Editorial Comment on: Device-Related Adverse Events During Percutaneous Nephrolithotomy: Review of the Manufacturer and User Facility Device Experience Database by Patel et al . (From: Patel NH,Schulman AA,Bloom JB,et al. J Endourol 2017;31:1007–1011)

The authors retrospectively reviewed the United States Food and Drug Administration's (FDA) Manufacturer and User Facility Device Experience (MAUDE) database for reported complications relating to devices used during percutaneous nephrolithotomy (PCNL) over a decade. ¹ There was a total of 218 incidents, mostly relating to lithotripters (24.3%), wires (19.7%), balloon dilators (13.8%), and occlusion balloons (12.8%). Patient harm resulting from these incidents included prolongation of procedures 4 (9.5%), retained stone fragments 7 (16.7%), subsequent endoscopic procedures (28.6%), bleeding (19.0%), ureteral injury (4.8%), and conversion to an open procedure (7.1%). Surprisingly, there were few or no incidents relating to nephroscopes (either rigid or flexible), stone baskets, or laser fibers. Using Gupta's classification system, most (81%) of the adverse events were mild or Level I, whereas 12% were moderate or Level II, 7% were severe or Level III, and there was even one case (0.5%) of life threatening or Level IV complication. ² Although most reported incidents were related to the use of lithotripters, incidents relating to balloon dilators were associated with higher risk of Levels II–IV complications when compared with other devices (odds ratio: 4.33, confidence interval: 1.978–9.493, p < 0.001). However, there was no one particular brand name that was associated with adverse events. When manufacturers of devices analyzed reported incidents, they found that the majority (54.8%) of reviewed incidents were related to misuse by the operator.

Although the MAUDE database has been used previously to assess reported adverse events relating to devices used in endourologic procedures, it has several shortcomings that need to be emphasized. ^3,4 First, incidents are under-reported. Second, the denominator or the exact number of PCNLs performed in the United Stated during the same period is unknown. According to one estimate, the overall annualized incidence of PCNLs performed in the United States was 5.34/100,000 people in 2014. ⁵ Since the U.S. population was 318.6 million in 2014, it is estimated that there were 17,013 PCNLs performed in the United States in 2014. This means that over a decade, about 170,130 PCNLs were performed. Therefore, the incidence of adverse events relating to devices used in PCNL is 218/170,130 = 0.13% or 1.3 adverse events per 1000 PCNLs performed. Therefore, the rate of adverse events relating to devices is rare.

Nevertheless, how can we improve patient safety?

I think this could be addressed at three levels. First, companies manufacturing these approved devices need to ensure adequate supply of these products, avoiding backorders and automatic replacements with other products that urologists may not be familiar with. Second, urologists using these devices need to be familiar with their use since the majority of the adverse events relating to devices used in this study were related to misuse by the surgeons. Training could take place on simulators and in wet laboratories. Finally, surgeons need to improve reporting of adverse events relating to devices so that the true incidence of adverse events relating to these devices is known.