The prediction of cause-specific and
overall mortality by co-morbidities in a registry-based cohort of colorectal cancer patients

Colorectal cancer (CRC) accounts for about 11% of all incident cancers in Switzerland with approximately 4000 new cases annually. Although survival is high when the cancer is detected at an early stage (90% 5-year survival rate for localized disease), only about 40% of all cases are diagnosed at an early stage. The presence of other chronic diseases at the time of diagnosis additionally affects the life expectancy of the patients, which, in turn, is – at least in theory – an important characteristic for treatment choice of the physician and the patient. Co-morbidity is defined as the co-existence of at least two chronic conditions or impairments in the same person. The Charlson co-morbidity index (CCI) is one of the most commonly used indices with respect to cancer. It has been developed with the aim of classifying co-morbid conditions that might alter the risk of mortality for use in longitudinal studies. Previous studies have shown that colorectal cancer patients with two or more co-morbidities have statistically significantly reduced survival compared to patients without co-morbidities, but most studies have been hospital-based, were small, confined to specific types of treatment, or were not able to examine whether the association between co-morbidity and survival was equally strong for all CRC stages.

The aim of this study is to examine the association of pre-diagnosis co-morbidities on survival after CRC diagnosis in patients of the Zurich Cancer Registry diagnosed in 2000/2001. The Zurich Cancer Registry was established in 1980 with the aim of collecting information on the diagnosis of cancer among inhabitants of the canton Zurich. The Zurich Cancer Registry is the largest registry in Switzerland with almost 7000 new cases per year. The cancer registry routinely collects information on type of cancer, name, date of birth, sex, place of residence, stage of the disease, information on treatment, and information on survival of the patient for selected types of cancer. Information on co-morbidities will be obtained from records from the hospitals, in which the patients have been treated. Including approximately 2000 CRC cases, Cox proportional-hazards regression will be used to estimate the hazard ratio (HR) and 95% confidence interval (CI) of death from causes other than CRC, all causes, or CRC, adjusting for age and pathologic stage. In addition, we will use relative survival models to estimate excess mortality among these patients, which has not yet been used to examine the effect of co-morbidities on survival of CRC patients. The Cancer Registry Zurich offers the opportunity to examine the association between co-morbidities and mortality in CRC patients in detail because of its long-standing collection of data and its collection of cases from a variety of hospitals with different patient characteristics.