Abstract
Background:
The Yale–Brown Obsessive Compulsive Scale (YBOCS) is the most widely used clinician-rated instrument for assessing the severity of obsessive-compulsive disorder (OCD). There are no validated translations of the scale in Indian languages. In a clinical sample, we evaluated the psychometric properties of the Bengali, Hindi, and Kannada versions of the Y-BOCS severity scale.
Methods:
The English version of Y-BOCS was translated into three Indian languages employing standard procedures. The Kannada, Hindi, and Bengali versions were administered to 71, 100, and 105 OCD patients, respectively, along with the National Institute of Mental Health-Global Obsessive-Compulsive Scale (NIMH-GOCS) and the Patient Health Questionnaire (PHQ-9). Cronbach’s alpha was assessed to measure internal consistency. Test-retest reliability was evaluated by re-administering the translated scales by the primary rater. Inter-rater reliability was evaluated by re-administering the scale by an independent rater within a week.
Results:
All three versions of the Y-BOCS severity scale had high Cronbach’s Alpha scores (>0.90) for total and subscale scores, suggesting good internal consistency. The intraclass coefficients for inter-rater and test-retest reliability were >0.90, suggesting excellent reliability. Although the total Y-BOCS score was significantly correlated with both NIMH-GOCS and PHQ-9, the correlation was stronger for the former, supporting the construct validity.
Conclusion:
The translated versions of Y-BOCS investigated in this study are psychometrically reliable and valid measures for the assessment of severity in the Indian population with OCD.
The Bengali, Hindi, and Kannada translated versions of the Yale–Brown Obsessive Compulsive Scale (Y-BOCS) are psychometrically reliable and valid measures for the assessment of severity in the Indian clinical population with obsessive-compulsive disorder (OCD).Key Messages:
The Yale–Brown Obsessive Compulsive Scale (Y-BOCS) 1 is the most widely used clinician-rated instrument for assessing the severity of obsessive-compulsive disorder (OCD) in clinical populations and research studies. Y-BOCS severity scores are used to define severity benchmarks and treatment response criteria for OCD.2,3 The Y-BOCS severity scale is a clinician-rated 10-item scale that rates the severity of obsessions and compulsions in time, interference, distress, resistance, and control. The original English version of the Y-BOCS has been found to have good inter-rater reliability, internal consistency, and construct validity. 1 A meta-analysis of studies evaluating the reliability of Y-BOCS (n = 128 studies) showed good reliability for the total scale (0.866 for Cronbach’s alpha; 0.848 for test-retest; 0.922 for inter-rater reliability). 4 The Y-BOCS has been translated into other languages, and its psychometric properties have been established.5,6
India, being the most populous country in the world, has a large population of people living with psychiatric disorders, including OCD. 7 There is burgeoning research from India on various aspects of OCD. 8 However, there are no psychometrically evaluated assessments for OCD available in Indian languages. India has a diverse population with 22 official languages. Translating instruments into multiple languages is necessary to cover a wider population. We attempted to bridge this lacuna in this study by evaluating the psychometric properties of Y-BOCS translations in three Indian languages, Bengali, Hindi, and Kannada, which are spoken in the eastern, northern, and southern regions of India. Although the Y-BOCS scale has been revised with a second edition, 9 we evaluated the first edition, as it is still the gold standard instrument used in clinical trials to establish empirical benchmarks for clinical evaluation.2,3
The study aimed to evaluate the reliability (internal consistency, test-retest, and inter-rater reliability) and validity (convergent and divergent validity) of the Bengali, Kannada, and Hindi versions of the Y-BOCS. We hypothesized that the three translated versions of the Y-BOCS would demonstrate good internal consistency, test-retest reliability, inter-rater reliability, and convergent and divergent validity.
Methods
After obtaining institutional ethics clearance, the psychometric evaluation of the three versions was conducted in three independent studies from September 2019 to October 2023. We used the Standards for Reporting Diagnostic Accuracy (STARD) checklist to write our report (online supplementary material). 10
Translation
The scales were translated according to standard guidelines. 11 The Y-BOCS symptom severity scale was translated into the respective languages, which were translated back to English by an independent translator. The translators were bilingual with proficiency in both languages. Two independent researchers (mental health professionals familiar with the Y-BOCS scale) compared the back-translated version with the original version for semantic, conceptual, and content equivalence. Discrepancies were rectified. The scale was further evaluated by experts (experienced mental health professionals proficient in the respective languages with expertise in OCD) for clarity of language and conceptual equivalence. This scale version was pilot-tested in 10 patients for clarity and understandability. The final version (online supplementary material) was evaluated for psychometric properties.
Psychometric Evaluation
Sample
Participants were recruited from a tertiary care hospital in India. Written informed consent was obtained from all study participants. Consenting participants who satisfied the following criteria were recruited. We recruited patients with a diagnosis of OCD as per DSM-5 criteria, 12 aged between 18 and 60 years, actively symptomatic at the time of presentation as per clinical examination, with the ability to speak and read the respective languages. Patients with a history of co-morbid psychotic disorders, bipolar disorder, mental retardation, dementia, or presenting with a current episode of severe depression or active substance use disorder other than nicotine dependence were excluded. We recruited 100, 105, and 71 participants for the validity assessment of the scale’s Bengali, Hindi, and Kannada versions, respectively.
Measures
The translated versions of the Y-BOCS were administered to all study participants by one primary bilingual rater for each version. The raters were trained in administering and scoring the scales through didactic teaching, observation, and co-administration of ratings with a trained clinician-rater. (first author: Bengali Version, second author: Hindi version, third author: Kannada version). The same raters re-administered the translated Y-BOCS within one week to assess retest reliability in a subset of participants (20, 78, and 35 participants for the Bengali, Hindi, and Kannada versions, respectively). In an independent subset of the sample (20, 27, and 22 participants for the Bengali, Hindi, and Kannada versions, respectively), the scale was re-administered by an independent rater within one week to assess inter-rater reliability (sixth author: Bengali version, fifth author: Hindi version, fourth author: Kannada version).
Patient Health Questionnaire-9 (PHQ-9) 13
The Bengali, Hindi, and Kannada versions of the PHQ-9 were administered to establish divergent validity. The PHQ-9 is a self-report inventory used for screening, diagnosis, and monitoring of depressive disorders.
National Institute of Mental Health—Global Obsessive-Compulsive Scale (NIMH—GOCS) 14
A single-item clinician-rated scale assessing the global severity of OC symptoms (scores ranging from 1 to 15) was administered to assess the convergent validity.
Sample Size Estimation
The estimated sample size for validity assessment was 100 participants for each version, based on the recommendations of at least 10 subjects per item of the scale. 11 We could not reach the desired sample size for the Kannada version due to the COVID-19 pandemic, but we achieved the desired sample size for the other two versions, which were conducted between 2022 and 23. We planned to recruit 20 participants each for test-retest and inter-rater reliability based on standard recommendations for sample size calculation for intraclass correlation coefficient (ICC) using a web-based application.15–17 A minimum requirement of 16 participants is required to obtain an ICC between two ratings of 0.9 with a power of 80% and a two-tailed significance of 0.05. We recruited the desired sample size to test the reliability of all three scale versions.
Statistical Analysis
Statistical analyses were conducted using SPSS version 26.0 (SPSS Inc., Chicago, USA). Cronbach’s α was measured for the whole scale and the obsessions and compulsions subscales to assess internal consistency. ICC was evaluated to assess the degree of concordance for test-retest reliability and inter-rater reliability. Pearson’s correlation coefficient was employed to assess convergent and divergent validity.
Results
The demographic and clinical characteristics of the three samples are summarized in Table 1.
Reliability
Internal Consistency
The Cronbach’s alpha for the total Y-BOCS scores for the Bengali, Hindi, and Kannada versions were 0.934, 0.930, and 0.928, respectively. The internal consistency estimates for the total/sub-scale scores are shown in Table 2. Cronbach’s alpha remained stable after removing each item, supporting the stability of the scale (Supplementary Table 1). The corrected item-total correlation for all items was >0.59 (Supplementary Table 2), well above the recommended cutoff of 0.3. 18
Demographic and Clinical Profile of the Sample.
OCD: Obsessive compulsive disorder; Y-BOCS: Yale Brown Obsessive Compulsive Scale; NIMH GOCS: National Institute of Mental Health - Global Obsessive-Compulsive Scale; PHQ-9: Patient Health Questionnaire-9; SSRI: Selective Serotonin Reuptake Inhibitors.
Reliability and Validity Estimates.
Y-BOCS: Yale–Brown Obsessive Compulsive Scale; NIMH GOCS: National Institute of Mental Health - Global Obsessive-Compulsive Scale; PHQ9: Patient Health Questionnaire-9.
*P value < .01.
Test-retest and Inter-rater Reliability
The ICCs for both measures for all three versions were greater than 0.90 (Table 2), suggestive of excellent reliability. 19
Construct Validity
The total Y-BOCS correlated more strongly with the NIMH-GOCS, a measure of OCD severity, than with the PHQ-9, a measure of depression severity (Table 2). This supports the scale’s convergent and divergent validity.
Discussion
All three versions of the Y-BOCS had good reliability, including internal consistency, test-retest reliability, and inter-rater reliability. Although the total score correlated with both global OCD severity and depression severity, the correlation was relatively stronger for the former, supporting the construct validity. The scales were translated, and their psychometric properties were evaluated employing rigorous processes recommended by guidelines. 11 The Cronbach’s alpha scores were >0.9, in the range of acceptable values. 20 Although extremely high values might suggest redundancy of items, Cronbach’s alpha of the English version of the Y-BOCS score has also been found to be around 0.9. 1 Cronbach’s alpha remained stable after removing each item, supporting the stability with good corrected item-total correlation. This is in contrast to some studies showing that the resistance against obsessions and compulsions items has a weaker correlation with the remainder of the scale.1,21 This led to removing the former item from the second edition of the Y-BOCS scale. 9 The item may be especially confusing in the context of cognitive-behavior therapy for OCD, where patients are advised to reduce their resistance to obsessions. However, this item showed a good corrected item-total correlation in our sample, which was predominantly treated with pharmacotherapy (Table 1). The scale showed excellent inter-rater and test-retest reliability. 19 The scales were administered by trained raters, who were post-graduate residents in psychiatry. This supports the ease of training and the potential for broader application. The Y-BOCS score had a stronger correlation with global OCD severity (assessed using the NIMH-GOCS scale) than PHQ-9 scores, which supported the convergent and divergent validity. There was a significant correlation between OCD and depression severity, as observed in previous studies.22,23
Limitations
We could not assess the construct validity across a broader range of scales due to the limited availability of validated scales in the studied Indian languages. The sample was recruited from a tertiary care center, and the ratings were administered by raters trained at a specialist center. However, this suggests that trained raters can administer the Y-BOCS reliably without compromising the fidelity in rating. 24 Further, we did not employ cognitive interviewing techniques, which could have helped assess the respondents’ understanding and interpretation of the scale. 25 The difference in sample sizes between the cohorts was due to the difficulties caused by the COVID-19 pandemic. Post-hoc analysis revealed that the sample size for each cohort was adequate to estimate a minimum ICC of 0.82 (±15%), Cronbach’s α of 0.6 (±15%) and Pearson’s correlation coefficient of 0.62 (±15%). 15
Conclusion
The Bengali, Hindi, and Kannada versions of the Y-BOCS retained the psychometric properties of the English version and can be used to assess OCD in clinical settings and research studies.
Supplemental Material
Supplemental material for this article available online.
Supplemental Material
Supplemental material for this article available online.
Footnotes
Data Availability
The datasets generated during the current study are available from the corresponding author upon reasonable request.
Declaration of Conflicting Interests
The authors declare no potential conflicts of interest with respect to the research, authorship, and publication of this article.
Declaration Regarding the Use of Generative AI
None used.
Ethical Considerations
The manuscript is based on three different studies, which were conducted as MD (Psychiatry) theses of the first three authors after obtaining approval from the NIMHANS ethics committee in June 2019, September 2021 and April 2022. All study participants provided written informed consent to participate in the study.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Prior Presentations and Publications
The data were partially presented as a poster presentation at the 75th Annual National Conference of the Indian Psychiatric Society 2024, Kochi and the 34th Annual Conference of the Indian Psychiatric Society Odisha State Branch 2024, Bhubaneshwar.
The manuscript has not been submitted for publication to any other journal or resource.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
