The field of geriatric assessment is crowded by a variety of assessment domains, a plethora of assessment tools, and research spanning diverse care settings. In their article published in this issue of the journal ONCOLOGY, Schubert, Gross, and Hurria have synthesized the evidence and propose a subset of commonly used functional assessment tools for assessing older adults with cancer.[1] Although the authors present a helpful summary of physical, cognitive, psychosocial, and other relevant domains and present a well-formed argument for their integration into the care of older cancer patients, their efforts represent only part of the information required for the translation of evidence into practice.
Aging is a highly individualized and complex process. Comprehensive geriatric assessment (CGA) is multidimensional and identifies a range of patient factors that are used to formulate an individualized care plan for clinical management. Studies of CGA interventions in various forms and settings have shown positive health effects in older populations.[2-5] Unfortunately, the efficacy data currently available regarding CGA only allow for analysis of indirect evidence supporting the use of cancer-specific CGA.[6,7] Without cancer-specific CGA data on such outcomes as choice of treatment, treatment tolerance, treatment completion, survival, disease-specific survival, quality of life, hospitalizations, and nursing home admissions, and without studies that determine which domains (for both patient and caregiver) and which measures are most useful, we simply do not have the knowledge base to translate the use of cancer-specific CGA into evidence-based practice.
Prospective Studies
Nearly a decade's worth of publications, including recommendations from the International Society of Geriatric Oncology (SIOG) task force on CGA, underscore the need for prospective studies to determine cancer-specific CGA's ability to predict relevant outcomes.[6,8-10] Notwithstanding, to our knowledge, there are few prospective outcome-based studies of cancer-specific CGA.[11-14] This is a notable start, but is insufficient to provide the evidence necessary for translation.
It is accepted that well-conducted randomized controlled trials (RCTs) provide the highest level of evidence to guide clinical management. However, conducting RCTs in vulnerable patient populations such as older cancer patients is challenging, and oncology treatment trials have documented low participation rates among older adults.[15,16] Barriers to participation and retention include study design; physician, patient, and logistic issues (eg, availability of caregivers, travel constraints); and financial costs. Additionally, the heterogeneous nature of older cancer patients necessitates large samples and/or increased duration of observation to achieve adequate study power.
Nonetheless, prospective clinical trials of cancer-specific CGA are feasible, albeit more difficult and expensive to conduct and complete. They are critically needed to provide the evidence required to redesign the care of older adults and improve outcomes.
Retrospective Studies
Outcome-based retrospective cohort and case-control studies evaluating the effectiveness of cancer-specific CGA are alternatives to RCTs, but are hitherto unrepresented in the literature. Retrospective studies circumvent the challenges of enrollment, retention, and attrition, as well as the high costs of prospective studies, by using existing data sources. If not properly designed, however, they can be more prone to confounding and bias.
In the case of cancer-specific CGA, the primary challenge of conducting retrospective studies is the scarcity of programs with sizeable numbers of patients for whom pre- and posttreatment data are available. This likely explains the absence of studies using these designs in the literature.
