Are you getting the most out of your patient data?

Last fall, I discussed in these pages the current reality of health care analytics and the promise of data to transform patient and population care in the future (“Data-driven care: Who it will benefit, and how,” September 2012, page 46). While meaningful data analysis for urologists is perhaps in its infancy, practices have been entering “data” into their information systems for years and are anxious to use this information.

One area of health care analytics that has practical application today is the accurate identification of certain patients within your practice, based on clinical characteristics and for a defined purpose. In this article, I describe three “use cases” for this simple form of descriptive analytics.

Clinical research: Your EHR can help

Clinical research involves many steps and demands many skills: study design, study approval, sponsorship and funding, subject recruitment, and project management, to name a few. Many urologists who have participated in clinical research activity understand that perhaps the biggest challenge-and determinant of success-is patient enrollment. Finding patients who meet inclusion and exclusion criteria may not be a priority in the clinician’s mind, and patient recruitment that relies on memory, sticky notes, e-mails, and other spontaneous processes is likely to fail.

Study criteria can often be translated into simple query statements against practice management data repositories to identify patients by age, race, ethnicity, gender, diagnosis, procedure, insurance, and other fields. Query language includes the ability to construct elaborate “and,” “or,” and “not” conditions to further refine searches.

The addition of clinical data from the electronic health record (EHR)-such as past medical history, past surgical history, family history, prostate symptom scores, prescriptions, refills, allergies, and other data not represented in the practice management system-can enrich the value and the accuracy of a search for potential study candidates. Systematically mining data from these sources can help the urologist decide whether to participate in a study by determining feasibility of recruitment, or assist in conducting an ongoing study.

Finally, EHRs may be “interfaced” with clinical trial software to both track patient events and keep the clinical research activity “in sync” with the non-research patient care activity.

Patient recall and tracking

The same tools and processes used to mine information for clinical trials can be applied to identify or track other types of patients who might otherwise go overlooked in the busy urology practice. For example, as new therapies for chronic disease like voiding dysfunction, urologic cancers, metabolic stone disease, and benign prostate disease become available, data-driven patient identification could be used to target patients for consultation or education.

Another example concerns the patient with the short- or long-term indwelling ureteral stent. Poor compliance with follow-up can represent a major liability for the patient (health risk) and the urologist (malpractice risk). Many urologists have developed manual processes to track these patients, and manual processes may be flawed in design or in execution (human error). Systematically identifying and tracking these patients with simple data mining tools is a safer and more reliable method. Analytics tools can also be used to create reports or export to simple patient registries and track patients with cancers, high-risk procedures, drug therapies, and other areas of interest.

In addition, analytics can be employed to identify patients who may have been prescribed certain drugs in the event of a recall notice, new adverse event, or other significant change in labeling that requires action by the physician. Most EHRs allow the entry of discrete lot numbers and other specific information that would assist such a use case, and urologists are advised to enter that information when possible.

Marketing: Internal and external

Business intelligence tools can easily be applied to internal and external marketing purposes in the urology practice. For example, a practice may wish to conduct an analysis of referring providers by geography, volume, patient types, insurance plans, and even clinical characteristics (chronic medical conditions) to determine referral value, consider opening an outreach clinic, evaluate new physician opportunity, determine the potential impact of a change in referral patterns, or take some other action.

Analytics might drive the appropriate targeting of patients for a newsletter, seminar, e-mail notification, patient portal push, or other purpose. For example, a query might be designed to find patients who are not deceased, have organic erectile dysfunction and Peyronie’s disease, have been prescribed phosphodiesterase type-5 inhibitors and penile injection therapy, and have not had corrective surgery. A mailing list thus generated and regularly updated could be the foundation for a quarterly seminar on male health.

Bottom line: EHR and practice management data is ready to be put to use. Despite the poor penetration of advanced business intelligence into physician practices in general and urology in particular, simple queries and tools using off-the-shelf analytics can be used to solve many practical problems today. Problems with data quality exist, but physicians are generally entering information about valuable clinical attributes like medical conditions, surgical conditions, and prescriptions if they have implemented an EHR.

Keep in mind the axiom, “garbage in, garbage out” when it comes to data entry. In other words, enter good data now and be rewarded with good information later.UT

Subscribe to Urology Times to get monthly news from the leading news source for urologists.