Supplemental modules pertaining to elder mistreatment should be included in existing comprehensive geographic health and social surveys, including ongoing longitudinal studies of aging populations. These studies will require the use of short instruments, or a series of questions, designed to identify likely victims of elder mistreatment. For reasons of economy, an alternative is to use these supplemental modules to target only selected forms of mistreatment, such as physical mistreatment, neglect, and financial exploitation. The unique contribution of such studies is to provide a large national sample from which reliable prevalence estimates can be drawn. Of equal importance, however, is the ability to use the longitudinal data to identify risk factors, further define health and social outcomes, and serve relevant policy needs. Elder mistreatment modules appended to existing national surveys can also serve as a test bed for new scientific approaches to data collection. Such piggy-backing of elder mistreatment items and instruments is logistically feasible in most contexts, and attempt should be made to further this application.
Once the measurement issues have been satisfactorily addressed, a comprehensive national prevalence study of elder mistreatment should be undertaken. The purpose of this study would be to generate useable national estimates of prevalence and the critical demographics for each of the principal forms of elder mistreatment (physical mistreatment, sexual mistreatment, emotional mistreatment, financial exploitation, and neglect).
Both the supplemental module studies and the national prevalence study must ultimately address family and nonfamily settings, including nursing homes and the full range of assisted living arrangements and other community-based locations in which vulnerable older persons reside. Without such information, policy makers and program developers have no empirical basis for assessing the needs of elder mistreatment victims or for deciding how much to invest in research and prevention programs.
In addition to improved household and geographically referent sampling techniques, new methods of sampling and identifying elder mistreatment victims in the community should be developed in order to improve the validity and comprehensiveness of elder mistreatment occurrence estimates. It is likely that household sampling, while extremely useful, will be incomplete to some degree because of difficulty in gaining access to those households and respondents most at risk of elder mistreatment. A particular problem is accessing and characterizing the wide variety of assisted living and related residential facilities where many vulnerable elders are located. Developing additional ways to approach and access these populations may require other sampling techniques, such as through social networks, institutions, or the health care system.
Research is needed on the phenomenology and clinical course of elder mistreatment. The clinical course, antecedents, and outcomes of the