Long-term association between seizure outcome and depression after resective epilepsy surgery.

OBJECTIVE: This study explored the association between long-term epilepsy surgery outcome and changes in depressive symptoms. METHODS: Adults were enrolled between 1996 and 2001 in a multicenter prospective study to evaluate outcomes of resective epilepsy surgery. The extent of depressive symptoms and depression case status (none, mild, or moderate/severe) were assessed using the Beck Depression Inventory (BDI) preoperatively and 3, 12, 24, 48, and 60 months postoperatively. A mixed-model repeated-measures analysis was performed, adjusting for covariates of seizure location, gender, age, race, education, and seizure control. RESULTS: Of the total 373 subjects, 256 were evaluated at baseline and 5 years after surgery. At baseline, 164 (64.1%) were not depressed, 34 (13.3%) were mildly depressed, and 58 (22.7%) had moderate to severe depression. After 5 years, 198 (77.3%) were not depressed, 20 (7.8%) were mildly depressed, and 38 (14.8%) were moderately to severely depressed. Five years after surgery, the reduction in mean change from baseline in BDI score was greater in subjects with excellent seizure control than in the fair and poor seizure control groups (p = 0.0006 and p = 0.02 respectively). Those with good seizure control had a greater reduction in BDI score than the poor seizure control group (p = 0.02) and borderline significant reduction compared with the fair seizure control group (p = 0.055). CONCLUSION: Although study participants had initial improvement in depressive symptoms, on average, after resective surgery, only patients with good or excellent seizure control had sustained long-term improvement in mood.

Expectations prior to epilepsy surgery: an exploratory comparison of men and women.

Although the clinical goal of resective epilepsy surgery is seizure freedom, patients have a wide set of expectations for this invasive procedure. The goal of this study was to evaluate potential gender differences in expectations among patients undergoing resective epilepsy surgery. Ratings of the importance of 12 potential impacts ("expectations") of resective surgery were analyzed in a seven-center cohort study including 389 adults aged 16 and older who underwent resective epilepsy surgery. Men and women both ranked anticipated changes in driving and memory as the most important presurgical expectations. Women rated driving, physical activity limitations, and economic worries as less important, and fatigue and pregnancy concerns as more important than did men (p's< or =0.05). Exploratory factor analysis indicated a different pattern of associations among the 12 importance items for men and women. Whether gender differences in presurgical values are associated with outcomes needs exploration.

Disparities of care in veterans with Parkinson's disease.

BACKGROUND: Disparities of Parkinson's disease (PD) care have not been assessed. METHODS: We examined the medical records of 309 (83%) non-Hispanic White and 65 (17%) non-White Los Angeles veterans with PD from 1998 to 2004 to determine if care quality as measured by 10 PD indicators different by race/ethnicity. RESULTS: In multivariate modeling, adherence to indicators was higher among non-Hispanic Whites (71% vs. 65%, risk ratio 1.15, 95% CI [1.07-1.32]) compared to non-Whites. Differences in adherence by race/ethnicity were greatest for depression treatment (p<0.05). CONCLUSIONS: We detected disparities in quality of PD care, particularly in depression treatment. Future research should determine causes for these so that interventions can be designed to reduce such disparities.

Worsening of quality of life after epilepsy surgery: effect of seizures and memory decline.

BACKGROUND: Surgery for intractable temporal lobe epilepsy usually controls seizures and improves health-related quality of life (HRQOL), but some patients experience continued seizures, memory decline, or both. The relative impact of these unfavorable outcomes on HRQOL has not been described. METHODS: We studied seizure control, memory change, and HRQOL among 138 patients in the Multicenter Study of Epilepsy Surgery (MSES), an ongoing, prospective study of epilepsy surgery outcomes. Seizure remission at 2 years and 5 years was prospectively determined based upon regularly scheduled follow-up calls to study patients throughout the follow-up period. HRQOL was assessed annually using the Quality of Life in Epilepsy Inventory (QOLIE-89). Memory decline was determined by change in verbal delayed recall from baseline to the 2- or 5-year follow-up. RESULTS: HRQOL improved in patients who were in remission at the 2-year or 5-year follow-up, regardless of memory outcome. Among those not in remission at both 2 and 5 years (25/138, 18%), HRQOL remained stable when memory did not decline (14/138, 10%), but HRQOL declined when memory did decline (11/138, 8%). These 11 patients had baseline characteristics predictive of poor seizure or memory outcome. Declines were most apparent on HRQOL subscales assessing memory, role limitations, and limitations in work, driving, and social activities. CONCLUSIONS: After temporal resection, health-related quality of life (HRQOL) improves or remains stable in seizure-free patients despite memory decline, but HRQOL declines when persistent seizures are accompanied by memory decline. These results may be useful in presurgical counseling and identifying patients at risk for poor psychosocial outcome following surgery.

Health care costs decline after successful epilepsy surgery.

BACKGROUND: Surgery is an effective, high-cost procedure used increasingly to treat refractory epilepsy. For surgery to be cost-effective, long-term cost savings from reduced health care use should provide some offset to the initial costs of evaluation and surgery. There is little information about how health care costs are affected by evaluation and surgery. OBJECTIVE: To determine whether health care costs change when seizures become controlled after surgery. METHODS: Health care costs for the 2 years prior to surgical evaluation and for 2 years afterward were calculated from medical records of 68 subjects with temporal lobe epilepsy (TLE) participating in a multicenter observational study. Costs were compared among patients who did not have surgery, patients who had persisting seizures after surgery, and patients who were seizure free after surgery. RESULTS: Antiepileptic drugs (AEDs) accounted for more than half of the costs of care in the pre-evaluation period. Total costs for seizure-free patients had declined 32% by 2 years following surgery due to less use of AEDs and inpatient care. Costs did not change in patients with persisting seizures, whether they had surgery or not. In the 18 to 24 months following evaluation, epilepsy-related costs were $2,068 to $2,094 in patients with persisting seizures vs $582 in seizure-free patients. CONCLUSIONS: Costs remain stable over 2 years post-evaluation in patients with temporal lobe epilepsy whose seizures persist, but patients who become seizure free after surgery use substantially less health care than before surgery. Further cost reductions in seizure-free patients can be expected as antiepileptic drugs are successfully eliminated.

Patient-perceived impact of resective epilepsy surgery.

OBJECTIVE: To evaluate the patient-perceived impact of resective epilepsy surgery, a key outcome to consider in evaluating such a highly invasive, elective procedure. METHODS: Impact measures obtained from 396 patients in a multicenter cohort study of resective epilepsy surgery included (1) willingness to undergo surgery if that decision could be made again and (2) the overall impact of surgery on the patient's life. Predictors of impact were analyzed using multivariate ordinal logistic regression. RESULTS: Of study participants, 73.8%, 77.4%, and 75.5% would definitely undergo surgery again and 78.2%, 80.2%, and 79.1% reported a very strong or strong positive overall impact of surgery at 3, 12, and 24 months. Multivariate ordinal logistic regression showed that seizure freedom predicted more positive perceptions at 3, 12, and 24 months (all p < 0.04). Becoming employed was uniquely associated with willingness to undergo surgery again and with overall impact at 24 months (all p < 0.05), but only a net 7% of the cohort improved their employment status. Right-sided resection (at 12 and 24 months, p < 0.005) and female gender (at 3 and 12 months, p = 0.006) were each positively associated with perceived overall impact. CONCLUSIONS: Most epilepsy surgery patients report a positive overall impact of the procedure on their lives and a high willingness to undergo surgery again if that choice could be made. Seizure-free individuals express consistently more positive perceptions of the procedure. Findings suggest that it is important to make early efforts to reintegrate epilepsy surgery patients into employment.

Validity and responsiveness of generic preference-based HRQOL instruments in chronic epilepsy.

Generic, preference-based instruments are recommended for assessing health-related quality of life (HRQOL) in cost-utility analyses (CUA). We aimed to determine which instrument is the most appropriate for CUA of epilepsy care, using established psychometric criteria. We compared validity and responsiveness of EQ5D (using both UK and US preferences), visual analog scale (VAS), Health Utilities Index Mark II (HUI-2) and Mark III (HUI-3) and SF6D in 165 adults evaluated for epilepsy surgery. SF6D had the strongest or next-strongest associations with seizure severity and seizure control. It was not associated with education or IQ. Only SF6D and HUI-3 discriminated between patients with and without seizures 2 years after baseline evaluation. SF6D was most or next-most responsive to being seizure-free for 2 years, in most responsiveness analyses. VAS was also responsive, but showed less evidence of validity. The QOLIE-89, an epilepsy-targeted profile instrument, had stronger evidence for validity and responsiveness than the preference instruments. SF6D has several key psychometric advantages over four other preference instruments in CUAs of epilepsy care. This may reflect better coverage of HRQOL dimensions affected by epilepsy, greater sensitivity at the upper end of the HRQOL continuum, or both. These findings may not generalize to other chronic conditions.

Changes in depression and anxiety after resective surgery for epilepsy.

OBJECTIVE: To determine changes in depression and anxiety after resective surgery. METHODS: Data from subjects enrolled in a prospective multicenter study of resective epilepsy surgery were reviewed with the Beck Psychiatric Symptoms Scales (Beck Depression Inventory [BDI] and Beck Anxiety Inventory [BAI]) and Composite International Diagnostic Interview (CIDI) up to a 24-month period. chi2 analyses were used to correlate proportions. RESULTS: A total of 358 presurgical BDI and 360 BAI results were reviewed. Moderate and severe levels of depression were reported in 22.1% of patients, and similar levels of anxiety were reported by 24.7%. Postoperative rates of depression and anxiety declined at the 3-, 12-, and 24-month follow-up periods. At the 24-month follow-up, moderate to severe levels of depression symptoms were reported in 17.6 and 14.7% of the patients who continued to have postoperative seizures. Moderate to severe depression and anxiety were found in 8.2% of those who were seizure-free. There was no relationship, prior to surgery, between the presence or absence of depression and anxiety and the laterality or location of the seizure onset. There were no significant relationships between depression or anxiety at 24-month follow-up and the laterality or location of the surgery. CONCLUSIONS: Depression and anxiety in patients with refractory epilepsy significantly improve after epilepsy surgery, especially in those who are seizure-free. Neither the lateralization nor the localization of the seizure focus or surgery was associated with the risk of affective symptoms at baseline or after surgery.

Predicting long-term seizure outcome after resective epilepsy surgery: the multicenter study.

BACKGROUND: In a seven-center prospective observational study of resective epilepsy surgery, the authors examined probability and predictors of entering 2-year remission and the risk of subsequent relapse. METHODS: Patients aged 12 years and over were enrolled at time of referral for epilepsy surgery, and underwent standardized evaluation, treatment, and follow-up procedures. The authors defined seizure remission as 2 years completely seizure-free after hospital discharge with or without auras, and relapse as any seizures after 2-year remission. The authors examined type of surgery, seizure, clinical and demographic variables, and localization study results with respect to prediction of seizure remission or relapse, using chi2 and proportional hazards analysis. RESULTS: Of 396 operated patients, 339 were followed over 2 years, and 223 (66%) experienced 2-year remission, not significantly different between medial temporal (68%) and neocortical (50%) resections. In multivariable models, only absence of generalized tonic-clonic seizures and presence of hippocampal atrophy were significantly and independently associated with remission, and only in the medial temporal resection group. Fifty-five patients relapsed after 2-year remission, again not significantly different between medial temporal (25%) and neocortical (19%) resections. Only delay to remission predicted relapse, and only in medial temporal patients. CONCLUSION: Hippocampal atrophy and a history of absence of generalized tonic clonic seizures were the sole predictors of 2-year remission, and only for medial temporal resections.

Do malpractice concerns, payment mechanisms, and attitudes influence test-ordering decisions?

Greater understanding is needed of nonclinical factors that determine neurologists' decisions to order tests. The authors surveyed 595 US neurologists and utilized demographic information, attitude scales, and clinical scenarios to evaluate the influence of nonclinical factors on test-ordering decisions. Greater test reliance, higher malpractice concerns, and receiving reimbursement for testing were all associated with a higher likelihood of test ordering. These findings have implications for training needs and suggest malpractice worries may inflate health care costs.

Initial outcomes in the Multicenter Study of Epilepsy Surgery.

OBJECTIVE: To obtain prospective data regarding seizures, anxiety, depression, and quality of life (QOL) outcomes after resective epilepsy surgery. METHODS: The authors characterized resective epilepsy surgery patients prospectively at yearly intervals for seizure outcome, QOL, anxiety, and depression, using standardized instruments and patient interviews. RESULTS: Of 396 patients who underwent resective surgical procedures, 355 were followed for at least 1 year. Of these, 75% achieved a 1-year remission at some time during follow-up; patients with medial temporal (77%) were more likely than neocortical resections (56%) to achieve remission (p = 0.01). Relapse occurred in 59 (22%) patients who remitted, more often in medial temporal (24%) than neocortical (4%) resected patients (p = 0.02). QOL, anxiety, and depression all improved dramatically within 3 months after surgery (p < 0.0001), with no significant difference based on seizure outcome. After 3 months, QOL in seizure-free patients further improved gradually, and patients with seizures showed gradual declines. By 12 and 24 months, overall QOL and its epilepsy-targeted and physical health domains were significantly different in the two outcome groups. (Anxiety and depression scores also gradually diverged, with improvements in seizure-free and declines in continued seizure groups, but differences were not significant.) CONCLUSION: Resective surgery for treatment of epilepsy significantly reduces seizures, most strikingly after medial temporal resection (77% 1 year remission) compared to neocortical resection (56% 1 year remission). Resective epilepsy surgery has a gradual but lasting effect on QOL, but minimal effects on anxiety and depression. Longer follow-up will be essential to determine ultimate seizure, QOL, and psychiatric outcomes of epilepsy surgery.

Early and late cognitive changes following temporal lobe surgery for epilepsy.

OBJECTIVE: To evaluate early and late postoperative cognitive changes in adult patients who had a standardized temporal lobe resection for control of intractable complex partial seizures and to relate quality of life measures at the long-term follow-up to seizure control and a verbal memory skill. METHODS: Patients in this longitudinal study had intractable epilepsy and had cognitive tests at the initial surgical evaluation, 1 year later, and at a long-term follow-up (at least >9 years later). Mean follow-up was 12.8 years. Participants were 44 patients who had an en bloc left (LTL) or right (RTL) temporal lobe resection and 8 nonsurgical patients with epilepsy. RESULTS: Patients with LTL surgery showed selective early decreases in verbal memory. At the long-term follow-up, further decreases in verbal memory and visual memory scores were seen for all patient groups. In general, the nonmemory scores remained stable over time. LTL surgery and initial high scores were predictors of verbal memory decreases seen at the early follow-up. Late memory declines were predicted by higher 1-year scores. LTL surgery was an additional predictor of late decline on a verbal memory task sensitive to the integrity of the left hippocampus. Higher quality of life scores of patients at the long-term follow-up were associated with both better seizure control and a higher everyday verbal memory skill: prose recall. CONCLUSIONS: Surgery-induced verbal memory deficits following LTL surgery continue 13 years after surgery. Late, possibly age-related, declines in multiple memory scores are seen in patients with temporal lobe surgery and nonsurgical patients with epilepsy. Patients with LTL surgery may be at risk for a more rapid decline in selective verbal memory skills. Evaluations of treatments for intractable epilepsy that compromise memory functions should consider the further quality of life impact of late age-related memory declines.

How long does it take for partial epilepsy to become intractable?

BACKGROUND: Much remains unknown about the natural history of intractable localization-related epilepsy, including how long it typically takes before intractability becomes evident. This information could guide the design of future studies, resolve certain discrepancies in the literature, and provide more accurate information about long-term prognosis. METHODS: Individuals evaluated for resective surgery for refractory localization-related epilepsy were prospectively identified at the time of initial surgical evaluation at seven surgical centers (between 1996 and 2001). The latency time between onset of epilepsy and failure of second medication and history of remission (>/=1 year seizure-free) before surgical evaluation were examined with respect to age at onset, hippocampal atrophy, febrile seizures, and surgical site. RESULTS: In the 333 patients included in the analysis, latency time was 9.1 years (range 0 to 48) and 26% reported a prior remission before surgery. A prior remission of >/=5 years was reported by 8.5% of study participants. Younger age at onset was strongly associated with longer latency time (p < 0.0001) and higher probability of past remission (p < 0.0001). In multivariable analyses, age at onset remained as the most important explanatory variable of both latency time and prior remission. CONCLUSIONS: A substantial proportion of localization-related epilepsy may not become clearly intractable for many years after onset. This is especially true of epilepsy of childhood and early adolescent onset. If prospective studies confirm these findings and the underlying mechanisms behind these associations become understood, this raises the possibility of considering interventions that might interrupt such a process and some day prevent some forms of epilepsy from becoming intractable.

Development of performance measures for acute ischemic stroke.

BACKGROUND AND PURPOSE: The purpose of the present study was to develop and rate performance measures for hospital-based acute ischemic stroke. METHODS: A national multidisciplinary panel of 16 individuals (2 stroke specialists, 2 general neurologists, 2 internists, 2 neuroscience nurses, 2 stroke advocacy organization representatives, 1 stroke rehabilitationist, 1 family practitioner, 1 emergency room physician, 1 neuroradiologist, 1 managed care organization director, and 1 hospital association representative) from 10 medical societies or lay organizations assisted in the development of 44 potential stroke performance measures. We developed evidence summaries for each of the performance measures and graded the level of evidence associated with each measure. The panel received a summary of the literature pertaining to each measure and rated the measures by use of a modified Delphi approach for 6 dimensions of quality, including validity of evidence, feasibility, impact on outcomes, room for improvement, plausibility, and an overall rating (little reason to do, could do, should do, and must do). RESULTS: Highly rated and agreed on performance measures for the overall rating include warfarin in atrial fibrillation, antithrombotics on hospital discharge, carotid imaging in appropriate patients, and use of stroke units. Additional measures notable for high agreement were heparins for deep-vein thrombosis prophylaxis and use of a stroke protocol. Panelists rated time-related thrombolytic measures such as head CT within 25 minutes highly on the room for improvement dimension but low on the overall dimension. Neurologists tended to rate measures lower than did nonneurologists (P<0.01) for all 9 measures pertaining to thrombolytic management. CONCLUSIONS: Highly rated and agreed on performance measures exist in all domains of hospital-based stroke care.

Training clinical researchers in neurology: we must do better.

The pace of scientific discoveries, the increasing complexity of managing patients, and the runaway cost of neurological services have created an urgent need for a wide range of clinical research in neurology. Despite increasing recognition of this need and recent increases in funding for training clinical investigators, neurologists conducting cellular and molecular investigations are more likely to join faculties, maintain research careers, and attain academic advancement. Because academic departments of neurology are successful in producing and nurturing basic science researchers, why aren't they just as triumphant in spawning clinical investigators? This crisis in the preparation of clinical investigators has been brought about by many factors: competing time demands for clinical service, lack of methodologically rigorous training in the disciplines necessary to conduct clinical research, and lack of mentorship. Neurology residents contemplating a clinical research career may observe junior faculty who lack career guidance, are ill-prepared as independent investigators, and must juggle patient demands while trying to write a research grant or conduct a study. Already burdened by medical school debts, is it any wonder that our neurology graduates don't leap to a career with a future that seems so insecure? Academic departments of neurology must develop full-scale clinical research training programs if they are to meet the pressing need for clinical research. As a starting point, they must free themselves from their dependence on providing clinical services to generate income. Following the model which has produced successful basic researchers, much greater effort must be given to establishing rigorous methodological training in collaboration with other departments, creating senior role models, and protecting time for clinical investigators to conduct research. Unless we create incentives to careers in clinical research, we will never answer the growing number of clinical research questions we face today.

Neurologists' use of ICD-9CM codes for dementia.

Neurologists' use of ICD-9CM codes in a sample of 181 consecutive, new patients evaluated for dementia by 48 of all 49 neurologists from the Rochester, NY, area over a 1-year period was examined. The specific code for AD, 331.0, was used for only 36.5% of patients judged by the neurologist to have AD as the most likely diagnosis. Other codes used were not inaccurate but would result in lower reimbursement. Variation in coding could affect validity of dementia research using claims data.

A review of health-related quality-of-life measures in stroke.

The objective of this review was to evaluate health-related quality of life (HR-QOL) measures for use with patients with stroke. HR-QOL measures are increasingly used for assessment in many health conditions; these measures may serve an important role in evaluating the impact of stroke and of stroke interventions. HR-QOL measures used in patients with stroke should: (i) cover the domains of HR-QOL that may be affected by stroke; (ii) have administration characteristics suitable for use in patients with stroke; and (iii) have undergone reliability and validity assessment in patients with stroke. The present study evaluates HR-QOL measures with reference to these requirements. A systematic literature review was conducted to identify and evaluate HR-QOL measures of potential use in studies of patients with stroke. Identified measures were assessed with regard to stroke-relevant domains covered, measure characteristics (e.g. self-administration versus administration by an interviewer, interviewer time to complete) and psychometric properties of reliability and validity. The measures evaluated vary widely on domains covered, and limited assessment of the performance of HR-QOL measures has been conducted in patients with stroke. No existing measure comprehensively covers all relevant domains or addresses fully the issues of obtaining and combining HR-QOL assessments in patients and proxies in many stroke populations. Additional psychometric testing in stroke populations is needed for existing HR-QOL measures. In addition, stroke-targeted HR-QOL measures need to be developed and evaluated with patients with stroke.

Factors related to agreement between self-reported and conventional Expanded Disability Status Scale (EDSS) scores.

BACKGROUND: Although the Expanded Disability Status Scale (EDSS) remains a widely used scale for evaluating impairments in people with multiple sclerosis (MS), EDSS assessments are infeasible in certain situations. A self-administered version of the EDSS would be potentially useful if it yielded similar results as the conventional physician-based version. METHODS: We developed a self-administered patient questionnaire to obtain ratings of neurologic impairments, and developed algorithms to estimate EDSS scores. We mailed the questionnaires to all new consecutive patients scheduled to be seen at an MS clinic. Questionnaires were completed prior to the visit and traditional EDSS ratings were made by one of two neurologists at the visit. One hundred and forty-six pairs of patient questionnaires and physician EDSS assessments were obtained. RESULTS: Kappa values for agreement between the physician's EDSS scores and the questionnaire-derived scores were 0.13 (for exact agreement), 0.39 (+/-0.5 EDSS steps), and 0.56 (+/-1.0 EDSS steps). A scatterplot showed that agreement was best at EDSS scores <3.0 and >5.0. Better agreement was obtained when patients had a higher level of education, and when the physician was more certain of the diagnosis of MS. CONCLUSIONS: While the self-assessed EDSS scores do not agree highly enough to take the place of conventional EDSS scores, they may be sufficient for MS trial screening or for assessing outcomes across broad categories of disability.

Management of multiple sclerosis across managed care and fee-for-service systems.

OBJECTIVE: To measure and compare care for adults with MS across managed care and fee-for-service (FFS) health systems. METHODS: The authors sampled adults with MS having physician visits over a 2-year period from a group model health maintenance organization (HMO) in southern California, from a midwestern independent practice association (IPA) model managed care plan, and from the FFS portion of the practices of a random sample of southern California neurologists. The authors mailed surveys to subjects in mid-1996; 930 of 1,164 (80%) of those eligible responded. The authors measured sociodemographic and clinical characteristics, management of recent changes in mobility, bladder control, and fatigue, use of a disease-modifying agent, assessment of general health symptoms and issues, and unmet information needs. The authors adjusted comparisons between systems for comorbidity, disease severity, and disease type. RESULTS: The groups differed on most sociodemographic and clinical characteristics. There were few differences in symptom management; differences that did exist tended toward more referrals or treatment for the FFS group. Access to the disease-modifying agent available at the time of the survey did not differ across systems, although patients' perceptions of the rationale for not using the drug did vary. General health issues and symptoms were more often assessed in the FFS and IPA systems than in the HMO, but improvement was needed across all three systems of care. There were substantial unmet information needs in all groups and especially high ones in the FFS and HMO samples. CONCLUSIONS: Strategies to improve care for people with MS should be developed and evaluated, particularly in areas like symptom assessment and meeting patient information needs. Where variations in service delivery exist, longitudinal studies are also needed to evaluate the potential impact on outcomes and to evaluate reasons for variation.

Systematic review of clinical prediction rules for neuroimaging in the evaluation of dementia.

BACKGROUND: Clinical practice guidelines for dementia do not recommend routine neuroimaging but vary in their recommended clinical prediction rules to identify patients who should undergo neuroimaging for potentially reversible causes of dementia. METHODS: Using a MEDLINE search supplemented by other strategies, we identified studies from January 1, 1983, through December 31, 1998, that evaluated the diagnostic performance of a clinical prediction rule. We calculated the sensitivity and specificity of each rule, then evaluated their diagnostic performance in a hypothetical cohort of 1000 patients with dementia, varying the prevalence of potentially reversible dementia from 1% to 15%. RESULTS: We identified 7 studies that evaluated at least 1 of 6 different clinical prediction rules. Only one rule consistently had high sensitivity (>85%) across all studies; none consistently had high specificity (>85%). Six of the 7 studies included less than 15 cases of potentially reversible dementia; thus the sensitivity and specificity for each rule had relatively wide confidence intervals. At a 5% prevalence of potentially reversible dementia, all rules had low positive predictive value (<15%) in our hypothetical cohort. Depending on the rule, our analysis predicts 6 to 44 of the 50 patients with potentially reversible dementia (5% prevalence in cohort of 1000 patients) would not undergo imaging. CONCLUSIONS: There is considerable uncertainty in the evidence underlying clinical prediction rules to identify which patients with dementia should undergo neuroimaging. Application of these rules may miss patients with potentially reversible causes of dementia.