Included in this study were female patients aged 21–45 years who applied to our family planning outpatient clinic between January 2021 to June 2021 for control visits or attending with any complaints. We only included patients who were referred to our family outpatient clinic for IUD insertion earlier and then attended their follow-up visits. Follow-up visits were scheduled in the 6th week, at 6 months, and then annually after the first year until the 5th year (Fig. 1). Patients were divided into two groups either coming to their routine control visit and having a normal IUD position or referring to our outpatient clinic with various complaints due to IUD usage and having a dislocated IUD during ultrasonographic evaluation. Exclusion criteria were: hospital admission within 6 weeks of insertion, IUD insertions outside of our hospital, space-occupying lesions (polyps, myomas), pelvic inflammatory diseases, cervical polyps, uterine myomas, adenomyosis, uterine anomalies, connective tissue diseases, descensus uteri, abnormal uterine bleeding, immune suppression, anticoagulant usage, copper allergy or copper metabolism disorders (Wilson disease), or previous treatment for Atypical Squamous Cells of Undetermined Significance (ASCUS), Low-grade squamous intraepithelial lesion (LSIL), High-grade squamous intraepithelial lesion (HSIL), Cervical intraepithelial neoplasia (CIN) lesions or malignancy by any method (cryotherapy, Loop electrosurgical excision procedure (LEEP) or cone biopsy).

Fig. 1.

A flow-chart diagram illustrating patient selection and scheduled visits during 5 years of period. IUD, Intrauterine device; PAP-Smear, Papanicolaousmear.

IUD insertions were performed by the same certified and trained midwife (SS) at the Family Planning Center, Health Sciences University Istanbul, Kanuni Sultan Suleyman Training and Research Hospital (Turkey). IUD insertions in our family planning outpatient clinic are performed on the 3rd day of the menstrual period, at the 6th week postpartum, and at the 6th week post-cesarean. On rare occasions we also inserted IUDs after menstrual regulation, after dilatation curettage, and after cesarean section. Patients were informed of the risks of IUD expulsion at each time period, and IUDs were initiated based on their personal preference. Gynecological examinations and The Papanicolaou-smear screening were performed prior to IUD insertion. After insertion, transvaginal ultrasonography (TV-USG) was performed to confirm the correct IUD placement. Follow-up visits were scheduled in the 6th week, at 6 months, and then annually after the first year. Speculum examination and TV-USG was performed on all patients at each follow-up visit. Patients were questioned for the presence of symptoms related to IUD use. Demographic characteristics, type of delivery, the last delivery time, insertion time, education status, gynecological symptom history (chronic pelvic pain, menorrhagia, dysmenorrhea, dyspareunia, leukorrhea), hygiene habits (vaginal douche), menstrual pattern, medical history and previous IUD displacement were documented. The presence of IUD thread, leukorrhea, vaginal bleeding, or foul-smelling discharge was investigated by examination of the speculum.

TV-USG was performed to ascertain the device position. Ultrasounds were performed vaginally using an ultrasound instrument (Aloka Prosound 6, 2008, Tokyo, Japan) with a 3.5 MHz vaginal transducer. Patients were examined in the lithotomy position and were asked to empty their bladder before examination. All measurements were performed by a single physician (OK) in order to ensure the standardization of ultrasonographic parameters and to minimize inter- and intra-observer variation. Measurements included the uterus position (anteversion, retroversion, anteflexion, retroflexion), cervical length, uterocervical angle (UCA), fundal length (FL), endometrial cavity length (ECL), the distance behind the intersection of the uppermost vertical arm of the IUD and the fundal endometrium (IUD-End), and uterine volume (cm${}^3$) (Fig. 2).

Fig. 2.

A schematic illustration of ultrasonographic uterine measurements. (A) Schematic illustration of the distance between the top of the intrauterine device (IUD) and the end of the endometrial cavity (IUD-End). (B) Schematic illustration of the ultrasonographic parameters defined for the uterus. A–B: IUD-fundus length (FL); B–C: endometrial cavity length (ECL); C–D: cervical length (CX); uterocervical angle (UCA); total uterine length (TUL): FL + ECL + CX [(A–B) + (B–C) + (C–D)].

The UCA is the angle measured in the triangular region between the anterior uterine segment and the cervical canal. For measurements made in the sagittal section, the first line forming the angle was drawn between the external os and the internal os along the endocervical canal. If the cervical canal was not linear, this first line was determined as the straight line drawn between the internal os and the external os. The secondary line forming the UCA was determined as the straight line extending from the beginning of the internal os to the top of the fundus along the anterior uterine segment [10].

Measurements of the uterine cervical longitudinal axis were performed after the image was enlarged to cover 3/4 of the screen. The internal cervical ostium, external cervical ostium, cervical canal, and endocervical mucosa were viewed simultaneously [11]. Three different images were obtained during the examination of each patient, and the ideal and shortest cervical length were recorded. Uterine volume was measured as follows: the uterine length was first calculated from the fundus to the internal os with a vaginal probe on a sagittal plane. The transducer was then moved 90° towards the transverse plane and adjusted to give the maximum anteroposterior diameter. Following this, the anteroposterior and transverse diameters were measured. Volume was calculated using the formula for a prolate ellipsoid: Volume = (0.52 $\times$ length $\times$ anteroposterior diameter $\times$ transverse diameter) [12]. The distance between the intersection of the uppermost vertical arm of the IUD and the junction between the endometrium and the uterine cavity (IUD-ED) was measured with TV-USG in the mid-sagittal plane. The maximum IUD-ED distance required to provide sufficient contraception is still controversial. To prevent unnecessary removal, we therefore defined displacement as an IUD-ED of $>$10 mm [13]. Normally positioned IUDs were defined as a distance of 10 mm between the top of the IUD and the end of the endometrial cavity, with both arms fully extended and parallel to the axis of the uterine cornua, and with the vertical part of the IUD lying centrally in the uterine cavity. “Displaced” IUDs were defined as any deviation from this normal position. Displacement was considered to be the non-fundal location of an IUD when the distance between the top of the IUD and the end of the endometrial cavity was $>$10 mm [13]. “Lower uterine segment” was defined as above the internal cervical ostium, or at the isthmus level. Patients whose IUDs were found to be displaced between the 6th week and up to 5 years after insertion and who attended our Family Planning Outpatient Clinic for control examination were divided into four subgroups as illustrated above (Fig. 3).

Fig. 3.

Illustration of abnormal copper intrauterine device positions. (A) Group 1: patients whose IUD fundus distance was $>$10 mm but did not extend to the lower uterine segment (low lying). (B) Group 2: patients whose IUDs were located in the lower uterine segment but did not extend to the cervix (partial expulsion). (C) Group 3: patients with cervical IUDs (cervical). (D) Group 4: patients whose IUDs were not detected (complete expulsion).

Statistical analyses were performed using SPSS v.21 (SPSS Inc., Chicago, IL, USA). The distribution of variables was tested for normality using histograms and the Kolmogorov Smirnov and Shapiro-Wilk W tests. Parametric continuous data are presented as the mean $\pm$ standard deviation, nonparametric continuous data as the median (min–max), and categorical variables as numbers (percentages). The t-test and Mann-Whitney U test were used to compare continuous data between two independent groups, while Chi-square and Fisher’s exact tests were used to compare categorical data. Factors affecting IUD displacement were analyzed using binary logistic regression. Age and variables thought to affect the displacement in univariate analysis (p 0.25) were added to the regression model. Variables that remain in the multiple regression model were determined by the Backward Logistic Regression (LR) method. The Hosmer-Lemeshow test was performed to evaluate the goodness-of-fit of the regression model. Receiver operator characteristic curve (ROC) analysis was performed to evaluate the predictive value of the UCA for IUD displacement. All tests were bilateral and the level of significance was set at a p-value of 0.05.

A total of 143 patients were enrolled in this study. Patients were categorized according to the position of their IUD as detected by ultrasonography during their 6-week follow-up visit after insertion. IUD displacement was observed in 67 patients (displaced group), while no IUD displacement was seen in 76 patients (control group). The mean age, body mass index (BMI), median parity, delivery type, timing of insertion, and IUD insertion time were not significantly different between these two groups. A comparison of the demographic characteristics of the two patient groups is shown in Table 1.

The clinical characteristics of patients are shown in Table 2. Patients with displaced IUDs showed more frequent menorrhagia (43% vs. 25%, p = 0.021) and irregular menstrual pattern (51% vs. 22%, p 0.001) compared to those without displacement.

Table 3 shows the incidence of symptoms during presentation at follow-up. The incidence of vaginal bleeding (28.4%) during speculum examination of patients with IUD displacement was significantly higher than in patients without displacement (5.3%) (p 0.001). No significant differences were observed between the two groups for the incidence of foul-smelling discharge, leukorrhea, or missing IUD thread in the speculum examination and abdominal examination (Table 3).

The uterine characteristics of the two patient groups are compared in Table 4. Patients with IUD displacement had significantly shorter cervical length (p = 0.005), longer endometrial cavity length (p 0.001), longer total uterine length (p 0.001), larger uterine volume (p 0.001), and larger UCA (p 0.001) compared to patients without displacement.

Ultrasonographic evaluation of the 67 patients with IUD displacement revealed that 64 (95.5%) had a displacement, one (1.5%) had a perforation, and two (3%) had IUD displacement and pregnancy. Of these 67 patients, the presentations for IUD displacement were: low but not extending to the cervix (62.7%), low and extending to the cervix (19.4%), cervical (14.9%), and complete expulsion (3%).

Patient age and all the variables identified in univariate analysis as being associated with IUD displacement with a p-value of 0.25 were added to the regression model. Variables that remained in the multiple regression model were determined using the backward LR method (Table 5). The variables remaining in the model (BMI, parity, and cervical length) had no significant effect on IUD displacement (p = 0.126, p = 0.069, and p = 0.097, respectively). However, the presence of vaginal bleeding upon the examination of speculum was associated with an approximately 86.9-fold (95% confidence interval (CI): 1.3–5885.3) increase in IUC displacement compared to patients without signs of bleeding (p = 0.038). Each one mm increase in the endometrial cavity length increased IUD displacement by 2.02-fold (95% CI: 1.03–3.96, p = 0.041). Each one mm increase in the total uterine length reduced IUD displacement by 0.52-fold (95% CI: 0.28–0.99, p = 0.045). Each one cm${}^3$ increase in the uterine volume increased IUD displacement by 2.32-fold (95% CI: 1.29–4.19, p = 0.005). Each one degree increase in the UCA increased IUD displacement by 1.31-fold (95% CI: 1.06–1.63, p = 0.012).

As shown in Fig. 4, ROC analysis revealed the UCA had significant predictive value for IUD displacement (area under the curve (AUC): 0.865, 95% CI: 0.80–0.93, sample entropy (SE): 0.032, p 0.001). The optimal UCA was found to be 139.5°, giving a sensitivity of 58%, specificity of 94.7%, positive predictive value of 90.7%, and negative predictive value of 72%.

Fig. 4.

Receiver operator characteristic curve and UCA degree averages for patients who underwent copper IUD insertion. (A) The receiver operator characteristic (ROC) area under the curve was 0.865 with a 95% confidence interval of 0.80–0.93, standard error of the mean (SEM) of 0.032 (p 0.001) and optimal cut-off value for the UCA of 139.5 degrees (p = 0.007). (B) Means of the UCA for the two groups. Error bars represent standard errors.