Osteoporosis identification among previously undiagnosed individuals with vertebral fractures.

Because osteoporosis is under-recognized in patients with vertebral fractures, we evaluated characteristics associated with osteoporosis identification. Most patients with vertebral fractures did not receive evaluation or treatment for osteoporosis. Black, younger, and male participants were particularly unlikely to have had recognized osteoporosis, which could increase their risk of negative outcomes. INTRODUCTION: Vertebral fractures may be identified on imaging but fail to prompt evaluation for osteoporosis. Our objective was to evaluate characteristics associated with clinical osteoporosis recognition in patients who had vertebral fractures detected on their thoracolumbar spine imaging reports. METHODS: We prospectively identified individuals who received imaging of the lower spine at primary care clinics in 4 large healthcare systems who were eligible for osteoporosis screening and lacked indications of osteoporosis diagnoses or treatments in the prior year. We evaluated characteristics of participants with identified vertebral fractures that were associated with recognition of osteoporosis (diagnosis code in the health record; receipt of bone mineral density scans; and/or prescriptions for anti-osteoporotic medications). We used mixed models to estimate adjusted odds ratios (ORs) and 95% confidence intervals (95% CIs). RESULTS: A total of 114,005 participants (47% female; mean age 65 (interquartile range: 57-72) years) were evaluated. Of the 8579 (7%) participants with vertebral fractures identified, 3784 (44%) had recognition of osteoporosis within the subsequent year. In adjusted regressions, Black participants (OR (95% CI): 0.74 (0.57, 0.97)), younger participants (age 50-60: 0.48 (0.42, 0.54); age 61-64: 0.70 (0.60, 0.81)), and males (0.39 (0.35, 0.43)) were less likely to have recognized osteoporosis compared to white participants, adults aged 65 + years, or females. CONCLUSION: Individuals with identified vertebral fractures commonly did not have recognition of osteoporosis within a year, particularly those who were younger, Black, or male. Providers and healthcare systems should consider efforts to improve evaluation of osteoporosis in patients with vertebral fractures.

Lumbar Spinal Stenosis Severity by CT or MRI Does Not Predict Response to Epidural Corticosteroid versus Lidocaine Injections.

BACKGROUND AND PURPOSE: Epidural steroid injections may offer little-to-no short-term benefit in the overall population of patients with symptomatic spinal stenosis compared with lidocaine alone. We investigated whether imaging could identify subgroups of patients who might benefit most. MATERIALS AND METHODS: A secondary analysis of the Lumbar Epidural Steroid Injections for Spinal Stenosis prospective, double-blind trial was performed, and patients were randomized to receive an epidural injection of lidocaine with or without corticosteroids. Patients (n = 350) were evaluated for qualitative and quantitative MR imaging or CT measures of lumbar spinal stenosis. The primary clinical end points were the Roland-Morris Disability Questionnaire and the leg pain numeric rating scale at 3 weeks following injection. ANCOVA was used to assess the significance of interaction terms between imaging measures of spinal stenosis and injectate type on clinical improvement. RESULTS: There was no difference in the improvement of disability or leg pain scores at 3 weeks between patients injected with epidural lidocaine alone compared with corticosteroid and lidocaine when accounting for the primary imaging measures of qualitative spinal stenosis assessment (interaction coefficients for disability score, -0.1; 95% CI, -1.3 to 1.2; P = .90; and for the leg pain score, 0.1; 95% CI, -0.6 to 0.8; P = .81) or the quantitative minimum thecal sac cross-sectional area (interaction coefficients for disability score, 0.01; 95% CI, -0.01 to 0.03; P = .40; and for the leg pain score, 0.01; 95% CI, -0.01 to 0.03; P = .33). CONCLUSIONS: Imaging measures of spinal stenosis are not associated with differential clinical responses following epidural corticosteroid injection.

Dynamic thresholds and a summary ROC curve: Assessing prognostic accuracy of longitudinal markers.

Cancer patients, chronic kidney disease patients, and subjects infected with HIV are routinely monitored over time using biomarkers that represent key health status indicators. Furthermore, biomarkers are frequently used to guide initiation of new treatments or to inform changes in intervention strategies. Since key medical decisions can be made on the basis of a longitudinal biomarker, it is important to evaluate the potential accuracy associated with longitudinal monitoring. To characterize the overall accuracy of a time-dependent marker, we introduce a summary ROC curve that displays the overall sensitivity associated with a time-dependent threshold that controls time-varying specificity. The proposed statistical methods are similar to concepts considered in disease screening, yet our methods are novel in choosing a potentially time-dependent threshold to define a positive test, and our methods allow time-specific control of the false-positive rate. The proposed summary ROC curve is a natural averaging of time-dependent incident/dynamic ROC curves and therefore provides a single summary of net error rates that can be achieved in the longitudinal setting.

Associations between diet and cardiometabolic risk among Yup'ik Alaska Native people using food frequency questionnaire dietary patterns.

BACKGROUND AND AIMS: In previous analyses, we identified three dietary patterns from food frequency questionnaire data among a sample of Yup'ik Alaska Native people living in Southwest Alaska: a "subsistence foods" dietary pattern and two market-based dietary patterns "processed foods" and "fruits and vegetables". In this analysis, we aimed to characterize the association between the dietary patterns and cardiometabolic (CM) risk factors (lipids, blood pressure, glucose, adiposity). METHODS AND RESULTS: We used multilevel linear regression to estimate the mean of each CM risk factor, comparing participants in the 4th to the 1st quartile of each dietary pattern (n = 637). Models were adjusted for age, sex, past smoking, current smoking, and physical activity. Mean log triglyceride levels were significantly higher among participants in the 4th compared to the 1st quartile of the processed foods dietary pattern (ß = 0.11). Mean HbA1c percent was significantly lower (ß = -0.08) and mean diastolic blood pressure (DBP) mm Hg was significantly higher (ß = 2.87) among participants in the 4th compared to the 1st quartile of the fruits and vegetables dietary pattern. Finally, mean log triglyceride levels and mean DBP mm Hg were significantly lower among participants in the 4th compared to the 1st quartile of the subsistence foods dietary pattern (ß = -0.10 and ß = -3.99 respectively). CONCLUSIONS: We found increased CM risk, as reflected by increased triglycerides, associated with eating a greater frequency of processed foods, and reduced CM risk, as reflected by lower triglycerides and DBP, associated with eating a greater frequency of subsistence foods.

Detecting signals in pharmacogenomic genome-wide association studies.

In a common pharmacogenomic scenario, outcome measures are compared for treated and untreated subjects across genotype-defined subgroups. The key question is whether treatment benefit (or harm) is particularly strong in certain subgroups, and therefore the statistical analysis focuses on the interaction between treatment and genotype. However, genome-wide analysis in such scenarios requires careful statistical thought as, in addition to the usual problems of multiple testing, the marker-defined sample sizes, and therefore power, vary across the individual genotypes being evaluated. The variability in power means that the usual practice of using a common P-value threshold across tests has difficulties. The reason is that the use of a fixed threshold, with variable power, implies that the costs of type I and type II errors vary across tests in a manner that is implicit rather than dictated by the analyst. In this paper we discuss this problem and describe an easily implementable solution based on Bayes factors. We pay particular attention to the specification of priors, which is not a straightforward task. The methods are illustrated using data from a randomized controlled clinical trial in which homocysteine levels are compared in individuals receiving low and high doses of folate supplements and across marker subgroups. The method we describe is implemented in the R computing environment with code available from http://faculty.washington.edu/jonno/cv.html.

Non-parametric estimation of a time-dependent predictive accuracy curve.

A major biomedical goal associated with evaluating a candidate biomarker or developing a predictive model score for event-time outcomes is to accurately distinguish between incident cases from the controls surviving beyond t throughout the entire study period. Extensions of standard binary classification measures like time-dependent sensitivity, specificity, and receiver operating characteristic (ROC) curves have been developed in this context (Heagerty, P. J., and others, 2000. Time-dependent ROC curves for censored survival data and a diagnostic marker. Biometrics 56, 337-344). We propose a direct, non-parametric method to estimate the time-dependent Area under the curve (AUC) which we refer to as the weighted mean rank (WMR) estimator. The proposed estimator performs well relative to the semi-parametric AUC curve estimator of Heagerty and Zheng (2005. Survival model predictive accuracy and ROC curves. Biometrics 61, 92-105). We establish the asymptotic properties of the proposed estimator and show that the accuracy of markers can be compared very simply using the difference in the WMR statistics. Estimators of pointwise standard errors are provided.

Time-dependent predictive accuracy in the presence of competing risks.

Competing risks arise naturally in time-to-event studies. In this article, we propose time-dependent accuracy measures for a marker when we have censored survival times and competing risks. Time-dependent versions of sensitivity or true positive (TP) fraction naturally correspond to consideration of either cumulative (or prevalent) cases that accrue over a fixed time period, or alternatively to incident cases that are observed among event-free subjects at any select time. Time-dependent (dynamic) specificity (1-false positive (FP)) can be based on the marker distribution among event-free subjects. We extend these definitions to incorporate cause of failure for competing risks outcomes. The proposed estimation for cause-specific cumulative TP/dynamic FP is based on the nearest neighbor estimation of bivariate distribution function of the marker and the event time. On the other hand, incident TP/dynamic FP can be estimated using a possibly nonproportional hazards Cox model for the cause-specific hazards and riskset reweighting of the marker distribution. The proposed methods extend the time-dependent predictive accuracy measures of Heagerty, Lumley, and Pepe (2000, Biometrics 56, 337-344) and Heagerty and Zheng (2005, Biometrics 61, 92-105).

Baseline pain and disability in the Investigational Vertebroplasty Efficacy and Safety Trial.

BACKGROUND AND PURPOSE: Multiple case series of vertebroplasty outcomes have been published, though no large, placebo controlled trial has yet been performed. Our aim was to report baseline characteristics for the Investigational Vertebroplasty Efficacy and Safety Trial (INVEST), a randomized blinded controlled study of vertebroplasty. MATERIALS AND METHODS: We compared baseline demographics, pain scores, and scores on the modified Roland-Morris Disability Scale (RMDS), a back pain-specific metric, between 2 groups. One group included subjects enrolled at the lead INVEST site (n = 27 to date). The second group consisted of eligible patients seen concurrently at the lead INVEST site, who declined enrollment (n = 70). Comparisons were made by using 2-sample t tests. RESULTS: Mean ages were similar between groups, averaging approximately 74 years among study participants and 77 years among nonenrolled eligible patients (P = .17). Approximately 75% of subjects were female in both groups. RMDS scores of enrolled patients at the lead site (18.0 +/- 4.2) were not statistically different from those of eligible nonenrolled patients at the lead site (18.6 +/- 3.6, P = .49). Pain scores in the enrolled subjects were measured as "average intensity over the prior 24 hours" with mean scores of 7.6 +/- 2.1 among enrolled patients at the lead site. Pain scores in eligible nonenrolled patients were measured as "pain at rest," with mean score of 3.4 +/- 3.3, and "pain with activity," with mean score of 8.5 +/- 2.0. CONCLUSIONS: Patient demographics among subjects enrolled in the INVEST are similar to those in a cohort of eligible nonenrolled patients. Back pain-specific disability was similar between subjects enrolled in the INVEST study and eligible nonenrolled patients at the lead site.

Dust and chemical exposures, and miscarriage risk among women textile workers in Shanghai, China.

INTRODUCTION: To investigate possible associations between miscarriage and occupational exposures in the Shanghai textile industry. METHODS: A retrospective cohort study of miscarriages among 1752 women in the Shanghai textile industry was conducted. Reproductive history was self-reported by women and occupational work histories were collected from factory personnel records. Occupational exposures were assigned by linking work history information to an industry-specific job-exposure matrix informed by factory-specific textile process information and industrial hygiene assessments. Estimates of cotton dust and endotoxin exposure were also assigned. Odds ratios (OR) and 95% CI were estimated by multivariate logistic regression, with adjustment for age at pregnancy, educational level, smoking status of the woman and her spouse, use of alcohol, and woman's year of birth. RESULTS: An elevation in risk of a spontaneously aborted first pregnancy was associated with exposure to synthetic fibres (OR 1.89, 95% CI 1.20 to 3.00) and mixed synthetic and natural fibres (OR 3.31, 95% CI 1.30 to 8.42). No increased risks were observed for women working with solvents, nor were significant associations observed with quantitative cotton dust or endotoxin exposures. Associations were robust and similar when all pregnancies in a woman's reproductive history were considered. CONCLUSIONS: Occupational exposure to synthetic fibres may cause miscarriages, and this possibility should be the subject of further investigation.

An evaluation of state firearm regulations and homicide and suicide death rates.

OBJECTIVE: To determine if any of five different state gun laws were associated with firearm mortality: (1) "shall issue" laws permitting an individual to carry a concealed weapon unless restricted by another statute; (2) a minimum age of 21 years for handgun purchase; (3) a minimum age of 21 years for private handgun possession; (4) one gun a month laws which restrict handgun purchase frequency; and (5) junk gun laws which ban the sale of certain cheaply constructed handguns. DESIGN: A cross sectional time series study of firearm mortality from 1979 to 1998. SETTING: All 50 states and the District of Columbia. SUBJECTS: All residents of the United States. MAIN OUTCOME MEASURES: Firearm homicides, all homicides, firearm suicides, and all suicides. RESULTS: When a "shall issue" law was present, the rate of firearm homicides was greater, RR 1.11 (95% confidence interval 0.99 to 1.24), than when the law was not present, as was the rate of all homicides, RR 1.08 (95% CI 0.98 to 1.17), although this was not statistically significant. No law was associated with a statistically significant decrease in the rates of firearm homicides or total homicides. No law was associated with a statistically significant change in firearm suicide rates. CONCLUSION: A "shall issue" law that eliminates most restrictions on carrying a concealed weapon may be associated with increased firearm homicide rates. No law was associated with a statistically significant reduction in firearm homicide or suicide rates.

MR nerve imaging in a prospective cohort of patients with suspected carpal tunnel syndrome.

OBJECTIVES: To evaluate the reliability and diagnostic accuracy of high-resolution MRI of the median nerve in a prospectively assembled cohort of subjects with clinically suspected carpal tunnel syndrome (CTS). METHODS: The authors prospectively identified 120 subjects with clinically suspected CTS from five Seattle-area clinics. All subjects completed a hand-pain diagram and underwent a standardized nerve conduction study (NCS). The reference standard for determining CTS status was a classic or probable hand pain diagram and NCS with a difference >0.3 ms between the 8-cm median and ulnar peak latencies. Readers graded multiple imaging parameters of the MRI on four-point scales. The authors also performed quantitative measurements of both the median nerve and carpal tunnel cross-sectional areas. NCS and MRI were interpreted without knowledge of the other study or the hand pain diagram. RESULTS: Intrareader reliability was substantial to near perfect (kappa = 0.76 to 0.88). Interreader agreement was lower but still substantial (kappa = 0.60 to 0.67). Sensitivity of MRI was greatest for the overall impression of the images (96%) followed by increased median nerve signal (91%); however, specificities were low (33 to 38%). The length of abnormal signal on T2-weighted images was significantly correlated with nerve conduction latency, and median nerve area was larger at the distal radioulnar joint (15.8 vs 11.8 mm(2)) in patients with CTS. A logistic regression model combining these two MR variables had a receiver operating characteristic area under the curve of 0.85. CONCLUSIONS: The reliability of MRI is high but the diagnostic accuracy is only moderate compared with a research-definition reference standard.

Common bile duct injury during laparoscopic cholecystectomy and the use of intraoperative cholangiography: adverse outcome or preventable error?

BACKGROUND: Common bile duct (CBD) injury is a serious complication of laparoscopic cholecystectomy (LC). Predictors of this adverse outcome have not been well documented. HYPOTHESIS: Surgeon experience and the use of intraoperative cholangiography (IOC) are associated with a decreased rate of major CBD injury during LC. DESIGN: A retrospective population-based cohort study. SETTING: Washington State hospital discharge database reports from 1991 through 1998. PATIENTS: Discharge reports were searched for International Classification of Diseases, Ninth Revision, procedure codes consistent with LC and then evaluated for procedure codes for CBD repair and reconstruction within 90 days of LC. MAIN OUTCOME MEASURE: The rate of CBD injury in patients undergoing LC based on the surgeon's experience and IOC use. RESULTS: In all, 30 630 LCs and 76 major CBD injuries (2.5/1000 operations) were identified in this analysis. There were no significant differences between injured and noninjured patients in demographics, disease, payer status, or hospital variables. A CBD injury occurred in 3.2 of 1000 LCs in the early case order of surgeons compared with 1.7 per 1000 at later points (P = .01) (relative risk, 1.81; 95% confidence interval, 1.44-2.88). The rate of injury in LCs performed without IOC was 3.3 per 1000 compared with 2.0 per 1000 in LCs with IOC (P = .02) (relative risk, 1.7; 95% confidence interval, 1.1-2.6). Surgeon's experience and IOC use were independent predictors of injury. CONCLUSIONS: The rate of CBD injury is significantly lower when IOC is used. This effect is magnified during the early experience of surgeons. Systematic use of IOC may significantly reduce the rate of CBD injury.

Pharmacologic stress testing for coronary disease diagnosis: A meta-analysis.

BACKGROUND: Although noninvasive pharmacologic stress tests are widely used, their relative performance is not clear. We compared the performance of pharmacologic stress tests combined with echocardiography or nuclear imaging for the diagnosis of coronary disease. METHODS: We performed a regression meta-analysis of published data. We included studies published between January 1975 and June 1999 in which subjects underwent echocardiographic or single-photon emission computed tomography (SPECT) stress testing with adenosine, dipyridamole, or dobutamine for diagnosis of coronary artery disease. All subjects also underwent coronary angiography. Two independent reviewers abstracted population characteristics, technical factors, methodologic factors, and results and calculated test sensitivity and specificity. RESULTS: Eighty-two studies met the inclusion criteria. The sensitivity of dipyridamole SPECT imaging, 89% (95% CI, 84%-93%), was higher than that of dipyridamole echocardiography, but the specificity of dipyridamole SPECT imaging, 65% (95% CI, 54%-74%), was lower than that of dipyridamole echocardiography. Dipyridamole and adenosine tests had similar sensitivities and specificities. The sensitivity of dobutamine echocardiography, 80% (95% CI, 77%-83%) was similar to that of dobutamine SPECT imaging, but dobutamine echocardiography had a higher specificity, 84% (95% CI, 80%-86%) than dobutamine SPECT imaging did. CONCLUSIONS: The findings of our study can be used to guide the selection of the optimal pharmacologic stress test for each patient. Maximum sensitivity can be attained by use of a vasodilator combined with SPECT imaging. Maximum specificity can be attained by use of a vasodilator with echocardiography. The highest combination of sensitivity and specificity can be attained with dobutamine echocardiography.

Can evidence change the rate of back surgery? A randomized trial of community-based education.

CONTEXT: Timely adoption of clinical practice guidelines is more likely to happen when the guidelines are used in combination with adjuvant educational strategies that address social as well as rational influences. OBJECTIVE: To implement the conservative, evidence-based approach to low-back pain recommended in national guidelines, with the anticipated effect of reducing population-based rates of surgery. DESIGN: A randomized, controlled trial. SETTING: Ten communities in western Washington State with annual rates of back surgery above the 1990 national average (158 operations per 100,000 adults). PARTICIPANTS: Spine surgeons, primary care physicians, patients who were surgical candidates, and hospital administrators. INTERVENTION: The five communities randomized to the intervention group received a package of six educational activities tailored to local needs by community planning groups. Surgeon study groups, primary care continuing medical education conferences, administrative consensus processes, videodisc-aided patient decision making, surgical outcomes management, and generalist academic detailing were serially implemented over a 30-month intervention period. OUTCOME MEASURE: Quarterly observations of surgical rates. RESULTS: After implementation of the intervention, surgery rates declined in the intervention communities but increased slightly in the control communities. The net effect of the intervention is estimated to be a decline of 20.9 operations per 100,000, a relative reduction of 8.9% (P = 0.01). CONCLUSION: We were able to use scientific evidence to engender voluntary change in back pain practice patterns across entire communities.

The Longitudinal Assessment of Imaging and Disability of the Back (LAIDBack) Study: baseline data.

STUDY DESIGN: Prospective cohort study of randomly selected Veterans Affairs (VA) outpatients. OBJECTIVE: To determine the prevalence of magnetic resonance imaging (MRI) findings in the lumbar spine among persons without current low back pain or sciatica and to examine which findings are related to age or previous back symptoms. SUMMARY OF BACKGROUND INFORMATION: Previous studies of patients without low back pain have not explored the possible association of various MRI findings to past symptoms. METHODS: We randomly selected an age-stratified sample of subjects without low back pain in the past 4 months from clinics at a VA hospital. We collected information on demographics, comorbidity, functional status, and quality of life. MR images were obtained using a standardized protocol through each of the five lumbar disc levels. RESULTS: Of 148 subjects, 69 (46%) had never experienced low back pain. There were 123 subjects (83%) with moderate to severe desiccation of one or more discs, 95 (64%) with one or more bulging discs, and 83 (56%) with loss of disc height. Forty-eight subjects (32%) had at least one disc protrusion and 9 (6%) had one or more disc extrusions. CONCLUSION: Many MR imaging findings have a high prevalence in subjects without low back pain. These findings are therefore of limited diagnostic use. The less common findings of moderate or severe central stenosis, root compression, and extrusions are likely to be diagnostically and clinically relevant.

Readiness for HIV vaccine trials: changes in willingness and knowledge among high-risk populations in the HIV network for prevention trials. The HIVNET Vaccine Preparedness Study Protocol Team.

Longitudinal data were analyzed to determine changes in willingness to participate in HIV vaccine efficacy trials and knowledge of vaccine trial concepts among populations at high risk of HIV-1 infection. Gay men (MSM), male and female injection drug users, and non-injecting women at heterosexual risk were recruited (n = 4892). Follow-up visits occurred every 6 months up to 18 months. Willingness was significantly lower at follow-up visits compared with at baseline. Knowledge levels increased for all study populations. Problematic concepts were possible effects of the vaccine on the immune system and lack of knowledge about efficacy of a vaccine before the start of a trial. For concepts concerning safety, blinding, and guarantees of future participation in trials, MSM men had significant increases in knowledge, but little to no change occurred for the other populations. An increase in knowledge was associated with becoming not willing, particularly among MSM with low knowledge levels. At least half of high-risk participants were consistently willing to participate in future vaccine efficacy trials and with basic vaccine education, knowledge levels increased. Continued educational efforts at the community and individual level are needed to address certain vaccine trial concepts and to increase knowledge levels in all potential study populations.

Multivariate continuation ratio models: connections and caveats.

We develop semiparametric estimation methods for a pair of regressions that characterize the first and second moments of clustered discrete survival times. In the first regression, we represent discrete survival times through univariate continuation indicators whose expectations are modeled using a generalized linear model. In the second regression, we model the marginal pairwise association of survival times using the Clayton-Oakes cross-product ratio (Clayton, 1978, Biometrika 65, 141-151; Oakes, 1989, Journal of the American Statistical Association 84, 487-493). These models have recently been proposed by Shih (1998, Biometrics 54, 1115-1128). We relate the discrete survival models to multivariate multinomial models presented in Heagerty and Zeger (1996, Journal of the American Statistical Society 91, 1024-1036) and derive a paired estimating equations procedure that is computationally feasible for moderate and large clusters. We extend the work of Guo and Lin (1994, Biometrics 50, 632-639) and Shih (1998) to allow covariance weighted estimating equations and investigate the impact of weighting in terms of asymptotic relative efficiency. We demonstrate that the multinomial structure must be acknowledged when adopting weighted estimating equations and show that a naive use of GEE methods can lead to inconsistent parameter estimates. Finally, we illustrate the proposed methodology by analyzing psychological testing data previously summarized by TenHave and Uttal (1994, Applied Statistics 43, 371-384) and Guo and Lin (1994).

Time-dependent ROC curves for censored survival data and a diagnostic marker.

ROC curves are a popular method for displaying sensitivity and specificity of a continuous diagnostic marker, X, for a binary disease variable, D. However, many disease outcomes are time dependent, D(t), and ROC curves that vary as a function of time may be more appropriate. A common example of a time-dependent variable is vital status, where D(t) = 1 if a patient has died prior to time t and zero otherwise. We propose summarizing the discrimination potential of a marker X, measured at baseline (t = 0), by calculating ROC curves for cumulative disease or death incidence by time t, which we denote as ROC(t). A typical complexity with survival data is that observations may be censored. Two ROC curve estimators are proposed that can accommodate censored data. A simple estimator is based on using the Kaplan-Meier estimator for each possible subset X > c. However, this estimator does not guarantee the necessary condition that sensitivity and specificity are monotone in X. An alternative estimator that does guarantee monotonicity is based on a nearest neighbor estimator for the bivariate distribution function of (X, T), where T represents survival time (Akritas, M. J., 1994, Annals of Statistics 22, 1299-1327). We present an example where ROC(t) is used to compare a standard and a modified flow cytometry measurement for predicting survival after detection of breast cancer and an example where the ROC(t) curve displays the impact of modifying eligibility criteria for sample size and power in HIV prevention trials.