Dynamic outcome prediction in a socio-demographically diverse population-based cohort of extremely preterm neonates.

OBJECTIVE: Accurate outcome prediction is crucial for counseling parents and providing individualized treatment to extremely premature infants. We sought to improve upon existing prediction model by using a diverse population-based cohort of extremely premature live births (⩽28 weeks' gestation) for survival and survival without severe neonatal morbidity at different times throughout the first week of life and to evaluate potential differences by race/ethnicity and maternal education. STUDY DESIGN: Retrospective cohort study of all California live births from 2007 through 2011 with linked birth, death and hospital discharge records. RESULTS: A total of 6009 infants were included. In the validation data set at time of delivery, the area under the receiver-operating characteristic curve for the model containing all predictors was 0.863 for survival and 0.789 for survival without severe morbidity. The marginal probability of survival without severe neonatal morbidity of an Asian infant born to a mother with <12 years of education compared with the reference (Caucasian infant, mother with ⩾12 years of education) was -0.23 (95% confidence interval (CI) -0.31 to -0.15) for all infants at time of birth and -0.28 (95% CI -0.39 to -0.18) for infants with attempted resuscitation. Notably, no other differences by racial/ethnic category and maternal education emerged. CONCLUSIONS: Probabilities of survival and survival without major morbidity change rapidly throughout the first week of life. Extremely premature infants born to Asian mothers with less than a high school education appear to have a lower probability to survive without significant morbidity compared with their Caucasian peers.

Risk of preterm birth among women using drugs during pregnancy with elevated α-fetoprotein.

OBJECTIVE: Examine the risk of preterm birth (PTB) among women who use drugs during pregnancy and have elevated α-fetoprotein (AFP). STUDY DESIGN: The sample included California singleton live births in 2005 to 2010 contained within a hospital discharge database linked to the Prenatal Screening Program. A selection of mothers who did not use drugs was selected at a ratio of 4:1. Risk of PTB was calculated using adjusted odds ratios and 95% confidence intervals (CIs) for women who did or did not use drugs by their AFP percentile. RESULTS: We identified 7190 women who used drugs and selected 28 760 women who did not. Of women using cocaine with AFP ⩾95th percentile, 43.8% delivered prematurely. Women using drugs with AFP ⩾95th percentile were 11 to 35 times as likely to deliver <32 weeks. CONCLUSION: The combination of drug use and elevated AFP results in high rates of PTB. This combination results in an additive risk.

Outbreak of cyclosporiasis associated with basil in Missouri in 1999.

During the summer of 1999, an outbreak of cyclosporiasis occurred among attendees of 2 events held on 24 July in different counties in Missouri. We conducted retrospective cohort studies of the 2 clusters of cases, which comprised 62 case patients. The chicken pasta salad served at one event (relative risk [RR], 4.25; 95% confidence interval [CI], 1.80-10.01) and the tomato basil salad served at the other event (RR, 2.95; 95% CI, 1.72-5.07) were most strongly associated with illness. The most likely vehicle of infection was fresh basil, which was included in both salads and could have been grown either in Mexico or the United States. Leftover chicken pasta salad was found to be positive for Cyclospora DNA by means of polymerase chain reaction analysis, and 1 sporulated Cyclospora oocyst was found by use of microscopy. This is the second documented outbreak of cyclosporiasis in the United States linked to fresh basil and the first US outbreak for which Cyclospora has been detected in an epidemiologically implicated food item.