Background Estimates suggest that around 6000 premature cancer deaths could be prevented each year in Britain if survival matched that of the rest of Europe. Earlier diagnosis may close this gap, particularly for cancers like myeloma, which require early intervention to prevent complications, such as bone destruction and renal failure. Unfortunately myeloma may be associated with prolonged time to diagnosis; multiple primary care consultations; fewer urgent referrals for suspected cancer; and more emergency presentations (EP). We investigated route to diagnosis of myeloma using data from a specialist UK population-based registry (www.hmrn.org), examining associations with socio-demographic characteristics and prognostic variables, and evaluating the impact on survival.

Methods All patients in the Haematological Malignancy Research Network have a core dataset (demographic, prognostic, treatment, outcome data) routinely abstracted from hospital records. Additional data were abstracted for myeloma patients diagnosed between July 2012 and December 2013, including types and dates of referrals, and specialities involved. Prognosis was based on CRAB criteria (calcium, renal failure, anaemia and bone lesions), and the International Staging System. Descriptive data are presented, with age and sex-specific background mortality rates obtained from national life tables; and relative survival estimates are given. Route to diagnosis was categorised from first referral (i.e. furthest from diagnosis) and survival was calculated from date of diagnosis.

Results With a median diagnostic age at diagnosis of 74 years, 442 patients (251 males and 191 females) were diagnosed with myeloma during the study period. EP was the most common route to diagnosis (29.2%), followed by GP urgent (20.8%), GP two-week wait (TWW) with suspected cancer (18.9%), GP routine (15.3%), and consultant-to-consultant referrals (7.9%); a further 7.9% were already being monitored by haematology. EP was associated with advanced stage disease, increased likelihood of complications and poorer 1-year relative survival, which was 84.4% (Confidence intervals 80.0–87.9) overall, compared to 71.7% (95% Confidence Interval 61.9–79.4%) for EP and 89.4% (95% CI 78.2–95.0%) for TWW referrals.

Conclusion The benefits of avoiding emergency presentation are obvious, in terms of both survival outcomes and prevention of complications. Difficulties arise in determining how best to achieve this in a cancer that is relatively infrequently encountered by GPs, and that often presents with symptoms (e.g. bone pain) that are comparatively common in the general population, and which may be erroneously attributed to the ageing process. Characterisation of events in primary care would improve understanding of referral practices and inform strategies to prevent emergency presentation.