Another look at the Dalkon Shield: Meta-analysis underscores its problems

Abstract

Numerous non-comparative clinical trials of the Dalkon Shield appear to provide ample evidence that the Dalkon Shield was an effective IUD; they seem to yield little evidence that it was a dangerous device, nor markedly different from its contemporaries. Equating the performance of the Dalkon Shield IUD to that of the Lippes Loop or of Copper IUDs is erroneous, however, with respect to rates of pregnancy, expulsion, pelvic infection, septic abortion, death with the device in situ, and tubal infertility. Randomized studies show the Dalkon Shield had approximately double the pregnancy rates of the Lippes Loop D or Copper IUDs (P < .05) and a significantly lower expulsion rate. These differences coupled together with the distinguishing multifilament tail of the Dalkon Shield underlay the higher rates of pregnancy complications, including septic abortions with the device in situ and deaths, associated with the Shield. A five-fold increased risk of hospitalized pelvic infection among Dalkon Shield users found in the Women's Health Study resulted not from ascertainment bias, but was related to the fact that Dalkon Shield users had more severe hospitalized PID than did other hospitalized women with PID and IUD use. Following cessation of distribution and of use of the Dalkon Shield, and following the FDA's recommendation to remove IUDs in case of pregnancy, there have been no deaths reported among pregnant American women with an IUD in situ in a 5-year period. Neither the IUDs of today nor those in use during 970–974 are equatable to the Dalkon Shield.