Article, Emergency Medicine

Descriptive study of drug-drug interactions attributed to prescriptions written upon discharge from the emergency department

a b s t r a c t

Objective: The primary purpose of this study was to identify the most common drug-Drug interactions (DDI’S) in patients prescribed medications upon discharge from the emergency department.

Methods: We conducted a respective chart review of patients discharged home with a prescription from an aca- demic emergency department. The study period was from August 1, 2015 to August 31, 2015. Patients will be ex- cluded if they meet the following criteria: age under 20 years; Discharge home without a prescription; inpatient hospital admission; transfer to another inpatient facility; or sign out against medical advice. The primary end- point is the identification and characterization of Drug-drug interactions caused by discharge prescriptions writ- ten by the treating physician.

Results: A total of 500 patient charts were included, with 38% having at least one DDI. Overall, there were 429 DDIs among 858 prescriptions written. 15.6% (n = 67) of the DDI’s were classified as B, no modification of ther- apy needed. 60% (n = 260) of the DDIs were risk-rating category C, requiring monitoring of therapy. 22% (n = 95) of the DDI’s identified were category D, which are consider modification of therapy. Lastly, we identified 1.6% (n = 7) category X DDI’s. The top 3 most commonly associated drugs were oxycodone/acetaminophen, ibu- profen, and ciprofloxacin.

Conclusion: DDIs are occurring upon discharge from a large, urban, tertiary care, academic medical center. Many of the DDI’s identified do not require any modification to therapy. However, 23.6% of identified DDI’s required modification or were contraindicated. A majority of the category X drug interactions involved QT prolongation.

(C) 2019


Adverse drug reactions (ADRs) occur in approximately 25% of pa- tients in the ambulatory care setting. Of these Adverse drug reactions, 39% are preventable or ameliorable with most preventable ADRs being the result of Drug-drug interactions (DDIs) [1]. DDIs result from a com- bination of at least two drugs leading to a change in potency, safety, or efficacy of one drug because of another drug [2].

The rate of DDI introduction in the emergency department (ED) is

estimated to occur anywhere from 3% to 47% [3-6]. DDIs occurring upon receipt of a prescription upon discharge from an ED are less likely to be monitored as patients may not make or keep follow up appoint- ments, and follow up typically occurs with a physician who did not pre- scribe the medication in the ED [3].

In 2002, Gaddis et al. conducted a retrospective chart review of 200 outpatient ED visits to evaluate if medication prescribed, administered,

* Corresponding author.

E-mail addresses: [email protected] (P.J. Bridgeman), [email protected] (J. Mele), [email protected] (G. Wei).

or dispensed by emergency medicine physicians had become the most common cause of DDIs. They evaluated patients at high-risk for DDIs, in- cluding patients >=60 years old taking three or more medications or pa- tients taking five or more medications regardless of age. On average patients were 64.5 (+-17.1) years old and were taking 7.2 medications. Out of the 200 patients evaluated, 50 had at least one DDI prior to the emergency department visit. Only 7 DDIs occurred in 140 patients who received a new medication from an emergency medicine physi- cian. The drugs implicated in these seven interactions were as follows: four incidences of meperidine-promethazine, one incidence of warfarin-aspirin, one incidence of theophylline-diazepam, and one inci- dence of aspirin-antacids [7].

Similarly, in 2013, De Paepe et al. evaluated DDIs and ADRs in pa- tients >=65 years old admitted to the ED to characterize drugs most fre- quently prescribed by emergency medicine physicians leading to ADRs. In total, 55 patients received a medication in the ED resulting in 39 interactions in 23 patients. Majority of the DDIs were risk rating C (62%) followed by B (31%) followed by D (7%). All D interactions were considered benefit greater than risk by an expert panel. None of the C interactions were considered unjustified by the same expert panel. 0735-6757/(C) 2019

The most common drugs occurring from most often to least often were as follows: beta2 agonists, ACE inhibitors, coumarin anticoagulants, di- goxin, and NSAIDs [8].

Given the focus of prior studies on high risk or elderly populations, this study was undertaken to estimate the occurrence and severity of DDIs resulting from medications prescribed in the ED to a more gener- alized population.


The objective of this study is to identify and characterize the most common DDIs attributable to medications prescribed upon discharge from the ED.


The study was a retrospective chart review performed at large, urban, tertiary care, academic medical center. The study was approved by the institutional review board. A patient list of all adult ED visits be- tween August 1st 2015 and August 31st 2015 which met study criteria was generated using a query of the electronic medical record. Patients who were discharged home from the ED with a prescription for medica- tion(s) were included in the query. Patients were excluded if they met the following criteria: discharged home without a prescription; inpa- tient hospital admission; transferred to another inpatient facility; or signed out against medical advice. Patients who left against medical ad- vice were excluded due to the inability to perform proper medication reconciliation prior to discharge. As this study was performed in an adult ED, charts belonging to patients under the age of 20 were not in- cluded in the analysis. Patients under the age of 20 were excluded based on logistics as they are seen in a separate pediatric ED at our facil- ity. In order to observe consistency in prescribing practices we chose to exclude this age group. A convenience sample of the first 500 charts was used.

Lexicomp interaction analysis was utilized to identify DDIs between at home and newly prescribed medications, and to determine DDI se- verity. DDI severity was classified per the risk-rating category, a risk rat- ing of A translates into no known interaction, a risk rating of B identifies an interaction which requires no clinical action, a risk rating of C iden- tifies an interaction that requires Clinical monitoring, a risk rating of D identifies the need to consider therapy modifications, and finally a risk rating of X represents therapies that are contraindicated (see Table 1). Baseline demographic data including age, gender, Home medications, and medications prescribed upon discharge were extracted from the medical record. All data was analyzed using descriptive statistics.


A patient list was generated using the Emergency Department Infor- mation Management Solution (EDIMS) to identify patients with the above inclusion and exclusion criteria from ED visits spanning August 1st 2015 to August 31st 2015. The first 500 charts from the generated list were evaluated. Patients were 42 years of age on average and mostly female (58.8%). Among the 500 patients, a total of 858 prescriptions were written upon discharge. Of the 500 patient charts reviewed, 38% (n = 190) of patients had at least one DDI of risk rating B or greater.

Of the 190 patients who had a DDI on review of their chart, the largest group, or 45.3% (n = 86) had only one DDI with the subsequent break- down illustrated in Fig. 1.

There were a total 429 DDIs identified among the 190 patients. Out of a total of 429 DDIs, 15.6% (n = 67) of the DDIs were classified as B, “requiring no intervention”, 60% (n = 260) of the DDIs classified as a C, “requiring therapeutic monitoring”, 22% (n = 95) of the DDIs identi- fied were category D, which are “consider modification of therapy.” Lastly, 1.6% (n = 7) category X DDIs or contraindicated combinations were identified. The study results of DDI classification by risk rating are highlighted in Fig. 2.

The top five drugs implicated in the 429 DDIs were oxycodone- acetaminophen (18%), ibuprofen (10%), ciprofloxacin (9%), prednisone (9%), and albuterol (9%). The top five drug interactions for category C and D interactions which require clinical attention and all category X in- teractions which are contraindicated are reported in Table 2.


To our knowledge, this is the first retrospective chart review to eval- uate DDIs occurring at discharge from an ED visit in a sample of patients who are not age or risk stratified. The first 500 adult patient visits were included regardless of age or number of home medications, giving an overall prevalence of DDIs occurring upon discharge. As such, this study provides a more generalizable assessment of the prevalence of DDIs resulting from emergency department prescriptions.

Themes were identified in the risk rating DDI severity, for example many of the top interactions identified involved prolongation of the QTc, increased bleeding risks, and CNS depression. Our study found that the top drug implicated was oxycodone/acetaminophen, which was not seen in the De Paepe study. This reflects the overall increase in opioid use in the United States which increased dramatically throughout the 1990’s through 2012. However, Opioid prescriptions ap- pear to have crested in the 2010-2012 period and started declining in 2012, furthermore, additional initiates have been launched in the last 2 years to further decrease opiate prescriptions [9]. Similar to the De Paepe study, NSAIDs and beta2 agonist were once again among the most common drugs involved in the DDIs. Furthermore, the DDI combi- nations identified in the 2002 study by Gaddis et al. have changed as the use of medications like theophylline and meperidine have decreased.

The majority of DDIs identified in this study fell into category C, with recommended monitoring of therapy. This finding highlights the impor- tance of considering a patient’s outpatient follow up when prescribing in the ED. The prevalence of category C DDIs in this and previous studies

Table 1 Lexicomp Online(R) Severity Scoring System using Risk Rating Category and subsequent DDI management recommendation.

Risk rating category Recommendation

A No known interaction

B No action needed

C Monitor therapy

D Consider therapy modification

X Avoid combination Fig. 1. Number of drug-drug interactions experienced by patients.

Fig. 2. Number of DDI per Risk-rating category.

such as the one by De Paepe evidence the necessity of reconciling a patient’s discharge medication according to their access to primary care in addition to their current medication list.

Category C drug interactions accounted for the largest proportion of interaction identified and may be encountered more frequently by emergency medicine physicians. Therefore, further discussion of the Clinical context of the drug interactions is warranted to assist emer- gency medicine physicians in performing a risk benefit analysis. Oxyco- done/acetaminophen and ciprofloxacin were identified as a category C drug interaction, with an increased risk of myoclonus being the main concern. This DDI is derived from a single case report of an 80 year old female initiated on ciprofloxacin and oxycodone and subsequently de- veloped myoclonus after 7 days of treatment [10]. Both opioids and fluoroquinolones been associated with neurologic adverse events such as seizures, delusions and hallucinations. Myoclonus may be a concern when these two medications are utilized together particularly in pa- tients at higher risk for neurologic adverse events. potential risk factors for higher risk of adverse neurologic effects are extremes of age, renal disease, history of neurologic disease, or impaired blood brain barrier integrity [11]. The potential clinical effect resulting from the ciprofloxa- cin and metronidazole drug interaction is additive qt prolongation.

Since completion of data collection and analysis the category C des- ignation was removed from the Lexicomp database. Additionally, the metronidazole and ciprofloxacin drug interaction we removed as a

Table 2

Top Five interactions for risk-rating category C & D and all X. Risk rating category Most common DDI

C o Metronidazole + ciprofloxacin

  • Oxycodone/acetaminophen + ciprofloxacin
  • Oxycodone/acetaminophen + HCTZ
  • Lisinopril + ibuprofen
  • Ondansetron + metronidazole

D o Oxycodone/acetaminophen + diazepam

  • Oxycodone/acetaminophen + cyclobenzaprine
  • Aspirin + ibuprofen
  • Ciprofloxacin + ondansetron
  • Ibuprofen + sertraline

X o Levofloxacin + citalopram

  • Ciprofloxacin + escitalopram
  • Ciprofloxacin + sotalol
  • Fluoxetine + ondansetron
  • Docusate + mineral oil
  • Celecoxib + naproxen
  • Dextromethorphan/quinidine + ondansetron

category C interaction [12]. However, both interactions are still listed as major drug interaction in a database by a separate publisher [13]. There is a case report that demonstrates metronidazole may prolong the qt interval and a separate case report where an occurrence of torsade’s des pointes (tdp) occurred which was attributed to concomi- tant administration of metronidazole and amiodarone. There have not been any documented case reports of tdp caused by concomitant ad- ministration of ciprofloxacin or ondansetron and metronidazole, al- though theoretically the risks does exist.

When evaluating the oxycodone/acetaminophen and hydrochloro- thiazide drug interaction there are three possible orthostatic hypoten- sion, hyponatremia and reduced efficacy of diuretics. Opioids may increase excretion of antidiuretic hormone causing a pharmacodynamic drug interaction possibly augmenting naturetic effects of diuretics and decreasing a diuretics effectiveness. Hyponatremia caused by opioids is an extremely rare side effect, but should be taken into consideration if a patients sodium begins to decrease while taking opioids [14]. The in- creased risk of orthostatic hypotension may predispose patients to a fall when these medications are taken together. Communicating this risk to patients will allow them to implement strategies at home that may de- crease their risk of fall if these medications are deemed required to be used concomitantly.

Lisinopril and ibuprofen was the final category C drug interaction identified in the study patient population. Overall the data supporting this interaction is more robust when compared to the other category C drug interactions identified. Increased rates of acute kidney injury and worsening of renal function have been identified with concurrent us of Non-steroidal anti-inflammatory drugs (NSAID’s) and ACE inhibitors. The risk of is further increased with concurrent use if diuretics [15,16]. Additionally, the antihypertensive effects of ACEi’s may be attenuated with concurrent use of NSAID’s. Therefore, the use of NSAID’s with ACE’s should be discouraged especially in patients with pre-existing renal dysfunction, the elderly, or those otherwise at risk for acute kidney injury. Alternative agents should be utilized for pain control such as acetaminophen. If the ACEi’s and NSAID’s are deemed necessary to be utilized together renal function and blood pressure should be assessed more frequently than if the agents.

The category D interactions should make clinicians consider selecting another equivalent therapy that avoids the identified interac- tion taking the clinical situation into account. The first two interaction identified were oxycodone/acetaminophen and either diazepam or cyclobenzaprine. In a safety statement from FDA concurrent use of opi- oid pain medications and benzodiazepines or CNS depressants is dis- couraged and should be avoided if possible. According to the FDA statement there is increased risk of overdose death when these types of medications are used together. Additionally, black box warnings are going to be added to the labeling for opioids and benzodiazepines. These statements are consistent with guidance provided by CDC regard- ing opioid and benzodiazepine or CNS depressant use [17]. Additionally, when diazepam or placebo in addition to naproxen was evaluated for the treatment of lower back no additional benefit was observed [18]. Another trial evaluated naproxen with either acetaminophen/oxyco- done or cyclobenzaprine or placebo and found no difference at 1 week in pain or functionality [19]. The category D interaction of ibuprofen or another NSAID and aspirin may be a frequently encountered drug in- teraction for emergency medicine physicians. When considering pre- scribing ibuprofen to patients currently taking aspirin there may be an increased risk of gastrointestinal bleeding. Additionally, NSAID’s may decrease aspirin’s cardioprotective efficacy. Thus, it is reasonable to avoid chronic use of NSAID’s for patients who also take aspirin for cardioprotection [20].

In the setting of an ED patient it may be reasonable for a short course

of an NSAID if required for pain control. However, an alternative agent such as acetaminophen should be considered for use in this patient pop- ulation. Patients should be advised to observe for any signs of gastroin- testinal bleeding.

Generally, the category X interactions identified should prompt the provider to prescribe an alternative agent that does not expose the pa- tient to unnecessary risk. Combinations of the identified medications should be avoided. A majority of the identified category X interactions involved prolongation of the qt interval and increased risk of tdp. The identified interaction of mineral oil and Colace is based on the possible risk for increased absorption of mineral oil leading to systemic toxicity. It is important to note this particular drug interaction is dependent on the route of mineral oil administration. It is not recommended to ad- minister the Oral formulation of mineral oil [21]. The combination of ibuprofen and sertraline has been identified to possibly increase the risk of bleeding through a number of cases and clinical trials. This inter- action may be generalizable to NSAID’s and SSRI’s. The data is conflicting regarding the actual occurrence of bleeding associated with concurrent use of these medications. Taking this information into account it is rea- sonable to avoid the combination of these two Medication classes [22]. There are important limitations of our study that need to be noted.

This study was confined to a single, urban, academic tertiary care facility and so the prevalence of DDIs noted in this study may not be generaliz- able across all practice settings. A multicenter study would be likely to provide a more accurate estimation of DDI prevalence and severity re- lating to medications prescribed in the emergency department.

We utilized only one drug interaction checker to evaluate discharge drug interactions. There are other programs available to evaluate drug interactions that may differ on their classification and severity. There- fore, the results may differ if another drug interaction checking software is utilized. Due to the fact this was a retrospective chart review we were not able to ascertain if patients actually filled the original discharge pre- scription. It is possible that an intervention to avoid a drug interaction occurred upon presentation to an outpatient pharmacy.

Our data may also be limited as we excluded patients b20 years old. At our institution patients in this age range are seen in a separate pedi- atric emergency department. We decided to exclude this patient popu- lation as different physicians see these patients and it may have introduced prescribing practices that were not reflective of the adult ED population at our facility.

In addition, medication reconciliation was performed using medica- tion lists provided by patients. Prior studies have demonstrated that el- derly patients and those with poor health literacy are often unable to provide complete Medication lists [4,23]. As such it is possible that our results are an underestimation of actual DDIs resulting from ED prescriptions.

We also did not convene an expert panel, as previous studies have done, to evaluate the clinical applicability of the identified drug interac- tions. Our overall goal was to make prescribers aware of the most com- mon DDI’s and allow them to provide education to patients regarding the need for increased monitoring for adverse events.

The results of this study suggest that both patients and providers in the emergency department may benefit from the implementation of a drug interaction screening check prior to discharge. As such, this study fills a current gap in the literature and elucidates an area of current prac- tice which can be improved to reduce adverse outcomes.


The authors declare no potential conflicts of interest.


  1. Gandhi TK, Weingart SN, Borus J, et al. adverse drug events in ambulatory care. N Engl J Med 2003;348(16):1556-64.
  2. Ellenhorn MJ, Sternad FA. Problems of drug interactions. J Am Pharm Assoc 1966;6

    (2):62-5 [passim].

    Beers MH, Storrie M, Lee G. Potential adverse drug interactions in the emergency room. An issue in the quality of care. Ann Intern Med 1990;112(1):61-4.

  3. Hancock D, Kennington JM, Beckner RR, Quick G. Emergency department medication and drug interaction evaluation. Hosp Pharm 1992;27(2):129-32.
  4. Herr RD, Caravati EM, Tyler LS, Iorg E, Linscott MS. Prospective evaluation of adverse drug interactions in the emergency department. Ann Emerg Med 1992;21(11): 1331-6.
  5. Goldberg RM, Mabee J, Chan L, Wong S. Drug-drug and drug-disease interactions in the ED: analysis of a high-risk population. Am J Emerg Med 1996;14(5):447-50.
  6. Gaddis GM, Holt TR, Woods M. Drug interactions in at-risk emergency department patients. Acad Emerg Med 2002;9(11):1162-7.
  7. De Paepe P, Petrovic M, Outtier L, Van Maele G, Buylaert W. Drug interactions and adverse drug reactions in the older patients admitted to the emergency department. Acta Clin Belg 2013;68(1):15-21.
  8. Guy Jr GP, Zhang K, Bohm MK, et al. Vital signs: changes in opioid prescribing in the United States, 2006-2015. MMWR Morb Mortal Wkly Rep 2017;66(26):697-704.
  9. Kango Gopal G, Hewton C, Pazhvoor SK. Myoclonus associated with concomitant ciprofloxacin and oxycodone in an older patient. Br J Clin Pharmacol 2014;77(5): 906-7.
  10. Grill MF, Maganti RK. Neurotoxic effects associated with antibiotic use: management considerations. Br J Clin Pharmacol 2011;72(3):381-93.
  11. Lexicomp Inc. Lexidrugs; 2018 [Accessed September 10th, 2018].
  12. Truven Health Analytics. Metronidazole and Ciprofloxacin drug interaction results.; 2018, Accessed date: 10 September 2018.
  13. Liamis G, Milionis H, Elisaf M. A review of drug-induced hyponatremia. Am J Kidney Dis 2008;52(1):144-53.
  14. Lapi F, Azoulay L, Yin H, Nessim SJ, Suissa S. Concurrent use of diuretics, angiotensin converting enzyme inhibitors, and angiotensin receptor blockers with non-steroidal anti-inflammatory drugs and risk of acute kidney injury: Nested case-control study. BMJ 2013;346:e8525 (Clinical research ed).
  15. Dreischulte T, Morales DR, Bell S, Guthrie B. Combined use of nonsteroidal anti- inflammatory drugs with diuretics and/or renin-angiotensin system inhibitors in the community increases the risk of acute kidney injury. Kidney Int 2015;88(2): 396-403.
  16. FDA. drug safety Communication: FDA warns about serious risks and death when combining opioid pain or cough medicines with benzodiazepines; requires its stron- gest warning [press release]. Augsut 31st, 2016; 2016.
  17. Friedman BW, Irizarry E, Solorzano C, et al. Diazepam is no better than placebo when added to naproxen for acute low back pain. Ann Emerg Med 2017;70(2):169-76 [e161].
  18. Friedman BW, Dym AA, Davitt M, et al. Naproxen with cyclobenzaprine, oxycodone/ acetaminophen, or placebo for treating acute low Back pain: a randomized clinical trial. JAMA 2015;314(15):1572-80.
  19. Lexicomp Inc. Ibuprofen and aspirin drug interaction results; 2018 [Accessed Sep- tember 10th, 2018].
  20. Langhorne, PA: Insight Pharmaceuticals Corp. Kondremul (mineral oil) [prescribing information]; 2013.
  21. Lexicomp Inc. Sertraline and ibuprofen drug interaction results; 2018 [Accessed Sep-

    tember 10th, 2018].

    Persell SD, Osborn CY, Richard R, Skripkauskas S, Wolf MS. limited health literacy is a barrier to medication reconciliation in ambulatory care. J Gen Intern Med 2007;22 (11):1523-6.

Leave a Reply

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