© The Author 2007. Published by Oxford University Press.
ARTICLES |
Potential Drug Interactions and Duplicate Prescriptions Among Cancer Patients
Affiliations of authors: Departments of Medical Oncology and Hematology (RPR, IFT, MKK) and Biostatistics (LW), Princess Margaret Hospital, Toronto, ON, Canada; Dendrix, Ltd, Sao Paulo, Brazil (EDS); Department of Public Health Sciences, University of Toronto, Toronto, ON, Canada (NAT)
Correspondence to: Monika K. Krzyzanowska, MD, MPH, Department of Medical Oncology and Hematology, Princess Margaret Hospital, 610 University Ave, Ste 5-227, Toronto, ON, M5G 2M9, Canada (e-mail: monika.krzyzanowska{at}uhn.on.ca).
 |
ABSTRACT |
|---|
 Top Abstract Context and CaveatsPatients and MethodsResultsDiscussionReferencesNotes |
|---|
Background: Cancer patients receive numerous medications, including antineoplastic agents, drugs for supportive care, and medications for comorbid illnesses. Therefore, they are at risk for drug interactions and duplicate prescribing.
Methods: A questionnaire eliciting information on demographics and medications taken in the previous 4 weeks was given to adult outpatients receiving systemic anticancer therapy for solid tumors. The Drug Interaction Facts software, version 4.0, was used to identify potential drug interactions and to classify them by level of severity (major, moderate, or minor) and the strength of scientific evidence for them (using categories [1–5] of decreasing certainty). Summary statistics and logistic regression were used to analyze the data. All statistical tests were two-sided.
Results: The survey was completed by 405 patients. We observed 276 potential drug interactions, and at least one potential interaction was identified in 109 patients (27%; 95% confidence interval [CI] = 23% to 31%). Of the potential interactions, 25 (9%) were classified as major and 211 (77%) as moderate. Nearly half (49%) of potential interactions were supported by level 1 or 2 scientific evidence. Most potential drug interactions (87%) involved non-anticancer agents such as warfarin, antihypertensive drugs, corticosteroids, and anticonvulsants, but some (n = 36, 13%) involved antineoplastic agents. In multivariable analysis, increased risk of receiving drug combinations in which there were potential drug interactions was associated with receipt of increasing numbers of drugs (odds ratio [OR] = 1.4 per additional drug, 95% CI = 1.26 to 1.58, P<.001 from the Wald chi-square test), type of medication (drugs to treat comorbid conditions versus supportive care medications only; OR = 8.6, 95% CI = 2.9 to 25, P<.001), and the presence of brain tumors. Thirty-two (8%) patients were exposed to duplicate medications, most often corticosteroids, proton pump inhibitors, or benzodiazepines.
Conclusion: Potential drug interactions were common among cancer patients and most often involved medications to treat comorbid conditions. Duplicate medications were infrequent.
Prior knowledge Cancer patients are often prescribed many medications concurrently; in addition to chemotherapeutic agents, they may receive drugs to treat non-neoplastic conditions as well as medicines for supportive care. The extent to which they are exposed to adverse drug interactions is unknown. Study design Information on individual exposure to various medications in a random sample of an outpatient population was collected by questionnaire, and potential drug interactions were identified using a software program that accessed published data on drug interactions. Contribution Potential interactions were identified in 27% of patients and classified according to severity, level of scientific evidence supporting the interaction, type of medications involved, type of cancer, and other parameters. Most of the potential interactions involved medications to treat comorbid conditions. Implications The results suggest the importance of further studies to determine the frequency of adverse drug interactions in cancer patients. Limitations The study was not designed to determine how often the potential drug interactions identified actually resulted in adverse clinical consequences for patients.
|
Cancer patients are particularly susceptible to drug interactions. One reason is that they often receive multiple medications—in addition to antineoplastic agents and drugs to treat comorbid conditions, cancer patients may receive medications to treat both therapy-induced toxicity and cancer-related syndromes, such as pain, seizures, and venous thrombosis. The risk of drug interactions is further heightened because the cancer patient's pharmacokinetic parameters may be altered. The change in pharmacokinetic parameters may be due to a number of factors: impaired drug absorption due to mucositis and malnutrition, variation in a drug's volume of distribution because of reduced levels of serum-binding proteins and generalized edema, or, in patients with renal and/or hepatic dysfunction, altered drug excretion.
There are three types of drug interactions: pharmacodynamic, pharmacokinetic, and pharmaceutical (1). Pharmacodynamic interactions usually result from combining two drugs with similar mechanisms of action (in which case they may behave in synergistic, additive, or antagonistic fashion) or when an electrolytic abnormality induced by one drug alters the net effect of another. A pharmacokinetic interaction takes place when a drug alters the absorption, distribution, metabolism, and/or excretion of another drug. Pharmacokinetic interactions via metabolic effects most often occur via drug interactions with cytochrome P450 enzymes; antineoplastic medications that are entirely or partly metabolized by such enzymes include cyclophosphamide, taxanes, etoposide, irinotecan, aromatase inhibitors, vinca alkaloids, bicalutamide, imatinib, gefitinib, and erlotinib (2–7). A pharmaceutical interaction occurs when two chemically incompatible drugs are mixed before intravenous administration resulting in inactivation of one or both drugs (1).
Several studies have evaluated the potential for drug interactions in general medicine. It was found in large surveys that approximately 60% of inpatients in general medical wards were at risk of drug interactions (8–10), and studies conducted in hospital emergency departments found that from 16% to 47% of patients were at risk of drug interactions (11–13). Among 103 outpatients screened by their family physician for the presence of drug combinations in which there was a potential for interaction, almost 70% had been exposed to such combinations (14). An analysis of more than 5 million prescriptions in the French national healthcare system found that 2% of outpatients were exposed to either absolutely or relatively contraindicated drug combinations (15). Finally, in a hospital-based retrospective study conducted in Norway, 18% of 732 deaths were associated, either directly or indirectly, with drug interactions (16).
By contrast, few studies have addressed drug interactions in patients with cancer. In the Norwegian study (17), 4% of cancer-related deaths in hospitalized patients were associated with severe drug interactions (16). In a previous study, we evaluated the frequency of drug combinations with potential to interact in cancer patients, but that study included only inpatients not currently receiving anticancer therapy (17). We found that 63% of patients were exposed to drug combinations in which there was a potential for interaction. Because of the possible negative impact of drug interactions on patients and the paucity of literature on this topic in patients with cancer, we designed a cross-sectional study to evaluate the epidemiology of exposure to drugs with a potential to interact among ambulatory cancer patients receiving systemic anticancer therapy. The secondary objective was to evaluate the frequency and risk factors for duplicate prescribing in this population.
|
Patients and Methods |
|---|
|
TopAbstractContext and CaveatsPatients and MethodsResultsDiscussionReferencesNotes |
|---|
Study Design and Participants
This was a cross-sectional study undertaken at the Princess Margaret Hospital, Toronto, during an 8-month period (from September 2005 to May 2006). Princess Margaret Hospital is Canada's largest cancer center, in which more than 12000 new patients were seen in 2005 across different clinics: radiation oncology, surgical oncology, medical oncology, malignant hematology, and palliative care. Our study was restricted to follow-up medical oncology clinics. Ambulatory adult patients with a diagnosis of solid malignancy and who were currently receiving standard systemic cancer-directed treatment were considered to be eligible. Patients receiving experimental agents, especially those in early-phase clinical trials, were excluded because little is known about drug interactions with newer agents. Patients were not considered to be ineligible because of poor performance status, language barriers, or any other issue. Because all patients were on active cancer-directed therapy, all patients included in the study had an Eastern Cooperative Oncology Group performance status of 2 or less. Translators were available for patients who did not speak English.
Patients were recruited by one of the authors (R. P. Riechelmann) and a research assistant. To obtain a patient sample that was representative of all ambulatory cancer patients treated at the hospital, the recruiters rotated among the several follow-up medical oncology clinics at Princess Margaret Hospital on a biweekly basis. At the beginning of each clinic, the recruiters reviewed the clinic list with the responsible staff oncologist to identify eligible patients and to obtain permission to invite the patient to participate in the study. Because all eligible patients were on systemic treatment, they were being seen frequently in clinic. Therefore, by attending the same clinics on multiple occasions, the recruiters were able to ask all eligible patients to participate.
The study was approved by the Institutional Research Ethics Board, and written informed consent was obtained from all participants. Consenting patients were asked to complete a questionnaire (available online as Supplementary Data), either by themselves or with help from their caregiver. The four-page questionnaire collected data on age, sex, cancer diagnosis, and comorbid illnesses, as well as details of anticancer treatment. It also asked patients to list all medications (both enteral and parenteral agents administered at home or in the hospital) taken in the previous 4 weeks, along with the names of the respective prescribing physicians. Patients who did not remember all their current medications while completing the questionnaire received telephone follow-up calls within a week. Chart review was performed to confirm and supplement information obtained from the questionnaire, such as anticancer regimen, treatment intent (palliative versus curative), and laboratory abnormalities. A laboratory abnormality was defined as an increase of 10% or greater above the upper normal limit in plasma levels of hepatic enzymes and creatinine measured within the prior 4 weeks (upper normal limits: aspartate transaminase 
35 U/L, alanine transaminase 40 U/L, alkaline phosphatase 110, bilirubin 22 µmol/L, creatinine 99 µmol/L). If more than one blood test was done during this period, we chose the most abnormal result for analysis.
Drugs were classified as either "supportive care agents" (defined as medications to treat cancer- and/or therapy-related symptoms) or "medications to treat comorbid conditions." A comorbid illness was defined as a noncancer clinical condition that required pharmacologic treatment. The number of medications for each patient was calculated by summing all medications except anticancer drugs; when a medication contained two or more pharmacologic compounds (e.g., acetaminophen combined with codeine), each drug was considered to be an individual drug in the analysis. However, when a patient was taking the same medication on more than one schedule (e.g., long- and short-acting morphine for pain control), that drug was counted only once.
Duplicate prescribing was considered to be present when two or more drugs from the same class were prescribed to treat the same condition (e.g., morphine and codeine prescribed as routine orders for pain) or different conditions (e.g., corticosteroids to prevent delayed nausea and as anti-inflammatory agents).
Potential drug interactions were identified using the Drug Interaction Facts software, version 4.0 (18). If a drug was not recognized by the program, potential interactions were identified manually by one of the authors (R. P. Riechelmann), using pharmacology textbooks. The Drug Interaction Facts software screens for potential drug interactions and, if one is identified, provides a description of the pharmacologic mechanisms for the interaction and classifies it as pharmacokinetic, pharmacodynamic, or, if there is not sufficient evidence to support a known underlying pharmacologic mechanism, unknown. It also classifies the interaction by level of severity and scientific evidence (Table 1). In terms of severity, drug interactions were classified as major, when the potential drug interaction could lead to permanent damage or risk of death; moderate, when the clinical consequence of an interaction required medical treatment; or minor, when small or no clinical effect was expected from the combination of two drugs. With respect to level of scientific evidence, potential drug interactions were classified on a 5-point scale, with level 1 evidence meaning that a drug interaction was supported by large clinical trials and level 5 meaning that the risk for interaction between two drugs was only theoretical. The Drug Interaction Facts software had been found in a previous study to have both sensitivity and specificity of 97% in detecting previously described/known potential drug interactions (19). We did not study potential interactions between drugs and complementary or alternative medications, herbs, or food because these were beyond the scope of this study and were not provided by the software. If a potentially serious drug interaction and/or duplicate medication was identified, the prescribing physician was contacted.
|
Statistical Analyses
There were no data about drug interactions in oncology on which to base our sample size. Therefore, 400 patients were predefined and chosen as a feasible sample, and we approached patients until this number was enrolled. Summary statistics were used to describe patient characteristics (such as age, sex, cancer types, treatment intent, type of anticancer therapy, comorbid conditions, number of drugs per patient, type of medication taken by patients, and laboratory abnormalities); frequency, types and classification of drug interactions; and frequency with which medications were duplicated. Logistic regression was used to identify risk factors associated with potential drug interactions. The dependent variable was the number of potential drug interactions for which there was reasonable supportive evidence (i.e., scientific evidence levels 1–3, Table 1). Explanatory variables were age, cancer type (breast, gastrointestinal, genitourinary, gynecologic, lung, brain, or other), treatment intent (palliative versus curative), treatment type (chemotherapy, hormone therapy, molecular agents, or combination), presence of at least one comorbid illness (yes/no), number of drugs (continuous variable), type of medications (supportive care agents, medications to treat comorbid conditions, or both), and presence of at least one laboratory abnormality (yes/no). Because some cancer types only occur in men or women, much of the information that sex would contribute to the regression analysis is implicit in cancer type; sex was not included as a potential risk factor, but cancer type was included. For binary variables, the group at lower risk of the outcome was chosen as the referent. Similarly, for other nominal variables, we also chose the group at least risk as the reference group. Variables with univariate P values that were less than .1 were entered into the multivariable model. Because of concerns that some variables may have been highly correlated, and therefore caused multicollinearity and violated a regression assumption, we constructed a correlation matrix and examined the variance inflation factors for each of the explanatory variables. These multicollinearity diagnostics did not indicate that multicollinearity was a problem in our final model. In multivariable analysis, both stepwise and backward selection techniques yielded the same results, so only results of the stepwise regression are presented. In the multivariable model, predictors were considered to be statistically significant if the P value was less than .05. The final model was adjusted for treatment intent as a proxy for disease severity. All statistical tests were two-sided.
|
Results |
|---|
|
TopAbstractContext and CaveatsPatients and MethodsResultsDiscussionReferencesNotes |
|---|
Characteristics of Patients
A total of 409 patients were initially identified as eligible; of these, two who had cognitive impairment and were considered to be ineligible by their staff oncologist were not asked to participate. Of the 407 patients invited to participate in the study, two subjects refused to participate. The remaining 405 patients completed the questionnaire and were included in the analysis; 399 (98.5%) of the questionnaires were fully completed. Six participants (1.5%) of 405 did not remember all their medications, and the missing information was collected by telephone within one week. The median age of participants was 58 years (range 21–88 years); 64% were female and 36% were male (Table 2). The most common cancer types were breast cancers (39%), gastrointestinal tumors (24%), and genitourinary cancers (16%). Sixty-two percent of patients were receiving anticancer therapy with palliative intent; 57% were receiving chemotherapy, 25% were being treated with hormone therapy, 7% were receiving molecular agents, and 11% were receiving more than one type of antineoplastic therapy. The median number of comorbid illnesses experienced by a patient was 1 (range = 0–5), with cardiovascular diseases, musculoskeletal disorders, hypothyroidism, and depression being the most frequent. The median number of medications per patient was 5 (range 0–23). Forty percent of patients were taking supportive medications only (mostly opioids, acetaminophen, antiemetic agents, steroids, stool softeners, and stomach protectors), 12% were receiving medications to treat comorbid illnesses exclusively, and 48% were receiving drugs for both purposes. Forty-six percent of patients had laboratory abnormalities, with liver dysfunction (affecting 28% of the patients) being the most common.
|
A total of 276 potential drug interactions were identified in 109 (27%; 95% confidence interval [CI] = 23% to 31%) patients (Table 3); interactions supported by at least several case reports (i.e., levels of scientific evidence < 3, Table 1) were identified in 18% of patients. The majority of potential drug interactions were of moderate severity (77%), and 49% of them were supported by levels 1 or 2 scientific evidence, i.e., evidence to suggest that adverse effects were probable. Approximately half (55%) of the potential interactions were classified as pharmacokinetic. Among the 276 potential interactions, 240 (87%) involved non-antineoplastic agents and 36 (13%) involved antineoplastic drugs.
|
The potential drug interactions between antineoplastic agents and common medications that were identified are presented in Table 4; most of them involved warfarin, hydrochlorothiazide, or quinolones. The drug that was identified most frequently as having potential for interaction with antineoplastic agents was warfarin; a potential for pharmacodynamic interaction of hydrochlorothiazide with cyclophosphamide and fluorouracil was also common. Because of the extensive number of different potential interactions between non-antineoplastic medications, only the ones identified most frequently are reported in Table 5; they involved, in order of decreasing frequency, antihypertensive agents (angiotensin-converting enzyme inhibitors [ACE inhibitors], beta-blockers, and hydrochlorothiazide), aspirin, warfarin, corticosteroids, phenytoin, and prochlorperazine.
|
|
A total of 390 patients were included in the logistic regression to determine risk factors associated with potential drug interactions; 15 patients who were currently not receiving other medications besides those used in systemic anticancer treatment were excluded from analysis because they were not at risk of potential drug interactions. Results of the univariate and multivariable analyses are presented in Table 6. In unadjusted analyses, older patients, patients receiving more drugs or drugs for comorbid illness, patients treated with palliative intent, those with comorbid illness, and those with certain types of cancer (genitourinary, brain, and gynecologic) were at increased risk of potential drug interactions. No statistically significant association was found between type of treatment or laboratory evidence of renal or liver dysfunction and the risk for drug interactions. In adjusted analyses, only increasing number of medications (odds ratio [OR] for every additional drug: 1.4, 95% CI = 1.26 to 1.58, P<.001), cancer type (OR, brain versus genitourinary tumors: 6.7, 95% CI = 2.0 to 23, P = .0025) and the type of medications that the patient was receiving (OR, drugs to treat comorbid conditions versus supportive care drugs: 8.6, 95% CI = 2.9 to 25, P<.001, OR, both types of medications versus supportive care drugs: 2.5, 95% CI = 1.2 to 5.5, P = .018) showed a statistically significant association with the drugs with potential to interact for which there was reasonable supportive evidence (i.e., scientific evidence levels 1–3, Table 1).
|
Duplicate Prescribing
Thirty-two (8%) patients were exposed to duplicate prescribing. Three of these patients had two instances of duplicate prescriptions. Most cases of duplicated medications were due to duplication of corticosteroids (16 cases), with the majority of these caused by prescription of dexamethasone for prevention of docetaxel-related reactions and prednisone to treat prostate cancer. Of the duplicate prescriptions not involving corticosteroids, seven were of proton pump inhibitors, six were of benzodiazepines, and four were of opioids. There was one duplication each of antidepressants and bisphosphonates.
|
Discussion |
|---|
|
TopAbstractContext and CaveatsPatients and MethodsResultsDiscussionReferencesNotes |
|---|
Our study demonstrates that the frequency of drug combinations in which there is potential for harmful interaction was high in the ambulatory cancer patients in our study, with about one-third of patients being exposed to at least one interaction. Drug combinations for which there were potential interactions supported by a level of scientific evidence of 3 or less (i.e., scientific evidence supported by at least several case reports) were prescribed to 18% of patients. These rates are of concern because 86% of all interactions were classified as major or moderate, and almost 50% of them were supported by level 1–2 scientific evidence (i.e., evidence from clinical trials). The large majority of potential drug interactions were due to the presence of non-anticancer agents (87%); most often these were antihypertensive agents, aspirin, warfarin, or anticonvulsants. Potential drug interactions due to the presence of anticancer agents most often also involved warfarin, whose anticoagulant effects are increased by anticancer agents (20–24), or hydrochlorothiazide, whose interaction with anticancer drugs may prolong neutropenia (25). Factors such as increasing number of medications, the presence of certain tumor types, and receipt of medications for comorbid illness as opposed to drugs for cancer- or treatment-related symptoms were associated with increased risk of potential drug interactions.
A retrospective pilot study undertaken by two of the authors of this study reported potential drug interactions in 63% of 100 consecutive inpatients not receiving anticancer therapy (17). In that study, which utilized the same electronic screening method as the present investigation, 75% of potential interactions were classified as major or moderate. A Norwegian study found that 4% of cancer-related deaths in hospitalized patients were likely to be associated with severe interactions (16). We are unaware of other studies that have evaluated the epidemiology of potential drug interactions among ambulatory cancer patients receiving antineoplastic therapy.
The finding that increasing number of medications was a risk factor for potential drug interactions in our population is consistent with previous studies (9,12,26) and is not surprising. Also, patients with brain tumors were more likely to be exposed to drug combinations for which there were potential drug interactions compared to patients with other tumor types; this was probably because brain tumor patients often use anticonvulsants. Since the majority of potential drug interactions involved antihypertensive agents, aspirin, and warfarin, it is not surprising that the odds of being exposed to a potential drug interaction were eight times greater for patients receiving medications to treat comorbid conditions than for patients receiving only drugs for supportive care. The type of antineoplastic therapy was not a predictive factor for potential drug interactions.
Our study is limited by the fact that it was performed in a single institution; thus, its external validity is unknown. However, the internal validity should be high because we systematically collected the names of all medications and tried to recruit consecutive eligible patients as often as possible to limit selection bias. Another limitation of our study is that, because of its high sensitivity (19), the screening software was capable of detecting the majority of drug interactions, including those supported by lower levels of evidence and those for which the clinical consequences are unknown.
The major limitation of our study is lack of information about the number instances in which drug combinations with potential interactions resulted in clinical consequences. We attempted to collect information on real drug interactions by asking patients about hospitalization and causes thereof during the 12 months before the completion of the questionnaire. However, it was difficult to obtain reliable information in part because Princess Margaret Hospital is a tertiary referral center and patients come from a wide encatchment area. Therefore, most admissions are to community hospitals closer to the patient's residence. Furthermore, there is an inherent bias in studies evaluating real drug interactions because if a drug interaction that led to serious clinical consequences has already occurred, adjustment in prescribing is likely to have taken place. Also, when we identified potentially life-threatening interactions, the primary oncologist was contacted, and, in most cases, treatment modifications were made to prevent adverse clinical events.
The most commonly used drug with potential for interaction with antineoplastic agents was warfarin. The combination of warfarin with fluorouracil, capecitabine, etoposide, carboplatin, paclitaxel, or gemcitabine may cause hemorrhage due to chemotherapy-induced serum protein displacement (higher volume of distribution of warfarin) and/or interference with warfarin hepatic metabolism (20–24). This finding is important because all of the above potential interactions are supported by level 2 scientific evidence for their occurrence and are of moderate severity. A frequently identified potential pharmacodynamic interaction was due to the concurrent administration of hydrochlorothiazide with cyclophosphamide and fluorouracil (Table 4). This potential interaction, although supported only by level 4 scientific evidence (i.e., a few case reports), may lead to prolonged and severe neutropenia (25). Among potential interactions with non-anticancer drugs, the most commonly identified were those between aspirin and ACE inhibitors or beta-blockers and between aspirin and corticosteroids, all supported by level 2 evidence. In the presence of aspirin, the hypotensive effects of ACE inhibitors and/or beta-blockers may be impaired because of decreased synthesis of prostaglandins caused by aspirin (27–29). The increased hepatic metabolism and renal excretion of aspirin induced by corticosteroids may lead to reduction of plasma levels of aspirin levels and consequently, less clinical effect (30).
Although identification of real drug interactions was not formally part of this study, we saw clear evidence for them. For instance, two patients who were taking long-term warfarin with a stable international normalized ratio of between 2.0 and 3.0 experienced an increase of more than 50% in the international normalized ratio after prescription of capecitabine (21). A patient receiving adjuvant fluorouracil for stage III colon cancer experienced prolonged neutropenia after being prescribed cimetidine during the fourth cycle of chemotherapy (31). Furthermore, a patient receiving a short-term nonsteroidal anti-inflammatory drugs together with a serotonin-selective reuptake inhibitor was admitted because of upper gastrointestinal bleeding (32). Descriptions of these probable drug interactions are detailed in Tables 4 and 5.
The best way to prevent drug interactions is unknown. Alert guidelines, such as electronic alerts that appear when health professionals enter patients' medication orders into the electronic medical record, or flyers to remind physicians and pharmacists of drugs with potential to interact could be developed to help identify potentially hazardous interactions. Computerized programs can increase recognition of such interactions and can provide an important tool for screening them (33). We suggest that patients at high risk, such as those with comorbid conditions or brain tumors and particularly those receiving warfarin, anticonvulsants, and antihypertensive medications be routinely screened for potential drug interactions. The development of medication databases and computerized physician medication order entry linked to screening electronic programs could help health professionals to identify dangerous drug combinations and monitor prescriptions of agents with high risks of interactions such as anticonvulsants and warfarin. These tools would also allow for easy identification of patients whose prescriptions contain numerous medications.
In our study, the frequency of duplicated medications was low and mostly due to coprescribing of dexamethasone and prednisone, ranitidine/famotidine and omeprazole, or benzodiazepines and opioids. The clinical impact of duplicate prescribing of steroids and stomach protectors is perhaps low, since dexamethasone was usually given for 2–3 days to prevent docetaxel-related hypersensitivity reactions or delayed nausea/vomiting. The coadministration of omeprazole and ranitidine/famotidine is not likely to lead to major consequences. However, duplicate prescribing of benzodiazepines and opioids can be hazardous.
In conclusion, potential drug interactions are frequent in oncology and many are clinically important. Population-based studies are needed to assess the prevalence of "real" drug interactions. Development of alert guidelines and computer-based screening would help physicians to recognize and prevent potentially dangerous drug interactions.
|
NOTES |
|---|
|
TopAbstractContext and CaveatsPatients and MethodsResultsDiscussionReferencesNotes |
|---|
All authors state that the funding agencies had no role in the design of this study, data collection, analysis and interpretation of the results, or the writing of the manuscript.
|
REFERENCES |
|---|
|
TopAbstractContext and CaveatsPatients and MethodsResultsDiscussionReferencesNotes |
|---|
(1) Beijnen JH, Schellens JH. Drug interactions in oncology. Lancet Oncol (2004) 5:489–96.[CrossRef][Web of Science][Medline]
(2) Hardman JG, Limbird LE, Gilman A, eds. Chemotherapy in malignant diseases. In: Goodman & Gilman's The Pharmacoligical Basis of Therapeutics (1996) 9th ed. Cedro (NM): McGraw-Hill Companies, Inc. 903–52.
(3) Lonning P, Pfister C, Martoni A, Zamagni C. Pharmacokinetics of third-generation aromatase inhibitors. Semin Oncol (2003) 30(Suppl 14):23–32.[Medline]
(4) Cockshott ID. Bicalutamide: clinical pharmacokinetics and metabolism. Clin Pharmacokinet (2004) 43:855–78.[CrossRef][Web of Science][Medline]
(5) O'Brien SG, Meinhardt P, Bond E, Beck J, Peng B, Dutreix C, et al. Effects of imatinib mesylate (STI571, Glivec) on the pharmacokinetics of simvastatin, a cytochrome p450 3A4 substrate, in patients with chronic myeloid leukaemia. Br J Cancer (2003) 89:1855–9.[CrossRef][Web of Science][Medline]
(6) Ranson M, Gefitinib Wardell S. a novel, orally administered agent for the treatment of cancer. J Clin Pharm Ther (2004) 29:95–103.[CrossRef][Web of Science][Medline]
(7) Hidalgo M, Bloedow D. Pharmacokinetics and pharmacodynamics: maximizing the clinical potential of Erlotinib (Tarceva). Semin Oncol (2003) 30(Suppl 7):25–33.[Web of Science][Medline]
(8) Egger SS, Drewe J, Schlienger RG. Potential drug-drug interactions in the medication of medical patients at hospital discharge. Eur J Clin Pharmacol (2003) 58:773–8.[Web of Science][Medline]
(9) Geppert U, Beindl W, Hawranek T, Hintner H. Drug interactions in clinical practice. A pilot project for quality assurance in prescribing. Hautarzt (2003) 54:53–7.[Web of Science][Medline]
(10) Glintborg B, Andersen SE, Dalhoff K. Drug-drug interactions among recently hospitalised patients–frequent but mostly clinically insignificant. Eur J Clin Pharmacol (2005) 61:675–81.[CrossRef][Web of Science][Medline]
(11) Karas S. The potential for drug interactions. Ann Emerg Med (1981) 10:627–30.[CrossRef][Web of Science][Medline]
(12) Herr RD, Caravati EM, Tyler LS, Iorg E, Linscott MS. Prospective evaluation of adverse drug interactions in the emergency department. Ann Emerg Med (1992) 21:1331–6.[CrossRef][Web of Science][Medline]
(13) 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) 14:447–50.[CrossRef][Web of Science][Medline]
(14) Davidson KW, Kahn A, Price RD. Reduction of adverse drug reactions by computerized drug interaction screening. J Fam Pract (1987) 25:371–5.[Web of Science][Medline]
(15) Guedon-Moreau L, Ducrocq D, Duc MF, Quieureux Y, L'Hote C, Deligne J, et al. Absolute contraindications in relation to potential drug interactions in outpatient prescriptions: analysis of the first five million prescriptions in 1999. Eur J Clin Pharmacol (2004) 59:899–904.[CrossRef][Web of Science][Medline]
(16) Buajordet I, Ebbesen J, Erikssen J, Brors O, Hilberg T. Fatal adverse drug events: the paradox of drug treatment. J Intern Med (2001) 250:327–41.[CrossRef][Web of Science][Medline]
(17) Riechelmann RP, Moreira F, Smaletz O, Saad ED. Potential for drug interactions in hospitalized cancer patients. Cancer Chemother Pharmacol (2005) 56:286–90.[CrossRef][Web of Science][Medline]
(18) Drug Interaction Facts. Version 4.0. 2006. Wolters and Kluwer Health. Available at: http://www.factsandcomparisons.com. [Last accessed; March 19, 2006.].
(19) Barrons R. Evaluation of personal digital assistant software for drug interactions. Am J Health Syst Pharm (2004) 61:380–5.
(20) Kolesar JM, Johnson CL, Freeberg BL, Berlin JD, Schiller JH. Warfarin-5-FU interaction–a consecutive case series. Pharmacotherapy (1999) 19:1445–9.[CrossRef][Web of Science][Medline]
(21) Camidge R, Reigner B, Cassidy J, Grange S, Abt M, Weidekamm E, et al. Significant effect of capecitabine on the pharmacokinetics and pharmacodynamics of warfarin in patients with cancer. J Clin Oncol (2005) 23:4719–25.
(22) Kinikar SA, Kolesar JM. Identification of a gemcitabine-warfarin interaction. Pharmacotherapy (1999) 19:1331–3.[CrossRef][Web of Science][Medline]
(23) Le AT, Hasson NK, Lum BL. Enhancement of warfarin response in a patient receiving etoposide and carboplatin chemotherapy. Ann Pharmacother (1997) 31:1006–8.[Abstract]
(24) Thompson ME, Highley MS. Interaction between paclitaxel and warfarin. Ann Oncol (2003) 14:500.
(25) Orr LE. Potentiation of myelosuppression from cancer chemotherapy and thiazide diuretics. Drug Intell Clin Pharm (1981) 15:967–70.[Abstract]
(26) 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) 112:61–4.
(27) Guazzi M, Brambilla R, Reina G, Tumminello G, Guazzi MD. Aspirin-angiotensin-converting enzyme inhibitor coadministration and mortality in patients with heart failure: a dose-related adverse effect of aspirin. Arch Intern Med (2003) 163:1574–9.
(28) Hall D, Zeitler H, Rudolph W. Counteraction of the vasodilator effects of enalapril by aspirin in severe heart failure. J Am Coll Cardiol (1992) 20:1549–55.[Abstract]
(29) Sziegoleit W, Rausch J, Polak G, Gyorgy M, Dekov E, Bekes M. Influence of acetylsalicylic acid on acute circulatory effects of the beta-blocking agents pindolol and propranolol in humans. Int J Clin Pharmacol Ther Toxicol (1982) 20:423–30.[Medline]
(30) Edelman J, Potter JM, Hackett LP. The effect of intra-articular steroids on plasma salicylate concentrations. Br J Clin Pharmacol (1986) 21:301–7.[Web of Science][Medline]
(31) Harvey VJ, Slevin ML, Dilloway MR, Clark PI, Johnston A, Lant AF. The influence of cimetidine on the pharmacokinetics of 5-fluorouracil. Br J Clin Pharmacol (1984) 18:421–30.[Web of Science][Medline]
(32) de Jong JC, van den Berg PB, Tobi H, de Jong-van den Berg LT. Combined use of SSRIs and NSAIDs increases the risk of gastrointestinal adverse effects. Br J Clin Pharmacol (2003) 55:591–5.[CrossRef][Web of Science][Medline]
(33) Glassman PA, Simon B, Belperio P, Lanto A. Improving recognition of drug interactions: benefits and barriers to using automated drug alerts. Med Care (2002) 40:1161–71.[CrossRef][Web of Science][Medline]
(34) Johnson EJ, MacGowan AP, Potter MN, Stockley RJ, White LO, Slade RR, et al. Reduced absorption of oral ciprofloxacin after chemotherapy for haematological malignancy. J Antimicrob Chemother (1990) 25:837–42.
(35) Cagnoni PJ, Matthes S, Day TC, Bearman SI, Shpall EJ, Jones RB. Modification of the pharmacokinetics of high-dose cyclophosphamide and cisplatin by antiemetics. Bone Marrow Transplant (1999) 24:1–4.[CrossRef][Web of Science][Medline]
(36) Ritchie LD, Grant SM. Tamoxifen-warfarin interaction: the Aberdeen hospitals drug file. BMJ (1989) 298:1253.
(37) Grossman SA, Sheidler VR, Gilbert MR. Decreased phenytoin levels in patients receiving chemotherapy. Am J Med (1989) 87:505–10.[Web of Science][Medline]
(38) Chin TW, Loeb M, Fong IW. Effects of an acidic beverage (coca-cola) on absorption of ketoconazole. Antimicrob Agents Chemother (1995) 39:1671–5.[Abstract]
(39) Gilbar PJ, Brodribb TR. Phenytoin and fluorouracil interaction. Ann Pharmacother (2001) 35:1367–70.
(40) Brummett RE. Ototoxicity resulting from the combined administration of potent diuretics and other agents. Scand Audiol Suppl (1981) 14(Suppl):215–24.[Medline]
(41) Costedoat-Chalumeau N, Amoura Z, Aymard G, Sevin O, Wechsler B, Cacoub P, et al. Potentiation of vitamin K antagonists by high-dose intravenous methylprednisolone. Ann Intern Med (2000) 132:631–5.
(42) Haidukewych D, Rodin EA. Effect of phenothiazines on serum antiepileptic drug concentrations in psychiatric patients with seizure disorder. Ther Drug Monit (1985) 7:401–4.[Web of Science][Medline]
(43) Tse CS, Iagmin P. Phenytoin and ranitidine interaction. Ann Intern Med (1994) 120:892–3.
(44) Vincent FM. Phenothiazine-induced phenytoin intoxication. Ann Intern Med (1980) 93:56–7.[Web of Science][Medline]
(45) Lackner TE. Interaction of dexamethasone with phenytoin. Pharmacotherapy (1991) 11:344–7.[Web of Science][Medline]
(46) Elbe DH, Chang SW. Moxifloxacin-warfarin interaction: a series of five case reports. Ann Pharmacother (2005) 39:361–4.
(47) Chesebro JH, Fuster V, Elveback LR, McGoon DC, Pluth JR, Puga FJ, et al. Trial of combined warfarin plus dipyridamole or aspirin therapy in prosthetic heart valve replacement: danger of aspirin compared with dipyridamole. Am J Cardiol (1983) 51:1537–41.[CrossRef][Web of Science][Medline]
(48) White WB. Hypotension with postural syncope secondary to the combination of chlorpromazine and captopril. Arch Intern Med (1986) 146:1833–4.
(49) Micossi P, Pontiroli AE, Baron SH, Tamayo RC, Lengel F, Bevilacqua M, et al. Aspirin stimulates insulin and glucagon secretion and increases glucose tolerance in normal and diabetic subjects. Diabetes (1978) 27:1196–204.[Abstract]
(50) Kubacka RT, Antal EJ, Juhl RP, Welshman IR. Effects of aspirin and ibuprofen on the pharmacokinetics and pharmacodynamics of glyburide in healthy subjects. Ann Pharmacother (1996) 30:20–6.[Abstract]
(51) Mahe I, Bertrand N, Drouet L, Simoneau G, Mazoyer E, Bal dit Sollier C, et al. Paracetamol: a haemorrhagic risk factor in patients on warfarin. Br J Clin Pharmacol (2005) 59:371–4.[CrossRef][Web of Science][Medline]
(52) Graham DY, Malaty HM. Alendronate and naproxen are synergistic for development of gastric ulcers. Arch Intern Med (2001) 161:107–10.
(53) Santos F, Smith MJ, Chan JC. Hypercalciuria associated with long-term administration of calcitriol (1,25-dihydroxyvitamin D3). Action of hydrochlorothiazide. Am J Dis Child (1986) 140:139–42.
Manuscript received July 12, 2006; revised February 7, 2007; accepted March 19, 2007.
CiteULike 
Connotea 
Del.icio.us What's this?
Find additional patient-related information at:
Editorial about this Article
- Patient Safety in Cancer Care: A Time for Action
- Peter G. Norton and G. Ross Baker
J Natl Cancer Inst 2007 99: 579-580.[Extract] [Full Text] [PDF]
Related Articles in JNCI
-
IN THIS ISSUE
J Natl Cancer Inst 2007 99: 577.[Extract] [Full Text] [PDF]
- Press Release: Cancer Patients Are at High Risk for Potential Drug Interactions
- Liz Savage
J Natl Cancer Inst 2007 99: 577.[Extract] [Full Text]
This article has been cited by other articles:
|
|
 |
|
  R. Yano, D. Tani, K. Watanabe, H. Tsukamoto, T. Igarashi, T. Nakamura, and M. Masada Evaluation of Potential Interaction Between Vinorelbine and Clarithromycin Ann. Pharmacother., March 1, 2009; 43(3): 453 - 458. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. E. Walsh, K. S. Dodd, K. Seetharaman, D. W. Roblin, L. J. Herrinton, A. Von Worley, G. Naheed Usmani, D. Baer, and J. H. Gurwitz Medication Errors Among Adults and Children With Cancer in the Outpatient Setting J. Clin. Oncol., February 20, 2009; 27(6): 891 - 896. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.-M. Wong, Y. Ko, and A. Chan Clinically Significant Drug-Drug Interactions Between Oral Anticancer Agents and Nonanticancer Agents: Profiling and Comparison of Two Drug Compendia Ann. Pharmacother., December 1, 2008; 42(12): 1737 - 1748. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. H Hanigan, B. L dela Cruz, D. M Thompson, K. C Farmer, and P. J Medina Use of prescription and nonprescription medications and supplements by cancer patients during chemotherapy: questionnaire validation Journal of Oncology Pharmacy Practice, September 1, 2008; 14(3): 123 - 130. [Abstract] [PDF]
|
|
|
|
 |
|
 F. Xu Caution in prescribing antidepressants for patients with cancer Am. J. Health Syst. Pharm., April 15, 2008; 65(8): 700 - 700. [Full Text] [PDF]
|
|
|
|
 |
|
 A J Munro Hidden danger, obvious opportunity: error and risk in the management of cancer Br. J. Radiol., December 1, 2007; 80(960): 955 - 966. [Full Text] [PDF]
|
|
|
|
 |
|
 P. G. Norton and G. R. Baker Patient Safety in Cancer Care: A Time for Action J Natl Cancer Inst, April 18, 2007; 99(8): 579 - 580. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||