Impact of delay in diagnosis on survival to head and neck carcinomas: a systematic review with meta-analysis

Objective: To address the contradictory information on the role of delay in diagnosis on head and neck cancer survival. Study design: Systematic review and meta-analysis. Search strategy: Search on MEDLINE (1966 to March 2011), EMBASE (1980 to March 2011) and ISI proceedings (from inception to March 2011). The terms used were ('Head and neck cancers') AND ('delay' OR 'prognostic' OR 'survival') both in MeSH terms and free-text words. The reference lists of the retrieved articles were also revised manually to identify other potentially relevant papers. All searches were independently undertaken by two clinicians and one epidemiologist, and the results merged. Setting: Primary and specialised care levels. Participants: Meta-analysis of data from papers on the subject published from 1966 to 2011. Main outcome measures: Survival. Methods: After search in Medline and other databases, we computed pooled relative risks (RR) and 95% confidence interval (95% CI) from the 10 studies retrieved. Results: The estimate of the relative risk of mortality related to any diagnostic delay (either patient or professional delay) was 1.34 (95% CI 1.12-1.61). Referral delay was associated with a three-fold increase in mortality. Total delay was marginally related to mortality (RR: 1.04, 95% CI: 1.01-1.07). By anatomic location, pharynx cancer shows the highest association (RR: 1.68, 95% CI: 1.22-2.31). Conclusions: Diagnostic delay is a moderate risk factor of mortality from head and neck cancer. However, part of the effect observed may be due to residual confounding (confounding from unknown variables that are not eliminated by adjustment).