I have made a long and in places mordant response to the current MyHealth Record Review, in part quoting laments by the person conducting the review when he was the Australian Information Commissoner.
'Understanding trust in digital health among communities affffected by BBVs and STIs in Australia' by Christy Newman, James MacGibbon, Anthony K J Smith, Timothy Broady, Deborah Lupton, Mark Davis, Brandon Bear, Nicky Bath, Daniel Comensoli Teddy Cook, Elizabeth Duck-Chong, Jeanne Ellard, Jules Kim, John Rule and Martin Holt at the UNSW Centre for Social Research in Health (2020) comments
Despite extensive government investment to expand digital health, minimal research has been conducted on community views of these systems in Australia. In particular, there has been scant attention to the perspectives on digital health of populations affected by blood-borne viruses (BBVs) and sexually transmissible infections (STIs) has received little attention.
The Trust in Digital Health study was conducted by the Centre for Social Research in Health in partnership with community organisations representing four of the priority populations in the current national BBV/STI strategies: people with HIV, trans and gender diverse people, sex workers, and gay and bisexual men.
Our methods included a national, online cross-sectional survey (April–June 2020) of the general population, including specific recruitment targets for the four priority populations. We also conducted semi-structured interviews with key informants (March–June 2020) with expertise in communities affected by BBVs/STIs, stigma and marginalisation.
The survey sample included 2,240 eligible participants, including 600 (26.8%) classified as members of one or more priority populations. Overall, priority populations reported the lowest levels of trust in digital technologies and in some health care services, and the most frequent experiences of stigma.
Priority populations were more likely to understand the potential benefits of My Health Record, but also to have opted out of having one. These groups were also more likely to have made use of digital services to access essential health care and medications during the COVID-19 response, and the least likely to be willing to share personal information with health authorities.
Key informants were keenly aware of the promise and benefits of digital health, but also concerned about the risks and consequences of communities affected by BBVs/STIs engaging with these systems. Specific issues related to different populations, but there was a shared focus on the harmful impacts of experiencing stigma and discrimination in health settings. Key informants also consistently reported that these communities typically fear that their personal information is more easily shared through digital means without the consent of the affected person, with a range of potential social, legal and economic consequences.
A range of mechanisms and conditions for building trust in digital health were also discussed, including the need for significant reforms in system design, in community consultation processes, and in the policy and legal contexts that shape the everyday lives, rights and wellbeing of these communities.
The variety of evidence we collected suggests that trust in digital health is influenced less by technical design or digital literacy, and more by the relational and structural factors which underpin trust in the institutions responsible for health system design and regulation.
The authors report
To address these concerns, we recommend finding new and more effective ways to ensure that consent is secured to collect, store and share health data, and that consent is specific, dynamic, and informed. Major investments in discrimination reduction strategies at every level of the health care system are also necessary to ensure that health care is accessible, competent, and safe. Resources should be directed towards remediating the legal and policy conditions that continue to discourage some communities from participating in digital health, and in supporting meaningful consultation with peer-based organisations who have the trust of communities affected by BBVs and STIs.
Access to and knowledge about My Health Record
Nearly a quarter of participants (23.3%) had an Australian electronic health record (My Health Record) and had accessed it at least once, while another quarter (25.8%) had a record but had never accessed it. More than a quarter of participants (29.4%) had opted out of or deleted their record. The remaining participants did not know if they had a record (18.0%) or were not eligible for one (3.5%). See Table F1. Priority populations were much more likely to have opted out of My Health Record than the general population sample. People with a university degree and people with one or more long- term health conditions were also more likely to have opted out, while people who had poor/fair health (self-rated) were less likely to have opted out (see Figure F1).
Less than a third of participants (28.8%) reported knowing ‘quite a bit’ or ‘a lot’ about My Health Record. People with a university degree, people with one or more long-term health conditions, and trans and gender diverse people were more likely to know more about My Health Record, while people who had poor/fair health were less likely to know about it (see Figure F2).
Common sources from which participants learned about My Health Record were media articles including social media (49.0%), government advertising campaigns (44.2%) and conversations with friends and family (30.2%). Overall, a relatively small proportion of participants (15.7%) had learned about My Health Record from information provided by community organisations; however, sex workers, people with HIV, trans and gender diverse people, and people with a university degree were more likely to have learned about it via this source (see Figure F3). Further information about participants’ access to and knowledge about My Health Record is shown in Table F1.
Use of My Health Record among participants who had a record
Among the 1,100 participants who had a My Health Record, frequency of use was relatively low. Small numbers of participants had accessed their record weekly (0.3%) or monthly (5.2%) during the past year. Over a third (34.4%) had accessed their record once or twice during the past year. Just over half (52.5%) of all participants who had a My Health Record had never accessed it, while a further 7.7% had accessed their record more than a year ago.
The most common reasons for accessing My Health Record were to know what was in the record (46.7%), followed by keeping informed (20.8%) and making sure the record was accurate (18.9%). Among participants who had accessed their record, the most common interactions on the platform were adding personal notes (28.5%), choosing what content could be seen (24.3%), setting access notifications to show when someone had viewed or changed the record (18.0%) and setting health care service access restrictions (17.2%).
Among participants who had a My Health Record, participants indicated that the most useful information in the record was test results (66.2%), medication history (58.7%), immunisation history (57.2%) and doctors’ notes (55.1%).
More than half of participants (54.0%) were unsure if their doctor had accessed or used the participant’s My Health Record. Nearly a quarter (24.7%) reported that their doctor(s) had added information to their record and nearly a fifth (19.8%) reported that their doctor(s) had accessed information from the record.
Relatively small numbers of priority population groups had a My Health Record and had accessed it, so further analyses of associations between priority population membership and different types of use of My Health Record were judged to be unreliable. Further details about the use of My Health Record and potential concerns about it are shown in Table F2.
Reasons for opting out of My Health Record
Reasons for opting out of or deleting My Health Record are shown in Table F3. Nearly all participants who had opted out of My Health Record indicated they had concerns about it. The most common reason was concern that the government could not adequately protect their privacy (79.3%), followed by concerns about data being shared between government agencies without consent (67.3%) and medical information being hacked or leaked (67.2%).
Priority populations were more likely to endorse a range of concerns relating to data privacy, data being shared without their consent, and concerns about being treated disadvantageously by the government and health care professionals. In particular, sex workers were much more likely to report opting out of or deleting My Health Record due to concern about health professionals not treating them with dignity and care, which was a concern shared by trans and gender diverse people and people with HIV (see Figure F4). Sex workers were also much more likely to report opting out due to concern about data being shared between government agencies without their consent and due to concern that the government might use their data in ways that disadvantaged them. These concerns were shared by gay and bisexual men and trans and gender diverse people (see Figures F8 and F10). Other significant associations are shown in Figures F4–F12.
Information that might have been useful among participants who had opted out of My Health Record
Among participants who had opted out of or deleted My Health Record, the information that was most commonly rated as useful if they had kept a record was their immunisation history (55.2%), test results (55.0%), medication history (50.8%) and doctors’ notes (47.0%). These types of information were also perceived as the most useful by participants who had a record (Table F2), which suggests that the perceived utility of the types of information stored in My Health Record was similar regardless of the concerns that some participants had about the platform.
Willingness to share My Health Record data with health care services
x Participants were asked how willing they would be to allow relevant information from an electronic health record to be shared with health care services. All participants were asked these items regardless of whether they indicated they had opted out of or deleted My Health Record, but participants who responded that they were not eligible for a record were not asked these items.
All participants were shown the items in relation to GP services, pharmacies, dentists, in-patient hospital and out-patient hospitals or specialist clinics. Participants were only asked about the other health care services if they had attended them in the past year, e.g., community-based or peer-led sexual health clinics, allied health services and mental health services. Responses were scored on a Likert-type scale from (1) ‘Not at all willing’ to (5) ‘Very willing’, and willingness was defined as having a score of four or more on the scale (i.e., Willing/Very willing). All aggregated data for these measures can be found in Table F4.
Overall, most participants (77.4%) indicated that they were willing for relevant information to be shared with their GP, followed by in-patient hospitals (73.9%), out-patient hospitals or specialist clinics (68.5%), dentists (56.9%) and pharmacies (56.0%). However, priority population groups were generally less willing to share relevant information from an electronic health record with health care services.
Specifically, sex workers, people with HIV and trans and gender diverse people were generally less willing to share relevant information from an electronic health record with GPs (seeFigure F13), pharmacies (see Figure F14), in-patient hospitals (see Figure F16), and out-patient hospitals or specialist clinics (see Figure F17). Sex workers and trans and gender diverse people were generally less willing to share relevant information from an electronic health record with dentists (see Figure F15). In contrast, gay and bisexual men were generally more willing to share relevant information when it came to GPs, in-patient and out-patient hospitals or specialist clinics (Figures F13, F16–F17).
Willingness to share My Health Record data with government agencies and industry bodies
Participants were also asked how willing they would be to allow relevant information from an electronic health record to be shared with health-related government agencies, non-health- related government agencies, health insurance companies, law enforcement and banking or financial institutions. All aggregated data for these measures can be found in Table F5.
Factors associated with willingness to share My Health Record data with government agencies and industry bodies are shown in Figures F18–F22. Priority populations, people with a university degree and one or more long-term health conditions were generally less willing to share My Health Record data with these organisations. Of note, trans and gender diverse people and sex workers were far less likely to support My Health Record data being shared with these organisations.
Support for sharing de-identified My Health Record data for research
Participants were asked how much they supported their de-identified health information from My Health Record being shared for non-commercial research, commercial research, research by health-related government agencies, and research by non-health-related government agencies. Examples of the types of institutions that were included are shown in Table F6. De-identified was defined as having one’s name, address, contact details and other identifying information removed. All participants were asked about these items regardless of whether they indicated they had opted out of or deleted My Health Record, but participants who responded that they were not eligible for a record were not asked these questions.
Responses were scored on a Likert-type scale from (1) ‘Strongly oppose’ to (5) ‘Strongly support’. Support was classified as a score of four or more on the scale (i.e., Support/Strongly support). Across the whole sample, support was highest for sharing de-identified data with health-related government agencies (44.7%), but support was lower among priority populations; this was followed by support for sharing data for non-commercial research by a university or research institute (43.7%), which was generally higher among priority populations. Support for de-identified data being used for research by commercial and non-health-related government agencies was generally low. All aggregated data for these measures can be found in Table F6.