Data Mining and Reporting with Smart Infusion Pumps


October 2010 - Vol. 7 No. 10 - Page #2

Angel Medical Center is a 25-bed, nonprofit, critical access hospital located in Franklin, North Carolina. Despite the challenges inherent to our rural setting and critical access designation, Angel Medical Center strives to maintain state-of-the-art facilities, technology, and services for our community, and this is reflected in our administration’s willingness to update our systems and in their encouragement to fully realize the potential of our new technology. We run a centralized pharmacy distribution model, and we are staffed with 2.4 pharmacist FTEs and 4.2 technician FTEs.

We began using our smart pump system in June 2007 after about six months of research and review of the available systems. As part of this process our administration and nursing departments narrowed the choices down to two companies and had them bring their pumps to our facility for a hands-on demonstration with nursing staff. While this demonstration was a positive step, the due diligence portion of the process was somewhat disconcerting, as the pharmacy department was not included. It goes without saying that we should have had a more significant role in the entire selection process, and it would behoove any pharmacy director to insist upon such participation with any automation or technological upgrades that will involve the pharmacy. In this case, there was a perception that this should be a nursing decision, as would have been more accurately the case with standard infusion pumps of the past. Nevertheless, I was introduced to the smart pumps we would be using only after the contract was signed. Fortunately, the system selection proved favorable and after a six-week implementation period, we were up and running with the pumps facility wide.

Establishing a Drug Library
From the onset, our vendor offered the option of using a pre-set, but customizable template versus building our drug library from scratch, and we decided to begin with a template. The template allowed for the establishment of up to six different units with a list of suggested drugs provided for each unit. As a small, critical access hospital we did not need an exhaustive drug library, and the predetermined library satisfied our needs at the time. Regardless of size, each facility should determine well in advance of product selection what their needs and desires are regarding the drug library, and match those with the options available on the various systems. Establishing and vetting a drug library can be a time consuming task, so plan accordingly within your overall project timeframe.

Once the pumps were in-house, representatives from our vendor assisted us with loading the library into the pumps and populating it with our minimums and maximums, as well as standard concentrations, allowing us to customize much of the library, with the exception of the drug names and unit titles. In hindsight, more upfront time spent researching our drug usage history, such as the units and rates most commonly used, would likely have resulted in the establishment of different rates, and more targeted choices of specific drugs for the different nursing units. Thus, this experience proved the old adage that an ounce of prevention is worth a pound of cure. In addition, I would recommend establishing standardized concentrations where possible, prior to smart pump implementation. When establishing a custom drug library, whether it is from a template or from scratch, standardized concentrations allow you to provide fewer choices, thus increasing nursing’s comfort level with the library and further ensuring patient safety. For example, with an epinephrine drip, this choice is only available in our ICU and ER libraries. By narrowing the list of choices, we are limiting opportunities for error. Furthermore, because our nursing staff buys in to the safety concepts behind dose standardization, we can avoid the risk of a nurse creating a new mix based on a possibly incorrect translation of the order.

Information Input, Data Extraction, and Reporting
One of the more prescient developments in smart pump technology over the last couple years that is becoming indispensable, particularly for larger hospitals, is the ability to push drug library changes to all pumps wirelessly. The pump system we have in place is unable to perform wireless library changes, so in our case, a drug representative has to come and manually download the change in the software into each pump. Fortunately, we have not had to perform many manual upgrades. When we do, however, we have to save up our list of changes to make this process as efficient as possible.

While we are limited insofar as wireless drug library changes, we can extract data from the pumps wirelessly, and manipulating and trending this data is where the real benefits to smart pumps lie. We currently have multiple departments taking advantage of this data. For example, the IV technician monitors the real-time status of running infusions at all times, and therefore can predict, based on how much is left in the bags, exactly when a replacement bag will be needed.

As director of pharmacy, I am the primary reviewer of the data. Once a week I review what alarms were triggered, what overrides took place, and the changes that were made in response to the alarms. My goal is to ascertain the issues behind any changes. Was a medication error caught? For example, if a pump went out of range and the nurse went back and changed it to the correct setting, that is documented as catching a near miss. If the nurse bypassed or overrode an alarm (most of our limits are soft stops), yet it was an actual error, this is also documented. In the case of an override wherein there was a legitimate reason for the higher dose, it is a trigger for me to review whether the drug library settings require an alteration. This is a perfect example of smart pump data influencing positive changes, thereby improving the entire process.

In addition to members of the pharmacy department, the nurse managers on the units using the pumps also receive override reports to better monitor the actions of their staff. The biomedical engineering staff has access to the software for scheduling and documenting preventative maintenance. In addition, asset utilization data is sent to the purchasing department. They trend data for usage efficiency—are we approaching maximum usage capacity during particular time periods, and therefore need more pumps? Using this data in many different ways allows us to stay apace of our drug administration demands and avoid down time.

Ultimately, the most significant quality metric we are tracking is staff compliance with the safety features of the pumps. We compare the total number of drugs administered with the number for which the drug library was accessed. We compile reports using this information that are reviewed at safety team meetings, and which ultimately go to our quality assurance and administration departments. By reviewing all medication errors and near misses, and then studying the underlying reasons for both, we can work to refine, and enforce compliance with, policies and procedures, thereby maximizing patient safety. The value of this analysis was evident upon our first drug library update, which resulted in an immediate and significant drop in overrides by nursing staff.

Integrating Bar Code Ordering
To further insure safety, we attached a secondary bar coding system to the pumps themselves that is separate from our regular bar code medication administration system. The IV admixtures that we send to the nurses include an additional bar code that can be scanned straight to the pump, automatically loading the drug name, concentration, and rate. It also verifies that the correct patient is receiving the infusion. One challenge with this secondary system is that it has no direct interaction with our pharmacy information system (PIS). Thus, while it is providing an electronic safety measure, it also requires the nurse to double scan the dose—once into the PIS, and once into the pump. Ideally, to prevent nurses from having to double scan an IV bag, two-way communication would exist between the pumps and the PIS. In addition, it is key to track how often nursing is using the bar code scan to program the pumps, as it can be challenging to ensure all of the nurses are compliant with the scanning approach.

Medication Error Reporting
As is the case with many facilities, our medication error-reporting program is voluntary. It relies on a staff person to self-report errors, and although it is clear there was vast underreporting of errors and near-miss errors in the past from all areas, we did get some reports of IV related errors. Now with the smart pump system in place, by reviewing the reports of alarms, we are assured of 100% reporting of incidences where a nurse has tried to use the pump, an alarm has triggered, and the pump has backed out. Knowing that we receive 100% of that data, as long as the nursing staff has been properly using the pump safety features we have not had any errors. This is the key to such a system. If nursing staff pay attention to the errors, take a moment to determine why an error or alarm was triggered, and take the proper, policy-delineated steps, errors are all but eliminated. This technology is very sound and we have not had a single pump fail from a mechanical standpoint.

Conclusion
As with any newly implemented technology, there are always staff members who struggle to adapt. Over the course of the last couple years, we naturally experienced various setbacks in the proper use of our smart pumps. There are always going to be some staff members that will engage the pumps without using the safety features, perhaps reverting to the ingrained practices they have used for the previous 20 or 30 years. But the fact of the matter is that the majority of our nursing staff appreciates having a modern, sophisticated, and safer pump. Consider your emergency room nurses, for whom time is of the essence. Mistakes are naturally mitigated when any individual’s variables are limited. It is also worth noting that one feature particularly appreciated by nursing is the modular nature of the pumps, allowing multiple pumps to be used for a single patient. This also provides portability, making it easy to transfer one or more pumps to another location.

When combining the front-end capabilities of smart pumps with a robust and multi-dimensional data and reporting system, you have a valuable resource for promoting medication safety. Remember to insist on involving all affected parties—pharmacy, nursing, and administration at the very least—from the beginning when choosing a system. Though the pharmacy department was not part of the selection process, we have been able to bridge the gaps in our system, and have evolved data collection and reporting processes that will continue to grow and support positive advancement specific to both IV medication administration, and to our medication safety processes as a whole.

Debra Cowan, BS Pharm, PharmD, is director of pharmacy at Angel Medical Center in Franklin, North Carolina. Dr. Cowan received her BS in pharmacy from the University of New Mexico and her PharmD from the University of Colorado.

Login

Like what you've read? Please log in or create a free account to enjoy more of what www.pppmag.com has to offer.

WHERE TO FIND: Smart Pumps

Current Issue

Enter our Sweepstakes now for your chance to win the following prizes:

Just answer the following quick question for your chance to win:

To continue, you must either login or register: