Development of a Natural Language Processing Engine to Generate Bladder Cancer Pathology Data for Health Services Research.

OBJECTIVE: To take the first step toward assembling population-based cohorts of patients with bladder cancer with longitudinal pathology data, we developed and validated a natural language processing (NLP) engine that abstracts pathology data from full-text pathology reports. METHODS: Using 600 bladder pathology reports randomly selected from the Department of Veterans Affairs, we developed and validated an NLP engine to abstract data on histology, invasion (presence vs absence and depth), grade, the presence of muscularis propria, and the presence of carcinoma in situ. Our gold standard was based on an independent review of reports by 2 urologists, followed by adjudication. We assessed the NLP performance by calculating the accuracy, the positive predictive value, and the sensitivity. We subsequently applied the NLP engine to pathology reports from 10,725 patients with bladder cancer. RESULTS: When comparing the NLP output to the gold standard, NLP achieved the highest accuracy (0.98) for the presence vs the absence of carcinoma in situ. Accuracy for histology, invasion (presence vs absence), grade, and the presence of muscularis propria ranged from 0.83 to 0.96. The most challenging variable was depth of invasion (accuracy 0.68), with an acceptable positive predictive value for lamina propria (0.82) and for muscularis propria (0.87) invasion. The validated engine was capable of abstracting pathologic characteristics for 99% of the patients with bladder cancer. CONCLUSION: NLP had high accuracy for 5 of 6 variables and abstracted data for the vast majority of the patients. This now allows for the assembly of population-based cohorts with longitudinal pathology data.

Early Stage Bladder Cancer: Do Pathology Reports Tell Us What We Need to Know?

OBJECTIVE: To assess a large national sample of bladder cancer pathology reports to determine if they contained the components necessary for clinical decision-making. METHODS: We examined a random sample of 507 bladder cancer pathology reports from the national Department of Veterans Affairs Corporate Data Warehouse to assess whether each included information on the 4 report components explicitly recommended by the College of American Pathologists' protocol for the examination of such specimens: histology, grade, presence vs absence of muscularis propria in the specimen, and microscopic extent. We then assessed variation in the proportion of reports lacking at least 1 component across Department of Veterans Affairs facilities. RESULTS: One hundred eight of 507 reports (21%) lacked at least 1 of the 4 components, with microscopic extent and presence vs absence of muscularis propria in the specimen most commonly missing (each in 11% of reports). There was wide variation across facilities in the proportion of reports lacking at least 1 component, ranging from 0% to 80%. CONCLUSION: One-fifth of bladder cancer pathology reports lack information needed for clinical decision-making. The wide variation in incomplete report rates across facilities implies that some facilities already have implemented best practices assuring complete reporting whereas others have room for improvement. Future work to better understand barriers and facilitators of complete reporting may lead to interventions that improve bladder cancer care.

Keratinized squamous metaplasia of the upper urinary tract resulting in recurrent renal colic.

A 60-year-old female with an extensive history of stone disease and shock wave lithotripsy presents with recurrent and increasingly severe renal colic. Work-up reveals obstruction with translucent debris that is found to be composed of keratin. Her history of chronic irritation of the collecting system has resulted in keratinizing squamous metaplasia (KSM) with hyperkeratosis that has sloughed from the upper urinary tract and has become lodged in the ureter. Because of the worsening of her symptoms on conservative management, the patient elected for a nephrectomy and her symptoms have since resolved. KSM of the renal pelvis is a relatively rare phenomenon and most often presents with irritative symptoms. It is thought to result from chronic irritation of the urothelium. KSM has been found to be coincident with squamous cell cancers of the urinary tract, though clear data implicating KSM as a premalignant lesion is lacking. We present a case of recurrent renal colic secondary to sloughing keratin debris from KSM.

Rate of dislodgment of ureteral stents when using an extraction string after endoscopic urological surgery.

PURPOSE: Ureteral stents are manufactured with an extraction string tethered to the distal end, which facilitates removal after urological surgery. However, the string may allow for stent dislodgment. We report the stent dislodgment rate in a multi-institutional series. MATERIALS AND METHODS: We retrospectively reviewed the records of ureteroscopy cases at 3 academic endourology practices. Demographic and operative data were obtained as well as string use and stent dislodgment data. Categorical variables were assessed with the Fisher exact test and the Student t-test was used to assess continuous variables. RESULTS: Of the 512 cases a string was used in 98 (19.1%) comprising 41 females (41.8%) and 57 males (58.2%). The stent was dislodged in 10 women and 3 men. No dislodgment occurred when a string was not used. When stratified by gender, 5.3% of men and 24.4% of women with a string experienced stent dislodgment (p = 0.013). Women were more than fourfold more likely to experience stent dislodgment than men (RR 4.6, 95% CI 1.36-15.8, p = 0.01). CONCLUSIONS: Almost 15% of patients in whom a stent is placed with a string sustain stent dislodgment and most of these patients will be women. We recommend considering the risks of dislodgment in each patient who undergoes ureteroscopy with stent placement and considering string removal if the surgeon believes that dislodgment could result in adverse events such as severe colic or obstruction.

Global robotic experience and the type of surgical system impact the types of robotic malfunctions and their clinical consequences: an FDA MAUDE review.

OBJECTIVES: To assess annual rates of robotic system malfunctions and compare the da Vinci S(®) system (dVS) and da Vinci(®) surgical system (dV). To assess the types of malfunctions and associated outcomes for robotic cases and determine the extent to which experience and technological improvements impact these. PATIENTS AND METHODS: This study is a retrospective review of the US Food and Drug Administration (FDA) MAUDE (Manufacturer and User Facility Device Experience) database, a publicly available, voluntary reporting system (http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfmaude/search.cfm). The database was searched using the two phrases 'da Vinci' and 'Intuitive Surgical' from 2003 to 2009. Malfunctions of the instruments, console, patient-side cart, camera and cannula were recorded. Data on intraoperative injuries, case delays and conversions were also collected. RESULTS: In all, 1914 reports were reviewed (991 dVS and 878 dV, 45 unclassified) with peak years for reports of 2008 for dVS (571) and 2007 for dV (211), P < 0.001. With respect to time, the proportion of console and patient-side cart malfunctions declined from 2007 onward compared with the proportions prior to 2007 (5.1% vs 9.4% and 6.6% vs 10.9%). Patient injury did not change with year of surgery (0.5-5.4% of malfunctions, P= 0.358), open conversions declined (21.3% of malfunctions before 2007 vs 9.9% from 2007 onward, P < 0.001) and patient deaths increased (0.0013% of cases before 2007 vs 0.0061% of cases from 2007 onward, P < 0.001). With regard to robotic system, console and patient-side cart malfunctions were more frequent with the dV than the dVS: 82/878 vs 39/991 and 100/878 vs 48/991, P < 0.001. Open conversion was more frequent with dV than dVS (19.3% vs 7.7% of reported malfunctions, P < 0.001), while patient injury was less with dV than dVS (3.5% vs 5.9%, P= 0.021). CONCLUSIONS: The dVS decreased console and patient-side cart errors relative to total malfunctions, which were also influenced by surgical year. Open conversions were reduced by increased robotic experience and newer surgical system. Differences in patient injury may reflect changes in reporting or case complexity.