Phenotypic Evidence from a Female Adult Pelvic Pain Clinic

Abstract

Listen

Translate article

Chronic pelvic pain (CPP) in women has complex etiology, multiple pain generators, and significant psychosocial sequalae. Deep phenotyping based on symptoms from multiple organ systems provides an optimal approach to patient-centered evaluation. This study sought to (1) identify distinct CPP profiles based on pain symptoms, cognition, and diagnoses, and (2) evaluate the clinical validity of the identified profiles by examining psychosocial histories and health care utilization. Two-hundred women seeking care for CPP in a tertiary outpatient pain clinic between 2017-2020 were included. Baseline data regarding pain intensity, interference, catastrophizing, acceptance, pain and psychiatric diagnoses were subject to a partition around medoids clustering algorithm to identify potential patient profiles. The number of profiles were determined by maximizing the average silhouette width followed by bootstrapped internal validation. Profiles were compared across demographic characteristics, social history, and subsequent treatment utilization. Two profiles with equal proportion were identified. Profile 1 was vulvodynia and myofascial pelvic pain-dominant characterized by lower pain burden and better psychological functioning. Profile 2 was visceral pain-dominant (e.g., irritable bowel syndrome, endometriosis) featuring higher pain intensity, interference and catastrophizing, lower pain acceptance, and much higher degree of psychiatric comorbidity. Patients in Profile 2 had longer pain duration, higher BMI, disability and substance use, and 2-4 times higher prevalence of childhood and adulthood abuse history (Ps<0.001). They were more likely to receive behavioral therapy (46% vs 27%, P = 0.005) and hormonal treatments (34% vs 21%, P = 0.04), and were prescribed more classes of medications for pain management (P = 0.045). Using a symptom and diagnosis-based phenotyping approach, women with CPP could be separated into two groups distinguished by pain clusters, pain cognition, and psychosocial comorbidities. The identified profiles were reflective of social history and health care utilization, and may inform future etiologic, mechanistic, and interventional research. Chronic pelvic pain (CPP) in women has complex etiology, multiple pain generators, and significant psychosocial sequalae. Deep phenotyping based on symptoms from multiple organ systems provides an optimal approach to patient-centered evaluation. This study sought to (1) identify distinct CPP profiles based on pain symptoms, cognition, and diagnoses, and (2) evaluate the clinical validity of the identified profiles by examining psychosocial histories and health care utilization. Two-hundred women seeking care for CPP in a tertiary outpatient pain clinic between 2017-2020 were included. Baseline data regarding pain intensity, interference, catastrophizing, acceptance, pain and psychiatric diagnoses were subject to a partition around medoids clustering algorithm to identify potential patient profiles. The number of profiles were determined by maximizing the average silhouette width followed by bootstrapped internal validation. Profiles were compared across demographic characteristics, social history, and subsequent treatment utilization. Two profiles with equal proportion were identified. Profile 1 was vulvodynia and myofascial pelvic pain-dominant characterized by lower pain burden and better psychological functioning. Profile 2 was visceral pain-dominant (e.g., irritable bowel syndrome, endometriosis) featuring higher pain intensity, interference and catastrophizing, lower pain acceptance, and much higher degree of psychiatric comorbidity. Patients in Profile 2 had longer pain duration, higher BMI, disability and substance use, and 2-4 times higher prevalence of childhood and adulthood abuse history (Ps<0.001). They were more likely to receive behavioral therapy (46% vs 27%, P = 0.005) and hormonal treatments (34% vs 21%, P = 0.04), and were prescribed more classes of medications for pain management (P = 0.045). Using a symptom and diagnosis-based phenotyping approach, women with CPP could be separated into two groups distinguished by pain clusters, pain cognition, and psychosocial comorbidities. The identified profiles were reflective of social history and health care utilization, and may inform future etiologic, mechanistic, and interventional research.