Diabetes mellitus (DM) is a serious consideration in orthopaedic surgery, from its influence on injury risk, to its effect on treatment outcomes. Although there is variability in bone mineral density based on type of DM, DM has been identified as an independent risk factor for hip fractures.1–3 The presence of DM has also been shown to increase morbidity and mortality in patients with hip fractures.4–8 DM has a well-established association with surgical site infection (SSI) and perioperative infection likely related to perioperative hyperglycemia.9–11 Described as “stress hyperglycemia,” posttrauma/postoperative hyperglycemia has been shown to be associated with more perioperative infections in nondiabetic patients after orthopaedic trauma.12–15

Owing to the rising prevalence of DM and its association with worse functional outcomes after acetabular fracture, patients with acetabular fractures and comorbid DM are an especially important study population.16 Although evidence exists for both DM and hyperglycemia associated with higher risk for orthopaedic infections, data are lacking regarding the potential effect on acetabular fracture patients. Similarly, the role of peri-injury hyperglycemia as it pertains to infection has predominantly been investigated in nondiabetic patients. Drawing from a 20-year patient registry, we aimed to investigate associations of DM and acetabular fracture fixation complications. We hypothesized that patients with DM would have more SSIs than nondiabetic patients and that diabetic patients with hyperglycemia during hospitalization would have greater risk of infection.

The study was done at a single, level 1 trauma center in the Midwest region of the United States. An IRB-approved review was done of 1141 patients who presented with acetabular fracture (AO/OTA type 62 injuries) between 1999 and 2019.17 All 134 diabetic patients in the registry were identified, along with a random sample of 345 nondiabetic patients as determined to be necessary by a priori power analysis. All patients were followed clinically for a minimum of 6 months. Overall, 83 diabetics and 270 nondiabetics were managed with fixation of their acetabulum fracture. Diabetic patients were characterized based on glycemic control, as measured by hemoglobin A1c and perioperative hyperglycemia. Patients with two or more inpatient blood glucose readings >200 mg/dL were considered to be hyperglycemic. Insulin dependence was also recorded. Patients treated exclusively with oral agents were considered non–insulin-dependent.

Data points included patient age, sex, American Society of Anesthesiologists (ASA) class, body mass index (BMI), and number of comorbidities excluding DM (categorized by cardiovascular disease, pulmonary disease, tobacco use, renal disease, gastrointestinal disease, hepatobiliary disease, rheumatologic/autoimmune disease, endocrine [non-DM], neurologic disease, or psychiatric disease). Acetabular fracture pattern, surgical approach, Injury Severity Score (ISS), and concomitant injuries were also recorded.17,18 Type of treatment was at surgeon discretion based on fracture pattern and patient risk factors.

Complications included SSIs (superficial and deep), deep vein thrombosis (DVT), pulmonary embolism (PE), pneumonia, urinary tract infection (UTI), and bacteremia or sepsis. Superficial infection was defined as patients with SSI managed with oral antibiotics and local wound care. Deep infection described SSIs that required surgical débridement and intravenous antibiotics. Early infections were identified during hospitalization and included SSIs, along with pneumonia, UTI, bacteremia, or sepsis. These infectious complications are described throughout the text as “early infection” when describing the whole cohort and “postoperative infection” when describing the surgically managed cohort.

All data were stored electronically and analyzed using the Statistical Package for Social Sciences (SPSS) (version 28; IBM SPSS). Demographic variables were assessed about both diabetes status and in-hospital infection using the Student t-test for continuous variables and the Pearson chi-square test for categorical variables. Multivariate binary logistic regression was done for any postoperative infection in patients managed surgically to account for possible confounders. Variables with univariate P < 0.2 were included in the analysis.

Four hundred seventy-nine patients with acetabular fractures were included in the study, and 28% had DM (n = 134) of whom 49% were insulin-dependent (Table 1). Those with DM were older (57.2 versus 43.2 years; P < 0.001) and had worse baseline heath status, with more medical comorbidities (1.9 versus 0.70; P < 0.001), higher BMI (33.6 versus 29.5; P < 0.001), and higher ASA class (2.9 versus 2.2; P < 0.001). Table 2 presents patients treated surgically versus nonsurgically for the entire cohort. Those who underwent fixation of their acetabular fracture were younger (45.6 versus 51.3; P = 0.003), had fewer comorbidities (0.9 versus 1.3; P = 0.004), and had higher ISS (18.5 versus 13.7; P < 0.001). Patients with associated hip dislocation (P < 0.001), femoral head injury (P = 0.04), or marginal impaction (P < 0.001) more frequently underwent surgical management. Frequency of surgical management based on Letournel fracture classification and other features of the fractures was also determined (Table 3).

Patients with DM were less often treated with surgery for their fracture (61.9 versus 78.2%; P < 0.001). Among diabetic patients, those who had surgery were significantly younger (54.6 versus 61.4 years; P = 0.01), had fewer comorbidities (1.7 versus 2.2; P = 0.04), and had higher ISS (17.9 versus 13.6; P = 0.01) (Table 3). Those treated surgically also had more associated hip dislocation (P < 0.001), femoral head injury (0.024), and marginal impaction (P = 0.003). A trend for lower peri-injury hemoglobin A1c (6.9 versus 7.6; P = 0.09) was noted among those treated surgically.

Superficial infection was defined as patients with SSI managed with oral antibiotics and local wound care. Deep infection described SSIs that required surgical débridement and intravenous antibiotics. Early infections were identified during hospitalization and included SSIs, along with pneumonia, UTI, bacteremia, or sepsis. Across the entire cohort, patients with DM were more likely than those without DM to develop any early infection (P = 0.049) (Table 1). UTI and bacteremia were more common in patients with DM (17.9% versus 11.6% [P = 0.07]; 6.7% versus 3.2% [P = 0.08], respectively); however, neither carried statistical significance. Deep SSIs were infrequent in both groups (1.5% and 2.0%, respectively). Rates of pneumonia, DVT, and PE were also comparable between patients with and without DM.

Of patients treated with open reduction and internal fixation for their acetabulum fracture (n = 353), 24% had DM (Table 4). Those with DM had higher BMI (34.2 versus 29.8; P < 0.001), more comorbidities (1.7 versus 0.70; P < 0.001), and higher ASA status (2.9 versus 2.2; P < 0.001). SSIs were infrequent in both diabetics and nondiabetics (2.4% and 3%, respectively). With the limited numbers of patients available, no association with type of surgical approach (anterior or posterior) was identified. Notably, anterior approaches were either ilioinguinal or anterior intrapelvic in combination with lateral and middle ilioinguinal windows. Posterior approaches were Kocher Langenbeck. No notable differences in other infections or other early complications were noted, comparing patients with and without DM. There was a trend for lower peri-injury hemoglobin A1c (6.9 versus 7.6; P = 0.09) for diabetics who underwent fixation. Surgically managed patients more frequently experienced hyperglycemia while in the hospital than nonsurgically managed patients (P = 0.004).

Surgically managed patients who developed SSIs and general postoperative infection were both compared with those who did not. Higher ASA class (P = 0.03), more baseline comorbidities (P = 0.04), and baseline pulmonary disease (P < 0.001) were associated with SSI on univariate analysis (Table 5). None of the recorded variables were found to be predictors of SSI on multivariate analysis. Several variables were associated with development of any postoperative infection, including ASA class, age, length of stay, ISS, and insulin use (Table 5). Notably, only older age (P = 0.03), longer length of stay (P < 0.001), baseline pulmonary disease (P = 0.02), and presence of abdominal injury (P < 0.001) were predictors for infectious outcome on multivariate analysis (Table 6). The relationship between insulin dependence and postoperative infection did not persist on multivariate analysis.