Syeda Quadri, MD

Patients with coronavirus disease 2019 (COVID-19) who are at high risk for serious illness are also at increased risk for drug-drug interactions (DDIs) that can have a negative impact on outcomes. In elucidating this issue, this article will discuss the following main points:

1. The reasons why patients at high risk for serious illness from COVID-19 are also at high risk for DDIs.

2. The extrapolation of results from pharmacoepidemiological studies to the population at risk for serious COVID-19 illness.

3. The mechanisms underlying DDIs including their frequency and their potential complexity, and how DDIs can present clinically.

4. The methods for preventing or mitigating DDIs.

5. An introduction to University of Liverpool drug interaction checker as a tool to reduce the risk of adverse DDIs when treating patients for COVID.

Why patients at high risk for serious COVID-19 illness are also at high risk for DDIs

Patients with multiple comorbid illnesses are likely to be taking multiple medications and seeing more than 1 prescriber for their health care needs. Obviously, the more medications a patient is taking, the more likely they are to experience a DDI, and the more likely that DDI will be complex, as illustrated by the case described in 

 This situation can lead to an increased probability of unplanned and adverse DDIs, which can contribute to or directly cause a poor outcome up to and including death.

There are no published studies looking at the number of medications taken by those at high-risk of a severe form of COVID-19. However, pharmacoepidemiological studies about the frequency and complexity of multiple medication use (MMU) in the United States and other industrialized countries have been published, and these data can be generalized to patients who are at high-risk for serious COVID-19 illness.

Findings from pharmacoepidemiological studies

A pharmacoepidemiological study by Preskorn et al

 examined MMU and the complexity of drug regimes in an adult outpatient population of United States veterans treated in Region 15 of the Veterans Affairs (VA) Integrated Service Network (VISN 15). The objective was to examine the extent and nature of MMU in relation to several factors, including patient age and number of prescribers treating the patient. The outcomes were the complexity and uniqueness of the drug regimens in relation to the numbers of different medications patients were taking. Medications counted in the study were limited to those that acted systemically or gastrointestinally (SG drugs) because these agents are of particular concern given their potential to interact with each other.

The drug regimens of 5003 patients were assessed, and the researchers found that a total of 394 different SG drugs were prescribed. Only 88 (22%) of these drugs were used in at least 1% of patients (ie, 50 or more patients). The remaining 306 drugs were used in fewer than 1% of the patients. Of the 5003 patients, 80% had current prescriptions for at least 2 SG drugs and 38% were receiving 5 or more drugs. 

There were 3819 different drug regimens used in this population, 3553 of which were used in only 1 patient. Thus, 71% of the patients were receiving a unique drug regimen (ie, only 1 patient in this population of 5003 individuals was taking that same total regimen of drugs regardless of dose or schedule). If dosing and scheduling were also considered, even more patients would have been considered unique with regard to their drug treatment. These results indicate that treatment for most of these patients was quite individualized—but based on what rationale and experience?

 illustrates the rate at which the prevalence of a specified drug combination decreases when the number of drugs in the combination increases. This figure shows the most common 4 drug combinationused in the population in this VA study. Of the 5003 patients, 26.5% received aspirin and 10.6% received furosemide, but only 4.2% received both aspirin and furosemide. Fewer than 2% of the patients took a combination of aspirin, furosemide, and digoxin (a third SG drug commonly used in this population). The percentage decreased further, to fewer than 1% of patients who were receiving a combination of those 3 SG drugs plus lisinopril, which was overall the second most commonly prescribed drug to this population of outpatients. Of the 28 patients taking this 4-drug combination, only 1 patient received only these 4 drugs, 2 others received the same regimen plus a fifth drug (simvastatin, another cardiovascular-related agent), and the remaining 25 were on unique drug regimens of 5 or more drugs.

The Slone survey, published by Kaufman et al4 in 2002, was also conducted in the United States. It obtained self-reported information about prescription, over the counter, and herbal drugs from adult outpatients aged 18 years and older. This general survey of 2590 adults found that 7% had taken 5 or more prescription medications in the previous week. Women aged 65 years and older were the highest consumers of medications. In this population, 94% took at least 1 medication, 57% took 5 or more medications, and 12% took 10 or more medications. Rates of prescription drug use were similarly high in men and women in the oldest age group.

The findings from the VISN 15 study were similar to those of pharmacoepidemiological studies done in Europe. A series of studies

 looked at reimbursement for prescription medications on an average day in Fünen County, Denmark in 1994.5 Bjerrum et al5 found that 8.7% (SD 0.2%) of the patients were taking to 2 to 4 drugs and 1.2% were taking 5 or more drugs. The proportion of patients receiving 5 or more medications increased with age up to 90 years, so that two-thirds of patients aged 70 years and older were receiving 5 or more medications.

 found that 26,337 elderly patients aged 70 years or older were receiving 21,293 different drug combinations and that the 10 most prevalent combinations were used in only 2.7% of the patients. In the third study, Bjerrum et al7examined drug combinations in 5443 patients aged 16 years or older who were receiving 5 drugs. They found a total of 3890 different drug combinations, of which the 10 most prevalent combinations occurred in only 3% of patients.

A pharmacoepidemiology study by Valent et al

 assessed the prevalence of MMU in the general population of Udine, Italy, during 2017. The results of this study were comparable to those of the VISN 15 study: 63.7% of the general population were prescribed at least 1 medication during the year. Compared with the rest of the adult population, MMU was more common among the elderly, especially those aged 80 years or older, with 31.7% of those aged 65 years or older prescribed 5 or more medications at least once during the study year.

Taken together, these studies demonstrate that the frequency and complexity of MMU are similar across different countries. 

 The fact that there are numerous medications available to treat any particular condition (ie, drug class) and that there are multiple options within each of most specific pharmacological classes of drugs (eg, multiple beta blockers and serotonin reuptake inhibitors) helps explain the multitude of different medication combinations that may be taken by an elderly patient or a patient with multiple comorbid conditions. Based on these studies, many patients are receiving a unique drug regimen, where “unique” refers to the total specific drug entities a patient is receiving, regardless of dose, formulation or administration schedule. In other words, no other patient in the population sampled will be taking the same regimen. That, in turn, means that no single prescriber is likely to have extensive clinical experience with even a small fraction of the multiple total drug regimens their patients are taking.

A DDI, by definition, occurs when the presence of a co-prescribed drug (the perpetrator) alters the nature, magnitude, or duration of effect of a given dose of another drug (the victim).

To understand DDIs, it is important to know how medications act, because their action determines whether and how they will interact pharmacodynamically with other medications. It is also important to understand their pharmacokinetics (ie, the mechanisms underlying their absorption into the body, distribution throughout the body, their metabolism, and their elimination from the body), because that is the second way that drugs can interact. As previously noted, medications that act systemically or gastrointestinally are of most concern because they have the greatest potential to interact. Therefore, identifying frequently used combinations of SG medications is the first step in addressing potentially hazardous combinations.

Drugs can interact in 2 ways: pharmacodynamically and/or pharmacokinetically. The word “pharmacodynamics” refers to the action of a drug on a specific site, and the biochemical and physiological effect it produces via that site of action. It is the body’s response to the drug. The word “pharmacokinetics” refers to the movement of a drug through the body: its absorption into the body from the site of administration (usually oral) into the central compartment (ie, blood), its distribution from the central compartment to peripheral compartments (ie, various organs including the brain), its metabolism (ie, its movement through the pathways of biotransformation), and finally its excretion from the body (usually via the kidneys).

DDIs can be therapeutic or adverse; they can be planned or unplanned. A planned DDI occurs when the prescriber knows that the addition of a medication to the existing treatment regime (ie, a combination regime) will enhance the efficacy or tolerability of both. An example of a planned DDI is adding an adjunctive medication such as bupropion to an SSRI to boost the antidepressant effect experienced by a patient.In other words, these DDIs are intentionally used to benefit patients.

To understand and avoid untoward DDIs, it is important to be familiar with equation 1 and equation 2 

 The 3 variables in equation 1 determine the effect a drug will produce in a patient.First, the drug must work on a site of action that is capable of producing the effect observed. The site of action for most drugs is a receptor, transporter (uptake pump), enzyme, or ion channel. By binding to the target(s), the drug can alter the target’s functional status and thus alter human physiology. The second variable is the drug’s pharmacokinetics, which is the ability of the drug to move through the body (ie its absorption, distribution, metabolism, and eventual elimination from the body). The third variable involves the interindividual differences among patients, which can shift the dose-response curve, making patients either more or less sensitive to the effect of the drug.

 illustrates that drug concentration is a function of the dosing rate a patient is taking, rather than what has been prescribed, in relation to their ability to clear the drug.

 The time course of potential DDIs varies depending on how long the drug(s) (or their effects) persist in the body. Hence, the potential for an interaction may last for days to even months after a medication has been discontinued. For example, a medication like fluoxetine can remain in the system for many weeks. Specifically, it can last 5 weeks or more depending on the half-life of the parent drug and its essentially equally active major metabolite (norfluoxetine) in a given patient.

A drug may also have induced a change in the body that may persist for weeks after the drug has been discontinued. Examples of such drugs are those that induce (turn on the promoter genes for) cytochrome P450 drug metabolizing enzymes (eg, carbamazepine) or drugs that covalently bind to and deactivate an enzyme (eg, most monoamine oxidase inhibitors).

Hence, it is important to elicit a detailed medication history, not only of the medications the patient is currently taking but also of medications the patient has taken in the past few months.

? DDIs can present in a myriad of ways ranging from common everyday problems (which are more common but more easily missed outcomes) to catastrophic adverse effects (which are much rarer but also more easily detectable outcomes). Catastrophic adverse events due to DDIs can include sudden death, seizures, cardiac rhythm disturbances, serotonin syndrome, malignant hypertension, neuroleptic malignant syndrome, and delirium to name a few. In terms of common everyday problems, DDIs can make a patient appear to have poor tolerability for a medication(s) or to be nonresponsive (ie, lack of efficacy) to the medication(s). They may also cause symptoms that mimic the worsening of a preexisting illness or the emergence of a new illness.

The pitfall with worsening of symptoms is that it may lead the prescriber to stop a potentially efficacious treatment due to the erroneous belief that the medication that has been added is itself causing the adverse effect, rather than identifying the DDI which is the true culprit leading to the untoward outcome. If the potential for a DDI is recognized, then the prescriber can make whatever adjustments are needed (eg, lowering the dose or switching to a different agent within a class of medications, such as changing from 1 SSRI to another that does not inhibit a specific cytochrome P450 enzyme) so that the patient can benefit from the efficacy of the medication without having the adverse effect(s). Alternatively, the emergence of new symptoms as a result of a DDI may lead to a misdiagnosis, which may result in the addition of new medications to treat the apparent worsening of an existing illness or the apparent emergence of a new illness.

Example of a patient receiving multiple medications

 shows an example of a 4-medication regimen that a patient was receiving from 4 different prescribers. This patient was one of the 5003 in the Preskorn et al VISN 15 study

 previously described. This patient was taking codeine for pain, erythromycin for an infection, paroxetine for depression, and metoprolol for hypertension. Parenthetically, the pharmacoepidemiological studies reviewed in this paper documented that patients receiving such multiple medication regimes are common in clinical practice. The clinically relevant questions are: Do 2 or more of these drugs interact? If so, how does that interaction present and what might 1 or more of the 4 prescribers do in reaction to that presentation?

 illustrates the known potential interactions among the 4 medications this patient was taking. Codeine is an inactive prodrug that must be converted by CYP 2D6 to morphine to produce analgesia. Metoprolol is a beta-blocker whose clearance is principally dependent on CYP 2D6 mediated biotransformation to polar metabolites that can subsequently be eliminated via the kidneys. Paroxetine substantially to completely inhibits CYP 2D6 at the usual dose needed to produce an antidepressant response. While paroxetine is metabolized by CYP 2D6 at low concentrations, it saturates this enzyme under usual dosing conditions. At higher concentrations, paroxetine is likely dependent on CYP 3A mediated biotransformation for its elimination. CYP 3A is substantially inhibited by erythromycin under usual dosing conditions. The inhibition of CYP 3A by erythromycin will produce an increased accumulation of paroxetine, which in turn would produce more inhibition of CYP 2D6, which in turn would lead to less conversion of codeine to morphine and more accumulation of metoprolol. This is an example of a complex or multiple DDI. Such DDIs can cause many adverse effects.

 also illustrates the complex and hidden way in which such a DDI can present. Due to the inability to convert codeine to morphine, the patient will have less than optimal pain control, which may be construed as the lack of efficacy of the codeine, opiate-seeking behavior by the patient, and/or worsening depression. The increased accumulation of metoprolol can cause hypotension and the patient may subjectively complain of increased fatigue, which can again be misconstrued as a worsening of depression. The increased accumulation of paroxetine can cause more insomnia, decreased emotional reactivity, and decreased libido, which can also look like worsening of depression. If the prescriber observes this increase in depression-like symptoms and concludes that the depressive episode is worsening, they may increase the dose of paroxetine, further worsening the problem. This scenario illustrates the hidden way in which a DDI can present and how it can perhaps lead the clinician (and their patient) dangerously astray.

Preventing or mitigating DDIs in patients with COVID-19

Why it is important to consider DDIs in those diagnosed with COVID-19.

 The discussion of multiple medications and DDIs presented in the preceding section is directly applicable to patients who are at high risk for having serious COVID-19 illness, because these individuals are likely be taking multiple medications and unique combinations of those medications. On top of their regular medication regimen, these patients will likely be prescribed additional medications to treat their COVID-19 illness, which will further increase the risk of experiencing DDIs. Therefore, it is important for prescribers to be aware of and understand the implications of: (1) the extent and complexity of multiple medication use, (2) how common it is for patients to be on unique medication regimens, and (3) how to help prevent immediate and long-term adverse effects, unexpected outcomes, and even mortality.

In addition, many of the emerging treatments for COVID-19 have the potential to interact with all medications the patient is taking. If the patient is in the intensive care unit, then the treating physician may simplify the medication regimen by stopping some or even all of the previous medications. If that approach is taken, the important questions are: Which medications can be abruptly and safely stopped and why? Which medications may need to be tapered to avoid withdrawal or rebound phenomena? The answer to these questions depends on the specific agent and the indication.

When patients with COVID-19 are hospitalized but are not in the intensive care unit, their usual medications may be continued as these patients may require more supportive treatment (eg, intravenous fluids) rather than the addition of specific anti-COVID-19 medications. Determination would be made on a case-by-case basis considering both the status of the patient and their medication list.

The challenge for prescribers and health care professionals in general is the rapid expansion of our knowledge about COVID-19 and its treatment, as illustrated by the exponential growth in clinical trials and clinical experience with this illness. For example, at the end of August 2020, the US Food and Drug Administration 

the indication for remdesivir to all patients hospitalized with COVID-19.

 When considering the wider use of this drug, it is important to note that it has the potential to interact with other drugs. While the usefulness of chloroquine phosphate or hydroxychloroquine sulfate remains uncertain, for example, some providers may consider using it in combination with remdesivir. If so, the prescriber should be aware of the potential for interaction between these 2 treatments. The fact sheet concerning remdesivir issued as part of the Emergency Use Authorization from the FDA

 that “In vitro, remdesivir is a substrate for drug metabolizing enzymes CYP2C8, CYP2D6, and CYP3A4, and is a substrate for Organic Anion Transporting Polypeptides 1B1 (OATP1B1) and P-glycoprotein (P-gp) transporters. In vitro, remdesivir is an inhibitor of CYP3A4, OATP1B1, OATP1B3, BSEP, MRP4, and NTCP.” With regard to DDIs, the fact sheet states “Drug-drug interaction trials of Veklury (remdesivir) and other concomitant medications have not been conducted in humans. Due to antagonism observed in vitro, concomitant use of Veklury with chloroquine phosphate or hydroxychloroquine sulfate is not recommended.”

Caveats about specific psychiatric medications. 

Caution should be considered with fluoxetine and long-acting depot antipsychotics when treating patients who have COVID-19 and are on multiple medications , because of the long time they persist in the body even after their use is discontinued. Certain anticonvulsants such as carbamazepine also should be a red flag because of their ability to induce specific CYP enzymes such as CYP 3A, with the inducted state persisting for at least a few weeks after the inducer has been stopped. For the same reason, it is important to determine if these types of medications had recently been discontinued before patients were hospitalized, because their effects may persist for days to weeks after the drugs are discontinued.

If some of the patient’s usual medications cannot be stopped because of potential adverse health consequences or because of the potential for clinically meaningful withdrawal symptoms, then the physician or health care provider has to consider how these medications can interact with those being added to treat the COVID-19 illness. In this regard, the University of Liverpool provides access to a tool to minimize DDIs when prescribing medications for COVID-19. This is a free drug-interaction checker that prescribers can access at 

 This website lists all of the medications currently being used to treat COVID-19 and potential DDIs with other medications. It is updated regularly as new treatment regimens for COVID-19 emerge. The checker is broken down in a table format showing interactions according to drug class.

This article has highlighted numerous important clinical issues regarding DDIs. First, clinicians must recognize the frequency and complexity of MMU, particularly in older patients and those with high levels of comorbid illnesses. The article also discussed how and why drugs may interact and how such interactions can present clinically. This information is important for patients with serious forms of COVID-19 because they are a population likely to be receiving MMU before they become ill with COVID-19 and then are likely be treated with additional medicines for COVID-19 illness. Factors to consider and general actions that can be taken to prevent or mitigate untoward DDIs have been discussed. The University of Liverpool website and tool was discussed so that readers can better understand how COVID-19 treatments may interact with other medications. Comments have also been provided concerning issues related to specific psychiatric medications and treatments currently associated with COVID-19.

This article was originally published in the 

and has been adapted here with permission from Lippincott Williams & Wilkins. The original version can be found online at 

is professor in the department of Psychiatry and Behavioral Sciences. The author notes that he has received grants/research support from or has served as a consultant, on the advisory board, or on the speaker’s bureau for Alkermes, BioXcel, Eisai, Janssen, National Institute of Mental Health, Sunovion, and Usona Institute. All clinical trial and study contracts were with and payments made to The University of Kansas Medical Center Research Institute, a research institute affiliated with The University of Kansas School of Medicine-Wichita. 

is a senior and immediate past chief resident psychiatrist in the department of psychiatry at the University of Kansas School of Medicine in Wichita, KS.

1. Preskorn SH, Silkey B. Multiple medications, multiple considerations. 

2. Preskorn SH, Silkey B, Shah R, et al. Complexity of medication use in the Veterans Affairs Healthcare System: Part I: Outpatient use in relation to age and number of prescribers. 

3. Silkey B, Preskorn SH, Golbeck A, et al. Complexity of medication use in the Veterans Affairs healthcare system: Part II. Antidepressant use among younger and older outpatients. 

4. Kaufman DW, Kelly JP, Rosenberg L, et al. Recent patterns of medication use in the ambulatory adult population of the United States: The Slone Survey. 

5. Bjerrum L, Rosholm JU, Hallas J, et al. Methods for estimating the occurrence of polypharmacy by means of a prescription database. 

6. Rosholm J-U, Bjerrum L, Hallas J, et al. Polypharmacy and the risk of drug-drug interactions among Danish elderly: a prescription database study. 

7. Bjerrum L, Søgaard J, Hallas J, et al. Polypharmacy: correlations with sex, age, and drug regimen: a prescription database study. 

8. Valent F, Polypharmacy in the general population of a Northern Italian area: analysis of administrative data. Ann 1st Super Sanita. 2019;55:233-239.

9. Preskorn SH. Drug-Drug Interactions With An Emphasis On Psychiatric Medications. Professional Communications; 2018.

10. U.S. Food and Drug Administration (FDA). COVID-19 Update: FDA Broadens Emergency Use Authorization for Veklury (remdesivir) to Include All Hospitalized Patients for Treatment of COVID-19. Silver Spring, MD: FDA Press Release; August 28, 2020. Accessed October 13, 2020. 

https://www.fda.gov/news-events/press-announcements/covid-19-update-fda-broadens-emergency-use-authorization-veklury-remdesivir-include-all-hospitalized

11. Fact sheet for health care providers: emergency use authorization (EUA) of Veklury® (remdesivir). Foster City, CA: Gilead Sciences; August 2020. 

Accessed October 13, 2020. https://www.gilead.com/-/media/files/pdfs/remdesivir/eua-fact-sheet-for-hcps.pdf

12. Pinkowski J. Drug-drug interactions could imperil COVID-19 treatment. Medscape May 10, 2020. Accessed October 13, 2020. 

https://www.medscape.com/viewarticle/930265

13. University of Liverpool. Interaction Checker: COVID-19 Drug Interactions. University of Liverpool; 2020. Accessed October 13, 2020. 

Articles

Drug-drug interactions are complicating COVID-19 treatment. Read what you need to know to keep yourself and your patients safe before and during COVID infection. 

Syeda Quadri, MD

Articles

Why Are Patients With COVID-19 at Risk for Drug-Drug Interactions?