Because of the stigma still associated with HIV disease, we initially thought that confidentiality might be a key barrier to HIE in HIV care settings. Instead, we found widespread acceptability for HIEs across patients, providers, and other stakeholders. Considering both our large quantitative patient dataset and qualitative interviews with stakeholders at the six sites part of HRSA’s demonstration project, our findings have broad implications for HIV treatment settings. They suggest that those who deliver and receive HIV care have grown more comfortable with integrating HIV-related patient data into modern technologies that may improve information flow and HIV care. These attitudes appear to be driven by growing comfort with information technologies, confidence in the security protocols utilized to protect data, trust in the providers and institutions who use the technologies, belief in the benefit to the patients, and awareness that electronic exchange represents an enhancement of data sharing already taking place by other means. Our quantitative data illustrate that patients were willing to share their information electronically with recipients who may help provide them with better care, such as medical providers, HIV support organizations, and government insurers. Thus, while patients still may be concerned about stigma and about people outside their healthcare finding out about their HIV status, they may not be that concerned about sharing their information with people who may impact their healthcare. The farther the data-recipient is from impacting their health directly, such as non-clinical staff or private insurers, the less willing were the patients to share their information with them, perhaps under the logical inference that there is no reason for those people to know. Although we found broad acceptability for electronic data sharing, the stakeholders’ attitudes continued to reflect an underlying expectation that access to information would be carefully monitored. Such perspectives included trust that the data were secured by encryption and other technological controls and that access rights to the systems had been well considered.
While the factors above acted as facilitators, our data suggest that barriers for electronic data sharing at the institutional level need to be carefully considered. Because some of the HIEs were high profile endeavors likely to attract attention from outside entities (e.g., media), the sites had to go through rigorous scrutiny at different institutional levels to assess their legality and potential liability. Many of the sites had to engage in long and formal reviews that thoroughly documented the design of the systems and the potential legal or public relations implications of implementing HIEs. These processes were further complicated by the very nature of the HIEs themselves, which link disparate institutions. The attitudes of institutional authorities were swayed by careful review of security procedures and governing laws. Those processes, however, were more cumbersome than the ones required to educate patients or providers. As such, the process of HIE development ultimately became one of forging shared vision and consensus among institutions.
The importance of consensus building and trust in the systems was also evidenced in the providers’ reflections on their own evolving attitudes regarding data sharing. The closer the stakeholders were to the process of HIE design and development, the greater their level of trust. System development and implementation required taking into account the attitudes and concerns of a wide variety of individuals potentially affected by a HIE. The often lengthy preparatory processes ended up serving as a crucial component of early success, because it allowed sites to address concerns, educate stakeholders about design choices, and build trust and support for system implementation. Without these processes, it is likely that stakeholder perspectives may have been very different.
HIE acceptability may depend on preexisting trust among patients, providers, and institutions. In our data, patient trust of the systems appeared to be driven in part by the confidence they already held in their providers and care facilities. It is important for further research to explore patients’ specific understanding of the data kept and currently shared in medical records , and their rationale behind supporting or not supporting enhanced information sharing . It would be particularly relevant to examine these issues with patients from communities that are likely to still be suspicious of or fear data sharing. Although stigma against HIV itself has declined , legal sanctions and social norms still may influence some populations affected by the disease . For example, publicly funded clinical settings provide critical safety net health coverage to undocumented workers who, because of the fear of deportation, may be far more hesitant to seek care if information about them were to be shared across regional health networks and government agencies, such as health departments.
In our findings, some users raised concerns that patients of certain ages or racial/ethnic backgrounds might object to, or might need additional reassurance, before they agreed to electronic data sharing. Our patient survey data did not permit us to test those hypotheses because the present study was not powered to detect attitudinal differences in patient subgroups. Further research will be needed to determine whether there are indeed specific groups of patients who are reluctant to share their personal health data electronically. If findings were to reveal that provider perceptions are not matched to actual racial/ethnic differences in patient attitudes about the acceptability of sharing health data, then it also would be important to identify the factors that lead to such a discrepancy.
Stakeholder perspectives were derived by the trust they placed in their collaborating HIE partners. The process of collaboration helped to leverage resources, providing opportunities for smaller agencies to benefit from the technical expertise and support provided by the IT staff affiliated with a HIE. Smaller agencies also received legal advice and health information systems capacity-building that they otherwise would not have been able to afford. In this sense, the HIEs helped to distribute resources more evenly across agencies and to enhance, access to technologies that could help address unmet needs.
Clinical environments should not anticipate easy implementation of information exchanges if there is existing distrust between patients and providers, or if partnering agencies have historically been suspicious of one another. Efforts must be made to forge common bonds before directing resources to setting up the technological infrastructures required for data exchange. Addressing attitudes on trust and confidentiality of sharing patient data electronically will influence the perceived usefulness, the perceived ease of use, and adoption of HIE. Gaining stakeholders and patients trust in the security and confidentiality of the data is absolutely essential when designing and implementing HIEs.
Our findings suggest that confidentiality concerns related to sharing data electronically will not be barriers to HIEs becoming integrated into patient care, at least among the general population of patients and providers in publicly funded clinics. Rather, the perception that the HIEs cause no harm and may in fact benefit patients and other stakeholders is likely to influence their use and acceptability. Nevertheless, the question remains whether human error or a breach in confidentiality, whether as part of the HIEs or a highly publicized event in the field of technological communication or online information exchange, may emerge to affect the expected benefits and the trust in the HIEs. Furthermore, acceptability and trust in the safety of personal information shared through HIEs may influence future policy issues, laws, and privacy rules related to a healthcare system and industry driven by new and ever changing technologies and electronic communication [57–60].
Our study has several limitations. First, the data were collected exclusively in public settings. Stakeholder attitudes may differ in other care environments. However, it is worth noting that the majority of HIV patients in care in the US currently receive services through at least one publicly-supported program (e.g., Ryan White Program, Medicaid, Medicare), which enhances the potential external validity of the findings. Second, the data are focused on efforts pre-implementation of the HIEs. Thus, they speak to the processes of building stakeholder trust and buy-in, but do not reflect on satisfaction or efficacy of the systems once in use. Third, the data collected with patients was limited to quantitative surveys. This curtails our ability to explore in detail the perspectives that drove those patients’ attitudes. In collaboration with the six HIE sites and based on our resources, we decided not to include patients in the qualitative sample. Interviewing them pre-implementation may not have yielded much specific information, but rather more potential scenarios, about the HIEs they were not familiar with. Providers were able to offer insights into some of the factors that likely affect patient attitudes. Given the many parallels between provider and patient attitudes and given that patients learn much of their information about systems from providers; we believe that those provider insights are relevant. They cannot, however, be considered definitive evidence about patients’ experiences with IT. Similarly, it is beyond the purpose of this paper to examine specific concerns of groups of patients. However, our patient quantitative data do not seem to indicate a difference according to ethnicity, gender, or sexual orientation in relationship to data sharing. Fourth, we did not interview institutional leaders outside the projects, who were more distal players within the partnering institutions. That was not part of our work, as we were tasked with evaluating the specific projects. Thus, we are not representing their personal opinions, but rather the institutional responses to the proposed projects based on the perceptions of project staff and providers. They offered good insights into the kinds of barriers at the institutional level, but those reports may not offer the same degree of nuance that would be found by interviewing those distal players.