Treatments of unilateral neglect: a review.

OBJECTIVES To review the existing literature on treatments of unilateral neglect, to synthesize findings, and to offer recommendations for future studies. DATA SOURCES Computerized databases including MEDLINE and PsychINFO. STUDY SELECTION All studies investigating treatment(s) of unilateral neglect. DATA EXTRACTION Authors reviewed design and other methodologic issues. DATA SYNTHESIS Unilateral neglect is a common consequence of right-hemisphere stroke. It is well recognized that the disorder is heterogeneous and has numerous subtypes. There have been numerous studies showing that arousal, hemispheric activation, and spatial attention treatments may all improve neglect, at least transiently. Despite these promising outcomes, little consensus exists as to whether 1 treatment is more efficacious than others, in part because cross-study differences in methodology render meta-analyses difficult, and in part because many studies fail to document duration of treatment effects or generalization to daily activities. One possibility is that these varied and diverse treatments may all be effective, reflecting redundancy in neural circuits devoted to attention and action in space, and consequent flexibility of the spatial processing system. It remains possible, however, that different subtypes of neglect may respond differentially to treatment of various sorts. Most existing studies of neglect have relied on very small populations of neglect patients, whose neglect is characterized only generally. CONCLUSION Methodologic shortcomings hinder assessment of the efficacy of various types of neglect treatment. In the future, these shortcomings could be addressed with larger studies of well-characterized patients that evaluate duration of treatment effects and include functional measures. In addition, the role of overarching variables, such as reduced arousal, requires consideration. The ultimate goal of these studies might be the development of triaging strategies wherein neglect patients are assigned to treatments of most likely benefit on the basis of neuroanatomic and behavioral profiles.