FROM JAMA SURGERY

Rates of postoperative hypocalcemia and recurrent laryngeal nerve (RLN) injury after thyroidectomy varied significantly from hospital to hospital in a recent large study of U.S. hospitals, suggesting to the authors that these measures could be used for quality improvement metrics.

In the study, published Nov. 29 in JAMA Surgery, hospitals with significantly lower rates of hypocalcemia were more likely to conduct postoperative parathyroid hormone level measurement as well as to prescribe vitamin D, calcium supplements, or both. Hospitals with lower RLN injury rates more frequently used energy devices and intraoperative nerve monitoring.

“Causation cannot be proven by this, but the confidence that these practice parameters are important is high,” senior author Bruce Hall, MD, PhD , vice president and chief quality officer at BJC Healthcare, and professor of surgery at Washington University, St. Louis, said in an interview. Dr. Hall is consulting director for the American College of Surgeon’s National Surgical Quality Improvement Program (NSQIP), which provided data for the analysis ( JAMA Surg. 2017 Nov 29. doi: 10.1001/jamasurg.2017.4593 ).

The work highlights the potential of hypocalcemia and RLN. “They could be used internally by hospitals for quality improvement now. However, the variables that we used in this study should now be refined,” coauthor Jason Liu, MD , said in an interview. Further variables might include, for example, whether the vocal cords were examined preoperatively or the date and time that parathyroid hormone level was checked. “NSQIP data are also available to participants for research purposes as Participant User Files, so investigators are encouraged to conduct their own studies,” added Dr. Liu, a clinical scholar at the ACS.

The researchers examined data from 14,540 patients who underwent thyroidectomies at 98 hospitals between Jan. 1, 2013, and Dec. 31, 2015. These included 13,242 operations at 96 hospitals with complete hypocalcemia data, 13,144 operations at 95 hospitals with complete RLN data, and 13,197 operations at 95 hospitals with complete hematoma data. The primary outcome was the 30-day incidence of hypocalcemia, RLN, and hematoma. The researchers also measured 30-day mortality, surgical site infections, and hospital readmissions.

A total of 3.3% of patients experienced clinically severe hypocalcemia (0.6% after partial thyroidectomy, 4.7% after total or subtotal thyroidectomy). Another 5.7% experienced RLN (4.2% after partial, 6.6% after total or subtotal). Hematoma occurred in 1.3% of cases, but there were no significant variations in rates of hematoma across participating institutions.

For hypocalcemia and RLN injury, there were hospital outliers both on the low end of complication rates and on the high end of complication rates, defined by odds ratios with 95% confidence ratios that were greater than 1 for high outliers, or lower than 1 for low outliers. There were no outliers with respect to hematoma, suggesting that it may not be a useful barometer of hospital performance.

With respect to hypocalcemia rates, four hospitals were low outliers, and seven were high. Eight hospitals were low outliers with respect to RLN injury, and 14 were high outliers.

In the analysis of postoperative hypocalcemia, both low and high outliers measured postoperative calcium with similar frequency (68.4% vs. 71.0%; P =.09). However, high performance outliers were more likely to prescribe postoperative calcium, vitamin D, or both (76.6% vs. 66.8%; P less than .001).

Among RLN outliers, intraoperative nerve monitoring was more common in the top performing hospitals (55.7% vs. 37.7%; P less than .001), as was the use of energy devices (69.1% vs. 55.2%; P less than .001).

There was one high outlier when it came to surgical site infections, and one high and one low outlier with respect to morbidity outcomes. There were no hospital readmission outliers.

No source of funding was disclosed. Dr. Liu and Dr. Hall reported having no financial disclosures.

acssurgerynews@frontlinemedcom.com

Ads