Background: There are over 16 000 nursing homes in the United States (US), among which approximately 70% of residents are cognitively impaired. Reflecting this, approximately 20% of US nursing homes maintain Special Dementia Care Units (SCUs). SCUs supposedly provide more staff time and more specialized staff assignments to residents than do traditional care units.
Aims of the Study:
This paper addresses the issues of staff time and assignment: do the costs of personal care inputs differ according to whether they are provided by SCUs or in traditional care settings? Related to this, are differences associated with the different settings, or are they accounted for by resident characteristics within the settings? Methods: Given the bias generally associated with collection of staff time data, the author developed (supported by the Health Care Financing Administration and the National Institute on Aging) and used in this study a barcode-based system ('InfoAide'). Using InfoAide, each provider automatically recorded task-and residentspecific time expenditure data which were subsequently monetized, using prevailing local wage rates. Individual resident personal characteristics and status data were provided by another simultaneous study of SCU impacts among the same residents. Regression analysis (MANCOVA for significantly correlated dependent variables) was used to examine the relationships between cost and SCU/traditional status, and individual resident characteristics, separately for each category of provider. Results: Controlling for resident characteristics, the cost of aide care is significantly (positively) related (p Υ
0.01) to SCU status. Cognitive impairment, ADL impairment and being restrained are also related to higher aide care cost (p Υ
0.05, p Υ
0.01, and p Υ
0.05, respectively). The same is generally true of Speech Therapy, Social Service and care by licensed practical nurses, although the differences between SCU and traditional care units are essentially trivial-and there are no SCU/traditional care differences for registered nurses. Discussion: SCU/traditional unit status, even when combined with the central resident covariates, explains very little variance in service costs, other than among nursing aides; in separate MR analyses in which monetized service time was the dependent variable, the cumulative adjusted R 2 among aides was 0.37; for