Predicting the probability of abortion in dairy cows: a hierarchical Bayesian logistic-survival model using sequential pregnancy data.

Although abortion contributes substantially to poor reproductive health of dairy herds, little is known about the predictability of abortion based on age, previous abortion or gravidity (number of previous pregnancies). A poor understanding of effects of maternal factors on abortion risk exists, in part, because of methodological difficulties related to non-independence of multiple pregnancies of the same cow in analysis of fetal survival data. We prospectively examined sequential pregnancies to investigate relationships between fetal survival and putative dam risk factors for 2991 abortions from 24,706 pregnancies of 13,145 cows in nine California dairy herds. Relative risks and predicted probabilities of abortion (PPA) were estimated using a previously described hierarchical Bayesian logistic-survival model generalized to incorporate longitudinal data of multiple pregnancies from a single cow. The PPA increased with increasing dam age at conception, with increasing number of previous abortions, and if the previous pregnancy was aborted >60 days in gestation. The PPA decreased with increasing gravidity and with increasing number of days open. For cows that aborted, the median time to fetal death decreased slightly as gravidity increased. The study considers several methodological issues faced in epidemiologic investigations of fetal health, including multi-modal hazard functions, extensive censoring and non-independence of multiple pregnancies. The model improves our ability to predict bovine abortion and to characterize fetal survival, which have important applications to herd health management.

Insulin regulation in AhR-null mice: embryonic cardiac enlargement, neonatal macrosomia, and altered insulin regulation and response in pregnant and aging AhR-null females.

The aryl hydrocarbon receptor (AhR) was originally characterized because of its high affinity binding of 2, 3, 7, 8-tetrachlorodibenzo-p-dioxin. However, studies using AhR-null mice have demonstrated the importance of this protein in normal physiology and development. Here we demonstrate that AhR-null embryos develop cardiac enlargement, and that this phenotype is dependent, at least in part, on the maternal genotype. Neonates born to AhR-null females had increased heart weights regardless of the neonatal genotype, an outcome also observed in gestational diabetes. The cardiac hypertrophy markers, beta-myosin heavy chain and atrial natriuretic factor, and the cardiac proliferative index were increased in AhR-null embryos, indicating that the cardiac enlargement is associated with myocyte hypertrophy and hyperplasia, which begin prior to birth. Importantly, two- to three-month-old pregnant and seven-month-old nonpregnant females, but not nonpregnant three-month-old AhR-null females had significantly decreased fasting plasma insulin levels and a reduced ability to respond to exogenous insulin compared to controls. Despite these alterations in insulin regulation and responsiveness, pregnant AhR females did not have abnormal glucose tolerance tests and did not develop hyperglycemia, classic characteristics of gestational diabetes. However, twenty-three percent of seven-month-old AhR-null females did have altered glucose tolerance tests, but did not show hyperglycemia or increased hemoglobin A1C concentration under normal feeding conditions. While the ultimate cause of the neonatal phenotype remains unclear, these studies establish that the AhR is required for normal insulin regulation in pregnant and older mice and for cardiac development in embryonic mice.

Mortality risk among hemodialysis patients receiving different vitamin D analogs.

Intravenous vitamin D is standard therapy for secondary hyperparathyroidism in hemodialysis (HD) patients. In for-profit dialysis clinics, mortality was higher for patients on calcitriol compared to paricalcitol. Doxercalciferol, a second vitamin D2 analog, is currently available. We assessed mortality associated with each vitamin D analog and with lack of vitamin D therapy in patients who began HD at Dialysis Clinic Inc. (DCI), a not-for-profit dialysis provider. During the 1999-2004 study period we studied 7731 patients (calcitriol: n=3212; paricalcitol: n=2087; doxercalciferol: n=2432). Median follow-up was 37 weeks. Mortality rates (deaths/100 patient-years) were identical in patients on doxercalciferol (15.4, 95% confidence interval (13.6-17.1)) and paricalcitol (15.3 (13.6-16.9)) and higher in patients on calcitriol (19.6 (18.2-21.1)) (P<0.0001). In all models mortality was similar for paricalcitol versus doxercalciferol (hazard ratios=1.0). In unadjusted models, mortality was lower in patients on doxercalciferol (0.80 (0.66, 0.96)) and paricalcitol (0.79 (0.68, 0.92)) versus calcitriol (P<0.05). In adjusted models, this difference was not statistically significant. In all models mortality was higher for patients who did not receive vitamin D versus those who did (1.2 (1.1-1.3)). Mortality in doxercalciferol- and paricalcitol-treated patients was virtually identical. Differences in survival between vitamin D2 and D3 may be smaller than previously reported.

A family of confidence intervals for the ratio of two binomial proportions.

The power divergence family of statistics introduced by Cressie and Read (1984, Journal of the Royal Statistical Society, Series B 46, 440-464) is used to obtain approximate confidence intervals for the ratio of two proportions. Intervals based on inverting the Pearson, likelihood-ratio, and Freeman-Tukey statistics are included in this family. Small-sample comparisons of the intervals are presented.

An empirical assessment of saddlepoint approximations for testing a logistic regression parameter.

Saddlepoint methods provide quick and easy approximations to significance levels for conditional tests of logistic regression parameters. We evaluate the accuracies of saddlepoint approximations for three well-known conditional tests: Bartlett's test for no three-factor interaction in a 2 x 2 x 2 table, the test for trend in a series of probabilities, and the exact test of no association in stratified 2 x 2 tables with a common odds ratio. General recommendations are suggested regarding the use of saddlepoint approximations for exact conditional significance levels.

Assessing the fit of the logistic regression model to individual matched sets of case-control data.

We develop exact conditional methods for checking the fit of a logistic regression model to individual matched sets in case-control studies. Justifications are given for preferring this approach to conventional maximum likelihood-based methods. Data from two studies are used to illustrate the techniques.

Estimating the Mahalanobis distance from mixed continuous and discrete data.

We present a method for estimating the Mahalanobis distance between two multivariate normal populations when a subset of the measurements is observed as ordered categorical responses. Asymptotic properties of the proposed estimator are developed. Two examples are discussed.

Bayesian accelerated failure time analysis with application to veterinary epidemiology.

Standard methods for analysing survival data with covariates rely on asymptotic inferences. Bayesian methods can be performed using simple computations and are applicable for any sample size. We propose a practical method for making prior specifications and discuss a complete Bayesian analysis for parametric accelerated failure time regression models. We emphasize inferences for the survival curve rather than regression coefficients. A key feature of the Bayesian framework is that model comparisons for various choices of baseline distribution are easily handled by the calculation of Bayes factors. Such comparisons between non-nested models are difficult in the frequentist setting. We illustrate diagnostic tools and examine the sensitivity of the Bayesian methods.

Mammography utilization after a benign breast biopsy among Hispanic and non-Hispanic women.

BACKGROUND: In spite of the effectiveness of mammography screening for early detection of breast carcinoma, the use of screening mammography varies widely across racial and ethnic groups. Recently, concerns have been raised about the potential adverse effect a benign breast biopsy may have on subsequent mammography utilization, including subsequent use among minority women. METHODS: Computerized health care claims data for 1991 through 1997 from a managed care organization were used to compare mammography use among Hispanic and non-Hispanic women who had had a mammogram followed by an incisional or excisional benign breast biopsy to women who had had a mammogram and no biopsy. Through survival analysis methods, the time-to-next mammogram was compared among these three groups. RESULTS: The sample included 693 (3.2%) and 289 (1.3%) women who had had a mammogram followed by an incisional biopsy or an excisional biopsy, respectively, and 20,540 (95.4%) women who had had a mammogram and no biopsy. Both Hispanic and non-Hispanic women with a biopsy returned sooner for subsequent mammograms than women without a biopsy (P < 0.0001). Hispanic women without a biopsy returned later than non-Hispanic women without a biopsy (P < 0.0001). However, Hispanic women with an excisional biopsy returned sooner than non-Hispanic women (P < 0.05). CONCLUSIONS: Within a managed care organization, both Hispanic and non-Hispanic women who had had a mammogram followed by a benign breast biopsy returned sooner for a subsequent mammogram than women who had had a mammogram and no biopsy. However, ethnic differences in time-to-next mammogram were observed for women without a biopsy and those with an excisional biopsy. Hispanic women without a biopsy returned later for a subsequent mammogram than non-Hispanic women in similar groups, but those with an excisional biopsy returned sooner.