Table 3.

Prescriber attitudes and concerns with warfarin use

Study	Study objective, (intervention/exposure and outcomes)	Study design, data source	Study population, study setting, time period	Results	Quality assessment, funding source
Dharmarajan et al. (2006) [27]	To evaluate the decision whether or not to anticoagulate among physicians in practice and in various levels of training (residents and fellows) for a specific, yet not unusual, case scenario in the nursing home	Design: cross-sectional study Data source: survey questionnaire based on an actual case from an LTC facility. The subject of the survey was an 87-year-old female LTC facility resident with dementia, AF, and history of hip fracture who suffered a recent fall without fracture	Population: 107 completed surveys were returned from 49 residents, 20 fellows, and 38 attending physicians Setting: a university teaching hospital in the Bronx, NY (US) Time period: survey dates not specified	The majority of physicians (85%) thought that long-term anticoagulation therapy was not indicated in the case patient. However, most (88%) said they would provide an antiplatelet agent (78% aspirin, 20% clopidogrel). The most cited reasons for not providing anticoagulation were risk of falls (98%), dementia (40%), and short life expectancy (32%). 92% of respondents said the patient was a candidate for short-term anticoagulation therapy. Responses to the questions were similar for all physicians (or faculty) irrespective of level of training or years in practice	Quality assessment for observational studies: 1) Unbiased selection of the cohort? Cannot be determined; cohort selection details not provided; non-response rate not disclosed 2) Selection minimizes baseline differences in prognostic factors? Cannot be determined 3) Sample size calculated/5% difference? No 4) Adequate description of the cohort? No; details of cohort other than practice specialty were not provided 5) Validated method for ascertaining exposure? No; reliability concern since limited to 1 case study; content validity of case study not described 6) Validated method for ascertaining clinical outcomes? No; validation assessment of response choices not performed 7) Outcome assessment blind to exposure? NA 8) Adequate follow-up period? Yes; cross-sectional 9) Completeness of follow-up? Yes 10) Analysis controls for confounding? Partial; cross-tabulations performed on responses by specialty 11) Analytic methods appropriate? Yes Funding: Geriatric Medicine Fellowship program
Harrold et al. (2002) [28]	To examine physician attitudes regarding the use of specialized anticoagulation services in the LTC setting	Design: cross-sectional study Data source: survey questionnaire	Population: 245 physicians asked to participate in the survey; 114 (47%) responded. 91 reported that they currently cared for residents in LTC facilities and thus completed the questionnaire Setting: 21 LTC facilities in Connecticut (US) Time period: Nov 1999 - Jan 2000	The majority of respondents agreed or strongly agreed that an anticoagulation service would reduce the workload on physicians (76%), and increase the percent of time that nursing home residents on warfarin are maintained in the target therapeutic range (54%). 53% disagreed or strongly disagreed with statements suggesting that this service would reduce the risk of warfarin-related bleeding. 45% of respondents agreed with a statement that this service would intrude on physician decision-making. 53% of the respondents said they might use an anticoagulation service for managing their LTC patients on warfarin. The most desirable aspects of an anticoagulation service were surveillance for drug interactions (65%), scheduling of laboratory tests (48%), management of warfarin dosing (45%), and risk assessment for bleeding (40%). The most frequently cited challenges to managing warfarin therapy in the nursing home setting were dealing with medications that interact with warfarin (59%), keeping patients within target therapeutic range (53%), and making dosage adjustments (30%)	Quality assessment for observational studies: 1) Unbiased selection of the cohort? Yes 2) Selection minimizes baseline differences in prognostic factors? Yes; performed analysis of non-responders 3) Sample size calculated/5% difference? No 4) Adequate description of the cohort? Yes 5) Validated method for ascertaining exposure? No; minimal description of anticoagulation services provided 6) Validated method for ascertaining clinical outcomes? Validation of new questionnaire not reported 7) Outcome assessment blind to exposure? NA 8) Adequate follow-up period? Yes; cross-sectional 9) Completeness of follow-up? Yes 10) Analysis controls for confounding? No; cross tabulation with subject attributes not performed 11) Analytic methods appropriate? No; statistical error (CIs) reported in only some findings Funding: Centers for Medicare and Medicaid Services, Department of Health and Human Services; AHRQ
Monette et al. (1997) [29]	To assess the knowledge and attitudes of physicians regarding the use of warfarin for stroke prevention in patients with AF in LTC facilities	Design: cross-sectional study Data source: survey questionnaire of 2 clinical scenarios with substantial contrasts in patient characteristics: 1) 94-year old male resident with chronic AF, ischemic heart disease, CHF and osteoarthritis, no history of falls, independent in activities of daily living; 2) 80-year old female with recent stroke with resulting hemiplegia and dysarthria, having chronic AF, CHF, CAD, hypertension, diabetes, and chronic renal insufficiency, with cognitive deficits and entirely nonambulatory	Population: 269 physicians were asked to participate in the survey; 182 (67.7%) completed the questionnaire Setting: 30 LTC facilities located in New England, Quebec, and Ontario (US and Canada) Time period: Feb 1995 to Jul 1995	Only 47% of respondents indicated that the benefits of warfarin greatly outweigh the risks in this setting; the remainder of physicians indicated that benefits only slightly outweigh the risks (34%) or that risks outweigh benefits (19%). The most frequently cited contraindications to warfarin use were: excessive risk of falls (71%), history of GI bleeding (71%), history of non-CNS bleeding (36%), and history of cerebrovascular hemorrhage (25%). Among the 164 physicians who reported using the INR to monitor warfarin therapy, 27% indicated a target range with a lower limit < 2.0, 71% indicated a target range between 2.0 and 3.0, and 2% indicated an upper limit > 3.0. Among respondents who answered questions about the clinical scenarios, estimates of the risk of stroke without warfarin therapy and the risk of intracranial hemorrhage with therapy varied widely	Quality assessment for observational studies: 1) Unbiased selection of the cohort? Yes 2) Selection minimizes baseline differences in prognostic factors? No; analysis of non-responders was not performed 3) Sample size calculated/5% difference? No 4) Adequate description of the cohort? Yes 5) Validated method for ascertaining exposure? Yes; conducted pre-testing to establish content validity 6) Validated method for ascertaining clinical outcomes? No further assessment validity conducted for new questionnaire 7) Outcome assessment blind to exposure? NA 8) Adequate follow-up period? Yes; cross-sectional 9) Completeness of follow-up? Yes 10) Analysis controls for confounding? Yes 11) Analytic methods appropriate? No; statistical error (CIs) reported in only some findings Funding: Dupont Pharma

AF, atrial fibrillation; AHRQ, Agency for Healthcare Research and Quality; CAD, coronary artery disease; CHF, congestive heart failure; CI confidence interval; CNS, central nervous system; GI, gastrointestinal; INR, international normalized ratio; LTC, long-term care; NA not available.