Prognostic Indices for Older Adults: A Systematic Review
What is the quality of prognostic indices for mortality for older adults?
This systematic review included literature searches from MEDLINE, EMBASE, Cochrane, and Google Scholar, from their inception through November 2011. Indices were included if they were validated, and predicted absolute risk of mortality in patients with an average of 60 years or greater. Indices that estimated intensive care unit, disease-specific, or in-hospital mortality were excluded. For each prognostic index, data on clinical setting, potential for bias, generalizability, and accuracy were collected.
A total of 21,593 titles were reviewed to identify 4,120 potentially relevant abstracts, from which 24 studies were identified. The authors identified 16 unique indices, which predicted risk of mortality from 6 months to 5 years for older adults in a variety of clinical settings, including the community (six indices), nursing home (two indices), and hospital (eight indices). At least one measure of transportability (the index is accurate in more than one population) was tested for all but three indices. No study was free from potential bias. Although 13 indices had C-statistics of 0.70 or greater, none of the indices had C-statistics of 0.90 or greater. Only two indices were independently validated by investigators who were not involved in the index’s development.
The authors concluded that several indices for predicting overall mortality in different patient groups exist. However, future studies need to independently test their accuracy in heterogeneous populations and their ability to improve clinical outcomes before their widespread use can be recommended.
As the authors point out, indices are often clinically useful and thus frequently incorporated into guideline recommendations. Understanding the quality and limitations of such indices improves their usefulness, but also assists in the development of more accurate indices. One key factor is the evaluation of indices in populations which reflect broader segments of the population.
Keywords: Prognosis, Intensive Care Units, Hospital Mortality, Cardiology, Bias (Epidemiology), Forecasting, Fibrinogen, Residence Characteristics, MEDLINE
< Back to Listings