In this article, Robert A. Dowling, MD, reviews the challenges of implementing active surveillance in a practice as well as ways to surmount these obstacles.
It has been more than 2 years since the U.S. Preventive Services Task Force released its final grade “D” recommendation concerning PSA-based screening for prostate cancer, and the impact of the recommendation on urologists, patients, and prostate cancer incidence, morbidity, and mortality is still being evaluated (JAMA Intern Med [online] Sept. 1, 2014).
The recommendation against screening, the basic research it drew upon, and the potential for overtreatment of low-risk disease have all served to heighten awareness about the side effects of definitive treatment for prostate cancer and the option of active surveillance.
In a series of articles, I will examine the challenges and the impact of conducting active surveillance in the context of the business of medicine-how a significant shift in approaching and treating patients with prostate cancer might be expected to affect health care operations, reimbursement, or income for the practicing urologist. In this first installment, I will review the challenges of implementing active surveillance in a practice as well as ways to surmount these obstacles.
Assessing the impact of a shifting role for active surveillance in the busy urology practice comes with some challenges. First, there is no easy way to identify how many and which patients in a practice-or in national benchmarks-are “on” active surveillance. No distinct procedural (CPT) or diagnostic (ICD-9 or ICD-10) code systems in common use today incorporate this concept, so these patients cannot be easily identified in billing or clinical data sets. One clinical terminology system-SNOMED CT (Systematized Nomenclature of Medicine-Clinical Terms)-does contain concepts that could be combined to create a meaning of “active surveillance for prostate cancer,” but the terminology is not in widespread use among urologists or their electronic system vendors today.
To face this and other challenges with the many options available to patients with prostate cancer, many urology and oncology practices have employed “patient navigators.” In some cases, the navigators also track patients (manually) on their journey and can report basic metrics on relative frequency of treatments for prostate cancer. For most, though, even establishing a baseline number of patients on active surveillance is a challenge.
A second challenge is a lack of established consensus on candidacy for active surveillance. According to research from the United Kingdom and United States, it can take 10 to 20 years for findings of research to be incorporated into common clinical practice (http://bit.ly/10wL5SQ). The work on identifying ideal candidates is arguably less than 10 years old and is rapidly evolving with the introduction of new genetic and molecular testing. The critical role of pathologists in helping determine candidacy for alternate management approaches has only recently been recognized and exposed another area ripe for consensus making (Arch Pathol Lab Med 2014; 138:1387-405). Perhaps the most relevant clinical guideline endorsed by AHRQ on the role of active surveillance in localized prostate cancer, published in late 2011, simply calls for more research-including on which men are appropriate candidates for surveillance (http://1.usa.gov/1ta7ShO).
A related challenge is that even if a firm consensus existed on candidacy for this management approach-for example, based in part on pathologic characteristics of the biopsy specimen-many clinical systems are not configured to receive, store, or interpret this information in a consistent and computable manner. Many lab interfaces are proficient at passing discrete data like PSA values and not so proficient at digitizing big documents like histology reports. Interfaces can be constructed to onboard information such as Gleason grade, Gleason score, percent core involvement, and number of positive cores, but in this author’s experience they are not in widespread use today.
Next: Clinical decision support tools needed
Web analytics: A primer for the busy urologist
Bundled payments are coming; how to prepare
EHRs: Costly replacement or ‘disruptive innovation’?
To speed the adoption of new management approaches-like active surveillance-based on clinical characteristics of the patient, urologists need clinical decision support (CDS) tools that can gather the information and “assist” clinicians with their tasks. These tools exist, but without the data in a computable format, they remain largely unused. The promise of a popup window that says “Hi, based on this patient’s recent new diagnosis of prostate cancer and assignment of risk based on their PSA, T stage, age, comorbidities, and biopsy characteristics, he may be a candidate for active surveillance” is mostly unrealized today.
A final challenge is that no consensus exists today on what services constitute “active surveillance.” Unanswered questions include: “Does every patient need to have a follow-up biopsy?”; “How and how often should patients be monitored?”; “What are the thresholds that trigger a recommendation to move from active surveillance to traditional treatment?”; and “What is the role of genetic and molecular testing in monitoring patients on active surveillance?”
Centers that have advanced the science in this area have well-established protocols, but community urologists who have begun to embrace active surveillance have not necessarily adopted a standard approach for the reasons mentioned above (delay in translating science to practice). Finally, it will be years before outcomes will be known that could inform the answers to these questions.
How should these challenges be addressed today, and what impact do they have on the office practice?
First, practices should be thoughtful about creating ways to identify and track their patients who are newly diagnosed with prostate cancer. Options include maintaining a list of patients in spreadsheet form (many EHRs allow the creation of a “patient list,” but a simple Excel spreadsheet will also suffice); creating a custom template in your documentation system that includes searchable keywords or metadata; assigning a custom ICD-9 or CPT code in your practice management system that signifies the patient is on active surveillance; or using the “problem list” functionality of your system to create a searchable custom problem-even better, attach a SNOMED CT concept or other clinical term to it. If your EHR doesn’t use SNOMED, you can “borrow” another term you are likely to never use.
Second, until a national guideline dictates otherwise, create your own criteria for identifying candidates and follow those criteria. Resist the temptation to only “copy forward” textual information you may have entered that could determine candidacy and is not computable. When the day arrives that information systems can process the information to help you, you will be prepared.
Encourage your vendor to be proactive about creating CDS tools in preparation for that day. If you already have CDS tools, create a rule and use it. If you haven’t already done so, consider investing in a lab information system interface that stores histology reports in computable formats (ie, discrete data).
Finally, standardize your active surveillance protocol and share it with your staff. Help them understand this is a plan just like a radical prostatectomy, and teach them how to schedule the appropriate tests, follow-ups, and reminders the same way each time. Develop patient handouts, web pages, or other artifacts to support and institutionalize your standard approach. Create an order set in your EHR that is specific to active surveillance (and use it!).
If you modify your approach, update your staff, your materials, and your electronic order sets. This is a significant change in the practice, and it needs to be actively managed.
Bottom line: Active surveillance is a new management paradigm that closely resembles trends in primary care, including “population health management,” “disease management,” and “case management.” All have in common an active approach to monitoring and managing according to evidence-based best practices. When usable and computable clinical guidelines on active surveillance have received national endorsement and are widely incorporated into clinical practice and CDS tools, today’s challenges and uncertainty will perhaps fade. In the meantime, urologists should see active surveillance as an opportunity to standardize approaches in the practice and continue to think about managing populations as well as individual patients.
In a future article, I will examine the possible impact of active surveillance on reimbursement and income in the urology practice.UT
Subscribe to Urology Times to get monthly news from the leading news source for urologists.