Toward semantic modeling of pharmacogenomic knowledge for clinical and translational decision support.

The purpose of this paper is to describe pilot work on a semantic model of the pharmacogenomics information found in drug product labels. The model's development is driven by a series of use cases that we have developed to demonstrate how structured pharmacogenomics information could be more effectively used to support clinical and translational efforts. Using an iterative process, the semantic model was field-tested by five pharmacists, who used it to manually annotate a subset of the sections that the Food and Drug Administration's Table of Pharmacogenomic Biomarkers in Drug Labels cites as containing pharmacogenomics information. The five pharmacists identified a total of 213 pharmacogenomics statements in 29 sections. The model showed the potential to make the unstructured pharmacogenomic information currently written in product labeling more accessible and actionable through structured annotations of pharmacogenomics effects and clinical recommendations.

Amoxicillin renal toxicity: how often does it occur?

OBJECTIVES: To determine the incidence of renal symptoms associated with amoxicillin, a retrospective review of exposures to amoxicillin in children younger than 6 years as reported to the National Poison Data System was done. METHODS: All ingestions of amoxicillin without coingestants in humans younger than 6 years reported to the National Poison Data System from 2004 through 2008 were analyzed. Data included age, sex, management site, outcome, symptoms, amount ingested, certainty of amount, chronicity, weight, and therapy. The study was approved by the institutional review board. Descriptive statistics were used to characterize the data. RESULTS: A total of 14,717 cases were identified. Related renal symptoms occurred in 5 patients (0.03%). In 1687 patients (9.6%), the total amount (in milligrams) was documented, and the median amount ingested was 1000 mg. In patients with a known amount (in milligrams) along with the child's weight (n = 1356), the median amount was 82.6 mg/kg. In this group, 213 ingested greater than 250 mg/kg (range, 251.4-1531.1 mg/kg; median, 366.5 mg/kg). Treatment sites for this group included the following: treated in the home, 129 (60.6%); treated and released from an health care facility, 63 (29.6%); treated while admitted, 2 (0.9%); refused a referral, 7 (3.3%); lost to follow-up, 9 (4.2%); and managed at other sites, 3 (1.4%). Within this group, 94 patients (44.1%) were followed up to a definitive outcome: 77 (81.9%) had no effect, 15 (16.0%) had minor symptoms, and 2 (2.1%) had moderate symptoms. CONCLUSIONS: Although renal toxicity may occur with amoxicillin ingestions, it is rare and does not seem to be dose related.

Evaluation of pharmacist utilization of a poison center as a resource for patient care.

The objective of this study was to evaluate pharmacist use of a Regional Poison Information Center (RPIC), identify potential barriers to utilization, and provide strategies to overcome these barriers. All calls placed to a RPIC by a pharmacist, physician, or nurse over a 5-year period were retrieved. These data were analyzed to assess the pharmacist utilization of the RPIC and the variation of call types. Additionally, a survey, designed to assess the past and future use of the RPIC by pharmacists, was distributed to pharmacists in the region. Of the 37,799 calls made to the RPIC, 26,367 (69.8%) were from nurses, 8096 (21.4%) were from physicians, and 3336 (8.8%) were from pharmacists. Among calls initiated by pharmacists, the majority involved medication identification (n = 2391, 71.7%). The survey had a 38.9% response rate (n = 715) and revealed a trend toward less RPIC utilization by pharmacists with more formal training but less practice experience. The utilization of the RPIC was lowest among pharmacists as compared to other health care professionals. This may be due to pharmacists' unfamiliarity with the poison center's scope of services and resources. Therefore, it is important that pharmacists are educated on the benefit of utilizing poison centers in clinical situations.

Cost minimization analysis comparing enteral N-acetylcysteine to intravenous acetylcysteine in the management of acute acetaminophen toxicity.

CONTEXT: Acetaminophen poisoning is one of the most common exposures and causes of poisoning-related fatalities as reported to U.S. poison information centers. Acetylcysteine is indicated for the antidotal treatment of acetaminophen poisoning to prevent or minimize acetaminophen-related hepatotoxicity. Available as either an enteral or intravenous (IV) formulation, both forms of acetylcysteine have been proven to be efficacious. Because of the differences in the acquisition costs and the length of treatment, it is unclear which treatment route is the most cost-effective. OBJECTIVE: The purpose of this study was to compare the total hospitalization charges associated with patients who received either enteral or IV acetylcysteine therapy. MATERIALS AND METHODS: A retrospective, IRB-approved cohort study of patients treated with either enteral or IV acetylcysteine at a university-related hospital for the treatment of acute acetaminophen overdose was conducted. Patients included were over 18 years of age, admitted during the 5-year periods of 1996-2000 (enteral) and 2004-2008 (IV), had an ICD-9 discharge diagnosis for acetaminophen overdose, had no transplant history, and were admitted within 24 h of the overdose. The primary endpoint was the total cost associated with the hospital stay. The Consumer Price Index (CPI) inflation calculator from the U.S. Bureau of Labor Statistics was used to adjust all monetary values to 2008 dollars. RESULTS: Of a total of 1,647 patients, 261 met the inclusion criteria with 70 patients being treated with enteral acetylcysteine and 191 patients treated with IV acetylcysteine. The associated cost was greater in the enteral group than in the IV group ($18,287.63 vs. $7,607.82; p < 0.001). The average length of stay was longer in the enteral group compared to the IV group (7 days vs. 4 days; p < 0.001). CONCLUSIONS: Patients who were treated with IV acetylcysteine had a decreased length of stay and cost of hospitalization compared with those patients who were treated with enteral acetylcysteine.

Effect of drug information request templates on pharmacy student compliance with the modified systematic approach to answering drug information questions.

BACKGROUND: The modified systematic approach to answering drug information questions is a technique used in drug information practice and in teaching pharmacy students to effectively provide drug information. Drug information request templates were developed to prompt students and other trainees to ask appropriate background questions and perform an effective search. OBJECTIVE: An evaluation was conducted to determine whether use of drug information templates by fourth-year pharmacy students during their drug information experiential rotation improved compliance with the modified systematic approach. METHODS: Fifty documented drug information requests, including 25 prior to template implementation (August 2005-August 2006) and 25 after template implementation (August 2007-August 2008), were randomly selected for evaluation. Each question was evaluated for completeness of background information obtained, categorization and identification of the ultimate question, completeness of references searched, and formulation of a concise response and an evidence-based recommendation. RESULTS: Background information was complete in 16% of pre-template questions and 92% of post-template questions (p < 0.001). Eighty-four percent of pre-template questions and 96% of post-template questions were appropriately categorized (p = 0.349). The requestor's ultimate question was clearly identified in 68% of pretemplate questions and 92% of post-template questions (p = 0.074). All necessary references were searched in 36% of pre-template questions and 88% of post-template questions (p < 0.001). A concise response was documented in 80% of pretemplate questions and 92% of post-template questions (p = 0.417). In questions determined to require a specific recommendation among the pre-template (n = 20) and post-template groups (n = 14), a clear and evidence-based recommendation was described in 40% (p = 0.038) and 79% (p = 0.038), respectively. CONCLUSIONS: Use of drug information request templates improves students' compliance with the modified systematic approach, most notably in obtaining background information and searching necessary references including primary literature.

Ordering off the menu: assessing compliance with a nonformulary medication policy.

BACKGROUND: Use of nonformulary medications (NFMs) may create added costs and potential safety issues. Institutional policies can support compliance with formulary systems and address these issues. However, periodic review is needed to assess the effectiveness of NFM policies. OBJECTIVE: To evaluate NFM use at a large university medical center and assess for compliance with the established policy as well as for prescribing errors, procurement procedures, and delivery times. METHODS: All new medication orders entered as NFM in the computerized order entry system during a 6-week period were evaluated and followed until administration of the first NFM dose. Orders were categorized as NFMs, formulary medications (FMs), or the patient's own medications. Each order was evaluated to determine the prescriber, service and location of the patient, and any prescribing errors. Additional information, such as the indication, location where the medication was obtained, and the delivery time, was obtained through patient chart review and the pharmacist intervention database. RESULTS: A total of 237 orders were entered as nonformulary; 72 (30%) were FMs and 165 (70%) were NFMs. Of the 165 NFM orders, 71 (43%) were orders for the patient's own medication. Of the remaining 94 NFMs, only 39% met the institution's established policy criteria. Overall, 28% of all orders entered as nonformulary had a prescribing error. Of these, 49% were incorrect dosages and 33% were misspelled orders. Seventy percent of NFMs were obtained from the institution's standard locations. The institution's nonformulary turnaround time was achieved with 96% of NFMs. CONCLUSIONS: The majority of NFM orders did not meet the institution's established criteria and these orders frequently contained prescribing errors. Routine assessment of NFM use and policy compliance can present opportunities for improvement.

Documenting drug information questions using software for poison information documentation.

PURPOSE: Current and alternative approaches to documenting drug information (DI) questions at DI centers in the United States and features of a software program used for documenting poison information are described. METHODS: A survey was distributed through an Internet listserver for DI specialists in institutional or academic practices. The survey requested information regarding the use of electronic databases to document DI inquiries and the type of databases used. Documented DI questions during a nine-month period using Visual Dotlab (Visual Dotlab Enterprises, Fresno, CA) at the University of Pittsburgh Medical Center (UPMC) DI center were analyzed to demonstrate the feasibility and efficiency of the software for documenting DI questions. RESULTS: Thirty-three DI centers responded to the survey, and 22 (67%) used electronic databases for question documentation. Of these, 15 (68%) were created with Microsoft Office Access (Microsoft Corporation, Redmond, WA). At the UPMC DI center, 841 DI questions were documented using Visual Dotlab. Questions from pharmacists, nurses, and physicians accounted for 654 (78%) of entries. Drug-drug interactions (12%) and dosage recommendations (12%) were the most common types of questions received. On average, DI specialists spent 46 minutes per response, which required an average of 1.7 follow-up calls per inquiry. Quality assurance was performed on 98% of questions documented. CONCLUSION: Visual Dotlab is not well-known and is currently not used widely by DI centers. However, it includes features that may help DI centers document and retrieve DI questions efficiently and comprehensively.