Developing and implementing national health identifiers in resource limited countries: why, what, who, when and how?

Figures & data

Figure 1. Ten WHO indicators assessing the ’90–90-90’ program showing linkages with other sources of data including case based surveillance and patient monitoring.

Box 1. The five basic functions and seven core elements of a NHID [20].

The five basic functions that a NHID should support are: ability to unambiguously identify individual patients at all care settings, from a clinical and administrative perspective; ability to link a variety and continuously evolving set of data elements across institutions, service providers and time, to constitute a lifelong view of the patient’s medical history, needed to deliver healthcare; ability to aggregate information across institutional boundaries for specific services and health outcomes, to track health program implementation and strengthening surveillance; protect the privacy, confidentiality and security of personal health information at the site where the information is held and by de-identification of records before they leave the primary site; reduce operational costs by supporting automated record management and information sharing.

The seven core elements of a NHID, required for the system to perform its functions consist of: an identifier system, consisting of alphanumeric characters with check-digit functionality, ensuring that numbers or characters are not transposed when typed, while the identifier should exclude characters that represent any aspect of the individual’s identity, including date of birth, gender or other personal identifiers identification information, including demographic or biometric data cross-references to local, site- or program-specific health identifiers – like facility medical record numbers or TB program identifiers – and other identifiers – including, civil identifiers, national identity documents, passports, drivers’ licenses, or others mechanisms to hide or encrypt identifiers so they don’t disclose any personal information software to register patients software to search, match, merge, encrypt or otherwise manipulate underlying information an administrative infrastructure, including human and financial resources that comprises the governance centre for managing and controlling the how NHID’s are allocated and used.

Figure 2. The transition from Tier 1 to Tier 3 facilities, sub-Saharan country 2000–2014.

Figure 3. Initial steps to develop a NHID system.

Box 2. Baseline assessment of existing national health information infrastructure.

Such a survey needs to include the following: a review of previously performed assessments such as Health Metric Network [23,24] or other surveys and secondary sources the number of facilities with and without with electrical power, with internet and percentage of ‘down time’, and the number of anticipated future new facilities an assessment of what power and telecom networks exist, what is being planned and time frames, checking estimates with internet service providers and national electrical and telecommunications provision identification of those parts of the country that are inaccessible during the year and how this affects power or telecommunications. an assessment of the number of facilities with existing computers, computer users and their skill levels, the presence of air conditioning and secure room for a server and identify the number of facilities with electronic medical or health records list and characterize each different EHR initiative at the level of data structures used and validations applied to the collected data.

Figure 4. Developing and implementing NHIDs and sections of the NHID system.

Box 3. Different models for implementing a NHID [20].

Heavily centralized model. This model typically leverages existing government services that occur at one central location, such as the capital city. Characteristics of a heavily centralized model include the following: Costs are minimized since the model can use enrolment locations that are already providing registration services and possibly are already issuing permits or licences. Typically, a single or small number of events induces the citizen to apply for inclusion in the identification system. These events might include birth, reaching a certain age, hospitalization, joining the military, or applying for a driving or marriage licence. The model lends itself to centralized administration and issuance of bar-code stickers or cards. Annual costs are typically lower than for other models due to the limited number of people applying on an annual basis, and use of existing infrastructure to receive applications. The costs of processing applications and administrative costs of the national registry are nearly the same per enrolee for all models.

If the country has predominantly more Tier 1 facilities instead of Tier 2 or 3, one has to solve the logistic issues of having a printer in the capital serve the label printing needs of all the rural clinics in the country, something that can be insurmountable. A compromise between highly centralized and highly distributed Tier 2 or 3 facilities is to print bar codes on initialized chip cards centrally, thereby assuring their uniqueness, so that the identifier on the chip equals the printed bar-code. Such cards can be centrally produced but locally assigned to clients.

Semi-distributed model. This model is similar to the heavily centralized model, but it leverages existing government services that typically occur at regional or provincial centres. Characteristics of a semi-distributed model include the following: Costs are typically somewhat higher than for a heavily centralized model but lower than for a highly distributed model. Typically, a single or small number of events induces the citizen to apply for inclusion in the identification system. These events are similar to those for a heavily centralized model. Typically, the application points have card printers and can issue cards upon administrative approval.

Highly distributed model. In this model, NHID administration occurs at the point of health service delivery throughout the country. Characteristics of a highly distributed model include the following: There is a much more complex rollout schedule, with a number of teams performing site readiness to meet an aggressive schedule, such as 1–2 years. Costs are higher due to the higher number of application and issuance points and the number of teams executing site start-ups. If an aggressive schedule is desired, a higher number of rollout teams will be required, and administrative costs will be higher in processing the increased number of applications. Many events will typically induce the citizen to apply for inclusion in the identification system, including all of the previously mentioned events plus community enrolment efforts and medical events such as hospitalization or clinic visits. A carefully thought-out and available communication system is required, with robust connectivity to at least the district level and probably also the larger site level. Rollout costs will be higher due to a larger rollout team, which includes providing equipment, hiring staff and training on the identification system. Unless there is a uniform national language, translation will be needed at least for the application forms and possibly also for the data-entry screens. This is a more complex NHID registry service model, due to the need to communicate to a wide variety of electronic systems, provide printouts for paper systems, and maintain highly synchronized pools of data to ensure no single point of failure in the national system.

Figure 5. Training requirements, upgrades and data stewardship.

Box 4. Potential benefits of identifier chip cards.

Patients carry and control their own basic medical information, and assurance of best practice information through information driven care. People with a chip card register for visit with card swipe, are served first to improve carry rates. The card has a problem list, reasons for all visits, provider information, medical history and other summary medical information for best care wherever client goes. The card contains all prescriptions and can be filled at a pharmacy. The card has all laboratory order requests and results, and vital signs. The card has all diagnoses and treatments, and can prevent harm from conflicting treatments. If a card is lost, place of last care can provide a full copy. It can provide a patient a review of his/her own record.

Box 5. Understanding of shared benefits, risks and actions.

Reflections of public and political will a broad and correct understanding of the system, its purpose, benefits and risks; broad popular support and trust of the system among civil society and all stakeholders. the existence of laws that support the system’s purpose and necessary procedures. laws, regulations and procedures that assure the safety of the system, that it protects the confidentiality and security of personal health information, that it is trusted and practical, and undergoes regular appraisals of its security. that there is a mechanism that assures the system will remain well-resourced long term. that a plan exists for decommissioning confidential information and tools, if the need exists.

Manifestations at the level of the system the existence of feedback control processes for all critical indicators of system operation, which provide information about the integrity of each of the elements of the system components and processes. providing this information in timely and actionable form to the persons responsible for any necessary interventions, and keeping alternative responders and supervisor abreast of developments. feedback control processes that are designed to maintain the homeostasis of certain operations at the correct balance, and in other cases are designed to optimize performance over time through ever refined system processes such as match algorithms. learning on the part of all persons that are part of the system, including not forgetting identity cards, so that the dependence on the matching algorithm is continually reduced. growing numbers of vital private and public health functions that are served by the system, in accord with wise and fair data use policies that both serve the public interest and continue to assure the security of the individual interest. For instance, the necessity for highly personalized cancer care or early interventions in disease with genetic markers may become standards of care, and the life-long records prove life-lengthening. that everyone carries their identity card, accepts a couple of biometrics, and appropriately trusts the data stewards with sensitive information where public interest has exercised exception to individual rights. Given the protective measures in place – including safety, justice, confidentiality and security of the system – and that these have met with public approval. that in the public realm the automated summaries of individual records to produce a national aggregated resource of information are demonstrated to be practical with reasonable precautions and controls. The pooled information can be used for training, education and research to develop public policies, with an ever growing use of such a valuable national information asset. the existence of interoperability at all levels of the system, between the potentially disparate clinical information systems which have evolved into standardized systems to provide continuity of information to the users of the service. that patient care can rely on the system and benefits from it, as do government and academic institutions and that business may profit by it, provided this is consented, the personal health information is protected and the work serves the wider public good.

UNAIDS/PEPFAR. Considerations and Guidance for Countries Adopting National Health Identifiers. Geneva (Switzerland); 2014. Available from: http://www.unaids.org/en/resources/documents/2014/national_health_identifiers

 Google Scholar

WHO. Health statistics and information systems. Geneva (Switzerland). Available from: http://www.who.int/healthinfo/country_monitoring_evaluation/documentation/en/

 Google Scholar

WHO. HMN country files. Geneva (Switzerland); 2016. Available from: https://www.dropbox.com/s/agme52z13pto136/countries.zip?dl=0

 Google Scholar