Data Sharing

Effective D2C requires collaboration between health department prevention, care, and surveillance programs. Data sharing agreements between collaborators ensure that D2C programs have access to the data needed to effectively locate and re-link individuals to HIV care while ensuring that client-level data are shared securely and confidentially between other programs and partners.

Many health department programs have internal data sharing in place for D2C. The internal flow of data may look different in different jurisdictions depending on the organizational structure and level of integration within the health department. If surveillance and prevention are located within the same division and both are supervised by the same director, there may be fewer barriers to data sharing. Additional levels of approval and documentation may be required in health departments where programs are organizationally siloed or have less history of collaboration. Local D2C programs at the city/county level may have difficulty accessing surveillance or care data housed at the state level without a formal agreement and working closely with the state health department.

Although many programs may share data freely within the health department without a data sharing agreement, having a formal agreement or policy is best practice to ensure consistency within your program and continuity if staff turnover occurs. Having a formal document describing the data flow practices for your D2C program can prevent the need to rebuild buy-in with new leadership and will make staff transitions smoother by ensuring that longstanding institutional knowledge and practices are documented in a policy (i.e., responsibilities for data flow should be job-specific and not person-specific).

For health departments interested in formalizing data practices within their health department, NASTAD has data sharing agreement templates available for RWHAP Part B/ADAP programs and HIV surveillance programs, state Medicaid programs, and RWHAP Part A funded Eligible Metropolitan Areas (EMA) and Transitional Grant Areas (TGA).

Data from NASTAD’s National HIV Prevention Inventory (NHPI) describe the types of programs and partners with which health department prevention programs have data sharing agreements for D2C-related activities. Among jurisdictions currently piloting or implementing D2C, 73% (n= 32) of programs reported having data sharing agreements in place with internal health department programs such as the Ryan White HIV/AIDS Program (RWHAP), the AIDS Drug Assistance Program (ADAP), and surveillance. Some health departments developed data sharing agreements with programs and partners external to the health department to facilitate re-linkage. The most common of these include local health departments, community-based organizations (CBOs), and health care providers. Other programs to consider sharing data with include major hospitals, correctional settings, and health care plans (including Medicaid) within your jurisdiction. Many D2C programs have reported difficulty receiving data from certain external entities. In particular, many health departments experience challenges receiving lab data from people living with HIV (PLWH) who are involved in clinical trial research or those who are receiving care at Department of Veteran's Affairs (VA) facilities. In most instances, health departments may need to develop relationships with the leadership and/or privacy officer at individual VA medical centers in their jurisdiction to develop data sharing policies to have this data reported regularly.

Data Sharing Agreements in Place with Health Departments for D2C (n=44)*

Entity Number of Jurisdictions Percent of Jurisdictions
Internal health department programs 32 73%
Local health departments 21 48%
Community based organizations (CBOs) 14 32%
Health care providers 13 30%
Hospitals 8 18%
Health care plans (public and/or private) 5 11%
Department of Veteran’s Affairs (VA) 3 7%
Tribal governments and/or tribally designated organizations 3 7%
Health Information Exchanges 1 2%

* data from NASTAD National HIV Prevention Inventory, 2017

The types of partners health departments share data with for D2C will depend on the specific D2C approach. Jurisdictions should carefully consider data security and confidentiality when exchanging data with entities outside the health department.

Sharing client-level data with non-governmental agencies including CBOs or AIDS Service Organizations (ASOs) can present challenges with confidentiality as they do not have the same public health authority as a health department. However, some health departments work with CBOs to assist with linkage if that not-in-care individual has previously been a client of that CBO or gives verbal or written consent to be linked to a specific CBO provider for case management.

Health departments should also consider what data are shared back with providers. For example, jurisdictions can look at adding consent language to allow the health department to inform testing providers whether an individual who tested positive with their CBO or health care practice was subsequently linked to care. The option to do this may depend on state or local regulations (i.e., the health department may be allowed legally to share data for purposes of continuity of care). Health departments should consult their local laws to determine specifics of what is allowable.

For a more in-depth analysis of issues related to public health data use and data sharing practices, also see HIV Data Privacy and Confidentiality: Legal & Ethical Considerations for Health Department Data Sharing.

Cross-Jurisdictional Data Sharing

Some health departments also develop arrangements to share data across state lines to confirm whether individuals on their not-in-care (NIC) list are living and/or receiving care in a neighboring jurisdiction. In implementing D2C, many health departments find that a significant proportion of individuals who appear on not-in-care lists have moved out of the jurisdiction. There are tools available to locate recent addresses (e.g., Accurint or other people-search tools), and there are processes in place to periodically de-duplicate surveillance data nationally (i.e., Routine Interstate Duplicate Review [RIDR] and Cumulative Interstate Duplicate Review [CIDR]). However, additional data exchange processes may make sense in some cases. In 2018, the Centers for Disease Control and Prevention (CDC) began funding a five-year program (PS18-1805) at Georgetown University to develop a secure data-sharing tool that would allow public health agencies to share surveillance data between jurisdictions. While such efforts are still underway, routine interjurisdictional data sharing may be most worthwhile in areas where individuals frequently relocate in and out of the jurisdiction, or where many PLWH live in one jurisdiction but receive care in another. Having procedures to share surveillance data in a timely manner can identify overlap in NIC lists to avoid duplication of efforts and prevent wasted staff time attempting to locate an individual who has moved out of the jurisdiction.

Health Department Example

Interjurisdictional Data Exchange in DC, Maryland, and Virginia

DC, Maryland, and Virginia state outlines

Within the larger D.C. metro area, many PLWH may live, work, and receive care in multiple jurisdictions. To address the surveillance data challenges this presents, the District of Columbia, Maryland, and Virginia departments of health share data interjurisdictionally through projects including the “Black Box” pilot and DMV (D.C., Maryland, and Virginia) HIV Surveillance Data Exchange. The Black Box pilot is described in this publication2 and is also featured in NASTAD’s Data Points resource. Following the pilot, the DMV HIV Surveillance Data Exchange was created in 2017 as a collaborative effort between health department staff from all three jurisdictions. Through this project, the health departments developed cross-jurisdictional data sharing agreements, created standardized SAS codes and variables, and developed a secure file transfer protocol (SFTP) site to conduct quarterly data exchange. After the data exchange between the three jurisdictions, there were 396 fewer PLWH estimated to be in DC each year over a five-year period, and the volume of cases needing Routine Interstate Duplicate Review (RIDR) decreased by 74% for DC-Maryland and 81% for DC-Virginia.3 As part of this project, staff from all three jurisdictions continue to participate on regular calls to discuss their surveillance data. Improving the efficiency and timeliness of data exchange between jurisdictions helped these health departments clean up their prevalence data, work with providers on improving reporting, and eliminate or de-duplicate individuals on their not-in-care list so their program can focus more attention on their own in-jurisdiction clients.

2 Ocampo JMF, Smart JC, Allston A, et al. Improving HIV Surveillance Data for Public Health Action in Washington, DC: A Novel Multiorganizational Data-Sharing Method. JMIR Public Health Surveill. 2016;2(1):e3. Published 2016 Jan 15. doi:10.2196/publichealth.5317

3 Hamp AD, Doshi RK, Lum GR, Allston A. Cross-Jurisdictional Data Exchange Impact on the Estimation of the HIV Population Living in the District of Columbia: Evaluation Study. JMIR Public Health Surveill. 2018;4(3):e62. Published 2018 Aug 13. doi:10.2196/publichealth.9800

Prioritization of Not-in-Care List Data

A health department’s decision about how to prioritize their NIC list will depend on circumstances unique to each jurisdiction, including local epidemiological data and the primary goals of the D2C program. However, programs should take several elements into consideration when determining not-in-care list prioritization:

  • Some key initial questions to develop criteria can include: What will make the largest impact on the epidemic within your jurisdiction? Should groups be prioritized by demographic categories, clinical characteristics (e.g., unsuppressed viral load, time since last evidence of care), transmission categories, or geographic areas?
  • While prioritization criteria are population-level, also consider the ethics and impact at the individual-level among those who may not be prioritized.
  • Consider internal staff capacity. If you have more individuals on the NIC list than your staff can follow up with, it may be beneficial to first determine the reasonable maximum staff caseload and work backwards to develop prioritization criteria based on how many individuals could reasonably be followed up with based on current available staff capacity. For example, see this sample scenario about using prioritization variables to estimate the number of individuals on the NIC list can be followed-up using existing health department resources.
  • Engage with your providers to determine their capacity to accept additional clients being re-linked to care through your D2C program (Example: see “Monitoring Medical Capacity” section of the Link-Up Detroit D2C Pilot Program Protocol). Ideally, D2C programs should try to make sure people are linked to providers that will successfully retain them in care to minimize clients repeatedly dropping out of care. Getting provider buy-in for your program is key and giving information back to your providers can help improve relationships.
  • Determine the potential need for, and availability of, wraparound services for clients that need linkage/re-linkage.
  • Programs should be intentional about field staff time when considering prioritization. For example, people diagnosed most recently will typically be easier to find. If your program chooses not to restrict the NIC list by date and prioritizes those who have been out of care for the longest period, there will likely be large numbers of individuals who were diagnosed many years ago and will be unlocatable. Programs should try to avoid sending staff into the field to find individuals who are unlocatable or already in care, as it is undermines staff morale and is an inefficient use of D2C resources.
  • Prioritization should be revisited and may need to change over time in response to your D2C program outcomes, staff resources, and local epidemic. Health departments should use their D2C program outcome data to go through an iterative process to finetune how the program prioritizes future NIC lists.

Data Collection and Management

Health departments use a variety of methods to collect and manage data related to D2C. This can include using or modifying existing health department data systems to track D2C data, or developing new systems designed specifically for their D2C program. Health departments should consider what types of data they will need to collect, including:

  • Demographic and locating information
  • Most recent CD4/viral load laboratory results
  • Contact notes from outreach efforts
  • Disposition codes (e.g., deceased, moved out of jurisdiction, confirmed not-in-care, etc.)
  • Information about additional services or referrals needed (e.g., housing, mental health services, etc.)
  • Reasons reported by not-in-care individuals as challenges or barriers to HIV care
  • Evaluation variables (e.g., time spent on investigation per client)

Other considerations for D2C data management include:

  • Do you need a system that Disease Intervention Specialists (DIS) and other D2C staff can access remotely?
  • How will be data be collected in the field (e.g., paper forms transported back to the office for data entry, or via secure tablets)?
  • How will your program identify individuals in the data system (e.g., unique client ID number)?
  • Which other databases and source(s) will your data collection system need to communicate with (eHARS, CareWare, state/local labs database, STD database, etc.)?
  • Who will have access to the NIC list in your system?
  • What format will you need data in for reporting?
  • Do you need the ability to set up contact reminders or other alerts in the system?

Additional benefits and challenges to using different formats of data systems for D2C are available in the NASTAD resource, Approaches to Data to Care Data Management.