Data Science
Background and Context
Immunization information systems (IIS) across the United States allow providers to use their electronic health record systems to obtain individual patient immunization records in near real-time through a single-message query and response. However, the COVID-19 pandemic highlighted the need to retrieve immunization information for large groups of patients/members, by providers and other partners, without relying on individual queries. Individual queries can be time consuming for users and adversely affect performance for IIS when sent in large volumes. During the pandemic’s vaccine response, not only did the number of queries rise, but the number of authorized stakeholders requesting information for groups of patients also increased, causing IIS to experience performance challenges. In addition to vaccine providers and schools, newly authorized stakeholders for group queries might include healthcare payers, and employers.
On average, an IIS receives twice as many queries as submissions. According to an AIRA member conversation from April 2021, in a single day one jurisdiction received 570,000 queries, dwarfing the approximate number of 17,000 queries per day for the average IIS pre-COVID-19 noted in an AIRA survey from March 2020. It is estimated that IIS are managing roughly a tenfold increase in queries during the COVID-19 response and that IIS are experiencing performance challenges created by the added demand to process the increased volume of single query messages.
The Immunization Integration Program (IIP), a multidisciplinary program striving to improve immunization interoperability, selected bulk query as a topic for the IIP Collaborative. A landscape analysis was conducted through discussions with more than 40 experts from EHR vendors, IIS, healthcare payers, provider organizations, school health providers, pharmacies, and public health organizations. The IIP convened a workgroup to analyze the issue and to recommend a path to improve queries in bulk to IIS. The workgroup included volunteer members representing IIS, EHR vendors, IIS technology partners, patient-focused portals, pharmacies, health information exchanges, provider organizations, healthcare payer organizations, clinicians, and technology standards experts.
Before developing its recommendations, the workgroup evaluated the features and limitations of commonly implemented approaches to requesting and retrieving immunization data for groups. The three most commonly implemented mechanisms were serial single-message queries, batch HL7 v2 messages, and flat or delimited file exchanges.
IIP Bulk Query Toolkit
The IIP Bulk Query Toolkit includes approaches to group data exchange, general requirements for sharing data and mitigating potential conflicts, specific recommendations, and details about individual use cases. Bulk immunization queries occur when an authorized entity (e.g. healthcare organization) requests immunization data for a group of people from an IIS. A bulk query is a single request containing a list of patient/member identifiers and demographics to be matched to records within the IIS. Typically, this request is submitted but the requester does not maintain an active connection while awaiting a response. Instead, once the data is available the requester retrieves or receives it from the IIS. More formally, bulk query is defined as:
A single request for the information for an entire group of people (patients, members, or students) known to and affiliated with a requestor. In this context, bulk query is an asynchronous technical approach for efficiently requesting a server to export a large amount of data to a pre-authorized client. While the term “bulk” suggests there may be a size requirement regarding the amount of data, size requirements and limitations should be published by the responder (receiver) system and discussed with the data requesters (senders).
Regardless of the approach used for requesting and retrieving immunization data for groups, key recommendations in the toolkit for immunization data requesters and IIS are summarized below:
- Prepare for eventual bulk query needs by evaluating principles of bulk queries, including needs, methods, timing, limits, policy and reconciliation
- Maintain communication amongst immunization data requesters and IIS to understand and agree upon data refresh frequency, planned outages and delays, changes in policy, and the available method(s) for submitting queries.
- While not widely implemented, consider using batch HL7 v2 message exchange in the interim as a standards-based until FHIR bulk query is fully developed and tested to request immunization-related data.
- For currently unmet group immunization query needs, avoid using serial single queries or implementation-specific file exchange methods.
- Limit duplicate queries (i.e., the same individuals within a short time period).
- Prioritize immunization record submission and employ strategies to minimize load on IIS production systems, such as:
- Use of parallel environments to support bulk query needs such as data lakes or query-specific environments.
- Scheduling queries for group immunization information at off-peak times.
- Throttling or limiting or group queries to a mutually agreeable rate (e.g., 100 single message queries per hour).
In addition to the above recommendations, the toolkit provides specific recommendations for moving forward with mechanisms for bulk query. The best practices are divided into business requirements and technical requirements. Some of the technical requirements address infrastructure but do not recommend specific architecture. Other technical requirements address software functionality and interoperability requirements. Best practices require all three — business, functionality, and interoperability requirements.
The workgroup identified several bulk query use cases to support stakeholders needing to:
- Assess immunization coverage for groups or entire populations.
- Plan and carryout vaccination activities.
- Provide routine health care using complete and accurate immunization records.
The IIP Bulk Query Toolkit also includes examples of how to apply the recommendations to use cases in the form of user stories for various stakeholder groups, decision flow diagrams to help IIS and querying organizations determine the best approach for the near-term while awaiting greater maturity, and examples of bulk FHIR approaches.
Future Work
The Helios FHIR Accelerator for Public Health is evaluating the FHIR Bulk Data standard to perform bulk queries of IIS. This work should further define specific guidance for the use cases defined in this toolkit. This will not require the replacement of existing bulk data efforts that work well for both IIS and querying entities. Further details and more specific testable requirements will become evident during testing and evaluation.
The workgroup also recognized the value of existing IIS efforts to use direct reporting capabilities, enabling authorized users or organizations to access data for their existing clients as a report accessible through the IIS web portal that uses data from a data warehouse. This effort, however, is an individual IIS approach and not a bulk query initiative. Further, several participants expressed a preference for subscribing to information about a patient cohort as new information becomes available. Such a subscription approach is different from bulk query.
Subscription services may offer benefits but also present unique challenges. Subscriptions require the requesters to regularly maintain their client listing; it may be difficult to review a roster within an IIS to add new patients and remove those who no longer have a relationship with the requester. Subscription services as an alternative to bulk query requires further analysis and consideration of emerging standards such as those in HL7 FHIR Subscriptions Framework and its proposed Subscription Resource.
Call to Action
Readers are encouraged to participate in or follow the Helios efforts through its Confluence site and volunteer for evaluation and testing options. Additionally, the IIP offers flexible technical assistance opportunities to partners seeking to implement IIP guidance and solutions. If your organization is interested in using the IIP Bulk Query Toolkit and has questions or would like assistance, please contact iip@himss.org. For more information about Helios, email helios@hl7.org.