Date of Award


Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)



First Advisor

Jonathan N. Grauer



Background – Total hip arthroplasty (THA) is one of the most frequently performed orthopaedic procedures. As the prevalence of diabetes mellitus (DM) continues to increase the burden of its sequelae and associated surgical complications have also increased. For these reasons, it is important to understand the associations between DM and the rates of perioperative adverse events in patients with DM who will undergo THA.

Methods – The American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) database records perioperative adverse events as well as patient factors including demographics and comorbidities. Patients who underwent THA between 2005 and 2014 were identified and characterized as having insulin dependent diabetes mellitus (IDDM), non-insulin dependent diabetes mellitus (NIDDM), or neither. Multivariate Poisson regression was used to assess the relative risk of multiple adverse events in the initial 30 postoperative days while controlling for demographic and comorbid factors.

Results – A total of 71,733 patients who underwent THA were identified (1,920 IDDM, 6,305 NIDDM, and 63,508 without DM). Relative to patients without diabetes, patients with NIDDM were at an increased relative risk for 3 of 17 adverse events studied. These were sepsis or septic shock, readmission to hospital within 30 days, and extended postoperative length of stay (LOS) (greater than 5 days). Patients with IDDM were at an increased relative risk for 11 of 17 adverse events studied. These included death, sepsis or septic shock, myocardial infarction, wound-related infection, unplanned intubation, renal insufficiency, return to the operating room, readmission, pneumonia, urinary tract infection, and extended LOS. Patients with IDDM and NIDDM were both at greater risk for sepsis or septic shock, readmission, and extended LOS. Patients with IDDM were at greater risk for all of these adverse events (sepsis or septic shock: relative risk [RR] = 3.53 versus 1.90, for IDDM and NIDDM respectively, readmission: RR = 2.11 vs. 1.28, and extended LOS: RR = 2.26 vs. 1.35).

Conclusions – Compared to patients with NIDDM, patients with IDDM are at greater risk for many more perioperative adverse events relative to patients without diabetes. These findings have important implications for patient selection, preoperative risk stratification, and postoperative expectations.


This is an Open Access Thesis.

Open Access

This Article is Open Access