Search PubMed⌕ Search

Biomedical subjects

D B Matchar

Publications and source records attributed to D B Matchar.

At least 91 records · Page 5Linked to original sources
PubMed1991

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.

Epidemiology↗
PubMed1991

Racial variations in ischemic stroke-related physical and functional impairments.

BACKGROUND AND PURPOSE: We sought to determine whether there are racial differences in physical and functional impairments resulting from an initial ischemic stroke. METHODS: We conducted a prospective, county-wide, multisite cohort study including a university hospital, a community hospital, and a Veterans Affairs hospital. The study population was an inception cohort of 145 patients hospitalized for ischemic stroke. Physical and functional impairments were measured using a modified form of the Fugl-Meyer test and the Barthel Index, respectively. Nurses trained to use these tests made assessments at admission and 5, 30, 90, and 180 days after admission. Patient and disease-specific data along with treatment data and vital status were collected. RESULTS: Forty-one patients (28%) were black. Compared with whites, black stroke patients were more likely to be widowed (51% versus 26%) and hypertensive (83% versus 63%) but less likely to be male (42% versus 69%) and alert on admission (66% versus 76%). There were no racial differences in mortality. Physical impairment was significantly more severe in black than in white patients at admission, and although physical impairment improved, it remained significantly worse in blacks. Functional impairment was also greater in black patients initially but was similar to that in white patients 90 days after the event. Multivariable analyses confirmed these findings. CONCLUSIONS: These results indicate that blacks may have greater residual physical deficits from stroke than whites.

Activities of Daily Living↗
PubMed1988

When and how to study the carotid arteries.

STUDY OBJECTIVE: To appraise critically the clinical indications for diagnostic evaluation of carotid artery disease and the efficacy, safety, limitations, and relative costs of commonly used noninvasive diagnostic tests. DESIGN: Formal decision analysis to determine the relative value of several diagnostic strategies and information synthesis of studies evaluating noninvasive diagnostic tests. SETTING: We did a secondary analysis of the literature, using available data to generate estimates for the probability of selected, clinically important events and the utilities of relevant outcomes observed in patients when carotid artery disease is present or absent, and we critically appraised studies reported since 1980 to estimate the sensitivity and specificity, safety, limitations, and costs of diagnostic tests. MAIN RESULTS: For patients with asymptomatic neck bruits who are and who are not being considered as candidates for other vascular surgery, no diagnostic testing is recommended. For symptomatic patients with anterior-circulation transient ischemic attacks or previous minor strokes, noninvasive diagnostic testing is recommended if surgical treatment is being considered. Patients who previously had minor strokes could be directly referred for selective carotid angiography because they have a higher likelihood of a surgically approachable carotid lesion. Patients with nonlateralizing signs and symptoms of ischemia do not need diagnostic evaluation of possible carotid artery disease unless their symptoms are related to ischemia in the carotid artery circulation. The preferred noninvasive diagnostic test is duplex ultrasound, which has a sensitivity of approximately 85% and a specificity of 90%. If duplex ultrasound is not available, an alternative noninvasive test is carotid Doppler ultrasound. CONCLUSIONS: We recommend a conservative diagnostic approach in patients with suspected carotid artery disease. Noninvasive diagnostic testing is indicated only for symptomatic patients with transient ischemic attacks in the anterior circulation who are considered candidates for carotid endarterectomy, when knowledge of the vessel anatomy is necessary.

Algorithms↗
PubMed1988

Routine intravenous pyelograms before hysterectomy in cases of benign disease: possibly effective, definitely expensive.

Intraoperative ureteral injuries are potentially serious iatrogenic complications. Routine preoperative intravenous pyelograms might decrease the risk of such injuries during hysterectomy in cases of nonmalignant disease, but no prospective studies have been conducted to determine their effectiveness. Since gynecologists must frequently decide whether to obtain pyelograms, we used decision analysis to determine how costly routine pyelograms would be and under what circumstance pyelograms might be especially justifiable. Our cost-effectiveness study was based on decision analysis techniques, with the best available estimates from the literature, and on expert opinion. These estimates were varied over a broad range during a sensitivity analysis so as not to bias the results. At a baseline ureteral incidence injury rate of 0.5%, the marginal cost-effectiveness ratio indicates 833 pyelograms would be obtained to prevent a single injury and approximately $3.33 million would be spent to prevent a single death. As the probability of an injury increases, the marginal cost-effectiveness ratio is less dependent on the test efficacy and is more modest. Since abnormal ureteral anatomy probably predicts ureteral injury, we suggest selectively obtaining preoperative pyelograms only when the probability of an abnormality is high.

Contrast Media↗
PubMed1987

Endarterectomy in carotid artery disease. A decision analysis.

Carotid endarterectomy is being performed with increasing frequency, now over 100,000 times annually in the United States. We used the methods of decision analysis to examine the question of when to perform carotid endarterectomy. We developed a model that simulates the possible outcomes for a cohort of patients at risk for stroke. Estimates of surgical risk, surgical efficacy, annual stroke rate, and nonstroke mortality were derived from the literature. Using sensitivity analysis, we found that surgical risk, surgical efficacy, and stroke risk are the most important factors in determining when surgery is appropriate. By examining a series of clinical scenarios, we constructed guidelines for carotid endarterectomy based on the estimated risk of future stroke. The analysis suggests that for patients with a risk of less than 3% per year, surgery is not indicated. For patients with risk between 3% and 5% per year, low-risk surgery can be expected to provide a benefit of at most three months of quality life, depending on the efficacy of surgery. For stroke risk between 5% and 10% per year, even high-risk surgery is favored if surgical efficacy is above 30%. Above a stroke risk of 10% per year, even high-risk, low-efficacy surgery should be considered. The challenge to advocates of carotid endarterectomy is to develop a cost-effective strategy for identifying patients at high risk for stroke.

Carotid Artery Diseases↗
PubMed1987

Isotope-dilution assay for urinary methylmalonic acid in the diagnosis of vitamin B12 deficiency. A prospective clinical evaluation.

Vitamin B12 deficiency is a frequently considered diagnosis for which there is no single, commonly available and accurate test. A urinary methylmalonic acid assay using gas chromatography-mass spectrometry has been proposed as the preferred test. We reviewed vitamin B12 assays on 1599 consecutive patients and prospectively studied all patients with low serum B12 levels (n = 75) and a random sample of patients with normal levels (n = 68). Of 96 evaluable patients, 7 had clinical deficiency. All 7 deficient patients had urinary methylmalonic acid levels greater than 5 micrograms/mg creatine (sensitivity, 100%; confidence interval, 65% to 100%). Of the 89 patients who were not clinically deficient, 88 had urinary methylmalonic acid levels less than or equal to 5 micrograms/mg creatinine (specificity, 99%). The overall test accuracy in this population was 99%. If the high sensitivity and specificity of the gas chromatography-mass spectrometry assay for urinary methylmalonic acid is supported by other clinical studies, the methylmalonic acid assay may become the reference standard for the diagnosis of vitamin B12 deficiency.

Aged↗
PubMed1987

Quantitative approaches to clinical diagnosis of cancer in elderly patients.

Recommendations for cancer detection in the elderly are based on even less evidence of efficacy than are recommendations for the general adult population. In this article, we present some quantitative measures of test performance--operating characteristics, effectiveness and cost effectiveness/cost benefit--and show examples of how to use these measures to adjust diagnostic strategies for factors especially relevant to the elderly.

Aged↗
PubMed1985

Regression models for prognostic prediction: advantages, problems, and suggested solutions.

Multiple regression models have wide applicability in predicting the outcome of patients with a variety of diseases. However, many researchers are using such models without validating the necessary assumptions. All too frequently, researchers also "overfit" the data by developing models using too many predictor variables and insufficient sample sizes. Models developed in this way are unlikely to stand the test of validation on a separate patient sample. Without attempting such a validation, the researcher remains unaware that overfitting has occurred. When the ratio of the number of patients suffering endpoints to the number of potential predictors is small (say less than 10), data reduction methods are available that can greatly improve the performance of regression models. Regression models can make more accurate predictions than other methods such as stratification and recursive partitioning, when model assumptions are thoroughly examined; steps are taken (ie, choosing another model or transforming the data) when assumptions are violated; and the method of model formulation does not result in overfitting the data.

Antineoplastic Combined Chemotherapy Protocols↗