mortality, older patients, electronic health record, predictive model
Background: A prognostication predictive model incorporated into the electronic health record (EHR) may be useful in assisting the health care team in accurately predicting mortality and may be used in appropriately allocating palliative care services.
Purpose: To systematically review and summarize current medical literature regarding the factors predictive of mortality in an inpatient population above 65 years of age.
Methods: Nondisease-specific prognostication indices that predict 1-year mortality in an inpatient population of adults over age 65 were included. We excluded studies that estimated intensive care unit, disease-specific or in-hospital mortality. A MEDLINE, CINAHL, Ovid and Cochrane literature search of English-language articles that developed and/or validated a prognostication index to predict mortality was performed. Review of 3,600 citations revealed 53 articles that reported variables associated with mortality. Based on the inclusion criteria, 9 studies were included in the final analysis. Data was extracted from the 9 studies using the following parameters: adequate method of description of population, nonbiased selection of patients, low loss to follow-up, adequate prognostic factor measurements, adequate outcome measurements and methods of validation. We performed qualitative analysis on 5 studies and 4 studies were pooled for a quantitative meta-analysis.
Results: The 1-year mortality rate for the 21,338 patients included in all the studies was 31% (95% confidence interval [CI]: 31.3–32.6); mean age was 80.6 years. Factors significantly associated with mortality included male sex (odds ratio [OR]: 1.25, 95% CI: 1.09–1.42; P < 0.001), congestive heart failure (OR: 0.41, 95% CI: 0.37–0.45; P < 0.001), chronic obstructive pulmonary disease (OR: 3.2, 95% CI: 0.42–24.9; P = 0.26), myocardial infarction (proportion 0.39; P < 0.001), and cerebrovascular disease (proportion 0.38, 95% CI: 0.32–0.44; P < 0.001).
Conclusion: One-year mortality for inpatients aged > 65 years was high and associated with male sex, chronic obstructive pulmonary disease and congestive heart failure. Generalization of these findings to all older adults should be made with caution because of insufficient published information. In the future, our results may be used to develop a prognostication tool that draws patient data in real time from the EHR to identify vulnerable older adults in the hospital with end-of-life needs.
Khan A, Maria A, Hocker J, Singh M, Simpson M. Prognostic Indices for Hospitalized Older Adults: A Meta-Analysis and Systematic Review. J Patient Cent Res Rev 2015;2:206-207. http://dx.doi.org/10.17294/2330-0698.1227