METHODS: We assessed the use of composite outcomes in neonatal RCTs included in Cochrane Neonatal reviews published till November 2017. Two authors reviewed the components of the composite outcomes to compare their patient importance and computed the ratios of effect sizes and event rates between the components, with an a priori threshold of 1.5, indicating a substantial difference. Descriptive statistics were presented.
RESULTS: We extracted 7,766 outcomes in 2,134 RCTs in 312 systematic reviews. Among them, 55 composite outcomes (0.7%) were identified in 46 RCTs. The vast majority (92.7%) of composite outcomes had 2 components, with death being the most common component (included 51 times [92.7%]). The components in nearly three-quarters of the composite outcomes (n = 40 [72.7%]) had different patient importance, while the effect sizes and event rates differed substantially between the components in 27 (49.1%) and 35 (63.6%) outcomes, respectively, with up to 43-fold difference in the event rates observed.
CONCLUSIONS: The majority of composite outcomes in neonatal RCTs had different patient importance with contrasting effect sizes and event rates between the components. In patient communication, clinicians should highlight individual components, rather than the composites, with explanation on the relationship between the components, to avoid misleading impression on the effect of the intervention. Future trials should report the estimates of all individual components alongside the composite outcomes presented.
OBJECTIVES: To evaluate the effectiveness of maintenance tocolytic therapy with oral nifedipine on the reduction of adverse neonatal outcomes and the prolongation of pregnancy by performing an individual patient data meta-analysis (IPDMA).
SEARCH STRATEGY: We searched PubMed, Embase, and Cochrane databases for randomised controlled trials of maintenance tocolysis therapy with nifedipine in preterm labour.
SELECTION CRITERIA: We selected trials including pregnant women between 24 and 36(6/7) weeks of gestation (gestational age, GA) with imminent preterm labour who had not delivered after 48 hours of initial tocolysis, and compared maintenance nifedipine tocolysis with placebo/no treatment.
DATA COLLECTION AND ANALYSIS: The primary outcome was perinatal mortality. Secondary outcome measures were intraventricular haemorrhage (IVH), necrotising enterocolitis (NEC), infant respiratory distress syndrome (IRDS), prolongation of pregnancy, GA at delivery, birthweight, neonatal intensive care unit admission, and number of days on ventilation support. Pre-specified subgroup analyses were performed.
MAIN RESULTS: Six randomised controlled trials were included in this IPDMA, encompassing data from 787 patients (n = 390 for nifedipine; n = 397 for placebo/no treatment). There was no difference between the groups for the incidence of perinatal death (risk ratio, RR 1.36; 95% confidence interval, 95% CI 0.35-5.33), intraventricular haemorrhage (IVH) ≥ grade II (RR 0.65; 95% CI 0.16-2.67), necrotising enterocolitis (NEC) (RR 1.15; 95% CI 0.50-2.65), infant respiratory distress syndrome (IRDS) (RR 0.98; 95% CI 0.51-1.85), and prolongation of pregnancy (hazard ratio, HR 0.74; 95% CI 0.55-1.01).
CONCLUSION: Maintenance tocolysis is not associated with improved perinatal outcome and is therefore not recommended for routine practice.
TWEETABLE ABSTRACT: Nifedipine maintenance tocolysis is not associated with improved perinatal outcome or pregnancy prolongation.
OBJECTIVE: The aim of this proof-of-concept study was to evaluate whether combining population pharmacokinetic and machine learning approaches could provide a more accurate prediction of the clearance of renally eliminated drugs in individual neonates.
METHODS: Six drugs that are primarily eliminated by the kidneys were selected (vancomycin, latamoxef, cefepime, azlocillin, ceftazidime, and amoxicillin) as 'proof of concept' compounds. Individual estimates of clearance obtained from population pharmacokinetic models were used as reference clearances, and diverse machine learning methods and nested cross-validation were adopted and evaluated against these reference clearances. The predictive performance of these combined methods was compared with the performance of two other predictive methods: a covariate-based maturation model and a postmenstrual age and body weight scaling model. Relative error was used to evaluate the different methods.
RESULTS: The extra tree regressor was selected as the best-fit machine learning method. Using the combined method, more than 95% of predictions for all six drugs had a relative error of < 50% and the mean relative error was reduced by an average of 44.3% and 71.3% compared with the other two predictive methods.
CONCLUSION: A combined population pharmacokinetic and machine learning approach provided improved predictions of individual clearances of renally cleared drugs in neonates. For a new patient treated in clinical practice, individual clearance can be predicted a priori using our model code combined with demographic data.