Multilevel Risk Factors Linked to Gynecologic Health Care Utilization in Patients With Chronic Pelvic Pain

INTRODUCTION: Chronic pelvic pain (CPP) affects 14% of women and is associated with high utilization of healthcare. CPP frequently results in referral to women’s health services and is the reason for 10% of gynecologist referrals, 12% of hysterectomies, and 40% of gynecologic-related laparoscopies. It is well-documented that patients with CPP are higher healthcare resource utilizers; however, patient-level factors for utilization of surgical gynecology resources has not yet been explored. OBJECTIVE: The objective of the study is to determine patient-level factors related to healthcare utilization in a surgical gynecology department for pelvic pain to develop early risk stratification and institutional interventions at initial consultation for pelvic pain. Early risk stratification can allow for improved patient’s care pathway, outcomes, and leading to optimized and potentially reduced utilization. METHODS: This is a retrospective cohort study using data from 07/01/2020 to 05/31/2023 at an academic tertiary care hospital. Our cohort included females >18 years old with a diagnosis code for pelvic floor tension myalgia (N94.89) or pelvic and perineal pain (R10.2), who received either an internal or external order for physical therapy (PT). Gynecologic surgery visit (GSV) was our primary outcome variable, measured as the number of surgical gynecologic appointments received after pelvic pain diagnosis over a 1-year time period. Age, BMI, ethnicity, race, relationship status, panel status, patient-reported outcomes measurement information system scores, along with the following comorbid conditions: pain, urinary and bowel dysfunction, pelvic organ prolapse, gynecologic conditions, coccydynia, endometriosis, and adenomyosis were pulled from the electronic health record (EHR). A “paneled patient” was a patient with primary care provider at the study institution. We fit a zero-inflated negative binomial (ZINB) regression model in STATA (version 18). Panel status was used in the logit part of the ZINB model to estimate the zero and non-zero count of GSV. All the other variables were used to predict the number of GSV. RESULTS: Among 1172 patients, the median number of GSV was 1 with the majority (67%) having at least one GSV (Figure 1). Panel status was not a significant predictor of zero or non-zero GSV. ZINB regression showed a variety of factors to positively influence the number of GSV including pain, urinary dysfunction, pelvic organ prolapse, and endometriosis (Table 1). ZINB showed coccydynia to negatively influence number of GSV. CONCLUSIONS: Our findings demonstrate that patients with comorbid gynecologic conditions, including pain, urinary dysfunction, pelvic organ prolapse, and endometriosis have greater utilization of GSV. These comorbidities are common overlapping pain conditions associated with CPP and have been consistently linked to greater utilization. Patients with coccydynia had less utilization of GSV; this could be due to more reliance on pelvic floor PT for treatment. Future interventions and research could prescreen for these comorbid risk factors and incorporate prompt referral services to pelvic PT and/or pain management programs for additional treatment. This would allow for improved patient care pathways and pain outcomes, adequate allocation of resources, decreased provider burden, and increased access to care for new consultation services among women seeking surgical gynecologic care.