Determinants of Risk-Aligned Bladder Cancer Surveillance-Mixed-Methods Evaluation Using the Tailored Implementation for Chronic Diseases Framework

Accurate Documentation Contributes to Guideline-concordant Surveillance of Nonmuscle Invasive Bladder Cancer: A Multisite Department of Veterans Affairs Study

OBJECTIVE

To determine if accurate documentation of bladder cancer risk was associated with a clinician surveillance recommendation that is concordant with AUA guidelines among patients with nonmuscle invasive bladder cancer (NMIBC).

METHODS

We prospectively collected data from cystoscopy encounter notes from four Department of Veterans Affairs (VA) sites to ascertain whether they included accurate documentation of bladder cancer risk and a recommendation for a guideline-concordant surveillance interval. Accurate documentation was a clinician-recorded risk classification matching a gold standard assigned by the research team. Clinician recommendations were guideline-concordant if the clinician recorded a surveillance interval that was in line with the AUA guideline.

RESULTS

Among 296 encounters, 75 were for low-, 98 for intermediate-, and 123 for high-risk NMIBC. 52% of encounters had accurate documentation of NMIBC risk. Accurate documentation of risk was less common among encounters for low-risk bladder cancer (36% vs 52% for intermediate- and 62% for high-risk, P < .05). Guideline-concordant surveillance recommendations were also less common in patients with low-risk bladder cancer (67% vs 89% for intermediate- and 94% for high-risk, P < .05). Accurate documentation was associated with a 29% and 15% increase in guideline-concordant surveillance recommendations for low- and intermediate-risk disease, respectively (P < .05).

CONCLUSION

Accurate risk documentation was associated with more guideline-concordant surveillance recommendations among low- and intermediate-risk patients. Implementation strategies facilitating assessment and documentation of risk may be useful to reduce overuse of surveillance in this group and to prevent unnecessary cost, anxiety, and procedural harms.

Examining longitudinal markers of bladder cancer recurrence through a semiautonomous machine learning system for quantifying specimen atypia from urine cytology

BACKGROUND

Urine cytology is generally considered the primary approach for screening for recurrence of bladder cancer. However, it is currently unclear how best to use cytological examinations for assessment and early detection of recurrence, beyond identifying a positive finding that requires more invasive methods to confirm recurrence and decide on therapeutic options. Because screening programs are frequent, and can be burdensome, finding quantitative means to reduce this burden for patients, cytopathologists, and urologists is an important endeavor and can improve both the efficiency and reliability of findings. Additionally, identifying ways to risk-stratify patients is crucial for improving quality of life while reducing the risk of future recurrence or progression of the cancer.

METHODS

In this study, a computational machine learning tool, AutoParis-X, was leveraged to extract imaging features from urine cytology examinations longitudinally to study the predictive potential of urine cytology for assessing recurrence risk. This study examined how the significance of imaging predictors changes over time before and after surgery to determine which predictors and time periods are most relevant for assessing recurrence risk.

RESULTS

Results indicate that imaging predictors extracted using AutoParis-X can predict recurrence as well or better than traditional cytological/histological assessments alone and that the predictiveness of these features is variable across time, with key differences in overall specimen atypia identified immediately before tumor recurrence.

CONCLUSIONS

Further research will clarify how computational methods can be effectively used in high-volume screening programs to improve recurrence detection and complement traditional modes of assessment.

Data-driven approach to implementation mapping for the selection of implementation strategies: a case example for risk-aligned bladder cancer surveillance

BACKGROUND

Implementation Mapping is an organized method to select implementation strategies. However, there are 73 Expert Recommendations for Implementing Change (ERIC) strategies. Thus, it is difficult for implementation scientists to map all potential strategies to the determinants of their chosen implementation science framework. Prior work using Implementation Mapping employed advisory panels to select implementation strategies. This article presents a data-driven approach to implementation mapping, in which we systematically evaluated all 73 ERIC strategies using the Tailored Implementation for Chronic Diseases (TICD) framework. We illustrate our approach using implementation of risk-aligned bladder cancer surveillance as a case example.

METHODS

We developed objectives based on previously collected qualitative data organized by TICD determinants, i.e., what needs to be changed to achieve more risk-aligned surveillance. Next, we evaluated all 73 ERIC strategies, excluding those that were not applicable to our clinical setting. The remaining strategies were mapped to the objectives using data visualization techniques to make sense of the large matrices. Finally, we selected strategies with high impact, based on (1) broad scope, defined as a strategy addressing more than the median number of objectives, (2) requiring low or moderate time commitment from clinical teams, and (3) evidence of effectiveness from the literature.

RESULTS

We identified 63 unique objectives. Of the 73 ERIC strategies, 45 were excluded because they were not applicable to our clinical setting (e.g., not feasible within the confines of the setting, not appropriate for the context). Thus, 28 ERIC strategies were mapped to the 63 objectives. Strategies addressed 0 to 26 objectives (median 10.5). Of the 28 ERIC strategies, 10 required low and 8 moderate time commitments from clinical teams. We selected 9 strategies based on high impact, each with a clearly documented rationale for selection.

CONCLUSIONS

We enhanced Implementation Mapping via a data-driven approach to the selection of implementation strategies. Our approach provides a practical method for other implementation scientists to use when selecting implementation strategies and has the advantage of favoring data-driven strategy selection over expert opinion.