Extending the sensory phenomenon assessment scale to pediatric obsessive–compulsive and related disorders: A preliminary psychometric investigation

Sensory phenomena (SP) refer to uncomfortable subjective experiences, such as pressure, itch‐like sensations, bodily tension, feelings of incompleteness, or “just‐right” perceptions, which may precede, accompany, or motivate repetitive behaviors. ¹ Although these experiences are most classically described in tic disorders (TD) as premonitory urges, related sensory experiences have also been reported across obsessive–compulsive and related presentations, including obsessive–compulsive disorder (OCD), body‐focused repetitive behaviors (BFRB), and other sensorimotor presentations, such as restless legs syndrome (RLS). ² , ³ , ⁴ Under these conditions, SP may function as a proximal driver of repetitive behavior. In TD, the premonitory urges typically intensify before a tic and are transiently relieved afterward. A similar sensory–behavioral pattern has been described in OCD, where some repetitive acts are performed not primarily to prevent feared consequences, but to reduce aversive internal sensations or restore a sense of completeness, symmetry, or “rightness”. ¹ , ⁵ , ⁶ , ⁷ One plausible mechanism through which such behaviors may be maintained is negative reinforcement: sensory urges or internal tension may intensify before a repetitive act and diminish afterward, which increases the likelihood that the behavior will recur when similar sensory states arise. ⁵ , ⁸ , ⁹ Consequently, SP may contribute to clinical burden both directly, through distressing internal experiences, and indirectly, through recurrent behaviors and associated functional interference. ¹⁰ Recognizing this sensory‐driven pathway may be particularly important in OCD and related presentations where conventional assessment frameworks have largely emphasized fear, threat reduction, and global symptom severity. However, SP remain incompletely characterized within conventional OCD assessment frameworks. Figure 1 summarizes the conceptual rationale for considering sensory experiences alongside repetitive behaviors, and their clinical consequences across obsessive–compulsive and related presentations.

FIGURE 1.

Conceptual framework for characterizing sensory phenomena across obsessive–compulsive and related presentations. SPAS, Sensory Phenomenon Assessment Scale.

SP in obsessive–compulsive and related presentations is phenomenologically heterogeneous and may not be adequately characterized by conventional measures of tic or obsessive–compulsive symptom severity alone. They can include bodily sensations or tension, urge‐like experiences, feelings of incompleteness, and internally or externally triggered “just‐right” experiences, with their relative prominence potentially differing across clinical presentations. In tic‐related obsessive–compulsive presentations, tic and obsessive–compulsive symptoms, and sensory experiences may overlap. ⁸ For example, non‐just‐right experiences have been reported to relate more closely to OCD than to tic severity in individuals with Tourette syndrome, and aversive bodily sensations or feelings of “not‐rightness” may temporally precede and motivate covert or overt repetitive behaviors. ¹¹ This phenomenological heterogeneity may also have a developmental dimension, as the expression and assessment of SP may vary across childhood and adolescence. Therefore, developmental differences in the recognition and reporting of internal sensory experiences may be particularly relevant when assessing SP in pediatric populations. However, existing assessment approaches remain limited in their ability to capture this phenomenological and developmental complexity, particularly in pediatric OCD or obsessive‐compulsive and related disorders (OCRDs). Most established measures focus primarily on tic severity, obsessive–compulsive symptom severity, or specific sensory experiences, rather than providing a multidimensional characterization of SP and their behavioral and functional consequences. For example, the Yale Global Tic Severity Scale (YGTSS) and the Children's Yale–Brown Obsessive Compulsive Scale (CY‐BOCS) primarily assess tic and obsessive–compulsive symptom severity, respectively, whereas the Premonitory Urge for Tics Scale (PUTS) focuses specifically on premonitory urges in TD. ¹² , ¹³ , ¹⁴ , ¹⁵ Although these measures are clinically valuable, they do not provide a unified multidimensional assessment of SP across sensory, behavioral, and functional features. Meanwhile, assessment may be further complicated in younger children, particularly those under 10 years of age, who may have greater difficulty identifying and verbally describing internal sensory states. ¹⁶ Hence, these phenomenological and developmental considerations highlight the potential value of a clinician‐guided, multidimensional approach for assessing SP in pediatric populations.

The Sensory Phenomenon Assessment Scale (SPAS) was developed as a clinician‐rated, semi‐structured instrument to characterize both the phenomenology and severity of sensory experiences. ¹⁶ The assessment begins with an eight‐item symptom checklist used to identify the type and bodily location of SP, including itching, sense of suffocation, pressure, sense of energy release, tension, sense of incompleteness, indescribable discomfort, and other sensory symptoms. When SPs are identified, the clinician rates five severity domains: number, frequency, tensity, degree of transformation, and functional impairment. Each domain was rated using six ordered response categories and then summed to generate the SPAS total severity score, with higher scores indicating greater severity. By integrating clinician guidance with information from both the child and caregiver, the SPAS may offer a structured approach to assess SP in pediatric populations. Initial evaluation in children with TD demonstrated good internal consistency, one‐month test–retest reliability, and a coherent one‐factor pattern of the severity indicators; SPAS scores were associated with PUTS scores but showed weaker associations with conventional tic and obsessive–compulsive symptom severity measures. ¹⁶ However, its psychometric performance and clinical correlates in pediatric OCD/OCRD remain insufficiently characterized.

Accordingly, the present study investigated the preliminary psychometric performance and clinical correlates of SPAS in a pediatric OCD/OCRD sample. A TD sample was included as a clinically informative reference because SP, particularly the premonitory urges, is well characterized in TD. This study was intended as an initial psychometric extension of the SPAS for pediatric OCD/OCRD. Specifically, we examined the internal consistency and dimensional coherence of the SPAS severity indicators and tested their associations with CY‐BOCS and Obsessive‐Compulsive Inventory‐Revised (OCI‐R) measures in the OCD/OCRD sample, and characterized SPAS‐related association patterns in the TD sample across measures of premonitory urges, tic severity, broader tic‐related symptoms, and quality‐of‐life burden. These complementary measures were included to distinguish the association of SPAS‐assessed SP with premonitory urges from its associations with tic severity, symptom burden, and quality‐of‐life impact.

We analyzed consecutive clinical assessment data at Beijing Children's Hospital. Participants were consecutively enrolled during the study period from May 1, 2022 to April 30, 2024, according to the prespecified eligibility criteria. The OCD/OCRD sample comprised 386 children and adolescents aged 6–17 years (median age [interquartile range {IQR}], 11 [10–13] years) who had a clinically significant obsessive–compulsive symptom burden, defined as a CY‐BOCS total score of ≥16, and received a consensus DSM‐5 diagnosis of OCD or another OCRD through a two‐stage clinical review process involving two attending child psychiatrists. The CY‐BOCS was used to quantify the severity of obsessive–compulsive symptoms rather than to establish a diagnosis. Of the 386 participants, 315 (81.6%) had OCD, two (0.5%) had trichotillomania or excoriation disorder, and 69 (17.9%) were classified as having other OCRDs. The TD reference sample comprised 307 patients aged 6–17 years with clinically diagnosed TD and clinically relevant premonitory urges (median age [IQR], 10 [8–12] years). Detailed diagnostic procedures, eligibility criteria, and sample characteristics are provided in Tables S1 and S2.

Associations between SPAS scores and clinical measures were examined using Spearman rank correlations, with Benjamini–Hochberg false discovery rate (FDR) correction applied within the corresponding families of correlation tests. In the OCD/OCRD sample, SPAS showed good internal consistency (Cronbach's α = 0.852; McDonald's ω = 0.920). The polychoric correlation matrix showed good factorability (Kaiser–Meyer–Olkin measure of sampling adequacy = 0.879; Bartlett's χ ² (10) = 1460.740, P < 0.001), and parallel analysis supported the retention of a single factor. Principal‐axis factor analysis yielded loadings ranging from 0.662 to 0.935, with corrected item–total correlations ranging from 0.522 to 0.788 (Table 1).

TABLE 1.

Psychometric properties and clinical correlates of the Sensory Phenomenon Assessment Scale (SPAS) in obsessive–compulsive disorder/obsessive–compulsive and related disorder (OCD/OCRD) sample

Domain	Measure or item	Statistic	Estimate	P‐value
Clinical associations	CY‐BOCS total score	Spearman's ρ	0.189	<0.001
	Obsession score	Spearman's ρ	0.190	<0.001
	Compulsion score	Spearman's ρ	0.106	0.047
	OCI‐R total score	Spearman's ρ	0.145	0.008
	Hoarding	Spearman's ρ	0.131	0.016
	Checking	Spearman's ρ	0.169	0.003
	Symmetry/ordering	Spearman's ρ	0.100	0.055
	Counting/numbers	Spearman's ρ	0.148	0.008
	Contamination/cleaning	Spearman's ρ	0.044	0.392
	Intrusive thoughts	Spearman's ρ	0.112	0.040
Internal consistency	Cronbach's α	Reliability coefficient	0.852	—
	McDonald's ω	Reliability coefficient	0.920	—
Factorability	KMO	Sampling adequacy	0.879	—
	Bartlett's test	χ 2(df)	1460.740 (10)	<0.001
Item‐level dimensional coherence	Number of sensory phenomena	Corrected item–total r	0.676	—
		Factor loading	0.842	—
	Frequency	Corrected item–total r	0.771	—
		Factor loading	0.935	—
	Tensity	Corrected item–total r	0.788	—
		Factor loading	0.905	—
	Degree of transformation	Corrected item–total r	0.522	—
		Factor loading	0.662	—
	Functional impairment	Corrected item–total r	0.663	—
		Factor loading	0.810	—

Note: Clinical associations are presented as Spearman's rank correlation coefficients (ρ) between the SPAS total score and the corresponding obsessive–compulsive symptom measures. The P‐values for clinical associations were adjusted for multiple comparisons using the Benjamini–Hochberg false discovery rate procedure, whereas the P‐value for Bartlett's test was unadjusted. Internal consistency was evaluated using Cronbach's α and McDonald's ω. Because the five SPAS severity indicators are ordinal, dimensional structure was examined using a polychoric correlation matrix. Parallel analysis supported retention of a single factor, followed by exploratory factor analysis using principal axis factoring. Corrected item–total correlations were calculated using the observed item scores. CY‐BOCS, Children's Yale–Brown Obsessive Compulsive Scale; KMO, Kaiser–Meyer–Olkin measure of sampling adequacy; OCI‐R, Obsessive‐Compulsive Inventory–Revised; OCD, obsessive–compulsive disorder; OCRD, obsessive–compulsive and related disorder; SPAS, Sensory Phenomenon Assessment Scale.

SPAS total scores showed modest positive associations with CY‐BOCS total scores (ρ = 0.189, FDR‐adjusted P < 0.001) and OCI‐R total scores (ρ = 0.145, FDR‐adjusted P = 0.008), whereas individual SPAS severity indicators showed related but non‐identical patterns of association with obsessive–compulsive symptom measures (Table 1; Figure 2A,C). Sensitivity analyses excluding participants with documented comorbid anxiety or depressive disorders yielded a highly similar pattern, with all correlations retaining the same direction and a maximum absolute change in Spearman's ρ of 0.018 (Supplementary Materials). These modest associations suggest a partial overlap between SPAS‐assessed SP and conventional obsessive–compulsive symptom severity, while leaving open the possibility that the SPAS captures complementary clinical information. The present correlational findings, however, did not establish SP as a distinct clinical dimension. Because the degree‐of‐transformation item may partly reflect the behavioral consequence of SP rather than the sensory severity itself, we repeated the psychometric analyses after excluding this item. The resulting four‐item solution showed similar internal consistency and dimensional coherence, indicating that its removal did not substantially alter the overall psychometric pattern. The comparatively lower loading of the transformation item may indicate some distinction from other severity indicators, although this interpretation requires further investigation (Table S3).

FIGURE 2.

Clinical association of SPAS in OCD/OCRD and TD samples. Panels A and B show lollipop plots of Spearman correlations between SPAS total score and clinical measures in the OCD/OCRD and TD samples, respectively. Points indicate Spearman's ρ, and labels show ρ with FDR‐adjusted significance. Panels C and D show bubble heatmaps of Spearman correlations between the five SPAS severity indicators and clinical measures in the OCD/OCRD and TD samples, respectively. Bubble color represents Spearman's ρ, bubble size represents the absolute magnitude of the correlation, and asterisks indicate FDR‐adjusted significance. The five SPAS severity indicators were the number of sensory phenomena, frequency, tensity, degree of transformation, and functional impairment. FDR, false discovery rate; OCD, obsessive–compulsive disorder; OCRD, obsessive–compulsive and related disorder. SPAS, Sensory Phenomenon Assessment Scale; TD, tic disorder.

The observed one‐factor pattern should be interpreted considering the phenomenological heterogeneity of SP. The five SPAS items assess severity dimensions rather than distinct sensory subtypes. ¹⁶ Therefore, coherence among these indicators does not imply that underlying experiences are homogeneous. Individuals with similar SPAS scores may differ in the predominant sensory experiences that contribute to those scores. Developmental variations may also influence SP expression and reporting. Previous developmental studies on premonitory urges have identified age‐related differences in psychometric characteristics and clinical correlates, suggesting that the assessment of sensory experiences may vary across childhood and adolescence. ¹⁴ , ¹⁷ Although the clinician‐guided and caregiver‐supported format of SPAS may facilitate pediatric assessment, age‐specific measurement properties were not examined in the present study. Future age‐stratified and longitudinal studies are needed to evaluate the developmental differences in the expression, reporting, and assessment of SP.

In the TD sample, the SPAS total score showed the strongest association with PUTS, whereas associations with tic severity, tic‐related symptom measures, and quality‐of‐life indices were comparatively smaller (Figure 2B,D; Table S4). Given that premonitory urges are a well‐characterized sensory feature of TD, this pattern provides a clinically interpretable reference for the TD sample. Importantly, additional within‐sample sensitivity analyses adjusting for age and sex did not materially alter the principal association patterns in either clinical sample (Supplementary Materials).

This study has several limitations. First, this retrospective, cross‐sectional study was conducted at a single pediatric outpatient center, limiting causal inference and generalizability. Second, diagnoses were established through routine clinical assessment and a two‐stage consensus review, rather than a standardized structured or semi‐structured interview. Third, although participants using psychotropic medication during the 6 weeks preceding the assessment were excluded, 163 participants (42.2%) had earlier psychotropic medication exposure. Pharmacotherapy can modify obsessive–compulsive symptom severity in pediatric OCD, but its effects on SP remain less well established. ¹⁸ , ¹⁹ , ²⁰ Thus, the potential influences of previous treatment exposure on obsessive–compulsive symptom severity or SP cannot be excluded. ¹⁸ , ¹⁹ , ²⁰ Fourth, subtype‐level diagnostic information was limited for non‐OCD OCRDs, and the sample was predominantly composed of patients with OCD, restricting conclusions regarding individual OCRD subtypes. Fifth, the demographic and clinical characteristics and eligibility criteria of the OCD/OCRD and TD samples are different, so cross‐sample differences should be interpreted descriptively rather than as evidence of transdiagnostic validity. Finally, the study did not provide a comprehensive validation of the SPAS, including test–retest reliability, convergent and discriminant validity, or prospective clinical utility.

In the pediatric clinical sample, the SPAS severity indicators showed good internal consistency and a coherent one‐factor pattern, while SPAS total scores were modestly associated with conventional obsessive–compulsive symptom measures. These findings provide preliminary support for further evaluation of SPAS as an approach to characterize SP in pediatric OCD/OCRD. Therefore, the present findings should be viewed as a preliminary psychometric extension of the SPAS to pediatric OCD/OCRD rather than as a validation of the SPAS as a transdiagnostic instrument. Independent prospective studies incorporating standardized diagnostic procedures, dedicated SP measures, age‐stratified and longitudinal assessments, and diagnostically stratified samples are needed to determine their validity, developmental measurement properties, and potential clinical utility across diagnostic contexts.

ETHICAL APPROVAL

The study was approved by the Ethical Committee of Beijing Children's Hospital, Capital Medical University (approval number: [2023]‐E‐105‐R). Informed consent was obtained from each participant and their main caregiver or legal guardian.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

Supporting information

Supporting Information