Introduction Little is known about hospital admissions (HAs) following a diagnosis of incurable lung cancer (LC). This study sets out to identify temporal trends for HA in this group of vulnerable patients in addition to exploring the reasons behind and potential risk factors for HA.

Methods All new LC diagnoses for 2009–2011 (n = 565) were identified, from which 1:4 were selected randomly (n = 142). Records were reviewed and those patients treated with curative intent or diagnosed and died in same HA were excluded from analysis. Basic demographic data were collected including co-morbidity score, stage, histology and LC was classified as either central (mediastinum to origin of lobar bronchi and vessels) or peripheral after review of the diagnostic CT scan. The presence or absence of significant pleural effusion (>1.c.m. depth) and extra-thoracic metastases was noted. Simple non-parametric tests were used to identify any risk factors for HA.

Results 84 patients (mean age 70.3 years, 42 males) were suitable for inclusion, accounting for 98 HAs with median length of stay of 6 days. Of the 59 patients with HA, 63%, 22%, 6% and 9% experienced 1, 2, 3 or ≥4 HAs. The HA: patient ratio fell with time from 1.44 in 2009, 1.23 in 2010 to 0.86 in 2011. Survival figures were 13.1%, 28.6%, 23.8% and 34.5% for <3, 3–6, 6–9 and >9 months respectively. 76% of HAs occurred within 3 months of death.

The primary cause of HA was determined to be infection (33%), breathlessness (16%), neurological (14%), pain (10%), gastrointestinal symptoms (10%), others (17%). No obvious clinical risk factors for HA were found when comparing those patients having HA to those without HA (Table 1).

Conclusions HAs in incurable LC are common but difficult to predict.

Future strategies designed to prevent HA may need to focus more on social factors in addition to providing rapid treatment of infection and symptom palliation in the last 3 months of life.