Menopause is rarely coded in the emergency department: A real-world study of 413,414 visits by women aged 40-65 years (the MIMOSA study).

OBJECTIVES: To quantify how often menopause is captured by diagnostic coding among women aged 40 to 65 attending the emergency department with potentially menopause-related symptoms, to characterise their presentation, and to identify factors associated with coding, early reattendance (within 72 h) and short-stay admission (under 48 h). STUDY DESIGN: Retrospective observational study, reported following an established reporting guideline for observational studies, using structured, real-world clinical data available in the MDClone platform, where data are de-identified and analyses are performed in aggregate, at Einstein Hospital Israelita, Sao Paulo (January 2017 to April 2025). Women aged 40 to 65 were selected by diagnostic codes for menopause-related symptoms and were assessed by internal medicine and emergency medicine physicians in the emergency department. The incidence cohort (each woman's first recorded visit) was the primary unit of analysis, with the visit level used for sensitivity analysis; cardiometabolic conditions were treated as covariates. Admission free-text notes were processed with natural language processing including negation handling, and the diagnoses at reattendances and admissions were used as indications of the nature of the presentations. MAIN OUTCOME MEASURES: Menopause-related diagnostic coding (code N95); emergency department reattendance within 72 h; short-stay admission under 48 h. RESULTS: We analysed 413,414 visits by 97,105 women (median age 45 years at entry). A menopause code was recorded for only 0.79% of women, peaking at ages 50 to 54 (1.6%) and lowest at 40 to 44 (0.34%), consistent with a missed window in early perimenopause. Coding was associated chiefly with the coexistence of other gynaecological codes, namely abnormal uterine bleeding (odds ratio 31.3, 95% confidence interval 26.0 to 37.5) and genitourinary syndrome of menopause (odds ratio 8.2, 95% confidence interval 4.5 to 15.0), together with sleep disorders (odds ratio 2.2), cardiometabolic comorbidity (odds ratio 1.7) and age, with no evidence of problematic multicollinearity. Negation handling substantially lowered text-based symptom frequencies (chest pain 20.5% to 5.6%; dyspnoea 30.3% to 6.8%). Among nonspecific cardiorespiratory presentations, only 1.5% of 72-h reattendances revealed serious cardiovascular disease. In the incidence cohort, reattendance within 72 h occurred in 3.7% of women and short-stay admission in 0.6%; these analyses were exploratory and showed limited discrimination. CONCLUSIONS: Menopause-related diagnostic coding is uncommon in the emergency department among midlife women with compatible symptoms, and it depends more on the presence of other gynaecological codes than on the clinical picture. Because coding is not equivalent to clinical recognition, the findings indicate low menopause-related coding and potential missed opportunities for recognition. The nonspecific presentation typical of early perimenopause receives little coded acknowledgement, at an age when vasomotor symptoms also signal future cardiovascular risk. Based on real-world data and best read as hypothesis-generating, the findings support training for emergency clinicians and active case-finding among midlife women with recurrent nonspecific complaints.