Biomedical subjects

Bernard Rosner

Publications and source records attributed to Bernard Rosner.

At least 19 recordsLinked to original sources

PubMed2007

Prospective study of vitamins C, E, and A and carotenoids and risk of oral premalignant lesions in men.

Case-control studies indicate that vitamins C, E, A and carotenoids decrease risk of oral premalignant lesions (OPLs) and oral cancer, but clinical trials have failed to find protective effects of beta-carotene and suggest that vitamin E may increase risk. The authors prospectively evaluated the association between intake of vitamins C, E, A and carotenoids and incidence of OPL. Participants were 42,340 men in the Health Professionals Follow-up Study who provided information on supplement use and diet every 2-4 years by food frequency questionnaire. The authors confirmed 207 clinically or histopathologically diagnosed OPL events occurring between 1986 and 2002 by medical record review. Multivariate-adjusted relative risks (RR) of OPL were calculated with proportional hazards models. Total intake of vitamin C, vitamin A or carotenoids was not significantly associated with OPL risk. Dietary vitamin C was significantly associated with reduced risk (quintile 5 vs. 1, RR = 0.52, 95% CI 0.31-0.85, p(trend) = 0.04), but no association with supplemental vitamin C was observed. Inverse associations were apparent for beta-cryptoxanthin and alpha-carotene intake. No clear relationship emerged with beta-carotene, lycopene or lutein/zeaxanthin. Vitamin E was associated with increased risk (quintile 5 vs. 1, RR = 1.86, 95% CI 1.06-3.19), particularly among current smokers and with supplemental intake (current-smokers, supplement dose tertile 3 vs. 1, RR = 3.07, 95% CI 1.28-7.34, p(trend) = 0.01). For current smokers, beta-carotene also increased risk. Vitamin C from dietary sources, but not supplements, was associated with a reduced risk of OPL. The observed increased risk for current smokers with high vitamin E or beta-carotene intake should be explored further.

Adult↗

PubMed2007

Interval estimation for rank correlation coefficients based on the probit transformation with extension to measurement error correction of correlated ranked data.

The Spearman (rho(s)) and Kendall (tau) rank correlation coefficient are routinely used as measures of association between non-normally distributed random variables. However, confidence limits for rho(s) are only available under the assumption of bivariate normality and for tau under the assumption of asymptotic normality of tau. In this paper, we introduce another approach for obtaining confidence limits for rho(s) or tau based on the arcsin transformation of sample probit score correlations. This approach is shown to be applicable for an arbitrary bivariate distribution. The arcsin-based estimators for rho(s) and tau (denoted by rho(s,a), tau(a)) are shown to have asymptotic relative efficiency (ARE) of 9/pi2 compared with the usual estimators rho(s) and tau when rho(s) and tau are, respectively, 0. In some nutritional applications, the Spearman rank correlation between nutrient intake as assessed by a reference instrument versus nutrient intake as assessed by a surrogate instrument is used as a measure of validity of the surrogate instrument. However, if only a single replicate (or a few replicates) are available for the reference instrument, then the estimated Spearman rank correlation will be downwardly biased due to measurement error. In this paper, we use the probit transformation as a tool for specifying an ANOVA-type model for replicate ranked data resulting in a point and interval estimate of a measurement error corrected rank correlation. This extends previous work by Rosner and Willett for obtaining point and interval estimates of measurement error corrected Pearson correlations.

Confidence Intervals↗

PubMed2006

Dietary fat and risk of postmenopausal breast cancer in a 20-year follow-up.

Dietary fat in midlife has not been associated with breast cancer risk in most studies, but few have followed women beyond one decade. The authors examined the relation of dietary fat, assessed by repeated questionnaires, to incidence of postmenopausal breast cancer in a cohort of 80,375 US women (3,537 new cases) prospectively followed for 20 years between 1980 and 2000. The multivariable relative risk for an increment of 5% of energy from total dietary fat intake was 0.98 (95% confidence interval: 0.95, 1.00). Additionally, specific types of fat were not associated with an increased risk of breast cancer. Furthermore, secondary analyses indicated no differences in breast cancer risk by estrogen receptor or progesterone receptor status. However, stratification by waist circumference indicated a significant decrease in breast cancer risk for participants with a waist circumference of 35 inches (88.9 cm) or greater (p-trend = 0.04). None of the latency intervals investigated were associated with an increased risk of breast cancer. In addition, fat intake before menopause was not related to risk of postmenopausal breast cancer. These results suggest a reduction in breast cancer risk for women with insulin resistance syndrome who consume high-fat diets and no association between specific sources of fat during midlife and risk of postmenopausal breast cancer.

Aged↗

PubMed2006

Extended work duration and the risk of self-reported percutaneous injuries in interns.

CONTEXT: In their first year of postgraduate training, interns commonly work shifts that are longer than 24 hours. Extended-duration work shifts are associated with increased risks of automobile crash, particularly during a commute from work. Interns may be at risk for other occupation-related injuries. OBJECTIVE: To assess the relationship between extended work duration and rates of percutaneous injuries in a diverse population of interns in the United States. DESIGN, SETTING, AND PARTICIPANTS: National prospective cohort study of 2737 of the estimated 18,447 interns in US postgraduate residency programs from July 2002 through May 2003. Each month, comprehensive Web-based surveys that asked about work schedules and the occurrence of percutaneous injuries in the previous month were sent to all participants. Case-crossover within-subjects analyses were performed. MAIN OUTCOME MEASURES: Comparisons of rates of percutaneous injuries during day work (6:30 am to 5:30 pm) after working overnight (extended work) vs day work that was not preceded by working overnight (nonextended work). We also compared injuries during the nighttime (11:30 pm to 7:30 am) vs the daytime (7:30 am to 3:30 pm). RESULTS: From a total of 17,003 monthly surveys, 498 percutaneous injuries were reported (0.029/intern-month). In 448 injuries, at least 1 contributing factor was reported. Lapse in concentration and fatigue were the 2 most commonly reported contributing factors (64% and 31% of injuries, respectively). Percutaneous injuries were more frequent during extended work compared with nonextended work (1.31/1000 opportunities vs 0.76/1000 opportunities, respectively; odds ratio [OR], 1.61; 95% confidence interval [CI], 1.46-1.78). Extended work injuries occurred after a mean of 29.1 consecutive work hours; nonextended work injuries occurred after a mean of 6.1 consecutive work hours. Injuries were more frequent during the nighttime than during the daytime (1.48/1000 opportunities vs 0.70/1000 opportunities, respectively; OR, 2.04; 95% CI, 1.98-2.11). CONCLUSION: Extended work duration and night work were associated with an increased risk of percutaneous injuries in this study population of physicians during their first year of clinical training.

Accidents, Occupational↗

PubMed2006

Statistical inference for correlated data in ophthalmologic studies.

In ophthalmologic studies, each subject usually contributes important information for each of two eyes and the values from the two eyes are generally highly correlated. Previous studies showed that test procedures for binary paired data that ignore the presence of intraclass correlation could lead to inflated significance levels. Furthermore, it is possible that asymptotic versions of these procedures that take the intraclass correlation into account could also produce unacceptably high type I error rates when the sample size is small or the data structure is sparse. We propose two alternatives for these situations, namely the exact unconditional and approximate unconditional procedures. According to our simulation results, the exact procedures usually produce extremely conservative empirical type I error rates. That is, the corresponding type I error rates could greatly underestimate the pre-assigned nominal level (e.g. (empirical type I error rate/nominal type I error rate) 0.8). On the other hand, the approximate unconditional procedures usually yield empirical type I error rates close to the pre-chosen nominal level. We illustrate our methodologies with a data set from a retinal detachment study.

Biometry↗

PubMed2006

Combined estrogen and testosterone use and risk of breast cancer in postmenopausal women.

BACKGROUND: The role of androgens in breast cancer etiology has been unclear. Epidemiologic studies suggest that endogenous testosterone levels are positively associated with breast cancer risk in postmenopausal women. Given the increasing trend in the use of hormone therapies containing androgens, we evaluated the relation between the use of estrogen and testosterone therapies and breast cancer. METHODS: We conducted a prospective cohort study in the Nurses' Health Study from 1978 to 2002 to assess the risk of breast cancer associated with different types of postmenopausal hormone (PMH) formulations containing testosterone. During 24 years of follow-up (1 359 323 person-years), 4610 incident cases of invasive breast cancer were identified among postmenopausal women. Information on menopausal status, PMH use, and breast cancer diagnosis was updated every 2 years through questionnaires. RESULTS: Among women with a natural menopause, the risk of breast cancer was nearly 2.5-fold greater among current users of estrogen plus testosterone therapies (multivariate relative risk, 2.48; 95% confidence interval, 1.53-4.04) than among never users of PMHs. This analysis showed that risk of breast cancer associated with current use of estrogen and testosterone therapy was significantly greater compared with estrogen-only therapy (P for heterogeneity, .007) and marginally greater than estrogen and progesterone therapy (P for heterogeneity, .11). Women receiving PMHs with testosterone had a 17.2% (95% confidence interval, 6.7%-28.7%) increased risk of breast cancer per year of use. CONCLUSION: Consistent with the elevation in risk for endogenous testosterone levels, women using estrogen and testosterone therapies have a significantly increased risk of invasive breast cancer.

Adult↗

PubMed2006

Prospective study of fruits and vegetables and risk of oral premalignant lesions in men.

The authors prospectively evaluated fruit and vegetable consumption and the incidence of oral premalignant lesions among 42,311 US men in the Health Professionals Follow-up Study. Diet was assessed every 4 years by food frequency questionnaires. The authors confirmed 207 cases of clinically or histopathologically diagnosed oral premalignant lesions occurring between 1986 and 2002. Multivariate-adjusted relative risks were calculated from proportional hazards models. Significant inverse associations were observed with citrus fruits, citrus fruit juice, and vitamin-C-rich fruits and vegetables, indicating 30-40% lower risks with greater intakes (e.g., citrus fruit juice quintile 5 vs. quintile 1 relative risk = 0.65, 95% confidence interval: 0.42, 0.99). Inverse associations with fruits did not vary by smoking status and were stronger in analyses of baseline consumption, with a 10-year lag time to disease follow-up (quintile 5 vs. quintile 1 relative risk = 0.41, 95% confidence interval: 0.20, 0.82; p = 0.01). No associations were observed with total vegetables or with beta-carotene-rich or lycopene-rich fruits and vegetables. For current smokers, green leafy vegetables (ptrend = 0.05) and beta-carotene-rich fruits and vegetables (ptrend = 0.02) showed significant linear trends of increased risk (one additional serving/day relative risk = 1.7). The risk of oral premalignant lesions was significantly reduced with higher consumption of fruits, particularly citrus fruits and juices, while no consistent associations were apparent for vegetables.

Adult↗

PubMed2006

Adult weight change and risk of postmenopausal breast cancer.

CONTEXT: Endogenous hormones are a primary cause of breast cancer. Adiposity affects circulating hormones, particularly in postmenopausal women, and may be a modifiable risk factor for breast cancer. OBJECTIVE: To assess the associations of adult weight change since age 18 years and since menopause with the risk of breast cancer among postmenopausal women. DESIGN, SETTING, AND PARTICIPANTS: Prospective cohort study within the Nurses' Health Study. A total of 87,143 postmenopausal women, aged 30 to 55 years and free of cancer, were followed up for up to 26 years (1976-2002) to assess weight change since age 18 years. Weight change since menopause was assessed among 49,514 women who were followed up for up to 24 years. MAIN OUTCOME MEASURE: Incidence of invasive breast cancer. RESULTS: Overall, 4393 cases of invasive breast cancer were documented. Compared with those who maintained weight, women who gained 25.0 kg or more since age 18 years were at an increased risk of breast cancer (relative risk [RR], 1.45; 95% confidence interval [CI], 1.27-1.66; P<.001 for trend), with a stronger association among women who have never taken postmenopausal hormones (RR,1.98; 95% CI, 1.55-2.53). Compared with weight maintenance, women who gained 10.0 kg or more since menopause were at an increased risk of breast cancer (RR, 1.18; 95% CI, 1.03-1.35; P = .002 for trend). Women who had never used postmenopausal hormones, lost 10.0 kg or more since menopause, and kept the weight off were at a lower risk than those who maintained weight (RR, 0.43; 95% CI, 0.21-0.86; P = .01 for weight loss trend). Overall, 15.0% (95% CI, 12.8%-17.4%) of breast cancer cases in this population may be attributable to weight gain of 2.0 kg or more since age 18 years and 4.4% (95% CI, 3.6%-5.5%) attributable to weight gain of 2.0 kg or more since menopause. Among those who did not use postmenopausal hormones, the population attributable risks are 24.2% (95% CI, 19.8%-29.1%) for a weight gain since age 18 years and 7.6% (95% CI, 5.9%-9.7%) for weight gain since menopause. CONCLUSIONS: These data suggest that weight gain during adult life, specifically since menopause, increases the risk of breast cancer among postmenopausal women, whereas weight loss after menopause is associated with a decreased risk of breast cancer. Thus, in addition to other known benefits of healthy weight, our results provide another reason for women approaching menopause to maintain or lose weight, as appropriate.

Adiposity↗

PubMed2006

CFH gene variant, Y402H, and smoking, body mass index, environmental associations with advanced age-related macular degeneration.

OBJECTIVES: We tested the hypothesis that modifiable lifestyle factors alter the genetic susceptibility associated with a common coding variant in the complement factor H (CFH) gene, Y402H, for the leading cause of blindness among the elderly, age-related macular degeneration (AMD). METHODS: In this case-control association analysis, Caucasian participants in the multicenter Age-Related Eye Disease Study with advanced AMD (n = 574 cases) or no AMD (n = 280 controls) were evaluated. AMD status was determined by grading of fundus photographs. Risk factors including cigarette smoking and body mass index (BMI) were assessed and DNA specimens were genotyped for the variant in the CFH gene. Unconditional logistic regression analyses were performed. Attributable risks and multivariable AMD risk scores were calculated. RESULTS: The number of risk alleles for Y402H was associated with advanced AMD, with odds ratios (OR) of 2.7 (95% confidence interval (CI) 1.8-3.8) for the CT heterozygous genotype and OR 7.4 (4.7-11.8) for the homozygous CC risk genotype, after controlling for demographic and behavioral risk factors. Current cigarette smoking (OR 5.1) and high BMI > or =30 (OR 2.1) were independently related to AMD, controlling for genotype. The association between AMD and BMI varied dependent on genotype (P interaction = 0.006 for the CT vs. TT genotype). The CC genotype plus higher BMI (OR 5.9) or smoking (OR 10.2) conferred the greatest risks. Gene plus environment risk scores provided an area under the receiver operating characteristic (ROC) curve of 0.70-0.75. CONCLUSIONS: Genetic and environmental factors are independently related to advanced AMD, and modifiable factors alter genetic susceptibility. The AMD risk score identifies a highly susceptible population.

Aged↗

PubMed2006

Unopposed estrogen therapy and the risk of invasive breast cancer.

BACKGROUND: Although short-term unopposed estrogen use does not seem to increase breast cancer risk, the effect of longer-term estrogen use remains unclear. We sought to assess the relationship between longer-term use of unopposed estrogen and the risk of invasive breast cancer over an extended follow-up period. METHODS: Within the Nurses' Health Study, a prospective cohort study, we observed 11 508 postmenopausal women who had a hysterectomy and reported information on estrogen use at baseline (1980). The study population was expanded every 2 years to include women who subsequently became postmenopausal and had a hysterectomy, so that 28 835 women were included in the final follow-up period (2000-2002). Estrogen use was assessed from self-reported data on biennial questionnaires. The main outcome was invasive breast cancer. RESULTS: A total of 934 invasive breast cancers were included in the analysis. Breast cancer risk increased with duration of unopposed estrogen use among longer-term users with the highest risk seen in cancers positive for estrogen receptor (ER+) and progesterone receptor (PR+). The multivariate relative risks (RRs) and 95% confidence intervals (CIs) for breast cancer with current use of unopposed estrogen for less than 5 years, 5 to 9.9 years, 10 to 14.9 years, 15 to 19.9 years, and 20 years or longer were, respectively, 0.96 (95% CI, 0.75-1.22), 0.90 (95% CI, 0.73-1.12), 1.06 (95% CI, 0.87-1.30), 1.18 (95% CI, 0.95-1.48), and 1.42 (95% CI, 1.13-1.77) (P for trend <.001). The risk of ER+/PR+ breast cancers was noted to be statistically significant after 15 years of current use (RR, 1.48; 95% CI, 1.05-2.07). CONCLUSION: Users of unopposed estrogen were at increased risk of breast cancer but only after longer-term use.

Aged↗

PubMed2006

Tubal sterilization in relation to breast cancer risk.

Tubal sterilization methods may damage surrounding tissue, potentially disrupting the ovarian blood supply and hormonal functioning, and may decrease breast cancer risk. We examined this hypothesis, within the Nurses' Health Study, among 77,511 women, aged 30-55 years and free of cancer at the start of follow-up in 1976. We documented 4,176 cases of invasive breast cancer from 1976 to 2000. Cox proportional hazards models, adjusting for multiple breast cancer risk factors, provided rate ratios (RR) and 95% confidence intervals (CI). Overall, tubal sterilization was not associated with breast cancer risk (RR=0.95, 95% CI=0.88-1.03). However, tubal sterilizations performed from 1970 to 1974 were inversely associated with risk (RR=0.84, 95% CI=0.73-0.97), while procedures performed in other years were not associated with risk. Among women with procedures performed in 1970-1974, those who were >or=35 years old at the time of sterilization were at the lowest risk (RR=0.81, 95% CI=0.66-0.98), while younger women had a suggested decreased risk (RR=0.87, 95% CI=0.72-1.06). Overall, tubal sterilization was not associated with breast cancer risk. However, a modest inverse association was observed at a time when the potentially destructive unipolar electrocautery method was commonly used, providing some support for an association between lower lifetime exposure to hormones and a decreased risk of breast cancer.

Adult↗

PubMed2006

Cigarette smoking, fish consumption, omega-3 fatty acid intake, and associations with age-related macular degeneration: the US Twin Study of Age-Related Macular Degeneration.

OBJECTIVE: To evaluate modifiable risk and protective factors for age-related macular degeneration (AMD) among elderly twins. METHODS: The US Twin Study of Age-Related Macular Degeneration comprises elderly male twins from the National Academy of Sciences-National Research Council World War II Veteran Twin Registry. To determine genetic and environmental risk factors for AMD, twins were surveyed for a prior diagnosis of AMD and underwent an eye examination, fundus photography, and food frequency and risk factor questionnaires. This environmental component of the study includes 681 twins: 222 twins with AMD (intermediate or late stages) and 459 twins with no maculopathy or early signs. Risk for AMD according to cigarette smoking and dietary fat intake was estimated using logistic regression analyses. RESULTS: Current smokers had a 1.9-fold increased risk (95% confidence interval, 0.99-3.68, P = .06) of AMD while past smokers had about a 1.7-fold increased risk (95% confidence interval, 1.2-2.6, P = .009). Increased intake of fish reduced risk of AMD, particularly for 2 or more servings per week (P trend = .04). Dietary omega-3 fatty intake was inversely associated with AMD (odds ratio, 0.55; 95% confidence interval, 0.32-0.95) comparing the highest vs lowest quartile. Reduction in risk of AMD with higher intake of omega-3 fatty acids was seen primarily among subjects with low levels (below median) of linoleic acid intake, an omega-6 fatty acid (P trend<.001). The attributable risk percentage was 32% for smoking and the preventive fraction was 22% for higher omega-3 intake. CONCLUSIONS: This study of twins provides further evidence that cigarette smoking increases risk while fish consumption and omega-3 fatty acid intake reduce risk of AMD.

Aged↗

PubMed2006

Smoking imputation and lung cancer in railroad workers exposed to diesel exhaust.

BACKGROUND: An association between diesel exhaust exposure and lung cancer mortality in a large retrospective cohort study of US railroad workers has previously been reported. However, specific information regarding cigarette smoking was unavailable. METHODS: Birth cohort, age, job, and cause of death specific smoking histories from a companion case-control study were used to impute smoking behavior for 39,388 railroad workers who died 1959-1996. Mortality analyses incorporated the effect of smoking on lung cancer risk. RESULTS: The smoking adjusted relative risk of lung cancer in railroad workers exposed to diesel exhaust compared to unexposed workers was 1.22 (95% CI = 1.12-1.32), and unadjusted for smoking the relative risk was 1.35 (95% CI = 1.24-1.46). CONCLUSIONS: These analyses illustrate the use of imputation in record-based occupational health studies to assess potential confounding due to smoking. In this cohort, small differences in smoking behavior between diesel exposed and unexposed workers did not explain the elevated lung cancer risk.

Adult↗

PubMed2006

Characteristics at diagnosis of type 1 diabetes in children younger than 6 years.

OBJECTIVE: To characterize the prodrome, presentation, family history, and biochemical status at diagnosis of type 1 diabetes mellitus (T1D) in children under age 6 years. STUDY DESIGN: This was a retrospective chart review of patients hospitalized at diagnosis with T1D from 1990 to 1999 in a children's hospital. RESULTS: A total of 247 children were hospitalized, 44% of whom presented in diabetic ketoacidosis (DKA). When stratified by 2-year age intervals, only total carbon dioxide (tCO(2)) was significantly lower in the youngest children (P = .02), and the duration of candidiasis was significantly longer in those children presenting in DKA (P = .004). Parents were more likely to recognize symptomatic hyperglycemia in children older than 2 years (P < .0001). Most parents sought care for their child suspecting that the child had diabetes; the other children were diagnosed when presenting with another concern. Only gender and tCO(2) were significantly correlated with hemoglobin A1c (HbA1c); age-adjusted HbA1c was 0.64% higher in girls compared with boys (P = .045), and each 1-mmol/L decrement in tCO(2) increased the age- and gender-adjusted HbA1c by 0.086% (P < .001). CONCLUSIONS: A high proportion of children under age 6 years present critically ill at the diagnosis of T1D. When any of the classic symptoms of diabetes or a yeast infection is present, a serum glucose level should be measured.

Blood Glucose↗

PubMed2006

Calcium and dairy intakes in relation to long-term weight gain in US men.

BACKGROUND: The role of calcium in the maintenance of body weight remains controversial. OBJECTIVE: We investigated the association between calcium and dairy intakes and 12-y weight change in US men. DESIGN: This study was conducted with the use of data from the Health Professionals Follow-up Study, a prospective cohort of men aged 40-75 y in 1986. Data on lifestyle factors and diet were updated biennially with self-administered questionnaires. The participants reported their body weight in 1986 and in 1998. The outcome in our study was 12-y weight change. We used multivariate linear regression to examine how baseline calcium intake (n = 23,504) and change in calcium intake (n = 19,615) were associated with weight change. Because dairy foods are the predominant source of calcium in the diet, we also evaluated a similar association with dairy intake. RESULTS: In a multivariate analysis with adjustment for potential confounders, baseline or change in intake of total calcium was not significantly associated with weight change. In addition, we did not find any association with dietary, dairy, or supplemental calcium intake when evaluated separately. The men with the largest increase in total dairy intake gained slightly more weight than did the men who decreased intake the most (3.14 compared with 2.57 kg; P for trend = 0.001). This association was primarily due to an increase in high-fat dairy intake. Low-fat dairy intake was not significantly associated with weight change. CONCLUSION: Our data do not support the hypothesis that an increase in calcium intake or dairy consumption is associated with lower long-term weight gain in men.

Adult↗

PubMed2006

The Wilcoxon signed rank test for paired comparisons of clustered data.

The Wilcoxon signed rank test is a frequently used nonparametric test for paired data (e.g., consisting of pre- and posttreatment measurements) based on independent units of analysis. This test cannot be used for paired comparisons arising from clustered data (e.g., if paired comparisons are available for each of two eyes of an individual). To incorporate clustering, a generalization of the randomization test formulation for the signed rank test is proposed, where the unit of randomization is at the cluster level (e.g., person), while the individual paired units of analysis are at the subunit within cluster level (e.g., eye within person). An adjusted variance estimate of the signed rank test statistic is then derived, which can be used for either balanced (same number of subunits per cluster) or unbalanced (different number of subunits per cluster) data, with an exchangeable correlation structure, with or without tied values. The resulting test statistic is shown to be asymptotically normal as the number of clusters becomes large, if the cluster size is bounded. Simulation studies are performed based on simulating correlated ranked data from a signed log-normal distribution. These studies indicate appropriate type I error for data sets with > or =20 clusters and a superior power profile compared with either the ordinary signed rank test based on the average cluster difference score or the multivariate signed rank test of Puri and Sen. Finally, the methods are illustrated with two data sets, (i) an ophthalmologic data set involving a comparison of electroretinogram (ERG) data in retinitis pigmentosa (RP) patients before and after undergoing an experimental surgical procedure, and (ii) a nutritional data set based on a randomized prospective study of nutritional supplements in RP patients where vitamin E intake outside of study capsules is compared before and after randomization to monitor compliance with nutritional protocols.

Humans↗

PubMed2006

Extension of the rank sum test for clustered data: two-group comparisons with group membership defined at the subunit level.

The Wilcoxon rank sum test is widely used for two-group comparisons for nonnormal data. An assumption of this test is independence of sampling units both between and within groups. In ophthalmology, data are often collected on two eyes of an individual, which are highly correlated. In ophthalmological clinical trials, randomization is usually performed at the subject level, but the unit of analysis is the eye. If the eye is used as the unit of analysis, then a modification to the usual Wilcoxon rank sum variance formula must be made to account for the within-cluster dependence. For some clustered data designs, where the unit of analysis is the subunit, group membership may be defined at the subunit level. For example, in some randomized ophthalmologic clinical trials, different treatments may be applied to fellow eyes of some patients, while the same treatment may be applied to fellow eyes of other patients. In general, binary eye-specific covariates may be present (scored as exposed or unexposed) and one wishes to compare nonnormally distributed outcomes between exposed and unexposed eyes using the Wilcoxon rank sum test while accounting for the clustering. In this article, we present a corrected variance formula for the Wilcoxon rank sum statistic in the setting of eye (subunit)-specific covariates. We apply it to compare ocular itching scores in ocular allergy patients between eyes treated with active versus placebo eye drops, where some patients receive the same eye drop in both eyes, while other patients receive different eye drops in fellow eyes. We also present comparisons between the clustered Wilcoxon test and each of the signed rank tests and mixed model approaches and show dramatic differences in power in favor of the clustered Wilcoxon test for some designs.

Biometry↗

PubMed2006

Impact of extended-duration shifts on medical errors, adverse events, and attentional failures.

BACKGROUND: A recent randomized controlled trial in critical-care units revealed that the elimination of extended-duration work shifts (> or =24 h) reduces the rates of significant medical errors and polysomnographically recorded attentional failures. This raised the concern that the extended-duration shifts commonly worked by interns may contribute to the risk of medical errors being made, and perhaps to the risk of adverse events more generally. Our current study assessed whether extended-duration shifts worked by interns are associated with significant medical errors, adverse events, and attentional failures in a diverse population of interns across the United States. METHODS AND FINDINGS: We conducted a Web-based survey, across the United States, in which 2,737 residents in their first postgraduate year (interns) completed 17,003 monthly reports. The association between the number of extended-duration shifts worked in the month and the reporting of significant medical errors, preventable adverse events, and attentional failures was assessed using a case-crossover analysis in which each intern acted as his/her own control. Compared to months in which no extended-duration shifts were worked, during months in which between one and four extended-duration shifts and five or more extended-duration shifts were worked, the odds ratios of reporting at least one fatigue-related significant medical error were 3.5 (95% confidence interval [CI], 3.3-3.7) and 7.5 (95% CI, 7.2-7.8), respectively. The respective odds ratios for fatigue-related preventable adverse events, 8.7 (95% CI, 3.4-22) and 7.0 (95% CI, 4.3-11), were also increased. Interns working five or more extended-duration shifts per month reported more attentional failures during lectures, rounds, and clinical activities, including surgery and reported 300% more fatigue-related preventable adverse events resulting in a fatality. CONCLUSIONS: In our survey, extended-duration work shifts were associated with an increased risk of significant medical errors, adverse events, and attentional failures in interns across the United States. These results have important public policy implications for postgraduate medical education.

Adult↗