Accurate constituent identification is a fundamental building block for every intervention across healthcare, finance, education, and more. Yet, one billion people have no formal identity, making them invisible in the eyes of the world. This fact is exacerbated in situations of refugee populations, undocumented immigrants, and in times of conflict.
Even in countries or populations where formal identification is high, some data, such as for health or political opposition or those in marginalised communities, privacy protection may require a “delinking” of their case data from a government-issued identification number.
Many traditional approaches for health identifiers have significant flaws.
Using personal identifiers including names, date of birth, and postcodes are often not culturally appropriate as many names are very common, people may not know their exact date of birth, many parts of the world do not use postcodes, and/or people may use different names in different contexts for cultural or privacy reasons. Programme-specific tracking tools like vaccination cards, physical QR codes, and patient booklets carried by beneficiaries are often lost or damaged, especially in conflict zones.
Identification Systems in Healthcare
In healthcare provision, many organisations use standardised treatment identifiers, such as a HIV treatment ID number or facility IDs. These alphanumeric IDs are used by health care facilities to identify new and return patients for case management.
These IDs are usually created by the enrolment facility after the first visit, using a combination of the facility code, date of enrolment, and some personal information (such as initials, month of birth). This code is then used in place of the patient’s name to connect all files across location and time.
While this approach is a good basis for a portable ID linking patient data together, the facility based manual enrolment approach can be very cumbersome and difficult to scale. Facilities may use different formats for their ID format, and countries often do not have a central data repository of all enrollees.
Many patients (especially key populations, refugees, and migrants) may visit multiple facilities, creating individual facility IDs for one person. Even when there is a central database, looking up the number may require the patient to remember the date of enrolment or the facility they enrolled in, making linking case management data very challenging and the risk of duplication high.
Biometric Identity Systems
Biometrics are a way to identify who you are or verify that you who you say you are, through measurements of biological characteristics. These characteristics can be physiological (e.g., fingerprints or iris) or behavioural (e.g., voice or gait). There are many different biometric methods depending on what is measured: fingerprints, face, iris, retina, palm, palm veins, voice, signature, gait, etc. Two or more different methods may be combined into a multimodal biometric system.
There is no single “best” biometric method as each offer has advantages and disadvantages, especially when working with specific populations and in specific contexts. It is critical to choose a technology that will work effectively in the specific contexts to prevent people from being misidentified or even excluded from services due to a failure of the biometric technology used.
Biometric Identity Benefits
Biometrics as a way to generate unique identifiers have been successful in connecting cases together, verifying delivery of interventions, and improving beneficiary tracking in “last mile” settings. Using a biometric does not require someone to remember information or keep track of a medical record.
It can also help link different records together by connecting the same biometric to multiple IDs, reducing fraud and duplications. In addition, biometrics can be used to build solid ID systems in short periods of time, especially in remote areas, sudden migration situations, or when ID cards cannot be a requirement for the delivery of services.
For example, following the deployment of biometric systems, a refugee camp in South Sudan recorded savings of $1 million a month. The World Food Program’s SCOPE project has registered 20 million refugees to biometrically verify the distribution of food aid, ensuring the right people are being reached, and also allowing for “better monitoring and risk control”.
Biometric Identity Challlenges
However, biometrics are not a silver bullet; safeguards, planning, and benefit/risk analysis need to be taken into account to determine whether biometric scanners are appropriate for a specific intervention and to track whether the biometrics continue to offer benefits in rapidly changing conditions.
For example, the accuracy of some (but not all) visual biometrics may be affected by unpredictable light levels. Some biometric methods, such as iris scanning, may require specialised hardware that may be harder to maintain in low resource environments. Fingerprint technology may be less accurate for the worn fingerprints of manual labourers, for example, or for people whose fingerprints have been damaged from regular handling of hot cooking implements. Some of the main facial recognition algorithms were shown to have radically different levels of accuracy with people of different racial backgrounds.
It is also important to consider the cultural acceptability of methods that may require users to touch a device (such as fingerprint), or which use an image of a person’s face, as this varies widely across different communities and cultures. A study in Bangladesh found that the majority of veiled Muslim women were willing to provide a fingerprint, although over 70% objected to having their iris scanned or photograph taken.
Conversely, less than half of a group of female sex workers in Zambia were comfortable providing their fingerprint, as fingerprints are often associated with law enforcement. The cultural context can also change rapidly due to outside factors, such as the use of biometrics for national identification programmes.
It is therefore important to choose biometric technologies that are calibrated for the population in question, monitor the collection of this data to identify potential exclusions, and try to choose a method which is least likely to lead to exclusion or discriminatory outcomes.
An edited synopsis of Using Biometrics for COVID-19 by Sarah Grieves, Simprints and Siobhan Green, IMC Worldwide
Sorry, the comment form is closed at this time.