Reducing Infection Rates with Enhanced Preoperative Diabetes Mellitus Diagnosis and Optimization Processes.

Given the increased risk of perioperative morbidity and mortality from hyperglycemia, the authors' practice identified a process for perioperative diagnosis of diabetes mellitus (DM) and glycemic optimization. Their practice incorporated a reflective hemoglobin A1c (HgA1c) for random plasma glucose ≥200 mg/dL. Patients with uncontrolled DM (ie, HgA1c ≥8%) or new-onset DM (ie, HgA1c ≥6.5%) underwent specialty consultation. The authors' process allowed for detection of undiagnosed diabetes mellitus, perioperative diabetic optimization, and lowered infection rates in the process. Objectives: Hyperglycemia and increased preoperative hemoglobin A1c (HbA1c) are associated with perioperative morbidity and death. For nonurgent operations, adequate glycemic control before surgery is recommended. Our surgical practice needed a process for preoperative diabetes mellitus (DM) diagnosis and glycemic optimization. Methods: Our review of the existing preoperative evaluation process found that patients without a DM diagnosis but with random plasma glucose ≥200 mg/dL received no additional screening. Patients with DM routinely receive neither preoperative HbA1c screening nor DM management when HbA1c is ≥8.0%. Results: A new preoperative evaluation process was designed. HbA1c screening was automatically performed for patients with random plasma glucose ≥200 mg/dL. For patients with a DM diagnosis, an HbA1c test was performed. Specialty consultation was prompted for patients with known DM and HbA1c ≥8.0% and those with no DM diagnosis but HbA1c ≥6.5%. In the first year postimplementation, 9320 patients received a basic metabolic panel; 263 had random plasma glucose ≥200 mg/dL that triggered an HbA1c check. In total, 123 patients (99 with and 24 without a DM diagnosis) were referred to endocrinology; 13 received a new DM diagnosis. Twenty patients had surgery delayed for DM treatment. All of the patients received individualized medication instructions for the perioperative period. Among patients with random plasma glucose ≥200 mg/dL, incidence rates for surgical site infection pre- and postimplementation were 47.8/1000 and 3.8/1000 population. Conclusions: The implemented process benefited patients scheduled for nonurgent procedures by optimizing glucose control and lowering infection rates through earlier preoperative DM diagnosis, glycemic management, and standardized patient medication instruction. [ABSTRACT FROM AUTHOR]