Predictors of effort test failure during neuropsychological evaluation in cases of traumatic brain injury

The application of suboptimal effort during neuropsychological assessment is frequently encountered. The variables that predict suboptimal effort are not well understood. To date individual variables that are associated with suboptimal effort have been examined in an ad hoc fashion and there has been no attempt to construct an empirical model of cognitive test taker effort. The aim of the current series of four studies was to examine the relationship between economic, demographic, psychological, and personal history variables and effort test failure (ETF), using a multivariable statistical technique (logistic regression) that has been only rarely employed in effort research. In study 1 the power of economic, demographic, psychological and behavioural variables to predict ETF was examined in an archival consecutive sample of mixed-severity adult traumatic brain injury patients (N = 555). In study 2 the predictive power of psychological and personal history variables to predict ETF was further examined while holding constant the statistical predictors identified in study 1. Study 3 comprised an exploration of the predictive relationship between a range of acculturation variables and ETF while holding constant the predictive variables identified in study 1. In study 4 the relationship between self-reported depressive symptomatology (SRDS) and ETF in compensation-seeking samples was examined by undertaking a systematic review of the literature between 1950 and 2012 (inclusive). A total of 9,501 articles were screened, of which 19 satisfied inclusion criteria. The results of study 1 revealed ETF to be significantly associated with compensation-seeking, low education, self-reported mood disorder, exaggerated displays of behavior, psychotic illness, being foreign-born, having sustained a workplace accident, and mild as compared to severe traumatic brain injury. In study 2 it was demonstrated that, holding study 1 variables constant, of a range of psychological and personal history variables examined, only self-reported depressive disorder was predictive of ETF. Scores on the Beck Depression Inventory-II were predictive of ETF holding compensation-seeking constant. Of the range of acculturation variables examined in study 3, only age at which English was learned was found to make a significant independent contribution to the predictive model established in study 1. The systematic review of the literature that examined SRDS and ETF revealed that studies were of high quality but typically afforded a low level of evidence. The results of those studies revealed a medium to large effect of SRDS on test taker effort. Psychological symptom reporting was found to be elevated in this population but frank malingering was not detected. Together, the studies indicate that ETF can be predicted by psychological symptom reporting, the display of abnormal behaviours, economic variables, demographic variables, injury-related variables, and workplace variables.