Validity of the Sims scales of neurologic impairment and amnestic disorder
Zack Cernovsky, Yves Bureau, James Mendonça, Varadaraj Velamoor, Stephan Mann, Gurpreet Sidhu, David M Diamond, Robbie Campbell, Emmanuel Persad, L Kola Oyewumi, Michel A Woodbury-Fariña
Background: The Structured Inventory of Malingered Symptomatology (SIMS) has never been scientifically validated to demonstrate that it adequately differentiates legitimate patients from malingerers. Method: We evaluated the content validity of the Amnestic Disorder (AM) and Neurologic Impairment (NI) scales of the SIMS via expert ratings by 11 experts (7 psychiatrists, 2 psychologists, and 2 neuroscientists). As the next step, SIMS scores of 23 patients injured in high impact car accidents were selected: seminal research of Bennet Omalu suggests that such collisions would rarely occur without cerebral micro vascular trauma and axonal shearing. Such patients typically suffer from neurological symptoms in the post-concussion whiplash spectrum. Criterion validity of AM and NI scales was assessed by statistical comparisons of AM and NI data of these 23 presumably legitimate patients to published SIMS data on instructed malingerers (30 malingerers of whiplash and 26 malingerers of post-concussive symptoms) and also to published SIMS data on relatively healthy normal controls responding honestly (N=34). Results: The review of item content of the AM and NI scales by the 9 expert clinicians and 2 neuroscientists indicated that all their items represent potentially legitimate medical symptoms: none was being judged as specific to malingering. The SIMS AM and NI scores diagnosed close to 75% of our 23 survivors of high impact motor vehicle accidents (MVAs) as “malingerers:” it is unlikely that there were so many true positives. An ANOVA of NI scores showed that neither malingerers of whiplash nor malingerers of post-concussive symptoms differed significantly from legitimate patients (p>.05) and that normal controls had significantly lower NI scores than both groups of malingerers and also lower scores than the legitimate patients (p<.0001). An ANOVA of AM scores showed that legitimate patients obtained higher AM scores than malingerers of whiplash (p=.0002), but lower scores than malingerers of post-concussive symptoms (p=.0008). In this ANOVA, normal controls had significantly lower AM scores than legitimate patients and also lower scores than malingerers of post-concussive symptoms (p<.001), but did not differ significantly from malingerers of whiplash (p>.05). Conclusions: The SIMS NI and AM scales obviously lack both in content and criterion validity, i.e., the NI and AM scales list legitimate neuropsychological symptoms and cannot adequately differentiate malingerers from legitimate patients.