0.006

84,048

0.29

b0.0001

2015

340

689

49.35

0.005

81,580

0.42

b0.0001

2016a

375

791

47.41

0.660

67,502

0.56

b0.0001

Overall

b0.0001

b0.0001

a 2016 Data only through 10/31/2016.

UMC ER Opioid Overall Risk Lubbock County Opioid Mortality*

*Information from the Lubbock County Medical Examiner

Fig. 1. Percent of patients presenting to UMC ED at high risk of opioid overdose vs. Lubbock County opioid mortality per 100,000 population.

There were several interesting findings in our data analysis. From 2013 to 2016 the rate of opioids on ED medication reconciliations did not increase significantly, but the rate of patients on 100 MME or greater daily did increase significantly. Several factors may influence this be- havior such as tolerance development through long term opioid ther- apy, but we cannot say why it appears the daily dosage of opioids increased significantly in this study.

Benzodiazepine prescriptions followed a similar trend as opioids. It appears the total number of new benzodiazepine prescriptions in this patient population did not increase, and the rate of patients taking a benzodiazepine along with an opioid did not increase over the study pe- riod. Once again, the reason why may by multifactorial, but the study cannot determine why the rate did not change. It is important to note that, although the rate of using a benzodiazepine and opioid concur- rently did not change significantly, an average of 39.7% of patients with a benzodiazepine on their ED medication reconciliation also had an opioid on the reconciliation. According to current CDC guidelines, this means that nearly 40% of the patients whose records we reviewed and who were also on benzodiazepines, should receive a prescription for naloxone as a rescue medication as well as receive naloxone educa- tion and training.

The data available from the drug urinalysis screening is difficult to interpret and is prone to bias. As a prospective measure, urine drug screens may be more beneficial when more information can be obtained directly from the patient and/or healthcare team to determine accuracy of results, including undocumented opioid administrations by emer- gency personnel. As a retrospective tool, too many confounding factors risk skewing the data.

Conclusions

Through this study we have determined that identifying patients visiting the emergency department who are at risk of opioid overdose is a realistic and achievable task. In our department, however, take- home naloxone was infrequently prescribed to at-risk patients.

Financial support

This research received no financial support.

Conflicts of interest

The authors do not have any conflicts of interest.

Author contributions statement

JRP and CFS conceived and designed the study. CFS supervised the conduct of the study as well as the data collection. CFS provided statis- tical advice and helped JRP analyze the data. JRP drafted the manuscript, and both authors contributed to its revision. CFS takes responsibility for the paper.

References

  1. Levi J, Segal L, Martin A. The facts hurt: a state-by-state injury prevention policy re- port. Trust for America’s Health. 64-65 http://www.healthyamericans.org/reports/ injuryprevention15/; June 2015, Accessed date: 15 December 2016.
  2. Rudd RA, Seth P, David F, Scholl L. Increases in drug and opioid-involved overdose deaths — United States, 2010-2015. MMWR Morb Mortal Wkly Rep 2016;65:1445-52.
  3. World Health Organization (WHO). Community management of opioid overdose. Geneva (Switzerland): World Health Organization (WHO). http://www.who.int/ substance_abuse/publications/management_opioid_overdose/en/; 2014. [Accessed

    December 15, 2016].

    Bohnert AS, Valenstein M, Bair MJ, et al. Association between opioid prescribing pat- terns and opioid overdose-related deaths. JAMA 2011;30(13):1315-21.

  4. Dunn KM, et al. Opioid prescriptions for chronic pain and overdose: a cohort study. Ann Intern Med 2010;152(2):85-92.
  5. Dowell D, Haegerich TM, Chou R. CDC guideline for prescribing opioids for chronic pain – United States, 2016. MMWR Recomm Rep 2016;65(1):1-49.
  6. Narcan (naloxone hydrochloride) [prescribing information]. Radnor, PA: Adapt Pharma; November 2015.
  7. Evzio (naloxone hydrochloride) [prescribing information]. Richmond, VA: Kaleo; April 2014.
  8. McDonald R, Strang J. Are take-home naloxone programmes effective? Systematic review utilizing application of the Bradford Hill criteria. Addiction Jul 2016;111 (7):1177-87.
  9. Walley AY, Xuan Z, Hackman HH, et al. Opioid overdose rates and implementation of overdose education and nasal naloxone distribution in Massachusetts: interrupted Time series analysis. BMJ 2013;346:f174. https://doi.org/10.1136/bmj.f174.
  10. Brady JE, Dimaggio CJ, Keyes KM, Doyle JJ, Richardson LD, Li G. Emergency depart- ment utilization and subsequent prescription drug overdose death. Ann Epidemiol 2015;25(8):613-619.e2.
  11. Kestler A, Buxton J, Meckling G, et al. Factors associated with participation in an emergency department-based take-home naloxone program for at-risk opioid users. Ann Emerg Med 2017;69(3):340-6.
  12. Acetaminophen and Codeine Tablets [prescribing information]. Huntsville, AL: Qualitest Pharmaceuticals; September 2016.

Leave a Reply

Your email address will not be published. Required fields are marked *