Postoperative Hysterectomy Complications: Comparing Novel Versus Traditional Hemoglobin A1c Cut-Offs.

STUDY

OBJECTIVE

To investigate the correlation between HbA1c and complications after hysterectomies.

DESIGN

Retrospective cohort analysis.

SETTING

Cerner Health Facts (519 million patient encounters from 750 hospitals).

PATIENTS: Patients undergoing laparoscopic ± robotic-assisted, vaginal, or abdominal hysterectomy, January 2010 to November 2018. We included patients with perioperative HbA1c values, including those diagnosed with diabetes or undergoing screening.

INTERVENTIONS: We performed univariable and multivariable logistic regressions with the outcome a composite of any post-op complication. We adjusted for HbA1c as continuous and categorical variables using HbA1c (>8.0% based on guidelines and >8.85% based on receiver operating characteristic analysis from current dataset) as cut-points.

MEASUREMENTS AND MAIN

RESULTS

Of 86,420 hysterectomies, 2693 (3.1%) had nonexcluded HbA1c, including 514 (19.1%) with complications and 2179 (80.9%) without. Median time to complication was 18 days. Patients were 75.2% white, 83.2% in urban hospitals. Route was predominantly minimally invasive: 30.5% laparoscopic, 25.6% vaginal, 43.9% abdominal, with 69.8% adnexa removed. Patients with complications were older (60.5 ± 15.9 vs 58.4 ± 14.2 years, p = .004), more likely from the Southern US (24.5% vs 18.5%, p = .006), used tobacco (26.7% vs 19.0%, p < .001), and were obese (32.3% vs 23.3%, p < .001). Most complications were infectious (87.4%). Higher HbA1c was associated with more complications, whether continuous (p = .03), HbA1c > 8.0% (p = .01), or HbA1c > 8.85% (p = .001). HbA1c > 8.85% optimized test characteristics (sensitivity = 0.14, specificity = 0.91, PPV = 0.27, NPV = 0.82). Predictive value improved with confounders (area under the curve = 0.71-0.72), and multivariable analysis demonstrated HbA1c > 8.85% increased complications (odds ratio [OR] = 1.71, 95% CI = 1.25-2.32). Obesity (OR = 1.31, 95% CI = 1.04-1.64) and urogynecological diagnoses, including urinary frequency, history of urinary tract infection, and dysuria (OR = 2.31, 3.71, 3.98, respectively), also increased complications.

CONCLUSION

HbA1c as a continuous variable impacted complications, so HbA1c reductions will likely decrease complications. For surgeons deciding to proceed with hysterectomy based on preoperative HbA1c, 8.0% and 8.85% are both acceptable cut-points. Multivariable models outperformed HbA1c alone, warranting further research to predict complications.