Algorithm Reduces Unnecessary Oophorectomies in Pediatric Patients

The use of a consensus-based preoperative risk stratification algorithm has been shown to significantly decrease unnecessary oophorectomies for benign ovarian neoplasms in pediatric patients, according to a study published in JAMA.

Surgeon holding scalpel near ovary | Image Credit: shidlovski – stock.adobe.com

The preoperative risk assessment algorithm can identify ovarian lesions with a strong likelihood of being non-cancerous and suitable for ovary-preserving surgery.

According to the study authors, the adoption of such an algorithm has the potential to mitigate the risk of unnecessary oophorectomy—the surgical removal of one or both ovaries, also referred to as ovariectomy—during adolescence and subsequent lifelong consequences of the procedure. Additional research is required to investigate the obstacles that may hinder compliance with this algorithm.

The study was conducted across 11 children's hospitals in the United States, including 519 patients aged between 6 and 21 years with a median (IQR) age of 15.1 (13-16.8) years. The intervention consisted of 3 phases: 6 months of preintervention assessment, followed by 6 months of intervention adoption, and 18 months of active intervention. There were 96 children included in the preintervention phase, 105 in the adoption phase, and 318 in the intervention phase. The intervention adoption group was excluded from statistical comparisons.

Nearly all children had a benign neoplasm, present in 93 (96.9%) children in the preintervention cohort and 298 (93.7%) in the intervention cohort.

The study demonstrated a substantial decrease in unnecessary oophorectomies during the intervention phase, with a notable absolute reduction of 7.7% (95% CI, 0.4%-15.9%; P = .03). While 15 of the 93 (16.1%) children with a benign neoplasm in the preintervention group received an unnecessary oophorectomy, only 25 of 298 (8.4%) children underwent the surgery during the intervention period.

The algorithm's test performance in identifying benign lesions showed a sensitivity of 91.6% (95% CI, 88.5%-94.8%) and a specificity of 90.0% (95% CI, 76.9%-100%), underlining its accuracy in distinguishing between benign and malignant cases. The researchers also noted a positive predictive value of 99.3% (95% CI, 98.3%-100%) and a negative predictive value of 41.9% (95% CI, 27.1%-56.6%).

During the intervention phase, there was a misclassification rate of 0.7%, referring to instances where malignant disease was mistakenly treated with ovary-sparing surgery. The algorithm demonstrated a strong adherence rate of 95.0% during this phase, indicating surgeons' willingness to adhere to the algorithm, though the lower fidelity level of 81.8% suggests there were obstacles hindering complete implementation.

“Adherent cases with incomplete fidelity with the algorithm were often due to a discrepancy between the initial and final interpretations of imaging,” the authors noted. “Additionally, many patients had an incomplete panel of tumor marker data available, which may have resulted from clinicians not having certain tumor markers evaluated based on clinical judgment.”

On this note, the authors said further investigation is needed to explore ways to enhance algorithm implementation for the benefit of a larger patient population. For example, a notable finding was that among the 25 unnecessary oophorectomies during the intervention phase, 9 occurred in patients with suspected torsion.

According to the authors, this implies that improved adherence to the algorithm's directives for detorsion and the thorough completion of preoperative risk stratification, followed by a definitive procedure, could potentially further decrease the incidence of unnecessary oophorectomies.

A short list of limitations was included in the study. First, the clinical practices at the participating sites started evolving during the study's planning phase, which led to a reduction in the anticipated percentage of unnecessary oophorectomies from an expected 27% to 16% in the preintervention cohort. Second, this study was conducted exclusively at tertiary children's hospitals with pediatric surgical subspecialists, which may limit the broader generalizability of the results.

Reference

Minneci PC, Bergus KC, Lutz C, et al. Reducing unnecessary oophorectomies for benign ovarian neoplasms in pediatric patients. JAMA. 2023;330(13):1247-1254. doi:10.1001/jama.2023.17183