Introduction
Methods
Eligibility criteria
Sources and search strategy
Data management
Selection process, data collection and data analysis
Qualitative methodology
Sl | Category | N |
---|---|---|
A. Rationale for research using EHRs | ||
1 | Public interest | 43 |
2 | Value of EHR research | 36 |
3 | Justice of the healthcare system | 8 |
4 | Private sector profits | 7 |
B. Factors affecting research use of EHRs | ||
5 | Obstructive regulations | 31 |
6 | Regulatory facilitations and institutional support | 27 |
7 | Technical difficulties with EHR database implementation | 25 |
8 | Factors hindering informing participants about research and obtaining consent | 22 |
9 | Public awareness, experience, opinions and attitudes | 24 |
10 | Researchers' attitude | 3 |
C. Data management | ||
11 | Safety and security | 36 |
12 | Levels of data identifiability and research | 41 |
13 | Special types of data and their protection | 13 |
14 | Data quality, quantity, and integrity | 26 |
15 | Data ownership, management, and curation | 15 |
16 | Meaningful data sharing | 19 |
17 | Data storage, extraction, and data transfer | 30 |
18 | Legitimacy of uses and users of EHR data | 22 |
D. Impact of digitalization on healthcare system: providers' operations and patient engagement | ||
19 | EHR-based research and changed professional relationships within health care institutions and between various institutions | 19 |
20 | EHR-based research and the practice-patient relationship | 11 |
21 | Medical staff and ethical responsibility in the context of EHR and EHR-based research | 17 |
22 | Additional workload for staff and patients | 12 |
23 | Pivotal role of patients’ contribution | 18 |
24 | Ideas for the future: digital patient-citizenship | 14 |
E. Risks, harms and burdens of research with EHRs | ||
25 | Risk to privacy | 39 |
26 | Compromised patients' autonomy | 21 |
27 | Dignitary harm—being wronged | 9 |
28 | Harmful or wrongful use of data | 11 |
29 | Legal harm | 2 |
30 | Psychological harm | 3 |
31 | Information risk | 13 |
32 | Risk of exploitation | 12 |
33 | Undue pressure to participate | 2 |
34 | Undermined trust | 14 |
35 | Compromised care | 5 |
36 | Group mediated risks | 15 |
37 | Financial conflict of interest | 9 |
38 | EHRs as a challenge and potentially negative influence on healthcare practices, staff, and patient-provider communication | 13 |
39 | Risks to healthcare provider | 3 |
F. Measures for subject protection | ||
40 | Independent review | 26 |
41 | Informed consent or authorization of data use | 43 |
42 | Legal, ethical and professional regulations | 19 |
43 | Risk assessment and risk minimization, non-maleficence | 14 |
44 | Primary care provider consent | 2 |
45 | Community/panel of patients consent | 2 |
46 | Provider consent mechanism | 2 |
G. Types of consent | ||
47 | Initiation of contact | 4 |
48 | Options on the continuum from no consent via opt-out, to explicit and informed consent | 27 |
49 | Verbal consent, written consent | 14 |
50 | Broad—narrow consent | 8 |
51 | Interactive consent and granularity | 16 |
52 | Retrospective consent | 6 |
53 | Proxy consent | 4 |
54 | Assent | 2 |
H. Content of consent | ||
55 | Data management: purpose, future use, storage and data sharing | 15 |
56 | Security measures | 12 |
57 | Benefits, risk and burdens | 10 |
58 | Commercial application of data | 9 |
59 | Communication of results | 4 |
60 | Sources of research funding | 2 |
I. Reasons and motives for participation in EHRs-based research | ||
61 | Benefits of making one's EHR available | 7 |
62 | Moral values and obligations | 13 |
63 | Trust in people and institutions inviting one to participate in EHR-based research | 13 |
64 | Personal cirumstances and characteristics | 14 |
J. Emotions experienced as a result of reflection on EHRs and/or participation in EHR-based research | ||
65 | Emotions indicating respondents' positive attitude towards EHR-based research | 9 |
66 | Emotions indicating respondents' negative attitude towards or doubts about EHR-based research | 12 |
67 | Emotional spectrum and ambivalence | 5 |
K. Ethical values, rights and obligations | ||
68 | Autonomy, control over personal data, dignity, confidentiality and privacy | 49 |
69 | Information, public education and engagement | 22 |
70 | Principle of beneficence | 17 |
71 | Principle of justice | 12 |
72 | Research integrity | 5 |
73 | Optimal health care and clinical judgment | 3 |
74 | Right to compensation | 1 |
First author, year | Journal | Country | Number of subjects, subjects characteristics | Research design | Main findings |
---|---|---|---|---|---|
Botkin et al. (2014) | Journal of Community Genetics | USA | 131, > 18, General population, | Qualitative study, focus groups—semi-structured discussion following the video on EHRs research | Initially the majority of participants did not know that EHRs were used for the purpose of research. When participants of the study were presented with educational materials by researchers, they expressed strong support towards research. Also the majority of participants preferred opt-out form of consent to research, referring to public benefits as justification. Nevertheless, some participants wanted to give informed consent before their EHRs are used in research. The participants stress the need for proper public education about the EHRs research |
Clerkin et al. (2012) | Family Practice—The International Journal for Research in Primary Care | Ireland | 35, > 18, patient in primary care setting, | Qualitative study, focus groups | Patients supported the idea of their records being used in research for the greater good. However patients are concerned about possible risk to their privacy and harmful effects, such as loss of job or insurance (mainly men) or social discomfort and embarrassment (women). Patients want to control their records, and they are willing to give their broad consent to a specific, chosen type of research; then they do not have to be informed on a study-by-study basis (broad consent) |
Geissbuhler (2013) | International Journal of Medical Informatics | 21 European countries | > 100 Participants of 2012 European Summit on Trustworthy Reuse of Health Data: decision makers, scholars, representatives of patients, industry, and European Commission | Participatory research: discussing scenarios in breakout sessions | Benefits of reuse of data outweigh possible risks. The most important challenge is to build a trustworthy framework for reuse of electronic health records. In such a framework involvement of citizens plays the essential role and they deserve to be “fully informed”. Governments should regulate and oversee reuse of patients’ data. Lack of a unified framework for reuse of electronic data in the EU entails increasing costs of research, missing economic opportunities for European countries (pharmaceuticals, health technology and devices, eHealth solutions), and undermines patients’ safety |
Grande et al. (2013) | JAMA Internal Medicine | USA | 3064, > 18, representative online panel | Quantitative study, Internet survey, experimental design using scenarios, likert-scale, check-box | Among the factors influencing participants’ willingness to share health information the purpose of use is more significant (63.4% importance weigh) than the type of data user (university hospital/drug company/public health department; 32,6% importance weigh) and the level of sensitivity of data (low sensitivity “medical history”, high sensitivity: “personal genetic test predicting risk of cancer”; 3.1% importance weigh). People favor research use of data over healthcare quality studies and marketing. People trust university hospitals more than the government, drug and insurance companies. Social and racial factors are associated with trust: non-Whites are more willing to share their health information for non-research purposes and those who do not have secure access to health care are less supporting of secondary use of data |
Grande et al. (2014) | Annals of Internal Medicine | USA | 3064, > 18, representative online panel | Quantitative study, internet survey, experimental design using scenarios, likert-scale, check-box | Respondents agree that the use of personal information for the sake of research is appropriate, but they have ambivalent feelings about research using personal data without consent. Participants rate the acceptability of use of EHRs for the purpose of research much higher than for the purpose of marketing. Results of the study indicate that the purpose of research using EHRs is more important than informed consent, because participants see unconsented research use as more appropriate than consented marketing use (5.65 vs. 4.52; difference = 1.13; 95% CI 0.87, 1.39, using the Likert scale: 1 = not at all appropriate, 10 = very appropriate) |
Jones et al. (2017) | Journal of Oncology Practice | USA | 32, > 18, cancer patients | Qualitative study, semi-structured in-depth interviews | Cancer patients felt uncomfortable with a permissive system that offers fewer notifications about when and how their data are used (care, research, marketing), by what kind of user (physician, researcher, insurance company), and if sensitive data are included. They preferred opt-in form of consent, as well as they favored stringent oversight especially over the use of data by insurance and pharmaceutical industries. They were more comfortable (more trustful) with the use of their data by physicians and researchers |
Kim et al. (2017) | BMC Medical Ethics | USA | 800, > 18 not fully representative sample | Quantitative survey, phone interview, questionnaire, likert-scale | The majority of respondents declare willingness to consent to sharing their medical information for the sake of research (74,8%), and a slight majority values more research benefits than privacy (50.6–46.8%). However, the majority of participants value control over their EHRs more than research benefits (69.8–26.9%) but at the same time they want to have control over their medical records. Participants' willingness to share their date for the sake of research is linked with education (higher level of education is positively correlated with higher likelihood of consent), race (Hispanic, black, Asian and others were less likely to consent than Whites), familiarity with EHR (those who already have EHR are more likely to consent) and belief that EHR improve quality of healthcare (those who believe in a link between EHR research use and quality of healthcare are more willing). Health status, as well as previous experience with technology are not predictors of consent |
King et al. (2012) | International Journal of Medical Informatics | Australia | 700, > 18, representative group | Qualitative survey, phone interview, questionnaire, likert-scale | Most respondents declare that research is important and are willing to share their health information. The majority of participants believe that their consent should be sought for use of their health information for any purpose other than medical treatment (92%). Majority of respondents declare concerns over privacy (66%) and only a slight minority of those who are concerned feel reassured when extra security measures are put in place (16%). Information that EHRs are de-identified reduces concern about consent. However, many respondents are not aware of a risk of re-identification. Younger, 18–19 y (51.4%) and older, 60 + years old (53.4%) respondents less frequently declare being concerned about privacy than those in between age groups 20–59 years old (73.3%) |
23, > 18, four groups: elderly patients, frequent receivers of health care; persons with higher education; persons “with different ethnic background”, a mixed group, > 18 | Qualitative study, focus groups | Participants expressed a wide range of attitudes towards EHR's secondary use: from approval and acceptance to privacy concerns. Majority of participants thought that medical research is important, although some were afraid that it may compromise their privacy. Also their view on consent requirement was diverse: some claimed that consent is not necessary, others required consent and very detailed information. Participants expected respect, they wanted to control their data, and to have the ability to track the users, in case something goes wrong | |||
Nair et al. (2004) | Journal of Health Services Research & Policy | Canada | 17, > 18 patients in primary care setting, | Qualitative study, semi-structured interviews | Patients generally did not know that their records were already used in research. Majority of patients (13) wanted to be asked for consent and be informed about the details of the study. Patients thought that control over their records and proper communication between them and their healthcare provider is crucial for being treated respectfully, not merely as an object of research. Patients recognized both the value of research and the need for balancing their consent preferences with time pressures in the clinical encounter, nevertheless they were concerned about the privacy of their records. They thought that health data should not be tradeable, especially to pharmaceutical and insurance companies |
Riordan et al. (2015) | International Journal of Medical Informatics | UK | 3157, > 18 | Questionnaire: yes/no, check-box | People want to be asked for consent before their EHR is accessed for research with identifiable data purposes (91%). A slight majority wants to be asked even before research use of their anonymized data (51%). 41% of the respondents were not aware of EHRs’ existence prior to taking part in the study. Support for explicit consent is associated with ethnicity and nationality (non-White and non-British), lower computer skills, poorer knowledge about the system and lower level of education |
Stevenson et al. (2012) | Family Practice—The International Journal for Research in Primary Care | UK | 50, > 18 patients, in Health Research Support Service Study; 6 focus groups, and 17 interviews with patients, and 6 interviews with staff | Qualitative study, focus groups and interviews | Patients and staff generally support research using EHRs, however opt-out type of participation elicited more split responses from both participants and staff. Patients and staff appreciated the benefits of the opt-out option (inclusiveness and effectiveness of research), although they were concerned whether it gives a fair opportunity to consent for participation. Patients expressed their concerns about the safety of data and possible misuse of data by private companies. Trust in healthcare institutions was an important factor for participation |
Stevenson (2015) | BMC Health Services Research | UK | 68, Patients—approached in Health Research Support Service Study (50), stakeholders (11), staff (7) | Qualitative study, focus groups and interviews | This is a qualitative evaluation of a pilot Health Research Support Service (HRSS) study, designed to assess feasibility of the Clinical Practice Research Datalink—the NHS (England) observational data and interventional research service. Topics emerged from thematic analysis were subsumed under four broader categories: understanding of the study, commitment and engagement, ‘work’ involved in participation, reflections on the process. Patients and staff had problems with understanding the study, e.g. patients did not know that they were not selected and that their action was required to opt-out. Patients and staff did not know that data will not be anonymized prior to leaving the practice system. Patients overlooked or did not pay attention to the information letter about the study, staff did not pay attention during the staff meetings. Patients and staff were committed to the study not due to engagement with HRSS, but because they trust NHS or their provider. Patients and staff acknowledged the social value of research, but were concerned about the fact that identifiable data are used and opt-out option is used as a proxy for consent. Stakeholder saw the value of the project for safety, efficacy and advances in medical sciences |
Weitzman et al. (2010) | Journal of Medical Internet Research | USA | 151, > 18 | Quantitative survey, Internet survey, questionnaire, check-box, Likert-scale | Users of Personal EHRs are not fully aware of the possibility of granular sharing of health information. Personal EHR allows a consumer to determine restrictions on sharing their record in terms of content (e.g. making available only some specific elements of a PHER) and time. Participants declared that they would be more willing to share their data if they were given the option to share it in a granular manner. Almost all respondents are willing to share their health information for the purpose of research (91%). Majority want to strictly control access to their records and want an opt-in form of consent for research (59%) |
30, > 18, patients, early users of personal EHRs, | Qualitative study, semi-structured focus groups and interviews | There was no clear preference for either opt-in or opt-out as a form of inclusion into research. Participants expressed their demand for information about and explanation of risk and benefits involved in research. Participants were willing to put some limits on information: time limits for use, or the limited scope (granularity). Guaranteeing privacy but not anonymity decreases willingness to share, however security audits increased trust and willingness to share data. Emergency research also increases willingness to share data | |||
Weitzman et al. (2012) | BMC Medical Informatics and Decision Making | USA | 261, > 18 or parents of patients, Personally controlled health record users, | Internet survey, questionnaire, check-box, open-ended question | People want to have personal EHRs and they like to have control over and access to their health information although they have limited knowledge about personal EHRs’ functioning. Respondents declare that they see value in EHRs in seeing records at any time (93.8%), monitoring health problems (68.9%), sending EHRs to other institutions (51.5%), participating in research (26.6%), participating in health/wellness activities (21.6%). They are more willing to share their information with public health authorities (63.3%) than with other providers (54.1%). The respondents are afraid of discrimination, lack of anonymity and government insensitivity. They do not want to share their sensitive (esp. financial, family and mental health status) information |
Willison et al. (2008) | Journal of the American Medical Informatics Association | Canada | 1230, > 18, a sample with different characteristic than a general population | Phone interview, scenarios, likert-scale | Most people support the use of EHRs and agree that beneficial research is more important than privacy. Respondents believe that “it is okay” to use EHR for the purpose of marketing campaigns (35%), to produce also profitable products (60%), improve quality of care (86%), track communicable diseases (89%). Majority want to be asked for their consent for access to health information: every time an EHR is accessed (32%), regularly renewed general permission (23%), notification (24%), and access without consent (12%). Among those who support the use of EHR, the minority (27%) think that consent is not necessary for automated extraction of data. The respondents declare trust in data institutes, hospitals, disease foundations, and university researchers, but tend not to trust insurance companies. Government and drug companies are trusted by half of respondents |