Impact of malignancy on In-hospital mortality, stratified by the cause of admission: An analysis of 67 million patients from the National Inpatient Sample

Objective To describe the patient characteristics and the reason for admission of patients with malignancy by malignancy, and to study mortality rates for the different causes of admissions among the different types of cancer. Patients and Methods Using the nationwide Inpatient Sampling (2015-2017) we examined the cause of admission and associated in-hospital mortality, stratified by presence and type of malignancy. Multivariable logistic regression models were used to examine the association between in-hospital mortality and malignancy sites for different primary admission causes. Results Out of 67 819 693 inpatient admissions, 8.8% had malignancy. Amongst those with malignancy, haematological malignancy was the most common (20.2%). The most common cause of admission amongst all cancers were malignancy-related admissions, where up to 57% of all colorectal admissions were malignancy-related. The most common non-malignancy cause of admission was infectious causes, which were most frequent among patients with haematological malignancy (18.4%). Patients with malignancy had higher crude mortality rates (5.7% vs 1.9%). Mortality rates were highest among patients with lung cancer (8.7%). Among all admissions, the adjusted rates of mortality were higher for patients with lung (OR 3.65, 95% CI [3.59-3.71]), breast (OR 2.06, 95% CI [1.99-2.13]), haematological (OR 1.79, 95% CI [1.76-1.82]) and colorectal (OR 1.71, 95% CI [1.66-1.76]) malignancies compared with patients with no malignancy. Conclusion Our work highlights the need to consider the burden of cancer on our hospital services and consider how the prognostic impact of different types of admissions may relate to the type of cancer diagnosis and understand whether these differences relate to disparities in clinical care/treatments.