Design, construction and evaluation of systems to predict risk in obstetrics.

We present a systematic, practical approach to developing risk prediction systems, suitable for use with large databases of medical information. An important part of this approach is a novel feature selection algorithm which uses the area under the receiver operating characteristic (ROC) curve to measure the expected discriminative power of different sets of predictor variables. We describe this algorithm and use it to select variables to predict risk of a specific adverse pregnancy outcome: failure to progress in labour. Neural network, logistic regression and hierarchical Bayesian risk prediction models are constructed, all of which achieve close to the limit of performance attainable on this prediction task. We show that better prediction performance requires more discriminative clinical information rather than improved modelling techniques. It is also shown that better diagnostic criteria in clinical records would greatly assist the development of systems to predict risk in pregnancy.

Ecological condition of coastal ocean waters of the U.S. continental shelf off South Florida: 2007

A study was initiated with field work in May 2007 to assess the status of ecological condition and stressor impacts throughout the U.S. continental shelf off South Florida, focusing on soft-bottom habitats, and to provide this information as a baseline for evaluating future changes due to natural or human-induced disturbances. The boundaries of the study region extended from Anclote Key on the western coast of Florida to West Palm Beach on the eastern coast of Florida, inclusive of the Florida Keys National Marine Sanctuary (FKNMS), and from navigable depths along the shoreline seaward to the shelf break (~100m). The study incorporated standard methods and indicators applied in previous national coastal monitoring programs — U.S. Environmental Protection Agency’s (EPA) Environmental Monitoring and Assessment Program (EMAP) and National Coastal Assessment (NCA) — including multiple measures of water quality, sediment quality, and biological condition. Synoptic sampling of the various indicators provided an integrative weight-of-evidence approach to assessing condition at each station and a basis for examining potential associations between presence of stressors and biological responses. A probabilistic sampling design, which included 50 stations distributed randomly throughout the region, was used to provide a basis for estimating the spatial extent of condition relative to the various measured indicators and corresponding assessment endpoints (where available). The study was conducted through a large cooperative effort by National Oceanic and Atmospheric Administration (NOAA)/National Centers for Coastal Ocean Science (NCCOS), EPA, U.S. Geological Survey (USGS), NOAA/Oceanic and Atmospheric Research (OAR)/Atlantic Oceanographic and Meteorological Laboratory in Miami, FKNMS, and the Florida Fish and Wildlife Conservation Commission (FWC). The majority of the South Florida shelf had high levels of dissolved oxygen (DO) in near-bottom water (> 5 mg L-1) indicative of “good” water quality.. DO levels in bottom waters exceeded this upper threshold at 98.8% throughout the coastal-ocean survey area. Only 1.2% of the region had moderate DO levels (2-5 mg/L) and no part of the survey area had DO <2.0 mg/L. In addition, offshore waters throughout the region had relatively low levels of total suspended solids (TSS), nutrients, and chlorophyll a indicative of oligotrophic conditions. Results suggested good sediment quality as well. Sediments throughout the region, which ranged from sands to intermediate muddy sands, had low levels of total organic carbon (TOC) below bioeffect guidelines for benthic organisms. Chemical contaminants in sediments were also mostly at low, background levels. For example, none of the stations had chemicals in excess of corresponding Effects-Range Median (ERM) probable bioeffect values or more than one chemical in excess of lower-threshold Effects-Range Low (ERL) values. Cadmium was the only chemical that occurred at moderate concentrations between corresponding ERL and ERM values. Sixty fish samples from 28 stations were collected and analyzed for chemical contaminants. Eleven of these samples (39% of sites) had moderate levels of contaminants, between lower and upper non-cancer human-health thresholds, and ten (36% of sites) had high levels of contaminants above the upper threshold.

Age and growth of the estuarine dolphin (Sotalia guianensis) (Cetacea, Delphinidae) on the Paraná Coast, southern Brazil

Teeth of 71 estuarine dolphins (Sotalia guianensis) incidentally caught on the coast of Paraná State, southern Brazil, were used to estimate age. The oldest male and female dolphins were 29 and 30 years, respectively. The mean distance from the neonatal line to the end of the first growth layer group (GLG) was 622.4 ±19.1 μm (n=48). One or two accessory layers were observed between the neonatal line and the end of the first GLG. One of the accessory layers, which was not always present, was located at a mean of 248.9 ±32.6 μm (n=25) from the neonatal line, and its interpretation remains uncertain.The other layer, located at a mean of 419.6 ±44.6 μm (n=54) from the neonatal line, was always present and was first observed between 6.7 and 10.3 months of age. This accessory layer could be a record of weaning in this dolphin. Although no differences in age estimates were observed between teeth sectioned in the anterior-posterior and buccal-lingual planes, we recommend sectioning the teeth in the buccal-lingual plane in order to obtain on-center sections more easily. We also recommend not using teeth from the most anterior part of the mandibles for age estimation. The number of GLGs counted in those teeth was 50% less than the number of GLGs counted in the teeth from the median part of the mandible of the same animal. Although no significant difference (P>0.05) was found between the total lengths of adult male and female estuarine dolphins, we observed that males exhibited a second growth spurt around five years of age. This growth spurt would require that separate growth curves be calculated for the sexes. The asymptotic length (TL∞), k, and t0 obtained by the von Bertalanffy growth model were 177.3 cm, 0.66, and –1.23, respectively, for females and 159.6 cm, 2.02, and –0.38, respectively, for males up to five years, and 186.4 cm, 0.53 and –1.40, respectively, for males older than five years. The total weight (TW)/total length (TL) equations obtained for male and female estuarine dolphins were TW = 3.156 × 10−6 × TL 3.2836 (r=0.96), and TW = 8.974 × 10−5 × TL 2.6182 (r=0.95), respectively.