ECRI Institute commends Memorial Sloan Kettering Cancer Center (New York, NY) for its project addressing the challenges associated with converting from a homegrown equipment management database to a commercial system. Memorial Sloan Kettering described the project in its submission for ECRI Institute's 13th Health Devices Achievement Award competition. Key components of this process included cleaning the data, maintaining data integrity in the new system, and centralizing the organization's recall management efforts.

The Health Devices Achievement Award recognizes outstanding initiatives undertaken by member healthcare institutions to improve patient safety, reduce costs, or otherwise facilitate better strategic management of health technology. For details about the winning submission and other finalists, see The Health Devices Achievement Award: Recognizing Exceptional Health Technology Management.

ECRI Institute congratulates Maryam Mehryar, Paul Frisch, PhD, and the rest of the Memorial Sloan Kettering team.
 

The Challenge

To transition the organization's medical device inventory from a homegrown equipment management database to a commercial computerized maintenance management system (CMMS)—with a particular goal of improving the organization's management of recalls and other medical device safety alerts.
 

The Context

The biomedical engineering team at Memorial Sloan Kettering had for decades used a homegrown database to manage its medical device inventory and service history. The flexibility afforded by the database had served the organization well, but its limitations eventually made the system untenable. The team identified two particular shortcomings that necessitated a change:

1. The database did not support the entry of data in a standardized format, leading to inconsistent entries. The database allowed medical device data to be entered in free-text fields; thus, users were able to enter any variation of a manufacturer or model name when adding a new medical device to the inventory. The use of different terms to describe the same device made it difficult for users to reliably retrieve all of the records related to any specific model. Efforts to clean (i.e., standardize) the data were made from time to time, but ultimately proved unsustainable. Because the database did not force the use of a standardized nomenclature, wording variations continued to proliferate.

2. The database did not support a centralized recall management process. The lack of data standardization, as noted above, was one concern. Another was that information was scattered among several stand-alone databases, folders, and files. As a result, users found it challenging to quickly and reliably identify whether the subject of a device recall, hazard report, or other safety notification—collectively referred to as "alerts"—matched a piece of equipment in the facility's inventory.

The team determined that it was time to make the leap from the homegrown equipment management database to a commercial CMMS.
 

The Process

The biomedical engineering team recognized that simply overlaying a commercial CMMS on their existing data and processes would be ineffective. They needed to invest the time and resources to select a system that met their needs and to implement tools and processes that would fix the underlying issues. The project focused on two key aspects:

1. Cleaning their data and implementing controls to maintain data integrity. Steps included:

a) Implementing a standardized, controlled medical device nomenclature

b) Cleaning existing data by matching device categories and manufacturer names to terms present in the standardized nomenclature, as well as by standardizing the terms used for model names

c) Establishing data-entry controls, so that entries for new equipment conform with the standardized terminology. New categories of devices now must be added by an administrator, using the approved nomenclature. This information can no longer be entered by general users or by using free-text fields.

2. Establishing a centralized recall management process. With access to clean data, users could more quickly and reliably determine whether a device identified in a particular alert was present in the facility's inventory. Additionally, with a new CMMS that could interface with other commercially available systems, the organization was now positioned to take advantage of workflow-improvement features offered by commercial alert management systems. The team decided to implement such a system.

Memorial Sloan Kettering reports that it used several ECRI Institute tools and services when making this transition. Specifically, the organization implemented the Universal Medical Device Nomenclature System (UMDNS) for its medical device nomenclature and Alerts Tracker with Automatch for its medical device alert management system. In addition, the organization used ECRI Institute's data cleaning service to convert its data to the UMDNS nomenclature. 
 

The Results

Though data cleaning was a daunting task, the team recognized the criticality of both normalizing existing data and creating an organizational plan to maintain data integrity. The value of data cleaning was perhaps most evident in the improvements in alert management that it helped facilitate. The use of a standardized naming convention across the entire system better enabled staff at all facilities to comprehend and act upon safety notices for specific types of devices.

The alert management process was further enhanced by the workflow improvements possible with the newly implemented alert management system, along with that system's ability to interface with the organization's new CMMS. The new systems automated the creation, assignment, and distribution of work orders from recall notifications—dramatically decreasing alert response time—and facilitated documentation of the process from start to finish.

Additionally, an inventory matching feature available with the alerts management system helped the organization clear out a large backlog of irrelevant alerts. This effort freed staff to focus on addressing current, actionable alerts.
 

Key Takeaways

The biomedical engineering team recognized that simply "throwing new technology" at the problem would be ineffective. Rather, they needed to invest the time and resources to select the right system and to implement tools and processes that would fix the underlying issues.

Cleaning their data, in particular, was an intimidating step—but a necessary one for achieving the benefits the organization desired. Data integrity is now recognized to be a central part of staff's job responsibilities.

Any organization making the leap from a homegrown database to a commercial CMMS will likely face challenges such as those tackled by the Memorial Sloan Kettering team. ECRI Institute takes pride in offering tools and services that helped the organization meet this challenge.