If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Human Reproduction Unit, Cruces University Hospital, Biocruces, University of the Basque CountryBiocruces Health Research Institute, University of the Basque Country, Bilbao, SpainInstituto Valenciano de Infertilidad, IVI Bilbao
Human Reproduction Unit, Cruces University Hospital, Biocruces, University of the Basque CountryBiocruces Health Research Institute, University of the Basque Country, Bilbao, Spain
The aim of this study was to ascertain the incidence of pelvic inflammatory disease (PID) after intrauterine insemination (IUI). A systematic review was conducted using three different approaches: a search of IUI registries; a search of published meta-analyses; and a search of prospective randomized trials. Search terms were ‘IUI’, ‘complications’, ‘infection’ and ‘PID’. Two IUI registers were identified that met the inclusion criteria, totalling 365,874 cycles, with 57 PID cases being reported. The post-IUI PID rate was 0.16/1000 (95% CI 0.2 to 0.3/1000). The frequency was higher in husband sperm cycles (0.21/1000) (28/135,839) than in donor sperm cycles (0.03/1000) (1/33,712) (P < 0.05; OR 6.95). Nineteen meta-analyses were retrieved, which included 156 trials, totalling 43,048 cycles, with no PID case being reported. Seventeen prospective clinical trials published between 2013 and 2014 were identified, totalling 4968 cycles; no PID case was reported. The reported rate of post-IUI clinical PID is low (0.16/1000), about 40% higher than reported in the general population of women during their reproductive life. No antibiotic prophylaxis should be recommended unless there is an associated risk factor.
Intrauterine insemination (IUI) is a widely used procedure. According to the European Society of Human Reproduction and Embryology register, more than 200,000 cycles are carried out in Europe every year, and about 175,000 of them correspond to the sperm of the husband or partner (
European IVF-Monitoring Consortium (EIM) for the European Society of Human Reproduction and Embryology (ESHRE) Assisted reproductive technology in Europe, 2012: results generated from European registers by ESHRE.
). Common complications of IUI include multiple pregnancy and ovarian hyperstimulation syndrome. Pelvic inflammatory disease (PID) has been reported as a rare complication of IUI, but no data about its expected frequency are available.
Severe complications are associated with PID, and some of them are especially relevant in women with unfulfilled fertility, such as tubal damage, tubo-ovarian abscess and even ovariectomy. About 15% of PID occurs after procedures that break the cervical mucus barrier (
). On occasions, PID may develop after certain procedures, such as hysterosalpingography, hysteroscopy, endometrial biopsy, oocyte recovery or even embryo transfer (
In some cases of IUI, the introduction of the catheter through the cervical canal and the release of washed sperm could prompt the ascension and spread of cervical microorganisms into the uterus and the tubes. It is well known that ascension of microorganisms is the main mechanism of PID genesis (
cite a frequency close to 1/500. The clinical impression, however, is that the frequency is much lower.
The occurrence of PID after IUI is a rare complication, and no reliable data on its frequency have been published; therefore, it is difficult to give exact information to patients, and to make decisions about prophylactic antibiotic therapy.
The aim of our study was to determine the frequency of PID after IUI on the basis of a literature search of large case studies, using three complementary analyses: data from IUI registers where PID is considered; meta-analyses of IUI reporting on complications; and all IUI prospective studies published between 2013 and 2014. Finally, we compare them with our centre's IUI complication register.
Materials and methods
A literature review was conducted on the reported frequency of PID, in relation to the number of cycles carried out and the use of husband or donor sperm. Three different literature searches were conducted in accordance with PRISMA and MOOSE guidelines (
Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis of observational studies in epidemiology (MOOSE) group.
We reviewed national and international registers on IUI, and identified cases in which PID had occurred after IUI. The search strategy was carried out in PubMed with the combination of any of the following MeSH descriptors: ‘ART register,’ ‘IUI register,’ ‘intrauterine insemination plus complications,’ ‘intrauterine insemination plus infection’ and ‘artificial insemination plus infection’. The time period for the search was 1965–2015. Only human reports were considered. The abstracts were reviewed by one investigator (KR), who reviewed full-text articles when needed. From that search, only one register was obtained, from California Cryobank (1986–2004) (
). A similar search was conducted in Google, both in English and Spanish, and the register of the Spanish Fertility Society (2002–2013) was obtained (Table 1).
Table 1Pelvic inflammatory disease after intrauterine insemination rate resulting from the published register analysis.
Sources: Spanish Fertility Society 2002–2013 and California Cryobank, 1986–2004.
The California Cryobank is a privately owned sperm bank established in 1977 that provides frozen donor semen worldwide. Physicians and recipients were asked to report any adverse outcomes resulting from insemination as well as pregnancy outcome. Most of the infection reports were made verbally by either the recipient or the insemination centre. Most reports were available days to weeks after the insemination (
, and each year includes IVF and IUI results of Spanish fertility centres. At that time, data could not be linked to the corresponding centres. The definition of PID was based on symptoms that required hospitalization. No reference was made to antibiotic prophylaxis in any of the registers.
Review of meta-analyses
A PubMed search was conducted using the key words ‘IUI meta-analysis’ from the period 2004–2014 (Table 2 and Supplementary Table S1). From the 50 meta-analyses retrieved, those not addressing IUI, or focusing on HIV, endometriosis or tubal patency, were excluded. Finally, 20 meta-analyses remained, which included 191 individual papers. One of the 20 reports on meta-analysis retrieved (
The prognostic profile of subfertile couples and treatment outcome after expectant management, intrauterine insemination and in vitro fertilisation: a study protocol for the meta-analysis of individual patient data.
) was excluded, as it corresponded to the design of a meta-analysis and not to one that had already been conducted (Figure 1). From the 191 remaining articles, 35 of these were excluded because their population was at least partially considered in more than one meta-analysis. Such cases were included only once; if a discrepancy was identified between the reported numbers of cycles, the publication that reported a higher number of cycles was included. No reference was made to antibiotic prophylaxis in any of the meta-analyses.
Table 2Pelvic inflammatory disease after intrauterine insemination resulting from the review of meta-analysis.
Effectiveness of GnRH antagonist in the management of subfertile couples undergoing controlled ovarian stimulation and intrauterine insemination: a meta-analysis.
The prognostic profile of subfertile couples and treatment outcome after expectant management, intrauterine insemination and in vitro fertilisation: a study protocol for the meta-analysis of individual patient data.
Ovarian stimulation protocols (anti-oestrogens, gonadotrophins with and without GnRH agonists/antagonists) for intrauterine insemination (IUI) in women with subfertility.
Human chorionic gonadotropin administration vs. luteinizing monitoring for intrauterine insemination timing, after administration of clomiphene citrate: a meta-analysis.
A PubMed search was carried out using the key words ‘IUI’ or ‘intrauterine insemination’. The search period was restricted to 2013–2014 (Table 3). Only randomized studies conducted on humans were considered. Studies not focusing on clinical outcomes were excluded. Eighteen studies were obtained. One of them (
An effective alternative to only gonadotrophin for controlled ovarian stimulation in unexplained infertility patients undergoing intra-uterine insemination: a clinical trial.
) was not included, as the original work could not be accessed, and the first author could not be contacted (Figure 2). No mention was found of antibiotic prophylaxis.
Table 3Pelvic inflammatory disease rate analysis in the intrauterine insemination prospective randomized trials in the period 2013–2014.
When is clomiphene or gonadotropin intrauterine insemination futile? Results of the Fast Track and Standard Treatment Trial and the Forty and Over Treatment Trial, two prospective randomized controlled trials.
The gradient technique improves success rates in intrauterine insemination cycles of unexplained subfertile couples when compared to swim up technique; a prospective randomized study.
An effective alternative to only gonadotrophin for controlled ovarian stimulation in unexplained infertility patients undergoing intra-uterine insemination: a clinical trial.
A total of 4968 cycles were included in the analysis after exclusion of 118 cycles reported in Chaudhury et al. (2013), as the study could not be accessed.
Effect of luteal phase support with vaginal progesterone in intrauterine insemination cycles with regard to follicular response: a prospective randomized study.
Advantages of recombinant follicle-stimulating hormone over human menopausal gonadotropin for ovarian stimulation in intrauterine insemination: a randomized clinical trial in unexplained infertility.
Eur. J. Obstet. Gynecol. Reprod. Biol.2013; 169: 244-247
A total of 4968 cycles were included in the analysis after exclusion of 118 cycles reported in Chaudhury et al. (2013), as the study could not be accessed.
An effective alternative to only gonadotrophin for controlled ovarian stimulation in unexplained infertility patients undergoing intra-uterine insemination: a clinical trial.
We reviewed our centre's IUI complication register during 1993–2016, finding 12,720 IUI cycles (8398 with husband sperm and 4322 with donor sperm). Most cycles were stimulated with gonadotrophin, and only one insemination was carried out per cycle. All patients had tubal patency assessed by hysterosalpingography. Antibiotic prophylaxis was never administered. Women received six inseminations, unless pregnancy was achieved before. The definition of PID was presence of clinical symptoms requiring hospitalization and antibiotic therapy. Data corresponding to 2002–2013 were already included in the Spanish Fertility Society register.
Results
Analysis of retrospective registers
The first register corresponded to 48,000 cycles from California Cryobank, resulting in a post-IUI PID rate of 0.17/1000 (8/48,000) (Table 1). Donor sperm were used in all these IUI.
The Spanish Fertility Society register included 317,874 IUI cycles: 135,839 corresponded to husband sperm and 33,712 to donor sperm, and for the remaining cycles, in the latest years, no differentiation was made between donor and husband sperm.
The combined analysis of the two registers resulted in a PID rate of 0.16/1000 cycles (57/365,874) (95% CI 0.2 to 0.3/1000). The frequency was 0.21/1000 (28/135,839) in cycles with the husband's sperm compared with 0.11/1000 in donor sperm cycles (9/81,712) (Yates chi squared = 2.23, non-significant). The relative risk was 1.87 (95% CI 0.88 to 3.97).
Given that the technique has changed significantly over time, and that there was a remarkable heterogeneity in the period covered by the two registers (1986–2004 for the California and 2002–2013 for the Spanish one), a second analysis was conducted, which was restricted to the Spanish register for the period 2002–2008 (as data were not subsequently subdivided into donor and husband sperm cycles).
The PID rate for husband sperm cycles was 0.21/1000 (28/135,839), which is significantly higher than the value of 0.03/1000 (1/33,712) in donor sperm cycles (Yates chi-squared = 3.94; P = 0.047). The relative risk was 6.95 (95% CI 0.95 to 51.06).
Ovarian stimulation protocols (anti-oestrogens, gonadotrophins with and without GnRH agonists/antagonists) for intrauterine insemination (IUI) in women with subfertility.
Human chorionic gonadotropin administration vs. luteinizing monitoring for intrauterine insemination timing, after administration of clomiphene citrate: a meta-analysis.
Effectiveness of GnRH antagonist in the management of subfertile couples undergoing controlled ovarian stimulation and intrauterine insemination: a meta-analysis.
) (Table 2 and Supplementary Table S1), resulting initially in 48,960 IUI cycles. A total of 5912 cycles, however, had been included in more than one meta-analysis, so the final analysis was restricted to 43,048 cycles. None of them reported on PID. In four cycles, no reference was made to any kind of complication (
). In the remaining 15, the complications described were multiple pregnancy, ectopic pregnancies, miscarriages and hyperstimulation syndrome. Therefore, the reported PID rate was 0/1000 (0/43,048) (95% CI 0 to 0.1/1000).
Review of prospective studies published between 2013 and 2014
For the period 2013–2014 (Table 3), we reviewed 17 reports (
The gradient technique improves success rates in intrauterine insemination cycles of unexplained subfertile couples when compared to swim up technique; a prospective randomized study.
When is clomiphene or gonadotropin intrauterine insemination futile? Results of the Fast Track and Standard Treatment Trial and the Forty and Over Treatment Trial, two prospective randomized controlled trials.
Advantages of recombinant follicle-stimulating hormone over human menopausal gonadotropin for ovarian stimulation in intrauterine insemination: a randomized clinical trial in unexplained infertility.
Eur. J. Obstet. Gynecol. Reprod. Biol.2013; 169: 244-247
Effect of luteal phase support with vaginal progesterone in intrauterine insemination cycles with regard to follicular response: a prospective randomized study.
The gradient technique improves success rates in intrauterine insemination cycles of unexplained subfertile couples when compared to swim up technique; a prospective randomized study.
), was the infection rate specifically mentioned (‘There were no cases of post-procedural infection’). In the remaining reports, a number of complications were addressed (multiple pregnancy, miscarriages, ectopic pregnancy, hyperstimulation syndrome), but no mention was made of PID. Therefore, the reported PID rate was 0/1000 (0/4968) (95% CI 0 to 0.7/1000).
Review of our center's records
Two cases of PID after IUI were reported, representing 1.6/1000 (2/12,720) (95% CI 0 to 0.6/1000). Both these cases occurred in husband sperm cycles (2/8398 = 0.24/1000) (95% CI 0 to 0.9/1000) with no case with donor sperm (0/4322) (95% CI 0 to 0.9/1000).
Discussion
The most relevant pathogenic mechanism in PID is the ascending spread of microorganisms from the vagina and cervix into the upper genital tract. Hypothetically, IUI could promote PID as it sweeps along the microorganisms from the cervix. The introduction of the catheter could act as a foreign body in the uterine cavity, and the deposit of processed sperm in the uterine cavity could also cause PID. A similar mechanism has been proposed for PID cases after hysterosalpingography, endometrial biopsy, curettage or IUD insertion.
). These reports, however, are based on relatively small populations and lack statistical power. Indeed, one cannot discount publication bias, as a study on complication rates, in which a relatively high complication rate is found, has more chances of being published than those not reporting any complications at all.
Hence, it is important that reviews analyse a large number of cases. Moreover, we estimated frequency using three different methods: analysis of two state/national retrospective registers; analysis of 19 prospective meta-analyses; and analysis of 17 individual prospective reports. Finally, we compared the obtained frequency with our centre's data.
From our ‘state register’ analysis, with more than 350,000 IUI cycles, the PID rate was 1.6/10,000, much lower than that usually reported. In this analysis, one cannot rule out an underreporting rate, as in all retrospective studies.
In the present meta-analysis, the PID rate was expected to be reported, as all important harms or unintended effects in each group should be reported (
); however, no case of PID was reported in over 40,000 cycles. We cannot, however, rule out that PID occurred after IUI in some studies but was not reported as it did not pertain to the aim of the study. The resulting reported PID rate was 0/10,000, slightly lower than that obtained in our state register analysis. When we conducted a similar study on all the prospective single reports on IUI during a 2-year period, again, no case was reported. Therefore, we cannot rule out underreporting for these reasons. In any case, the upper limit of the confidence interval in both analyses was not too far from the confidence intervals from our register study.
Finally, from our own centre's IUI register, we obtained a PID rate of 1.6/10,000, exactly the same rate that we reported in the ‘retrospective register’ analysis. We also observed a non-significant trend to higher PID rates in husband sperm cycles.
In a recent US nationally representative population of sexually experienced women of reproductive age, the reported prevalence of women with a history of PID increased from 2.9% among women aged 18–24 years to 6.7% among women aged 40–44 years (
). Therefore, in the general population, the expected per month rate should be that difference (3.8%) divided by 338 (13 cycles per year multiplied by 26 years, the difference between the maximum and the minimum age). The resulting per month PID rate would be 1.12/10,000, which does not differ significantly from the value of 1.6/10,000 from our register population (44/395,798 (
and the population of infertile women in the present study differ remarkably. Infertile women probably have more undiagnosed tubal conditions, less current use of barrier contraceptive methods and a lower number of current partners. In comparing both figures, it could be stated that PID after IUI is roughly 40% higher than in a normal cycle in the unselected population. Our report does not address subclinical PID, of which the repercussions are difficult to quantify. Subclinical PID was not addressed in the WWM report (
In the register analysis, when the sperm source was considered, a non-significant trend to higher PID rates was observed in husband sperm insemination (2/10,000) than in donor insemination (1.1/10,000). Between 1986 and 2001, all the cases corresponded to donor cycles, demonstrating remarkable chronologic heterogeneity. Therefore, when the analysis was restricted to the period 2002–2008 (the only time period were the exact number of donor and husband sperm cycles could be ascertained), the PID rate in husband sperm cycles (2/10,000) was significantly higher than in donor sperm cycles (0.3/10,000), with a relative risk of 6.95 (95% CI 0.95 to 51.06). It was not, however, possible to ascertain from our analysis whether this was caused by a reduced risk linked with donor sperm (microbiological screening of donors or perhaps due to the freezing procedure) or with a different profile among women undergoing husband or donor sperm insemination. We could speculate that women undergoing husband sperm insemination could have a higher proportion of mild tubal factor, that could facilitate PID. Indeed, a number of women receiving donor sperm are either single or lesbians and perhaps have a different cervicovaginal microbiological status. In this context, it has been reported that sexually transmitted diseases are transmitted more efficiently during heterosexual intercourse, than during lesbian intercourse (
). This could also be attributed to a reporting bias by a commercial cryobank.
In conclusion, on the basis of our analysis, the reported PID rate after IUI is low, close to 1.6/10,000 cycles, roughly 40% higher than in one ovarian cycle in the unselected population. Prophylactic antibiotic therapy is not recommended, unless there is a pre-existing risk factor.
Appendix. Supplementary material
The following is the supplementary data to this article:
Ovarian stimulation protocols (anti-oestrogens, gonadotrophins with and without GnRH agonists/antagonists) for intrauterine insemination (IUI) in women with subfertility.
An effective alternative to only gonadotrophin for controlled ovarian stimulation in unexplained infertility patients undergoing intra-uterine insemination: a clinical trial.
The gradient technique improves success rates in intrauterine insemination cycles of unexplained subfertile couples when compared to swim up technique; a prospective randomized study.
When is clomiphene or gonadotropin intrauterine insemination futile? Results of the Fast Track and Standard Treatment Trial and the Forty and Over Treatment Trial, two prospective randomized controlled trials.
Human chorionic gonadotropin administration vs. luteinizing monitoring for intrauterine insemination timing, after administration of clomiphene citrate: a meta-analysis.
Effectiveness of GnRH antagonist in the management of subfertile couples undergoing controlled ovarian stimulation and intrauterine insemination: a meta-analysis.
Advantages of recombinant follicle-stimulating hormone over human menopausal gonadotropin for ovarian stimulation in intrauterine insemination: a randomized clinical trial in unexplained infertility.
Eur. J. Obstet. Gynecol. Reprod. Biol.2013; 169: 244-247
Effect of luteal phase support with vaginal progesterone in intrauterine insemination cycles with regard to follicular response: a prospective randomized study.
Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis of observational studies in epidemiology (MOOSE) group.
The prognostic profile of subfertile couples and treatment outcome after expectant management, intrauterine insemination and in vitro fertilisation: a study protocol for the meta-analysis of individual patient data.
Roberto Matorras, M.D. is Professor and Head of Human Reproduction Unit at the Department of Obstetrics and Gynecology of Cruces Hospital, Vizcaya, Spain. He is involved with several societies, has arranged national and international conferences and has authored numerous books and publications.
Key message
The frequency of pelvic inflammatory disease after intrauterine insemination is low. Prophylactic antibiotic therapy is not recommended, unless the patient has a pre-existing risk factor.
Article info
Publication history
Published online: November 24, 2017
Accepted:
November 8,
2017
Received in revised form:
November 3,
2017
Received:
May 23,
2017
Declaration: The authors report no commercial or financial conflicts of interest.
30620-X&id=rbmo1851-fig-0001.jpg){kind=link}
30620-X&id=rbmo1851-fig-0002.jpg){kind=link}