Background Growth in hospice utilisation has been accompanied by an increase in the proportion of hospice patients who die in an inpatient hospice setting rather than at home.

Objective To determine whether this increase in inpatient utilisation is consistent with patient preferences.

Design Retrospective cohort study.

Setting Seven hospices in the Coalition of Hospices Organised to Investigate Comparative Effectiveness (CHOICE) network.

Patients 70 488 patients admitted between 1 July 2008 and 31 May 2012.

Measurements We measured changes in patients’ stated preferences at the time of admission regarding site of death, including weights to adjust for non-response bias. We also assessed patients’ actual site of death and concordance with patients’ preferences.

Results More patients died receiving inpatient care in 2012 as compared to 2008 (1920 (32.7%), 2537 (18.5%); OR 1.21; 95% CI 1.19 to 1.22; p<0.001). However, patients also expressed an increasing preference for dying in inpatient settings (weighted preferences 27.5% in 2012 vs 7.9% in 2008; p<0.001). The overall proportion of patients who died in the setting of their choice (weighted preferences) increased from 74% in 2008 to 78% in 2012 (p<0.001).

Limitations This study included only seven hospices, and results may not be representative of the larger hospice population.

Conclusions Although more patients are dying while receiving inpatient care, these changes in site of death seem to reflect changing patient preferences. The net effect is that patients in this sample were more likely to die in the setting of their choice in 2012 than they were in 2008.