doi: 10.1161/hs1001.097097
(Stroke. 2001;32:2299.)
© 2001 American Heart Association, Inc.
Original Contributions |
Effect of Antiplatelet and Anticoagulant Agents on Risk of Hospitalization for Bleeding Among a Population of Elderly Nursing Home Stroke Survivors
From the Department of Community Health (B.J.Q., V.M., K.L.L.) and Center for Gerontology and Health Care Research (K.L.L., V.M.), Brown University, Providence; and School of Medicine, Department of Family Medicine (C.B.E.), Brown University, Pawtucket, RI.
Correspondence to Kate L. Lapane, PhD, Center for Gerontology and Healthcare Research, Brown University, Box G-B222, Providence, RI 02912. E-mail Kate_Lapane{at}brown.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background and Purpose—— Anticoagulants and antiplatelet agents are underutilized in the nursing home setting, perhaps because trials demonstrating treatment efficacy excluded people resembling those with long-term care needs. We sought to quantify the effect of antiplatelet and anticoagulant agents on risk of hospitalization for bleeding among an elderly nursing home population.
Methods—— We used a case-control design and identified first hospitalizations for bleeds using Medicare claims data from 1992 to 1997 as potential cases. Cases had at least one minimum data set (MDS) assessment within the 6 months before that hospitalization and a diagnosis of stroke. We identified up to 5 controls residing in the same facility during the same year and quarter as the case with a diagnosis of stroke. Our sample consisted of 3433 cases and 13 506 controls. Using the MDS, we identified standing orders for aspirin, dipyridamole, ticlopidine, or warfarin and used conditional logistic regression modeling to estimate the effect of these agents on risk of hospitalization for a bleed.
Results—— After adjustment, use of warfarin (odds ratio [OR], 1.26; 95% CI, 1.11 to 1.43) and combination therapy (OR, 1.34; 95% CI, 0.99 to 1.82) were associated with an increased risk of hospitalization for a bleed compared with nonusers. The odds of aspirin use was greater among cases than controls (OR, 1.07; 95% CI, 0.96 to 1.18) after adjustment.
Conclusions—— Although present, the risk associated with use of these agents is small. The numbers needed to treat to harm 1 resident with aspirin and warfarin were 467 and 126, respectively.
Key Words: anticoagulants • antiplatelet agents • case-control studies • cerebellar hemorrhage • gastrointestinal hemorrhage • nursing homes • stroke
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Stroke is the third leading cause of death among Americans.1 Six hundred thousand persons suffer a stroke each year, of whom 450 000 survive.2 The prevalence of stroke increases with advancing age3,4 and often requires nursing home placement in the 5-year poststroke period.5
Randomized controlled trials and meta-analyses have demonstrated the efficacy of both anticoagulants6 and antiplatelet agents7–10 in preventing instances of subsequent stroke. Despite this evidence, these drugs are underutilized in nursing homes.11 Underuse of these agents may occur because this population differs markedly from participants included in clinical trials in that nursing home residents often have multiple comorbid conditions, cognitive and physical impairment, and polypharmacy.12 Using the data contained in the Systematic Assessment of Geriatric Drug Use Via Epidemiology (SAGE) database from 1992 to 1997, we quantified the effect of antiplatelet and anticoagulant agents on the increased risk of hospitalization for bleeding among elderly nursing home stroke survivors.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We used a case-control study design with data derived from the SAGE database.13,14 In 1987 Congress mandated that all US Medicare/Medicaid-certified nursing facilities perform uniform, comprehensive assessments of all residents as a means to improve care planning. To achieve this goal, the Health Care Financing Administration (HCFA) developed a standardized tool, the Resident Assessment Instrument (RAI), which includes the minimum data set (MDS), a care assessment component, and a drug inventory component.15
Case Identification
We identified cases by inpatient hospitalizations in which the primary diagnosis was for selected bleeds occurring between January 1, 1992, and June 30, 1998 using International Classification of Diseases, Ninth Revision codes detailed in Table 1. We identified 135 631 hospitalizations for bleeding events among nursing home residents, using the first bleeding hospitalization to define case status among persons with multiple hospitalizations. Since the MDS provided exposure history, we identified residents with at least 1 assessment in the 6 months preceding their hospitalization in which drug data were collected with an MDS diagnosis of stroke (n=14 495). For residents with >1 qualifying MDS assessment in the 6 months before their hospitalization, we used the most proximal.
|
Control Identification
We identified residents who successfully cross-linked to the HCFA Health Insurance Skeleton Eligibility Write-off file as potential controls. We matched persons without incident hospitalizations for bleeding events residing within the same facility during the same year and quarter to potential cases such that every case could be matched to a maximum of 5 controls, since hospitalization rates may vary by facility-level factors present in the nursing home.16 We required that potential controls also have a diagnosis of stroke and medication use documented on an MDS assessment within the previous 6 months. In instances in which the potential number of controls residing in the same facility exceeded the number of cases, we randomly selected 5 controls for each case. We did not include cases for whom we could not identify at least 1 eligible control (n=72). Our final matched sample consisted of 3433 cases and 13 506 controls.
Exposure Classification
Trained nursing home staff recorded the name, frequency, and route of administration, whether the order was scheduled or pro re nata (prn), and the National Drug Code (NDC) for up to 18 drugs actually taken by the resident in the previous 7 days.15 We linked the NDC codes to the Master Drug Data Base,17 allowing subclassification into individual drug agents. We classified agents commonly used for the secondary prevention of stroke, although no information was available as to the actual indication for these agents. Among these were standing orders for antiplatelets (aspirin, dipyridamole, or ticlopidine) and anticoagulants (warfarin). With the exception of persons taking aspirin and dipyridamole in combination, we considered users of multiple agents in a separate category. Clopidogrel was not available during the time period from which we obtained exposure information (1992–1997). To evaluate a potential dose-response effect associated with aspirin use, we calculated a daily dose for aspirin for >80% of users based on the dose and frequency of administration.
Analytic Approach
We compared characteristics of cases and controls that might potentially confound the observed association between use of secondary stroke prevention agents and likelihood of hospitalization for a bleed. We hypothesized a priori that the decision to treat, hospitalize, and detect adverse events among elderly stroke survivors may be influenced by both the physical and mental sequelae of ischemic stroke. We examined these domains using 2 previously validated summary scales, the Activities of Daily Living (ADL) scale18 and the Cognitive Performance Scale (CPS).19,20 The ADL scale is a 5-item, 6-level scale based on dependency (dressing, eating, toileting, bathing, locomotion, transferring, and incontinence). We defined limitations as mild (0 to 1 activities impaired), moderate (2 to 3 activities impaired), or severe (4 to 5 activities impaired). The CPS is a summary scale of cognitive impairment.21 We categorized the CPS score as mild (0 to 1), moderate (2 to 3), or severe (4 to 6) impairment. These categories correspond to the Mini-Mental State Examination values of 23 to 24, 12 to 17, and 1 to 6, respectively.20,22 We also considered age,23 alcohol,24 and tobacco use,25 use of nonsteroidal anti-inflammatory drugs (NSAIDs),26,27 use of corticosteroids,28 and use of selective serotonin reuptake inhibitors.29 We considered concurrent antibiotic use, formerly demonstrated to increase the risk of bleeding in persons concomitantly taking warfarin.30
We used conditional logistic regression models to quantify the effect of drug agents on likelihood of hospitalization for bleeding. We constructed these models with hospitalization for a bleed as the outcome variable (ie, case or control status) and derived odds ratios (ORs) and 95% CIs from these models. We also performed subgroup analyses evaluating the effect of these agents on specific bleed etiologies. Finally, we calculated the number needed to treat for harm using methods proposed by Bjerre and LeLorier31 for case-control study designs. Calculation of such a measure requires estimation of the incidence rate of outcome among the unexposed population. Because we were unable to directly estimate this rate from our sample, we estimated the rate from the study base of newly admitted stroke survivors from which our sample arose. These background rates were 30.6 for any serious bleed (95% CI, 28.2 to 32.8), 5.2 for central nervous system bleeds (95% CI, 4.31 to 6.15), and 24.4 for GI bleeds (95% CI, 22.3 to 26.4) per 1000 person-years of follow-up.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The results of analyses examining distributions of demographic and clinical characteristics of the sample at the index MDS assessment are presented in Table 2, stratified by case or control status. Overall, our sample was predominantly female (68.8%) and of white, non-Hispanic descent (80.8%). The percentage of women was slightly higher in the control group than in cases (69.8% versus 64.7%). The largest age group represented in our sample consisted of persons aged >75 years, such that 40.2% of the sample were between the ages of 75 and 84 years and 41.6% were >85 years, comparably distributed across cases and controls. Both physical and cognitive impairments were common, with >58% of the sample demonstrating severe physical impairments and >33% demonstrating severe cognitive impairment. Severe physical impairment was slightly more common in cases than in controls. Concurrent use of NSAIDS, antibiotics, and gastrointestinal (GI) protectants were all more prevalent in cases than in controls (Table 2).
|
Included in Table 3 are the estimates of the effects of other known risk factors for bleeding. Men were more likely to be hospitalized for a bleeding episode than women (OR, 1.30; 95% CI, 1.19 to 1.41). Persons aged 75 to 84 years and participants aged >85 years were both more likely to have a bleed than were younger study participants. Residents developing a bleed were 1.38 times more likely to be dependent in their ADL (95% CI, 1.16 to 1.65) than were controls. Finally, use of NSAIDS, GI protective agents, and corticosteroids were predictive of hospitalizations for bleeding events. Persons taking NSAIDS had a nearly 2-fold increased likelihood of developing an adverse bleeding event (OR, 1.93; 95% CI, 1.68 to 2.23), while persons taking GI protectants were 1.70 times as likely to be hospitalized for a bleed than were persons not taking these agents (95% CI, 1.55 to 1.86). Finally, persons hospitalized for a bleed were 1.23 times more likely to be taking corticosteroids (95% CI, 0.99 to 1.52) than were controls.
|
Findings from crude and multivariable analyses of the main effects are also presented in Table 3. Overall, the use of aspirin and dipyridamole in combination, dipyridamole alone, and ticlopidine alone were relatively rare, with comparable use by both cases and controls. The prevalence of warfarin use was higher among all study participants, with higher percentages of use among cases (12.0%) versus controls (10.2%). Nineteen percent of all study participants had aspirin use recorded on their index MDS assessment (19.0% of cases versus 19.1% of controls). In crude analyses, warfarin and combination users of antiplatelet and anticoagulant agents were associated with a slightly increased likelihood of hospitalization for an adverse bleeding event. These effects persisted after controlling for potentially confounding covariates. The likelihood of hospitalization for a bleed was 1.26 times higher for warfarin users (95% CI, 1.11 to 1.43) and 1.34 times higher in combination users (95% CI, 0.99 to 1.82) compared with nonusers. After controlling for confounders, cases were 1.07 times more likely to have taken aspirin than controls (95% CI, 0.96 to 1.18). Number needed to treat for harm calculations estimate that 467 (95% CI, 182 to 
), 126 (95% CI, 76 to 297), and 96 (95% CI, 40 to ) persons need to be treated with aspirin, warfarin, and combination therapy, respectively, to seriously harm (hospitalization for a bleed) 1 resident. Of those with a known daily aspirin dose, 84% took at least 325 mg/d of aspirin, with similar daily dosages used by both cases and controls. Multivariable effects examining the effect of dose were similar among persons taking at least 325 mg/d (OR, 1.07; 95% CI, 0.95 to 1.20) and among those taking <325 mg/d (OR, 1.15; 95% CI, 0.90 to 1.48) compared with non-users.
Our analyses of specific bleed etiologies are presented in Table 4 and indicate that the likelihood of having taken antiplatelet or anticoagulant agents was increased for cases with central nervous system bleeds relative to matched controls. The odds of aspirin use was 1.36 (95% CI, 1.05 to 1.78) and 1.64 (95% CI, 1.19 to 2.26) for warfarin use in cases compared with controls. The number needed to treat for harm values for central nervous system bleeds were 534 (95% CI, 214 to 3846) for aspirin and 301 (95% CI, 153 to 1012) for warfarin. Findings for GI bleeds alone were less clear, despite the larger sample size (n=2937 cases and n=11 579 matched controls). Cases with GI bleeds were more likely to have taken warfarin (OR, 1.18; 95% CI, 1.03 to 1.36; number needed to treat for harm, 228; 95% CI, 114 to 1366), but our sample did not demonstrate an increased odds of aspirin use among cases hospitalized for GI bleeds relative to matched controls (OR, 1.01; 95% CI, 0.91 to 1.14).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
To our knowledge, this is the first study to comprehensively investigate the risk of hospitalization for bleeding associated with secondary stroke prevention agents in an elderly nursing home population. We found that there is an increased likelihood of being hospitalized for a bleed among older stroke survivors in the nursing home taking aspirin, warfarin, or combination therapy. This effect persisted despite high background rates of GI bleeding events26 and, presumably, central nervous system bleeds in nursing home residents.
There was a slightly elevated occurrence of bleeding events associated with aspirin use, which remained after adjustment. Although not directly comparable, our data are concordant with findings by Petty et al32 demonstrating that bleeding complication rates for aspirin were 3.5 per 100 person-years of observation in a community-dwelling population. In the Swedish Aspirin Low-Dose Trial (SALT), persons taking 75 mg/d of aspirin after an ischemic stroke or transient ischemic attack experienced higher rates of any bleeding event than those on placebo (7.2% versus 3.2%).8 Similarly, in the European Stroke Prevention Study (ESPS), 8.2% of persons taking 50 mg/d of aspirin as secondary prevention had a bleeding event (at any site) compared with only 4.5% in the placebo group.7 A meta-analysis of aspirin randomized controlled trials for both ischemic stroke and myocardial infarction shows an absolute risk of hemorrhagic stroke of 12 events for every 10 000 persons treated.33 Yet, persons included in these trials are by design those least likely to experience adverse events,32 making their generalizability to the clinically complex elderly nursing home population questionable.
Considering the effect of daily aspirin dose is of interest since we could not confirm the commonly held belief that aspirin is safer in lower doses. Use of high-dose aspirin (
325 mg/d) was more common in our sample than use of lower daily doses, with comparable distribution of use across cases and controls. The risk of experiencing an adverse bleed was elevated in both high- (>325 mg/d) and low-dose (
325 mg/d) groups in our sample. Data from the Physicians Health Study showed that male physicians taking 325 mg of aspirin every other day were at a 2-fold increased risk of having a hemorrhagic stroke compared with nonusers.34 As previously discussed, persons randomized to receive lower doses of aspirin in both the SALT8 and ESPS7 randomized controlled trials were at increased risk of adverse bleeding events. Our study comprehensively evaluated the effect of aspirin dose on risk of any adverse bleeding event, although selection bias may exist in these analyses because we were unable to ascertain dose for 16% of aspirin users.
We found there to be an excess of hospitalization for bleeding among persons taking warfarin compared with nonusers. The European Atrial Fibrillation Study (EAFT) assessed the effect of both aspirin and warfarin compared with placebo among persons with nonrheumatic atrial fibrillation and a prior ischemic event on rates of ischemic stroke. During this trial, 8% of participants taking warfarin experienced a major adverse bleeding event compared with 1.4% of those taking placebo.6 By virtue of study design, however, persons with contraindication to warfarin, including concomitant use of NSAIDs or other medications affecting coagulation, were not eligible for randomization to warfarin.6
The international normalized ratio (INR) is a laboratory test used to assess both the efficacy and toxicity of a warfarin regimen. Dosing for warfarin is guided by the desired level of the INR, ultimately driven by the indication for use. Information was not available on either the target INR or the actual INR of persons taking warfarin within our study. Research shows an association between higher target INR and an increased risk of bleeding.30 Samsa and colleagues35 showed that INR values below the therapeutic level are common, possibly indicating a reluctance to aggressively treat with warfarin. The reluctance to adequately dose warfarin may be representative of issues challenging physicians when making the decision to treat. Because underdosing was probable in our sample, the risk of hospitalization for bleeding from warfarin may be higher with more aggressive treatment regimens.
Our data suggest that use of 2 stroke prevention agents simultaneously may also confer an excess risk of bleeding outcomes. The majority of the combination therapy group took aspirin and warfarin (37.6%) or aspirin and ticlopidine (54.7%) in combination. The heterogeneity of this group may preclude the meaningful interpretation of any relative measure of effect. Furthermore, we do not dismiss the possibility that confounding by indication has inflated our estimates. It is probable that persons requiring more intensive therapeutic regimens are sicker and as such are at greater risk for bleeding events.
In addition to our main findings, we demonstrate a greater risk of requiring hospitalization for a bleeding episode among persons taking both NSAIDs and GI protectants. Lapane and colleagues26 previously demonstrated a risk of hospitalization for bleeding among a similar population of elderly nursing home residents taking NSAIDs. It is unclear from our data whether persons taking GI protective agents are at higher risk of developing a bleed or if the prescription of one of these agents is an intermediary between use of antiplatelets or anticoagulants and hospitalization for a bleed.
Limitations
While the study design afforded a unique opportunity to look at the likelihood of hospitalization for a bleeding event in relation to antiplatelet and anticoagulant use, certain methodological concerns must be addressed. The data presented are from 5 states and might have limited generalizability to the institutionalized population of stroke survivors living in US nursing homes.
Since we used hospitalization claims for outcome identification, we may have missed less severe cases for inclusion into our study. Although hemorrhagic strokes are serious events, GI effects may range from minor to life threatening, and therefore the potential for misclassification may be more profound. Regardless, we attempted to create comparability for hospitalization and treatment practices between cases and controls by matching on facility, quarter, and year. Additionally, case status may be a marker for overall better quality of care, including better pharmacotherapy and increased recognition of bleeding events. Conversely, controls may be representative of older persons whose conditions are not closely monitored on an ongoing basis and therefore not recognized or treated.
We based our classification of stroke prevention therapy on one point in time and were unable to determine exposure history for persons before the index MDS assessment or initiation or discontinuation patterns. In addition, we did not have information on the actual indication for use of these agents and presume that they are being used for secondary stroke prevention. A review of the effects of aspirin published by the Fourth American College of Chest Physicians Consensus Conference on Antithrombotic Therapy suggests that the effect of aspirin on bleeding can be both acute and chronic.36 Here we presumably measure an acute effect, since we collected exposure history a maximum of 6 months before index date for cases and controls. We have no information on whether the resident was actually taking the exposure of interest at the time of hospitalization. Exposure misclassification is likely to be similar among the cases and controls, thereby resulting in an attenuation of our measures of effect.
By virtue of study design, we excluded residents with a known hospitalization for bleeding from the sample of potential controls. Furthermore, among the cases we used the first recorded hospitalization for a bleeding event within the study period as the event of interest. Despite this attempt to minimize confounding by prior history of bleeding, it is possible that persons were either hospitalized or treated for a bleed before the time for which claims data were available, potentially diluting our measures of effect.
Conclusions
We previously demonstrated that underuse of secondary stroke prevention agents is common within the elderly nursing home population.11 The risk of recurrent stroke and death within this population is high, thereby making studies that quantify the risk and benefit of treatment vital. Given the present study, the number needed to treat for harm is 467 for aspirin and 126 for warfarin for 6 months, making the absolute risk of bleeding associated with treatment low.
|
Acknowledgments |
|---|
This work was supported by a grant from the Agency for Healthcare Research and Quality (R03-HS11256-01). The authors would like to thank Dr E. Paul Larrat for his thoughtful comments on the manuscript.
|
Footnotes |
|---|
Dr Eaton is on the Speakers Bureaus for Kos Pharmaceuticals, Merck Pharmaceuticals, and Pfizer Pharmaceuticals.
Received April 16, 2001; revision received June 25, 2001; accepted July 5, 2001.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1. Murphy SL. Deaths: Final Data for 1998. National Vital Health Statistics Reports; 48 (11). Hyattsville, Md: National Center for Health Statistics; 2000.
2. American Heart Association. 2001 Heart and Stroke Statistical Update. Dallas, Tex: American Heart Association; 2000.
3.
Sivenius J, Heinonen OP, Pyorala K, Salonen J, Riekkinen P. The incidence of stroke in the Kuopio area of East Finland. Stroke. 1985; 16: 188–192.
4.
Bougousslavsky J, Van Melle G, Regli F. The Lausanne Stroke Registry: analysis of 1,000 consecutive patients with first stroke. Stroke. 1988; 19: 1083–1092.
5.
Leibson CL, Ransom JE, Brown RD, O’Fallon WM, Hass SL, Whishant JP. Stroke-attributable nursing home use: a population-based study. Neurology. 1998; 51: 163–168.
6. EAFT Study Group. Secondary prevention in non-rheumatic atrial fibrillation after transient ischemic attack or minor stroke. Lancet. 1993; 342: 1255–1261.[Medline] [Order article via Infotrieve]
7. Forbes CD, for the ESPS Investigators (European Stroke Prevention Study). Secondary stroke prevention with low-dose aspirin, sustained release dipyridamole alone and in combination. Thromb Res. 1998; 92 (suppl 1): S1–S6.[Medline] [Order article via Infotrieve]
8. SALT Collaborative Group. Swedish Aspirin Low-Dose Trial (SALT) of 75 mg of aspirin as secondary prophylaxis after cerebrovascular ischemic events. Lancet. 1991; 338: 1345–1349.[Medline] [Order article via Infotrieve]
9. Gent M, Blakely JA, Easton DJ, Ellis DJ, Hachinski VC, Harbison JW, Panak E, Roberts RS, Sicurella J, Turpie AG. The Canadian American Ticlopidine Study (CATS) in thromboembolic stroke. Lancet. 1989; 1: 1215–1220.[Medline] [Order article via Infotrieve]
10. CAPRIE Steering Committee. A randomized, blinded, trial of clopidogrel versus aspirin in patients at risk of ischemic events (CAPRIE). Lancet. 1996; 348: 1329–1338.[Medline] [Order article via Infotrieve]
11.
Quilliam BJ, Lapane KL. Clinical correlates and drug treatment of residents with stroke in long-term care. Stroke. 2001; 32: 1385–1393.
12. Gambassi G, Landi F, Peng L, Brostrup Jensen C, Calore K, Hiris J, Lipsitz L, Mor V, Bernabei R, for the SAGE Study Group (Systematic Assessment of Drug Use Via Epidemiology). Validity of diagnostic and drug data in standardized nursing home resident assessments: potential for geriatric pharmacoepidemiology. Med Care. 1998; 36: 167–179.[Medline] [Order article via Infotrieve]
13. Bernabei R, Gambassi G, Lapane KL, Sgadari A, Landi F, Gatsonis C, Lipsitz L, Mor V, for the SAGE (Systematic Assessment of Geriatric Drug Use Via Epidemiology) Study Group. Characteristics of the SAGE database: a new resource for research on outcomes in long-term care. J Gerontol A Biol Sci Med Sci. 1999; 54: M25–M33.
14. Bernabei R, Gamabssi G. The SAGE database: introducing functional outcomes in geriatric pharmaco-epidemiology. J Am Geriatr Soc. 1998; 46: 251–252.[Medline] [Order article via Infotrieve]
15. Morris JN, Hawes C, Fries BE, Phillips CD, Mor V, Katz S, Murphy K, Drugovich ML, Friedlob AS. Designing the national resident assessment instrument for nursing homes. Gerontologist. 1990; 30: 293–307.[Abstract]
16. Intrator O, Castle NG, Mor V. Facility characteristics associated with hospitalization of nursing home residents: results of a national study. Med Care. 1999; 37: 228–237.[Medline] [Order article via Infotrieve]
17. First Data Bank Inc. Master Drug Database (MDDB) Documentation Manual. Indianapolis, Ind: First Data Bank Inc; 1995.
18. Hawes C, Morris JN, Phillips CD, Mor V, Fries BE, Nonemaker S. Reliability estimates for the minimum data set for nursing home resident assessment and care screening (MDS). Gerontologist. 1995; 35: 172–178.[Abstract]
19. Hartmaier SL, Sloane PD, Guess HA, Koch GG, Mitchell CM, Phillips CD. Validation of the Minimum Data Set Cognitive Performance Scale: agreement with the Mini-Mental State Examination. J Gerontol A Biol Sci Med Sci. 1995; 50: M128–M133.
20. Hartmaier SL, Sloane PD, Guess HA, Koch GG. The MDS Cognition Scale: a valid instrument for identifying and staging nursing home residents with dementia using the minimum data set. J Am Geriatr Soc. 1994; 42: 1173–1179.[Medline] [Order article via Infotrieve]
21. Morris JN, Fries BE, Mehr DR, Hawes C, Phillips C, Mor V, Lipsitz LA. MDS Cognitive Performance Scale. J Gerontol A Biol Sci Med Sci. 1994; 49: M174–M182.
22. HCFA. Minimum Data Set V 2 User’s Manual. Natick, Mass: Eliot Press; 1999.
23.
Gorter JW. Major bleeding during anticoagulation after cerebral ischemia: patterns and risk factors: Stroke Prevention in Reversible Ischemia Trial (SPIRIT). Neurology. 1999; 53: 1319–1327.
24. Parsells-Kelly J, Kaufman D, Koff RS, Laszlo A, Wiholm BE, Shapiro S. Alcohol consumption and the risk of major upper gastrointestinal bleeding. Am J Gastroenterol. 1995; 90: 1058–1064.[Medline] [Order article via Infotrieve]
25.
Svanes C, Soreide JA, Skarstein A, Fevang BT, Bakke P, Vollset SE, Svanes K, Soreide O. Smoking and ulcer perforation. Gut. 1997; 41: 177–180.
26. Lapane KL, Spooner JJ, Mucha L, Strauss WL. Effect of nonsteroidal anti-inflammatory drug use on the rate of gastrointestinal hospitalizations among people living in long-term care. J Am Geriatr Soc. 2001; 49: 577–584.[Medline] [Order article via Infotrieve]
27. Garcia-Rodriguez LA, Jick H. Risk of upper gastrointestinal bleeding and perforation associated with individual non-steroidal anti-inflammatory drugs. Lancet. 1994; 343: 769–772.[Medline] [Order article via Infotrieve]
28.
Ellershaw JE. Corticosteroids and peptic ulceration. Palliat Med. 1994; 8: 313–319.
29.
De-abajo FJ, Garcia-Rodriguez LA, Montero D. Association between selective serotonin reuptake inhibitors and upper gastrointestinal bleeding: a population based case-control study. BMJ. 1999; 319: 1106–1109.
30. Panneerselvam S, Baglin C, Lefort W, Baglin T. Analysis of risk factors for over-anticoagulation in patients receiving long-term warfarin. Br J Haematol. 1998; 103: 422–424.[Medline] [Order article via Infotrieve]
31.
Bjerre LM, LeLorier J. Expressing the magnitude of adverse effects in case-control studies: "the number needed to be treated for one additional patient to be harmed." BMJ. 2000; 320: 503–506.
32.
Petty G, Brown R, Whisnant J, Sicks J, O’Fallon WM, Wiebers D. Frequency of major complications of aspirin, warfarin, and intravenous heparin for secondary stroke prevention. Ann Intern Med. 1999; 130: 14–22.
33.
He J, Whelton PK, Vu B, Klag MJ. Aspirin and risk of hemorrhagic stroke: a meta-analysis of randomized controlled trials. JAMA. 1998; 280: 1930–1935.
34. Steering Committee of the Physicians’ Health Study Research Group. Final report on the aspirin component of the ongoing Physicians’ Health Study. N Engl J Med. 1989; 321: 129–135.[Abstract]
35.
Samsa GP, Matchar DB, Goldstein LB, Bonito AJ, Lux LJ, Witter DM, Bian J. Quality of anticoagulant management among patients with atrial fibrillation. Arch Intern Med. 2000; 160: 967–973.
36.
Hirsh J, Dalen JE, Fuster V, Harker LB, Patrono C, Roth G. Aspirin and other platelet active drugs: the relationship among dose, effectiveness, and side effects. Chest. 1995; 108 (suppl 4): 247S–257S.
This article has been cited by other articles:
 |
|
 |
|
  R. M. Holt, F. L. Schwartz, and J. H. Shubrook Diabetes Care in Extended-Care Facilities: Appropriate intensity of care? Diabetes Care, June 1, 2007; 30(6): 1454 - 1458. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. I. Shireman, J. D. Mahnken, P. A. Howard, T. F. Kresowik, Q. Hou, and E. F. Ellerbeck Development of a contemporary bleeding risk model for elderly warfarin recipients. Chest, November 1, 2006; 130(5): 1390 - 1396. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. N. Papanikolaou, G. D. Christidi, and J. P.A. Ioannidis Comparison of evidence on harms of medical interventions in randomized and nonrandomized studies Can. Med. Assoc. J., February 28, 2006; 174(5): 635 - 641. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. B. Christian, K. L. Lapane, and R. S. Toppa Racial Disparities in Receipt of Secondary Stroke Prevention Agents Among US Nursing Home Residents Stroke, November 1, 2003; 34(11): 2693 - 2697. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. M. Bjerre, J. LeLorier, B. J. Quilliam, and K. L. Lapane Timeless * Response Stroke, July 1, 2002; 33(7): 1743 - 1744. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||