How do you improve patient safety worldwide? With visual intelligence and zero-miss barcode reading automation.
One of the fundamental goals in healthcare is securing patient safety. Patient Safety can include anything from the transfer of information to administering the right medication. With so many factors blindly contributing to the daily battle, it is comforting to know that ZiuZ Medical has placed the identification and validation of automatically dispensed medication firmly in the completed column worldwide.
ZiuZ Medical is a visual intelligence specialist, based in the Netherlands. ZiuZ helps pharmacies and hospitals make their automatic dose dispensing systems safer. Once a pill leaves a bottle, it also loses its identity and from one staff member to another the ability to remember exactly what it is–especially when many pills can look so similar.
Automatic dose dispensing came to the rescue with the ability to individually package pills together so that only the medicines intended for a specific patient are in fact delivered to that patient. But what happens if during that automation process the wrong pills get placed into the package? By using 13 parameters, near-infrared (NIR) image analysis technologies, and intelligent algorithms, ZiuZ is able to identify each pill or capsule within each bag. This automation process moves at a rate of 6 inches per second, giving each medication pack an approval or rejection, and ensuring only the correct medication is placed in each bag.
Getting a Grip on Quality
As you can probably imagine, one of the most crucial elements for patient safety would be the dispensing of the right drugs each and every time no matter the workflow. Each patient’s needs are unique, making the treatment process varied and often complicated. Ensuring that each patient gets the correct dosage of the correct drug is vital, and errors can be fatal.
In a study titled, “Medication errors: the importance of safe dispensing”, published by the British Journal of Clinical Pharmacology, states that the most common causes of dispensing errors were: being busy, being short-staffed, being subject to time-constraints, fatigue of healthcare providers, interruptions during dispensing, and look-alike/sound-alike medicines (1).
It is clear that automatic dose dispensing and dispensing machines have been recommended as one potential mechanism to improve efficiency and patient safety, and they are now widely used in many hospitals (2). There is no doubt that these machines can enhance the efficiency of medication distribution, but their capacity to reduce medication errors is controversial and depends on many factors. ZiuZ Medical created a visual intelligent solution incorporating the industries best image-based barcode scanning software from TachyonIQ™. A patient’s life can depend on having a solution that automatically verifies the contents of each pill pack, ensuring that they contain the correct medication.
Testing, Testing, 1 2 3
The task at hand was to find a way not only to identify each individual pill within every given medication pack assembled but to also find a way to compare that information with what should be located inside each package pointing out errors. This process would need to include difficult medicines and capsules, and situations with broken pills or crumbs.
The ZiuZ intelligent system started with a detailed photograph of the transparent side of each bag to show the contents of the pouch, but it needed the ability to reliably scan and decode the 1D and 2D barcodes printed on the back of each pack containing the details of what items should be found inside. This part played a crucial role. Without it, their technology would not be able to then perform the verification/comparison to deliver a match or error result. The need to find an enterprise-grade, high-performing scanning solution proved a challenge – that is until they discovered TachyonIQ’s CortexDecoder.
In healthcare, like in a lot of industries, time is a valuable resource. However, unique to healthcare life and death decision are made in seconds. The right scanning solution would make checking packs of medication fast and accurate while reinforcing a superior product reputation. ZiuZ turned to TachyonIQ’s barcode scanning software, CortexDecoder. Integrating CortexDecoder into the med pack checking device has delivered a scanning experience that exceeded all expectations.
Prior to finding TachyonIQ, ZiuZ Medical was using a decoder that had an error rate of .05%, this meant 1 in every 200 barcodes was misread, and the wrong medication could slip by. Finding this unacceptable, it led to the search for a decoding algorithm that would be able to keep up. The decoding software that ZiuZ needed had to deliver on three essential things:
- The error read rate still needed to be lower.
- It needed to be able to read 2D barcodes moving at a speed of about 6 inches per second
- It had to flawlessly decode from the reflective surface of the shiny, thin foil packs that tended to wrinkle fairly easily
It was a tall order for any barcode scanning software to hit even one of the three items on their list. When ZiuZ found a decoder with the ability to exceed all three, they were amazed.
ZiuZ tested a few decoding platforms before discovering the power and versatility of TachyonIQ, finding issues with each along the way. One platform’s business model didn’t match that of ZiuZ Medical with a licensing structure that was too complex. And others that simply didn’t perform as well as they needed falling short on several items. After hearing about TachyonIQ and the power of CortexDecoder, they knew they had discovered the decoding solution they had been searching for. A simple test drive of the enterprise SDK and it was confirmed.
In all accounts, the CortexDecoder’s performance exceeded every expectation. The error rate experienced with the previous decoder was all but completely eliminated. CortexDecoder was able to scan thousands of barcodes before even one could be identified as unread if finding any at all.
The 2D barcode that prints on the back of the medication packets hold the information of what pills and how many are supposed to be in every pack, and that information needs to match the photograph that the machine takes of the transparent side of the bag. These dispensing machines only have one chance of reading each barcode that zooms by. CortexDecoder reads these barcodes with zero-miss reliability, allowing the verification process to happen. ZiuZ Medical has found that by integrating CortexDecoder into their system, it has made their machines more efficient and helped them move closer to improving patient safety worldwide.
“CortexDecoder provides the scan performance we have been looking for,” says Johan van Duijne, Product Manager for ZiuZ Medical. “Not only did the decoder perform better in our machines than we expected, but the customer support provided by TachyonIQ made choosing their decoder even easier.”
Decoding in Action
For OEM manufacturers looking to integrate enterprise-grade barcode scanning into their equipment, there is a solution that will improve efficiency and exceed every expectation. CortexDecoder is a high-performing decoding algorithm bundle that works on any platform or operating system.
By pairing CortexDecoder with the latest imaging technologies – you can deliver a market-leading product that will provide your customers with true enterprise-grade barcode scanning capabilities and zero-miss performance even in less than ideal scanning environments or on damaged barcodes.
Interested in test driving our tech for yourself, or learning more about the SDK features? Check out some of our great videos showcasing our SDK in action.
- Cheung, Ka-Chun, Marcel L Bouvy, and Peter A G M De Smet. “Medication Errors: The Importance of Safe Dispensing.” British Journal of Clinical Pharmacology6 (2009): 676–680. PMC. Web. 12 July 2018. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2723208/
- Guidance on the interdisciplinary safe use of automated dispensing cabinetsHorsham (PA)Institute for Safe Medication Practices; 2008[cited 2009 Sep 8]. Available from: http://www.ismp.org/Tools/guidelines/ADC_Guidelines_Final.pdf