Most research on drug treatment of human behavioral disorders involves only measurement of changes in behavioral rate or severity. The outcomes of analyses based solely on these changes do not permit an analysis of the actions of drugs on the environmental variables that may be exerting control on the target behavior of interest. Thus, little is known about the behavioral mechanisms of actions of psychotropic drugs. Accumulated evidence suggests that biological (e.g. opioid peptides) and environmental (e.g. social interaction) variables influence self-injurious behavior (SIB), a common form of behavioral disorder in persons with neurodevelopmental disorders. To date, however, no studies have examined directly whether pharmacological treatment alters the environmental mediation of self-injury.

In this study, the sequential dependency between staff behavior and self-injury was examined during naltrexone and placebo treatment conditions in a sample of 10 adults with severe self-injury secondary to neurodevelopmental disorders. For all subjects, naltrexone effects were examined in the context of an ABAB placebo, double-blind design. During each phase, we directly observed each subject two to three times per day in fifteen-minute sessions for a mean total of 136 fifteen-minute sessions per subject (range: 130-142), or approximately 34 hours per subject (range: 33.0-35.5). Direct observational data were collected in real time by trained observers using hand-held microcomputers (Psion HC3As). The Observer and MOOSES software packages were used for downloading and analysing data on a Dell PC.

Subject and staff behavioral data were analyzed for frequency and duration, inter-observer agreement and sequential dependencies. For sequential analysis, sequential dependencies were analyzed as either event- or time-based. The sequential analysis option on MOOSES is designed to calculate sequential dependencies between identified antecedent and target events. The analysis allows the counting of occurrences of a single or combined behavior code, designated as the target, within a time window or event lag from another selected code (i.e. the 'given' or antecedent). The event-based method counts the number of times codes follow other codes at discrete event steps or lags. The time-based method tallies the frequency with which codes follow other codes within designated intervals of time. In general, this analysis procedure determines whether the probability that one event or behavior leads to another is significantly different from that expected by chance. Conditional probabilities, Allison-Liker Z scores and Yule's Q were computed for the resulting frequency tables for each analysis.

Naltrexone-related reductions (>33%) in the daily rate of SIB were observed for 50% of the subjects. For the subjects who displayed a positive response to naltrexone, the magnitude of the sequential dependency between staff behavior and subject self-injury was significantly greater during treatment with naltrexone than during placebo. These results are discussed in relation to the utility of observational research methods for determining potential behavioral mechanisms of action-regulating medication effects for behavioral disorders.

Paper presented at Measuring Behavior 2002 , 4^th International Conference on Methods and Techniques in Behavioral Research, 27-30 August 2002, Amsterdam, The Netherlands

© 2002 Noldus Information Technology bv