Factors associated with diagnostic and treatment intervals in colorectal cancer: A linked data study.

Abstract

This research aimed to assess the length of intervals before diagnosis and treatment for colorectal cancer in Australia using linked datasets, and to determine any factors associated with interval length. A colorectal cancer clinical registry was linked to general practice electronic medical record data and routinely collected hospital referral datasets to determine the length of four key intervals in the time before first treatment. Cox proportional hazards regression was used to assess associations between individual characteristics (sociodemographic variables such as age and sex, and disease characteristics such as cancer subtype and treatment approach) and the length of each interval. Sample sizes available for analysis varied by interval, ranging from 99 to 9359. The median interval length ranged from 21 (IQR 5-38) days for the time between diagnosis and treatment to 63 (IQR 24-218) days for the time between first presentation and diagnosis. Overall, few measured characteristics were associated with the lengths of any of the intervals. Of note, shorter diagnostic intervals were associated with presenting to the general practitioner with alarm symptoms, and people proceeding to surgery as initial treatment had shorter times to treatment than any other treatment modality. Given disease and medical system factors were associated with interval length, broad improvements to the overall efficient functioning of the healthcare system are likely to improve timeliness. More targeted interventions could focus on processes at the transitions between different levels of the healthcare system and implementing recommended maximum lengths of intervals along the diagnostic and treatment pathway.