Anticipated Impact on Veterans' Healthcare It is generally accepted that diabetes patients have a high risk for post-operative infections, but evidence specifically on infection risks in the dental surgical field is currently scant. The growing number of veterans with diabetes eligible for VA dental services warrants strengthening the evidence base to guide their dental care and prevent possible morbid complications. Project Background/Rationale Currently much is feared but little is known about post-operative complications of oral surgery. Existing dental education and practice guidelines cite excess infectious risk among diabetic patients. However, empiric evidence for such assertions is lacking. Nevertheless, dentists commonly prescribe antibiotics for dental surgical procedures that invade bone, with 20% of patients receiving antibiotics on the day of an invasive procedure in the VA, and as many as 60% outside it. This practice may contribute to the rising problem of microbial resistance and increase overall healthcare costs. Project Objectives We propose a retrospective cohort study using VA data of dental osseous procedures at VA Medical Centers.
Our specific aims are: 1. Among all patients undergoing osseous dental surgery, (approximately 180,000 between 1998-2008), to test the hypothesis that (H1) incidence of infectious complications is higher in patients diagnosed with diabetes than in non-diabetic patients. 2. Among diabetic patients undergoing osseous dental surgery (approximately 50,000 between 1998-2008), to test the following hypotheses: (H2) Uncontrolled diabetes and/or microvascular complications are independent risk factors for post-operative infection following dental osseous surgery. (H3) Diabetic patients who received prophylactic antibiotics have similar risk for infectious complications as those who did not. Methods This retrospective cohort study on patients who have undergone VA dental osseous procedures (identified using ICD-9, ADA and CPT codes) will examine infectious complications (identified through a new antibiotic prescription on days 4-30 post-operatively, or a revisit with an associated ICD9 code for infection). We will utilize national VA and data, including PBM, from 1998-2008 to carry out the aims and to test H1 and H2. We will confirm our detection methods using medical record review of 1000 records of Birmingham VAMC patients. We will test our hypotheses using multivariable logistic regression modeling, including patient characteristics available in VA data, and adjusting for clustering of patients within VA dental clinics using hierarchical modeling. Results from this study will not only provide important evidence to minimize risks of a common complication following oral surgical procedures, but will serve as the foundation for future studies examining sources of practice variation, as well as intervention studies to reduce risks among diabetes patients, should they be at elevated risk.
Dentists do not know whether diabetic patients who have dental surgery needed antibiotics routinely. Antibiotic resistance is a growing problem, so this information would be useful. The mouth has especially good circulation;it possible that infection is not a big enough risk to warrant routine use of antibiotics. This study will use existing information in the VA's medical records to study whether antibiotics are needed, and if so, which types of patients, including diabetes patients, may benefit the most.
