Risks of preoperative anemia in women undergoing elective hysterectomy and myomectomy.

BACKGROUND

Hysterectomy is one of the most common surgeries performed worldwide. Identification of modifiable risk factors for complications or readmissions could lead to targeted interventions to improve patient care and reduce health care costs. Preoperative anemia has been identified as a risk factor for adverse postoperative outcomes following noncardiac surgery. However, studies have not focused on young and healthy surgical populations, such as women undergoing gynecologic surgery for benign indications.

OBJECTIVE

The purpose of this study was to evaluate whether preoperative anemia in women undergoing elective hysterectomy or myomectomy for benign indications was associated with increased 30 day postoperative morbidity and mortality.

STUDY DESIGN

In this retrospective, population-based cohort study, we followed up adult women (≥18 years of age) who underwent elective hysterectomy or myomectomy (laparoscopic/laparotomy) between the years 2013 and 2015 for benign indications in Ontario, Canada. We used linked administrative data from a government-administered, single-payer provincial health care system using Canadian Classification of Health Interventions intervention codes, International Classification of Diseases, 10th revision, diagnostic codes, physician billing codes, and laboratory data from both community and hospital laboratories across the province. Our exposure of interest was preoperative anemia, defined as a hemoglobin value <12 g/dL on the complete blood count measured closest to the date of surgery. Our primary outcome was the composite of 30 day postoperative morbidity and mortality. Secondary outcomes were 5 individual components of the primary outcome: death, transfusion, surgical site infection, venothromboembolism, and return to the hospital within 30 days. To adjust for confounding, we generated a propensity score using a multiple logistic regression model in which the presence of anemia was regressed on all baseline characteristics. We matched anemic to nonanemic patients on the logit of the propensity score. Using an unadjusted log-binomial model estimated using generalized estimating equations to account for the matched pairs, we calculated the relative risk, 95% confidence intervals, and P values to evaluate the effect of anemia on outcomes.

RESULTS

Of the 16,218 women in the cohort, 3664 (22.6%) had anemia. After propensity matching, standardized differences in all baseline characteristics (n = 3261 per group) were <0.10. In the matched cohort, the primary outcome (death, complications, or readmission) occurred in 41.2% of anemic patients and 36.2% of nonanemic patients (relative risk, 1.14, 95% confidence interval, 1.07-1.21, P < .0001; absolute risk reduction, 5.03%, 95% confidence interval, 2.70-7.36; (number needed to harm = 20). The risk of transfusion was significantly higher in anemic patients (relative risk, 3.25, 95% confidence interval, 2.67-3.95, P < .0001; absolute risk reduction, 8.34%, 95% confidence interval, 7.06-9.63; number needed to harm = 12). There was no difference in other secondary outcomes. In a subgroup analysis (women >55 years vs ≤55, n = 736), older women were at increased risk of the primary outcome (relative risk, 1.40, 95% confidence interval, 1.12-1.76, P = .004), transfusion (relative risk, 4.20, 95% confidence interval, 1.65-10.72, P = .003), surgical site infection (relative risk, 1.35, 95% confidence interval, 1.01-1.81, P = .04), and return to the hospital (relative risk, 2.36, 95% confidence interval, 1.54-3.62, P < .0001).

CONCLUSION

Preoperative anemia in women undergoing elective hysterectomy/myomectomy was common and is an independent risk factor for 30 day postoperative adverse outcomes, especially in older women.