Factors associated with improved survival among older colorectal cancer patients in the US: a population-based analysis
1 Boston Health Economics, Inc, Waltham, USA
2 Health Economics and Outcomes Research, GE Healthcare, Waukesha, USA
3 Institute for Clinical and Evaluative Sciences, Sunnybrook Health Sciences Center, Toronto, Canada
4 Harvard Medical School, Boston, USA
BMC Cancer 2009, 9:227 doi:10.1186/1471-2407-9-227Published: 13 July 2009
The purpose of this study was to estimate the relative impact of changes in demographics, stage at detection, treatment mix, and medical technology on 5-year survival among older colorectal cancer (CRC) patients.
We selected older patients diagnosed with CRC between 1992 and 2000 from the SEER-Medicare database and followed them through 2005. Trends in demographic characteristics, stage at detection and initial treatment mix were evaluated descriptively. Separate multivariate logistic regression models for colon (CC) and rectal cancer (RC) patients were estimated to isolate the independent effects of these factors along with technological change (proxied by cohort year) on 5-year survival.
Our sample included 37,808 CC and 13,619 RC patients (combined mean ± SD age: 77.2 ± 7.0 years; 55% female; 87% white). In recent years, more CC patients were diagnosed at Stage I and fewer at Stages II and IV, and more RC patients were diagnosed at Stage I and fewer at Stages II and III. CC and RC patients diagnosed in later years were slightly older with somewhat better Charlson scores and were more likely to be female, from the Northeast, and from areas with higher average education levels. Surgery alone was more common in later years for CC patients while combined surgery, chemotherapy, and radiotherapy was more common for RC patients. Between 1992 and 2000, 5-year observed survival improved from 43.0% to 46.3% for CC patients and from 39.4% to 42.2% for RC patients. Multivariate logistic regressions indicate that patients diagnosed in 2000 had significantly greater odds of 5-year survival than those diagnosed in 1992 (OR: 1.35 for CC, 1.38 for RC). Our decomposition suggests that early detection had little impact on survival; rather, technological improvements (e.g., new medical technologies or more effective use of existing technologies) and changing demographics were responsible for the largest share of the change in 5-year survival in CC and RC between 1992 and 2000.
Technological advances and changes in patient demographics had the largest impact on improved colorectal cancer survival during the study period.