1. Introduction
Chronic pain in children is a common problem,11,18,20 frequently occurring as headaches, abdominal pain, and limb pain, with many patients experiencing widespread chronic pain. Pain is also pervasive in many pediatric chronic diseases, which can vary in severity, frequency, and quality of pain. The International Association of Pain (IASP)22 definition of pain specifically notes that pain is a “sensory and emotional” experience, such that pain comprises multiple facets of human experience and is much more than nociception. Research in pediatric pain measurement has focused primarily on pain intensity. However, a single construct does not describe the multifaceted experience of pain.
Multifaceted assessment of pain can improve communication about pain and help clinicians characterize, differentiate, and treat a patient's unique experience of pain. Pain quality is an important domain of pain, and represents the subjective sensory experiences associated with pain as well as the affective experiences of pain.27 Two pediatric instruments that include pain quality assessment are the Adolescent Pediatric Pain Tool (APPT),25 which includes a checklist of pain quality items, and the Varni–Thompson Pediatric Pain Questionnaire (PPQ),30 which measures the intensity, sensory, emotional, and evaluative components of pain. The PPQ is proprietary, which limits its use, and the utility of the APPT is limited by its scoring system, which simply counts the number of pain qualities reported on the measure. To enhance the assessment of pediatric pain quality, a measure is needed that (1) is developed with psychometrically rigorous methods specifically for children and adolescents with input from pediatric patients experiencing pain, (2) assesses the degree of discomfort associated with a particular pain quality, improving scoring precision, (3) standardizes the multifaceted assessment of pain by scoring the measure in the same metric as other pain domains, and (4) is freely available.
The National Institutes of Health Patient-Reported Outcomes Measurement Information System (PROMIS) initiative3 was created to measure physical, mental, and social health for adult and pediatric populations.4,9,23 Rigorous psychometric methods were deployed to achieve reliable, responsive, and valid assessments of health domains in a common metric. The initial steps for the development of pain quality candidate items included (1) evaluating existing measures, (2) incorporating feedback from expert clinicians and pediatric pain researchers, (3) soliciting qualitative information with focus groups and semistructured interviews of youth with chronic pain, (4) translatability review, and (5) final refinement of items after cognitive review14 of items with patients. The methodology and results of this iterative qualitative process are comprehensively detailed in a prior publication16 and resulted in 59 candidate items of pain quality for large-scale quantitative testing. The aim of the current study was to quantitatively evaluate the measurement properties of these 59 PROMIS pediatric pain quality candidate items. Consistent with the IASP definition, we hypothesized that we would observe 2 dimensions of pain qualities—sensory and affective. We also aimed to develop short form versions, reflecting sensory and affective experiences of pain.
2. Methods
2.1. Procedures
Patients were recruited through outpatient clinics at 3 pediatric medical centers in Ohio, Pennsylvania, and Georgia. This study was approved by the institutional review boards at each institution, and patients and parents provided written informed assent/consent for participation in the study. Study coordinators identified candidates meeting eligibility criteria (see below) based on review of electronic medical records. All assessments were completed in person in an outpatient medical setting. Caregivers answered sociodemographic and health history questions using pencil and paper forms. Children responded to candidate pain quality self-report items as part of a larger pool of PROMIS items pertaining to their pain experience in addition to self-report measures of pain interference, depressive symptoms, fatigue, average pain intensity, and pain catastrophizing. All candidate items and surveys were administered on a laptop computer. Data collection procedures were standardized across the 3 study sites.
2.2. Participants
Participants included children and parents of children with a chronic pain condition (eg, sickle cell disease [SCD], juvenile fibromyalgia [JFM], and juvenile idiopathic arthritis [JIA]). These populations were chosen because of the clinical and content expertise of the investigators, and the variable nature of pain across these disease groups: episodes in SCD; more chronic widespread pain with flares in JFM; and intermittent, but often daily, localized pain in children with JIA. The intent was to have a sample with a diverse pain experience to assess whether PROMIS items were appropriate measures for use across chronic/recurrent pain conditions. To participate, children had to be between 8 and 18 years of age (inclusive), fluent in English, and diagnosed by a physician with a chronic pain condition. Individuals were ineligible if concurrent medical, psychiatric, or cognitive condition(s) would interfere with their participation. During recruitment, candidates were screened for whether they had experienced pain in the past 7 days to ensure that those with recurrent pain conditions were currently or within a few days of experiencing an episode of pain because the recall period of PROMIS measures is typically 7 days.
2.3. Measures
2.3.1. Pediatric pain quality item bank
The initial pediatric pain quality item bank consisted of 59 items that assessed specific physical sensations and affective components associated with pain. Initial development of these items is comprehensively detailed in the study by Jacobson et al.,16 but summarized here. Items were developed or revised based on a literature review, expert feedback, cognitive interviews, focus groups, and semistructured interviews conducted with N = 34 youth aged 8 to 18 years with chronic pain (for full demographic detail on this sample, see the study by Jacobson et al.15,16). Clinical and research experts in the field of pain assessed and revised the initial pool of items generated from the literature review for clarity of concept and language, and to adhere to PROMIS formatting. Interviewers asked participants about the subjective meaning of the domain, each item, and about the participants' own experiences (eg, “What kinds of things come to mind when I say the word______?”). Participants also completed the candidate items for the measure and were asked about comprehension and preferred word choice. Participants discussed difficulties they had completing the items, what they believed the item means, and how they chose their responses. The goal was that the items and response options were written in a way that was mutually understood by patients and the research team. Items were modified or excluded based on the participants' feedback. The pain quality domain contains attributes such as perceived temperature (eg, hot, cold), sensations (eg, throbbing, tingly), and perceived affective qualities of pain (eg, uncomfortable, distressing). Twenty-three of the items measured “affective” aspects of pain quality and were in the format of “In the past 7 days, did your pain ever feel ______” (eg, miserable, unpleasant), with dichotomous response options (yes/no). The remaining 32 items assessed the “sensory” aspect of pain quality and were in the format “In the past 7 days, did your pain ever feel ______” (eg, sharp, achy), with a 5-point ordinal response scale (“not at all” to “very much”). These response options were consistent with the PROMIS adult neuropathic pain quality measure.1 The 7 day recall period was chosen to be consistent with other pediatric pain-related PROMIS measures.
2.3.2. Other measures
2.3.2.1. Patient-reported outcomes measurement information system pediatric measures
This study also measured (1) pain interference (8-item short form v1.0)29 ie, difficulties in completing daily activities, emotional difficulties, and impairment in physical functioning because of pain, (2) depressive symptoms (8-item short form v1.0),13 (3) fatigue (10-item short form v1.0),17 and pain behavior (8-item short form v1.0),8 ie, behaviors that typically communicate to others that an individual is experiencing pain that include observable displays of pain and verbal reports of pain. These measures assessed presence of symptoms in the past 7 days via a 5-point ordinal scale. Response options were “0 = never,” “1 = almost never,” “2 = sometimes,” “3 = often,” and “4 = almost always.” The pain behavior items also contained a “had no pain” response option. Higher scores indicated a higher level of the construct being measured. Cronbach's alphas for the measures in the current study ranged from α = 0.89 (pain behavior) to α = 0.97 (fatigue).
2.3.2.2. Pain intensity numeric rating scale
Overall self-reported pain intensity in the past week was collected using a four-item 0-to-10 numeric rating scale (NRS) based on the Brief Pain Inventory.6 Participants were asked to rate their worst pain, average pain, and least pain in the last week, and pain right now. Mean scores were calculated for the 4 items. Internal consistency of these 4 items for the current study was α = 0.898.
2.3.2.3. Pain catastrophizing
The Pain Catastrophizing Scale—Child Version (PCS-C),7 contains 13 ordinal items assessing pain appraisals that include rumination, magnification, and feelings of helplessness. Response options include not at all (0), mildly (1), moderately (2), severely (3), and extremely (4). Total scores range from 0 to 52, with higher scores reflecting greater pain catastrophizing. Cronbach's alpha for this measure in the current study was α = 0.955.
2.4. Statistical analyses
All analyses were conducted in Mplus 8.419 and Stata 16.26 Descriptive statistics summarizing the demographic characteristics of the sample were calculated. Descriptive statistics (mean, SD, % of responses in each category, and % missing per item) were calculated for each of the pain quality items as well as average pain intensity, pain behavior, pain catastrophizing, depressive symptoms, and fatigue scores.
2.4.1. Item calibration
We conducted confirmatory factor analysis (CFA) to assess the dimensionality of the 59 pain quality items in the initial item bank. A weighted least square mean and variance (WLSMV) estimator was used because of the categorical response options. Model fit was evaluated using empirically validated fit indices and levels suggested by Hu and Bentler12: root mean square error of approximation (RMSEA) < 0.05 for good fit or < 0.08 for adequate fit, comparative fit index (CFI) and Tucker–Lewis Index (TLI) > 0.95 for good fit or > 0.90 for adequate fit, and standardized root mean square residual (SRMR) < 0.08. Item responses were then calibrated using Samejima's graded response model for ordered categorical item responses.24 Item response theory (IRT) parameters (ie, discrimination and thresholds) were estimated for each item. Any item with discrimination parameters that were less than 1.0 or were not statistically significant were excluded from the final item bank.
2.4.2. Short-form development
The goal was to develop 2 versions of the pain quality short form: one pain quality short form focusing on the affective aspect of pain quality, and one version focusing on the sensory aspect of pain quality. Two lists of items (one with a focus on sensory items and one with a focus on affective items) were compiled based on the psychometric properties of the items. Specifically, we selected items with high discrimination parameters (ie, strongly related to the underlying construct), and items that covered a wide threshold range (ie, comprehensive coverage across the range of the construct). After these 2 lists of items were compiled based on the psychometric properties, we solicited feedback from pain management clinicians and researchers using a modified Delphi consensus-building methodology. Six experts in the areas of anesthesia/pain management, pain psychology, hematology/oncology, and rheumatology participated in this consensus-building process. The experts were asked to select a “Top 8” (typical length of the longer PROMIS pediatric short forms) and a “Top 4” (subset of the Top 8) items from each list (ie, list of affective focused items and list of sensory focused items) based on their expert opinion on which of the items were most informative, given their experience (research or clinical) with pediatric patients experiencing pain. Numerical rank order of importance was not required unless the expert participants chose to do so. The results of the first round were compiled together. A second round of voting occurred if the item selection from the first round was nonoverlapping.
2.4.3. Convergent validity and reliability
To assess the convergent validity of the PROMIS pediatric pain quality item bank and the 2 versions of the short form, we examined bivariate correlations with the PROMIS pediatric pain interference short form, pain behavior short form, depressive symptoms short form, fatigue short form, as well as scores on the pain catastrophizing scale and the pain intensity NRS scale. Internal consistency was estimated for both the full item bank and both versions of the short form using Cronbach's alpha.
3. Results
3.1. Sample characteristics
A total of 448 children (mean age = 13.5 years, SD = 2.8) participated in the study. One participant had missing data on all the pain quality items and was excluded from the analyses, resulting in an analytical sample size of N = 447 (see Table 1 for demographics and diagnoses). The majority of patients with JIA and JFM were White and those diagnosed with SCD were predominantly Black. Overall, there were a larger proportion of female participants (71.4%), which is consistent with the higher prevalence of chronic pain in girls, particularly adolescents, and the known female predominance of persons who develop JIA. The subgroup of patients with SCD was gender-balanced (52% female). There were other statistically significant differences in group characteristics, including higher mean age of participants with JFM and higher average pain intensity scores compared with participants with JIA and SCD. In addition, participants with SCD had higher mean pain behavior scores overall, and higher mean pain interference scores compared with participants with JIA.
Table 1 -
Demographic and clinical characteristics by diagnosis groups (N = 448).
|
Juvenile fibromyalgia |
Juvenile idiopathic arthritis |
Sickle cell disease |
Missing diagnosis |
| N |
151 |
175 |
115 |
7 |
| Mean age (SD) |
14.83 (1.96) |
12.83 (2.94) |
12.97 (2.89) |
12.29 (3.35) |
| % Female |
89.4% |
69.1% |
52.2% |
57.1% |
| % White |
84.1% |
89.1% |
0% |
0% |
| % African American |
6.6% |
4.6% |
97.4% |
85.7% |
| % Hispanic |
5.3% |
3.4% |
3.5% |
0% |
| Mean pain NRS score (SD) |
5.13 (2.61) |
2.91 (2.11) |
4.38 (2.55) |
3.17 (1.83) |
| Mean pain behavior score* (SD) |
50.67 (9.21) |
44.94 (9.29) |
56.43 (7.88) |
57.51 (5.52) |
| Mean pain interference score* (SD) |
53.13 (9.43) |
44.50 (9.09) |
54.09 (8.18) |
54.93 (8.82) |
| Mean depressive symptoms score* (SD) |
53.82 (10.55) |
45.83 (7.65) |
51.45 (10.23) |
49.07 (8.65) |
| Mean fatigue score* (SD) |
53.99 (9.84) |
44.31 (8.17) |
53.46 (8.67) |
51.17 (9.50) |
| Mean pain catastrophizing score (SD) |
23.30 (14.00) |
11.50 (11.18) |
24.66 (13.16) |
20.5 (10.82) |
*Reported as T-scores from Patient-Reported Outcomes Measurement Information System pediatric short forms.
NRS, numeric rating scale.
3.2. Item calibration
Contrary to our hypothesis that we would observe 2 domains of pain quality (sensory and affective), instead the unidimensional model fit the data adequately: CFI = 0.934, TLI =0.931, RMSEA = 0.056, and SRMR = 0.074. Based on discrimination parameters less than 1.0, 3 items were excluded from the final item bank. The IRT discrimination parameters for the remaining 56 items ranged from 1.05 to 3.81 and were statistically significant. The category threshold parameters for these 56 items ranged from −1.02 to 3.66. Item response theory parameters for all items are presented in Table 2. Comprehensive item-level descriptive statistics for all 56 items can be found in Appendix A (Supplemental Materials, available at https://links.lww.com/PAIN/B684).
Table 2 -
Discrimination and threshold item response theory parameters from the unidimensional model of the final 56 items included in the patient-reported outcomes measurement information system pediatric pain quality item bank.
| Item stem |
a
|
b1 |
b2 |
b3 |
b4 |
| Unpleasant?*
|
2.06 |
−0.93 |
|
|
|
| Miserable?*
|
2.69 |
0.01 |
|
|
|
| Worrying?*
|
1.69 |
0.30 |
|
|
|
| Agonizing?*
|
3.16 |
0.32 |
|
|
|
| Distressing?*
|
2.30 |
0.25 |
|
|
|
| Excruciating?*
|
3.10 |
0.39 |
|
|
|
| Nauseating?*
|
1.68 |
0.75 |
|
|
|
| Dreadful?*
|
2.70 |
0.28 |
|
|
|
| Horrible?*
|
3.39 |
−0.02 |
|
|
|
| Awful?*
|
3.81 |
−0.12 |
|
|
|
| Sickening?*
|
1.70 |
0.47 |
|
|
|
| Tiring?*
|
1.83 |
−0.48 |
|
|
|
| Uncomfortable?*
|
2.11 |
−1.07 |
|
|
|
| Unending?*
|
2.65 |
0.15 |
|
|
|
| Stressful?*
|
2.41 |
−0.22 |
|
|
|
| Intolerable?*
|
2.49 |
0.31 |
|
|
|
| Cruel?*
|
2.50 |
0.53 |
|
|
|
| Torturing?*
|
3.38 |
0.50 |
|
|
|
| Unbearable?*
|
3.22 |
0.27 |
|
|
|
| Annoying?*
|
2.10 |
−0.90 |
|
|
|
| Nagging?*
|
2.04 |
−0.05 |
|
|
|
| Weird?*
|
1.05 |
0.49 |
|
|
|
| Uncontrollable?*
|
2.03 |
0.10 |
|
|
|
| Tight?†
|
1.91 |
−0.55 |
0.23 |
0.78 |
1.45 |
| Numb?†
|
1.75 |
0.37 |
0.98 |
1.87 |
2.37 |
| Cramping?†
|
1.87 |
−0.14 |
0.61 |
1.15 |
1.87 |
| Pulsing?†
|
2.74 |
−0.06 |
0.46 |
0.97 |
1.60 |
| Tingly?†
|
1.91 |
0.07 |
0.81 |
1.52 |
1.87 |
| Sore?†
|
1.74 |
−1.20 |
−0.39 |
0.24 |
1.03 |
| Hot?†
|
1.53 |
0.09 |
0.97 |
1.49 |
2.39 |
| Sharp?†
|
2.35 |
−0.52 |
0.16 |
0.55 |
1.22 |
| Squeezing?†
|
2.29 |
0.13 |
0.71 |
1.20 |
1.78 |
| Throbbing?†
|
2.14 |
−0.41 |
0.10 |
0.69 |
1.29 |
| Stabbing?†
|
2.53 |
−0.03 |
0.42 |
0.83 |
1.26 |
| Spreading?†
|
2.64 |
−0.08 |
0.46 |
1.06 |
1.53 |
| Stinging?†
|
2.09 |
0.25 |
0.84 |
1.45 |
2.10 |
| Tearing (ripping)?†
|
2.11 |
0.49 |
1.10 |
1.54 |
2.05 |
| Pounding?†
|
2.63 |
−0.17 |
0.34 |
0.87 |
1.38 |
| Shoot to a different part of your body?†
|
1.41 |
−0.11 |
0.59 |
1.35 |
1.90 |
| Splitting?†
|
2.56 |
0.38 |
0.96 |
1.43 |
1.89 |
| Sensitive?†
|
2.28 |
−0.53 |
0.11 |
0.62 |
1.18 |
| Like pins and needles?†
|
2.13 |
0.02 |
0.61 |
1.10 |
1.50 |
| Deep?†
|
2.22 |
−0.40 |
0.13 |
0.65 |
1.25 |
| Constant?†
|
2.38 |
−0.70 |
−0.10 |
0.41 |
0.86 |
| Pulling?†
|
2.24 |
0.30 |
0.93 |
1.49 |
1.99 |
| Electrical?†
|
2.55 |
0.79 |
1.22 |
1.52 |
1.90 |
| Drilling?†
|
2.41 |
0.46 |
0.94 |
1.34 |
1.88 |
| Pinching?†
|
2.27 |
0.28 |
0.86 |
1.50 |
2.15 |
| Steady?†
|
1.63 |
−0.54 |
0.16 |
0.88 |
1.58 |
| Achy?†
|
1.95 |
−1.03 |
−0.40 |
0.22 |
1.03 |
| Move to a different part of your body?†
|
1.61 |
−0.23 |
0.44 |
1.02 |
1.63 |
| Itchy?†
|
1.32 |
1.30 |
1.85 |
2.61 |
3.66 |
| Cold?†
|
1.22 |
1.09 |
1.69 |
2.52 |
3.27 |
| Tugging?†
|
2.32 |
0.36 |
0.93 |
1.60 |
2.04 |
| Tender?†
|
1.94 |
−0.46 |
0.17 |
0.76 |
1.46 |
| Burning?†
|
1.90 |
0.42 |
1.01 |
1.51 |
2.06 |
*Response options: 0 = no; 1 = yes.
†Response options: 1 = not at all; 2 = a little bit; 3 = somewhat; 4 = quite a bit; 5 = very much.
a = discrimination (slope) parameter; b1-b4 = threshold (location) parameters.
3.3. Short-form development
Although we found that the overall pain quality measure is unidimensional, clinicians and researchers may still be interested in assessing affective pain qualities in their pediatric patient populations for treatment planning or to track outcomes. Thus, we opted to create 2 versions of the pediatric pain quality short form—one based on affective pain qualities and one based on sensory pain qualities. Based on statistical factors (high discrimination parameters, ie, strongly related to the underlying construct, and wide threshold range, ie, comprehensive coverage across the range of the construct) and investigative team input (S.K.-Z., K.R.G., D.D.S., E.M.M., and A.C.C.), the list of items to be considered by the Delphi group in the creation of the short forms was narrowed to 33 items (18 sensory items and 15 affective items—see Table 3). After compilation of the votes from the first round of ranking, there was mutual agreement for the 8 items of pain quality focusing on affective aspects (“unpleasant,” “miserable,” “worrying,” “horrible,” “awful,” “unending,” “unbearable,” and “annoying”), and for 7 of 8 items of pain quality focusing on the sensory aspect (“tingly,” “sore,” “sharp,” “throbbing,” “stabbing,” “achy,” and “burning”). To complete the selection of the final item sensory version of the pain quality short form, experts were requested to rank the following 3 items (that had strong but not unanimous consensus in the first round) in order of recommendation for inclusion as the eighth item on the sensory version of the short form: “cramping,” “stinging,” and “tight.” This final round of voting resulted in the eighth item of this short-form version being “tight.” The correlation between the 2 short-form versions was r = 0.83, P < 0.001. The final items included in the 2 versions of the PROMIS pediatric pain quality short form are bolded in Table 3.
Table 3 -
Items considered for inclusion in the affective and sensory versions of the pain quality short form.
| Item stem |
Quality descriptor |
| Affective pain quality items |
|
| In the past 7 days, did your pain ever feel |
Distressing? |
|
In the past 7 days, did your pain ever feel
|
Miserable?
|
| In the past 7 days, did your pain ever feel |
Sickening? |
|
In the past 7 days, did your pain ever feel
|
Awful?
|
| In the past 7 days, did your pain ever feel |
Uncontrollable? |
|
In the past 7 days, did your pain ever feel
|
Unpleasant?
|
| In the past 7 days, did your pain ever feel |
Excruciating? |
|
In the past 7 days, did your pain ever feel
|
Horrible?
|
|
In the past 7 days, did your pain ever feel
|
Worrying?
|
|
In the past 7 days, did your pain ever feel
|
Unbearable?
|
| In the past 7 days, did your pain ever feel |
Torturing? |
|
In the past 7 days, did your pain ever feel
|
Annoying?
|
|
In the past 7 days, did your pain ever feel
|
Unending?
|
| In the past 7 days, did your pain ever feel |
Tiring? |
| In the past 7 days, did your pain ever feel |
Stressful? |
| Sensory pain quality items |
|
|
In the past 7 days, did your pain feel
|
Sharp?
|
| In the past 7 days, did your pain feel |
Like pins and needles? |
| In the past 7 days, did your pain feel |
Squeezing? |
|
In the past days, did your pain feel
|
Achy?
|
|
In the past 7 days, did your pain feel
|
Burning?
|
|
In the past 7 days, did your pain feel
|
Tingly?
|
|
In the past 7 days, did your pain feel
|
Sore?
|
|
In the past 7 days, did your pain feel
|
Tight?
|
| In the past 7 days, did your pain feel |
Numb? |
| In the past 7 days, did your pain feel |
Cramping? |
|
In the past 7 days, did your pain feel
|
Throbbing?
|
|
In the past 7 days, did your pain feel
|
Stabbing?
|
| In the past 7 days, did your pain feel |
Sensitive? |
| In the past 7 days, did your pain feel |
Steady? |
| In the past 7 days, did your pain feel |
Cold? |
| In the past 7 days, did your pain feel |
Stinging? |
| In the past 7 days, did your pain feel |
Electrical? |
| In the past 7 days, did your pain feel |
Tender? |
Experts were asked to select their top 8 and their top 4 items (a subset of their top 8).
Final items included on each version of the short form are bolded.
The following items overlap with the Patient-Reported Outcomes Measurement Information System adult nociceptive pain quality measure: “achy” and “sore.”
The following items overlap with the Patient-Reported Outcomes Measurement Information System adult neuropathic pain quality measure: “tingly.”
3.4. Convergent validity and reliability
Scores from the 56-item PROMIS pediatric pain quality item bank were significantly correlated with scores on PROMIS pediatric pain behavior, r = 0.73, P < 0.001, PROMIS pediatric pain interference, r = 0.77, P < 0.001, PROMIS pediatric depressive symptoms, r = 0.58, P < 0.001, and PROMIS pediatric fatigue, r = 0.71, P < 0.001. The 56-item pain quality item bank scores were also significantly correlated with scores on the NRS, r = 0.74, P < 0.001, and the PCS, r = 0.68, P < 0.001. Correlations between these measures and the 2 versions of the pain quality short form were similar in magnitude, as expected (Table 4).
Table 4 -
Bivariate correlations between the patient-reported outcomes measurement information system pediatric pain quality affective and sensory short forms and other measures for convergent validity.
|
Sensory version of short form |
Pain interference*
|
Depressive symptoms*
|
Fatigue*
|
Pain intensity NRS |
Pain catastrophizing scale |
Pain behavior*
|
| Affective version of short form |
0.82 |
0.76 |
0.59 |
0.68 |
0.74 |
0.68 |
0.72 |
| Sensory version of short form |
— |
0.74 |
0.52 |
0.68 |
0.72 |
0.61 |
0.71 |
All correlations significant at P < 0.001.
*Represents scores on a Patient-Reported Outcomes Measurement Information System short form.
NRS, numeric rating scale.
Internal consistency for the full 56-item bank, as estimated by Cronbach's alpha, was α = 0.97. Internal consistency for the sensory version of the short form was α = 0.88 and for the affective version of the short form was α = 0.87.
3.5. Scoring
All PROMIS measures are scored on a T-score metric, where 50 reflects the mean of the calibration sample and 10 reflects the SD of the calibration sample. For all PROMIS measures, a higher score indicates more of the concept being measured. A global score of pain quality (from the item bank or from either version of the short form), rather than from any single item, provides a reliable index of pain quality that can be used to assess pain quality over time. The recommended scoring for all PROMIS measures follows standard expected a posteriori scoring for response patterns,2 where PROMIS measures are scored using item-level calibrations. The most accurate way to score any PROMIS measure is to use the HealthMeasures Scoring Service (https://www.assessmentcenter.net/ac_scoringservice) or a data collection tool that automatically calculates scores (eg, Assessment Center, REDCap auto-score), or to generate them using Mplus19 or in R21 using Firestar.5 Expected a posteriori scores are then transformed to the PROMIS T-score metric (T = 10 × score + 50). This method is particularly advantageous when there are item-level missing data.
A second option for scoring involves look-up tables. For this option, first, the total raw score of this measure is calculated by simply summing the values of the responses to each of the 56 questions. Then, the score conversion table provided in Appendix B, https://links.lww.com/PAIN/B684, translates the total raw score into a standardized T-score. Similarly, to score the short-form measures, simply take the sum of the responses to either of the 8-item short-form version, and Appendix C and D, https://links.lww.com/PAIN/B684, provide the raw score to T-score look-up tables for the affective and sensory versions of the short form, respectively. All questions must be answered to produce a valid score using the look-up tables. More information about scoring and interpreting PROMIS measures can be found at http://www.healthmeasures.net/score-and-interpret/interpret-scores.
4. Discussion
This study expands the suite of PROMIS measures currently available for pediatric pain assessment (pain intensity, pain behavior, and pain interference) to include a new PROMIS pediatric pain quality measure that assesses the specific sensory and affective components of pain.
We found that the PROMIS pediatric pain quality item bank was best explained by a unidimensional model—in other words, pain qualities in a pediatric population seem to be a unitary construct. This was an unexpected finding based on the IASP characterization of the pain experience and study findings in adult pain research subjects. Our hypothesis of bidimensionality had been based in part on the fact that PROMIS adult pain quality measures found that the sensory pain qualities were even further multidimensional—distinguishing between nociceptive pain and neuropathic pain sensations.1 Neuropathic pain conditions are far more prevalent in adult populations, affecting estimated 6% to 8% (eg, diabetic neuropathy, postherpetic neuralgia, and chemotherapy-related). Thus, a priori, we did not anticipate finding that there would be differentiation of neuropathic pain in our pediatric sample as neuropathy is a rare finding in the pediatric age group and not expected in our sample. Nonetheless, we did include neuropathic pain items in the pediatric pain quality item bank for those patients who may have this condition.
Our attention in pediatrics was focused on any differences between sensory experience of pain and affective or emotional experience of pain. This question was based on the IASP characterization of multifaceted pain experience, and a clinical and research question of whether being able to characterize pain by strength of sensory component or existence of a strong affective component could guide appropriate treatment. For example, would patients with a strong affective component of pain be more likely to have depression, or pain catastrophizing and therefore preferentially benefit from psychological interventions such as CBT? Would patients with more intense sensory component benefit from alterative treatment, eg, more aggressive control of underlying chronic condition? Could patients with high pain intensity be triaged in clinical practice to receive a pain quality scale and treatment be guided by the result, and impact of treatment followed longitudinally? In fact, we did not find there to be 2 dimensions of pain quality according to our psychometric analysis. Nor did we find any significant difference in correlation between depression, pain catastrophizing, and the 2 different pain quality short forms.
One potential reason for a unidimensional result in youth may be due to developmental differences in expressive language skills where nuances in pain quality may not be conveyed in the same way as adults. Developmentally, youth may not be able to separate the feeling that their pain is achy from the misery of the experience or the worry associated with having a painful condition. Another potential reason is that the neurobiological underpinnings of the pain experience itself may be different in the still developing brain. Emerging research suggests that adolescents seem to be more sensitive to experimental pain than adults and they also recruit more areas of the brain in response to noxious stimuli (including brain areas aligned with the default mode network and frontoparietal network) that are responsible for modulating the pain response to a sensory pain stimulus.28
Although we found that the pain quality measures in pediatrics fit a unidimensional model, clinicians and researchers may benefit from having a bridge between the unidimensional expressions of pain and the multidisciplinary treatment model in their pediatric patient populations to better inform treatment planning or outcomes assessment. Thus, we used consensus-building methods to provide 2 versions of an 8-item PROMIS pediatric pain quality short form: one version focusing on the sensory aspects of pain quality and one version that focuses on the affective aspects of pain quality. Clinicians and clinical researchers may find it practical to track either the affective or sensory cluster of items when only one set seems particularly affected for their patient population. As with the other PROMIS short forms, this tailoring of outcomes measures could decrease the burden on patients who often find themselves answering multiple questionnaires. The item bank could be used in research studies, or if a clinic wished to implement computer adaptive testing to generate pain quality scores. The pain quality item bank and the 2 versions of the pain quality short form were developed specifically for children and adolescents with chronic pain using psychometrically rigorous, modern measurement methods. As with all PROMIS measures, input from stakeholders, including pediatric pain patients, was solicited during item development, which ensured developmentally appropriate and clinically relevant wording. In addition, the PROMIS pediatric sensory-focused items go beyond the tallying up of yes/no responses to various pain qualities currently used in other measures to assess the degree of discomfort associated with a particular pain quality. This report of large-scale psychometric testing among children with chronic painful conditions supports reliability of the measure and convergent validity with other self-reported measures.
The 2 versions of the pediatric pain quality short forms could help advance studies on pain mechanisms such as the neurobiology of pain pathways to map sensory vs emotional components of pain processing in the brain, which could lead to further clinical and translational studies on targeting pain therapeutics and outcome measurement in trials. For example, some therapies may have a differential impact on the sensory aspect of pain (eg, analgesic drugs) whereas others (eg, cognitive behavioral therapy, mindfulness) may alter affective processing. Having 2 versions of the pediatric pain quality short forms allows researchers and clinicians to select from items that may be more relevant to their patient population or responsive to intervention. However, given the unidimensionality of the item bank, a single short form can be administered to derive a pain quality score (both do not need to be administered to get a representative score).
A multifaceted assessment of the pain experience of children suffering from a variety of recurrent and chronic pain conditions is critical to help clinicians characterize and differentiate a patient's unique experience of pain. Understanding a patient's full pain experience can reveal different pain phenotypes that may be associated with different underlying conditions and etiologies.31 A pain quality–specific measure complements the existing PROMIS pain measures (ie, pain behavior, pain intensity, and pain interference), giving researchers and clinicians a comprehensive, unified, multidimensional assessment of pain in pediatric populations.
5. Limitations and directions for future research
This study was limited to a convenience sample of English-speaking U.S. participants and patients with 3 types of chronic pain conditions. Further validation of this measure with a more diverse sample of youth with a variety of acute, recurrent, and chronic pain conditions would increase generalizability. Cross-cultural validation and translation into other languages is also needed for broader use, and a next step in this research would be to understand clinical meaningfulness of scores. In addition, more research will be needed to understand whether pain quality differs substantively across pain diagnoses, for example, those with neuropathic vs musculoskeletal or visceral pain. We did not assess or account for potential social desirability as a source of response bias.10 Indeed, this may be a pertinent source of response bias across all PROMIS measures, given the nature of the health-related outcomes assessed by PROMIS. For example, it is possible that participants respond in a manner that minimizes their symptoms. This might be even truer of pediatric populations who may not want their caregivers to worry. And conversely, patients may express themselves in a manner supporting secondary gain or the maladaptive coping approach of catastrophizing. Further research that captures and accounts for this potential source of response bias is needed. Finally, this measure can be linked with the PROMIS adult pain quality measure in studies that include both adult and pediatric pain populations, or for longitudinal research assessing the trajectories of pain characteristics across the life span.
6. Conclusion
We have developed and psychometrically evaluated a pediatric pain quality measure for self-report use among children and adolescents aged 8 to 18 years. Development of the item content and response options included input from children and adolescents with chronic pain. The resulting measure shows good reliability and validity, and is recommended for use in research and clinical care with pediatric populations. Two versions of the short form are presented that measure the pertinent affective and sensory aspects of pain quality to enhance the feasibility of assessing pain quality in clinical settings.
Conflict of interest statement
The authors have no conflicts of interest to declare.
Appendix A. Supplemental digital content
Supplemental digital content associated with this article can be found online at https://links.lww.com/PAIN/B684.
Supplemental video content
A video abstract associated with this article can be found at https://links.lww.com/PAIN/B685.
Acknowledgements
The authors are indebted to C. Jeffrey Jacobson, PhD, for his extensive contributions to the qualitative work that underlay development of the PROMIS pediatric pain quality measure. This project would not have been possible without the excellent research assistance from Jennifer Farrell Miller, PhD, Kimberly Barnett, PhD, Jenna Tress, and Leann Schilling, and the authors thank Caravella McCuistian for assistance with patient recruiting, administration of questionnaires, and regulatory compliance. Finally, the authors are very grateful to the patient and parent participants in this research study.
Data Sharing: All data and analysis code are available from the corresponding author upon request.
The Patient-Reported Outcomes Measurement Information System (PROMIS) is an NIH Roadmap initiative to develop a computerized system measuring PROs in respondents with a wide range of chronic diseases and demographic characteristics. PROMIS II was funded by cooperative agreements with a Statistical Center (Northwestern University, PI: David Cella, PhD, 1U54AR057951), a Technology Center (Northwestern University, PI: Richard C. Gershon, PhD, 1U54AR057943), a Network Center (American Institutes for Research, PI: Susan (San) D. Keller, PhD, 1U54AR057926), and thirteen Primary Research Sites, which may include more than one institution (State University of New York, Stony Brook, PIs: Joan E. Broderick, PhD and Arthur A. Stone, PhD, 1U01AR057948; University of Washington, Seattle, PIs: Heidi M. Crane, MD, MPH, Paul K. Crane, MD, MPH, and Donald L. Patrick, PhD, 1U01AR057954; University of Washington, Seattle, PIs: Dagmar Amtmann, PhD and Karon Cook, PhD, 1U01AR052171; University of North Carolina, Chapel Hill, PI: Darren A. DeWalt, MD, MPH, 2U01AR052181; Children's Hospital of Philadelphia, PI: Christopher B. Forrest, MD, PhD, 1U01AR057956; Stanford University, PI: James F. Fries, MD, 2U01AR052158; Boston University, PIs: Stephen M. Haley, PhD and David Scott Tulsky, PhD (University of Michigan, Ann Arbor), 1U01AR057929; University of California, Los Angeles, PIs: Dinesh Khanna, MD and Brennan Spiegel, MD, MSHS, 1U01AR057936; University of Pittsburgh, PI: Paul A. Pilkonis, PhD, 2U01AR052155; Georgetown University, PIs: Carol M. Moinpour, PhD (Fred Hutchinson Cancer Research Center, Seattle) and Arnold L. Potosky, PhD, U01AR057971; Children's Hospital Medical Center, Cincinnati, PI: Esi M. Morgan DeWitt, MD, MSCE, 17 1U01AR057940; University of Maryland, Baltimore, PI: Lisa M. Shulman, MD, 1U01AR057967; and Duke University, PI: Kevin P. Weinfurt, PhD, 2U01AR052186). NIH Science Officers on this project have included Deborah Ader, PhD, Vanessa Ameen, MD, Susan Czajkowski, PhD, Basil Eldadah, MD, PhD, Lawrence Fine, MD, DrPH, Lawrence Fox, MD, PhD, Lynne Haverkos, MD, MPH, Thomas Hilton, PhD, Laura Lee Johnson, PhD, Michael Kozak, PhD, Peter Lyster, PhD, Donald Mattison, MD, Claudia Moy, PhD, Louis Quatrano, PhD, Bryce Reeve, PhD, William Riley, PhD, Ashley Wilder Smith, PhD, MPH, Susana Serrate-Sztein, MD, Ellen Werner, PhD, and James Witter, MD, PhD.
Dedication: This paper is dedicated to our colleague and coauthor of this paper, Dennis Revicki, PhD, who passed away in May 2021. The authors are deeply saddened by this loss. He generously shared his attention, wisdom, and wit with our group in the pursuit of advancing patient-reported outcome measures (PROs) in pediatric populations. Dennis had an incredibly productive career and was a world-renowned expert in outcomes measurement and research. He has left an indelible mark on the field of PROs and pain measurement, and his professional legacy will endure through those who had the privilege of collaborating with him.
References
[1]. Askew RL, Cook KF, Keefe FJ, Nowinski CJ, Cella D, Revicki DA, Morgan DeWitt EM, Michaud K, Trence DL, Amtmann D. A
PROMIS measure of neuropathic pain quality. Value Health 2016;19:623–30.
[2]. Bock RD, Mislevy RJ. Adaptive EAP estimation of ability in a microcomputer environment. Appl Psychol Meas 1982;6:431–44.
[3]. Cella D, Riley W, Stone A, Rothrock N, Reeve B, Yount S, Amtmann D, Bode R, Buysse D, Choi S, Cook K, Devellis R, DeWalt D, Fries JF, Gershon R, Hahn EA, Lai JS, Pilkonis P, Revicki D, Rose M, Weinfurt K, Hays R;
PROMIS Cooperative Group. The Patient-Reported Outcomes Measurement Information System (
PROMIS) developed and tested its first wave of adult self-reported health outcome item banks: 2005-2008. J Clin Epidemiol 2010;63:1179–94.
[4]. Cella D, Yount S, Rothrock N, Gershon R, Cook K, Reeve B, Ader D, Fries JF, Bruce B, Rose M;
PROMIS Cooperative Group. The Patient-Reported Outcomes Measurement Information System (
PROMIS): progress of an NIH Roadmap cooperative group during its first two years. Med Care 2007;45(5 suppl 1):S3–S11.
[5]. Choi SW. FIRESTAR: computerized adaptive testing (CAT) simulation program. 1.2.2 ed. Department of Medical Social Sciences Northwestern University’s Feinberg School of Medicine, Chicago, IL, 2009.
[6]. Cleeland CS, Ryan KM. Pain assessment: global use of the Brief pain inventory. Ann Acad Med Singap 1994;23:129–38.
[7]. Crombez G, Bijttebier P, Eccleston C, Mascagni T, Mertens G, Goubert L, Verstraeten K. The child version of the pain catastrophizing scale (PCS-C): a preliminary validation. PAIN 2003;104:639–46.
[8]. Cunningham NR, Kashikar-Zuck S, Mara C, Goldschneider KR, Revicki DA, Dampier C, Sherry DD, Crosby L, Carle A, Cook KF, Morgan EM. Development and validation of the self-reported
PROMIS pediatric pain behavior item bank and short form scale. PAIN 2017;158:1323–31.
[9]. DeWalt DA, Rothrock N, Yount S, Stone AA;
PROMIS Cooperative Group. Evaluation of item candidates: the
PROMIS qualitative item review. Med Care 2007;45(5 suppl 1):S12–21.
[10]. Ferrando PJ, Anguiano-Carrasco C. Acquiescence and social desirability as item response determinents: an IRT-based study with the Marlowe-Crowne and the EPQ Lie scales. Pers Individ Dif 2010;48:596‐600.
[11]. Goodman JE, McGrath PJ. The epidemiology of pain in children and adolescents: a review. PAIN 1991;46:247–64.
[12]. Hu LT, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct Equ Model 1999;6:1–55.
[13]. Irwin DE, Stucky B, Langer MM, Thissen D, Dewitt EM, Lai JS, Varni JW, Yeatts K, DeWalt DA. An item response analysis of the pediatric
PROMIS anxiety and depressive symptoms scales. Qual Life Res 2010;19:595–607.
[14]. Irwin DE, Varni JW, Yeatts K, DeWalt DA. Cognitive interviewing methodology in the development of a pediatric item bank: a patient reported outcomes measurement information system (
PROMIS) study. Health Qual Life Outcomes 2009;7:3.
[15]. Jacobson CJ, Farrell JE, Kashikar-Zuck S, Seid M, Verkamp E, Dewitt EM. Disclosure and self-report of emotional, social, and physical health in children and adolescents with chronic pain—a qualitative study of
PROMIS pediatric measures. J Pediatr Psychol 2013;38:82–93.
[16]. Jacobson CJ Jr, Kashikar-Zuck S, Farrell J, Barnett K, Goldschneider K, Dampier C, Cunningham N, Crosby L, DeWitt EM. Qualitative evaluation of pediatric pain behavior, quality, and intensity item candidates and the
PROMIS pain domain framework in children with chronic pain. J Pain 2015;16:1243–55.
[17]. Lai JS, Stucky BD, Thissen D, Varni JW, DeWitt EM, Irwin DE, Yeatts KB, DeWalt DA. Development and psychometric properties of the
PROMIS(®) pediatric fatigue item banks. Qual Life Res 2013;22:2417–27.
[18]. McGrath PJ, Finley GA. Chronic and recurrent pain in children and adolescents. Vol. 13. International Assn for the Study of Pain, Washington, DC, 1999.
[19]. Muthén LK, Muthén BO. Mplus User's Guide. 8th ed. Los Angeles, CA: Muthén & Muthén, 1998–2017.
[20]. Perquin CW, Hazebroek-Kampschreur AAJM, Hunfeld JAM, Bohnen AM, van Suijlekom-Smit LWA, Passchier J, van der Wouden JC. Pain in children and adolescents: a common experience. PAIN 2000;87:51–8.
[21]. R Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing, 2021.
[22]. Raja SN, Carr DB, Cohen M, Finnerup NB, Flor H, Gibson S, Keefe FJ, Mogil JS, Ringkamp M, Sluka KA, Song XJ, Stevens B, Sullivan MD, Tutelman PR, Ushida T, Vader K. The revised International Association for the Study of Pain definition of pain: concepts, challenges, and compromises. PAIN 2020;161:1976–82.
[23]. Reeve BB, Hays RD, Bjorner JB, Cook KF, Crane PK, Teresi JA, Thissen D, Revicki DA, Weiss DJ, Hambleton RK, Liu HH, Gershon R, Reise SP, Lai JS, Cella D;
PROMIS Cooperative Group. Psychometric evaluation and calibration of health-related quality of life item banks—plans for the patient-reported outcomes measurement information system (
PROMIS). Med Care 2007;45:S22–31.
[24]. Samejima F. Estimation of latent ability using a response pattern of graded scores. ETS Res Bull Ser 1968;1968:i-169.
[25]. Savedra MC, Holzemer WL, Tesler MD, Wilkie DJ. Assessment of postoperation pain in children and adolescents using the adolescent pediatric pain tool. Nurs Res 1993;42:5–9.
[26]. StataCorp. Stata Statistical Software: Release 16. College Station, TX: StataCorp LLC, 2019.
[27]. Talbot K, Madden VJ, Jones SL, Moseley GL. The sensory and affective components of pain: are they differentially modifiable dimensions or inseparable aspects of a unitary experience? A systematic review. Br J Anaesth 2019;123:e263–72.
[28]. Tong H, Maloney T, Payne M, Kashikar-Zuck S, Coghill R, Lopez-Sola M. Augmented pain-evoked brain responses account for greater pain sensitivity in healthy adolescents. J Pain 2021;22:605.
[29]. Varni JW, Stucky BD, Thissen D, Dewitt EM, Irwin DE, Lai JS, Yeatts K, Dewalt DA.
PROMIS Pediatric Pain Interference Scale: an item response theory analysis of the pediatric pain item bank. J Pain 2010;11:1109–19.
[30]. Varni JW, Thompson KL, Hanson V. The Varni/Thompson Pediatric Pain Questionnaire. I. Chronic musculoskeletal pain in juvenile rheumatoid arthritis. PAIN 1987;28:27–38.
[31]. Victor TW, Jensen MP, Gammaitoni AR, Gould EM, White RE, Galer BS. The dimensions of pain quality: factor analysis of the Pain Quality Assessment Scale. Clin J Pain 2008;24:550–5.
