SLR - August 2015 - Dae Sik Alex Kim
Association Between Diabetes, Obesity, and Short-Term Outcomes Among Patients Surgically Treated for Ankle Fracture
Reference: Cavo MJ, Fox JP, Markert R, Laughlin RT. Association Between Diabetes, Obesity, and Short-Term Outcomes Among Patients Surgically Treated for Ankle Fracture. J Bone Joint Surg Am. 2015 Jun 17;97(12):987-94.
Scientific Literature Review
Reviewed By: Dae Sik Alex Kim, DPM
Residency Program: Wyckoff Heights medical Center
Podiatric Relevance: In the United States, an ankle fracture is one of the more commonly encountered orthopedic injuries, with about 25 percent of these injuries undergoing surgical fixation. Obesity can be seen in up to 50 percent of diabetic adults. The relationship between obesity and postoperative complications following surgical fixation of ankle fractures is unclear. Although surgical fixation of ankle fracture is safely accomplished, up to 8 percent of cases are complicated by mechanical and infectious wound. Poor postoperative outcomes are associated with concurrence of venous stasis, peripheral neuropathy, peripheral vascular insufficiency, and diabetes. This study aims to see if there is a significant relationship between obesity and diabetes and their effect on in-hospital length of stay, complication frequency, and hospital costs.
Methods: Using NIS data from 2001-2010, a cross-sectional study using data from 1000 nonfederal hospitals, was conducted to evaluate the relationship between diabetes and obesity in hospital admissions with a primary diagnosis of ankle fracture or dislocation. Patients were divided into four categories: Group A (patients without diabetes or obesity), Group B (obesity only), Group C (diabetes only), and Group D (both diabetes and obesity). Patients who exhibited poly-trauma were excluded from the study. The three outcomes that were examined in this study were: length of stay, in-hospital complications, and extrapolated in-hospital costs. Covariates to describe the patient population were also collected in order to run statistical analyses. These patient demographics were age, sex, race, quartile of income based on zip code, type of hospitalization, and concurrent diagnosis of venous stasis disease. To calculate the odds ratio (with a 95 percent confidence interval) for the association between diagnostic categories and in-hospital complications, a logistic regression model was constructed. In order to compare the hospital costs and length of stay between the diagnostic groups, a linear regression model was constructed. For both statistical analyses, a multivariable logistic regression model was used to adjust for the different demographic features. In the linear regression model for hospital costs, an additional variable was added to account for variances in regional rages. A p-value of 0.05 was considered significant.
Results: The final sample consisted of 148,483 discharges for a surgically treated ankle fracture between January 2001 and December 2010. Of the total sample population, 78.4 percent qualified for group A (neither diabetes nor obesity), 5.0 percent for Group B (obesity alone), 13.6 percent for Group C (diabetes alone), and 3 percentfor Group D (both obesity and diabetes). The median age was 53.0 years with 62.2 percent being female and 62.2 percent had either a closed bimalleolar or trimalleolar fracture. The ankle fracture subtype was not explicitly coded in 20.9 percent of the discharges. A concurrent diagnosis of venous stasis, peripheral neuropathy, or peripheral vascular disease was uncommon. The frequency of in-hospital complications was the highest for Group D (both diabetes and obesity) at 6.5 percent and lowest for Group A (neither diabetes nor obesity) at 2.6 percent. The mean length of stay and hospital costs also increased from Group A to Group D. After adjusting for sociodemographic factors, there was an increased likelihood of developing a complication of care if patients were in Groups B, C, or D. Older patients with peripheral vascular disease and other co-morbidities were also found to have experienced greater complication of care. Female sex was associated with fewer complications.
Conclusion: Patients who had any combination of diabetes and obesity experienced more frequent in-hospital visits, longer length of stays, and higher hospital bills compared to patients who were non-diabetic or non-obese. Despite the diagnoses of diabetes and obesity, this study found that the more co-morbidities a patient had, the more complications they experienced. The frequency of complications was also increased for patients undergoing different surgical procedures such as open fracture and dislocation compared to a closed unimalleolar fracture. Undiagnosed diabetes remains a huge confounder in this study, since symptoms can remain dormant for years before a diagnosis of diabetes is actually made. Due to high rate of associated comorbid medical conditions in this population, the ability to clearly define the relationship between obesity and postoperative outcomes is made difficult.
Due to the high prevalence of diabetic and obese patients, the purpose of this study, in trying to establish a relationship between ankle fractures and hospital complications, is very relevant. One of the strengths of this study was running the multivariable logistic regression models in an attempt to decrease confounders. However, due to the latency of the onset of diabetes and disease prognosis, there is the possibility that some of the subjects in the non-diabetic group actually had diabetes at the time of this study. Another limitation of this study was the inability to stratify patients according the severity of their ankle fractures. Further investigation into categorizing obesity into different BMI categories is worth noting for future reference. Overall, this study found that obesity and diabetes were linked to greater hospital complications, longer hospital stays, and greater hospital bills.