Determining clinically important differences in health status measures: a general approach with illustration to the Health Utilities Index Mark II

The objective of this article was to describe and illustrate a comprehensive approach for estimating clinically important differences (CIDs) in health-related quality-of-life (HR-QOL). A literature review and pilot study were conducted to determine whether effect size-based benchmarks are consistent with CIDs obtained from other approaches. CIDs may be estimated based primarily upon effect sizes, supplemented by more traditional anchor-based methods of benchmarking (i.e. direct, cross-sectional or longitudinal approaches). A literature review of articles discussing CIDs provided comparative data on effect sizes for various chronic conditions. A pilot study was then conducted to estimate the minimum CID of the Health Utilities Index (HUI) Mark II, and to compare the observed between-group differences observed in a recent randomised trial of an acute stroke intervention with this benchmark. The use of standardised effect size benchmarks has a number of advantages-for example, effect sizes are efficient, widely accepted outside HR-QOL, and have well accepted benchmarks based upon external anchors. In addition, our literature review and pilot study suggest that effect size-based CID benchmarks are similar to those which would be obtained using more traditional methods. For most HR-QOL instruments, we do not know the changes in score which constitute CIDs of various magnitudes. This makes interpretation of HR-QOL results from clinical trials difficult, and having a benchmarking process which is relatively straightforward would be highly desirable.

Likelihood ratios with confidence: sample size estimation for diagnostic test studies

Confidence intervals are important summary measures that provide useful information from clinical investigations, especially when comparing data from different populations or sites. Studies of a diagnostic test should include both point estimates and confidence intervals for the tests' sensitivity and specificity. Equally important measures of a test's efficiency are likelihood ratios at each test outcome level. We present a method for calculating likelihood ratio confidence intervals for tests that have positive or negative results, tests with non-positive/non-negative results, and tests reported on an ordinal outcome scale. In addition, we demonstrate a sample size estimation procedure for diagnostic test studies based on the desired likelihood ratio confidence interval. The renewed interest in confidence intervals in the medical literature is important, and should be extended to studies analyzing diagnostic tests.

Accuracy of the Papanicolaou test in screening for and follow-up of cervical cytologic abnormalities: a systematic review

PURPOSE

To evaluate the accuracy of conventional and new methods of Papanicolaou (Pap) testing when used to detect cervical cancer and its precursors.

DATA SOURCES

Systematic search of English-language literature through October 1999 using MEDLINE, EMBASE, other computerized databases, and hand searching.

STUDY SELECTION

All studies that compared Pap testing (conventional methods, computer screening or rescreening, or monolayer cytology) with a concurrent reference standard (histologic examination, colposcopy, or cytology).

DATA EXTRACTION

Two reviewers independently reviewed selection criteria and completed 2 x 2 tables for each study.

DATA SYNTHESIS

94 studies of the conventional Pap test and three studies of monolayer cytology met inclusion criteria. No studies of computerized screening were included. Data were organized by cytologic and histologic thresholds used to define disease. For conventional Pap tests, estimates of sensitivity and specificity varied greatly in individual studies. Methodologic quality and frequency of histologic abnormalities also varied greatly between studies. In the 12 studies with the least biased estimates, sensitivity ranged from 30% to 87% and specificity ranged from 86% to 100%.

CONCLUSIONS

Insufficient high-quality data exist to estimate test operating characteristics of new cytologic methods for cervical screening. Future studies of these technologies should apply adequate reference standards. Most studies of the conventional Pap test are severely biased: The best estimates suggest that it is only moderately accurate and does not achieve concurrently high sensitivity and specificity. Cost-effectiveness models of cervical cancer screening should use more conservative estimates of Pap test sensitivity.

Measurement of motor recovery after stroke. Outcome assessment and sample size requirements

BACKGROUND AND PURPOSE

The purpose of this study was to analyze recovery of motor function in a cohort of patients presenting with an acute occlusion in the carotid distribution. Analysis of recovery patterns is important for estimating patient care needs, establishing therapeutic plans, and estimating sample sizes for clinical intervention trials.

METHODS

We prospectively measured the motor deficits of 104 stroke patients over a 6-month period to identify earliest measures that would predict subsequent motor recovery. Motor function was measured with the Fugl-Meyer Assessment. Fifty-four patients were randomly assigned to a training set for model development; 50 patients were assigned to a test set for model validation. In a second analysis, patients were stratified on basis of time and stroke severity. The sample size required to detect a 50% improvement in residual motor function was calculated for each level of impairment and at three points in time.

RESULTS

At baseline the initial Fugl-Meyer motor scores accounted for only half the variance in 6-month motor function (r2 = 0.53, p less than 0.001). After 5 days, both the 5-day motor and sensory scores explained 74% of the variance (p less than 0.001). After 30 days, the 30-day motor score explained 86% of the variance (p less than 0.001). Application of these best models to the test set confirmed the results obtained with the training set. Sample-size calculations revealed that as severity and time since stroke increased, sample sizes required to detect a 50% improvement in residual motor deficits decreased.

CONCLUSIONS

Most of the variability in motor recovery can be explained by 30 days after stroke. These findings have important implications for clinical practice and research.

Mathematical model for the natural history of human papillomavirus infection and cervical carcinogenesis

The authors constructed a Markov model as part of a systematic review of cervical cytology conducted at the Duke University Evidence-based Practice Center (Durham, North Carolina) between October 1997 and September 1998. The model incorporated states for human papillomavirus infection (HPV), low- and high-grade squamous intraepithelial lesions, and cervical cancer stages I-IV to simulate the natural history of HPV infection in a cohort of women from ages 15 to 85 years. The age-specific incidence rate of HPV, and regression and progression rates of HPV and squamous intraepithelial lesions, were obtained from the literature. The effects of varying natural history parameters on cervical cancer incidence were evaluated by using sensitivity analysis. The base-case model resulted in a lifetime cervical cancer risk of 3.67% and a lifetime cervical cancer mortality risk of 1.26%, with a peak incidence of 81/100,000 at age 50 years. Age-specific distributions of precursors were similar to reported data. Lifetime risk of cancer was most sensitive to the incidence of HPV and the probability of rapid HPV progression to high-grade lesions (two- to threefold variations in risk). The model approximates the age-specific incidence of cervical cancer and provides a tool for evaluating the natural history of HPV infection and cervical cancer carcinogenesis as well as the effectiveness and cost-effectiveness of primary and secondary prevention strategies.

Similar motor recovery of upper and lower extremities after stroke

BACKGROUND AND PURPOSE

This study examined the validity of the clinical tenet that poststroke recovery of the upper extremity is less rapid and complete than poststroke recovery of the lower extremity. Previous studies comparing upper and lower extremity recovery have evaluated disability rather than motor impairment. Individuals with lower extremity impairments may be more functional and appear less disabled than individuals with upper extremity impairments. Function of the upper extremity requires finer motor control, for which the patient can less readily compensate. Therefore, impairments and disability would predictably be more highly correlated in this area. We tested the hypothesis that upper and lower extremity motor recovery are similar.

METHODS

The 95 patients selected for this study were enrolled in the Durham County Stroke Study and had been diagnosed with anterior circulation ischemic stroke. Each subject received Fugl-Meyer assessments within 24 hours of admission and then 5, 30, 90, and 180 days after stroke. We used these assessments to compare the time course and patterns of motor function of the upper and lower extremities.

RESULTS

Repeated-measures ANOVA revealed that percent maximal motor recovery was significantly (P < .001) affected by time after stroke but not by extremity (upper extremity versus lower extremity) (P = .32). When stroke severity level is controlled, the upper and lower extremities continue to show no difference in percent motor recovery (P = .19).

CONCLUSIONS

In patients with anterior circulation ischemic stroke, the severity of motor impairment and the patterns of motor recovery are similar for the upper and lower extremities. The most rapid recovery for both extremities occurs within 30 days.

Inaccuracy of the International Classification of Diseases (ICD-9-CM) in identifying the diagnosis of ischemic cerebrovascular disease

In administrative databases the International Classification of Diseases, Version 9, Clinical Modification (ICD-9-CM) is often used to identify patients with specific diagnoses. However, certain conditions may not be accurately reflected by the ICD-9 codes. We assessed the accuracy of ICD-9 coding for cerebrovascular disease by comparing ICD-9 codes in an administrative database with clinical findings ascertained from medical record abstractions. We selected patients with ICD-9 diagnostic codes of 433 through 436 (in either the primary or secondary positions) from an administrative database of patients hospitalized in five academic medical centers in 1992. Medical records of the selected patients were reviewed by trained medical abstractors, and the patients' clinical conditions during the admission (stroke, TIA, asymptomatic) were recorded, as well as any history of cerebrovascular symptoms. Results of the medical record review were compared with the ICD-9 codes from the administrative database. More than 85% of those patients with the ICD-9 code 433 were asymptomatic for the index admission. More than one-third of these asymptomatic patients did not undergo either cerebral angiography or carotid endarterectomy. For ICD-9 code 434, 85% of patients were classified as having a stroke and for ICD-9 code 435, 77% had TIAs. For code 436, 77% of patients were classified as having strokes. Limiting the identifying ICD-9 code to the primary position increased the likelihood of agreement with the medical record review. The ICD-9 coding scheme may be inaccurate in the classification of patients with ischemic cerebrovascular disease. Its limitations must be recognized in the analyses of administrative databases selected by using ICD-9 codes 433 through 436.

Improving the reliability of stroke subgroup classification using the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) criteria

BACKGROUND AND PURPOSE

We sought to improve the reliability of the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) classification of stroke subtype for retrospective use in clinical, health services, and quality of care outcome studies. The TOAST investigators devised a series of 11 definitions to classify patients with ischemic stroke into 5 major etiologic/pathophysiological groupings. Interrater agreement was reported to be substantial in a series of patients who were independently assessed by pairs of physicians. However, the investigators cautioned that disagreements in subtype assignment remain despite the use of these explicit criteria and that trials should include measures to ensure the most uniform diagnosis possible.

METHODS

In preparation for a study of outcomes and management practices for patients with ischemic stroke within Department of Veterans Affairs hospitals, 2 neurologists and 2 internists first retrospectively classified a series of 14 randomly selected stroke patients on the basis of the TOAST definitions to provide a baseline assessment of interrater agreement. A 2-phase process was then used to improve the reliability of subtype assignment. In the first phase, a computerized algorithm was developed to assign the TOAST diagnostic category. The reliability of the computerized algorithm was tested with a series of synthetic cases designed to provide data fitting each of the 11 definitions. In the second phase, critical disagreements in the data abstraction process were identified and remaining variability was reduced by the development of standardized procedures for retrieving relevant information from the medical record.

RESULTS

The 4 physicians agreed in subtype diagnosis for only 2 of the 14 baseline cases (14%) using all 11 TOAST definitions and for 4 of the 14 cases (29%) when the classifications were collapsed into the 5 major etiologic/pathophysiological groupings (kappa=0.42; 95% CI, 0.32 to 0.53). There was 100% agreement between classifications generated by the computerized algorithm and the intended diagnostic groups for the 11 synthetic cases. The algorithm was then applied to the original 14 cases, and the diagnostic categorization was compared with each of the 4 physicians' baseline assignments. For the 5 collapsed subtypes, the algorithm-based and physician-assigned diagnoses disagreed for 29% to 50% of the cases, reflecting variation in the abstracted data and/or its interpretation. The use of an operations manual designed to guide data abstraction improved the reliability subtype assignment (kappa=0.54; 95% CI, 0.26 to 0.82). Critical disagreements in the abstracted data were identified, and the manual was revised accordingly. Reliability with the use of the 5 collapsed groupings then improved for both interrater (kappa=0.68; 95% CI, 0.44 to 0.91) and intrarater (kappa=0.74; 95% CI, 0.61 to 0.87) agreement. Examining each remaining disagreement revealed that half were due to ambiguities in the medical record and half were related to otherwise unexplained errors in data abstraction.

CONCLUSIONS

Ischemic stroke subtype based on published TOAST classification criteria can be reliably assigned with the use of a computerized algorithm with data obtained through standardized medical record abstraction procedures. Some variability in stroke subtype classification will remain because of inconsistencies in the medical record and errors in data abstraction. This residual variability can be addressed by having 2 raters classify each case and then identifying and resolving the reason(s) for the disagreement.

Incidence and occurrence of total (first-ever and recurrent) stroke

BACKGROUND AND PURPOSE

It has recently been hypothesized that the figure of approximately half a million strokes substantially underestimates the actual annual stroke burden for the United States. The majority of previously reported studies on the epidemiology of stroke used relatively small and homogeneous population-based stroke registries. This study was designed to estimate the occurrence, incidence, and characteristics of total (first-ever and recurrent) stroke by using a large administrative claims database representative of all 1995 US inpatient discharges.

METHODS

We used the Nationwide Inpatient Sample of the Healthcare Cost and Utilization Project, release 4, which contains approximately 20% of all 1995 US inpatient discharges. Because the accuracy of International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) coding is suboptimal, we performed a literature review of ICD-9-CM 430 to 438 validation studies. The pooled results from the literature review were used to make appropriate adjustments in the analysis to correct for some of the inaccuracies of the diagnostic codes.

RESULTS

There were 682 000 occurrences of stroke with hospitalization (95% CI 660 000 to 704 000) and an estimated 68 000 occurrences of stroke without hospitalization. The overall incidence rate for occurrence of total stroke (first-ever and recurrent) was 259 per 100 000 population (age- and sex-adjusted to 1995 US population). Incidence rates increased exponentially with age and were consistently higher for males than for females.

CONCLUSIONS

We conservatively estimate that there were 750 000 first-ever or recurrent strokes in the United States during 1995. This new figure emphasizes the importance of preventive measures for a disease that has identifiable and modifiable risk factors and for the development of new and improved treatment strategies and infrastructures that can reduce the consequences of stroke.

Quality of anticoagulation management among patients with atrial fibrillation: results of a review of medical records from 2 communities

BACKGROUND

Most treatment of patients at risk for stroke is provided in the ambulatory setting. Although many studies have addressed the proportion of eligible patients with atrial fibrillation (AF) receiving warfarin sodium, few have addressed the quality of their anticoagulation management.

OBJECTIVE

As a comprehensive assessment of quality, we analyzed the proportion of eligible patients receiving warfarin, the proportion of time their international normalized ratios (INRs) were within the target range, and, when an out-of-target range INR value occurred, the time until the next INR measurement was made.

METHODS

Retrospective review of the medical records of 660 patients with AF managed by general internists and family practitioners in Rochester, NY, and the Research Triangle area of North Carolina.

RESULTS

Only 34.7% of eligible patients with AF received warfarin. The INR values were out of the target range approximately half the time, and the response to these values was not always timely. For all the measures considered, both Rochester practices with access to an anticoagulation service had higher (albeit not ideal) quality of warfarin management than the remaining practices.

CONCLUSIONS

We found significant deficiencies in the practice of warfarin management and suggestive evidence that anticoagulation services can partially ameliorate these deficiencies. More research is needed to describe the quality of anticoagulation management in typical practice and how this management can be improved.

Impact of social support on outcome in first stroke

BACKGROUND AND PURPOSE

The purpose of this study is to examine the impact of social support on outcome after first stroke in a prospective cohort study. Although modest evidence exists for the importance of several psychosocial factors, studies have failed to use widely recognized measures of outcome and social support, have failed to control for time since onset, and have not used longitudinal techniques.

METHODS

Forty-six surviving patients were followed for 6 months after stroke. Recovery was measured using repeated measures of functional status as indicated by the Barthel Index of activities of daily living. Perceived social support was measured at 1, 3, and 6 months after onset. Repeated-measures multivariate analysis of variance was used to analyze changes in functional status.

RESULTS

Significant differences were found across levels of social support in trajectories of functional status (p = 0.002). A significant three-way interaction between stroke severity, social support, and outcome was also found (p = 0.012). Patients with more severe stroke and the largest amount of social support attained an average Barthel Index that was 68 points (65%) higher than the group reporting the least support.

CONCLUSIONS

High levels of social support were associated with faster and more extensive recovery of functional status after stroke. Social support may be an important prognostic factor in recovery from stroke. Socially isolated patients may be at particular risk for poor outcome.

Health status of individuals with mild stroke

BACKGROUND AND PURPOSE

Diminished quality of life and limitations in higher levels of physical functioning are often underestimated in stroke and are not fully captured by measures such as the Barthel Index and the Rankin Outcome Scale. This study used additional measures to assess the health status of 304 persons with mild stroke and to compare these individuals with 184 persons with transient ischemic attack and 654 persons without history of stroke/transient ischemic attack but at elevated risk for stroke (asymptomatic group).

METHODS

Subjects were recruited from the Academic Medical Center Consortium (inpatients), the Cardiovascular Health Study (population-based sample of community-dwelling persons 65 years and older), and United HealthCare (inpatients and outpatients typically younger than 65 years). Subjects were interviewed by telephone or in person to assess activities of daily living (Barthel Index), depression (Center for Epidemiological Studies Depression Scale), health status (MOS-36), and utility for current health state.

RESULTS

Most respondents were independent on all Barthel items. The stroke group was more impaired on the MOS-36 than the asymptomatic group but similar to the group with transient ischemic attack. Health-related quality of life was lowest for persons with stroke. While symptom status and Barthel Index score were the strongest predictors of health status, the Barthel Index showed a consistent ceiling effect when compared with the physical function subscale of the MOS-36.

CONCLUSIONS

The consequences of even mild stroke affect all dimensions of health except pain. Standardized assessment of persons with stroke must evaluate across the entire continuum of health-related functions.

Predicting complications of carotid endarterectomy

BACKGROUND AND PURPOSE

Carotid endarterectomy has been shown to be beneficial in patients with high-grade carotid stenosis and ipsilateral transient ischemic attack or stroke. This benefit will be realized only if the operation is performed safely. We sought to determine the extent to which clinically significant adverse events occurring after carotid endarterectomy can be predicted from clinical data available before surgery.

METHODS

Eleven hundred sixty patients were randomly selected from all patients who underwent carotid endarterectomy and were discharged during the calendar years 1988, 1989, and 1990 in 12 academic medical centers in 10 states. Clinical data abstracted from hospital charts were analyzed retrospectively. A model was developed and validated to predict the occurrence of stroke, myocardial infarction, or death during the postoperative period of hospitalization.

RESULTS

Eight patients (6.9%) suffered at least one adverse event. Rates for individual complications were as follows: death, 1.4%; nonfatal stroke, 3.4%; nonfatal myocardial infarction, 2.1%; and nonfatal stroke or death, 4.8%. Significant predictors of adverse events were age 75 years or older, symptom status (ipsilateral symptoms versus asymptomatic or nonipsilateral symptoms), severe hypertension (preoperative diastolic blood pressure of greater than 110 mm Hg), carotid endarterectomy performed in preparation for coronary artery bypass surgery, history of angina, evidence of internal carotid artery thrombus, and internal carotid artery stenosis near the carotid siphon. The presence of two or more of these risk factors was associated with a nearly twofold increase in risk of an adverse event (relative risk, 1.7; 95% confidence interval, 1.0 to 3.0).

CONCLUSIONS

Clinical data can be used to stratify patients undergoing carotid endarterectomy according to risk of postoperative in-hospital stroke, myocardial infarction, or death.

Epidemiology of recurrent cerebral infarction: a medicare claims-based comparison of first and recurrent strokes on 2-year survival and cost

BACKGROUND AND PURPOSE

Because recurrent strokes will tend to leave patients with greater disability than first strokes, patients with recurrent strokes should have poorer outcomes on average than those with first strokes. The extent of this difference has, however, not yet been estimated with precision.

METHODS

Using a random 20% sample of Medicare patients aged 65 years and older admitted with a primary diagnosis of cerebral infarction during calendar year 1991, we used historical data from the previous 4 years to classify patients as having either first or recurrent stroke and followed survival and direct medical costs for 24 months after stroke. First and recurrent stroke groups were compared with the log-rank test (survival) and t test (cost) and also multivariate modeling.

RESULTS

Survival from first stroke is consistently better than that for recurrent stroke: 24-month survival was 56.7% versus 48.3%, respectively. Costs were similar for the initial hospital stay and in months 1 to 3 after stroke. During months 4 to 24 after stroke, total costs were higher among those with recurrent stroke by approximately $375/mo across all patients, with this difference being greatest for younger patients and least for patients aged 80 years or older. Most of the difference in total monthly cost was attributable to nursing home utilization (averaging approximately $150/mo) and acute hospitalization (averaging approximately $120/mo).

CONCLUSIONS

Patients with recurrent stroke have, on average, poorer outcomes than those with first stroke. To be as accurate as possible, clinical policy analyses should use different estimates of health and cost outcomes for first and recurrent stroke.

Noninvasive carotid artery testing. A meta-analytic review

PURPOSE

To compare the operating characteristics of six noninvasive tests for carotid artery stenosis.

DATA SOURCES

A structured search was done using MEDLINE, reference lists from selected articles, and bibliographies from neurology textbooks that focused on the diagnosis of carotid artery stenosis in humans. The search yielded 568 articles.

STUDY SELECTION

Articles were selected if the noninvasive test results they presented used carotid angiography as the reference standard for comparison, if carotid artery occlusion was considered as a separate category, and if contingency tables could be constructed.

DATA EXTRACTION

At least two physicians reviewed all selected articles. Items abstracted included patient demographics, study design, sites of patient enrollment, whether the interpretation of test outcomes was blinded, and specific results. Sensitivity, specificity, receiver operating characteristic (ROC) curves, and summary measures of effectiveness for each test were calculated.

RESULTS

Carotid duplex ultrasonography, carotid Doppler ultrasonography, and magnetic resonance angiography have sensitivities between 0.82 and 0.86, specificities at 0.98, and test-effectiveness measures at or exceeding 3.0 when predicting 100% occlusion. For carotid stenosis of 70% or more, these three tests and supraorbital Doppler ultrasonography all have sensitivities of 0.83 to 0.86, specificities of 0.89 to 0.94, test-effectiveness measures approaching 3.0, and composite ROC areas of 0.91 to 0.92. Limiting analysis to studies that enrolled consecutive patients and those in which interpretation of the noninvasive tests was independent of the angiograms did not substantially change our results.

CONCLUSIONS

Carotid duplex ultrasonography, carotid Doppler ultrasonography, and magnetic resonance angiography are all similarly successful at predicting 100% carotid artery occlusion and 70% stenosis. Other factors, such as cost, availability, and local experience may influence the decision to use these tests to screen for carotid artery atherosclerosis that may respond to surgery.

Health values of the seriously ill. SUPPORT investigators

OBJECTIVE

To assess 1) the health values and health ratings of seriously ill hospitalized patients, their surrogate decision makers, and their physicians; 2) the determinants of health values; and 3) whether health values change over time.

DESIGN

Prospective, longitudinal, multicenter study.

SETTING

5 academic medical centers.

PARTICIPANTS

1438 seriously ill patients with at least one of nine diseases who had a projected overall 6-month mortality rate of 50%; their surrogates; and their physicians.

MEASUREMENTS

Time-tradeoff utilities (reflecting preferences for a shorter but healthy life) and health ratings.

RESULTS

At study day 3, patients had a mean time-tradeoff utility of 0.73 +/- 0.32 (median [25th, 75th percentile], 0.92 [0.63, 1.0]), indicating that they equated living 1 year in their current state of health with living 8.8 months in excellent health. However, scores varied widely; 34.8% of patients were unwilling to exchange any time in their current state of health for a shorter life in excellent health (utility, 1.0), and 9.0% were willing to live 2 weeks or less in excellent health rather than 1 year in their current state of health (utility, 0.04). Health rating scores averaged 57.8 +/- 24.0 (median [25th percentile, 75th percentile], 60 [50, 75]) on a scale of 0 (death) to 100 (perfect health). The patients' mean time-tradeoff score exceeded that of their paired surrogates (n = 1041) by 0.08 (P < 0.0001). Time-tradeoff scores were related to psychosocial well-being; health ratings; desire for resuscitation and extension of life rather than relief of pain and discomfort; degree of willingness to live with constant pain; and perceived prognosis for survival and independent functioning. Scores of surviving patients increased by an average of 0.06 after 2 months (P < 0.0001) and 0.08 after 6 months (P < 0.0001).

CONCLUSIONS

Health values of seriously ill patients vary widely, are higher than patients' surrogates believe, are related to few other preference and health status measures, and increase over time.

Relationship between test frequency and outcomes of anticoagulation: a literature review and commentary with implications for the design of randomized trials of patient self-management

BACKGROUND

Patient self-management (PSM) of anticoagulation, which is primarily based upon the premise that more frequent testing will lead to tighter anticoagulation control and thus to improved clinical outcomes, is a promising model of care. The goals of this paper are (1) to describe the strength of evidence correlating more frequent testing with improved outcomes; and (2) to discuss implications of these findings for the design of randomized controlled trials (RCTs) assessing the effectiveness and cost-effectiveness of PSM.

METHODS

We performed two literature reviews: one examining the strength of the relationship between time in target range (TTR) and the clinical outcomes of major bleeding and thromboembolism; and the second examining the strength of the relationship between frequency of testing and TTR.

RESULTS

We found that (1) the relationship between TTR and clinical outcomes is strong, thus supporting use of TTR as a primary outcome variable; and (2) more frequent testing seems to increase TTR, although the studies supporting this latter conclusion were relatively few and not definitive. Statistical analysis suggested that a study which uses clinical event rates as its primary outcome would need to be much larger than a comparable study which is based upon TTR.

CONCLUSIONS

When designing randomized trials of PSM, the design should (1) use as its control group high quality anticoagulation management rather than usual care; (2) include the maximum possible amount of self-management in the intervention group; (3) include different testing intervals in the intervention group; (4) use TTR as the primary outcome variable and event rates as a secondary outcome; and (5) base the sample size calculations upon a 5-10% absolute improvement in TTR. Additional RCTs are needed in order to determine how the promise of PSM can best be fulfilled.

Multicenter review of preoperative risk factors for endarterectomy for asymptomatic carotid artery stenosis

BACKGROUND AND PURPOSE

The benefit of carotid endarterectomy is highly dependent on surgical risk. However, little data are available concerning factors affecting the risk of endarterectomy performed for asymptomatic carotid artery stenosis outside the setting of a randomized controlled trial. The purpose of this study was to analyze the impact of potential preoperative risk factors on the frequency of postoperative complications in patients undergoing the operation for asymptomatic disease in academic medical centers.

METHODS

Data regarding postoperative complications were systematically abstracted from the medical records of a random sample of patients who underwent carotid endarterectomy at 12 academic medical centers.

RESULTS

Of 1160 procedures reviewed, 463 (40%) were performed for asymptomatic disease. Postoperative stroke or death occurred in 13 (2.8%), and myocardial infarction occurred in 8 (1.7%). The rate of postoperative stroke or death was lower in asymptomatic patients than in those with a history of cerebrovascular symptoms in a different vascular distribution, but the difference was not significant (1.8% versus 4.2%; P=.21). There were no significant differences in these rates based on race, a history of angina, recent myocardial infarction, chronic obstructive pulmonary disease, hypertension, the degree of stenosis of the contralateral or ipsilateral carotid artery, or the presence of angiographically recognized ulceration, intraluminal thrombus, or siphon stenosis in the ipsilateral vessel (chi(2); P>.05). Postoperative stroke or death was more frequent in women (5.3% versus 1.6% in men; P=.02), in those aged 75 years or older (7.8% versus 1.8% in those younger than 75 years; P=.01), and in those with a history of congestive heart failure (8.6% versus 2.3% in those without a history of congestive heart failure; P=.03). The risk of stroke or death was higher in the 16 patients who had carotid endarterectomy performed in combination with coronary artery bypass surgery than in those who had only endarterectomy (18.7% versus 2.1%; P<.001).

CONCLUSIONS

The overall risk of postoperative stroke or death was nearly twice that reported by Asymptomatic Carotid Atherosclerosis Study (ACAS) investigators in the setting of a clinical trial but was within acceptable guidelines. Women were at higher postoperative risk than men, which supported ACAS findings. Additional high-risk groups were those aged 75 years or older, those with a history of congestive heart failure, and those undergoing prophylactic endarterectomy for asymptomatic stenosis in combination with coronary surgery. Knowledge of these rates may help to better assess an individual's postoperative risk and therefore the anticipated benefit of surgery.

Intermediate, indeterminate, and uninterpretable diagnostic test results

Diagnostic tests do not always yield positive or negative results; sometimes the results are intermediate, indeterminate, or uninterpretable. No consensus exists for the incorporation of such results into data assessment. Conventional Bayesian analysis leads investigators to either exclude patients with non-positive, non-negative results from their studies or categorize such results into inappropriate cells of the standard four-cell decision matrix. The authors propose a standardized method for reporting results in studies dealing with diagnostic test use and discuss how researchers should expand the four-cell matrix to six cells when non-positive, non-negative results occur. They suggest that the six-cell matrix with new operational definitions of sensitivity, specificity, likelihood ratios, and test yield should be adopted routinely. In addition, they define the different types of non-positive, non-negative results and demonstrate how clinicians can use tree-structured decision analysis from the six-cell matrix. While their method does not solve all problems posed by non-positive, non-negative results, it does suggest a standard method for reporting these results and utilizing all the data in decision making.

Performing cost-effectiveness analysis by integrating randomized trial data with a comprehensive decision model: application to treatment of acute ischemic stroke

A recent national panel on cost-effectiveness in health and medicine has recommended that cost-effectiveness analysis (CEA) of randomized controlled trials (RCTs) should reflect the effect of treatments on long-term outcomes. Because the follow-up period of RCTs tends to be relatively short, long-term implications of treatments must be assessed using other sources. We used a comprehensive simulation model of the natural history of stroke to estimate long-term outcomes after a hypothetical RCT of an acute stroke treatment. The RCT generates estimates of short-term quality-adjusted survival and cost and also the pattern of disability at the conclusion of follow-up. The simulation model incorporates the effect of disability on long-term outcomes, thus supporting a comprehensive CEA. Treatments that produce relatively modest improvements in the pattern of outcomes after ischemic stroke are likely to be cost-effective. This conclusion was robust to modifying the assumptions underlying the analysis. More effective treatments in the acute phase immediately following stroke would generate significant public health benefits, even if these treatments have a high price and result in relatively small reductions in disability. Simulation-based modeling can provide the critical link between a treatment's short-term effects and its long-term implications and can thus support comprehensive CEA.

Knowledge of risk among patients at increased risk for stroke

BACKGROUND AND PURPOSE

Patients who recognize their increased risk for stroke are more likely to engage in (and comply with) stroke prevention practices than those who do not. We describe perceived risk of stroke among a nationally diverse sample of patients at increased risk for stroke and determine whether patients' knowledge of their stroke risk varied according to patients' demographic and clinical characteristics.

METHODS

Respondents were recruited from the Academic Medical Center Consortium (n = 621, five academic medical centers, inpatients of varying age); the Cardiovascular Health Study (n = 321, population-based sample of persons aged 65+ years); and United HealthCare (n = 319, five health plans, inpatients and outpatients typically younger than 65 years). The primary outcome was awareness of being at risk for stroke.

RESULTS

Only 41% of respondents were aware of their increased risk for stroke (including less than one half of patients with previous minor stroke). Approximately 74% of patients who recalled being told of their increased stroke risk by a physician acknowledged this risk in comparison with 28% of patients who did not recall being informed by a physician. Younger patients, depressed patients, those in poor current health, and those with a history of TIA were most likely to be aware of their stroke risk.

CONCLUSIONS

Over one half of patients at increased risk of stroke are unaware of their risk. Healthcare providers play a crucial role in communicating information about risk, and successful communication encourages adoption of stroke prevention practices. Educational messages should be targeted toward patients least likely to be aware of their risk.

Multicenter review of preoperative risk factors for carotid endarterectomy in patients with ipsilateral symptoms

BACKGROUND AND PURPOSE

Randomized clinical trials have shown that carotid endarterectomy decreases the risk of subsequent stroke in patients with high-grade carotid stenosis and ipsilateral transient ischemic attack or minor stroke. The benefit of surgery is highly dependent on surgical risk. We previously found that patients with ipsilateral hemispheric symptoms were at greater risk of carotid endarterectomy complications compared with those who were asymptomatic or had nonipsilateral symptoms. The goals of the present study were (1) to identify preoperative clinical factors that may increase the risk of complications after carotid endarterectomy in patients with ipsilateral hemispheric symptoms and (2) to develop a risk index based on this patient-level data.

METHODS

Records from 1160 carotid endarterectomies performed at 12 academic medical centers composed the primary data set. Hospital charts for the admission during which carotid endarterectomy was performed were systematically reviewed by abstractors using a defined protocol. The present analysis was carried out on data from the subset of patients who had carotid endarterectomy for ipsilateral hemispheric symptoms. Candidate variables were identified based on univariate Fisher's exact tests or chi 2 tests. A risk index was then developed using those variables with a greater than 90% probability of being associated with adverse outcomes.

RESULTS

Of the 697 patients with ipsilateral symptoms, 8.5% had either stroke, myocardial infarction, or died during the postoperative period of hospitalization. Those over the age of 75 had a greater risk of myocardial infarction (6.6% versus 2.3%, P = .024) but not of stroke or death (P > .10). The overall frequencies of adverse outcomes were also higher in the 5 patients with complete ipsilateral carotid occlusions (40% versus 8.2%, P < .01), the 28 patients with ipsilateral intraluminal thrombus (17.9% versus 8.1%, P = .07), and the 65 patients with ipsilateral carotid siphon stenosis (13.9% versus 7.9%, P = .10). There were no differences in adverse outcomes among those with different degrees of ipsilateral stenosis (30% to 49%, 50% to 69%, and 70% to 99%). Adverse outcome rates were similar regardless of the type of symptom (transient ischemic attack, recent ipsilateral minor stroke, remote ipsilateral minor stroke). There were no significant differences in adverse outcome rates based on sex, race, history of angina, recent myocardial infarction, congestive heart failure, chronic obstructive pulmonary disease, hypertension, degree of stenosis of the contralateral carotid artery, or presence of ulceration in the ipsilateral artery (Fisher's exact tests, P > .10). A count of variables with greater than 90% probability of being associated with adverse outcomes (age > or = 75 years or angiographic evidence of ipsilateral carotid occlusion, stenosis in the region of the carotid siphon, or intraluminal thrombus) was used to form a simple risk index. "High-risk" patients (one or more risk factors) had more than two times the risk of complications compared with "low-risk" patients who had no risk factors (odds ratio, 2.18; 95% confidence interval, 1.25 to 3.81).

CONCLUSIONS

Certain preoperative clinical variables may place patients with ipsilateral symptoms at greater risk of perioperative complications after carotid endarterectomy. Prospective validation of a simple risk index would provide an additional method for assessing preoperative risk in endarterectomy candidates.