The mismatch study: exploring the discrepancies between clinician's diagnosis, structured interview outcomes and patient self-report in PTSD

Abstract

A valid and reliable PTSD diagnosis is requisite t o give PTSD patients the treatment they need. Therefore, obtaining the right diagnosis is crucial . Multiple studies demonstrate that discrepancies exist between the several diagnostic instruments an d their outcomes. These discrepancies, or ‘mismatches’, might be partially explained by patie nt as well as clinicians' factors. The objective of the current study was to explore the amount of mism atch between the PSS-SR (self-report), SCID-I (standardized measure) and the clinician’s diagnosi s (clinical interview), as well as the possible explanations for this mismatch. This was done by p erforming a two-stage project. Study 1. The first stage was the pilot-study. Data were obt ained from the Jelgersma Clinic (Oegstgeest, The Netherlands). In a sample of borde rline patients (N=28) only a fair agreement (Cohen’s Kappa .267) was found between the clinicia n’s diagnosis and the PSS-SR, with the PSS-SR diagnosing PTSD more often than the clinicians' jud gment. Furthermore, the pilot study indicated that someone suffering from additional anxiety symptoms is more likely to be diagnosed with PTSD according to the PSS-SR. Study 2. The objective of Study 2 was to replicate and exte nd the findings from study 1 in a large scale group, with additional structured interview data on PTSS diagnoses, by investigating the amount of mismatch between the PSS-SR, SCID-I and the clinici an’s diagnosis. As a second objective, explanations were sought at the level of patient an d clinician characteristics for potential mismatch. The participants for Study 2 were all newly present ed patients at the Altrecht Academic Anxiety centre (Utrecht, The Netherlands) who received a pr imary PTSD diagnosis according to the SCID-I and on whom data were available on clinicians' diag nosis at intake and on PSS-SR (N=184). Between the SCID-I and the PSS-SR a mismatch of 8% was foun d and between the clinicians' diagnosis and SCID-I a mismatch of 31% was found. No agreement (C ohen’s Kappa .056) was found between the clinicians' diagnosis and the PSS-SR, with the PSS- SR diagnosing PTSD more often than the clinician. Study 2 did not find an indication that someone suffering from anxiety symptoms is more prone to receive a PTSD diagnosis. However, results indicated that suffering from depressive symptoms might enhance receiving a PTSD diagnosis a ccording to the PSS-SR and the clinicians' opinion. Educational level of the clinician did not have any influence on the degree of mismatch. However, clinicians with 0-1 years and >20 years of experience showed the highest degree of mismatch (in comparison to the SCID-I), with respec tively 50% and 42% mismatch. Overall, these findings are worrying with respect to the accuracy of diagnoses that clinicians make. The current project has been a first step to address this impor tant issue.