Re. Myers et al., ANALYSIS OF COLORECTAL-CANCER STAGE AMONG HMO MEMBERS TARGETED FOR SCREENING, Archives of internal medicine, 157(17), 1997, pp. 2001-2006
Background: This study is a retrospective analysis of data collected f
rom patient medical records, a fecal occult blood test (FOBT) screenin
g program, and computerized health maintenance organization (HMO) clai
ms and encounters records. Objective: To identify factors associated w
ith a diagnosis of early (Dukes A and B) colorectal cancer among older
adults targeted for annual FOBT screening.Methods: Study subjects wer
e insured by the former US Healthcare Inc (Blue Bell, Pa), an independ
ent practice association-type HMO. The HMO was recently integrated int
o Aetna-US Healthcare. Before diagnosis, subjects were eligible for fr
ee annual FOBT screening through the HMO's colorectal cancer screening
program. The study subjects included men and women (N = 222) who were
aged 50 years or older and had a diagnosis of colorectal cancer betwe
en 1987 and 1990. Variables considered were patient age, gender, socio
economic status, medical history, screening history, length of enrollm
ent in the HMO, and stage of disease at diagnosis. Results: Univariate
analyses indicate that colorectal cancer diagnosis due to FOBT screen
ing (P = .03), frequency of FOBT screening (P = .09), and length of HM
O membership (P = .10) were positively related to being diagnosed as h
aving early stage colorectal cancer. Multivariable analysis shows that
hating a screen-detected colorectal cancer was significantly and posi
tively related (P = .03) to being diagnosed as having early stage dise
ase. Conclusions: Findings support annual FOBT screening among older a
dults. Results illustrate the value of applying standard methods to th
e collection and analysis of patient data in a managed care context. T
he study also highlights a need for research on patient adherence to s
creening and physician follow-up of abnormal screening test results.