The Journals of Gerontology Series A: Biological Sciences and Medical Sciences 60:323 (2005)
© 2005 The Gerontological Society of America
AUTHORS' RESPONSE
Marilyn J. Rantz,
Lanis Hicks,
Gregory Petroski,
Richard Madsen,
David Mehr,
Vicki Conn,
Mary Zwygart-Stauffacher and
Meridean Maas
MU MDS and Quality Research Team University of MissouriColumbia
Address correspondence to Marilyn J. Rantz, PhD, RN, Sinclair School of Nursing, S406, MU, University of MissouriColumbia, Columbia, MO 65211. E-mail: rantzm{at}missouri.edu
We appreciate Bruce Robinson's additional analysis of our method of classifying facilities into groups that achieve predominately good, average, and poor resident outcomes. His analysis confirmed our findings illustrated in Figure 1 of our article that a 5-month average window between sample selection and time of observation resulted in many facilities "moving" from one outcome group to another. In our follow-up analyses of statewide data explained in the article, we highly recommend that if a method such as the one we designed is used to classify facilities, the method should then require two consecutive 6 months of Minimum Data Set data to ensure more stable group classification.
However, we do not share his view that our results confirm that Minimum Data Set-based outcome data should not be provided to consumers. Our classification method used 23 quality indicators and attempted to use them in a composite measure for a research project. Aggregation of quality indicators into a single quality measure is a challenging and, based on our results, not exact science. Our results do not address individual quality indicator reliability. In fact, as displayed in Table 2, 10 of the 23 quality indicators were sensitive measures that discriminated well between the facilities that consistently were classified between groups with predominately good, average, or poor resident outcomes. Whether or not to use Minimum Data Set quality indicators in public reporting is beyond the scope of our article and does not follow from Robinson's analysis.