Dimensions of the endometrial cavity and intrauterine device expulsion or removal for displacement: a nested case–control study

Authors

  • H Liang,

    1. Department of Reproductive Epidemiology and Social Science, National Population and Family Planning Key Laboratory of Contraceptive Drugs and Devices, Shanghai Institute of Planned Parenthood Research, Shanghai, China
    Search for more papers by this author
  • L Li,

    1. National Research Institute for Family Planning, Beijing, China
    Search for more papers by this author
  • W Yuan,

    1. Department of Reproductive Epidemiology and Social Science, National Population and Family Planning Key Laboratory of Contraceptive Drugs and Devices, Shanghai Institute of Planned Parenthood Research, Shanghai, China
    Search for more papers by this author
  • Y Zou,

    1. National Research Institute for Family Planning, Beijing, China
    Search for more papers by this author
  • E-S Gao,

    1. Department of Reproductive Epidemiology and Social Science, National Population and Family Planning Key Laboratory of Contraceptive Drugs and Devices, Shanghai Institute of Planned Parenthood Research, Shanghai, China
    Search for more papers by this author
  • JMN Duffy,

    1. Women's Health Research Unit, Centre for Primary Care and Public Health, Blizard Institute, Barts and The London School of Medicine and Dentistry, London, UK
    Search for more papers by this author
  • S-C Wu

    Corresponding author
    1. National Research Institute for Family Planning, Beijing, China
    • Correspondence: Prof S-C Wu, National Research Institute for Family Planning, Beijing 100081, China. Email wu.shang.chun@yeah.net

    Search for more papers by this author

Abstract

Objective

To evaluate the effect of the dimensions of the uterine cavity in relation to the expulsion, or removal for displacement, of intrauterine devices (IUDs) MLCu375 and TCu380A.

Design

A case–control study nested in a multicentre clinical trial.

Setting

Eighteen family planning clinics in China.

Population

Forty-eight pairs of TCu380A users and 118 pairs of MLCu375 users.

Methods

The women were classified as cases if IUD expulsion or displacement occurred during the first year of follow-up after insertion. One control was randomly selected for each case matched by IUD model, centre, age, service provider for insertion, and date of insertion. Axial length of the uterine cavity (LUC) and the largest transverse diameter of the coronal section of the uterine cavity (LTD) were measured using abdominal ultrasound. We used multivariate conditional logistic regressions to estimate the adjusted odds ratios (aORs) of expulsions or displacements among women with different sizes of uterine cavity according to the quartile of LUC and LTD, respectively.

Results

Among MLCu375 users, women with LTD ≥ 27 mm had a higher risk of expulsion or displacement (aOR 2.40; 95% confidence interval, 95% CI, 1.02–5.63), after adjusting for the volume of menstrual flow, dysmenorrhoea, parity, uterine position, MLCu375 type, and LUC. Among TCu380A users, the association between LTD ≥ 37 mm and expulsion or displacement (aOR 4.98; 95% CI 1.01–22.49) was statistically significant, after adjusting for LUC and potential confounding factors.

Conclusion

Our study suggests that LTD should be considered when making the decision of which IUD model to use.

Introduction

Intrauterine devices (IUDs) are the most widely used reversible method of contraception, particularly in less developed regions such as Africa, Asia (excluding Japan), and Latin America, where 16.5% of women of reproductive age use this device. In contrast, only 9.4% women use IUDs in more developed regions.[1] The IUD has been the most commonly used method of contraception in China since the 1950s. Currently, it is used by more than 50% of Chinese women of reproductive age.[2]

TCu380A and MLCu375, which are commonly used in China, are highly effective and have few side effects.[3-5] A Cochrane review reported expulsion rates to range from 1.9 to 4.6% for TCu380A, and from 1.6 to 5.0% for MLCu375, during the first year after insertion.[6] If expulsion goes unnoticed, women may unexpectedly become pregnant. In addition, the displacement of the IUD in the uterine cavity may increase the risk of expulsion.[7, 8] When the distance between the upper end of the IUD to the external uterine fundus is >2 cm, as identified by ultrasonography within the first 5 days of the menstrual cycle, the IUD will be removed to protect women from an unexpected pregnancy, which is a routine practice in China. In a large multicentre clinical trial conducted in China, the expulsion (including complete and partial expulsion) rates of MLCu375 and TCu380A were 0.64/100 and 0.61/100 women-years, respectively.[9] The removal rate for displacement (downward dislocation), however, was 12.7/100 and 6.0/100 women-years, respectively.[9]

Young maternal age, heavy menstruation, dysmenorrhoea, uterine position, and the service provider's experience with regard to insertion were reported to be associated with expulsion or removal for displacement.[10-15] One hypothesis for IUD expulsion or displacement is dimensional incompatibility between the uterine cavity and the IUD. Castro found that the expulsion rate within 1 and 2 years of MLCu375 was more than four times higher for women with an endometrial cavity length of ≥45 mm, compared with those with a cavity length of <45 mm (P < 0.01)[16]; however, two prospective studies failed to find any association between endometrial cavity length and the expulsion rate of IUDs.[17, 18] The relationship between uterine cavity dimensions and IUD expulsion remains unclear.

Therefore, using data collected in a multicentre clinical trial,[9] we performed a nested case–control study to explore the association of axial and transverse lengths of the uterine cavity and expulsion or displacement of MLCu375 and TCu380A IUDs.

Methods

Study population

Data in this study were obtained from a multicentre randomised clinical trial that compared TCu380A and MLCu375 IUDs with other IUDs. Detailed information on the design of the clinical trial has been provided elsewhere.[9] Briefly, 18 family planning clinics in China used identical case record forms and protocols to collect information on IUD insertion and use between 2004 and 2005. All of the women were 20–40 years of age, had at least one live birth, had normal menstruation (21–42 days between two cycles; ≤7 bleeding days for each cycle), had an axial length of the uterine cavity (LUC) ranging from 5 to 9 cm, and had no contraindications for IUD use. IUDs were randomly assigned to participants according to sealed random-allocation envelopes, preprinted before study initiation, and opened at the time of IUD insertion. The IUDs were inserted within 7 days of the end of menstruation. The last surgical or medical abortion (at <14 weeks of gestation) had to have occurred at least 1 or 2 months prior to IUD insertion, respectively. The last caesarean section had to have occurred at least 6 months prior to IUD insertion. Follow-up visits were scheduled after 1, 3, 6, and 12 months of IUD insertion. At each visit, women were interviewed to obtain information on side effects of IUDs and were scanned by ultrasound to examine whether expulsion, displacement, or any other abnormality had occurred. There were 852 women who were enrolled to receive TCu380A, and 860 women who were enrolled to receive MLCu375 in the clinical trial.

Sizes of the MLCu375 and TCu380A IUDs used in the study

TCu380A has only one size, with the vertical and horizontal arms being 36 and 32 mm in length, respectively. The MLCu375 device comes in two sizes: short (length, 29 mm; width, 18 mm) and standard (length, 36 mm; width, 18 mm).

Definition of displacement

Displacement is considered to occur when the distance between the upper end of the IUD to the external uterine fundus is >2 cm, as identified by ultrasonography within the first 5 days of the menstrual cycle.

Definitions of cases and controls

Among all of the participants in the clinical trial, women were classified as cases if IUD expulsion or removal for displacement occurred during the first year after insertion. Simultaneously, controls were chosen among women using the same model of IUD and in whom no pregnancy, expulsion, or removal occurred. One control was randomly selected for each case, matched by centre, IUD model, age (±2 years), service provider for insertion, and the date of IUD insertion (±1 month).

Figure 1 shows the study population in this case–control study. Fifty-one cases with TCu380A and 120 cases with MLCu375 were matched by controls. Three pairs of women with TCu380A and two pairs of women with MLCu375, for whom information on the dimensions were missing, were excluded, and the remaining 48 pairs with TCu380A (nine expulsions and 39 displacements) and 118 pairs with MLCu375 (15 expulsions and 103 displacements) were included in the current analyses.

Figure 1.

The study population for the case–control study.

Measurement of uterine cavity dimensions

The axial length of the uterine cavity (LUC; the distance between the endometrium of the uterine fundus and the cervical internal os) and the length of the largest transverse diameter of the coronal section of the uterine cavity (LTD; the maximum distance between the two edges of the coronal section of the uterine cavity) were measured by abdominal ultrasound (Figure 2).

Figure 2.

Figure showing the axial length of the uterine cavity (LUC; the distance from b to c) and the length of the largest transverse diameter of the coronal section of the uterine cavity (LTD; the distance from i to j).

For the cases, measurements were performed within the first 7 days of the menstrual cycle of women following the removal of the IUD for displacement or following confirmation of its absence in the uterine cavity. For women in the control group, measurements were performed within the first 7 days of the menstrual cycle after they were confirmed as controls. In each centre the measurements were made by the same service provider for matched cases and controls, who was trained to take measurements by the study team.

Statistical analysis

All data handling and statistical analyses were performed using sas 9.2 (SAS Institute Inc., Cary, NC, USA).

Expulsion and displacement may have the same mechanism, in which myometrial contraction expels the IUD and leads to displacement and eventually to expulsion.[7, 8] Therefore, we analysed expulsion and displacement together.

We examined the association between the endometrial cavity size and expulsion or displacement by MLCu375 and TCu380A separately, because the association may be different for these two IUDs because of differences in shape and size. In this study, LUC and LTD were measured to explore the relationship between dimensions of the uterine cavity and expulsion or displacement of the IUDs.

Although there are two types of MLCu375 (short and standard), the type of MLCu375 used was not matched between cases and controls (which might be an important confounding factor for studying the relationship between uterine dimensions and IUD expulsion). Therefore, we generated a new variable (D1), which was defined as the length of the uterine cavity minus the length of the vertical arm of MLCu375, in order to measure the effect of the length difference between the IUD and the uterine cavity on expulsion or displacement.

First, a paired Student's t-test was used to compare the mean lengths of LUC, LTD, and D1 (for MLCu375 only) between cases and controls, as these measurements are normally distributed.

Second, LUC, LTD, and D1 were all treated as dichotomous variables in the following analyses. We categorised women into two groups using 45 mm as the cut-off point of LUC, based on Castro's study,[16] when analysing the effect of LUC among MLCu375 users. However, according to the literature there are no widely accepted cut-off points of LUC for TCu380A, of LTD for either model, and of D1. Women were categorised into two groups (small or big uterine cavity) according to the quartile of LUC (39 mm for TCu380A), LTD (27 mm for MLCu375 and 37 mm for TCu380A), and D1 (15 mm for MLCu375), respectively. Then conditional logistic regressions were used to estimate the odds ratios (ORs) of expulsions or displacement between women with different sizes of uterine cavity according to LUC, LTD, and D1, respectively.

The volume of menstrual flow (less than normal or normal), dysmenorrhoea (no or yes), uterine position (anteverted, midposition, or retroverted), and parity (1 or ≥2), which were possible risk factors according to the literature,[8-13] were considered as potential confounding factors. In the first multivariate analysis, the above covariates were adjusted to estimate the adjusted odds ratio (aOR). For MLCu375, the type of MLCu375 was further adjusted in the model when analyzing the effect of LUC and LTD, respectively. The effect of LUC on expulsion in the present analyses was comparable with those reported in the literature, where just LUC was measured and LTD was not considered. In the second multivariate analysis, LUC or LTD was further adjusted to assess whether the association of LTD or LUC was independent of LUC and LTD, respectively.

Third, we performed a subgroup analysis among pairs of women who were both using the short MLCu375 (53 pairs). We did not perform the same analyses in pairs for the standard MLCu375 because of the small sample size (19 pairs).

Finally, as most of the cases in this study were displacements, with few expulsions, we performed a sensitive analysis after expulsion cases were excluded.

Results

There were no statistically significant differences in the characteristics of the women in the case and the control groups, regardless of MLCu375 or TCu380A use (Table 1).

Table 1. Characteristics of the study population
 MLCu375 (%)TCu380A (%)
Cases (n = 118)Controls (n = 118)Cases (n = 48)Controls (n = 48)
  1. a

    Mean (SD).

Age a 28.7 (4.7)28.4 (4.7)28.6 (4.9)29.3 (4.5)
20–25 years202697
25–30 years53542021
30–35 years28221211
35–40 years171679
Parity
1100984543
≥2182035
Volume of menstrual flow
Less than normal152353
Normal103954345
Dysmenorrhoea
No89953635
Yes29231213
Outcome of last pregnancy
Vaginal delivery64632518
Caesarean section40361622
Abortion141978
Uterine position
Anteverted50442015
Midposition202389
Retroverted48512024
Time between last pregnancy outcome and insertion of IUD
≤6 months17221010
6–12 months33301311
>12 months68662427

For those with MLCu375, women in the case group had a greater mean length of LTD than those in the control group, whereas there were no significant differences in LUC and D1 between the two groups. For women with TCu380A, similar results were observed (Table 2).

Table 2. Uterine cavity dimensions in cases and controls
DimensionsCasesControls t a P
  1. a

    Paired Student's t-test.

MLCu375 n = 118n = 118  
Largest transverse diameter of coronal section of uterine cavity (LTD)32.0 ± 6.2 mm30.6 ± 6.9 mm–2.430.02
Length of uterine cavity (LUC)41.7 ± 8.4 mm40.7 ± 8.0 mm–1.580.12
Length of uterine cavity – the length of vertical arms of the IUD (D1)9.9 ± 8.3 mm9.6 ± 8.1 mm–0.630.53
TCu380A n = 48n = 48  
Largest transverse diameter of coronal section of uterine cavity (LTD)33.1 ± 6.4 mm31.5 ± 6.9 mm–2.150.04
Length of uterine cavity (LUC)43.3 ± 5.2 mm42.6 ± 5.8 mm–0.820.42

Table 3 describes the association between the uterine dimensions of women and the expulsion or displacement of MLCu375. Women with an LTD of ≥27 mm had a higher risk of expulsion or displacement (aOR 2.66; 95% confidence interval, 95% CI, 1.16–6.11) after adjusting for the volume of menstrual flow, dysmenorrhoea, parity, uterine position, and MLCu375 type. The association remained significant after further adjusting LUC in the model (aOR 2.40; 95% CI 1.02–5.63). There was a higher risk of expulsion or displacement (aOR 2.40; 95% CI 1.03–5.58) in women with an LUC of ≥15 mm than the axial length of the MLCu375 (D1) used after adjusting for volume of menstrual flow, dysmenorrhoea, parity, and uterine position; however, the risk was not significant after further adjusting for LTD (aOR 2.17; 95% CI 0.90–5.20). There was no significant difference in the risk of expulsion or displacement between the groups according to LUC in each model. Similar results were observed when the aforementioned analyses were restricted to the pairs who had short MLCu375 IUDs inserted (data not shown).

Table 3. The effect of uterine dimensions on the expulsion or displacement of the MLCu375 IUD
DimensionsCase n (%)Control n (%)ORaaOR1 (95% CI)baOR2 (95% CI)
  1. a

    Conditional logistic regression.

  2. b

    Adjusted for volume of menstrual flow, dysmenorrhoea, uterine position, parity, and type of MLCu375 IUD; for D1, adjusted for volume of menstrual flow, dysmenorrhoea, uterine position, and parity.

  3. c

    LTD was further adjusted besides the variables listed in model 1.

  4. d

    D1 was further adjusted besides the variables listed in model 1.

Length of uterine cavity (LUC)
<45 mm79 (66.9)87 (73.7)111
≥45 mm39 (33.1)31 (26.3)1.621.45 (0.69–3.02)1.32 (0.63–2.79)c
Length of uterine cavity – the length of vertical arms of the IUD (D1)
<15 mm86 (72.9)95 (80.5)111
≥15 mm32 (27.1)23 (19.5)2.002.40 (1.03–5.58)2.17 (0.90–5.20)c
Largest transverse diameter of coronal section of uterine cavity (LTD)
<27 mm16 (13.6)31 (26.3)111
≥27 mm102 (86.4)87 (73.7)2.882.66 (1.16–6.11)2.40 (1.02–5.63)d

For women with TCu380A, those with LTD ≥ 37 mm had a higher risk of expulsion or removal for displacement (aOR 4.64; 95% CI 0.97–22.24) after adjusting for volume of menstrual flow, dysmenorrhoea, parity, and uterine position, although the association was not significant; however, similar associations were obtained (aOR 4.98; 95% CI 1.01–24.49) after further adjusting for LUC in the model. There was no statistically significant difference in the risk of expulsion or displacement between the groups according to LUC in each model (Table 4).

Table 4. The effect of uterine dimensions on the expulsion or displacement of the TCu380A IUD
DimensionsCase n = 48 (%)Control n = 48 (%)ORaaOR1 (95% CI)baOR2 (95% CI)
  1. a

    Conditional logistic regression.

  2. b

    Adjusted for volume of menstrual flow, dysmenorrhoea, uterine position, and parity.

  3. c

    Adjusted for volume of menstrual flow, dysmenorrhoea, uterine position, parity, and LTD.

  4. d

    Adjusted for volume of menstrual flow, dysmenorrhoea, uterine position, parity, and LUC.

Length of uterine cavity (LUC)
<39 mm9 (18.8)12 (25.0)111
≥39 mm39 (81.2)36 (75.0)1.501.46 (0.50–4.25)1.66 (0.54–5.08)c
Largest transverse diameter of coronal section of uterine cavity (LTD)
<37 mm32 (66.7)39 (81.3)111
≥37 mm16 (33.3)9 (18.7)4.504.64 (0. 97–22.24)4.98 (1.01–24.49)d

The results did not change significantly when the multivariate analyses were restricted to the cases of displacements and matched controls, regardless MLCu375 or TCu380A use (data not shown).

Discussion

Main findings

In this nested case–control study, we found that women with a greater LTD (≥27 mm for MLCu375 and 37 mm for TCu380A) had a higher risk of expulsion or removal for displacement, even after adjusting for the effect of the axial LUC. Among MLCu375 users, an LUC of 15 mm or greater than the maximum length of the IUD was associated with an increased risk of expulsion or displacement; however, this association was not significant after further adjusting for LTD.

Strengths and limitations

This study has several methodological strengths. To our knowledge, this was the first study to investigate the effect of axial and transverse diameters of the uterine cavity on expulsion or displacement of IUDs. This was a nested case–control study within a multicentre clinical trial. The information on IUD type and other covariates were collected before the outcome occurred, thereby reducing recall bias. The cases were confirmed prospectively as the clinical trial continued, which made the classification of cases and controls less likely to be disturbed by misclassification. We collected detailed information on gynaecological factors and reproductive history, which made it possible to adjust for potential confounding factors.

This study also has limitations. First, compared with the previous study on the relationship between the length of uterine cavity and expulsion, the current study analysed expulsion and displacement together because of the small sample size for expulsion. Because most cases in the current study comprised IUD removal for displacement, the results may have better implications for displacement. This was further confirmed by the similar results after sensitivity analyses restricted to displacement cases only. Although we assumed that displacement and expulsion had a similar mechanism, removal for displacement may not lead to expulsion, and an IUD facing downwards in the uterine cavity may still be able to protect against pregnancy. Second, the type of B-ultrasound device used to measure the dimensions of the uterine cavity is different in each centre; however, because the measurements were performed by the same service provider with the same B-ultrasound device in each centre, it was reasonable to believe that the misclassification would be non-differential, which would thereby reduce any bias towards null. Moreover, the mean LUC and LTD in this study were similar to those in other studies performed in China,[19, 20] which indicated the validity of the clinical measurements in our study. Third, although this case–control study was nested in a large multicentre clinical trial, relatively low numbers of TCu380A pairs resulted in wide 95% CIs around the estimates. In any further study, the sample size should be enlarged to corroborate the observed association.

Interpretation

Our results that women with a uterine cavity length of 15 mm or greater than the maximum length of MLCu375 (D1) had a higher risk of expulsion or displacement presented a similar association as reported in a previous study, in which Cavimeter was used to measure LUC.[16] The study demonstrated that when the difference between uterine cavity length and the maximum length of MLCu375 was 10 mm or greater, the expulsion rate increased significantly (P > 0.01). Our study provided additional evidence on the association of LUC with expulsion or displacement in the Chinese population.

Our results demonstrate that the association between expulsion or displacement of TCu380A and LUC was not statistically significant, which is consistent with the results of several previous studies that did not show an association between length of the endometrial cavity or total uterine length and the risk of expulsion of T-shaped Cu.[12, 17, 18]

Our study found that women with a greater LTD had a higher risk of expulsion or displacement among MLCu375 or TCu380A users. To the best of our knowledge, few studies have investigated the association between expulsion and LTD, as measured by ultrasound. The axial or transverse diameter was related to the shape or size of the uterine cavity; this indicates that women with uterine cavities of the same axial length may not have the same transverse diameter. Some studies used different measuring devices, including Wing Sound II, uterometer, and three-dimensional ultrasonography, to measure the axial and transverse diameters of the uterine cavity in order to simulate its shape.[21-23] Based on measurements of axial LUC and three transverse diameters of the coronary section by uterometer, Gan et al. categorised the uterine cavity into three different shapes: isosceles triangle; narrow long triangle; and wide short triangle.[22] Based on the shapes and diameter values of the coronary section, as measured by three-dimensional ultrasonography, which shows the coronal section of the uterine cavity so that the diameters could be measured clearly, Lu et al. categorised the uterine cavity into four forms: isosceles triangle; equilateral triangle; slab-sided triangle; and curve triangle.[23] Women with different shapes of uterine cavity will have different transverse diameters, even if they have identical lengths. This could partly explain why the length of the transverse diameter, but not LUC, was associated with expulsion or displacement. This study suggested that the transverse diameter might be a better index than the axial length for studying the association between uterine cavity dimensions and side effects of IUD. Further study is needed for exploring the relationship.

Conclusion

The largest transverse diameter of the coronal section of the uterine cavity (LTD) was associated with expulsion or displacement of MLCu375 and TCu380A IUDs. Our study suggests that the transverse diameter of the uterine cavity should be considered when deciding which IUD model to use; however, more research with larger sample sizes is needed to generate the cut-off points of LTD for the clinical practice of IUD insertion.

Disclosure of interests

The authors declare no potential conflicts of interest with respect to the authorship and/or publication of this article.

Contribution to authorship

HL analysed the data, interpreted the results, and wrote the first draft of the article. LL collected data, and assisted with data analysis and the interpretation of results. WY and ESG contributed to the study design, interpretation of results, and the revision of the article. YZ contributed to analysing data and interpreting the results, and to the revision of the article. JMND contributed to the interpretation of results and revised the article. SCW conceived the idea, designed and initiated the study, and revised the article. All authors read and approved the final version.

Details of ethics approval

Ethics approval was given by the Institutional Review Board of the National Research Institute for Family Planning, Beijing, China (reg. no. 12005). Written informed consent was obtained from all women participating in the study.

Funding

The current study was part of the project entitled Investigation on the Reason and Prevention Strategy for Failure of Intrauterine Device (2004BA720A31), which was funded by the National Key Technology Programme, China.

Acknowledgements

We gratefully acknowledge Dr Miao Maohua for her help in revising the article, and acknowledge the EBM-CONNECT Collaboration. The EBM-CONNECT (Evidence-Based Medicine Collaboration: network for systematic reviews and guideline development research and dissemination) collaboration (in alphabetical order by country) includes: L. Mignini (Centro Rosarino de Estudios Perinatales, Argentina); P. von Dadelszen, L. Magee, and D. Sawchuck (University of British Columbia, Canada); E. Gao (Shanghai Institute of Planned Parenthood Research, China); B.W. Mol and K. Oude Rengerink (Academic Medical Centre, the Netherlands); J. Zamora (Ramon y Cajal, Spain); C. Fox and J. Daniels (University of Birmingham, UK); and K.S. Khan, S. Thangaratinam and C. Meads (Barts and the London School of Medicine, Queen Mary University of London, UK).

Ancillary