Home | Cochrane Pregnancy and Childbirth



Planning your analysesExample plan for data and analysisThis example is adapted from a protocol on Prenatal administration of progestogens for preventing spontaneous preterm birth in women with a multiple pregnancy. The authors were asked to make decisions about whether it is valid to pool certain types of interventions together or present as separate analyses. The options for data capture and analysis are set out below. Primary outcomes: maternal mortality; preterm birth <34 weeks; perinatal mortality; Secondary outcomes: prelabour spontaneous rupture of membranes, preterm birth < 37 weeks, preterm birth < 28 weeks; adverse drug reaction; caesarean sectionPlanned subgroup analysis: planned subgroup analysis should be kept to a minimum and generally restricted to the review’s primary outcomes (or the 7 GRADE outcomes): Time of treatment commencing (before 20 weeks' gestation versus after 20 weeks' gestation)Different dosage regimens (divided arbitrarily into a cumulative dose of less than 250 mg per week versus a dose of greater than or equal to 250 mg per week)There are two different options for planning your data and analysis – these are outlined below. Please use either option 1 or option 2.Option 1With this option, where you plan to include subgroups, you don’t have a separate main analysis but all data must be accounted for in the planned subgroup analysis. This is achieved by creating a third subgroup where the data without the specific subgroup characteristics are parison 1 - Intramuscular (IM) progesterone versus no treatment or placeboOutcome or Subgroup1.1 Maternal mortality (subgroup by dose)?? 1.1.1 Low dose (250 mg weekly or less)?? 1.1.2 High dose (greater than or equal to 250 mg weekly) 1.1.3 Mixed or unclear1.2 Maternal mortality (subgroup by timing)?? 1.2.1 Commencing before 20 weeks' gestation?? 1.2.2 Commencing after 20 weeks' gestation 1.2.3 Mixed or unclear1.3 Preterm birth less than 34 weeks (subgroup by dose)?? 1.3.1 Low dose (250 mg weekly or less)?? 1.3.2 High dose (greater than or equal to 250 mg weekly) 1.3.3 Mixed or unclear1.4 Preterm birth less than 34 week (subgroup by timing)?? 1.4.1 Commencing before 20 weeks' gestation?? 1.4.2 Commencing after 20 weeks' gestation?? 1.4.3 Mixed or unclear1.5 Perinatal mortality (subgroup by dose)?? 1.5.1 Low dose (250 mg weekly or less)?? 1.5.2 High dose (greater than 250 mg weekly)?? 1.5.3 Mixed or unclear1.6 Perinatal mortality (subgroup by timing)?? 1.6.1 Commencing before 20 weeks' gestation?? 1.6.2 Commencing after 20 weeks' gestation?? 1.6.3 Mixed or unclear1.10 Prelabour spontaneous rupture of membranes1.11 Preterm birth less than 37 weeks1.12 Preterm birth less than 28 weeks1.13 Adverse drug reaction1.11 Caesarean sectionComparison 2 - Vaginal progesterone versus no treatment or placebo (list analyses – as above)Option 2With this option, you have the main analysis (i.e. no subgroups) and then have subsequent subgroup analysis listed beneath the main analysis. Note, that in this example, you would perform interaction tests for subgroup differences – however, it would not make sense to display a pooled effect in your subgroup analysis (e.g. 1.2). The only relevant pooled effect should come from the main analysis (e.g. 1.1). Comparison 1 - Intramuscular (IM) progesterone versus no treatment or placeboOutcome or Subgroup1.1 Maternal mortality1.2 Maternal mortality (subgroup by dose)?? 1.2.1 Low dose (250 mg weekly or less)?? 1.2.2 High dose (greater than or equal to 250 mg weekly)1.3 Maternal morta lity (subgroup by timing)?? 1.3.1 Commencing before 20 weeks' gestation?? 1.3.2 Commencing after 20 weeks' gestation1.4 Preterm birth less than 34 weeks1.5 Preterm birth less than 34 weeks (subgroup by dose)?? 1.5.1 Low dose (250 mg weekly or less)?? 1.5.2 High dose (greater than or equal to 250 mg weekly)1.6 Preterm birth less than 34 week (subgroup by timing)?? 1.6.1 Commencing before 20 weeks' gestation?? 1.6.2 Commencing after 20 weeks' gestation1.7 Perinatal mortality1.8 Perinatal mortality (subgroup by dose)?? 1.8.1 Low dose (250 mg weekly or less)?? 1.8.2 High dose (greater than 250 mg weekly)1.9 Perinatal mortality (subgroup by timing)?? 1.9.1 Commencing before 20 weeks' gestation?? 1.9.2 Commencing after 20 weeks' gestation1.10 Prelabour spontaneous rupture of membranes1.11 Preterm birth less than 37 weeks1.12 Preterm birth less than 28 weeks1.13 Adverse drug reaction1.11 Caesarean sectionComparison 2 - Vaginal progesterone versus no treatment or placebo (list analyses – as above) ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download