Term PROM Research Articles

A05 Use of a sequential method for monitoring safety in patients enrolled in the multicenter acute stroke trial

s 4 0 1 A05 USE OF A SEQUENTIAL METHOD FOR MONITORING SAFETY IN PATIENTS ENROLLED IN THE MULTICENTER ACUTE STROKE TRIAL F. Boutitie, J.P. Boissel, and M. Hommel Unit6 de Pharmacologic Clinique Lyon, France The Multicenter Acute Stroke Trial is a double-blind placebo-controlled randomized clinical trial aimed at evaluating the relative efficacy of a thrombolytic agent (streptokinase) in patients with a recent episode of acute ischemic stroke. The safety profile of the active treatment in this indication is not well known, but severe bleeding, including intracranial hemorrhage, can be expected. Hence, it was judged ethically essential to monitor patients' safety as closely as possible (with symptomatic cerebral hemorrhage as a primary criterion), while keeping the overall type I error rate at its predefined level. Sequential monitoring procedures are especially adapted for this purpose. The rationale for the choice of the Sequential Probability Ratio Test will be presented, as well as the design specifications. Also, results of power calculations obtained using simulated data will be shown, to demonstrate the effect of choosing expected rates of events which are not representative of the true rates in both study groups at the design stage. A06 CHANGING THE STUDY DESIGN DURING THE RECRUITMENT PHASE OF AN INTERNATIONAL PERINATAL MULTICENTRE CLINICAL TRIAL Jufie Weston, Mary Hannah and Arne Ohlsson Women's College Hospital Toronto, Canada Objective: To outline the process of changing the trial design during the recruitment phase of a mnlticontre clinical trial and to outline problems encountered. Material: The Term PROM Study was initially designed to compare induction of labour with expectant management in pregnant patients with prelabour rupture of membranes at term. Patients in the induction group received either intravenous oxytocin or vaginal prostaglandin to induce labour. Patients in the expectant group awaited spontaneous labour unless induction was indicated for medical reasons. Al~r recruitment into the trial had begun information became available suggesting that there may be differences in neonatal infection, caesarian section rates and maternal satisfaction between the two induction methods. Results: Following discussions with centre personnel, MRC and experts outside the trial, changes were felt to be necessary to create four study groups which would allow separate analysis of the two induction methods. These changes to the study design were made with a resulting increase in sample size to 5000 women. Conclusions: Many difficulties are encountered when the trial design is changed after recruitment has begun. However, after new information becomes available changes may be necessary in order to maintain the trial's clinical relevance. A07 DATA MANAGEMENT WORKLOAD ASSESSMENT FOR MEDICAL ONCOLOGY B. Troehanowski, C. Houston and C. Begg Memorial Sloan-Kettering Cancer Center New York, New York Investigators at our cancer center prepare many grant and contract proposals throughout the year. In order to improve our ability to project data management needs, we have developed a formula for assessing workload. This formula permits the estimation of the l?ercentage effort of a full time employee required to complete the data management activities. This ts of course a difficult task and we began by analyzing the work components of a medical oncology dsts manager's responsibilities. This includes a number of distinct phases as patients follow their treatment course. They are: accrual of new patients, active follow-up, and off-study follow-up. Each phase of a patient's treatment is assigned a multiplier based on how many hours per month the activity should require. However, these

Previous pregnancy outcomes and subsequent risk of preterm rupture of amniotic sac membranes.

To test the hypothesis that previous unfavourable pregnancy outcomes increase the risk for premature birth, with (PP) or without (PTB) premature rupture of the amniotic sac (PROM) at the index pregnancy and that multiple undesirable outcomes increase risk. Case control study. Two university hospitals. Four hundred sixty-three women aged 15 to 45 years who were delivered preterm PROM, full term PROM and preterm without PROM matched with 463 women who delivered full term. All women included in the study had at least one previous pregnancy. Odds ratio of previous adverse pregnancy outcome among index cases. Compared with controls, PP cases had odds ratios of 95 for previous preterm birth, 186 for abortion and prematurity and 158 for fetal loss, abortion/prematurity after controlling for confounding variables. Compared with controls, PTB cases had an odds ratio of 96.5 for previous preterm delivery, 84 for abortion and prematurity, and 320 for fetal loss/abortion and prematurity after controlling for confounding variables. Previous preterm delivery, abortion and prematurity and fetal loss/abortion and prematurity all increase risk for subsequent preterm birth with or without PROM.