Routes to diagnosis and the association with the prognosis in patients with cancer - A nationwide register-based cohort study in Denmark.

Abstract

BACKGROUND\nThe prognosis of cancer is related to how the cancer is identified, and where in the healthcare system the patient presents, i.e. routes to diagnosis (RtD). We aimed to describe the RtD for patients diagnosed with cancer in Denmark by using routinely collected register-based data and to investigate the association between RtD and prognosis measured as one-year all-cause mortality.\n\n\nMETHODS\nWe conducted a population-based national cohort study by linking routinely collected Danish registry data. We categorised each patient into one of eight specified RtD based on an algorithm using a stepwise logic decision process. We described the proportions of patients with cancer diagnosed by different RtD. We examined associations between RtD and one-year all-cause mortality using logistic regression models adjusting for sex, age, cancer type, year of diagnosis, region of residence, and comorbidity.\n\n\nRESULTS\nWe included 144,635 cancers diagnosed in 139,023 patients in 2014-2017. The most common RtD were cancer patient pathway from primary care (45.9 %), cancer patient pathway from secondary care (20.0 %), unplanned hospital admission (15.8 %), and population-based screening (7.5 %). The one-year mortality ranged from 1.4 % in screened patients to 53.0 % in patients diagnosed through unplanned hospital admission. Patients with an unplanned admission were more likely to die within the first year after diagnosis (OR\u2009=\u20093.38 (95 %CI: 3.24-3.52)) compared to patients diagnosed through the cancer patient pathway from primary care.\n\n\nCONCLUSION\nThe majority of cancer patients were diagnosed through a cancer patient pathway. The RtD were associated with the prognosis, and the prognosis was worst in patients diagnosed through unplanned admission. The study suggests that linking routinely collected registry data could enable a national framework for RtD, which could serve to identify variations across patient-, health-, and system-related and healthcare factors. This information could be used in future research investigating markers for monitoring purposes.