Was the Dalkon Shield a safe and effective intrauterine device? The conflict between case-control and clinical trial study findings.

OBJECTIVE: To compare the findings of the case-control and cohort studies used to indict the Dalkon Shield (A.H. Robins Company, Inc., Richmond, VA) with the findings of the Dalkon Shield clinical trials. DATA IDENTIFICATION: All published reports on the Dalkon Shield were identified through MEDLARS system (United States National Library of Medicine) searches and by cross checking all references in these reports. The same approach was used to identify all case-control and cohort studies of the purported relationship between intrauterine devices (IUDs) and pelvic inflammatory disease (PID). STUDY SELECTION: Only studies of interval patients that included 50 or more women and 6 or more months of follow up that computed standard IUD event rates (rates of pregnancy and expulsion and removal for pain and bleeding) were selected for this study. All case-control and cohort studies identified were included except two case-control studies that included women with sterile chronic salpingitis. RESULTS: The 16 case-control and 2 cohort studies found or suggested that the Dalkon Shield increased the risk of PID. The 71 clinical trials of the Dalkon Shield show that when this device is inserted by an experienced clinician it is a safe and effective contraceptive method, comparable with other IUDs used at the time. There was no evidence of an increased risk of PID found in these clinical trials. CONCLUSIONS: This study offers convincing evidence that the indictment of the Dalkon Shield was a mistake. Additionally, this study shows that physician skill and experience is far more important to successful IUD insertion than previously recognized, a finding with considerable implications for IUD study designs and for marketing strategies.     

True and false regarding the risk of infection in contraception

Contraception is certainly less responsible for the recrudescence of venereal diseases than are other social factors, such as the increased sexual freedom among adolescents. The condom and all topical spermicidal agents provide a valuable barrier against the risk of venereal infections. Genital infection, especially moniliasis, is more frequent among oral contraception (OC) users than among nonusers. The exact etiology of this phenomenon is not known; it is probable that OC exercises an influence on vaginal pH, which varies between 3-4.5 in OC users, and between 5.5-6.5 among nonusers. Risk of infection in IUD users is increased by prior genital and pelvic infections, which remain an absolute contraindication to IUD use. Copper IUDs have the lowest rate of infection, and the Dalkon Shield has the highest; rate of infection is 2 times higher in wearers for over 5 years. Since pelvic infection can have a negative influence on tubal permeability, IUD use is not recommended for nulliparous women.     

Operative laparoscopy: time for acceptance

This reassessment of IUDs as a risk factor for PID contrasts study results from prospective and case-control studies. Except for a transient increased risk after IUD insertions, increased risk is seen mainly in case-control studies after 1973, the time of a major PID epidemic and adverse publicity related to the Dalkon Shield. Removal of oral contraceptive users from case-control studies generally reduces their reported odds ratios, frequently to borderline statistical significance for studies with high accuracy of PID diagnosis. The remaining elevated odds ratios likely relate to diagnostic bias as primary care clinicians are more likely to refer patients wearing an IUD to referral centers where case-control studies are conducted. Biased case-control studies provided evidence for expensive litigation, resulting in removal of IUDs from the American market. The Food and Drug Administration requirements for postmarketing surveillance are inadequate to provide prospective data to avoid this kind of loss.     

The intrauterine device, pelvic inflammatory disease, and infertility: the confusion between hypothesis and knowledge.

An earlier editorial made many false claims against the IUD and women. In many cases, the author extrapolated experience from the Dalkon Shield to today's IUDs. He even said the modern contraception has not reduced fertility, but data from at least 7 developing countries clearly refute that statement. Moreover he claims that the most female ambition is motherhood. Thus he blamed women for high fertility rates without acknowledging that women in developing countries have little control over fertility. Husbands determine when and how often to have intercourse and whether to use contraceptives or not. Women do not always have access to contraceptives. The author boldly stated that the IUD poses a threat to future fertility because it causes pelvic infections. The real threat is intercourse by which upper genital tract infections are transmitted. It is these sexually transmitted diseases that cause pelvic infections. The studies the author quoted have been found to have considerable bias and/or methodologic flaws. Some studies used as the comparison group women using contraceptive which lower the pelvic infection risk. Contemporary studies find no increased risk of infertility in copper and hormone IUD users, but instead show increased risk for multiple sex partners. Some studies do indicate an increased risk shortly after insertion which is due to insertion instrumentation used. The author even claimed that a Swedish study concluded that IUDs promote pelvic infection in patients with gonorrhea and chlamydia. But the Swedish study found no difference in pelvic infection risk between IUD and non-IUD, nonhormonal contraceptive users. Based on hundreds of millions of woman-years of worldwide experience, WHO claims modern copper and hormone releasing IUDs, when used as directed, may be the most effective and reliable reversible contraceptive method. This editorial based on flawed conclusions and misinformation does an injustice to the IUD and to women.     

Copper in the menstrual and physiological cycle cervical mucus of copper IUD users (author's transl)

The article studies the concentration of copper in the cervical mucus of users and nonusers of copper IUDs. Of the 170 samples taken 90 were basal, i.e. taken from nonusers, and 80 were taken from users of No-Gravid, Dalkon Shield, Gravigard, Copper T, and Progestasert. Concentration of copper was much higher in IUD wearers than in nonwearers, even after 2 years from insertion, except in wearers of Dalkon Shield and of Progestasert. Copper concentration in wearers of Gravigard and Copper T were lower than concentrations found in wearers of No-Gravid.     

Early postpartum insertion of an intrauterine device

Over a 2-year period in London, Dalkon Shield IUDs were inserted in 162 patients during the postpartum period, usually on the 5th day. To insert the IUD, a sterile bivalve speculum was inserted in the vagina and the cervix swabbed with .5% aqueous chlorhexidine. The anterior lip of the cervix was grasped with a sponge holding forceps and the depth of the uterine cavity was assessed with uterine sound. A Dalkon Shield was inserted at the fundus of the uterine cavity using a standard type carrier. The device was correctly placed in 86.4% of the 162 patients at the 6-week postnatal examination. 22 patients expelled the device completely or partially, 3 patients were diagnosed as being pregnant with the device in situ, and 1 patient was pregnant due to expulsion. At the first menstrual period, 32 women noted heavier bleeding and 2 patients complained of increased discomfort; however, there was no undue discomfort at subsequent menses. This pilot study assessed the practicability of establishing an IUD during a woman's stay on the postnatal ward. however, the rate with Lippes Loop and Saf-T-Coil was unacceptably high. The Dalkon Shield is no longer in use, but this type of IUD with its low expulsion characteristics may successfully be applied to early puerperium. A clinical trial of the Multi-load 250, similar to the Dalkon Shield, is commencing in the near future.     

An association between the Dalkon Shield and complicated pregnancies among women hospitalized for intrauterine contraceptive device--related disorders.

A nationwide mail survey of virtually all physicians likely to be involved with intrauterine contraception resulted in 3,502 unduplicated reports of intrauterine contraceptive device (IUD)--related hospitalizations during the first six months of 1973. Dalkon Shield use was significantly more frequent among women hospitalized for a complicated pregnancy than those hospitalized for a non-pregnancy-related disorder. Although the observed association was not substantially altered by stratifications of the mail survey reports by the patient's age, race, or geographical region, the association did not apply to those women whose IUD's were explicitly reported to be of the nulliparous size. Interviews conducted with a probability sample of physicians who had not responded to the survey confirmed that the association between the Dalkon Shield and complicated pregnancy also existed in their experience. An association between the standard Dalkon Shield and complicated pregnancies might reflect an increased rate of pregnancy with this device, an increased rate of complications occurring after zygotic implantation, or perhaps both. Whatever the explanation, the observed association is sufficiently widespread to require further investigations.     

The benefit of endometrial biopsy before IUD insertion

131 patients aged 25-35 followed from 1969-84 underwent endometrial biopsies under antibiotic cover preliminary to insertion of IUDs. 125 subsequently had IUDs inserted, including 10 Dalkon Shields, 50 Gravigards, and 66 copper 200, Gyne-T, or Nova T devices. All IUDs were inserted in the 1st 10 cycle days under a 3-day antibiotic cover. The 131 biopsies revealed 18 simple endometrial hyperplasias, 1 glandulo- cystic hyperplasia, and 5 endometrial polyps. The hyperplasias were treated with a progestin on cycle days 15-25 for 3-6 months and a curettage was performed for the polyps. No cases of endometritis were found in nulligestes. The 11 patients with latent endometritis were treated with antibiotics for 10 days and were vaccinated with an antipyogenic. A repeat endometrial biopsy was done to confirm disappearance of the endometritis before the IUD was inserted. The 126 IUD users were followed for 6 years on average. 2 cases of salpingitis and an adnexal reaction were observed. 2 of the 3 were treated with antibiotics and had no complications. The 3rd underwent tubal microsurgery after which she became pregnant. The IUD was removed in all 3 cases at the onset of symptoms. 1 of the 3 patients did not have a preliminary endometrial biopsy before the IUD was inserted. Despite the small sample size, it appears that a preliminary endometrial biopsy can reduce the infection rate among IUD users to that of the general female population.     

Long-term use of intrauterine contraceptive devices in a private practice

Clinical experience with 1504 insertions of intrauterine contraceptive devices (Lippes Loop, Dalkon Shield, Cu-7) in a private practice is evaluated. Expulsion rates and removal rates for bleeding and/or pain and personal reasons differed for the three types of devices. Pregnancy rates associated with use of the three types of devices were similar. Pelvic inflammatory disease rates were similar for Dalkon Shield and Cu-7 users but were higher than the rate for Lippes Loop users. Use of intrauterine contraceptive devices did not appear to compromise future fertility based on the experience of women who had the devices removed. The study shows that long-term IUD use (up to 5 years) provides a safe and effective method of contraception.     

Pelvic pain associated with an unsuspected intraomental intrauterine device placed 30 years earlier: a case report

BACKGROUND: Recent-onset chronic pelvic pain thought to originate from an intraomental Dalkon Shield intrauterine device (IUD) (A. H. Robbins, Inc., Richmond, Virginia) that had remained asymptomatic for over 30 years is unusual. Case reports and patient series suggest that intraomental IUDs remain asymptomatic for long periods of time and are usually not associated with chronic pelvic pain. CASE: A 49-year-old woman with an unrecollected history of Dalkon Shield placement 30 years earlier was evaluated after 6 months of pelvic pain. A negative workup and failure of conservative management prompted an abdominal hysterectomy with bilateral salpingooophorectomy, which failed to relieve the symptoms. Revaluation identified a mobile, intraabdominal Dalkon Shield. Real-time x-ray fluoroscopy and reverse Trendelenburg positioning were used to laparoscopically retrieve the IUD from the patient's benign-appearing omentum. The pelvic pain remained resolved for over 1 year after removal of the IUD. CONCLUSION: An intraomental IUD that remained asymptomatic for 30 years was the most likely source of chronic pelvic pain in this perimenopausal patient. Increased intraabdominal fat deposition associated with the perimenopause may have contributed to this patient's change in clinical status.     

Coitus during menses. Its effect on endometriosis and pelvic inflammatory disease

Four hundred ninety-eight consecutive new female patients in a private obstetrics-gynecology practice were asked about their coital habits during menses (SxMs/F = frequent, SxMs/O = occasional, SxMs/N = never) and history of pelvic inflammatory disease (PID). Endometriosis and tubal infertility factor were diagnosed on laparoscopy or laparotomy. The frequency of endometriosis was higher (P less than .05) in patients with coitus during menses (SxMs/F + O) (17.5%) than in the SxMs/N group (10.9%). The frequency of PID was not different between the SxMs/F + O (10.8%) and SxMs/N (9.7%) groups. Thirty-five percent of the infertility patients had a tubal infertility factor. The frequency of tubal infertility factors was higher in the PID group (78.4%) than in the non-PID group (30.4%) (P less than .001). Endometriosis was associated with an increase in tubal infertility factors (47.9% vs. 33.3% in the nonendometriosis group) but to a lesser degree than PID. The frequency of PID in the group with a tubal infertility factor was higher (22.7% vs. no tubal factor, 3.4%) (P less than .001), and the frequency of endometriosis was higher in the group with a tubal infertility factor (19.3% vs. no tubal factor, 11.5%) (P less than .05). Coitus during menses appeared to increase the chances of endometriosis but not of PID. The rate of tubal infertility factors increased in both PID and endometriosis.     

Selecting patients and inserting the Progestasert

In this article, the author discusses what is presently known about the Progestasert IUD, a progesterone delivery system. Progestasert delivers progesterone on a daily basis and in continuous fashion into the endometric cavity. It has no systemic effects, although there is increasing evidence of its effects on hypermenorrhea na dysmenorrhea. Evidence suggests that Progestasert may avoid one of the major problems associated with former IUDs (particularly the Dalkon Shield): salpingitis. More than 10 years ago, physicians observed during sterilization procedures that a large number of women wearing inert or copper IUDs had salpingitis. Data from 27 Progestasert users show that none had salpingitis at the time of sterilization. Progestasert may also carry a lower risk of pelvic inflammatory disease associated with IUD use. A study in Helsinki found only 3 cases of pelvic inflammatory disease in 20,000 women years use of progestin (Levonorgestrel) impregnated IUDs, compared to the expected incidence (according to the Center for Disease Control) of 1/100 woman years. Additionally, in a study involving over 300 patients, the use of prophylactic antibiotics after insertion twice a day for 5 days and inserting Progestasert during the time ovulation rather at the time of menstruation produced no cases of the pelvic inflammatory conditions usually associated with the insertion procedure. Concerning the risks of ectopic pregnancy, evidence suggests that progestasert tends towards more ectopics when compared to the copper IUDs, but towards less when compared to not using any contraception. Although Progestasert is not contraindicated in nulliparous patients, caution should be taken with such patients.     

Fertility after contraception or abortion

There is a very small correlation, if any, between the prior use of OCs and congenital malformations, including Down's syndrome. There are few, if any, recent reports on masculinization of a female fetus born to a mother who took an OC containing 1 mg of a progestogen during early pregnancy. However, patients suspected of being pregnant and who are desirous of continuing that pregnancy should not continue to take OCs, nor should progestogen withdrawal pregnancy tests be used. Concern still exists regarding the occurrence of congenital abnormalities in babies born to such women. The incidence of postoperative infection after first trimester therapeutic abortion in this country is low. However, increasing numbers of women are undergoing repeated pregnancy terminations, and their risk for subsequent pelvic infections may be multiplied with each succeeding abortion. The incidence of prematurity due to cervical incompetence or surgical infertility after first trimester pregnancy terminations is not increased significantly. Asherman's syndrome may occur after septic therapeutic abortion. The pregnancy rate after treatment of this syndrome is low. The return of menses and the achievement of a pregnancy may be slightly delayed after OCs are discontinued, but the fertility rate is within the normal range by 1 year. The incidence of postpill amenorrhea of greater than 6 months' duration is probably less than 1%. The occurrence of the syndrome does not seem to be related to length of use or type of pill. Patients with prior normal menses as well as those with menstrual abnormalities before use of OCs may develop this syndrome. Patients with normal estrogen and gonadotropin levels usually respond with return of menses and ovulation when treated with clomiphene. The rate for achievement of pregnancy is much lower than that for patients with spontaneous return of menses. The criteria for defining PID or for categorizing its severity are diverse. The incidence of PID is higher among IUD users than among patients taking OCs or using a barrier method. The excess risk of PID among IUD users, with the exception of the first few months after insertion, is related to sexually transmitted diseases and not the IUD. Women with no risk factors for sexually transmitted diseases have little increased risk of PID or infertility associated with IUD use. There appears to be no increased risk of congenital anomalies, altered sex ratio, or early pregnancy loss among spermicide users. All present methods of contraception entail some risk to the patient. The risk of imparied future fertility with the use of any method appears to be low.(ABSTRACT TRUNCATED AT 400 WORDS)     

Postmenopausal bleeding secondary to a Dalkon Shield retained for 33 years: a case report

BACKGROUND: Although the Dalkon Shield intrauterine device (IUD) was removed from the market in the early 1970s, some women currently in their late 50s and early 60s may still have the Shield in their uteri. An IUD retained for a long time is very rare but can be a cause of postmenopausal bleeding. CASE: A 58-year-old, postmenopausal woman presented with vaginal bleeding and spotting of 8 months' duration. Careful investigation of her history revealed that she had undergone IUD placement 33 years earlier. She was not sure about the type and whether the IUD was removed or expelled subsequently. There was no string visualized in the vagina. An office endometrial biopsy revealed Actinomyces and an associated polymorphonuclear cell exudate. A retained Dalkon Shield was diagnosed and removed hysteroscopically. CONCLUSION: Providers should include "retained IUD" in the differential diagnosis of postmenopausal bleeding. In addition to endometrial biopsy, hysteroscopic evaluation of the uterine cavity is a valuable tool in the diagnosis and treatment of such bleeding. A retained IUD may cause postmenopausal bleeding secondary to chronic endometritis or migration into the uterine wall. Hysteroscopic dislodgement and removal may be challenging due to chronic embedding within the uterine wall.     

Unilateral tubo-ovarian abscess and intrauterine contraceptive devices.

The association of unilateral tubo-ovarian abscess and the presence or use of an intrauterine contraceptive device (IUD) appears to be a definite clinical entity. Four cases of unilateral tubo-ovarian abscess in patients using the IUD are presented. Three patients had a Dalkon Shield IUD and one had a Lippes Loop. Two patients had unilateral salpingo-oophorectomy while the other 2 had total abdominal hysterectomy and bilateral salpingo-oophorectomy. The differential diagnosis, possible etiology, route and mode of infection, and management are discussed.     

Intrauterine contraceptive devices and risk of pelvic inflammatory disease: standard of care in high STI prevalence settings

The intrauterine contraceptive device (IUD) is highly effective and cost-effective. IUD use is limited in some regions, however, due to concerns about increased risk of pelvic inflammatory disease (PID) and subsequent complications such as infertility and ectopic pregnancy. Recent reviews suggest that the overall risk of PID with modern IUDs is lower than previously thought, at least in regions with a low prevalence of sexually transmitted infections (STIs). Risk of PID may be higher, however, in places where gonorrhoea and chlamydia are prevalent, where screening for STIs is limited and where aseptic conditions for insertion are difficult to ensure. A World Health Organization multi-centre study and other studies have confirmed regional differences in STI prevalence, and the WHO study established that PID risk is temporally related to IUD insertion procedures. Studies of the effectiveness of antibiotic prophylaxis to prevent infectious complications are inconclusive due at least in part to use of sub-therapeutic regimens for pathogens commonly implicated in PID. In summary, the IUD can be safe and effective if inserted under aseptic conditions in women free of cervical infection. Further study is needed to define appropriate standards of care for IUD insertion where STI prevalence is high and ability to rule out infection is limited. Even with safe insertion, IUD promotion in areas of high STI/HIV prevalence must address women's needs for dual protection from infection and unwanted pregnancy.     

Unilateral tuboovarian abscess: A distinct entity

In a 30 month interval at the Los Angeles County-University of Southern California Medical Center, 85 patients had tuboovarian abscesses removed that were unrelated to complications of pregnancy. Thirty-seven patients (44 per cent) of the 85 had unilateral abscesses. Twenty patients (54 per cent) of the 37 patients with a unilateral abscess were using an intrauterine contraceptive device (IUD). In 13 (65 per cent) of the 20 women using an IUD, the foreign body was a Dalkon Shield. Unilateral pelvic abscesses can occur with or without the presence of an IUD.     

Tubal infertility in relation to prior induced abortion

One hundred twenty-seven women who had been given diagnoses of tubal infertility between 1979 and 1981 in King County, Washington, yet previously had been pregnant, were interviewed to determine their prior history of legally induced abortion. Their responses were compared with those of 395 women who conceived a child at the same time the infertile women began their unsuccessful attempt to become pregnant. In making the comparison, we adjusted for the effects of variables that in this population were related both to having an induced abortion and to the occurrence of infertility, i.e., age, number of prior pregnancies, number of sexual partners, cigarette smoking habits, Dalkon Shield (A. H. Robins Company, Richmond, VA) use, and whether the woman worked outside the home. The risk of tubal infertility in women who had had an induced abortion was not increased above that of other women (relative risk, 1.15; 95% confidence interval, 0.70 to 1.89). For women with two or more abortions, the relative risk was 1.29 (95% confidence interval, 0.39 to 4.20). When only the most recent pregnancy was considered, the relative risk was 1.19 (95% confidence interval, 0.72 to 1.97). Our results suggest that legal abortion, as performed during the past decade in the United States, does not carry an excess risk for future tubal infertility.     

Intra-uterine contraceptive devices (IUDs) in relation to uterine histology and microbiology

Histological and microbiological examinations were made of 90 uteri taken in association with abdominal hysterectomy. Thirty-one women had an intra-uterine contraceptive device (IUD) in utero at the time of operation, 16 women had previously used an IUD, while 43 had never used one. A chronic inflammatory reaction was seen significantly more often in the fallopian tubes of women with an IUD in situ (21/31, 68%) than in those who had never used an IUD (7/43, 16%). In contrast, non-active post-inflammatory tubal changes were found less frequently in women with an IUD in situ (6%) than in women who had never used an IUD (16%). All tubes were sterile, expect for two with apathogenic microbes. An endometrial inflammatory reaction was found in 4 women with an IUD in situ and in one woman who had never used an IUD. The possibility of tubal infertility must be borne in mind when planning IUD contraception.     

Correlation between use of IUD and pelvic inflammatory disease (PID) as independent variables and risk of occurrence of ectopic pregnancy

The relationship between IUD and ectopic pregnancy was evaluated in a study involving 189 controls and 69 women with ectopic pregnancy. The patients involved were between the ages of 18 and 45, sexually active, without previous tubal ligation; they were interviewed about previous pregnancies, diseases, abortions and their socio-economic status. Diagnosis of ectopic pregnancy was made using serum pregnancy tests, needle biopsies of the Douglas pouch and/or ultrasound and/or laparoscope. The results show that there is no statistical difference between patients using the IUD with a history of previous PID and those with previous PID who had never used an IUD; instead, there is a 10-fold risk in women affected by PID, regardless if IUD use. Women who had used an IUD for less than 3 years showed a smaller risk than those who had used the IUD for a greater period of time. Finally there was no difference in women not affected by PID, regardless of IUD use. This ultimately correlates with the notion that one of the mechanisms of action of the IUD consists of an endometrial inflammation as seen in foreign-body reactions.