Biomedical subjects

Patrick J Heagerty

Publications and source records attributed to Patrick J Heagerty.

At least 37 records · Page 2Linked to original sources

PubMed2005

Directly parameterized regression conditioning on being alive: analysis of longitudinal data truncated by deaths.

For observational longitudinal studies of geriatric populations, outcomes such as disability or cognitive functioning are often censored by death. Statistical analysis of such data may explicitly condition on either vital status or survival time when summarizing the longitudinal response. For example a pattern-mixture model characterizes the mean response at time t conditional on death at time S = s (for s > t), and thus uses future status as a predictor for the time t response. As an alternative, we define regression conditioning on being alive as a regression model that conditions on survival status, rather than a specific survival time. Such models may be referred to as partly conditional since the mean at time t is specified conditional on being alive (S > t), rather than using finer stratification (S = s for s > t). We show that naive use of standard likelihood-based longitudinal methods and generalized estimating equations with non-independence weights may lead to biased estimation of the partly conditional mean model. We develop a taxonomy for accommodation of both dropout and death, and describe estimation for binary longitudinal data that applies selection weights to estimating equations with independence working correlation. Simulation studies and an analysis of monthly disability status illustrate potential bias in regression methods that do not explicitly condition on survival.

Activities of Daily Living↗

PubMed2005

Survival model predictive accuracy and ROC curves.

The predictive accuracy of a survival model can be summarized using extensions of the proportion of variation explained by the model, or R2, commonly used for continuous response models, or using extensions of sensitivity and specificity, which are commonly used for binary response models. In this article we propose new time-dependent accuracy summaries based on time-specific versions of sensitivity and specificity calculated over risk sets. We connect the accuracy summaries to a previously proposed global concordance measure, which is a variant of Kendall's tau. In addition, we show how standard Cox regression output can be used to obtain estimates of time-dependent sensitivity and specificity, and time-dependent receiver operating characteristic (ROC) curves. Semiparametric estimation methods appropriate for both proportional and nonproportional hazards data are introduced, evaluated in simulations, and illustrated using two familiar survival data sets.

Biometry↗

PubMed2005

Partly conditional survival models for longitudinal data.

It is common in longitudinal studies to collect information on the time until a key clinical event, such as death, and to measure markers of patient health at multiple follow-up times. One approach to the joint analysis of survival and repeated measures data adopts a time-varying covariate regression model for the event time hazard. Using this standard approach, the instantaneous risk of death at time t is specified as a possibly semi-parametric function of covariate information that has accrued through time t. In this manuscript, we decouple the time scale for modeling the hazard from the time scale for accrual of available longitudinal covariate information. Specifically, we propose a class of models that condition on the covariate information through time s and then specifies the conditional hazard for times t, where t > s. Our approach parallels the "partly conditional" models proposed by Pepe and Couper (1997, Journal of the American Statistical Association 92, 991-998) for pure repeated measures applications. Estimation is based on the use of estimating equations applied to clusters of data formed through the creation of derived survival times that measure the time from measurement of covariates to the end of follow-up. Patient follow-up may be terminated either by the occurrence of the event or by censoring. The proposed methods allow a flexible characterization of the association between a longitudinal covariate process and a survival time, and facilitate the direct prediction of survival probabilities in the time-varying covariate setting.

Acquired Immunodeficiency Syndrome↗

PubMed2005

Medicaid acceptance and availability of timely follow-up for newborns with Medicaid.

OBJECTIVE: Decreased physician participation in Medicaid has been shown to affect adversely timeliness of adult acute care and pediatric specialty care, but it is not clear whether this is the case for newborn follow-up. The objectives of this study were to determine whether there is a difference within clinics in the timeliness of follow-up appointments that are given to newborns with Medicaid compared with newborns with private insurance and to determine whether there is a difference between clinics that do and do not accept Medicaid in the timeliness of appointments that are given for newborn follow-up. METHODS: A randomized crossover study was conducted among general pediatric clinics and practices that were identified from the yellow pages and Internet searches of hospitals and health departments in 8 metropolitan areas from September 2003 to March 2004. A simulated parent telephoned clinics to find the earliest available appointment for a 1-day-old infant who needed routine follow-up after discharge that day. Clinics were randomly assigned to receive a first call from a patient with either Medicaid or private insurance; each clinic received the same call at least 3 weeks later with the patient's insurance status reversed. The main outcome measure was whether the appointment was timely (< or =2 days from the day of the call). RESULTS: Of 401 participating clinics, 22% did not accept Medicaid. Among clinics that accepted Medicaid, availability of a timely appointment for a newborn with Medicaid was similar to that for a newborn with private insurance (87% vs 90%, respectively). Appointments that were provided to privately insured newborns were as likely to be timely in clinics that accept Medicaid as in clinics that do not accept Medicaid (89.5% vs 93.4%, respectively). However, providing timely appointments was significantly less likely in clinics that were in high-poverty locations compared with clinics that were not (86.1% vs 92.7%, respectively). CONCLUSIONS: Although newborns with Medicaid did not have access to >20% of clinics because of their insurance, among clinics that did accept Medicaid, timeliness of available follow-up was similar for newborns with Medicaid compared with newborns with private insurance and similar between clinics that did and did not accept Medicaid. However, to the extent that care for newborns with Medicaid is concentrated in clinics in high-poverty areas, some newborns with Medicaid may not be able to receive timely appointments.

Community Health Services↗

PubMed2004

Marginalized transition models for longitudinal binary data with ignorable and non-ignorable drop-out.

We extend the marginalized transition model of Heagerty to accommodate non-ignorable monotone drop-out. Using a selection model, weakly identified drop-out parameters are held constant and their effects evaluated through sensitivity analysis. For data missing at random (MAR), efficiency of inverse probability of censoring weighted generalized estimating equations (IPCW-GEE) is as low as 40 per cent compared to a likelihood-based marginalized transition model (MTM) with comparable modelling burden. MTM and IPCW-GEE regression parameters both display misspecification bias for MAR and non-ignorable missing data, and both reduce bias noticeably by improving model fit.

Antipsychotic Agents↗

PubMed2004

The association between pain and disability.

A clearer understanding of how pain intensity relates to disability could have important implications for pain treatment goals and definitions of treatment success. The objectives of this study were to determine the optimal pain intensity rating (0-10 scale) cutpoints for discriminating disability levels among individuals with work-related carpal tunnel syndrome (CTS) and low back (LB) injuries, whether these cutpoints differed for these conditions and for different disability measures, and whether the relationship between pain intensity and disability was linear in each injury group. Approximately 3 weeks after filing work injury claims, 2183 workers (1059 CTS; 1124 LB) who still had pain completed pain and disability measures. In the LB group, pain intensity rating categories of 1-4, 5-6, and 7-10 optimally discriminated disability levels for all four disability measures examined. In the CTS group, no pain intensity rating categorization scheme proved superior across all disability measures. For all disability measures examined, the relationship between pain intensity and disability level was linear in the CTS group, but nonlinear in the LB group. Among study participants with work-related back injuries, when pain level was 1-4, a decrease in pain of more than 1-point corresponded to clinically meaningful improvement in functioning, but when pain was rated as 5-10, a 2-point decrease was necessary for clinically meaningful improvement in functioning. The findings indicate that classifying numerical pain ratings into categories corresponding to levels of disability may be useful in establishing treatment goals, but that classification schemes must be validated separately for different pain conditions.

Adult↗

PubMed2004

The effect of alcohol consumption on the prevalence of iron overload, iron deficiency, and iron deficiency anemia.

BACKGROUND & AIMS: Our aim was to investigate the relationship between alcohol consumption and iron overload, iron deficiency, or iron deficiency anemia in the U.S. population. METHODS: Adult participants of the Third National Health and Nutrition Examination Survey who did not consume alcohol (n = 8839) were compared with participants who consumed < or =1 (n = 4976), >1 to < or =2 (n = 1153), or >2 (n = 915) alcoholic drinks/day during the preceding 12 months. We examined the following markers of iron overload: elevated serum transferrin-iron saturation (TS) level (>45%, >50%, and >60%), elevated serum ferritin level (>300, >400, >500, and >600 ng/mL), and combinations of both elevated serum TS and ferritin levels. Iron deficiency was defined as the presence of at least 2 of the following: serum ferritin level <12 ng/mL, serum TS level <15%, and erythrocyte protoporphyrin level >1.24 micromol/L. Iron deficiency anemia was defined as the presence of both iron deficiency and anemia. RESULTS: Compared with nondrinkers, the prevalence of all markers of iron overload was significantly elevated among those who consumed >2 alcoholic drinks/day after adjusting for potential confounders. Consumption of any amount of alcohol was associated with a 40% reduction in the risk of iron deficiency anemia. CONCLUSIONS: Consumption of up to 2 alcoholic drinks/day seems to be associated with reduced risk of iron deficiency and iron deficiency anemia without a concomitant increase in the risk of iron overload. Consumption of >2 alcoholic drinks/day is associated with a significant elevation in the risk of iron overload.

Adolescent↗

PubMed2004

Marginal modeling of multilevel binary data with time-varying covariates.

We propose and compare two approaches for regression analysis of multilevel binary data when clusters are not necessarily nested: a GEE method that relies on a working independence assumption coupled with a three-step method for obtaining empirical standard errors, and a likelihood-based method implemented using Bayesian computational techniques. Implications of time-varying endogenous covariates are addressed. The methods are illustrated using data from the Breast Cancer Surveillance Consortium to estimate mammography accuracy from a repeatedly screened population.

Adult↗

PubMed2004

Semiparametric estimation of time-dependent ROC curves for longitudinal marker data.

One approach to evaluating the strength of association between a longitudinal marker process and a key clinical event time is through predictive regression methods such as a time-dependent covariate hazard model. For example, a Cox model with time-varying covariates specifies the instantaneous risk of the event as a function of the time-varying marker and additional covariates. In this manuscript we explore a second complementary approach which characterizes the distribution of the marker as a function of both the measurement time and the ultimate event time. Our goal is to extend the standard diagnostic accuracy concepts of sensitivity and specificity so as to recognize explicitly both the timing of the marker measurement and the timing of disease. The accuracy of a longitudinal marker can be fully characterized using time-dependent receiver operating characteristic (ROC) curves. We detail a semiparametric estimation method for time-dependent ROC curves that adopts a regression quantile approach for longitudinal data introduced by Heagerty and Pepe (1999, Applied Statistics, 48, 533-551). We extend the work of Heagerty and Pepe (1999, Applied Statistics, 48, 533-551) by developing asymptotic distribution theory for the ROC estimators where the distributional shape for the marker is allowed to depend on covariates. To illustrate our method, we analyze pulmonary function measurements among cystic fibrosis subjects and estimate ROC curves that assess how well the pulmonary function measurement can distinguish subjects that progress to death from subjects that remain alive. Comparing the results from our semiparametric analysis to a fully parametric method discussed by Etzioni et al. (1999, Medical Decision Making, 19, 242-251) suggests that the ability to relax distributional assumptions may be important in practice.

Adolescent↗

PubMed2004

A comparison between audio computer-assisted self-interviews and clinician interviews for obtaining the sexual history.

OBJECTIVE: The objective of this study was to compare reporting between audio computer-assisted self-interview (ACASI) and clinician-administered sexual histories. GOAL: The goal of this study was to explore the usefulness of ACASI in sexually transmitted disease (STD) clinics. STUDY: The authors conducted a cross-sectional study of ACASI followed by a clinician history (CH) among 609 patients (52% male, 59% white) in an urban, public STD clinic. We assessed completeness of data, item prevalence, and report concordance for sexual history and patient characteristic variables classified as socially neutral (n=5), sensitive (n=11), or rewarded (n=4). RESULTS: Women more often reported by ACASI than during CH same-sex behavior (19.6% vs. 11.5%), oral sex (67.3% vs. 50.0%), transactional sex (20.7% vs. 9.8%), and amphetamine use (4.9% vs. 0.7%) but were less likely to report STD symptoms (55.4% vs. 63.7%; all McNemar chi-squared P values <0.003). Men's reporting was similar between interviews, except for ever having had sex with another man (36.9% ACASI vs. 28.7% CH, P <0.001). Reporting agreement as measured by kappas and intraclass correlation coefficients was only moderate for socially sensitive and rewarded variables but was substantial or almost perfect for socially neutral variables. ACASI data tended to be more complete. ACASI was acceptable to 89% of participants. CONCLUSIONS: ACASI sexual histories may help to identify persons at risk for STDs.

Adolescent↗

PubMed2003

Paraoxonase activity, but not haplotype utilizing the linkage disequilibrium structure, predicts vascular disease.

OBJECTIVE: The effects of paraoxonase (PON1) activity and of genetic variation in the PON1 promoter and coding region on carotid artery disease (CAAD) were investigated. METHODS AND RESULTS: We identified functional promoter polymorphisms and examined their effects in a cohort with and without CAAD. We used the full sequences in 23 white subjects to determine the linkage disequilibrium (LD) structure of the PON1 region and to direct the grouping of haplotypes for disease association testing. There are several discrete regions of the PON1 gene with strong local LD, but the useful levels of LD do not extend across the entire gene. Indeed, PON1-162/-108/55/192 haplotype did not predict additional variation in PON1 activities compared with the 4 genotypes separately. PON1 hydrolysis activity predicted CAAD status, but this was not attributable to the promoter or coding region polymorphisms or haplotype or to the effects of smoking or statin use on PON1 activity. CONCLUSIONS: PON1 does not have LD across the gene, and use of haplotypes in association studies should consider the LD structure. PON1 activity predicts CAAD, yet 4 functional polymorphisms do not. Additional investigations of genetic and environmental factors that influence PON1 activity as a risk factor for vascular disease are warranted.

Adult↗

PubMed2003

Motor cortex organization after stroke is related to side of stroke and level of recovery.

BACKGROUND AND PURPOSE: The present study hypothesized that side of stroke and level of recovery influence motor system organization after stroke. METHODS: Functional MRI was performed on 14 control subjects and 21 patients with chronic stroke during index finger tapping (control subjects, right; patients, recovered side). RESULTS: On functional MRI, stroke patients with right arm involvement showed (1) significantly smaller activation in contralateral motor cortexes compared with control subjects; (2) smaller ipsilateral (nonstroke) premotor and larger contralateral (stroke-side) sensorimotor activation compared with patients with left arm involvement, although electromyogram across groups was similar; and (3) 2.7-fold-larger contralateral sensorimotor cortex activation, ventrally, in those with full recovery compared with those with partial recovery, despite similar tapping force, frequency, range of motion, and electromyogram between groups. Supplementary motor area activation was unrelated to level of recovery. CONCLUSIONS: After stroke that causes mild to moderate initial impairment and mild residual hand weakness, cortical organization varies with side of injury and with final motor status. The findings may have implications for treatment after stroke.

Arm↗

PubMed2003

Safety and acceptability of the Reality condom for anal sex among men who have sex with men.

OBJECTIVES: To assess safety and acceptability of Reality condoms for anal sex among men who have sex with men. METHODS: Crossover study among HIV-seroconcordant (33 HIV-negative and 5 HIV-positive) monogamous male couples, randomized to latex male and Reality condom use with anal sex. RESULTS: Slippage with removal was reported more frequently with Reality than male latex condoms [odds ratio (OR), 2.7; 95% confidence interval (CI), 1.2-5.8 for receptive partners and OR, 34.1; 95% CI, 13.8-84.1 for insertive partners]. Receptive partners more frequently reported pain or discomfort (OR, 5.0; 95% CI, 2.6-9.4) and rectal bleeding (OR, 1.9; 95% CI, 0.9-4.1) with Reality condoms than male condoms. Over 20% reported willingness to use the Reality condom in the future with a partner of unknown HIV status; willingness was associated with past problems with male condoms and no problems with Reality condoms among receptive partners, and with past use of Reality condoms and HIV seropositivity among insertive partners. CONCLUSIONS: Men reported more frequent problems with Reality condoms than male latex condoms used for anal intercourse, particularly slippage, discomfort, and rectal bleeding. Design modifications, training, and research on the clinical significance of safety outcomes are needed for use of Reality condoms with anal sex.

Adult↗

PubMed2003

Apolipoprotein E genotype and neurodevelopmental sequelae of infant cardiac surgery.

BACKGROUND: There has been increasing recognition of adverse neurodevelopmental sequelae in some children after repair of congenital heart defects. Even among children with the same cardiac defect, significant interindividual variation exists in developmental outcome. Polymorphisms of apolipoprotein E have been identified as a risk factor for worse neurologic recovery after central nervous system injury. METHODS: A single-institution prospective study of patients <or=6 months of age undergoing cardiopulmonary bypass for repair of congenital heart defects was undertaken to evaluate the association between apolipoprotein E genotype and postoperative neurodevelopmental dysfunction. Developmental outcomes were evaluated at 1 year of age by using the Bayley Scales of Infant Development. RESULTS: One-year evaluation was performed in 244 patients. After adjustment for preoperative and postoperative covariates-including gestational age, age at operation, sex, race, socioeconomic status, cardiac defect, and use of deep hypothermic circulatory arrest-the apolipoprotein E epsilon2 allele was associated with a worse neurologic outcome as assessed by the Psychomotor Developmental Index of the Bayley Scales of Infant Development (P =.036). Patients with the apolipoprotein E epsilon2 allele had approximately a 7-point decrease in the Psychomotor Developmental Index. CONCLUSIONS: Apolipoprotein E epsilon2 allele carriers had significantly lower Psychomotor Development Index scores at 1 year of age after infant cardiac surgery. The effect was independent of ethnicity, socioeconomic status, cardiac defect, and use of deep hypothermic circulatory arrest. An effect of the apolipoprotein E epsilon4 allele was not detected. Genetic polymorphisms that decrease neuroresiliency and impair neuronal repair after central nervous system injury are important risk factors for neurodevelopmental dysfunction after infant cardiac surgery.

Alleles↗

PubMed2003

Racial differences in the relationship between hepatitis C infection and iron stores.

Black race and increased hepatic iron stores predict poor response to interferon treatment for chronic hepatitis C virus (HCV) infection. We tested the hypothesis that these 2 observations are linked by investigating whether HCV-infected African-Americans have increased iron stores relative to uninfected persons. Using data from the third National Health and Nutrition Examination Survey (NHANES III), we determined the risk of having increased iron stores, defined as elevation of both serum ferritin and transferrin-iron saturation (TS), in HCV-RNA-positive blacks (n = 100) and nonblacks (n = 126) relative to HCV-RNA-negative blacks (n = 4,002) and nonblacks (n = 10,943). HCV-positive blacks were 5.4 times (95% CI, 1.2 to 24) more likely to have increased iron stores than HCV-positive nonblacks. The proportion of HCV-positive blacks who had increased iron stores was 16.4% among those with abnormal liver enzymes and 2.8% among those with normal liver enzymes, compared with only 0.6% among HCV-negative blacks. After adjustment for age, alcohol intake, gender, menopausal status, education, body mass index, and poverty index, HCV-positive blacks with abnormal liver enzymes had an elevated risk of having increased iron stores (odds ratio, 17.8; 95% CI, 5.1 to 63). In contrast, among persons of other races, there was a much smaller difference in the proportion of persons with increased iron stores between HCV-positive persons with (3.4%) or without (1.4%) abnormal liver enzymes and HCV-negative persons (0.9%). In conclusion, a greater proportion of blacks than persons of other races respond to HCV infection with an increase in iron stores. This finding may partly explain the reduced response of HCV-positive African-Americans to antiviral treatment.

Adult↗

PubMed2003

Daily weight-bearing activity does not increase the risk of diabetic foot ulcers.

PURPOSE: This study had two purposes: to identify characteristics associated with participants' usual weight-bearing activity, and to determine whether weight-bearing activity increased the risk of foot ulcer among persons with diabetes and prior foot ulcer. METHODS: We conducted a prospective cohort study of 400 participants with diabetes and a prior history of foot ulcer. Participants were from the Veterans Affairs Puget Sound Health Care System and Group Health Cooperative in Seattle, WA, and were originally assembled for a randomized controlled trial of footwear. Feet were examined and demographic and health history information was collected at enrollment. Daily weight-bearing activity was reported at enrollment and every 17 wk thereafter for 2 yr. All incident foot lesions were recorded. RESULTS: Weight-bearing activity was significantly higher among women. Activity was lower among participants who were older, unmarried, or who had a history of congestive heart failure, respiratory disease, stroke, or depression. Activity decreased significantly over the study period but remained similar in those with or without insensate feet. After adjustment for foot-related and health-status characteristics, moderately active participants (4.5-7.4 weight-bearing h.d-1) were at substantially but nonsignificantly reduced risk of foot ulcer compared with "least active" participants (<4.5 weight-bearing h.d-1) (OR 0.50, 95% CI 0.22-1.16). "Most active" participants (>or=7.5 weight-bearing h.d-1) were at significantly reduced risk (OR 0.20, 95% CI 0.04-0.87). Weight-bearing activity exerted similar effects on foot ulcer risk in participants with insensate versus sensate feet . CONCLUSION: Increased weight-bearing activity did not increase the risk of foot re-ulceration. Further studies are needed to explore weight-bearing activities that are safe for those with diabetes and insensate feet.

Aged↗

PubMed2002

Vitamin C and E intake is associated with increased paraoxonase activity.

OBJECTIVE: Paraoxonase (PON1), an esterase physically associated with high density lipoprotein, has been shown to inhibit atherogenic low density lipoprotein and high density lipoprotein oxidation. PON1 activity appears to be primarily under genetic control with some environmental modification and is a predictor of vascular disease. Vitamins C and E, dietary antioxidants, scavenge free-oxygen radical products that may depress PON1 activity. Therefore, we evaluated the relationship between dietary vitamin C and E intake and PON1 activity. METHODS AND RESULTS: The vitamin C and E intakes of male white subjects (n=189) were estimated by using a standardized food frequency survey. With covariates, vitamin C or E intakes were found to be significant positive predictors of PON1 activity for the hydrolysis of paraoxon and diazoxon with the use of linear regression. Smoking and use of statins were independent predictors of PON1 activity. CONCLUSIONS: PON1 activity, which is primarily genotype dependent, varies with antioxidant vitamins, cigarette smoking, and statin drug use. Because PON1 activity is a better predictor of vascular disease than is the currently described genetic variation in PON1, further studies of the environmental influences on PON1 activity and additional PON1 genetic variants are warranted.

Aged↗

PubMed2002

Effect of therapeutic footwear on foot reulceration in patients with diabetes: a randomized controlled trial.

CONTEXT: Many people with diabetes experience lower-limb ulcers. Footwear has been implicated as a primary cause of foot ulcers, yet research is limited on the efficacy of shoe and insert combinations to prevent reulceration. OBJECTIVE: To determine whether extra-depth and -width therapeutic shoes used with 2 types of inserts reduce reulceration in diabetic individuals with a history of foot ulcer. DESIGN, SETTING, AND PARTICIPANTS: Randomized clinical trial of 400 diabetes patients with history of foot ulcer in 2 Washington State health care organizations who did not require custom shoes for foot deformity and were enrolled between August 1997 and December 1998 and followed up for 2 years. Data collected at regular intervals documented physical, foot, and diabetes characteristics; footwear use; foot lesions; and ulcers. INTERVENTIONS: Participants were randomly assigned to receive 3 pairs of therapeutic shoes and 3 pairs of customized medium-density cork inserts with a neoprene closed-cell cover (n = 121); to receive 3 pairs of therapeutic shoes and 3 pairs of prefabricated, tapered polyurethane inserts with a brushed nylon cover (n = 119); or to wear their usual footwear (controls; n = 160). MAIN OUTCOME MEASURE: Foot reulceration, compared among the 3 groups. RESULTS: Two-year cumulative reulceration incidence across the 3 groups was low: 15% in the cork-insert group, 14% in the prefabricated-insert group, and 17% in controls. In the intent-to-treat analysis, patients assigned to therapeutic shoes did not have a significantly lower risk of reulceration compared with controls (risk ratio [RR] for the cork-insert group, 0.88; 95% confidence interval [CI], 0.51-1.52 and RR the for prefabricated-insert group, 0.85; 95% CI, 0.48-1.48). All ulcer episodes in patients assigned to therapeutic shoes and 88% wearing nonstudy shoes occurred in patients with foot insensitivity. CONCLUSIONS: This study of persons without severe foot deformity does not provide evidence to support widespread dispensing of therapeutic shoes and inserts to diabetic patients with a history of foot ulcer. Study shoes and custom cork or preformed polyurethane inserts conferred no significant ulcer reduction compared with control footwear. This study suggests that careful attention to foot care by health care professionals may be more important than therapeutic footwear but does not negate the possibility that special footwear is beneficial in persons with diabetes who do not receive such close attention to foot care by their health care providers or in individuals with severe foot deformities.

Aged↗