Background/Purpose:   Dermatomyositis (DM) and polymyositis (PM) are debilitating inflammatory myopathies with five-year mortality rates estimated to be 33%.  However, the leading causes of inpatient mortality in these conditions have not been fully explored.  In this study, we used a multi-state sample from the Healthcare Cost and Utilization Project (HCUP) to evaluate the sociodemographic and clinical predictors of inpatient mortality in patients with DM/PM.

Methods:   Using the 2009 HCUP State Inpatient Databases for California, New York, Florida, Washington, and Utah, we performed a retrospective study of individuals >18 years of age who met a validated administrative definition of DM/PM.  The primary outcome was death during hospitalization.  Principal diagnoses were determined according to Clinical Classification Software (CCS) codes and variables for interstitial lung disease (ILD), infection, malignancy, and cardiovascular disease (CVD) were generated based on any primary or secondary ICD-9 code falling within these respective diagnostic categories.  A modified Charlson Index was calculated excluding the above diagnoses.  Logistic regression was used to investigate the relationship between inpatient mortality and sociodemographic characteristics (age, gender, race, income), admission characteristics (length of stay, weekend admission), and associated diagnoses (ILD, infection, malignancy, CVD).

Results:   3,300 admissions with PM/DM were identified and inpatient mortality was 3.9% (128 deaths).  Subjects had an average age of 60 years (SD 17), 68% were female, 62% had DM and 38% had PM. In unadjusted analyses, age (65 years in those who died vs. 60 years in those who survived), weekend admission (27% vs. 19%), length of stay (17 vs. 7 days), more comorbidities (Charlson score of 1.4 vs. 1.3), ILD (22% vs. 12%), infection (76% vs. 36%), malignancy (21% vs. 9%), and CVD (51% vs. 29%) were all associated with mortality.  Leading principal diagnoses in subjects who died were septicemia (24%), respiratory failure (17%), pneumonia (9%), aspiration pneumonitis (9%), and acute cerebrovascular disease (5%).   In adjusted logistic regression, age (OR 1.09, 95% CI 1.02-1.17), weekend admission (OR 1.5, 95% CI 1.0-2.3), length of stay (OR 1.05, 95% CI 1.03-1.07), ILD (OR 2.0, 95% CI 1.2-3.1), infection (OR 4.3, 95% CI 2.8-6.5), malignancy (OR 3.2, 95% CI 1.9-5.2), and CVD (OR 2.4, 95% CI 1.6-3.5) were independently associated with mortality (see Table).

Conclusion:   Among hospitalized patients with DM/PM, infection, CVD and factors associated with the disease process (ILD, malignancy) were leading causes of mortality.    Sepsis was the leading cause of death in these patients, suggesting that additional attention to infection prevention in both the inpatient and outpatient settings may significantly improve outcomes.