Search

For many diseases or medical conditions, there is no perfect diagnostic test. The simplest possible LCM assumes the observed results arise from a mixture of two latent classes (or states) labeled "disease" and "disease free" within which the four observed tests are independent. For example, when evaluating a new test against an imperfect reference test with poor sensitivity, it is possible that both tests give a false negative result for a less severe case of infection. It also complicates evaluation of a new diagnostic test for the disease, as it is unknown whether an individual with a positive result on the new test and a negative result on the best available test is a true positive or a false positive. [Extracted from the article]

When multiple imperfect dichotomous diagnostic tests are applied to an individual, it is possible that some or all of their results remain dependent even after conditioning on the true disease status. The estimates could be biased if this conditional dependence is ignored when using the test results to infer about the prevalence of a disease or the accuracies of the diagnostic tests. However, statistical methods correcting for this bias by modelling higher-order conditional dependence terms between multiple diagnostic tests are not well addressed in the literature. This paper extends a Bayesian fixed effects model for 2 diagnostic tests with pairwise correlation to cases with 3 or more diagnostic tests with higher order correlations. Simulation results show that the proposed fixed effects model works well both in the case when the tests are highly correlated and in the case when the tests are truly conditionally independent, provided adequate external information is available in the form of fixed constraints or prior distributions. A data set on the diagnosis of childhood pulmonary tuberculosis is used to illustrate the proposed model. [ABSTRACT FROM AUTHOR]

Background Streptococcus pneumoniae (SP) pneumonia is often treated empirically as diagnosis is challenging because of the lack of a perfect test. Using BinaxNOW-SP, a urinary antigen test, as an add-on to standard cultures may not only increase diagnostic yield but also increase costs. Objective To estimate the sensitivity and specificity of BinaxNOW-SP and subsequently estimate the cost-effectiveness of adding BinaxNOW-SP to the diagnostic work-up. Design We fit a Bayesian latent-class meta-analysis model to obtain estimates of BinaxNOW-SP accuracy that adjust for the imperfect accuracy of culture. Meta-analysis results were combined with information on prevalence of SP pneumonia to estimate the number of patients who are correctly classified under competing diagnostic strategies. Taking into consideration the cost of antibiotics, we determined the incremental cost of adding BinaxNOW-SP to the work-up per case correctly diagnosed. Results The BinaxNOW-SP test had a pooled sensitivity of 0.74 (95% credible interval [CrI], 0.67-0.83) and a pooled specificity of 0.96 (95% CrI, 0.92-0.99). An overall increase in diagnostic accuracy of 6.2% due to the addition of BinaxNOW-SP corresponded to an incremental cost per case correctly classified of $582 Canadian dollars. Conclusions The methods we have described allow us to evaluate the accuracy and economic value of a new test in the absence of a perfect reference test using an evidence-based approach. [ABSTRACT FROM AUTHOR]

When two imperfect diagnostic tests are carried out on the same subject, their results may be correlated even after conditioning on the true disease status. While past work has focused on the consequences of ignoring conditional dependence, the degree to which conditional dependence can be induced has not been systematically studied. We examine this issue in detail by introducing a hypothetical missing covariate that affects the sensitivities of two imperfect dichotomous tests. We consider four forms for this covariate, normal, uniform, dichotomous and trichotomous. In the case of a dichotomous covariate, we derive an expression showing that the conditional covariance is a function of the product of the changes in test sensitivities (or specificities) between the subgroups defined by the covariate. The maximum possible covariance is induced by a dichotomous covariate with a very strong effect on both tests. Through simulations, we evaluate the extent to which fitting a latent class model ignoring each type of covariate but including a general covariance term can adjust for the correlation induced by the covariate. We compare the results to when the conditional dependence is ignored. We find that the bias because of ignoring conditional dependence is generally small even for moderate covariate effects, and when bias is present, a model including a covariance term works well. We illustrate our methods by analyzing data from a childhood tuberculosis study. Copyright © 2016 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Composite reference standards (CRSs) have been advocated in diagnostic accuracy studies in the absence of a perfect reference standard. The rationale is that combining results of multiple imperfect tests leads to a more accurate reference than any one test in isolation. Focusing on a CRS that classifies subjects as disease positive if at least one component test is positive, we derive algebraic expressions for sensitivity and specificity of this CRS, sensitivity and specificity of a new (index) test compared with this CRS, as well as the CRS-based prevalence. We use as a motivating example the problem of evaluating a new test for Chlamydia trachomatis, an asymptomatic disease for which no gold-standard test exists. As the number of component tests increases, sensitivity of this CRS increases at the expense specificity, unless all tests have perfect specificity. Therefore, such a CRS can lead to significantly biased accuracy estimates of the index test. The bias depends on disease prevalence and accuracy of the CRS. Further, conditional dependence between the CRS and index test can lead to over-estimation of index test accuracy estimates. This commonly-used CRS combines results from multiple imperfect tests in a way that ignores information and therefore is not guaranteed to improve over a single imperfect reference unless each component test has perfect specificity, and the CRS is conditionally independent of the index test. When these conditions are not met, as in the case of C. trachomatis testing, more realistic statistical models should be researched instead of relying on such CRSs. [ABSTRACT FROM AUTHOR]

The prevailing lack of consensus about the comparative harms and benefits of cancer screening stems, in part, from the inappropriate calculations of the expected mortality impact of a sustained screening programme. There is an inherent, and often substantial, time lag from the time of screening until the resulting mortality reductions begin, reach their maximum and ultimately end. However, the cumulative mortality reduction reported in a randomised screening trial is typically calculated over an arbitrarily defined follow-up period, including follow-up time where the mortality impact is yet to realise or where it has already been exhausted. Because of this, the cumulative reduction cannot be used for projecting the mortality impact expected from a sustained screening programme. For this purpose, we propose a new measure, the time-specific probability of being helped by screening, given that the cancer would have proven fatal otherwise. This can be decomposed into round-specific impacts, which in turn can be parametrised and estimated from the trial data. This represents a major shift in quantifying the benefits due to a sustained screening programme, based on statistical evidence extracted from existing trial data. We illustrate our approach using data from screening trials in lung and colorectal cancers. [ABSTRACT FROM AUTHOR]

Heterogeneity in diagnostic meta-analyses is common because of the observational nature of diagnostic studies and the lack of standardization in the positivity criterion (cut-off value) for some tests. So far the unexplained heterogeneity across studies has been quantified by either using the I2 statistic for a single parameter (i.e. either the sensitivity or the specificity) or visually examining the data in a receiver-operating characteristic space. In this paper, we derive improved I2 statistics measuring heterogeneity for dichotomous outcomes, with a focus on diagnostic tests. We show that the currently used estimate of the 'typical' within-study variance proposed by Higgins and Thompson is not able to properly account for the variability of the within-study variance across studies for dichotomous variables. Therefore, when the between-study variance is large, the 'typical' within-study variance underestimates the expected within-study variance, and the corresponding I2 is overestimated. We propose to use the expected value of the within-study variation in the construction of I2 in cases of univariate and bivariate diagnostic meta-analyses. For bivariate diagnostic meta-analyses, we derive a bivariate version of I2 that is able to account for the correlation between sensitivity and specificity. We illustrate the performance of these new estimators using simulated data as well as two real data sets. Copyright © 2014 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Diagnostic tests rarely provide perfect results. The misclassification induced by imperfect sensitivities and specificities of diagnostic tests must be accounted for when planning prevalence studies or investigations into properties of new tests. The previous work has shown that applying a single imperfect test to estimate prevalence can often result in very large sample size requirements, and that sometimes even an infinite sample size is insufficient for precise estimation because the problem is non-identifiable. Adding a second test can sometimes reduce the sample size substantially, but infinite sample sizes can still occur as the problem remains non-identifiable. We investigate the further improvement possible when three diagnostic tests are to be applied. We first develop methods required for studies when three conditionally independent tests are available, using different Bayesian criteria. We then apply these criteria to prototypic scenarios, showing that large sample size reductions can occur compared to when only one or two tests are used. As the problem is now identifiable, infinite sample sizes cannot occur except in pathological situations. Finally, we relax the conditional independence assumption, demonstrating in this once again non-identifiable situation that sample sizes may substantially grow and possibly be infinite. We apply our methods to the planning of two infectious disease studies, the first designed to estimate the prevalence of Strongyloides infection, and the second relating to estimating the sensitivity of a new test for tuberculosis transmission. The much smaller sample sizes that are typically required when three as compared to one or two tests are used should encourage researchers to plan their studies using more than two diagnostic tests whenever possible. User-friendly software is available for both design and analysis stages greatly facilitating the use of these methods. Copyright © 2010 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Verification bias arises in diagnostic test evaluation studies when the results from a first test are verified by a reference test only in a non-representative subsample of the original study subjects. This occurs, for example, when inclusion probabilities for the subsample depend on first-stage results and/or on a covariate related to disease status. Reference standard bias arises when the reference test itself has imperfect sensitivity and specificity, but this information is ignored in the analysis. Reference standard bias typically results in underestimation of the sensitivity and specificity of the test under evaluation, since subjects that are correctly diagnosed by the test can be considered as misdiagnosed owing to the imperfections in the reference standard. In this paper, we describe a Bayesian approach for simultaneously addressing both verification and reference standard bias. Our models consider two types of verification bias, first when subjects are selected for verification based on initial test results alone, and then when selection is based on initial test results and a covariate. We also present a model that adjusts for a third potential bias that arises when tests are analyzed assuming conditional independence between tests, but some dependence exists between the initial test and the reference test. We examine the properties of our models using simulated data, and then apply them to a study of a screening test for dementia, providing bias-adjusted estimates of the sensitivity and specificity. Copyright © 2010 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Background Clopidogrel is an antiplatelet drug that is prescribed after percutaneous coronary intervention (PCI) to prevent stent thrombosis. Previous studies have suggested that some statins may inhibit the antiplatelet effects of clopidogrel via competitive metabolism of its activating enzyme cytochrome P450 3A4 (CYP3A4). Objectives To investigate a possible interaction between statins and clopidogrel after a PCI procedure in a population-based cohort study. Methods A population-based cohort study was carried out between January 2001 and December 2004 using the health insurance databases from Quebec, Canada. The primary endpoint was a composite of death from any cause, myocardial infarction (MI), unstable angina, repeat revascularization and cerebrovascular events. PCI patients ≥ 66 years of age were followed from their initial post-discharge clopidogrel prescription until the earliest of study endpoint occurrence, end of clopidogrel exposure or end of study (90 days post discharge). Time-dependent Cox regression analysis was performed. Results We identified 10 491 patients who were prescribed clopidogrel post-PCI and 43.5% were also prescribed statins at the baseline discharge. During 1793 patient years of follow-up, 623 composite endpoints were observed. Compared to the reference group (non-CYP3A4-metabolized statins), the co-prescription of CYP3A4-metabolized statins (hazard ratio (HR) 1.16, 95% confidence interval (CI) 0.91-1.47), or no statin use (HR 1.22, 95%CI 0.93-1.59) were not statistically associated with an increase in adverse outcomes. Conclusions In this PCI cohort, the association of clopidogrel with CYP3A4-metabolized statins did not demonstrate an increased early risk of adverse cardiovascular events, although a small risk could not be completely excluded. Copyright © 2009 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Applications of latent class analysis in diagnostic test studies have assumed that all tests are measuring a common binary latent variable, the true disease status. In this article we describe a new approach that recognizes that tests based on different biological phenomena measure different latent variables, which in turn measure the latent true disease status. This allows for adjustment of conditional dependence between tests within disease categories. The model further allows for the inclusion of measured covariates and unmeasured random effects affecting test performance within latent classes. We describe a Bayesian approach for model estimation and describe a new posterior predictive check for evaluating candidate models. The methods are motivated and illustrated by results from a study of diagnostic tests for Chlamydia trachomatis. Published in 2008 by John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

To describe the prevalence of and characteristics associated with major and minor depression in older medical inpatients and to compare associated characteristics by sex and history of depression. Cross-sectional study of two patient samples, with and without a screening diagnosis of major or minor depression. The medical services of two acute care hospitals. Medical admissions of people aged 65 and older with at most mild cognitive impairment (N=380). Diagnoses of major and minor depression (Diagnostic Interview Schedule), cognitive impairment (Mini-Mental State Examination), premorbid disability, sociodemographic variables (including social networks and support), comorbidity, severity of illness, history of depression. The prevalence of major depression differed by hospital, ranging from 14.2% (95% confidence interval (CI)=11.7–17.1) in Hospital A to 44.5% (95% CI=33.1–56.4) in Hospital B. The prevalence of minor depression was similar in the two hospitals, ranging from 9.4% (95% CI=7.4–11.9) in Hospital A to 7.9% (95% CI=2.9–16.3) in Hospital B. After adjustment for hospital, the same characteristics (history of depression, premorbid disability, cognitive impairment, perceived adequacy of support, and visits from friends) were associated with major and minor depression, although most of these associations tended to be weaker for minor depression. Most of these factors were also associated with depression in multivariate analyses. The most important characteristics in women were premorbid disability, history of depression, and adequacy of emotional support; in men they were history of depression, cognitive impairment, and adequacy of emotional support. A cerebrovascular or other cardiovascular diagnosis did not explain the association between depression and cognitive impairment. Major and minor depression occur frequently in older medical inpatients and are associated with similar patient characteristics. A history of depression and the patient's sex should be considered in the identification and interpretation of these associated factors. [ABSTRACT FROM AUTHOR]

To assess the reliability, validity, and responsiveness of an instrument for measuring the severity of delirium, the Delirium Index (DI).Prospective cohort study, with repeated patient assessments at multiple points in the hospital, at 8 weeks after discharge, and at 6 and 12 months after admission.The medical services of a primary acute-care hospital.Medical admissions aged 65 and older: 165 with delirium and dementia, 57 with delirium only, 55 with dementia only, and 41 with neither.Severity of delirium symptoms was measured using the DI. Delirium was diagnosed using the Confusion Assessment Method. Other measures included the Mini-Mental State Examination, Informant Questionnaire on Cognitive Decline in the Elderly, Barthel Index (BI), premorbid instrumental activities of daily living, Charlson Comorbidity Index, Clinical Severity of Illness scale (CSI), and the Acute Physiology Score (APS).The intraclass correlation coefficient of interrater reliability was 0.98. Two measures of fluctuation were significantly higher in patients with delirium than in those without delirium. At baseline, the DI was correlated with the BI, APS, and CSI in delirious patients with (correlation coefficient (r)=−0.43, 0.17, and 0.36, respectively) or without (r=−0.44, 0.39, 0.22, respectively) dementia. At 8 weeks, in delirious patients with and without dementia, internal responsiveness as measured by effect sizes was−0.60 and−0.74, respectively, and the standardized response mean for both groups was−0.64. Low to good levels of external responsiveness were found.The DI appears to be a reliable, valid, and responsive measure of the severity of delirium, in patients with delirium, with or without dementia. [ABSTRACT FROM AUTHOR]

Objectives To examine the relationship between depression among medically ill, frail elders and family caregivers' hours of care, health status, and quality of life. Design and Methods A cross-sectional study of 193 family caregivers of seniors treated in the emergency department (ED) was conducted. Measures included patient depression (Geriatric Depression Scale—15), and caregivers' hours of care, mental health and physical functioning (SF-36), and quality of life (EQ-5D). Results Mean caregiver age was 60.0 ± 16.1 years and 70.5% were female. More caregivers of depressed seniors provided more care in the previous month (37.3% vs 22.4%, p = 0.03), had poor mental health (63.5% vs 47.0%, p = 0.03), and poor perceived quality of life (63.5% vs 50.4%, p = 0.04) compared to caregivers of non-depressed seniors. Multiple logistic regression analyses indicated that patient depression was associated with poor caregiver quality of life (OR = 3.15, 95% CI 1.48, 6.73), and poor mental health in spousal and adult child caregivers (OR = 2.72, 95% CI = 0.88, 8.39, and OR = 3.29, 95% CI = 1.10, 9.86, respectively). Conclusions Psychosocial support may be needed for caregivers of depressed seniors. Copyright © 2004 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

To evaluate the validity of the Identification of Seniors at Risk (ISAR) screening tool for detecting severe functional impairment and depression and predicting increased depressive symptoms and increased utilization of health services. Four university-affiliated hospitals in Montreal. Data from two previous studies were available: Study 1, in which the ISAR scale was developed (n=1,122), and Study 2, in which it was used to identify patients for a randomized trial of a nursing intervention (n=1,889 with administrative data, of which 520 also had clinical data). Patients aged 65 and older who were to be released from an emergency department (ED). Baseline validation criteria included premorbid functional status in both studies and depression in Study 2 only. Increase in depressive symptoms at 4-month follow-up was assessed in Study 2. Information on health services utilization during the 5 months after the ED visit (repeat ED visits and hospitalization in both studies, visits to community health centers in Study 2) was available by linkage with administrative databases. Estimates of the area under the receiver operating characteristic curve (AUC) for concurrent validity of the ISAR scale for severe functional impairment and depression ranged from 0.65 to 0.86. Estimates of the AUC for predictive validity for increased depressive symptoms and high utilization of health services ranged from 0.61 to 0.71. The ISAR scale has acceptable to excellent concurrent and predictive validity for a variety of outcomes, including clinical measures and utilization of health services. [ABSTRACT FROM AUTHOR]

To determine the effects of prevalent and incident delirium on length of hospital stay. Prospective cohort study, comparing (1) length of stay after admission in cases of prevalent delirium versus controls without prevalent delirium with (2) length of stay after diagnosis in cases of incident delirium versus controls matched by day of diagnosis. The medical services of a primary, acute care hospital. Medical admissions of patients aged 65 and older from the emergency department with delirium diagnosed during the first week in hospital. Patients admitted to intensive care or oncology and those with a primary diagnosis of stroke were excluded. A sample of those without delirium was also enrolled. Delirium was diagnosed using the Confusion Assessment Method. Data on length of stay and diagnosis-related groups (DRGs) were abstracted from administrative data. Measures of covariates included the Informant Questionnaire on Cognitive Decline in the Elderly, the Delirium Index, the instrumental activities of daily living questionnaire from the Older American Resources and Services project, the Charlson Comorbidity Index, the Clinical Severity Scale, and the Acute Physiology Score. The study sample comprised 359 patients: 204 with prevalent delirium, 37 with incident delirium, and 118 without delirium. After controlling for covariates, prevalent delirium was not associated with a significantly longer hospital stay, but incident delirium was associated with an excess stay after diagnosis of 7.78 days (95% confidence interval=3.07, 12.48). Similar results were obtained using log-transformed or DRG-adjusted estimates of length of stay. In older medical inpatients, incident but not prevalent delirium is an important predictor of longer hospital stay. Interventions to prevent incident delirium may reduce length of stay. [ABSTRACT FROM AUTHOR]

OBJECTIVES: To determine the prognostic significance of subsyndromal delirium (SSD) presentations. DESIGN: Cohort study. SETTING: University-affiliated primary acute care hospital. PARTICIPANTS: One hundred sixty-four elderly medical inpatients who did not meet Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised (DSM-III-R) criteria for delirium during the first week after admission were classified into three mutually exclusive groups. The first group, prevalent SSD, included those who had two or more of four core symptoms of delirium (clouding of consciousness, inattention, disorientation, perceptual disturbances) at admission. The second group, incident SSD, included those who did not meet criteria for prevalent SSD but displayed one or more new core symptoms during the week after admission. The third group had no prevalent or incident SSD. The three groups were followed up at 2, 6, and 12 months. MEASUREMENTS: Outcomes (length of stay, symptoms of delirium (Delirium index), cognitive (Mini-Mental State Examination) and functional status (instrumental activities of daily living), and mortality) were compared using univariate techniques and multivariate regression models that adjusted for age, sex, marital status, living arrangements before admission, comorbidity, clinical and physiological severity of illness, and dementia status and severity. RESULTS: Patients with prevalent SSD had longer acute care hospital stay, increased postdischarge mortality, more symptoms of delirium, and a lower cognitive and functional level at follow-up than patients with no SSD. Most of the findings for incident SSD were similar but not statistically significant. Patients with prevalent or incident SSD had risk factors for DSM-defined delirium. CONCLUSION: SSD in elderly medical inpatients appears to be a clinically important syndrome that falls on a continuum between no symptoms and DSM-defined delirium. [ABSTRACT FROM AUTHOR]

Objective. To examine whether availability of cardiac services at the admitting hospital affects case–selection for angiography and one–year survival following angiography, within groups of patients who have similar clinical need for angiography according to published criteria. Study Setting. Elderly Medicare beneficiaries (37,788) discharged with a diagnosis of acutemyocardial infarction (AMI) from hospitals in seven U.S. states between February 1994 and July 1995.We focused on patients who were eligible to receive angiography 12 or more hours after symptom onset. Data Collection. Data were abstracted from patient’s medical records, Medicare National Claims Standard Analytic Files, Health Care Financing Administration (HCFA) Provider of Service File and Health Insurance Master File. Methods. Admitting hospitals were classified as offering no cardiac services, angiography only, or revascularization. Case–selection differences across these three types of hospitals were examined by comparing relative risk of receiving angiography for various patient and hospital characteristics. Relative differences in one–year survival rate, comparing patients who received angiography to those who did not, were estimated within each hospital type and clinical need category (necessary, appropriate, or uncertain) after matching on propensity to receive angiography. Principal Findings. Compared to patients for whom angiography was deemed necessary, the relative risk of receiving angiography among those for whom it was deemed of uncertain benefit was 0.58, 0.79, and 0.92 (p–value of homogeneity test < 0.001) at hospitals offering no cardiac services, angiography only, and revascularization, respectively. There was no significant difference in survival following angiography across hospital types, overall as well as within clinical need categories. Conclusions. Despite increased case selection at hospitals with on–site cardiac services, there was no evidence of increase in the survival rate associated with angiography use at these hospitals. [ABSTRACT FROM AUTHOR]

Aims There is evidence that large doses of inhaled corticosteroids lead to an increased risk of glaucoma, cataracts and other problems associated with oral corticosteroid use. However, no formal investigation so far has been conducted into the relationship between inhaled corticosteroids and diabetes. Methods Our nested case-control design studied the association between current use of inhaled corticosteroids and the risk of using antidiabetic medications among a cohort of 21 645 elderly subjects. We also investigated the possibility of a dose–response relationship in users of beclomethasone. Data were obtained from the medical and pharmaceutical databases of the Regie de l'assurance maladie du Québec. Results Within the cohort, we identified 1494 cases and we selected 14 931 controls using density sampling. The unadjusted rate ratio (and 95% confidence interval, CI) for developing diabetes among current users of inhaled corticosteroids was 1.4 (1.2, 1.5). After adjusting for covariates, the rate ratio (95% CI) decreased to 0.9 (0.8, 1.1). The loss of statistical significance was due in large part to adjusting for the current use of oral corticosteroids. We also did not observe a statistically significant increase in risk among users of high-dose beclomethasone compared to nonusers, after adjusting for covariates. Conclusions Our results do not indicate an increased risk of diabetes among current users of inhaled corticosteroids. [ABSTRACT FROM AUTHOR]

Background Studies of auditory hallucinations are scant. Method To determine the prevalence and phenomenology of auditory hallucinations among elderly subjects with hearing impairment. Objective We surveyed 125 men and women aged 65 years and over referred to the Audiology department of a university-affiliated primary acute-care hospital. Results The prevalence of auditory hallucinations was 32.8% (95% confidence interval (CI) 24.7–41.8). These hallucinations represented a spectrum of phenomenology from elementary personal impressions to complex percepts. The types of auditory hallucinations included humming or buzzing (35.9%), shushing (12.8%), beating or tapping (10.6%), ringing (7.7%), other individual sounds (15.4%), multiple sounds (12.6%), voices (2.5%) or music (2.5%). Subjects with any type of hallucination were younger and had poorer discrimination scores in the left ear and impaired binaural discrimination with lip-reading. Subjects with hallucinations that had more qualities of a true percept heard different types of sounds and had lower reflex thresholds and better air conduction in the right ear. Conclusion Auditory hallucinations are frequent in elderly subjects with hearing impairment and seem to be associated with younger age and asymmetrical hearing impairment. Copyright © 2002 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

OBJECTIVES:To determine the effectiveness of a two-stage (screening and nursing assessment) intervention for older patients in the emergency department (ED) who are at increased risk of functional decline and other adverse outcomes. DESIGN:Controlled trial, randomized by day of ED visit, with follow-up at 1 and 4 months. SETTING:Four university-affiliated hospitals in Montreal. PARTICIPANTS:Patients age 65 and older expected to be released from the ED to the community with a score of 2 or more on the Identification of Seniors At Risk (ISAR) screening tool and their primary family caregivers. One hundred seventy-eight were randomized to the intervention, 210 to usual care. INTERVENTION:The intervention consisted of disclosure of results of the ISAR screen, a brief standardized nursing assessment in the ED, notification of the primary care physician and home care providers, and other referrals as needed. The control group received usual care, without disclosure of the screening result. MEASUREMENTS:Patient outcomes assessed at 4 months after enrollment included functional decline (increased dependence on the Older American Resources and Services activities of daily living scale or death) and depressive symptoms (as assessed by the short Geriatric Depression Scale). Caregiver outcomes, also assessed at baseline and 4 months, included the physical and mental summary scales of the Medical Outcomes Study Short Form-36. Patient and caregiver satisfaction with care were assessed 1 month after enrollment. RESULTS:The intervention increased the rate of referral to the primary care physician and to home care services. The intervention was associated with a significantly reduced rate of functional decline at 4 months, in both unadjusted (odds ratio (OR) = 0.60, 95% confidence interval (CI) = 0.36–0.99) and adjusted (OR = 0.53, 95% CI = 0.31–0.91) analyses. There was no intervention effect on patient depressive symptoms, caregiver outcomes, or satisfaction with care. CONCLUSION:A two-stage ED intervention, consisting of screening with the ISAR tool followed by a brief, standardized nursing assessment and referral to primary and home care services, significantly reduced the rate of subsequent functional decline. J Am Geriatr Soc 49:1272–1281, 2001. [ABSTRACT FROM AUTHOR]