Reasons for undertreatment with oral anticoagulants in frail geriatric outpatients with atrial fibrillation: a prospective, descriptive study.

OBJECTIVES: The main aims of the study were to explore whether oral anticoagulation (OAC) for atrial fibrillation (AF) in geriatric outpatients is prescribed in accordance with international (American College of Cardiology/American Heart Association/European Society of Cardiology [ACC/AHA/ESC]) and Dutch national guidelines for the general practitioner (GP) and to identify whether age and selected co-morbid conditions are associated with undertreatment. As a secondary objective, we wanted to establish how many patients discontinue OAC because of major bleeding. METHODS: In 2004, at the first visit of all patients to the geriatric day clinic of the Slotervaart Hospital in Amsterdam, the Netherlands, demographic data, Mini-Mental State Examination score, medical history, Charlson Comorbidity Index score, and data on medication use and changes were documented. The presence of AF was established by assessment of medical history information obtained by the GP, the history taken from patients and their caregivers, and the results of clinical evaluation, including ECG findings. Associations between the use of OAC, demographic data and co-morbid conditions registered in the Dutch NHG (Nederlands Huisartsen Genootschap [Dutch College of General Practitioners]) standard for GPs as risk factors for stroke or contraindications to the use of OAC were analysed. The reasons for discontinuing OAC were assessed after 4 years by requesting the information from the anticoagulation services or the GP. RESULTS: At the time of the initial visit, 17.5% of the 807 outpatients had chronic AF (n = 135) or were known to have paroxysmal AF (n = 6). The mean age of the 141 patients in this cohort was 84.3 years (SD 6.2 years). Co-morbid conditions increasing the risk of stroke were present in 129 patients (91.5%). Contraindications to the use of OAC were observed in 118 patients (83.7%). Of the 116 patients with AF in their history before their visit, 57.8% were being treated with OAC at the time of their visit. After comprehensive geriatric assessment, 73 (51.8%) of the 141 patients with chronic or paroxysmal AF were continued on OAC. Of the 141 patients with chronic or paroxysmal AF, 110 (78.0%) had both extra stroke risk factors and contraindications to the use of OAC. Only increasing age was significantly and independently associated with not being prescribed anticoagulants (p < 0.001). At the 4-year follow-up, OAC had been discontinued in 5.5% of patients because of major bleeding; three patients (4.1%) taking OAC had died as a result of major bleeding, and one other patient had discontinued treatment because of a major, non-lethal bleeding episode. CONCLUSION: Applying the NHG standard for appropriate prescription, and disregarding age as a risk factor or contraindication, in this population, 14 of 141 patients (9.9%) were inappropriately prescribed OAC, salicylates or no prophylaxis. Since only patient age was associated with not prescribing OAC in this study, higher age still seems to be considered the most important contraindication to anticoagulation therapy. Implementation of better models for stratifying bleeding risk in the frail elderly is needed. After 4 years, the cumulative rate of bleeding causing discontinuation of anticoagulation therapy in this usual-care study of frail older patients was not alarmingly higher than in other usual-care studies.

Drug-drug interactions in a geriatric outpatient cohort: prevalence and relevance.

BACKGROUND: The prevalence of drug-drug interactions (DDIs) in a geriatric population may be high because of polypharmacy. However, wide variance in the clinical relevance of these interactions has been shown. OBJECTIVES: To explore whether adverse drug reactions (ADRs) as a result of DDIs can be identified by clinical evaluation, to describe the prevalence of ADRs and diminished drug effectiveness as a result of DDIs and to verify whether the top ten most frequent potential DDIs known to public pharmacies are of primary importance in geriatric outpatients in the Netherlands. METHOD: All adverse events classified by the Naranjo algorithm as being a possible ADR and drug combinations resulting in diminished drug effectiveness were identified prospectively in 807 geriatric outpatients (mean age 81 years) at their first visit. The setting was a diagnostic day clinic. The Medication Appropriateness Index (MAI) and Beers criteria were used to evaluate drug use and identify possible DDIs. The ten most frequent potential interactions, according to a 1997 national database of public pharmacies ('Top Ten') in the Netherlands, and possible adverse events as a result of other interactions, were described. The effects of changes in medication regimen were recorded by checking the medical records. RESULTS: In 300 patients (44.5% of the 674 patients taking more than one drug), 398 potential DDIs were identified. In 172 (25.5%) of patients taking more than one drug, drug combinations were identified that were responsible for at least one ADR or which possibly resulted in reduced effectiveness of therapy. Eighty-four of the 158 possible ADRs resulting from enhanced action of drugs forming combinations listed in the 'Top Ten' were seen in 73 patients. Only four DDIs resulting in less effective therapy that involved drug combinations in the 'Top Ten' were identified. Changes in drug regimens pertaining to possible interactions were proposed or put into effect in 111 of the 172 (65%) patients with possible DDIs. Sixty-one (55%) of these patients returned for follow-up. Of these, 49 (80%) were shown to have improved after changes were made to their medication regimen. CONCLUSION: In this study, nearly half of the geriatric outpatients attending a diagnostic day clinic who were taking more than one drug were candidates for DDIs. One-quarter of these patients were found to have possible adverse events or diminished treatment effectiveness that may have been at least partly caused by these DDIs. These potential interactions can be identified through clinical evaluation. In the majority of patients (99 of 172) the potential interactions resulting in possible ADRs or diminished effectiveness were not present in the 'Top Ten' interactions described by a national database of public pharmacies, a finding that emphasizes that the particular characteristics of geriatric patients (e.g. frequent psychiatric co-morbidities) need to be considered when evaluating their drug use. At least 7% of all patients taking more than one drug, and 80% of those with possible drug interactions whose drug regimen was adjusted, benefited from changes made to their drug regimens.