Racial and ethnic differences in risk for hysterectomy following insertion of a levonorgestrel-releasing intrauterine device: A retrospective cohort study

The data analyzed in this study are owned and administered by the North Carolina Translational and Clinical Sciences Institute (NC TRACS). Access to the data for research purposes can be requested at: https://tracs.unc.edu/index.php . Questions about data access may be directed to [email protected] .

Our cohort included patients who were White (58.1%), Black (26.6%) and Hispanic (15.3%) and approximately half (48.2%) were aged 35 years or over ( Table 1 ; See Appendix Table A.3 for patient characteristics by race and ethnicity). There were 76 hysterectomies observed during follow-up (9.7% of patients). Median follow-up time (to either hysterectomy or censoring) was 1.69 years (interquartile range 0.8–2.9 years, 90th percentile: 4.0 years). Minimum follow up time was 0 days (<10 patients had an insertion on the last day of the study period). In no instances did follow-up time differ significantly where hysterectomy rates did not, suggesting that follow-up rates were similar regardless of patient characteristics. The cumulative risk of hysterectomy within one year following LNG-IUD insertion was 7.0% and this increased to 14.9% by four years post-insertion. There was little evidence of rates differing by race and ethnicity ( Table 1 ), but several other factors were associated with significantly higher hysterectomy rates, including older age, higher severity uterine bleeding and bulk symptoms, being multiparous (compared to nulliparous), having fibroids, and being on Medicaid or being uninsured (compared to being privately insured). Hysterectomy rates were also significantly higher among patients who had used hormonal IUDs before than among patients with no recorded prior use and the largest magnitude differences were at three (16.0% vs 12.9%) and four (21.6% vs 15.0%) years post-insertion. Table 2 presents the unadjusted and multivariable-adjusted hazard ratio estimates from Cox regression models. In unadjusted models, higher age at treatment, more severe uterine bleeding and bulk symptoms, diagnostic codes for fibroids, and being multiparous or on Medicaid were associated with increased hysterectomy risk, similar to findings from Table 1 . In adjusted models, only bleeding severity remained significantly associated with hysterectomy rates (HR: 1.05[1.00–1.10]), though age at treatment and bulk severity were borderline significant with estimates of meaningful magnitude (age HR: 1.04[1.00–1.09]; bulk HR: 1.15[1.00–1.32]). Notably also, while confidence intervals for associations with pelvic pain symptoms over-lapped the null, the associations were of similar magnitude and direction as those for symptoms of uterine bleeding. We tested for interactions between race and ethnicity and each other covariate but none were significant at the p<0.10 level and in our final model with all covariates, including interactions with race and ethnicity for all variables did not improve model fit (p=0.287).

We use a retrospective cohort design among patients treated for benign gynecological conditions with an LNG-IUD. We identified eligible patients using the North Carolina Translational and Clinical Sciences Institute (NC TraCS) “data warehouse” that leverages structured data from electronic health records (EHR). Data pertained to a large and expanding healthcare system in the US South, with an extensive financial assistance program for underinsured and uninsured patients. Available data primarily came from hospitals and office-based gynecology providers who were employees of the system’s academic medical center. We abstracted additional unstructured clinical data from surgical notes and other medical records over four months preceding LNG-IUD insertion [ 22 – 24 ] for matching with structured EHR data. Eligible people were aged 18–44 years, and had an LNG-IUD insertion associated with a diagnostic code for a benign gynecological condition between April 4 th 2014 and September 18 th 2019 (specifically: abnormal vaginal bleeding, fibroids, endometriosis, and gynecological pain; see Appendix Table A.1 for medical codes). Patients with LNG-IUDs inserted for other reasons who later developed benign gynecological conditions were not considered within scope. The data warehouse extraction identified 1,081 patients across the hospital system with an LNG-IUD insertion and an associated diagnostic code for a benign gynecological condition within the period studied. This was reduced to an analytical sample of 783 LNG-IUD patients following exclusions, which were largely related to inaccurate or incomplete data, or LNG-IUD placement for primary reasons other than benign gynecological conditions, such as cancer or post-partum contraception. Patients missing data on race and ethnicity were excluded as were patients identifying as Non-Hispanic and either American Indian, Alaska Native, Asian, Native Hawaiian, or Pacific Islander, or identifying with multiple non-White races. No single one of these groups contained enough patients to produce stable estimates and avoid risk of identifiability. Figure 1 details numbers of patients excluded for specific reasons. The remaining cohort includes all patients who met inclusion criteria. Structured sociodemographic data including age, race, Hispanic ethnicity, marital status and insurance status were extracted from EHR. Age in years was centered on its mean value (33.2 years) for regression models and categorized into ages 35+ and <35 years for estimating censoring-adjusted hysterectomy rates. Race and ethnicity, reported by patients at registration, were categorized as: Non-Hispanic Black, Hispanic, Non-Hispanic White. Insurance status was coded as: Private (including Tricare for members of the US military), Medicaid, Medicare (including Medicare Advantage), or uninsured (recorded as “self-pay”). Further variables indicated presence of diagnostic codes for fibroids or endometriosis. Manual abstraction of unstructured EHR data (with a pre-specified protocol) [ 22 – 24 ] provided information on prior IUD treatment, parity and gynecological symptom severity. Prior IUD treatment was coded as: no prior use, prior use of an IUD (type unknown), and prior use of a hormonal IUD. This represents past (i.e., non-current) IUD use, as current users, who were having an IUD removed as the new one was being inserted, had already been excluded (see Figure 1 ). Parity was coded as: not recorded, nulliparous, primiparous or multiparous, based on the number of prior deliveries. No data were available on LNG-IUD dosage. We coded gynecological symptom severity using adapted versions of symptom severity indices previously validated among hysterectomy patients [ 25 ]. Symptom indices for hysterectomy patients drew on unstructured physician notes, diagnostic and procedure billing codes, pre-operative physician notes, laboratory and pathology test results, and pharmacy prescriptions. Appendix Table A.2 details how these indices were adapted to LNG-IUD patients for whom we had no pre-operative notes, laboratory or pathology test results or pharmacy prescription data. Three distinct scales represented severity of uterine bleeding symptoms (range 0–21; median: 4), pelvic pain symptoms (range: 0–16; median: 3), and symptoms related to uterine bulk (range: 0–7; median: 1). Higher scores indicate more severe symptoms. Binary classifications of high vs low-moderate severity were used in estimation of hysterectomy rates (high severity categorized based on 75 th percentile as scores of 7+ for bleeding, 5+ for pelvic pain, and 4+ for bulk). Further treatments identified (in administrative or abstracted data) as occurring between LNG-IUD insertion and study end included further LNG-IUD-related encounters and related surgical procedures: myomectomy, uterine artery embolization (UAE), endometrial ablation (EA), and hysterectomy (see Figure 2 and Appendix Table A.1 for medical codes). We focus on hysterectomy as very few other surgical procedures were observed during follow up (<10 EAs, myomectomies, or UAEs) and we were not able to capture IUD expulsions and removals comprehensively. Patient follow up was until hysterectomy or study end (September 18, 2019), whichever came first, irrespective of LNG-IUD removal, or other procedures being performed. Analyses therefore assess risk of re-intervention with hysterectomy specifically, and no other potential measures of treatment success or failure. We used a Kaplan-Meier estimator to calculate cumulative incidence of hysterectomy at one, two, three and four years following LNG-IUD insertion (with days as the unit of time). We did similar calculations of hysterectomy rates stratified by race and ethnicity, symptom severity, having diagnostic codes for fibroids or endometriosis, prior IUD use and insurance status, and used a log-rank test to compare rates between groups. We used Cox regression models to further investigate associations between patient characteristics and hysterectomy rates. We estimated unadjusted associations for race and ethnicity and each covariate, before testing for interactions between race and ethnicity and each other covariate. We considered main effects significant at the p<0.05 level but used a more liberal p<0.10 threshold for interactions as statistical power may have been low. A final model included all covariate main effects, regardless of significance (i.e. race and ethnicity, age at treatment, bleeding, pain and bulk severity scores, diagnostic codes fibroids and endometriosis, prior IUD use, parity and insurance status) and any significant interactions.

Among patients aged 18–44 treated with LNG-IUDs for non-cancerous conditions we did not find statistically significant differences in hysterectomy risk by race and ethnicity over a median follow-up time of 1.7 years. Cumulative incidence rates for hysterectomy were 7.0% within the first year following LNG-IUD placement, and 14.9% at four years post-insertion. Hysterectomy rates were higher among patients who were older, had more severe pre-insertion bleeding and bulk symptoms, had diagnostic codes for fibroids, were multiparous, and were Medicaid beneficiaries. Associations with these patient characteristics did not appear to be moderated by race and ethnicity. Bleeding severity remained significantly associated with hysterectomy after adjustment for other patient characteristics in Cox regression models, and there were also borderline associations with hysterectomy risk for age and bulk severity. Comparisons of LNG-IUDs and hysterectomy as treatments for benign gynecological conditions, found LNG-IUDs to be relatively cost effective, even though patients randomized to LNG-IUD treatment often eventually undergo hysterectomy (e.g. in one large trial starting in 1994 in Finland, 20% underwent hysterectomy after one year, and 42% after five years) [ 12 , 19 , 26 ]. We observed lower rates which could be due to a number of factors including our observational design, larger sample size, differing medical system and differing geographic, social and historical context. Evidence is sparse on rates and patient characteristics associated with hysterectomy among large observational samples of patients receiving non-randomized LNG-IUD treatment. While observational data may lack measurement of factors that are important for understanding causal pathways of non-randomized exposures, detailed description of naturalistic, “real-world” treatment may still help identify potential structural or systemic barriers contributing to healthcare disparities. Our study contributes to such detailed description, with rich clinical measures of bleeding, pain and bulk symptoms, combined with sociodemographic information such as age, parity and insurance status. Although we did not find statistically significant evidence of disparities by race and ethnicity in risk for hysterectomy following LNG-IUD insertion, we cannot rule out an association. Following adjustment for symptom severity and other covariates, point estimates indicated considerably lower risk of hysterectomy for patients who were Black or Hispanic than for those who were White (HRs were 0.58–0.71), but confidence intervals were wide and we may have been under-powered to detect this association. A study of re-intervention rates following uterine-sparing surgeries (rather than LNG-IUD placement) found little evidence for rates differing by race and ethnicity [ 27 ]. Their study focused on fibroid patients in an integrated healthcare delivery system but findings were consistent in not showing increased risk of hysterectomy for patients who were Black or Hispanic. This contrasts with overall hysterectomy rates which are higher among Black and Hispanic than White populations in the United States [ 5 – 7 ]. Disparities may arise earlier in the care pathway, but there is little evidence they arise via less favorable responses to treatment for patients who are Black or Hispanic compared to those who are White. Treatment recommendations emphasize consideration of the relative cost and burden of treatment options, symptom severity and response to treatments, patient preferences, e.g. with regards to preservation of fertility, and patient age in relation to the expected onset of menopause [ 21 , 28 , 29 ]. Patients from different racial and ethnic groups may value and prioritize these factors differently, for example, medical costs and interference with work capacity may be more salient concerns among minoritized racial and ethnic groups, who also experience socioeconomic disadvantages [ 30 , 31 ]. We did not find evidence that age, symptom severity, diagnoses of fibroids or endometriosis, prior IUD use, parity or insurance status were differentially associated with hysterectomy following LNG-IUD placement for patients who were White, Black and Hispanic. However, our study may have been under-powered to detect such interactions and they should not be ruled out for future investigations in larger samples. Our study has some limitations: it is observational and potential relationships between patient characteristics and LNG-IUD treatment are unknown (we only observed the treated). The lack of clear evidence for racial or ethnic disparities in hysterectomy risk following LNG-IUD placement does not preclude potential disparities occurring earlier in the care pathway. For example, there is evidence that Black and Hispanic patients have more severe symptoms and are more likely to be treated directly with hysterectomy, without first attempting alternative treatments such as the LNG-IUD [ 17 ]. With no reliable measure for all patients of how effectively LNG-IUDs controlled symptoms, we could not determine whether Black and Hispanic patients simultaneously experience worse symptom control with LNG-IUDs but greater barriers to accessing hysterectomy treatment (balancing out to no clear evidence of differences in hysterectomy risk). Moreover, we primarily captured data from hospitals rather than primary care, so observed LNG-IUD patients may have been those receiving advanced or specialty care, potentially using the LNG-IUD as a stop-gap while awaiting surgery. Analyses assumed patients were only treated within the same hospital system. Observation of hysterectomy could have been less likely for patients receiving care from multiple healthcare systems, which potentially underestimates hysterectomy rates. We may have differentially underestimated rates if patients with particular characteristics were more likely to receive additional care outside the study healthcare system. Hysterectomy was approximately twice as likely for patients aged 35 or more, with fibroids, or more severe bleeding and bulk symptoms. Clinicians may want to consider that LNG-IUDs may be less successful at delaying hysterectomy for older patients with severe symptoms. Lesser effectiveness with bulk symptoms is consistent with previous findings that LNG-IUDs are more likely expelled for patients with large fibroids [ 1 , 14 ]. Multiparous patients also had higher hysterectomy rates, but this was less evident when controlling for covariates, potentially suggesting that parity relates to hysterectomy via associations with age and symptom severity. Indeed, age and parity might both influence patient desires for future fertility.

Levonorgestrel-releasing intrauterine devices (LNG-IUD) are among recommended initial treatments for abnormal uterine bleeding (AUB; i.e. uterine bleeding with irregular frequency, duration or volume) and similar benign conditions such as endometriosis, adenomyosis and uterine fibroids [ 1 – 4 ]. These conditions are common, especially among Black and Hispanic populations [ 5 – 8 ]. For example, prevalence estimates for AUB for reproductive-aged women range between 10–30% [ 9 – 11 ]. AUB is associated with considerable quality of life impacts and economic costs [ 9 ]. LNG-IUDs can be effective at controlling symptoms [ 1 , 12 – 14 ], but may be under-utilized relative to more invasive treatments such as hysterectomy [ 3 , 15 , 16 ], especially for Black and Hispanic patients [ 17 ]. After LNG-IUD placement, re-intervention with hysterectomy is common. For instance, one study found that 46% of patients had undergone a hysterectomy within 10 years of insertion [ 18 ]. Furthermore, LNG-IUDs are more likely to fail and be expelled when fibroids are present [ 1 , 14 ]. Evidence on risk of hysterectomy following LNG-IUD placement primarily consists of small studies with non-generalizable patient samples, and few studies follow patients for more than 1–2 years [ 12 , 19 , 20 ]. Information for patients or clinicians on the clinical and sociodemographic patient characteristics that are associated with response to LNG-IUD treatment or need for surgical re-intervention, and whether these differ by race and ethnicity, is scarce. Differential progression to more invasive treatment may be expected among minoritized patients. For example, social and economic disadvantages more common among minoritized groups may incentivize permanent, curative resolution of symptoms [ 21 ]. This could motivate patient preference for hysterectomy as an initial treatment, but also for more rapid progression to hysterectomy if an LNG-IUD is inserted but not immediately effective in managing symptoms. A higher prevalence of fibroids among minoritized patients [ 8 ] might also mean LNG-IUDs are less effective [ 14 ] and increase risk of progression to hysterectomy. Despite these known patterns, there are sparse data on which LNG-IUD patients eventually undergo hysterectomy, and whether hysterectomy risk factors differ by race or ethnicity. Therefore, our primary aim is to assess whether risk of hysterectomy following LNG-IUD placement differs by race and ethnicity. Our secondary aim is assessing whether associations between the clinical and sociodemographic characteristics of the patient (age, symptom severity, prior IUD use, parity and insurance status) and hysterectomy risk following LNG-IUD placement differ by race and ethnicity.