To Name :
To Email :
From Name :
From Email :
Comments :

Evidence-Based Reviews

Drug-drug interactions: Avoid serious adverse events with mood stabilizers

In clinical practice, prescribed drug combinations are often uncontrolled experiments, with unknown potential for toxic effects.

Vol. 4, No. 5 / May 2005

Drug-drug interactions (DDIs) can be viewed as physiologic combat wherein a “perpetrator” drug affects a “victim” drug’s pharmacokinetics or pharmacodynamics. Your challenge is to deter that interaction in patients taking two or more medications.

This article—first in a series—discusses polypharmacy risk factors that increase the likelihood of detrimental DDIs, then focuses on DDIs in patients taking mood stabilizers for bipolar disorder. We also offer practical tips to reduce DDI risk. Future articles will discuss DDI risks with antidepressants, antipsychotics, and anxiolytics.

To predict DDIs, you need to know psychotropics’ mechanism of action, metabolism, and effects on cytochrome P-450 (CYP) enzymes. Our discussion is not exhaustive because the data base is massive and new interactions continue to be discovered. Our aim is to equip you to anticipate and prevent DDIs when prescribing.


An adverse event (AE) is any undesirable experience that occurs when a patient uses a medical product, whether or not the product caused the event. The FDA says an “undesirable experience” may be:

  • an unfavorable and unintended symptom or sign
  • an abnormal lab or radiographic finding
  • a disease that is temporarily associated with the medical product.

A temporal relationship is all that is required, although preexisting conditions and events clearly related to other causes are not usually considered adverse events.

An AE becomes “serious” (an SAE) when its duration, intensity, and/or frequency leads to death, a life-threatening condition, initial or prolonged hospitalization, disability, or congenital anomaly. Reporting is voluntary, but we strongly recommend that you report all SAEs to the FDA.

These definitions can help you confirm that a patient has experienced an SAE, but the task becomes more complicated when you try to attribute an SAE to a drug interaction. In the absence of an FDA definition, we assert that DDIs are responsible for SAEs when a perpetrator drug affects the pharmacokinetics or pharmacodynamics of a victim drug and exacerbates a known untoward event of the victim drug (Box 1).1-5 Which drug is the perpetrator and which is the victim is not always clear, and sometimes a medication—such as carbamazepine—can be both at once.

Box 1

Drug-drug interactions: Taking a toll

More than 100,000 possible detrimental DDIs have been documented in medical literature and pharmaceutical company data. This number is likely to grow with increased scrutiny, as <10% of adverse events from DDIs are reported.

DDIs cause morbidity, mortality, and increased health care costs. More than 106,000 Americans die each year from properly prescribed, correctly taken medications. Polypharmacy is associated with extended hospital stays, and using >6 drugs is an independent predictor of death. DDIs contribute to the cause of death in acute overdoses and can be responsible for false-positive suicide diagnoses.

In clinical practice, DDI-associated toxicity may be mistaken for a new disease process, or a disease may be incorrectly perceived as progressing when a medication is rendered ineffective.

Source: References 1-5


Individuals with psychiatric illnesses are at particular risk for DDIs (Box 2). Patients seen by psychiatrists, for example, are six times more likely than patients seen by primary care physicians to be taking multiple medications.6

Polypharmacy increases the risk of adverse events, nonadherence, medication errors, and drug interactions.7 FDA’s MedWatch Web site lists more than 630 DDI warnings.8 The more medications a patient is taking, the greater the risk for detrimental DDIs and cumulative toxicity,9 which often lead to DDI-induced AEs.10

A study of DDIs in 5,125 mostly older outpatients11 found that:

  • 1,594 (31%) had at least one interacting drug combination (average 1.6)
  • subjects with one or more DDIs were taking an average 8.1 drugs, compared with 5.2 drugs in those without DDIs—a significant difference
  • 155 (3%) had interactions of “major clinical significance.”

‘Uncontrolled experiments.’ Drug combinations often are “uncontrolled experiments” with unknown potential for toxic effects.12 Studies have linked polypharmacy and DDIs as well as DDIs and AEs:

  • Although drug interactions are responsible for only 3.8% of emergency department visits, patients with DDIs are usually admitted to the hospital.13
  • Preventable drug interactions cause approximately one-third of all AEs in hospitalized patients and account for one-half of all AE costs.14

DDI risk is increasing over time as the number of medications used to treat psychiatric patients has grown. For example, 3.3% of patients discharged between 1974 and 1979 from the National Institute of Mental Health Biological Psychiatry Branch were taking 3 or more medications, compared with more than 40% of patients discharged between 1990 and 1995—a 12-fold increase.15

Box 2

Psychiatric patients: High risk for DDIs

Symptom-based prescribing. Patients with psychiatric illness are often prescribed >1 medication to manage symptoms and signs, rather than a single medication targeting a specific psychiatric disorder.

Multiple prescribers. Patients with anxiety and depressive disorders may see multiple providers, which increases the risk for polypharmacy, drug-drug interactions, and adverse events.

Medical comorbidity. Persons with psychiatric illness are at increased risk for concomitant medical illness, and persons with medical illness are at increased risk for psychiatric illness.

Psychiatric comorbidity. Persons with one psychiatric illness are at increased risk for other psychiatric illnesses.

Source: Adapted from reference 6.


Use the acronym “LISTEN” (Table 1) to minimize DDI risk in patients taking combination therapies.16 The 6 steps in LISTEN can help you determine which drug or drugs you may discontinue before adding another.

We also recommend that you monitor therapeutic and toxic effects by checking serum drug levels, especially for drugs with a low therapeutic index. Lithium, for example, requires close mentoring of plasma concentration every 2 to 6 months and during dosage adjustments to avoid toxicity.17 Therapeutic drug monitoring has been shown to prevent adverse events from DDIs.16 For added safety, encourage patents to purchase all medications at one pharmacy and to enroll in that pharmacy’s DDI monitoring program.18

Keep in mind that systemic conditions may require a dosage change:

  • Increased volume of distribution, as in patients who gain weight or total water volume, requires higher doses to maintain a constant therapeutic effect.
  • Reduced clearance, as in patients with decreased renal or hepatic function, will likely require lower doses to prevent toxicity.19

Table 1

LISTEN: 6 tips to minimize DDI risk


Listeach drug’s name and dosage in the patient’s chart and in a note given to the patient.


Each drug should have a clear indication and well-defined therapeutic goal; discontinue any drug not achieving its goal


Make the regimen as simple as possible, with once- or twice-daily dosing.


When possible, treat multiple symptoms with a single drug, rather than multiple symptoms with multiple drugs


Educatepatients about polypharmacy, DDIs, and adverse events; assess all medications—including vitamins, minerals, herbs, dietary supplements, nonprescription products—and address potential DDIs


Avoid prescribing medications with a narrow therapeutic window.


Diagnoses of schizophrenia, anxiety disorders, and affective disorders are major risk factors for polypharmacy.20 DDIs are a particular concern in patients with bipolar disorder, given their complex treatment regimens.21

Interactions occur with the most commonly prescribed bipolar medications, including lithium and anticonvulsants (Table 2).17.21-25 Although atypical antipsychotics are also considered mood stabilizers in bipolar disorder, we will discuss their potential DDIs in a future article.

Table 2

Some drug-drug interactions with mood stabilizers

Mood stabilizer

Drug interactions


↑plasma clomipramine, phenytoin, primidone

↑risk of neurotoxic side effects and confusional states with lithium

Alters thyroid function with anticonvulsants

↓anticoagulant concentrations and↑bleeding risk

↓oral contraceptive reliability; can cause false-negative pregnancy tests

↑metabolism and may ↓efficacy of cancer chemotherapy (docetaxel, estrogens, paclitaxel, progesterone, cyclophosphamide)

↑aprepitant, granisetron metabolism and ↓efficacy

↑glipizide, tolbutamide metabolism


NSAIDs (ibuprofen, indomethacin, piroxicam) and COX-2 inhibitors ↑plasma lithium

ACE inhibitors ↑plasma lithium

Calcium channel blockers and carbamazepine ↑lithium neurotoxicity

SSRIs ↑diarrhea, confusion, tremor, dizziness, and agitation

Acetazolamide, urea, xanthine preparations, alkalinizing agents such as sodium bicarbonate ↓plasma lithium

Metronidazole ↑lithium toxicity

Encephalopathic syndrome possible with haloperidol


↑concentration of carbamazepine’s epoxide metabolite

Carbamazepine, phenytoin, phenobarbital ↓plasma lamotrigine 40% to 50%

↑plasma sertraline

↓plasma valproic acid 25%; valproic acid doubles plasma lamotrigine and ↑rash risk


↑valproic acid concentrations 11%; valproic acid ↓plasma topiramate 14%

↑plasma phenytoin up to 25%; phenytoin, carbamazepine ↓plasma topiramate by 40% to 48%

↓digoxin bioavailability

↓oral contraceptive efficacy

Valproic acid

↑plasma phenobarbital, primidone

↓phenytoin clearance, volume distribution and ↑breakthrough seizure risk

↑serum concentration of antiepileptics, such as lamotrigine; absence status possible with clonazepam

↑= Increases ↓= Decreases

ACE = angiotensin-converting enzyme; COX = cyclooxygenase

NSAIDs = nonsteroidal anti-inflammatory drugs; SSRIs = selective serotonin reuptake inhibitors

Source: References 17, 21-25


Lithium is excreted via the kidneys, so be cautious when using lithium in patients taking diuretics.17,22 Drugs that can lower serum lithium concentrations by increasing urinary lithium excretion include acetazolamide, urea, xanthine preparations, and alkalinizing agents such as sodium bicarbonate.17

Combining lithium with selective serotonin reuptake inhibitors can cause diarrhea, confusion, tremor, dizziness, and agitation.17 An encephalopathic syndrome has occurred in a few patients treated with lithium plus haloperidol.

Monitor lithium levels closely when bipolar patients start or stop nonsteroidal anti-inflammatory drugs (NSAIDs). Nonprescription ibuprofen can cause serious and even life-threatening serum lithium elevations by affecting lithium’s rate of tubular reabsorption.26 Indomethacin, piroxicam, and selective cyclooxygenase-2 (COX-2) inhibitors also increase plasma lithium concentrations.25

For patients taking lithium with heart drugs, angiotensin-converting enzyme (ACE) inhibitors may increase plasma lithium levels,17 and calcium channel blockers may increase the neurotoxicity risk.17,22 Using the anti-infective metronidazole with lithium may provoke lithium toxicity.


Drugs that affect the expression of hepatic enzymes—especially glucuronosyltransferase—may increase clearance of valproic acid and its derivatives. Phenytoin, carbamazepine, or phenobarbital, for example, can double valproic acid clearance.

On the other hand, drugs that inhibit CYP-450 (such as antidepressants) have little effect on valproic acid concentration. Valproate can decrease plasma clearance of amitriptyline, so consider monitoring this tricyclic’s blood levels in patients also taking valproate.17

Because valproic acid can increase serum phenobarbital, monitor barbiturate concentrations when using these two drugs. A similar interaction occurs with primidone, which is metabolized into a barbiturate. Breakthrough seizures may occur with phenytoin, as valproic acid can reduce phenytoin clearance and apparent volume distribution by 25%.22

Using valproic acid with clonazepam may produce absence status in patients with a history of absence-type seizures.17 Valproic acid also displaces diazepam from its plasma albumin binding sites and inhibits its metabolism.

Concomitant use of valproic acid can increase serum concentrations of other antiepileptic drugs. For example, lamotrigine levels may double,24 and felbamate’s peak concentration may increase and require dosage reduction. Valproic acid may also interact with nonpsychiatric medications:

  • Subtherapeutic valproic acid levels have been reported when co-administered with the antibiotic meropenem.
  • In patients with HIV infection, valproic acid can decrease clearance of the antiretroviral zidovudine by 38%.
  • Patients receiving rifampin for tuberculosis may need a dosage adjustment, as oral rifampin’s clearance can increase 40% with concomitant valproic acid.


Metabolized by CYP 3A4, carbamazepine may induce its own metabolism as well as the CYP 3A4 isoenzyme. Therefore inhibitors and inducers of CYP 3A4 may affect carbamazepine plasma levels.

Carbamazepine can increase plasma levels of other psychotropics including clomipramine, phenytoin, and primidone.17,22 When used with lithium, it may increase the risk of neurotoxic side effects and confusion.23 It can alter thyroid function when used with other anticonvulsants.

For bipolar patients with diabetes, carbamazepine can cause hyperglycemia by inducing the metabolism of oral sulfonylureas such as glipizide and tolbutamide. In women, carbamazepine decreases the reliability of oral contraceptives17 and can cause false-negative pregnancy tests.24

For cancer patients, concurrent carbamazepine may induce metabolism of chemotherapy drugs such as docetaxel, estrogens, paclitaxel, progesterone, and cyclophosphamide, decreasing their efficacy.21 It can increase metabolism of aprepitant and granisetron—used to treat chemotherapy-related nausea—reducing plasma concentrations and possibly efficacy. Carbamazepine’s additive dopamine blockade can increase the risk of extrapyramidal symptoms when used with docetaxel or the antiemetic/antivertigo agents chlorpromazine, metoclopramide, or prochlorperazine.

Did you miss this content?
An under-recognized epidemic of elder abuse needs your awareness and action