Incidence, prevalence, and trends in endometriosis diagnosis: a United States population-based study from 2006 to 2015

Endometriosis, defined as the presence of extrauterine endometrial tissue, often has serious symptoms that significantly impact the lives of those affected and is associated with annual health care costs of approximately $16,500. Despite the significant societal and individual burden of endometriosis, prevalence and incidence data in the United States are lacking with prevalence rates ranging from 0.8% to 11% and incidence ranging from 4.2 to 35 per 10,000 women-years. It is unclear as to why these rates differ so greatly between studies, although diagnostic parameters, practice specialty, race and ethnicity, age, and socioeconomic status seem to influence estimates and may play a role. This 10-year retrospective cohort study aimed to define endometriosis incidence and prevalence in a nonselected cohort of women in the United States and identify trends and influencing factors in diagnosis pattern over time. Clinical and demographic patient data were collected from the Kaiser Permanente Washington (KPWA) database from January 2006 to December 2015. Women continuously enrolled at KPWA, between the ages of 16 and 60, and with at least 1 medical encounter during the 2 years before entering the study cohort were included. For estimation of incidence rate, women were excluded who had hysterectomy within 2 months or recorded history of endometriosis within 2 years of study cohort entry. All diagnoses of endometriosis recorded before January 2016 were included in prevalence estimates. Endometriosis was considered present if evidence was noted in an operative report, pathology report, imaging report, and/or medical encounter within 60 days of ICD coding date, and positive predictive value of ICD coding was determined by chart review. Incidence rates per 10,000 woman-years over all women were calculated using direct standardization method to the 2015 study population. Poisson regression analysis was used to determine secular trends in incidence by race/ethnicity groups, diagnosis modalities, and provider type. Overall and age-specific prevalence was calculated for the 2015 population. A total of 332,056 women contributed person-years to the analysis cohort, within which 2863 cases of endometriosis were identified resulting in an overall average incidence rate of 24.3 per 10,000 woman-years. Prevalence rate was calculated at 1.9% for the 2015 cohort. The positive predictive value was found to be 80% for identification of incident endometriosis cases using ICD code. Among all participants, incidence rate per 10,000 woman-years declined from 30.2 to 17.4 ( P < 0.001) during the study period. Incidence rates were highest among women aged 36 to 45 years in most years and were similar across race/ethnicity groups. The only race/ethnicity group that did not experience a significant decline in incidence rate were Hispanic women ( P = 0.07). Diagnosis modality or diagnosing provider type did not differ by race/ethnicity. Surgical and clinical diagnosis rate per 10,000 person-years declined from 13.4 and 16.1 to 7.4 and 8.9, respectively ( P < 0.001). Incidence of diagnosis by an OBGYN and PCP per 10,000 person-years decreased from 22.9 and 5.1 to 13.1 and 2.5, respectively ( P < 0.001). Frequency of chronic pelvic pain diagnosis increased from 3.0% to 5.6% between 2006 and 2015. This study shows that, among a nonselected population of US women, incidence rates of endometriosis declined uniformly across age groups, race/ethnicities, diagnostic modalities, and provider types and corresponded to an increase in the rate of chronic pelvic pain diagnosis.