Sas system software

Statistical analysis was performed using SAS system software (version 9.1, SAS Institute, Cary, NC, USA). All the data in each experiment was subjected to one-way analysis of variance (ANOVA) and means were compared by Duncan's multiple range test. Values of P < 0.05 were considered statistically significant.

We used Fisher exact test to determine the difference between patients with or without hypoglycemia, according to patient characteristics. After calculating differences in hypoglycemia incidence and its 95% confidence intervals (CI) among subgroups, we calculated number needed to harm (NNH) by the inverse of incidence difference between subgroups. Among subgroups having statistically significant differences, we calculated multivariate adjusted relative risk estimates and 95% CI by using negative binomial regression model. All statistical tests were two-sided at α = 0.05. All statistical calculations were performed using SAS system software, version 9.1.3 (SAS Institute Inc., Cary, NC, USA) and R 2.15.2.

Biochemical parameters, BCS, DIM, parity and milk fatty acids data were analyzed using the SAS system software (version 9.4; SAS Institute Inc., Cary, North Carolina, USA). A one-way ANOVA was used to evaluate the differences of the total lipid fraction within the two groups (BHB0 vs. BHB1). This test was used to select the parameters with a predictive power in order to diagnose subclinical ketosis or hyperketonemia. The hypotheses of linear model on the residuals were graphically assessed.
The Receiver Operating Characteristic (ROC) (MedCalc Sofware Ltd., Ostend, Belgium) curves were performed to establish threshold value of each predictive biochemical parameter or forecast milk fatty acids. The ROC curves were derived from the analysis of data of all experimental animals (BHB0-BHB1) with the discriminant of ewes with BHB > 0.86 mmol/L (BHB1) [9 ,11 ]. The area under the curve (AUC) shows the diagnostic power of the test. Therefore, the optimal cut-off of each parameter based on Youden criterion, was calculated to discriminate the subclinical ketosis or hyperketonemia groups in ewes in early phase of lactation. Statistical significance was set at p-value ≤ 0.05.

Patients will be randomly assigned to one of the two arms at a ratio of 1:1. The randomization list will be computer-generated by the study statistician using SAS system software (version 9.2, SAS Institute, Inc., Cary, NC, USA). The randomization process will be centralized through a secured website managed by the Bordeaux University Hospital Clinical Trial Unit (CTU) («Unité de Soutien Méthodologique à la Recherche Clinique et Epidémiologique (USMR)»). After confirmation of the patient’s eligibility criteria, the investigator will access the website of the CTU, which will provide the patient’s unique allocation number and randomization group. Access to the final dataset will be limited to the investigators.

Computations for all experiments were performed using SAS system software [25 ]. A randomized complete block design with factorial arrangements was used for each group of insects as follow: 2 × 5 × 3 (young adults) and 3 × 5 × 3 (old adults) for mortality (strains: NI8 and GHA; concentrations: n × 10⁷, n × 10⁶, n × 10⁵, n × 10⁴ spores/mL; and evaluation times: 3, 5, and 10 days after sprayed) and 2 × 5 (young adults) and 3 × 5 (old adults) for sporulation (strains: NI8, GHA, and KUDSC; and concentrations: same as above). Each treatment combination was repeated four times. Nonparametric estimates of the survival function of kudzu bugs were compared between treatments using PROC LIFETEST [25 ]. Statistical differences in the survival of M. cribraria were declared based on the log-rank statistic. Mortality and infection were analyzed by using PROC GLM to detect differences between treatments for each group of insects. Mortality and sporulation data for each group of insects and each strain were analyzed by PROBIT [25 ] using common logarithm (log to the base 10) of the concentration value.