Bridging from clinical endpoints to estimates of treatment value for external decision makers

C. W. Zhu, C. Leibman, R. Townsend, T. McLaughlin, N. Scarmeas, M. Albert, J. Brandt, D. Blacker, M. Sano, Y. Stern

Research output: Contribution to journalArticlepeer-review

9 Scopus citations

Abstract

Aim: While clinical endpoints provide important information on the efficacy of treatment in controlled conditions they often are not relevant to decision makers trying to gauge the potential economic impact or value of new treatments. Therefore, it is often necessary to translate changes in cognition, function or behavior into changes in cost or other measures, which can be problematic if not conducted in a transparent manner. The Dependence Scale (DS), which measures the level of assistance a patient requires due to AD-related deficits, may provide a useful measure of the impact of AD progression in a way that is relevant to patients, providers and payers, by linking clinical endpoints to estimates of cost effectiveness or value. The aim of this analysis was to test the association of the DS to clinical endpoints and AD-related costs. Method: The relationship between DS score and other endpoints was explored using the Predictors Study, a large, multi-center cohort of patients with probable AD followed annually for four years. Enrollment required a modified Mini-Mental State Examination (mMMS) score ≥30, equivalent to a score of approximately ≥16 on the MMSE. DS summated scores (range: 0-1) were compared to measures of cognition (MMSE), function (Blessed Dementia Rating Scale, BDRS, 0-17), behavior, extrapyramidal symptoms (EPS), and psychotic symptoms (illusions, delusions or hallucinations). Also, estimates for total cost (sum of direct medical cost, direct non-medical cost, and cost of informal caregivers' time) were compared to DS scores. Results: For the 172 patients in the analysis, mean baseline scores were: DS: 5.2 (SD: 2.0), MMSE: 23.0 (SD: 3.5), BDRS: 2.9 (SD: 1.3), EPS: 10.8%, behavior: 28.9% psychotic symptoms: 21.1%. After 4 years, mean scores were: DS: 8.9 (SD: 2.9), MMSE: 17.2 (SD: 4.7), BDRS: 5.2 (SD: 1.4), EPS: 37.5%, behavior: 60.0%, psychotic symptoms: 46.7%. At baseline, DS scores were significantly correlated with MMSE (r=-0.299, p<0.01), BDRS (r=0.610, p<0.01), behavior (r=.2633, p=0.0005), EPS (r=0.1910, p=0.0137) and psychotic symptoms (r=0.253, p<0.01); and at 4-year follow-up, DS scores were significantly correlated with MMSE (r=-0.3705, p=0.017), BDRS (r=0.6982, p<0.001). Correlations between DS and behavior (-0.0085, p=0.96), EPS (r=0.3824, p=0.0794), psychotic symptoms (r=0.130, ns) were not statistically significant at follow-up. DS scores were also significantly correlated with total costs at baseline (r=0.2615, p=0.0003) and follow-up (r=0.3359, p=0.0318). Discussion: AD is associated with deficits in cognition, function and behavior, thus it is imperative that these constructs are assessed in trials of AD treatment. However, assessing multiple endpoints can lead to confusion for decision makers if treatments do not impact all endpoints similarly, especially if the measures are not used typically in practice. One potential method for translating these deficits into a more meaningful outcome would be to identify a separate construct, one that takes a broader view of the overall impact of the disease. Patient dependence, as measured by the DS, would appear to be a reasonable choice - it is associated with the three clinical endpoints, as well as measures of cost (medical and informal), thereby providing a bridge between measures of clinical efficacy and value in a single, transparent measure.

Original languageEnglish (US)
Pages (from-to)256-259
Number of pages4
JournalJournal of Nutrition, Health and Aging
Volume13
Issue number3
DOIs
StatePublished - Mar 2009

ASJC Scopus subject areas

  • Medicine (miscellaneous)
  • Nutrition and Dietetics
  • Geriatrics and Gerontology

Fingerprint

Dive into the research topics of 'Bridging from clinical endpoints to estimates of treatment value for external decision makers'. Together they form a unique fingerprint.

Cite this