Introduction

With the implementation of the Bundled Payments for Care Improvement Initiative (BCPI), there is increasing interest in prudent patient selection for shoulder arthroplasty procedures. No consensus exists on which patient comorbidities contribute most to the risk of post-operative healthcare utilization, cost, and complications. To potentially avoid this and incentivize providers to care for these higher risk patients, it is critical to identify patient and pathologic factors that are associated with higher rates of health care utilization, costs, and complications. This study aimed to explore relationships between preoperative comorbidities and inpatient hospital costs, complications, non-home discharge, and length of stay (LOS) for shoulder arthroplasty procedures to aid in risk stratification reimbursement rates as a part of the BCPI.

Methods

The National Inpatient Sample (NIS) was queried from 2011 to 2015 for all patients, aged 40 to 95, undergoing elective anatomical, reverse, or partial shoulder arthroplasty. Patient comorbidity status was assessed through the use of clinical comorbidity software on the NIS database. A comorbidity was included in our model if it was represented in >1% of the sample. A multivariate linear model utilizing gamma regression with a log-link function was used to assess the contribution of each comorbidity on inflation-adjusted cost, while controlling for patient and hospital specific factors. Poisson Regression was used to calculate the relative risk of each comorbidity on the development of major and minor peri-operative complications, non-home discharge, and an extended LOS past 3 days, while controlling for patient and hospital specific factors.

Results

300,933 patients aged 40-95 years old were identified as undergoing an elective shoulder arthroplasty between 2011 and 2015. The median age was 70 years old and 55% of patients were female. The most common comorbidities were hypertension (67.9%), diabetes mellitus (20.6%), chronic pulmonary disease (18.3%), obesity (16.4%), and hypothyroidism (16.3%). The greatest contributors to cost were coagulopathy ($3,017), pre-operative electrolyte abnormality ($2,279), congestive heart failure ($2,082), anemia ($1,167), and a psychotic disorder ($1,038). The comorbidities with the greatest risk of non-home discharge destination included psychotic disorders (RR: 2.12), alcohol abuse (RR: 1.78), and congestive heart failure (RR: 1.45). The comorbidities most associated with an extended LOS included an electrolyte abnormality (RR:1.81), psychotic disorder (RR:1.55), and coagulopathy (RR: 1.54). The comorbidities with the highest risk of a minor complication included electrolyte abnormalities, coagulopathy, and renal failure. Additionally, the comorbidities that put patients at the highest risk of major complications included valvular heart disease, congestive heart failure, and coagulopathy. When accounting for prevalence of comorbidity, the greatest contributors to overall yearly health care expenditures were a pre-operative electrolyte abnormality ($13,674,367), obesity ($10,167,822), and anemia ($7,884,980)

Discussion

Our data identifies numerous comorbidities associated with increased healthcare utilization, cost, and complication rates. Furthermore, different comorbidities are associated with varying degrees of contribution to these metrics. Understanding comorbidity contribution to cost can aid health care administrators to adjust reimbursement according to risk. Additionally, clinicians should understand comorbidities association with complications to help guide the decision-making process for choosing to proceed with elective shoulder arthroplasty.