Is Undiagnosed Dysglycemia Common in Orthopaedic Surgery and Is it Associated With 30-day Outcomes? A Large-database Study

Background Given that patients undergoing orthopaedic surgery frequently share risk factors that are associated with metabolic syndrome, the patients orthopaedic surgeons treat may be especially likely to have undiagnosed dysglycemia. Although its broader impact is documented in other specialties, its specific prevalence and association with postoperative outcomes after orthopaedic procedures remains incompletely defined. Quantifying this hidden burden is important to developing individual risk profiles, refining perioperative decision-making, and establishing targeted preoperative medical management strategies for patients considering orthopaedic surgery. Questions/purposes In a large-database study, we asked: (1) In adults who underwent orthopaedic surgery, what proportion of patients had previously undiagnosed dysglycemia? (2) Was undiagnosed dysglycemia associated with increased odds of 30-day readmission and reoperation compared with patients with normoglycemia? (3) What was the relationship between preoperative HbA1c levels and the 30-day risk of complications in patients who underwent orthopaedic surgery? Methods Among the 2,990,063 procedures recorded in the American College of Surgeons National Surgical Quality Improvement Program database (2021-2023), 25% (735,259 of 2,990,063) were orthopaedic. After excluding 75% (2,254,804 of 2,990,063) of nonorthopaedic procedures and 16% (493,272 of 2,990,063) of orthopaedic procedures without documented preoperative HbA1c, 8% (241,987 of 2,990,063) of all procedures were analyzed. Patients were assigned to one of eight glycemic groups (normoglycemic, undiagnosed prediabetes, undiagnosed diabetes, or diagnosed diabetes with near-normal, very good, good, poor, and very poor glycemic control) based on diabetes diagnosis and HbA1c level. The study cohort included patients undergoing a range of procedures, primarily consisting of total joint arthroplasty, fracture fixation, and spine surgery. The mean ± SD patient age was 66 ± 12 years and 56% [134,741 of 241,987] of patients were female. To address the first research question, the percentage of patients with previously undiagnosed dysglycemia in each cohort was measured, supplemented by a best-case sensitivity analysis to provide a conservative estimate for the entire surgical population. The second and third research questions investigated the association of undiagnosed dysglycemia on 30-day odds of readmission and reoperation and the relationship between preoperative HbA1c and 30-day odds of complications, using multivariable logistic regression adjusted for demographic, comorbidity, and procedural factors. To illustrate variation across operations, the 30-day complication rate of patients across HbA1c strata was evaluated for the 10 most frequent primary Current Procedural Terminology–coded procedures. Results We found that 22% (52,609 of 241,987) of patients with a recorded preoperative HbA1c had undiagnosed prediabetes and 4% (10,690 of 241,987) of patients met the criteria for undiagnosed diabetes. In a best-case sensitivity analysis in which we assumed that all patients without an HbA1c measurement did not have diabetes and had normoglycemia, the smallest percentage of patients with previously undiagnosed dysglycemia would have been 9% (63,299) across all 735,259 patients who underwent orthopaedic surgery. After adjusting for potentially confounding variables, neither prediabetes nor diabetes was independently associated with increased odds of 30-day readmission (adjusted OR 0.95 [95% CI 0.88 to 1.0]; p = 0.12 and adjusted OR 0.99 [95% CI 0.84 to 1.2]; p = 0.94) or reoperation (adjusted OR 0.94 [95% CI 0.85 to 1.0]; p = 0.25 and adjusted OR 0.99 [95% CI 0.84 to 1.2]; p = 0.94) compared with patients with normoglycemia. The adjusted ORs for any 30-day complication were higher in patients with undiagnosed diabetes (adjusted OR 1.1 [95% CI 1.0 to 1.2]; p = 0.03) and increased across HbA1c strata in those with diagnosed diabetes (adjusted OR 1.2, 1.1, 1.2, 1.4, and 2.0; all p ≤ 0.009) compared with patients with normoglycemia. Procedure-level analyses showed consistent patterns with variable magnitude: 30-day rates of any complication in THA rose from 5% (1425 of 27,812) to 17% (49 of 294) across HbA1c strata, whereas 30-day complication rates peaked at 30% (203 of 675) in hinged TKA and at 22% (11 of 49) in revision TKA. On the other hand, 30-day complication rates in shoulder procedures and knee arthroscopy remained uniformly low (< 6%) with minimal variation by HbA1c. Conclusion Unrecognized dysglycemia is common among patients undergoing orthopaedic surgery and is associated with postoperative morbidity. Selective preoperative HbA1c assessment and procedure-specific targets may improve perioperative risk stratification. Examining these effects in clinical studies with procedure-specific preoperative HbA1c screening to identify and appropriately treat patients at risk could help to further validate these findings. Level of Evidence Level III, therapeutic study.