The publication data currently available has been vetted by Vanderbilt faculty, staff, administrators and trainees. The data itself is retrieved directly from NCBI's PubMed and is automatically updated on a weekly basis to ensure accuracy and completeness.
If you have any questions or comments, please contact us.
BACKGROUND - A protective effect of breastfeeding on childhood lymphoma has been indicated but supportive evidence is limited.
METHOD - Data from a population-based case-control study of childhood cancer in Shanghai, including 82 lymphoma cases and 159 acute leukaemia cases and their age- and sex-matched community controls, were analysed.
RESULTS - After adjustment for potentially confounding variables, a slight, although non-significant, reduction in risk of lymphoma was observed among children who were breastfed as infants versus those who were not (odds ratio [OR] = 0.69; 95% CI: 0.3-1.7). The reduction was somewhat greater for children who had been breastfed longer and appeared to pertain primarily to Hodgkin's disease and to cases diagnosed before the age of 6 years. As expected, there was no reduction in risk of acute leukaemia associated with breastfeeding.
CONCLUSIONS - Although providing neither strong support for nor refuting the study hypothesis, these data suggest that if breastfeeding does reduce the risk of lymphoma, its protective effect among Chinese children is likely modest in magnitude and concentrated in certain subgroups defined by length of breastfeeding, age at diagnosis and histological subtype of cancer.
The probability score (PS) or Brier score has been used in a large number of studies in which physician judgment performance was assessed. However, the covariance decomposition of the PS has not previously been used to evaluate medical judgment. The authors introduce the technique and demonstrate it by analyzing prognostic estimates of three groups: physicians, their patients, and the patients' decision-making surrogates. The major components of the covariance decomposition--bias, slope, and scatter--are displayed in covariance graphs for each of the three groups. The decomposition reveals that whereas the physicians have the best overall estimation performance, their bias and their scatter are not always superior to those of the other two groups. This is primarily due to two factors. First, the physicians' prognostic estimates are pessimistic. Second, the patients place the large majority of their estimates in the most optimistic category, thereby achieving low scatter. The authors suggest that the calculational simplicity of this decomposition, its informativeness, and the intuitive nature of its components make it a useful tool with which to analyze medical judgment.
In the preceding companion article in this issue, an optical dye and a nitroxide radical were combined in a new dual function probe, 5-SLE. In this report, it is demonstrated that time-resolved optical anisotropy and electron paramagnetic resonance (EPR) data can be combined in a single analysis to measure rotational dynamics. Rigid-limit and rotational diffusion models for simulating nitroxide EPR data have been incorporated into a general non-linear least-squares procedure based on the Marquardt-Levenberg algorithm. Simultaneous fits to simulated time-resolved fluorescence anisotropy and linear EPR data, together with simultaneous fits to experimental time-resolved phosphorescence anisotropy decays and saturation transfer EPR (ST-EPR) spectra of 5-SLE noncovalently bound to bovine serum albumin (BSA) have been performed. These results demonstrate that data from optical and EPR experiments can be combined and globally fit to a single dynamic model.
The analysis and interpretation of the data collected in SUPPORT provide great potential for understanding the relationships among treatment choices, patient and physician values and preferences, perceptions about the risks and benefits of treatments, institutional characteristics, and outcomes (as measured by quality of life, survival, and satisfaction). The complicated analyses required to elucidate these relationships will pose many technical challenges in dealing with longitudinal observational data collected from seriously ill patients at multiple sites. Major challenges include the handling of incomplete data, proper parameterization of treatment effects, strategies to avoid various potential biases, validating predictive models, and constructing endpoints that combine survival with quality of life. Within the structure of the SUPPORT study, mechanisms have been established to guide the analyses and to ensure their quality and validity.
Recruitment of disease-free subjects into cohort studies and measurement of their time from exposure/infection to disease selectively excludes individuals (the unseen sample) who had earlier exposure and who have shorter times to disease. The unseen and observed samples may differ in other characteristics in addition to incubation period and exposure/infection time. For data with known truncation times, we develop non-parametric maximum likelihood estimates of the size, exposure/infection dates and distribution of incubation time in the unseen sample. We provide procedures to estimate and compensate for the biasing effects due to exclusion of the unseen sample in descriptive and survival analysis. We give consistency properties of these estimates and assess variability using bootstrap methods. One can use imputation to derive the above estimates from data with unknown truncation times that have been estimated parametrically. Application is made to an AIDS cohort study of over 5000 homosexual men. Important estimates obtained from this application are the annual seroconversion rates from 1978 to 1983, not otherwise obtainable in this study population.