Reanalysis of a multireported trial on home uterine activity monitoring

A randomized multicenter study of home uterine activity monitoring using the "Term Guard," is widely cited in support of the Term Guard monitoring system. Critical appraisal of the data in the six reports of the study showed that all reports scored poorly on avoidance of allocation, exclusion, and assessment bias. In all of them the numbers of women for whom outcomes were reported were different, and none provided information on infant outcomes. Comparison of analyses and subanalyses showed inconsistencies that defied the laws of arithmetic. Reanalysis of the data, as well as could be done from the available information, showed that use of the Term Guard system was compatible with both an increase and a decrease in the frequency of preterm delivery (95% CI +2.2% to -16.5%). It is concluded that the reports of the multicenter study contain such high potential for bias and so many internal inconsistencies that they should be dismissed entirely.