Developing and Implementing a Cloud-Based Software Solution for Tracking Ureteral Stents: A Pilot Study

Introduction

Unlike other temporary urologic devices such as indwelling urethral catheters, ureteral stents are internalized. Consequently, urologists who place stents and the patients in whom they are placed can lose track of them. ¹ In fact, as many as 12.5% of stents are maintained too long. ² Retained stents increase a patient's risk for complications, including urinary tract infection and obstruction, jeopardizing the kidney that the stent was intended to protect. ³ Moreover, since additional procedures are often needed for retained stents, their associated costs tend to be greater than those of stents removed on time. ⁴ Taken together, tracking systems that reduce rates of retained stents are expected to improve health and lower treatment expenditures.

However, despite recognition that retained stents are a problem, no solution has emerged as a clear winner. Most practices still rely on paper-based systems for tracking, but they are error prone. To simplify tracking, one stent manufacturer developed a mobile application that generates removal reminders for practices. ⁵ Adoption has been limited because users must perform activities outside of their normal workflows to maintain stent logs. To minimize provider burden, some practices have created automated systems that “scrape” information from the electronic health record (EHR). ⁶ Although these systems have been shown to reduce rates of retained stents, they are not extensible to other EHR platforms.

To address these issues, members of our study team (J.M.H. and M.W.N.) developed the Health Insurance Portability and Accountability Act (HIPAA)-compliant, cloud-based software that automatically registers patients who undergo stent placement. We then piloted the software in our practice to assess its performance. Our software's automation eliminates errors that can result from manual case entry. By leveraging information collected as part of routine patient care, it also minimizes the burden on providers for registry maintenance. Furthermore, our software uses electronic messages that are generated during the execution of administrative and clinical tasks, which are standardized across EHRs, making it potentially compatible with all major platforms.

Materials and Methods

Software solution

Our tracking software is based on a set of international standards for the transfer of administrative and clinical data between EHR applications, known as Health Level-7 (HL7). Every time an event occurs and data need to flow between two applications (e.g., between a hospital's health information and inventory management systems), small text-based HL7 messages about the event are generated and sent in the EHR. Using an integration engine as an interpreter, our software “listens” for these HL7 messages and pulls out information from them to identify patients undergoing stent placement. Our software then puts these patients into a registry and follows them through their subsequent health care encounters.

The stent registry can be viewed through a secure web interface that is accessible by providers or designated clinic staff (Fig. 1). For each patient, a unique profile with name and medical record number is created. In addition, the date of stenting and the provider who placed the stent are recorded by our software. Patients are grouped (by our software) into one of four mutually exclusive categories: (1) in need of an appointment, (2) with a scheduled appointment, (3) attended appointment, and (4) stent removed at appointment. As appointments are made, attended, canceled, or rescheduled, patients move between the first three categories automatically. Visual coding allows for the identification of patients who have exceeded the intended stent dwell time. For the software version tested in this pilot, manual reconciliation was required of the provider (or a designee) to move a patient from category #3 to #4.

OPEN IN VIEWER

FIG. 1.

User interface.

Results

Table 1 displays patient- and procedure-specific data for our study population. During our 1-month pilot study, our manual review of the EHR revealed that 51 ureteral stents (including those with and without a tether) were placed during 49 procedures performed on 46 patients. Fifty-two percent of these patients were women. The average patient age at the time of surgery was 54.7 (standard deviation, 17.1) years. Seventy-eight percent of procedures occurred at our university hospital. The overwhelming majority were unilateral (96%). Ureteroscopy with laser lithotripsy accounted for most stent-related procedures (65%). A tether was used in a minority of stents placed (39%).

Figure 2 highlights our software's performance. It effectively captured all 49 procedures during which a stent was placed. Our software properly distinguished 31 procedures where the patient who underwent stenting had follow-up scheduled before surgery (either in clinic or for a planned second-look procedure in the operating room). More importantly, it alerted our schedulers to 18 procedures for which no return visit had been scheduled. Furthermore, the software was able ascertain follow-up attendance, correctly identifying 10 procedures where patients did not arrive for follow-up appointments.

OPEN IN VIEWER

FIG. 2.

Results of pilot study.

Discussion

We developed, implemented, and validated a novel cloud-based, HIPAA-compliant software solution for tracking ureteral stents. Our software effectively captured all stent placements during a 1-month pilot. More importantly, it identified patients lacking follow-up appointments after surgery, as well as those who failed to attend their postoperative visits. Had these patients not been identified, their stents may have been forgotten. Advantages of our software include its automated data capture, agnosticism to device manufacturer, and seamless EHR integration.

Electronic systems to track ureteral stents are well described and have been shown to lower rates of retained devices. ⁷ However, their widespread adoption has been limited. Many required manual data entry for maintenance and periodic review to ensure device reconciliation. Lynch and colleagues described a system that they integrated into their hospital's EHR, which produced automated reminders; however, stent placements still had to be logged with a barcode scanner by staff. ⁸ Ziemba and colleagues published their experience with Ureteral Stent Tracker^TM (Visible Health, Inc., Boston Scientific)—a software application that can be downloaded onto a smartphone. ⁹ Despite the ease of this mobile application, it functions outside the EHR, requiring urologists to use separate systems for tracking and reconciliation. More recently, the Yale group reported on a stent log that they created within their instance of Epic with automated data capture on device placement. ¹⁰ While this program could potentially be ported to other Epic sites, it is not extensible to other platforms.

Our study must be considered in the context of some limitations. First, although our software solution currently generates provider alerts when patients are not scheduled for follow-up appointments or fail to arrive for their appointments, we are yet to incorporate secure patient messaging. Use of short message service texts or emails could help reinforce timely stent removal. Second, at present, our software can only read the EHR, but not write data to it. While we are actively working to achieve two-integration, a provider (or his/her designee) still has to log into our application outside of the EHR to confirm ureteral stent removal. Third, our software was built on the common backbone of all EHR platforms (HL7) to allow for integration at any hospital, but future pilots at non-Epic sites are needed. Finally, our study was designed to assess the feasibility of developing and implementing a software solution for automated ureteral stent tracking. While it provides compelling pilot data, a longer and larger one, which potentially involves multiple sites, is still needed to evaluate whether deployment of our software significantly reduces the number of retained ureteral stents.

Notwithstanding these limitations, our software solution has important clinical implications. Retained ureteral stents can have potentially catastrophic downstream consequences for patients. Severely encrusted stents are difficult to remove and sometimes require complex intervention, leading to considerable morbidity and added financial costs. They also have substantial medico-legal ramifications for providers, as forgotten ureteral stents represent a large proportion of malpractice claims filed against urologists for nononcologic care. ^11,12 Despite the success of electronic tracking systems at reducing rates of retained stents, their use has been limited. Our solution has several advantages that may help with its dissemination. Namely, it offers automated, high-fidelity device tracking with alerts. It does not require a dedicated health care provider for maintenance. It is agnostic to device manufacturer and can be seamlessly integrated into the EHR.

Conclusions

We developed and effectively piloted a high-fidelity software solution for automated tracking of ureteral stents that is agnostic to the device manufacturer and can be seamlessly integrated into the EHR, causing minimal disruption to provider workflows. In addition to tracking ureteral stents, this solution also has potential application for other temporary implantable medical devices (e.g., inferior vena cava filters, intrauterine devices), for which timely retrieval or exchange is also important.