ObjectivesIn the International Classification of Diseases, Tenth Edition (ICD-10), hypochondriasis (illness anxiety disorder) and dysmorphophobia (body dysmorphic disorder) share the same diagnostic code (F45.2). However, the Swedish ICD-10 allows for these disorders to be coded separately (F45.2 and F45.2A, respectively), potentially offering unique opportunities for register-based research on these conditions. We assessed the validity and reliability of their ICD-10 codes in the Swedish National Patient Register (NPR).DesignRetrospective chart review.MethodsSix hundred individuals with a diagnosis of hypochondriasis or dysmorphophobia (300 each) were randomly selected from the NPR. Their medical files were requested from the corresponding clinics, located anywhere in Sweden. Two independent raters assessed each file according to ICD-10 definitions and Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision and Fifth Edition criteria. Raters also completed the Clinical Global Impression–Severity (CGI-S) and the Global Assessment of Functioning (GAF).Primary outcome measurePer cent between-rater agreement and positive predictive value (PPV). Intraclass correlation coefficients for the CGI-S and the GAF.ResultsEighty-four hypochondriasis and 122 dysmorphophobia files were received and analysed. The inter-rater agreement rate regarding the presence or absence of a diagnosis was 95.2% for hypochondriasis and 92.6% for dysmorphophobia. Sixty-seven hypochondriasis files (79.8%) and 111 dysmorphophobia files (91.0%) were considered ‘true positive’ cases (PPV=0.80 and PPV=0.91, respectively). CGI-S scores indicated that symptoms were moderately to markedly severe, while GAF scores suggested moderate impairment for hypochondriasis cases and moderate to serious impairment for dysmorphophobia cases. CGI-S and GAF inter-rater agreement were good for hypochondriasis and moderate for dysmorphophobia.ConclusionsThe Swedish ICD-10 codes for hypochondriasis and dysmorphophobia are sufficiently valid and reliable for register-based studies. The results of such studies should be interpreted in the context of a possible over-representation of severe and highly impaired cases in the register, particularly for dysmorphophobia.

Background The 14-item Short Health Anxiety Inventory (SHAI-14) is a common measure of health anxiety but its screening properties have not been studied. The aims of this study were to evaluate the SHAI-14 as a screening instrument, identify cut-offs for clinically significant health anxiety and investigate which scores correspond to different severity levels. Method The study included 1729 psychiatric patients and 85 healthy controls. Participants completed the SHAI-14 and underwent a diagnostic interview. Cut-off scores were evaluated in three scenarios to approximate screening 1) in a psychiatric clinic, 2) in a low prevalence setting and, 3) of healthy volunteers (cut-off for remission). Receiver operating characteristics were used. Classification of severity was based on the distribution of SHAI-14 scores reported by patients with clinically significant health anxiety. Results The area under the curve (AUC) values were high in all scenarios (above 0.95). The optimal cut-off scores on the SHAI-14 were 22 in the psychiatric context, 29 in a setting with low prevalence of psychiatric disorders and 18 versus healthy controls. SHAI-14 scores of 0–27 represented no or mild health anxiety, 28–32 moderate health anxiety and 33–42 substantial health anxiety. Conclusion Brief self-report measures used as screening instruments are a simple way of gathering information about the presence of specific symptoms and thus a way to detect the likelihood of a diagnosis. The SHAI-14 shows evidence of good diagnostic utility in both clinical and non-clinical settings. However, which cut-off score is to be used, depends on the intended purpose and the setting where the cut-off is used.

No abstract