[The author responds:]
John Hirdes' thoughtful comments deserve clarification. The epistemology of health outcomes in institutions is as yet poorly defined, although we generally agree on what should not happen to nursing home patients. What is surprising is the lack of empirical data supporting the effectiveness of interventions for common clinical problems, [1, 2] although fortunately this is changing. [3] Can the MDS help in this regard? It does provide standardized, risk-adjusted outcome data that can be used to compare facilities and specific patient subgroups. However, it does not further quality or quantify the interventions being carried out on each patient, nor does it support prognostication even at a crude level. Therein lies the problem. The mere existence of a care process does not ipso facto lend support to its effectiveness.
The MDS provides a sound, validated, systematic approach to care planning, costing and outcome evaluation. It should be broadly implemented not only in chronic care facilities but in home care, nursing home and ambulatory care settings. It is, however, only part of the answer. Formal research is needed to improve clinical care in these settings. Methinks inferences on the effectiveness of specific interventions on the basis of MDS data smack uncomfortably of deus ex machina.
Jean A. Chouinard, MD
Sisters of Charity at Ottawa Health Service; Ottawa, Ont.