Validation of a simple body map to measure widespread pain in urologic chronic pelvic pain syndrome: A MAPP Research Network study

Tables 1 and 2 present the prevalence and reliability of widespread pain by body region, separately for regions in the MAPP II body map pain severity threshold (≥1, ≥4, ≥7). Overall, 76.7% of the participants (73.2%% of the men, 78.6% of the women) reported extra-pelvic pain in at least one site. Pain was most common in the lower back, front of the head, legs, and arms regardless of the pain severity cutoff. The ICC was predominantly greater than 0.50 (moderate correlation) with lower values only for the chest and head at all three severity cutoffs. However, the ICC was higher (0.71–0.76) for the 12-point Widespread Pain Index. This indicates that location of the pain in an individual participant may vary from week-to-week, but the degree of widespread pain overall remains relatively stable over the 4-week run-in period. Figure 3 compares the results of a single baseline measure measured at week 4 regressed on the propensity score over the four run-in measures (weeks 0–3) for the 12-point widespread pain scale. The two results are highly correlated (r=0.85, p<0.0001). Figure 4 illustrates the associations between the three body map scores to baseline outcomes of interest. The level of association between outcome and the three body maps scores were similar. All associations between body map scores and outcomes were significant with the exception of the MAPP I body map score with urinary symptoms severity. Figure 5 summarizes predicted probability curves of having one or more COPCs given the MAPP II body map ordinal score based on clinical diagnosis (CP/CPPS or IC/BPS) (left panel) or by sex (right panel). Either model predicts that female participants with 1–2 extra pelvic pain regions had ~70% chance of having a COPC (compared to ~50% for men), and with three or more extra pelvic pain regions had a ~95% chance of having a COPC (compared to ~75% for men). These differentials by UCPPS diagnoses and sex were statistically significant (p<0.01) for each predictive model. Table 2 compares results for the MAPP II body map with or without a pain severity requirement of ≥4. Using a pain severity cutoff of ≥4, a total of 235 participants (42%) would be classified as having pelvic pain only (no widespread pain). However, almost half of these (n=103, or 18% of the total sample) reported ‘subthreshold’ pain (pain present but ≤3 in severity), and most had this subthreshold pain in multiple body regions. At the bottom of the table, we propose a scoring system that distinguishes this subthreshold group from those with no widespread pain whatsoever.

Details of the two MAPP multisite longitudinal protocols, the Epidemiology and Phenotyping Study conducted in the first phase of MAPP (MAPP I) and the Symptom Patterns Study conducted in the second phase of MAPP (MAPP II), have been previously described. [ 3 ; 14 ] In MAPP I, 233 women and 191 men with UCPPS participated in a 12-month cohort study and completed body map assessments indicating the degree of pain experienced throughout the body. In MAPP II, 378 women and 190 men with UCPPS completed body map pain assessments as part of a 36-month cohort study. Detailed baseline data of the participants, including demographics, urological symptoms, non-urological symptoms and psychosocial co-morbidities have been previously published [ 3 , 14 ]. The MAPP I body map included 45 body sites that were collapsed into 8 anatomic regions (1 pelvic region and 7 extra-pelvic regions; see Figure 1 ).[ 12 ] The degree of widespread pain (Widespread Pain Index) was defined as the sum of all non-pelvic regions with pain (range 0–7). All participants completed a baseline body map pain assessment that inquired about pain that had been present in the past week. This body map did not assess pain severity, and it included a large ‘pelvic’ region that encompassed the entire abdomen, including the upper abdominal quadrants. The MAPP investigators considered these potential shortcomings in designing the MAPP II body map. We modified the 74-site CHOIR body map by dividing the abdominal region into upper and lower quadrants; this yielded a total of 76 body sites for the MAPP II body map ( Figure 2 ). To simplify the map, these 76 sites were collapsed into 13 anatomic regions (1 pelvic region, 12 extra-pelvic regions; see Figure 6 ). The degree of widespread pain (Widespread Pain Index) was defined as the sum of all non-pelvic regions with pain (range 0–12). Participants also rated the pain severity (0–10) for each painful site. Two thresholds of pain severity (≥1, ≥4) were used to define the presence of pain. The threshold of ≥4 was chosen because these scores have correlated with moderate or severe pain-related interference with functioning in patients with chronic pain.[ 1 ] In the MAPP II study, we administered the body map at four weekly contacts (weeks 0,1,2,3) during study run-in, prior to the baseline visit (week 4), to assess the short-term reliability of the measurements. The intra-class correlation coefficient (ICC) was calculated across the 4 run-in periods for each of the 12 regions, as well as for the composite 12-point Widespread Pain Index. For each MAPP II study participant, the region-specific propensity for pain was estimated as the conditional proportion of times during the run-in period (weeks 0,1,2,3) that the participant repeatedly reported pain in each region, and the propensity for the Widespread Pain Index (0–12) was estimated as the average body map index score during the run-in period. Three body map versions were analyzed, including the 8-region MAPP I body map and the 13-region MAPP II body map (both with any level of pain), and the 13-region MAPP II body map requiring at least a pain severity score ≥4. To calculate associations of baseline body map scores to other baseline measures, generalized linear models that included baseline body map score as predictor were used. The baseline body map score in the models did not include the pelvic region and was scored as a continuous measure between 0–7 for MAPP I and 0–12 for both MAPP II body map scores. The slope coefficients for the body map scores represent the change in the outcome given a one-point increase in body map score. All outcomes used for comparison (shown in Figure 4 ) included a question about global non-urologic pain (“Please rate the overall severity of any persistent pain symptoms that were NOT UROLOGIC OR PELVIC PAIN SYMPTOMS (e.g., back pain, headache, etc.) over the past week” (0–10)), and the Complex Multi-Symptom Inventory (CMSI).[ 21 ] The CMSI is a 41-item symptom checklist specific to functional syndromes. The total CMSI score represents the overall symptom burden from somatic symptoms or bodily pain. Three CMSI subscores can be calculated, which are not mutually exclusive. A total of 34 items assess symptoms outside of the pelvis (Non-Urologic subscale), including 18 items that cover non-specific somatic or functional symptoms (Somatic Awareness subscale), and 4 items measure sensory sensitivity to non-painful environmental stimuli (Sensory Sensitivity subscale). The CMSI also includes diagnostic modules for other functional syndromes, including fibromyalgia (FM), chronic fatigue syndrome (CFS), irritable bowel syndrome (IBS), migraine, and temporomandibular joint disorder (TMD). The probability of having at least one of these 5 chronic overlapping pain conditions (COPCs) given the ordinal categorical body map score of 0 Regions, 1–2 Regions, or 3 or more regions of pain was calculated. The pelvic region was not included in determining the value of the ordinal variable. The associations between body map scores were also tested with baseline measures of UCPPS symptom severity (pelvic pain severity and urinary symptom severity)[ 7 ], and with psychosocial measures (SF-12, Patient Reported Outcomes Measurement Information System (PROMIS) sleep disruption, PROMIS fatigue scale, Perceived Stress Scale (PSS), and the Hospital Anxiety and Depression Scale (HADS).[ 2 ; 5 ; 19 ; 20 ] Intraclass Correlation Coefficients (ICCs) were estimated for each of the 12 body map regions, as well as for the overall non-pelvic composite Widespread Pain Index (0–12), using indicators of pain reported at thresholds of ≥1, ≥4, ≥7 at each of the four run-in weeks, by implementing a one-way components of variance model for categorical data, in which participant is incorporated as a random effect. An ICC for repeated measures of the same outcome above 0.6 is considered an indicator of substantial reliability, and conversely, an indicator of low within-participant variability.[ 13 ] The propensity for widespread pain was estimated as the average Widespread Pain Index (0–12) during the run-in period (weeks 0,1,2,3). As a result, a simple linear regression of the baseline level of widespread pain (week 4) on the propensity for widespread pain (weeks 0–3) was estimated to indicate the strength of association, as well as the slope estimate indicating the attenuation of the baseline index relative to the average index during the run-in period. Correlation between body map scores and other baseline measures were assessed with generalized linear models in which the outcomes of interest were treated as continuous measures. The associations between the outcome and each of the three body map scores were assessed in separate models in which the body map score of interest was included as the only predictor in the model. The slope estimate for each model, as well as the corresponding 95% confidence intervals are presented in Figure 4 . The interpretation of the slope coefficient is the change in the outcome given a one unit increase in the body map score. Confidence intervals that do not overlap the line at x=0 in Figure 4 are indicative of statistically significant associations between outcome and body map score. Finally, the likelihood of each participant having at least one of the 5 COPCs was estimated by fitting a logistic regression model with the 3-level ordinal body map index as predictor, coded 0: none, 1: 1–2 regions and 2: 3+ regions to an indicator variable coded 0: no COPCs or 1: Any of the 5 COPCs. Differential widespread pain effects for participants with UCPPS diagnoses of CP/CPPS Only vs IC/BPS and Male vs Female Sex were also investigated.

Our analysis demonstrates that a single assessment of widespread pain with a 13-region body map has adequate reliability over four weeks and substantial validity when compared to other widespread pain measures. More specifically, the exact locations of pain did vary from week to week (ICC values ~0.50–0.65), but the degree of overall widespread pain was stable (ICC values ~0.71–0.76). Furthermore, the widespread pain scores obtained using a propensity score methodology were highly correlated (r=0.85) with a single baseline value. These findings indicate that a single body map measurement is sufficient to characterize widespread pain in UCPPS patients in the clinical setting. Given the variability in the location of pain from week to week, researchers may want to consider obtaining repeated body map assessments if the exact pain location is important to their research questions.[ 6 ] We found that the addition of pain severity ratings (pain score ≥4) to the body map assessment provided only limited additional information (slightly higher correlations) that may pertain more to research than the clinic. The additional effort needed to collect levels of pain does not appear to be necessary in the clinical management of UCPPS patients. From a research perspective, some of the associations with other baseline measures were made slightly greater after adding pain severity to the body map calculations ( Figure 4 ). It is also notable that almost 20% of participants reported pain below the threshold value of 4. This ‘subthreshold’ group may be different from the 23% of participants who reported no widespread pain whatsoever. Since some of these subtle differences may be important in research studies that are exploring novel pathophysiologic mechanisms, we would encourage the use of the 76-site, 13-region body map with pain severity for future research studies, where more precise measures of pain location and/or severity are needed (e.g., correlating widespread pain with biomarker or neuroimaging studies). However, the addition of pain severity to the body map does not appear necessary for routine clinical care, as the streamlined body map works well to identify clinically important widespread pain. This is underscored by the finding that the presence of widespread pain (regardless of pain severity) predicted the presence of a clinically significant COPC ( Figure 5 ). In Figure 6 , we provide our recommendation of a simplified version of the CHOIR body map that could be used by clinicians who treat UCPPS. This body map may be freely used using this paper as the citation. The body map can essentially serve as a brief review of symptoms focused on pain location. It can be complementary to the history in that it assesses for current symptom rather than prior diagnoses which may be subject to recall bias and/or may not reflect current symptoms. It could also potentially be used to screen for pelvic pain in non-UCPPS patients. Our analyses indicate that the MAPP I body map (45 sites, 8 regions) performed almost as well as the MAPP II body map (76 sites, 13 regions) while offering the potential advantage of less complexity. However, we feel that the MAPP II version is preferable. The MAPP II body map ( Figures 2 and 6 ) was slightly modified from a body map that was developed in community-based individuals with diverse chronic pain diagnoses and included both treatment-seeking and non-treatment seeking individuals as part of its development.[ 18 ] The CHOIR body map[ 18 ] from which the MAPP II body map was developed is now considered the reference standard body map for pain research, so UCPPS pain research may be compared to other chronic pain studies. Furthermore, the MAPP II body map has better face validity, as the individual sites (e.g., abdominal quadrants, shoulders) are more consistent with commonly recognized clinical pain sites. Finally, the increased complexity and granularity of data from the MAPP II body map provides more research and phenotyping opportunities, and findings from the 76-site body map can easily be translated to the 13-region clinical version.

These data support the use of a body map to identify the presence of widespread pain and to screen for COPCs in patients with UCPPS. This information is becoming increasingly important in guiding the management of UCPPS. Clinicians are encouraged to routinely use the MAPP II body map and to make appropriate referrals or initiate global treatments for UCPPS patients who are identified to have widespread pain.

Urologic chronic pelvic pain syndrome (UCPPS) is an umbrella term that encompasses interstitial cystitis/ bladder pain syndrome (IC/BPS) in men and women, and chronic prostatitis/ chronic pelvic pain syndrome (CP/CPPS) in men. Approximately 70–80% of UCPPS patients report pain outside the pelvic region.[ 9 ; 12 ; 15 ; 17 ] The presence of widespread pain is associated with more severe UCPPS symptoms, worse psychosocial functioning, increased prior traumatic life experiences, poorer quality of life, and a worsening UCPPS symptom trajectory.[ 9 ; 12 ; 16 ] Furthermore, UCPPS patients with widespread pain demonstrate unique findings on neuroimaging and quantitative sensory testing.[ 4 ; 8 ; 10 ; 11 ] “Widespreadness,” taken as a distinct clinical phenotype in the UCPPS patient may have an impact on treatment response. In this study we describe ongoing efforts by the Multidisciplinary Approach to the Study of Chronic Pelvic Pain (MAPP) Research Network to characterize and standardize the measurement of widespread pain in patients with UCPPS. A 74-site body map was previously developed by the Collaborative Health Outcomes Information Registry (CHOIR)[ 18 ] and is now considered the reference standard body map for pain research. We adapted the CHOIR body map for UCPPS patients, and specifically assessed: 1) the test-retest reliability of body map measurements, 2) the value of including a moderate pain threshold (≥4 out of 10) for each region, 3) the relationship of the body map determination of widespread pain to other pain and illness impact variables, and 4) the validity of a simplified body map for clinical use.