Examining Timeliness of Total Knee Replacement Among Patients with Knee Osteoarthritis in the U.S.: Results from the OAI and MOST Longitudinal Cohorts.

BACKGROUND: Patients with knee osteoarthritis may undergo total knee replacement too early or may delay or underuse this procedure. We quantified these categories of total knee replacement utilization in 2 cohorts of participants with knee osteoarthritis and investigated factors associated with each category. METHODS: Data were pooled from 2 multicenter cohort studies that collected demographic, patient-reported, radiographic, clinical examination, and total knee replacement utilization information longitudinally on 8,002 participants who had or were at risk for knee osteoarthritis and were followed for up to 8 years. Validated total knee replacement appropriateness criteria were longitudinally applied to classify participants as either potentially appropriate or likely inappropriate for total knee replacement. Participants were further classified on the basis of total knee replacement utilization into 3 categories: timely (indicating that the patient had total knee replacement within 2 years after the procedure had become potentially appropriate), potentially appropriate but knee not replaced (indicating that the knee had remained unreplaced for >2 years after the procedure had become potentially appropriate), and premature (indicating that the procedure was likely inappropriate but had been performed). Utilization rates were calculated, and factors associated with each category were identified. RESULTS: Among 8,002 participants, 3,417 knees fulfilled our inclusion and exclusion criteria and were classified into 1 of 3 utilization categories as follows: 290 knees (8% of the total and 9% of the knees for which replacement was potentially appropriate) were classified as "timely", 2,833 knees (83% of the total and 91% of those for which replacement was potentially appropriate) were classified as "potentially appropriate but not replaced", and 294 knees (comprising 9% of the total and 26% of the 1,114 total knee replacements performed) were considered to be "likely inappropriate" yet underwent total knee replacement and were classified as "premature". Of the knees that were potentially appropriate but were not replaced, 1,204 (42.5%) had severe symptoms. Compared with the patients who underwent timely total knee replacement, the likelihood of being classified as potentially appropriate but not undergoing total knee replacement was greater for black participants and the likelihood of having premature total knee replacement was lower among participants with a body mass index of >25 kg/m and those with depression. CONCLUSIONS: In 2 multicenter cohorts of patients with knee osteoarthritis, we observed substantial numbers of patients who had premature total knee replacement as well as of patients for whom total knee replacement was potentially appropriate but had not been performed >2 years after it had become potentially appropriate. Further understanding of these observations is needed, especially among the latter group. CLINICAL RELEVANCE: Undergoing total knee replacement too early may result in little or no benefit while exposing the patient to the risks of a major operation, whereas waiting too long may cause limitations in physical activity that in turn increase the risk of additional disability and chronic disease; however, little is known about timing of this surgery. We quantified the extent of premature, timely, and delayed use, and found a high prevalence of both premature and delayed use.

Diagnosing potential noncompliance. Physicians' ability in a behavioral dimension of medical care.

Predictions by housestaff physicians of whether or not patients would return for follow-up visits and of the amount of prescribed medication they would take were compared to measured compliance for 187 patients discharged from the medical service. Although physicians were able to predict visit compliance better than by chance alone, at best they could accurately predict only 35% of the noncompliers and one half of their predictions of noncompliance were incorrect. In predicting medication compliance, less than one half of physician predictions correctly discriminated between compliant and noncompliant patients and three fourths of their predictions of noncompliance were inaccurate. Because of physician limitations in this important clinical area, medical education should expand efforts to develop physician skills in diagnosing and managing sociobehavioral aspects of illness, and efforts to improve quality of care cannot ignore these "nontechnical" factors.

Hypertension, vascular headaches, and seizures after carotid endarterectomy. Case report and therapeutic considerations.

The following three unusual complications of carotid endarterectomy developed in a 64-year-old woman: vascular headaches, delayed hypertension, and seizures. beta-Blockade with propranolol hydrochloride had no beneficial effect on either the hypertension or the headaches. Peripheral alpha-blockade with prazosin hydrochloride resulted in prompt improvement in the patient's condition. Altered autonomic vascular control secondary to carotid sinus trauma is proposed as a possible causative mechanism for all three complications. The therapeutic implications are discussed. Based on this experience, we would caution against the use of unopposed beta-blockade in this setting and suggest that the role of prazosin in the initial management of postendarterectomy hypertension and vascular headache be explored.

Routine laboratory testing for medical disorders in psychiatric inpatients.

We examined the clinical utility of routine admission laboratory testing for medical disorders in 250 psychiatric inpatients by using clinical criteria to classify laboratory abnormalities as true- or false-positive results. The mean number of tests per patient was 27.7. The mean percentage of true-positive results was 1.8%; the mean predictive value was 12%. When three clinically defined subgroups were examined, both measures of test performance varied in direct proportion to the pretest probability of medical disease. Eleven patients (4%) had important medical problems discovered through routine laboratory testing. A testing battery consisting of nine tests in women and 13 in men would have identified all of these patients. Our results suggest that extensive, routine testing for medical disorders in this setting is unnecessary and that more efficient and accurate testing strategies, based on clinical information, can and should be developed.

A double-blind clinical trial in weight control. Use of fenfluramine and phentermine alone and in combination.

We performed a double-blind, controlled clinical trial comparing phentermine resin (30 mg in the morning), fenfluramine hydrochloride (20 mg three times a day), and a combination of phentermine resin (15 mg in the morning) and fenfluramine hydrochloride (30 mg before the evening meal), and placebo. We combined low doses of the two drugs to maintain efficacy while diminishing adverse effects. Eighty-one people with simple obesity (130% to 180% of ideal body weight) participated. Individualized diets were prescribed and discussed again during the 24-week study period. Weight loss in those receiving the combination (8.4 +/- 1.1 kg; mean +/- SEM) was significantly greater than in those receiving placebo (4.4 +/- 0.9 kg; Scheff é's test) and equivalent to that of those receiving fenfluramine (7.5 +/- 1.2 kg) or phentermine (10.0 +/- 1.2 kg) alone. Adverse effects were less frequent with the combination regimen than with other active treatments. Thirty-seven participants dropped out of the study, 18 for reasons related to drug treatment. Combining fenfluramine and phentermine capitalized on their pharmacodynamic differences, resulting in equivalent weight loss, fewer adverse effects, and better appetite control.

Pneumococcal vaccine strategy. Feasibility of a vaccination program directed at hospitalized and ambulatory patients.

A prospective analysis of patients hospitalized with pneumococcal infection at two Rochester (NY) hospitals was performed during an 18-month period to assess the feasibility of a pneumococcal vaccination program directed at inpatients. Of the 113 patients hospitalized with pneumococcal infection, only nine patients (8%) had previously received pneumococcal vaccine and 76 patients (67%) had been hospitalized at least once in the previous three years. Of the 104 patients who had a source of regular medical care, 100% had made at least one outpatient visit in the previous three years. The association between the acquisition of a serious pneumococcal infection and hospitalization within the previous three years suggests that a vaccination strategy directed at hospitalized high-risk patients would be potentially effective. Also, vaccination of high-risk ambulatory patients remains a viable strategy to prevent serious pneumococcal infection.

Determinants of response to anorexiants.

We analyzed data from an 81-patient clinical trial of anorexiant medication, searching for the predictors of response. In the trial we assigned treatments to participants by minimization, a process that decreases differences between treatment groups. Based on the literature, the investigators' experience, and the pharmacologic properties of the medications, we selected 25 factors for use in the minimization process. Retrospective examination by contingency-table analysis indicated that the most important predictors of weight loss in this study were weight loss during the 3-wk diet-only run-in period (chi 2, p less than 0.001), physician estimation of patients motivation (chi 2, P less than 0.003), participant eating habits (nighttime "binge" eaters responded best; chi 2, P less than 0.003), adherence to treatment (chi 2, P less than 0.01), and type of treatment (chi 2, P = 0.05). When multiple regression analysis was applied, several other factors aided in explaining the variance in the weight loss results. For example, duration of obesity was inversely related to weight loss and weight loss during the study increased with family income. Depending on the treatment used, anxiety and depression visual analogue scores also explained some of the variance. Failure to account for factors such as these in treatment assignment may account for the low discriminant power of clinical trials of anorexiants.

The value of diagnostic information to patients with suspected multiple sclerosis. Rochester-Toronto MRI Study Group.

OBJECTIVE: To determine the value of diagnostic information to patients with suspected multiple sclerosis (MS). Because treatment choices would be only minimally affected by earlier diagnosis for most patients with this clinical problem, this study assessed the "nondecisional" value of diagnosis. DESIGN: Prospective survey of patients before and after diagnostic workup, including imaging with magnetic resonance scanning. We assessed the effect of diagnostic information on patients' sense of well-being, as well as direct measures of the utility of information (using time trade-off and willingness-to-pay techniques). SETTING: Patients referred from primary care practices for diagnostic workup for suspected MS to neurology clinics and practices. PATIENTS: Sixty-eight individuals, mean age 37.5 years, 53 female and 15 male. Thirty-one patients were classified as having "probable MS," and 37 were classified as having "possible MS" by the examining neurologist before workup. MEASUREMENTS: Present and future health perception, uncertainty about diagnosis-prognosis, and level of anxiety. Willingness to pay for diagnostic information, quality of life as measured by the time trade-off technique, and psychological state of the patient before and after diagnosis. RESULTS: Diagnostic uncertainty fell significantly as a result of the diagnostic workup. Most patients (59/62) said that they were better off having received diagnostic information. Although anxiety seemed to be reduced by testing, overall anxiety levels did not decrease as much as anticipated. Patients also became less optimistic about their future health after testing. On average, patients were willing to forgo 4.5 quality-adjusted life days to receive an earlier diagnosis and their quality of life after diagnosis improved slightly. Subgroups of patients differed in their response to diagnostic information. Those in whom no definitive diagnosis emerged tend to be more anxious rather than being reassured by the "negative" workup. Individuals with "positive" workups became less anxious and expressed favorable feelings about the diagnostic workup even though they often faced a chronic disease. CONCLUSIONS: Overall, the diagnostic workup seemed to benefit patients and improve their sense of well-being. However, whether the effects were beneficial or not depended on the results of the diagnostic workup itself. In clinical practice the decision to undergo testing in situations in which definitive treatment is unavailable should be individualized. The potential for negative as well as positive consequences should be recognized.

Carotid endarterectomy trends in the patterns and outcomes of care at academic medical centers, 1990 through 1995.

OBJECTIVE: To evaluate whether the patterns of inpatient care and patient characteristics have changed for patients undergoing a carotid endarterectomy across a group of academic medical centers from 1990 through 1995. If changes occurred, we investigated whether they had an impact on patient outcomes. DESIGN: Retrospective evaluation of patients undergoing a carotid endarterectomy using a hospital discharge data set compiled by the Academic Medical Center Consortium. SETTING: Ten academic medical centers. PATIENTS: A total of 7019 hospital admissions for patients who had 1 carotid endarterectomy performed as a principal procedure from January 1990 to December 1995. MAIN OUTCOME MEASURES: Trends in patient demographics, comorbidities, length of stay, days in the intensive care unit, and inpatient cerebral arteriogram use were determined. Patient outcomes included inpatient mortality, discharge to an institution, 30-day readmission rate, and selected diagnoses (postoperative hemorrhage, infection, or seizure; acute myocardial infarction; or cranial nerve palsy) and postprocedure diagnostic tests (computed tomography and magnetic resonance imaging of the head and electroencephalogram) indicative of complications. RESULTS: Over the 6-year study period, the number of carotid endarterectomies performed more than doubled and the percentage of hospital admissions for patients 65 years or older increased from 65% to 75%. The mean and median length of stay halved and the percentage of admissions with transfers to the intensive care unit decreased from 56% to 26% of cases. In addition, the percentage of cases with a cerebral arteriogram during the same admission but prior to the day of the carotid endarterectomy decreased from 52% to 27%. There were no trends in inpatient mortality, discharge to an institution, or 30-day readmission rate. There were no significant trends indicative of poorer quality of care as measured by the frequency of secondary diagnoses or postprocedure diagnostic test use. CONCLUSIONS: Despite dramatic changes that have occurred in patient characteristics and in hospital management practices for patients undergoing a carotid endarterectomy from 1990 to 1995, we were unable to detect any measurable impact on patient outcomes. These data have implications for monitoring and evaluating the impact of systemwide change on the overall quality of care for patients undergoing a carotid endarterectomy.

Clinical diagnosis of multiple sclerosis. The impact of magnetic resonance imaging and ancillary testing. Rochester-Toronto Magnetic Resonance Study Group.

OBJECTIVE: Magnetic resonance imaging, computed tomography, cerebrospinal fluid analysis, and evoked potential testing are used to assist in the diagnosis of patients suspected to have multiple sclerosis (MS). The impact of these tests on a clinician's diagnosis of patients suspected to have MS has not been studied systematically. DESIGN: Clinicians made a diagnosis of each patient following clinical evaluation, again after reviewing the results of magnetic resonance imaging, and finally after reviewing information from other laboratory testing. These diagnoses were compared with the criterion standard of a masked "gold standard" panel reviewing all information after a mean follow-up of 0.9 year. SETTING: The General Neurology Clinic and Multiple Sclerosis Clinic of the University of Rochester (NY). PATIENTS: A consecutive sample of 62 patients diagnosed as having either possible or probable MS following clinical evaluation. MAIN OUTCOME MEASURE: Changes in diagnostic certainty of clinicians following incremental presentation of new laboratory data and the accuracy of such diagnoses. RESULTS: Clinicians used magnetic resonance imaging findings to diagnose definite MS or to eliminate MS from diagnostic consideration in 44% of cases. In these cases, further laboratory testing did not alter clinicians' decisions. In the remaining 56% of cases, in which magnetic resonance imaging did not lead to a diagnosis of definite MS or eliminate MS from diagnostic consideration, further laboratory testing led to such diagnoses in an additional 13% of cases. Gold standard diagnoses were in agreement with the clinician's assessments. CONCLUSIONS: Magnetic resonance imaging aids in the evaluation of patients suspected to have MS; other subsequent studies (computed tomography, cerebrospinal fluid analysis, and evoked potential testing) have less impact. After all studies are performed, about half of such patients still have a tentative diagnosis.

Intracranial mass lesions in acquired immunodeficiency syndrome: using decision analysis to determine the effectiveness of stereotactic brain biopsy.

We studied the effectiveness of performing a stereotactic brain biopsy in the individual with acquired immunodeficiency syndrome (AIDS) and an intracranial mass lesion who failed 2 weeks of antitoxoplasmosis therapy. We used a decision analysis to compare two different treatment strategies: biopsy and no biopsy. The analysis estimates the average life expectancy for each choice and investigates the sensitivity of these results by varying parameters within the model. In the base case analysis (diagnostic yield of biopsy, 0.89; operative mortality, 0.015; life expectancy of lymphoma untreated and treated, 42 and 120 days), the life expectancy of the biopsy strategy was 98 days compared with 67 days for the no-biopsy strategy, for a net survival benefit of 31 days. Sensitivity analyses revealed that the life expectancy of the biopsy strategy remained greater than the no-biopsy strategy for a wide range of variable specifications. The net survival benefit, however, was sensitive to the diagnostic success rate, the operative mortality, the likelihood of a lymphoma diagnosis, and the life expectancy of patients being diagnosed and treated for lymphoma. These data allow AIDS patients and physicians to learn more about the potential outcomes of the alternative management strategies when an individual fails to respond to empiric antitoxoplasmosis therapy.

The effect of spectrum bias on the utility of magnetic resonance imaging and evoked potentials in the diagnosis of suspected multiple sclerosis.

BACKGROUND: Spectrum bias is the tendency for the effectiveness of a test (or treatment) to vary as a function of clinical, pathologic, or comorbid variables including disease severity. Our null hypothesis was that within the population of suspected multiple sclerosis (MS) patients, there would be no difference in MRI and evoked potential (EP) sensitivity and specificity between those with mild MS versus those with more severe clinical signs of the disease, i.e., that spectrum bias is not a factor. METHODS: Three hundred three patients with suspected MS were evaluated by a board-certified neurologist, then scanned with MRI. Two hundred four patients also received EP testing. RESULTS: The sensitivity of MRI in patients with suspected MS was 58 percent with a false-positive rate of 9%. The overall sensitivity was 64% in the probable and 45% in the possible group. In the low pretest probability group sensitivity was 20%, and it was 70% in the high pretest probability group. These differences in sensitivity are statistically significant (p < 0.03). In contrast, the specificity between groups did not differ significantly. EP sensitivity was 69% in the high probability subgroup and 5% in the low probability subgroup. (p < 0.01). CONCLUSION: In this study, both EP and MRI results demonstrated an association between disease frequency, disease severity, and test sensitivity with greater disease frequency and intensity suggesting more impressive diagnostic test performance. The distorting effect of the variable clinical severity on MRI and EP sensitivity in suspected MS underscores that diagnostic tests perform differently in different groups of patients.

The prostate: decreasing cost-effectiveness of biopsy with advancing age.

RATIONALE AND OBJECTIVES: The purpose of this study was to determine the cost-effectiveness of prostate biopsy at different excess prostate-specific antigen (PSA) levels as a function of age. METHODS: Medical decision analysis was performed with standard software (SMLTREE) to determine marginal effectiveness in quality adjusted life years (QALYs) and marginal cost-effectiveness in dollars per QALY of immediate prostate biopsy at different excess PSA levels between 0 ng/mL and 20 ng/mL. The probability of clinically significant cancer with a positive biopsy (pD+Bx+) was assumed to decrease with age from 1.0 at age 50 to 0.2 at age 70. Costs were based on charges at our hospital and were considered over a 2-year time frame. RESULTS: With our base case assumptions there was a decrease in QALYs and an increase in costs doing an immediate prostate biopsy at all excess PSA levels between 0 ng/mL and 20 ng/mL, compared with not biopsying the prostate at > or = 70 years. Doubling pD+Bx+ from 0.2 to 0.4 in the 70-and-older age group resulted in a small increase in QALYs in biopsying the prostate at excess PSA levels between 0 ng/mL and 20 ng/mL. However, the marginal cost-effectiveness of prostate biopsy was very high, ranging from $275,000/QALY biopsying at an excess PSA level of 0 ng/mL to $68,000/QALY biopsying at an excess PSA level of 20 ng/mL. This compared with it being more effective and less costly to biopsy at all excess PSA levels > or = 0 ng/mL in 50-year old patients. CONCLUSION: Immediate prostate biopsy is not cost-effective and can be detrimental in patients > or = 70 years of age at all excess PSA levels between 0 ng/mL and 20 ng/mL.

Cost-effectiveness of routine antenatal varicella screening.

OBJECTIVE: To evaluate the cost-effectiveness of routine antenatal varicella serologic screening of pregnant women with negative or indeterminate varicella histories. METHODS: Routine antenatal varicella screening was evaluated using a decision analytic model. Outcomes were varicella cases, deaths, and life-years. Probabilities were derived from the literature, and sensitivity analysis was performed when data were imprecise or subject to variation. The analysis was repeated to include the effect of a policy of routine screening and vaccination of all adults. RESULTS: Routine antenatal varicella screening of history-negative women was not cost-effective unless the cost of screening was decreased six-fold, varicella exposure rates were greater than 6%, or there was a greater than three-fold decrease in varicella exposure in women testing nonimmune compared with unscreened women. These results were not sensitive to alterations in varicella-zoster immunoglobulin (Ig) effectiveness, varicella communicability, rates and timing of contact reporting, costs (per case, pneumonia, and death), or serologic test performance. If performed as part of a policy of universal screening of all history-negative adults (with vaccination of the majority of those testing nonimmune), routine antenatal varicella testing became cost-effective. CONCLUSION: Routine antenatal varicella screening of all pregnant women with negative or indeterminate varicella histories is not cost-effective. It could be cost-effective in groups of women with increased exposure risk, or if part of a policy of screening and vaccination of all adults.

Estimating the accuracy of screening mammography: a meta-analysis.

OBJECTIVE: To estimate the accuracy of mammographic screening. DESIGN: A meta-analysis of published literature. DATA SOURCES: Published English-language randomized controlled trials, case-control studies, and demonstration projects involving screening mammography were identified using recent review articles. We found additional references using MEDLINE searches combining the MeSH terms "mammography," "screening," and/or study authors and locations. STUDY SELECTION: We included all studies that provided information to calculate the true-positive rate (TPR) and the false-positive rate (FPR) for breast cancer screening. DATA EXTRACTION: Reported data were reviewed independently by the authors; calculations were compared and discrepancies resolved. We calculated the sensitivity as the number of breast cancers detected during the first round of screening (true positives) divided by the sum of the true positives and the false negatives (defined as cancer discovered within 1 year of screening). False-positives were determined by biopsy. DATA SYNTHESIS: TPR and FPR values from each study were plotted in receiver operating characteristic (ROC) space. Tests of homogeneity were performed to assess the validity of using summary ROC curves or a single point to summarize the data. The reported TPR and FPR of mammography ranged from 83% to 95% and 0.9% to 6.5%, respectively. The sensitivity of mammography is higher in women over the age of 50 years. CONCLUSIONS: The accuracy of mammography should be recognized and included in discussions about policies for screening for breast cancer. This meta-analysis, by quantifying the expected TPR/FPR, should assist program planners, physicians, and women to understand better the cost and clinical implications of such screening programs.

The Johns Hopkins ambulatory-care coding scheme.

A classification and coding system for ambulatory-care problems has been developed at the Johns Hopkins Medical Institutions and three affiliated institutions. The provider's statement of the patient's problem, as recorded on an encounter from, is kept in a computer file. Codes from the classification scheme, based on those used in four existing schemes, are automatically assigned to diagnoses, symptoms, well-care services, and treatment procedures categorized by physiological system and subsystem. About 85 percent of recorded problems are machine-codable; the remainder are alphabetized for efficient manual coding. The coding system is integrated with an overall information system that allows linkage of coded problem data to diverse data on patient and provider characteristics. Examples are given of the uses and limitations of the linked data for care evaluation, management, and clinical research.

Do quality report cards play a role in HMOs' contracting practices? Evidence from New York State.

OBJECTIVE: To answer two related questions: (1) Do managed care organizations (MCOs) in New York State (NYS) consider quality when they choose cardiac surgeons? (2) Do they use information about risk-adjusted mortality rates (RAMR) provided in the New York State Cardiac Surgery Reports? DATA SOURCES: (1) Telephone interviews with and contracting data from the majority of MCOs licensed in NYS; (2) RAMR, quality outlier designation, and procedure volume for all cardiac surgeons, as reported in the Cardiac Surgery Reports. STUDY DESIGN: Interview data were analyzed in conjunction with patterns revealed by contracting data. Null hypotheses that MCOs' contracting choices were random with respect to the information published in the Cardiac Surgery Reports were tested. PRINCIPAL FINDINGS: Sixty percent of MCOs ranked the quality of surgeons as most important in their contracting considerations. Although 64 percent of MCOs indicated some knowledge of the NYS Cardiac Surgery Reports, only 20 percent indicated that the reports were a major factor in their contracting decision. Analyses of actual contracting patterns show that in aggregate, the hypothesis of random choice could be rejected with respect to high-quality outlier status and high procedure volume but not for RAMR or poor-quality outlier status. The panel composition of the majority of MCOs (80.2 percent) was within two standard deviations of the expected mean under the null hypothesis. CONCLUSIONS: Despite a professed preference for high-quality surgeons, the use of publicly available quality reports by MCOs is currently low, and contracting practices for the majority of MCOs do not indicate a systematic selection either for or against surgeons based on their reported mortality scores. This study suggests that policy initiatives to increase the effective use of report cards should be encouraged.

Focusing technology assessment using medical decision theory.

Combining medical decision theory and epidemiologic information, the authors have developed a strategy to assess diagnostic technologies. For any patient, patient utilities with new diagnostic information are compared with the preferred fallback action absent that diagnostic information. After determination of whether the expected value of diagnostic information (EVDI) justifies its cost, the method adds across the eligible population to determine whether the global EVDI justifies the technology's deployment, employing a screen (Hurdle 1) that assumes that the diagnostic device has perfect accuracy. This preliminary evaluation relies on published data on treatment efficacy, population probabilities of illness, etc., but not on new clinical trials. If the technology is not sufficiently cost-effective, even with this optimistic assumption, the strategy recommends against its use. Otherwise, the next step is Hurdle II, in which the critical clinical studies, identified by the decision-theory model, are undertaken. These commonly include measuring the actual diagnostic accuracy of a device, with which the cost-effectiveness is recalculated. These studies in general do not require randomized controlled trials.

Targeting assessments of magnetic resonance imaging in suspected multiple sclerosis.

Decision-analytic methods can be valuable for targeting research in technology assessment. They can indicate whether further evaluation of a technology is warranted, and if so, which variables are key determinants of its clinical utility and cost-effectiveness. This approach was tested on a salient issue--whether magnetic resonance imaging (MR) should be used in evaluating patients with mild neurologic symptoms who might have multiple sclerosis (MS). The authors developed a decision-analytic model to assess the expected utility and costs associated with immediately using MR in this situation, compared with waiting for further symptoms to emerge before testing. Sensitivity analyses demonstrated that priorities for technology assessment research include estimating the value of information to patients in resolving uncertainty, evaluating the impact on patients of being labeled with a diagnosis of MS, and measuring the test characteristics of MR.

An evaluation of clinicians' subjective prior probability estimates.

The degree of consensus and the accuracy of subjective prior probability estimates made by 104 clinicians were examined. The clinicians' estimates were compared with objective prior probabilities obtained from published sources and actual patient outcomes. Each clinician made seven estimates based upon written case summaries abstracted from patient records. Consensus was measured by calculating estimate ranges and standard deviations. The clinicians' estimates varied widely: the smallest range was 80 (2%-82%); four of the seven probability ranges were greater than 90. The average standard deviation was 19.5. Using these prior probabilities and Bayes' theorem, widely varying posttest probabilities would result after many common diagnostic tests. Accuracy was measured using the Brier score, which ranges from 0 to 1; a score of 0 indicates perfect accuracy. The clinicians' Brier scores ranged from 0.05 to 0.57. The objectively determined probabilities achieved a Brier score of 0.11, better than that of 96% of the clinicians. Clinical experience did not consistently affect estimate accuracy or consensus. The clinicians' subjective estimates were inaccurate measures of the prior probability of disease. There was little consensus regarding disease likelihood among the clinicians. Objective prior probabilities were more accurate and less variable.