Assessing Juveniles Who Commit Murder

Psychiatric Times

May 2005

Vol. XXII

Issue 6

In their evaluations of 18 males in Texas condemned to death for homicides committed before their 18th birthdays, Dorothy Otnow Lewis, M.D., and colleagues relied on neuropsychiatric, neuropsychological, and educational examinations and assessments (J Am Acad Psychiatry Law 2004; 32(4):408-429).

"The psychiatric evaluation was based on a modified form of the Bellevue Adolescent Interview Schedule and the Dissociative Disorders Interview Schedule [DDIS]," Lewis told Psychiatric Times. "We used a cheat sheet to remind us of the questions from the interviews, but followed a more conversational approach in delivering the questions."

Lewis and colleagues noted that topics in the psychiatric evaluation included histories of neuropsychiatric symptoms, psychiatric treatment, medical history, characteristics of temper, family mental health histories, and histories of child physical and sexual abuse and family violence. An inmate's face, head and body were also examined for scars. Before the modified Bellevue Adolescent Interview was used in Texas, Lewis told PT, it was pre-tested at a residential treatment facility for delinquent children in another state.

"We trained the staff in its use. They liked it so much, they are using it as part of their intake," she said.

The neurological examination not only included mental status; cranial nerve function; and motor, sensory and cerebellar function, it also checked for the presence of frontal lobe deficits. Several tests were used in the neuropsychological assessments. Among them were the Test of Memory Malingering (TOMM), the Personality Assessment Inventory (PAI), the Weschsler Adult Intelligence Scale-III (WAIS-III) and tests of frontal lobe functioning. One tool, Lewis said, that was particularly helpful was the Iowa Gambling Task, an unstructured test of executive function developed by Bechara and colleagues (Cognition 1994; 50(1-3):7-15).

"We used the computerized version of Iowa Gambling Task even though it was still in the developmental stage," Lewis said. "It tests one's ability to learn and modify one's behavior in response to negative consequences."

In the task, the player chooses from any of four decks of cards to win fake money. The object of the game is to maximize winnings and minimize losses. At the outset, the player is told that with two of the decks (decks A and B) they can get big wins but also terrible losses and those decks should be avoided. When the game was completed, the juveniles were asked to describe their decision-making strategy.

"Some of those we tested told us they recognized they should avoid the risky decks, but couldn't stop themselves from pulling from them," she said. "One described it this way: 'Deck B was paying so good, it was hard to get off. C and D paid chump change, but they weren't losing that much. A and B, they were really snatching me!' As a result of frontal lobe impairment, they seemed unable to control their immediate impulses and act in accordance with their intellectual understanding."

The majority of the men (84%) exhibited significant impairment on the Iowa Gambling Task. They tended to select cards from the risky decks with about the same frequency at the end of the game as at the beginning.

While the comprehensive assessments took a long time, Lewis and her colleagues emphasized their importance.

"The diagnostic evaluation of a violent person is a weighty responsibility, but the evaluation for purposes of determining guilt, innocence, or mitigating factors in murder cases is weightier still," they wrote. "The heaviest burden of all is the diagnostic evaluation of a violent juvenile who has become involved with the adult criminal justice system."