Checklist for standardized reporting of drug-drug interaction management guidelines.

PURPOSE: Inconsistencies and omissions in drug-drug interaction (DDI) management guidelines may lead to harm and suboptimal therapy. The purpose of this study was to define a checklist for DDI management guidelines to help developers produce high-quality guidelines that will support healthcare providers in clinical practice. METHODS: We carried out a two-round Delphi process with an international panel of healthcare providers, most of whom are pharmacists involved in providing DDI information, in order to select those items that should be addressed in DDI management guidelines (including grading systems that could be used). RESULTS: Twenty-three panellists reached consensus on 19 items in two main domains. These were consolidated into a checklist of 15 elements for standardized reporting in management guidelines. For each element a description is provided to specify what information should be documented in that specific element. CONCLUSIONS: It was possible to reach a broad consensus on which relevant items should be included in a checklist for the development of DDI management guidelines.

Ineffectiveness and adverse events of nitrofurantoin in women with urinary tract infection and renal impairment in primary care.

PURPOSE: To determine whether treatment with nitrofurantoin in women with urinary tract infection (UTI) and renal impairment in primary care is associated with a higher risk of ineffectiveness and/or serious adverse events than in women without renal impairment. METHODS: A cohort of 21,317 women treated with nitrofurantoin and a cohort of 7,926 women treated with trimethoprim, identified from the Pharmo Record Linkage System, were analysed. The primary outcome was ineffectiveness of treatment of nitrofurantoin defined as the start of a second antibacterial within 1 month after the start of nitrofurantoin. The secondary outcome was the occurrence of serious adverse events of nitrofurantoin leading to hospitalization within 90 days. A cohort of trimethoprim users was used to determine if the associations found for nitrofurantoin were mainly related to nitrofurantoin itself. The association between renal impairment and the risk of these outcomes was determined with Cox regression and expressed as hazard ratios (HRs). RESULTS: Overall, the incidence density for ineffectiveness was 5.4 per 1,000 person-days, and moderate renal impairment was not associated with ineffective treatment [HR 1.1, 95 % confidence interval (CI) 0.74-1.51]. The overall incidence density for adverse events was 0.02 per 1,000 person-days. In patients with renal impairment (<50 ml/min/1.73 m²) the risk of pulmonary adverse events leading to hospitalization was significantly increased (HR 4.1, 95 % CI 1.31-13.09) CONCLUSIONS: Nitrofurantoin treatment was not associated with a higher risk of ineffectiveness in women with UTI and moderate renal impairment (30-50 ml/min/1.73 m²). However, we did find a significant association between renal impairment (<50 ml/min/1.73 m²) and pulmonary adverse events leading to hospitalization.

Information comparison of the effects of drugs on laboratory tests in drug labels and Young's book.

BACKGROUND: The effects of drugs on laboratory tests may lead to misinterpretation of laboratory data, unnecessary tests, higher costs and missed diagnoses. This study compared the information on drug-laboratory effects (DLE) described in 200 drug labels with that in Young's book. METHODS: Information on DLE was searched in the drug labels of 200 frequently prescribed drugs using the keywords 'interfer*', 'influence', and 'laborator*'. This information was compared with the information in Young's book. Each item of information scored 1 point if it was specific and exactly the same. Primary outcome was the percentage of DLE with completely the same information. RESULTS: In 23 (11.5%) of the 200 drug labels 83 DLE were described. Most DLE were described in drug labels of contraceptives (71%) and antibacterials (15%). The most frequently affected laboratory tests were adrenal gland (17%), urine tests (15%), liver tests (10%) and renal function tests (10%). Comparison of six DLE with Young's book was not possible because the information was not described in the book. Twelve (14.5%) DLE of the information in the drug label was identical to that in Young's book. Detailed information about nature of the effect, strength of the effect and body fluid was not described in the drug labels. CONCLUSIONS: In a limited number of DLE in the drug labels the information was the same as in Young's book. Overall, the information on DLE provided in drug labels is unclear, inconsistent and incomplete and does not support healthcare professionals in making evidence-based monitoring decisions.

A pharmacy medication alert system based on renal function in older patients.

BACKGROUND: Patients with diabetes or cardiovascular disease are at risk of reduced renal function and frequently use drugs that interact with renal function. GPs monitor renal function in these patients. Computerised prescription systems produce alerts in patients labelled as having chronic kidney disease, but alerts are often ignored. If pharmacists use a pharmacy medication alert system (PMAS) based on renal function, they can provide the GP with therapeutic advice to optimise the medication. The extent of this advice and the feasibility in the clinical context are unknown. AIM: To assess the therapeutic advice formulated by pharmacists with help of a PMAS based on the renal function of patients aged ≥70 years with diabetes or cardiovascular disease. DESIGN AND SETTING: Observational study in primary health care in the Netherlands. METHOD: GPs provided pharmacists with the renal function of older patients with diabetes or cardiovascular disease who were using target drugs, that is, drugs requiring therapeutic advice in patients with reduced renal function. With the help of a PMAS, pharmacists assessed the actual medication. The GP weighed the advice in relation to the clinical context of the individual patient. RESULTS: Six hundred and fifty patients were prescribed 1333 target drugs. Pharmacists formulated 143 therapeutic recommendations (11% of target drugs) concerning 89 patients (13.7% of study population). In 71 recommendations in 52 patients (8.0% of study population), the GP agreed immediately. CONCLUSION: The use of a PMAS resulted in therapeutic advice in 11% of the target drugs. After weighing the clinical context, the GP agreed with half of the advice.

Instructions on laboratory monitoring in 200 drug labels.

BACKGROUND: Monitoring drug treatment is important to assess the therapeutic effects and to prevent adverse drug reactions. Unfortunately, the clinical evidence for monitoring is often missing. To attain evidence-based laboratory monitoring and to improve patient safety it is mandatory for the clinical chemist to develop effective and rational methods for monitoring. The legal source for this evidence-based information is the drug label. We analysed frequency, nature, and applicability of instructions on laboratory monitoring described in 200 drug labels. METHODS: The applicability of instructions was assessed with an adapted Systematic Information for Monitoring score. Seven items of information were evaluated: why to monitor, what to monitor (essential), when to start or stop monitoring, how frequently to monitor, critical value (essential) and how to respond (essential). Each item scored one point when information was described specifically, otherwise the score was zero. Instructions were applicable if all three essential items scored. RESULTS: In 131 drug labels, 566 instructions on laboratory monitoring were identified, an average of 2.8 per drug label. Kidney, liver, electrolyte, and drug monitoring were important biomarker categories (71%). The median applicability score was 2.1 (0-6) and 95 (17%) instructions were applicable. Six determinants were associated with applicable instructions: kidney (OR 7.0; 95% CI 4.4-11.3), creatine phosphokinase (4.5; 1.5-13.6), drug selection (6.8; 4.0-11.7), dose adjustments (2.4; 1.5-3.7), year on the market 2000-2007 (2.6; 1.1-6.1) and statins (4.8; 2.5-9.0). CONCLUSIONS: Drug labels frequently describe instructions on laboratory monitoring, but these are ambiguous and incomplete and clinical applicability for the professional is limited.

Laboratory tests in the clinical risk management of potential drug-drug interactions: a cross-sectional study using drug-dispensing data from 100 Dutch community pharmacies.

BACKGROUND: Patient safety and the life cycle of a drug are negatively influenced by the still increasing occurrence of potential drug-drug interactions (DDIs). Clinical risk management of potential DDIs is required in patients using drugs to influence the benefit-risk profile positively. Information about laboratory test results, in particular, may be useful in the assessment of potential DDIs for the individual patient. OBJECTIVE: The objective of this study was to examine the frequency and nature of laboratory tests required for the assessment of the clinical relevance of potential DDIs in Dutch community pharmacies. In addition, the nature and clinical relevance of these potential DDIs is analysed. METHODS: All patients from 100 Dutch community pharmacies using, according to dispensing information, two or more drugs concomitantly on a specified date (Wednesday, 4 April 2007), were included (n = 223,019). The anonymous dispensing data of the included patients were analysed against a list of DDIs requiring laboratory tests for the assessment of their clinical relevance. The number of patients at risk for these potential DDIs with severe adverse reactions was calculated. The frequency of potential DDIs requiring laboratory tests were stratified by age, sex and degree of polypharmacy. RESULTS: Of the included patients, 24.4% had one or more potential DDIs (n = 54,427). In 9.0% of the included patients, one or more laboratory tests for the assessment of clinical relevance of the potential DDI were required (n = 19,968). The frequency of DDIs requiring laboratory tests increased with increasing age and number of drugs, but was not related to sex. The most commonly required laboratory tests were for renal function (42.2%), electrolytes (20.1%) and coagulation (13.1%). The percentage of patients at risk for potential DDIs requiring laboratory tests with adverse reaction category F (serious, irrecoverable disablement or death) was 2.5%; category E (increased risk of failure of life-saving therapy) was 0.6%; and category D (inconvenience with residual symptom and failure of therapy concerning serious but non-fatal diseases) was 3.8%. CONCLUSIONS: A large number of patients in Dutch community pharmacies are at risk for potential DDIs requiring laboratory tests for the assessment of the clinical relevance of the interaction. There is a strong relationship between the frequency of DDIs requiring laboratory tests and age and the number of drugs concomitantly used. In the clinical risk management of potential DDIs, information about laboratory test results is of additional value. Future research is necessary in order to obtain more evidence on using laboratory tests in terms of which tests should be linked to pharmacy data, in which patients they should be done, how often and what actions should be taken when an abnormal value is found.