Data-driven diagnosis and clinical presentation of high-tone pelvic floor dysfunction

In this cross-sectional study of patients presenting to a referral center for chronic pelvic pain, summative score of ≥12/60 tenderness to palpation on a standardized examination of 6 pelvic floor muscle groups (bilateral pubococcygeus, iliococcygeus, obturator internus) demonstrated good sensitivity and specificity for presence of HTPFD, corresponding to patient-reported reproduction of pain with palpation of pelvic floor muscles. This method showed good performance for classification of HTPFD compared to previously described methods. Patients with HTPFD were more likely to report more persistent pain (≥14 days per month) that was exacerbated by physical activity, radiated to other regions of the body, and demonstrated substantially worse sexual function compared to patients without HTPFD. This analysis builds on prior work standardizing assessment of pelvic floor muscles to develop a data-driven assessment strategy for HTPFD. Meister et al described a standardized examination technique with excellent reproducibility and agreement among 4 examiners. 13 This technique assessed 2 muscle groups bilaterally, including obturator internus muscles and single site along the levator ani muscles. Given focus on describing examination technique and reproducibility, no recommendation was made regarding findings that warranted diagnosis of HTPFD. There is substantial overlap with the technique described in Meister and the technique used in this analysis, with the notable difference that we examined 2 discrete sites of levator ani complex (pubococcygeus/anterior levator ani and iliococcygeus/posterior levator ani) to align our technique with other validated work. 11 , 14 , 15 , 22 Diagnosis of HTPFD varies widely in existing literature, with substantial variability in number and location of pelvic floor muscle groups and minimum threshold of pain intensity to assign diagnosis. 8 , 9 , 11 , 13 , 15 Improving clinical recognition of this condition depends on development of a data-driven consensus around assessment strategy, diagnostic criteria, and clinical presentation of HTPFD. The ≥12/60 pelvic floor myofascial tenderness threshold identified in this analysis was more strongly associated with patient-reported pain reproduction than previously described methods. This analysis is the most comprehensive description of primary patient-reported symptoms associated with HTPFD published to date. Consistent with existing literature, patients with HTPFD were much more likely to report dyspareunia and restriction of sexual activity compared to those without HTPFD. It should be noted that we did find broad overlap in symptom patterns despite statistically significant differences, such that symptom patterns alone are not sufficient to distinguish between patients with and without HTPFD. Our finding that rates of other chronic pelvic pain conditions, such as endometriosis, IC/BPS, and irritable bowel syndrome, did not differ among patients with and without HTPFD is notable. If HTPFD were merely a manifestation of other pelvic pain conditions, we would expect to see that rates of these overlapping pain conditions to be substantially higher among those with HTPFD. Similar rates of other pelvic pain conditions support that HTPFD is a separate and distinct condition rather than a symptom or clinical phenotype of another condition. HTPFD very frequently co-occurs with other chronic pelvic pain conditions, including endometriosis (60% prevalence) 8 , 9 and IC/BPS (80% prevalence). 11 However, existing literature has often considered pelvic myofascial tenderness as a symptom of endometriosis or IC/BPS rather than a stand-alone diagnosis. 8 , 9 , 11 Inadequate identification of all contributing pain conditions, including HTPFD, represents a missed opportunity to identify important confounding factors when evaluating treatment efficacy. Our findings add further support to the relationship between HTPFD and nociplastic pain, which has been described in existing literature. In patients with chronic pelvic pain, degree of nociplastic pain strongly corresponded to severity of pelvic floor myofascial tenderness to palpation. 28 Patients with IC/BPS who reported pelvic floor myofascial tenderness to palpation demonstrate greater degree of nociplastic pain and widespread pain. 11 Patients with high degree of nociplastic pain may not respond as robustly to peripherally directed treatments, such as physical therapy or surgery, and may benefit from multimodal treatment approach with addition of centrally directed strategies. 32 – 35 Future research in HTPFD should consider nociplastic pain as potential phenotypic factor that may guide personalized treatment strategies. Strengths of this study include large patient series drawn from tertiary referral center with extensive expertise in management of chronic pelvic pain. Pelvic examination was standardized due to teaching and supervision from a single senior physician with physical examinations performed by 10 different physicians during the study period, which increases validity and generalizability of the examination technique. However, our tertiary referral population may not be generalizable, and thus the threshold for HTPFD should be replicated in other clinic populations. It should be noted that this study design did not allow for the assessment of inter-rater reliability as each patient underwent a single examination with a single provider. However, there were no significant differences between providers in post hoc pairwise comparison testing. An additional limitation is that we used a semi-quantitative guideline for the application of pressure during palpation of pelvic floor muscles, instructing examiners to use enough pressure to cause their nailbed to blanch slightly. Inherent variability is to be expected between examiners based on this method, although we felt that this was as standardized a method as was possible in a clinical setting. For reference, capillary refill test (pressure applied to fingertip to cause blanching) typically requires 10 N of pressure, corresponding to approximately 1 kg. Tu et al 36 demonstrated that approximately 1.5 kg of pressure (measured by fingertip algometer) was needed to elicit a pressure-pain detection threshold on palpation of pelvic floor muscle in women without chronic pelvic pain. While quantification of applied pressure was outside of the scope of this study, we suspect that most examiners used between 1 and 1.5 kg of pressure based on these data. Finally, we have chosen to use the term “high-tone pelvic floor dysfunction” to describe this condition, although we did not formally assess muscle tone as part of our physical examination. This condition suffers from the use of multiple different titles (pelvic myofascial pain, pelvic floor myofascial tenderness, etc.). We felt that it was important to use a term that emphasizes dysfunction of the pelvic myofascial structures rather than solely tenderness or pain. Consensus regarding terminology is desperately needed and would improve ability of clinicians, researchers, and patients to more easily find information regarding this condition.

Between July 2019 and January 2023, 684 patients attended a new patient visit. Of these, 42 opted to defer pelvic examination and 30 patients requested that examination be stopped prior to completion and were therefore excluded from this analysis, leaving 612 patients with complete pelvic examination ( Figure 3 ). The ROC curve used to identify pelvic myofascial tenderness diagnostic threshold score that corresponded to pain reproduction with pelvic floor muscle palpation demonstrated an AUC of 0.900 (95% CI, 0.873–0.926), indicating good discriminative ability. Youden’s index was used to identify optimal threshold which maximized sensitivity and specificity. Summative score of ≥12/60 corresponded to sensitivity 82.3% and specificity 79.1%. Secondary ROC analysis assessed diagnostic threshold score that corresponded to physician diagnosis of HTPFD. The resulting model demonstrated an AUC of 0.858 (95% CI, 0.816–0.900) and Youden’s index indicated an identical threshold score. The ≥12/60 cutoff score corresponded to 76.2% sensitivity and 85.5% specificity ( Figure 4 ). The study design did not allow for assessment of inter-rater reliability as each patient underwent examination with a single provider. However, we assessed for effect of examiner on tenderness ratings via Kruskal-Wallis test. We performed post hoc pairwise comparisons using Mann-Whitney U tests with Bonferroni adjustment for multiple comparisons, which indicated no significant pairwise differences in average tenderness score between providers after adjustment (all P >.05). Performance of this ≥12/60 tenderness score was then compared to 3 previously described diagnostic thresholds for HTPFD relative to patient-reported pain reproduction. The current method showed the strongest relationship with pain reproduction on examination by all examined metrics, except for specificity, although it should be noted that the CI around the AUC for the current method and that used by Shafrir et al heavily overlapped ( Table 1 ). The current method showed the strongest agreement with methods used by Shafrir et al 9 (tenderness ≥4/10 on 2 of 6 muscle groups), followed by much lower agreement between the “any tenderness ≥1/10” methods ( Supplemental Table 1 provides measures of agreement). Using pelvic myofascial tenderness score of ≥12/60 to define presence of HTPFD, 389 (63.6%) patients in this cohort were categorized as having HTPFD and 223 (36.4%) as no HTPFD. Patients with HTPFD demonstrated mean pelvic myofascial tenderness score of 28.2±11.68 compared to 4.27±3.81 for patients without HTPFD ( P <.001). Patients with HTPFD were younger (34.9±8.9 vs 38.1±9.2, P <.001) and more likely to be nulliparous (58.9% vs 49.8%, P =.029) than patients without HTPFD. Patients with HTPFD were more likely to report prior diagnosis of low back pain (34.7% vs 22.0%, P =.001) or migraine/chronic tension headache disorder (39.8% vs 30.0%, P =.017). However, there were no differences in prior diagnoses of chronic pelvic pain conditions, including endometriosis, interstitial cystitis/bladder pain syndrome (IC/BPS), and irritable bowel syndrome. Patients with HTPFD had higher degrees of nociplastic pain as assessed by Fibromyalgia survey score (11.65±4.19 vs 9.56±4.3, P <.001) and were more likely to report widespread pain (82.0% vs 67.7%, P <.001) ( Table 2 ). We observed differences across a broad range of pain descriptive characteristics and pain exacerbating and alleviating factors ( Table 3 ). Patients with HTPFD were more likely to describe pain as heavy feeling in the pelvis (OR, 2.00; 95% CI, 1.43, 2.79; P <.001) and experience persistent pain (≥14 pain days/month; OR, 2.32; 95% CI, 1.60, 3.35; P <.001). They were more likely to report pain exacerbation with physical activity, including exercise (OR, 1.69; 95% CI, 1.21, 2.32; P =.002) and walking (OR, 1.85; 95% CI, 1.32, 2.60; P <.001), and less likely to report pain alleviation with nonsteroidal anti-inflammatory drugs (OR, 0.55; 95% CI, 0.39, 0.78; P =.001). They were more likely to report that pain radiated to other body regions and experience nausea and vomiting associated with pain (all P <.001) ( Figure 5 ). Patients with HTPFD reported clinically and statistically significantly worse pelvic pain and sexual pain symptoms ( Table 4 ). Relative to sexual pain symptoms, patients with HTPFD were more likely to report that they “always” experience pain with intercourse (OR, 2.20; 95% CI, 1.45, 3.35; P <.001) and that pain that occurred during intercourse and in the 24 hours following intercourse (OR, 2.06; 95% CI, 1.46, 2.92; P <.001) compared to those without HTPFD. Severity of pain during intercourse was greater for patients with HTPFD than those without (6.55±2.54 vs 5.46±2.60, P <.001). Patients with HTPFD were more likely to report that they had interrupted/stopped or avoided intercourse due to pain in the last year (all P <.001) ( Figure 4 ).

This is a cross-sectional study of consecutive patients who presented to a chronic pelvic pain referral center between July 2019 and January 2023. Patients completed a detailed questionnaire through the electronic medical record portal prior to scheduled visit, which included medical/surgical history, comprehensive symptom assessment, and multiple validated questionnaires assessing pain symptoms and phenotypic factors.Questionnaires were completed electronically and results were automatically transferred to a secure encrypted Research Electronic Data Capture database. 20 Patients reported self-identified race and ethnicity. Patients reported prior diagnoses of chronic overlapping pain conditions. 21 Diagnosis of endometriosis was based on reported history of endometriosis confirmed either by surgery or imaging (endometrioma or magnetic resonance imaging indicating deep infiltrating endometriosis). Clinical documentation for visits occurs via standardized electronic medical record template used by all providers. Pelvic examination findings and physician-reported diagnoses were abstracted from individual charts using standard case report form. We included all new patients who had completed pelvic floor examination at their first visit, only excluding those who deferred examination or requested that examination be stopped prior to completion. To explore differences in clinical presentation associated with this condition, we first used a data-driven approach to identify diagnostic threshold of pelvic floor myofascial tenderness to categorize the presence of HTPFD in binary fashion. Patients underwent detailed physical examination at their first visit, which included standardized pelvic floor myofascial assessment. Patients were able to decline pelvic examination or ask that examination be stopped early, in which case they were excluded from analysis. Pelvic floor muscles ( Figure 1 ), including bilateral pubococcygeus (anterior levator ani), iliococcygeus (posterior levator ani), and obturator internus, were palpated transvaginally with a single digit using enough pressure to cause the nailbed to blanch slightly. We have included a figure ( Figure 2 ) from Meister et al 13 which nicely demonstrates examination technique. This figure demonstrates palpation of the (A) obturator and (B) iliococcygeus muscles. Pubococcygeus muscle is the area just distal to the iliococcygeus in the B figure. We elected not to include evaluation of coccygeus or piriformis muscles because the average distance from introitus to examiner’s index fingertip is unlikely to reliably reach these muscles in most patients, particularly those with larger body habitus. For each of the 6 muscle groups, patients were asked to indicate whether they experienced pressure or pain, and intensity if pain was present (range, 1–10). Scores for all 6 pelvic floor muscle sites were summed (range, 0–60). Patients were asked whether palpation of any muscle group reproduced typical pain symptoms (pain reproduction, yes/no). This examination technique strongly aligns with the standardized technique described in Meister et al, 13 with the notable difference that we assessed 2 discrete locations along the levator ani complex to align examination technique with that used in Multidisciplinary Approach to the Study of Chronic Pelvic Pain Research Network, World Endometriosis Research Foundation examination consensus statement, and other validated work. 11 , 14 , 15 , 22 All providers (n=10) who completed pelvic examinations over the study period had completed fellowship or were in fellowship training for minimally invasive gynecologic surgery. All providers were trained on this standardized examination by a single senior physician. Fellowship trainees completed pelvic examinations with direct supervision from this senior physician to ensure consistent examination procedures. All providers use standardized template to document examination findings. Physicians listed diagnoses suspected to contribute to each patient’s pain symptoms in clinical documentation. During the study period, physician diagnosis of HTPFD was determined based on clinical judgment incorporating patient history, symptoms, and examination findings, but was not based on any threshold score on pelvic floor examination. To identify a pelvic myofascial tenderness score that distinguished between presence and absence of HTPFD, we performed receiver operating characteristic (ROC) curve analyses, electing to define diagnosis of HTPFD based on patient-reported pain reproduction in response to palpation of pelvic floor muscles. Summative pelvic floor tenderness score obtained from standardized exam detailed above was used as test variable. Area under the curve (AUC) was computed to assess the ability of pelvic floor tenderness score to accurately distinguish pain reproduction on palpation of pelvic floor muscles. Youden’s index was examined to identify optimal diagnostic threshold score that maximized both sensitivity and specificity in the model. Based on these analyses, we classified patients above (with HTPFD) and below (without HTPFD) the threshold score, using these groups for comparison in subsequent analyses. We secondarily aimed to evaluate how the pain reproduction score corresponded to the physician’s diagnosis of HTPFD. We conducted identical ROC curve analyses with physician diagnosis of HTPFD as a test variable. We then compared performance of the pelvic myofascial tenderness score derived above to 3 previously described diagnostic thresholds for HTPFD relative to patient-reported pain reproduction: (1) any tenderness (≥1/10) on any of 4 muscle groups (bilateral pubococcygeus, iliococcygeus), 8 , 16 (2) any tenderness (≥1/10) on any of 6 muscle groups (bilateral pubococcygeus, iliococcygeus, obturator internus), 11 and (3) tenderness ≥4/10 on 2 of 6 muscle groups (bilateral pubococcygeus, iliococcygeus, obturator internus). 9 We constructed binary variables for HTPFD according to each of these methods and pelvic myofascial tenderness threshold examined in the current analyses. We compared classification accuracy with pain reproduction on examination as the “gold standard” for each of these methods. These included Kappa measure of agreement and overall AUC, sensitivity, and specificity. Patients described pain pattern, characteristics, exacerbation factors, and timing and severity of pain related to sexual activity on new patient questionnaire. Scores (0–10) regarding pain with bladder function, bowel function, menses, and intercourse were reclassified as categorical (≥4/10 vs ≤3/10), based on threshold established in prior studies as indicative of clinically meaningful levels of pain. 23 , 24 Interstitial cystitis symptom index was assessed in continuous fashion (higher scores, worse urinary symptoms). 25 Nociplastic pain, also referred to as central sensitization, was assessed by American College of Rheumatology Fibromyalgia survey score (0–31, comprised of widespread pain index [0–19] and symptom severity score [0–12]). 26 This measure assesses nociplastic pain on a continuum, with higher scores reflecting greater degree of nociplastic pain activity, 27 and has previously demonstrated excellent reliability in a chronic pelvic pain population. 28 We also measured the prevalence of widespread pain, defined as reported pain in ≥3/7 nonpelvic body regions. Widespread pain is a key feature of nociplastic pain and has been used as a simple clinical proxy for nociplastic pain in chronic pelvic pain conditions. 29 , 30 Bivariate comparisons were conducted to explore differences in symptom presentation associated with HTPFD. To evaluate whether patients with HTPFD differed statistically ( P <.05) from patients without HTPFD, we used t tests and chi-squared tests for continuous and categorical data, respectively. For categorical variables with more than 2 response options (ie, pain pattern, timing of pain with intercourse, “select all that apply” variables), we created proxy variables for comparisons between those who endorsed a given response option and those who endorsed all other valid response options (omitting missing responses “prefer not to answer” and “not applicable”). Distributions of continuous variables were examined (eg, skew, kurtosis) to verify that they met assumptions necessary for analysis. For example, we used cutoffs of |2| and |3| for skewness and kurtosis, respectively, and applied Levene’s test to assess for presence of heteroscedasticity. Welch’s t test was used in place of Student’s if the latter was present. Unadjusted odds ratios (ORs) with 95% confidence intervals (CIs) were calculated to measure the magnitude, direction, and statistical significance of association between the presence of HTPFD and pain characteristics. This study was approved by the Institutional Review Board of the University of Michigan (IRB# HUM00003797). All analyses were performed with SPSS version 29.0.1.0. 31 Study reporting adhered to Strengthening the Reporting of Observational Studies in Epidemiology guidelines.

This analysis builds on prior work in standardizing assessment of HTPFD and makes important advancements toward a data-driven diagnostic evaluation for this underdiagnosed condition. HTPFD should be considered when patients report summative tenderness of ≥12/60 on palpation of 6 pelvic floor muscles, although future studies should evaluate whether this criterion is valid in other patient populations and whether it corresponds to treatment response to pelvic floor physical therapy and other treatment modalities. HTPFD is associated with more persistent pain that may radiate to other regions of the body and is exacerbated by physical activity. Dyspareunia and impaired sexual function appear to be hallmark symptoms of HTPFD.

High-tone pelvic floor dysfunction (HTPFD) has long been recognized as a condition that frequently contributes to pelvic pain symptoms 1 – 3 and has been referred to using various terms including pelvic myofascial pain, pelvic floor muscle tenderness, levator ani syndrome, and myofascial pelvic floor dysfunction, among others. HTPFD is characterized by pelvic floor muscles that are hyper-contractile, tender to palpation, and unable to perform a normal range of dynamic physiologic actions, such as contraction, relaxation, and appropriate coordination. 4 HTPFD may be idiopathic with seemingly spontaneous presentation, but may occur in response to painful stimuli within the pelvis or abdominal core. 4 Prevalence of HTPFD in the general population is unknown and is likely underdiagnosed, but has been reported in approximately 60% of patients who present with chronic pelvic pain symptoms and frequently co-occurs with other pelvic pain conditions. 4 – 11 Despite widespread agreement that this condition is highly prevalent and impactful, there is no consensus as to standardized assessment strategy or diagnostic criteria. A recent systematic review summarized the substantial heterogeneity in assessment. 12 While several authors have detailed their examination technique, 13 , 14 there is currently no data-driven consensus regarding examination findings that indicate the presence of HTPFD. Number and location of pelvic floor muscle groups and minimum threshold of pain intensity to assign this diagnosis are highly variable across existing literature. Furthermore, there are scant primary data regarding clinical presentation of HTPFD. Dyspareunia has been strongly associated with HTPFD in several studies, 8 , 15 , 16 but most literature regarding other symptoms is based on expert opinion rather than primary or patient-reported data. 17 , 18 Lack of data-driven diagnostic assessment and patient-reported symptoms associated with HTPFD create substantial challenges relative to improving clinical identification of this condition and evaluating efficacy of treatment strategies. 4 A recent systematic review noted that the lack of data-driven consensus in defining and diagnosing HTPFD was a major barrier to progress for both clinical research and treatment of this condition. 19 Our objectives were to (1) identify a pelvic myofascial tenderness score that corresponded to patient-reported reproduction of pain, with the goal of standardizing assessment and diagnosis of HTPFD, and (2) compare clinical presentation for patients with and without HTPFD.