Making sense of ICD-11 diagnostic criteria of compulsive sexual behavioural disorder

Dear Editor,

Compulsive sexual behaviour disorder (CSBD) is identified by persistent, intense sexual urges that lead to repetitive activities, causing significant distress or impairment over 6 months or more.¹ It is characterised by neglecting important responsibilities, repeated unsuccessful attempts to control the behaviour and continuation despite negative consequences.¹ Crucially, the distress linked to CSBD is not due to moral judgments about sexual behaviour.¹

The inclusion of CSBD in the International Classification of Diseases (ICD-11) in 2018 marks a significant development, offering much-needed clarity for diagnosis and treatment.² This clarity has the potential to enhance clinical practice and advance research within the mental health sector. A clear classification of CSBD can significantly influence funding, public health and clinical guidelines, making standardised classifications like the ICD-11 essential for designing interventions and allocating resources.³ Such clarity facilitates the implementation of standardised treatments, improves the quality and availability of care, supports the effective budgeting of scarce healthcare resources, and reduce the stigma associated with CSBD.⁴

The study team has decided to embark on research aimed at evaluating previous clinical case notes to determine whether earlier clinicians’ assessments align with the latest CSBD criteria. This study seeks to break new ground by involving 3 independent reviewers who will collect data from these notes, with inter-rater analysis conducted based on their evaluations. The objective is to address existing challenges in classification and diagnosis, thereby contributing to improved clinical practices, more effective resource allocation and informed policy development, ultimately benefitting individuals affected by CSBD.

The dataset for this review was extracted from a local addiction treatment centre’s treatment outcome monitoring database, selecting patients who sought treatment for sexual-related issues or diagnoses between 2013 and 2018. Three psychiatrist reviewers performed a pilot test on 10 random cases to ensure familiarity with the new ICD-11 CSBD guidelines and to align their expectations and interpretations between the guidelines and the clinical cases. This exercise was intended to help them become familiar with the new ICD-11 CSBD guidelines and to ensure their expectations and interpretations were consistent with both the guidelines and the clinical cases. Their expertise and prior experience in the field played a crucial role in aligning their assessments with the latest diagnostic criteria. Any disagreement was resolved, with support from the research assistants, through formal discussion. All other following case reviews were conducted independently by them.

To measure inter-rater reliability (IRR), percentage agreement and Fleiss’ kappa were adopted for analysis because percentage agreement provides a straightforward measure of how well 2 or more observers agree on categorical data, with an acceptable level defined at 70% and above.³ Fleiss’ kappa is ideal for calculating IRR involving 3 or more raters, especially when handling categorical data, with interpretation as follows: ≤0 (none), 0.01–0.20 (slight), 0.21–0.40 (fair), 0.41–0.60 (moderate), 0.61–0.80 (substantial) and 0.81–1.00 (almost perfect).³

Table 1. Results.

A total of 78 cases (Table 1) were reviewed, with a mean age of 28.29 years (standard deviation [SD]=11.2). The sample consisted of 95% males, majority were of Chinese ethnicity (69.2%), and 56.7% were married. Most cases were from 2018 (26.9%) and 2015 (23.1%). Item 3d (Table 1) revealed to have the highest percentage agreement (86%), while the other items fell below the acceptable threshold. The overall percentage agreement on whether cases satisfied the CSBD criteria was 58%, indicating a lower-than-acceptable agreement level among reviewers. Fleiss’ kappa analysis revealed none-to-slight agreement for items 1 (κ=0.137; z=2.09, P=0.036) and 3d (κ=0.104; z=1.59, P=0.0111). However, items 2, 3a, 3b and 3c showed fair agreement, and the overall agreement on whether cases met the CSBD criteria was also fair (κ=0.304; z=4.46, P<0.001).

The inclusion of CSBD in the ICD-11 in 2018 introduces unique challenges, particularly when reviewing retrospective cases. In our analysis, we found that only 8% (n=6) of the cases reviewed met the CSBD criteria, with only 2 of these cases documented from 2018 onwards. Considering the relatively recent introduction of this diagnosis, we recognise that there is some uncertainty which may lead to confusion due to the subjective interpretation of the diagnostic criteria since their introduction in 2018. Additional factors that may complicate the assessment include: (1) a lack of comprehensive understanding and training regarding the new diagnostic criteria, (2) insufficient corroborative information, and (3) superficial patient information, which could lead to potential under-reporting of CSBD symptoms. This significantly affects the identification and assessment of key disorder criteria across cases. The inconsistency and bias from subjective interpretation and under or overreporting of symptoms may lead to incomplete assessment and inaccurate diagnosis, thereby affecting treatment decision. Additionally, the results identified substantial variability in CSBD diagnoses, with agreement percentages ranging from 46% to 86%. Fleiss’ kappa values indicated slight to fair consistency, with values between κ=0.104 and κ=0.345. Overall agreement was found to be 58%, with a Fleiss’ kappa value of κ=0.304. The fair to moderate IRR highlights the need for standardised diagnostic procedures to improve consistency in identifying CSBD.

To address inconsistencies in identifying and assessing CSBD, this study recommends regular training for clinicians to standardise clinical diagnoses and ensure proper documentation.⁶ Standardising the practices in care and treatment for CSBD can enhance reliability, resulting in more accurate diagnoses in both clinical and research settings.⁴ One approach is to utilise a manualised method for deriving these criteria, which provides a structured framework that promotes consistency and precision. This method not only streamlines the assessment process but also improves the clarity of diagnoses, benefiting both practitioners and patients. Additionally, given the sensitivity of the condition, establishing positive rapport⁸ and therapeutic trust with patients during their first visit can assist clinicians in formulating concise case notes and appropriately assessing patients.⁷

Our study provides valuable insights into ICD-11’s recent inclusion of CSBD, highlighting the need for clinician training to standardise clinical diagnoses and ensure proper documentation, so as to improve treatment outcomes, ultimately benefitting both clinicians and patients.

Acknowledgments

The team would like to express their gratitude to Dr Edimansyah Abdin, Principal Biostatistician from the Research Division at the Institute of Mental Health for his essential advice and expertise, particularly regarding the statistical direction of this work. His sage advice was the most important tool in finishing this manuscript.

We would especially like to thank National Addictions Management Service for providing the resources needed to make this study possible.

REFERENCES

1. Kraus SW, Krueger RB, Briken P, et al. Compulsive sexual behaviour disorder in the ICD-11. World Psychiatry 2018;17:109-10.
2. Lew-Starowicz M, Coleman E. Mental and sexual health perspectives of the International Classification of Diseases (ICD-11) Compulsive Sexual Behavior Disorder. J Behav Addict 2022;11:226-9.
3. Aguilar-Yamuza B, Trenados Y, Herruzo C, et al. A systematic review of treatment for impulsivity and compulsivity. Front Psychol 2024;15:1430409.
4. Briken P, Bőthe B, Carvalho J, et al. Assessment and treatment of compulsive sexual behavior disorder: a sexual medicine perspective. Sex Med Rev 2024;12:355-70.
5. Shabankhani B. Assessing the inter-rater reliability for nominal, categorical and ordinal data in medical sciences. Arch Pharm Prac 2020;11:144-8.
6. Beauchemin M, Cohn E, Shelton RC. Implementation of Clinical Practice Guidelines in the Health Care Setting. ANS Adv Nurs Sci 2020;42:307-24.
7. Butt MF. Approaches to building rapport with patients. Clin Med (Lond) 2021;21: e662-3.