So, you’re interested in leveraging biometrics for patient identification. That’s great news! But before you make your investment, are you asking the right questions?
A. Cleanliness and hygiene are critical in a healthcare environment. Some biometric readers (fingerprint, finger vein, palm vein) require a patient to make physical contact with the hardware device to capture a quality scan for matching purposes. Contact with a device requires a healthcare facility to sanitize the hardware device before the next patient arrives. For an average facility, this can mean spending tens of thousands per year on these sanitary wipes.
A. This could perhaps be the most important question to ask when investigating the feasibility of a biometric patient identification system. It is widely known in the healthcare industry that duplicate medical records are a drain on both money and resources as medical facilities struggle to discover a solution that will actively prevent their creation. A biometric patient identification solution can indeed prevent the creation of new duplicate medical records, but only if it is a system based on what is called one-to-many identification (1:N). Learn more.
A. There are certain human physiological characteristics that stay the same from birth and others that change over time. Depending on what biometric hardware modality you choose for your patient identification system you may or may not have to re-enroll children over time as they get older. In addition, some biometric systems can’t be used at all on young children or infants.Here is a breakdown of common biometric hardware modalities used for patient identification and their enrollment specifications:
Palm vein –
Palm vein patterns are formed during the first eight weeks of gestation in a chaotic manner, influenced by environment in a mother’s womb. This is why the vein pattern is unique to each individual, even to twins. Veins grow with a person’s skeleton and do not reach a stable, “adult” pattern until growth stops. Consequently, patients using palm vein may have to periodically re-enroll over time.
Like a snowflake, the iris – which constitutes the extremely visible colored ring around the pupil – of every human is absolutely unique, exhibiting a distinctive pattern that forms randomly in the womb during a process called chaotic morphogenesis. The iris pattern and structure exhibit long term stability and do not change from 10 months of age throughout the entire life cycle, allowing even infants to be reliably identified. Iris recognition is considered the most accurate form of biometric recognition and the chances of two irides being identical is 10⁷⁸.
A. Closed architecture biometric patient identification systems cause healthcare facilities to rely exclusively on one biometric hardware vendor for the life of the system. In other words, only one piece of biometric hardware can be used with the software and if the hardware vendor decided to change their technology down the road, every single patient enrolled in the system would have to go through a time-consuming, laborious re-enrollment process. If this were the case, you could stand the chance of losing years of data collection.Open architecture systems classify biometric hardware as a peripheral and allow healthcare facilities to choose from a variety of biometric hardware devices to protect their investment. Open systems offer much more flexibility and can work with many biometric devices.
Independent testing and certification of biometric hardware and software technology is key to measuring the quality and integrity of the product. Independent testing and certification demonstrates that the software has passed rigorous, third-party testing and supports the features and benefits advertised by the manufacturer. It is particularly important to check which organizations have performed any testing on a biometric algorithm, device or system. For example, the National Institute of Standards and Technology (NIST) is a benchmark testing and certification organization for biometric technology and they have years of experience working with many different applications from scores of biometric software research and development companies.If biometric hardware and/or software has not been independently tested and certified then accuracy and performance metrics provided by the vendor cannot be benchmarked and verified. Considering the weight and depth of the investment in a biometric patient identification solution, is the absence of independent certification and testing by a reputable third party organization worth the risk?
A. Please review Q2 for some background information on 1:N, 1:1 or 1:Few searches. In addition to clarifying the biometric search type, another important question to ask is how quickly you can expect to receive a search result after a patient has scanned their biometric information.In the biometric industry, search speeds are normally based on three items:
- The number of biometric templates being searched
- The size of the template
- The power of the underlying search algorithm
While it isn’t critical to understand the science behind the technology in order to factor this information into your purchasing decision, asking this question will help to determine the architecture of your biometric patient identification system and whether or not you may need to invest in additional hardware or software to satisfy your expectations of how quickly you expect to receive a match when a patient scans their biometric data.