Background: A geriatric screening tool would be valuable to identify elderly cancer patients who might benefit from a comprehensive geriatric assessment (CGA). This study evaluated the accuracy of the cardiovascular health study (CHS) instrument in predicting abnormality in CGA. The vulnerable elders' survey-13 (VES-13) was also evaluated.
Patients and methods: Patients aged ≥ 70 years with solid tumors underwent a CGA after being screened with the CHS and VES-13. Analyses were conducted for the overall population and according to the disease status (early or advanced) and type of early cancer (breast or gastrointestinal, GI).
Results: Of 259 patients, 75% were impaired according to the CHS and 47% according to the VES-13. CGA impairment was reported in 171 patients (66%). In the overall population, overall accuracy, sensitivity and specificity of CHS in identifying CGA impairments were 74%, 87% and 49%, respectively. The corresponding figures for the VES-13 were 68%, 62% and 81%. Sensitivity and specificity of CHS in predicting CGA impairments in subgroups were early 81% and 55%, advanced 98% and 29%; early breast 78% and 69%, early GI 87.5% and 19%.
Conclusions: The CHS compared favourably with VES-13 for sensitivity. However, the great variability in specificity observed with the CHS within subgroups limits its applicability in the global population.