|Author, location, date of data collection||Methodology, sampling, analysis||Health condition/setting||No. of participants (N)||Participant demographics n (%)||Outcomes, result(s)||QualSyst score|
|, Canada, October 2015 and November 2016||Survey, mean (SD or IQR) and Fisher exact test||General medicine /cardiology patients||590||
65, IQR 8
27 (4.6), primary school
101 (17.1), did not complete HS
144 (24.4), HS
77 (13), professional training
142 (24), University or college
77 (13), Masters or PhD
23 (3.8), prefer not to disclose
535 (91.6), Caucasian
15 (3.6), Black & Caribbean
3 (0.5), First nations
12 (2.0), Asian
5 (0.8), Other
17 (2.9), prefer not to disclose
97.4% allowed researchers access to their record in the administrative database for the purposes of conducting a clinical trial.|
75.3% would allow the use of their initials and DOB to link administrative record to their medical charts.
53.7 preferred the use of administrative records for follow-up care compared with 30.9% for face-to-face appointments.
|, USA, not reported||Survey, not reported||General medicine/patients with asthma or stable angina||3429||No reported||
335 (9.8) refused consent to data collection from their medical records for research.|
Information relating to stable angina or asthma was not considered particularly sensitive.
Consent rates were not dissimilar by disease, sex, mean age, severity of symptoms, or satisfaction with care.
|, USA, not reported||Survey, logistic regression||General medicine/general public||2271||Not reported||
96.3% consented to data linkage. Only two variables were significant: government insurance (OR = .332[.442–.775], p = .011) and health literacy (OR = 1.35[1.09–1.67], p = .006).|
Those with government insurance were less likely to consent to data linkage of survey data with clinical data.
|, USA, not reported||Survey or telephone interview, logistic regression||General medicine/general public||1574||
232 (15), 18– < 35
512 (33), 35– < 50
465 (30), 50– < 65
365 (23), 65+
86.2% of respondents authorised data linkage of survey and health administrative data.|
Non-authorisers (n = 217) were significantly younger (born between 1975 and 1993), had less comorbidities, and were less likely to have visited clinic (< 3 visits) in one year.
|, Australia, April and June 2016||Survey, absolute numbers and frequency, Fisher exact test and ordinal logistic regression||General medicine/hospitalised patients||919||
177 (19), 18–30
287 (31), 31–50
261 (28), 51–70
158 (17), 71–85
36 (4), > 85
180 (20), did not complete HS
266 (29), HS
192 (21), vocational
206 (22), bachelor’s
75 (8), master’s/doctoral
Primary language, English
421 (46) assumed that their data was already used for medical research (without explicit consent).|
810 (88) approved or were indifferent to their medical data used for treatment purposes to be reused for research purposes.
800 (87), found opt-out consent acceptable or were indifferent.
Preferred consent models
• 494 (54), opt-out
• 172 (19), no preference
• 253 (28), opt-in
• Sex, employment status, and cognitive status did not impact the acceptability of opt-out consent.
• Opt-out had higher rates of support in refugee (93%) and those who were illiterate in their primary language (92%).
To withdraw consent: 431 (40) preferred a telephone voicemail message and 330 (30) preferred email correspondence.
|, Britain, October to December 2004||Questionnaire, raw figures with percentages||Recently discharged medical and surgical patients/hospital||166||Not reported||
21% definitely wanted or preferred to be asked for permission to use anonymised medical history and 20% for reasons for treatment.|
The proposed use of information (e.g., research, audit) made little difference to the overall approval percentages (range 10–12%).
13% definitely wanted to be asked for permission for use of information; most had no preference or preferred not to be asked.
The most popular method for asking was signing a form while in hospital, rather than a request later. Being informed each time material was used, but not being asked for your permission, was the second most popular option.
• 7% and 5% wanted to be asked permission when their age or gender was required respectively. 14% wanted to be asked when sexual orientation was asked.
• 10% wanted to be asked if information was to be used for clinical audit.
• ~ 10% preferred to be asked for permission when information was intended to help future patients or to teach healthcare professionals.
• 12% wanted or preferred to be asked permission when being used for publication in medical journals.
|, Europe, March to May 2018||Survey, descriptive statistics, and Chi squared test for independence||Rare diseases/patients and their families with rare diseases||2013||
2 (0), 15–17
70 (3), 18–24
293 (15), 25–34
852 (42), 35–49
644 (32), 50–64
152 (8), ≥ 65
1775 (88), EU
238 (12), non-EU
1909 (95), diagnosed
104 (5), undiagnosed
Number of rare disease (n = 1909)
1664 (87), 1
174 (9), 2
44 (2), 3
13 (1), 4
14 (1), ≥5
Reasons for consenting to research using respondent’s health data included: development of new treatments (97%), better understanding of their disease (97%), improved diagnosis (97%), receiving additional specialist advice (95%), and improved research and care for their disease (90%).|
• Only 51% consented when data was to be used for research not related to the medical field.
Respondents over 65 years of age were less willing to consent to share data for research on other diseases (84%).
Respondents with lower levels of education were more willing to share (59%) compared with those with higher levels of education (48%).
Disease severity—those with more severe rare disease were more likely to share data (64%) compared with those with less severe disease (40%).
Incentives to increase participation included: more information about the disease, be informed of the results of research, capacity to withdraw from the research, and being informed of any data breaches.
Perceived sensitivity of data (disability, genetic information, physiological data, symptoms, name of disease) all included the rates of consent.
• 51% believe that information on disability was sensitive, the lowest rate was seen for the name of the disease (33%).
Willingness to share data is not precluded by wanting to control access.
• Only 1% did not request any control over their data; 47% wanted full control on access.
• Women were more likely to request control (48%) compared with men (42%).
Reasons for not consenting included: fear of discrimination (34%), fraud (32%), concerns about personal safety (20%), use of data without knowing (35%), and used in a context that they do not approve (47%).
• 50% of respondents wanted control to restrict access to third parties.
49% were willing to allow an ethics committee to decide about access on their behalf; 43% were against this.
Communication is essential and should include how the data is to be used, and the results of this research.
|, UK, January 2009 to December 2010||Panel survey, descriptive analysis, t-test, logistic regression||General medicine/general public||50,994||
6897 (16.3), 16–24^
7495 (17.7), 25–34
8300 (19.6), 35–44
7846 (18.6), 45–54
6809 (16.1), 55–64
4959 (11.7), 64–74
9819 (23.2), none
18, 744 (44.3), secondary level
13,745 (32.5), higher/degree
70.7% of respondents consented to record linkage.|
75% of respondents age 16–24 years consented.
• Similar proportions 71% (males) and 70% (females) consented to data linkage.
• Younger age, marriage, employment status, car ownership, and education were all significantly associated with consent.
Small increases in consent rates in individuals with poor health (self-assessed) [adjusted OR 51.11; 95% CIs: 1.06, 1.16], less so when defined by General Health Questionnaire score (adjusted OR 5 1.05; 95% CIs: 1.00, 1.10).
Larger differences were observed for those of non-white ethnicity who were 38% less likely to consent (adjusted OR 50.62; 95%CIs: 0.59, 0.66).
Consent was higher in Scotland than England (adjusted OR 5 1.17; 95% CIs: 1.06, 1.29) but lower in Northern Ireland (adjusted OR 5 0.56; 95% CIs: 0.50, 0.63).
|, Europe, 2012||Questionnaire, results expressed in percentages||Leukodystrophy/leukodystrophy patients and family||195 (149, family; 46, patients)||
121 (62), 40–64
130 (66.7), France
24 (12.3), Italy
9 (4.6), Belgium
6 (3.1), Spain
26 (13.3), Germany
Patients and their families were strongly in favour of participating in the registry.|
A process combining initial broad consent with ongoing information was appropriate.
Information should be given to the patient at initial consent.
|, UK, 1996–2000||Survey (seven), random effects model||Epidemiological research/general public||27,797||
2396 (8.6), 18–29
1792 (6.4), 30–39
2317 (8.3), 40–49
7714 (27.6), 50–59
6824 (24.5), 60–69
5116 (18.4), 70–79
1838 (6.6), 80+
Responders < 50 years of age 75–95% gave consent to follow-up.|
This fell in older people, particularly females. In one study consent fell to < 50% in women ≥ 80 years of age.
Males, younger people, and subjects reporting the symptom under investigation were more likely to give consent.
• Responders who had the symptom being investigated had ~ 1.5 times the odds of consenting to review of their medical records as those not reporting symptoms.
|, Finland, not reported||Survey, descriptive statistics||General medicine/general public||418||
44 (10.5), ≤ 30
47 (11.2), 31–40
57 (13.6), 41–50
88 (21.1), 51–60
126 (30.1), 61–70
49 (11.7), > 70
7 (1.7), missing
74 (17.7), PS
44 (10.5), SS
165 (39.5), HS
80 (19.1), University of Applied Sciences
or bachelor’s degree
52 (12.4), Master’s degree or higher
3 (0.7), missing
83% had a positive or very positive opinion about health research in general.|
49% (positive) or 19% (very positive) opinion about using administrative health registries for research purposes. 3% had a negative or very negative opinion.
61% were willing to provide identifiable health information for research.
• 8% would like to be informed when their information is used.
• 86% favoured broader consent methods - one consent covering a certain register or a research topic.
• 55% required ethical evaluation from register-based research addressing a sensitive issue.
When asked about Finnish legislation, 35% of the study population wanted to tighten the law for some parts, 28% were satisfied with the current practices, and 25% wanted to liberalize the law to advance scientific research.
Supported research purposes for health registries included:
• 76% aetiologic studies.
• 60% disease monitoring.
• 53% assessing the effectiveness of health care.
• 38% any research use.
• 48% (positive) and 12% (very positive) towards the use of their own health information in register-based research.
• Respondents with higher education were more often inclined than those with a lower education level to allow. Of the participants with the highest education level, 82% had a positive or very positive opinion compared with participants with the lowest education level (41%) (p < 0.001).
Use of hospital medical records
• 75% would like to be informed of the possible research use of their medical records on admission to hospital.
• 52% would like to be able to limit research use of their medical records.
• 11% forbid it altogether.
• 29% thought that everyone’s medical records should be accessible for research.
• 32% would like to be able to limit linkage of certain registries.
• 34% felt that everyone’s information should be available for researchers in every national register.
• 15% would forbid record linkage altogether.
• 30% indicated no need for informed consent in register-based research.
• 30% similar proportion felt consent should be obtained for every study.
• 39% thought consent necessary, in some situations, such as studies addressing a sensitive study topic.
• Men found informed consent unnecessary more often than women (37 % vs. 27 %), whereas women more frequently thought that in special cases informed consent should be required (44 % vs. 31 %).
|, Canada, August 2014 and May 2015||Survey, descriptive statistics, and Student’s t-tests and nonparametric tests||Cancer/outpatient clinic||569||
109 (19.2), breast
86 (15.1), gastrointestinal
83 (14.6) genitourinary
70 (12.3), thoracic
73 (12.8), hematologic
73 (12.8), head and neck
63 (11.1), gynaecologic
12 (2.1), other
Clinical trial participation, yes
346 (60.8), university, college, professional
39 (6.9), vocational, technical, diploma
169 (29.7), elementary, HS
15 (2.6), prefer not to answer or missing
452 (79.4), white
68 (12), Asian
49 (8.6), other/prefer not to answer, missing
93%, (cohort 1) would allow long-term access to their information and allow personal information to be used to match clinical trial with administrative data.|
68% (cohort 2) preferred to make additional clinical information available through linkage with administrative databases.
9% preferred to have no further information made available to researchers.
No significant differences were found in the subset of patients who were part of a clinical trial compared with those who had never participated (p = 0.65).
Canadian Primary Care Sentinel Surveillance Network has established policy governing the protection of privacy and use of health information for research where no individual patient consent is required.
|, Taiwan, 2001||Survey, multiple logic regression||General medicine/general population||14,611||
1769 (12.1), 20–24
3263 (22.3), 25–34
3436 (23.5), 35–44
2782 (19), 45–54
1606 (11), 55–64
1214 (8.3), 65–74
541 (3.7), 75+
3776 (25.8), ≤ college
4366 (29.9), senior high
2266 (15.5), junior high
2942 (20.1), elementary
1261 (8.3), illiterate
2911 (88%) gave consent to link their questionnaire to their NHI records.|
Age over ≤ 65 years, married, illiterate, those with a monthly household income < 30,000 New Taiwan (NT) dollars, or were living in a suburban area were less likely to consent.
• Non-consenters had relatively lower mean scores in all eight physical and mental functional status domains of SF-36.
No difference in gender and self-reported health was between individuals who consented and those who refused was noted.
|, UK, September to December 2009||Survey, bivariate associations and estimate multivariate logistic regression||General medicine/BHPS Wave 18||13,454||
1965 (14.6), 16–24^
3200 (23.8), 25–39
2546 (18.9), 40–49
1984 (14.7), 50–59
3758 (27.9), ≥ 60
6633 (49.3), England
2357 (17.5), Wales
2282 (17), Scotland
2182 (16.2), Northern Ireland
425 (3.2), higher degree
1593 (11.8), first degree
4220 (31.4), diploma in higher education
1711 (12.7), A-levels
2130 (15.8), O-level or equivalent
663 (4.9), GCSE
67 (0.5), Commercial qualification, no O-level/GCSE
2395 (17.8), none/still at school
12,077 (89.8), British/Irish White
1377 (10.2), other
Health problems related to: (note, not all reported)
1793 (13.2), chest
2528 (18.8), heart
1046 (7.8), stomach
642 (4.8), diabetes
1119 (8.3), anxiety
208 (1.5), cancer
Note: Additional data taken from publication supplementary tables|
41% consented to health data linkage.
• 99% of those who gave consent to link to health data to the BHPS also gave consent to link to the NHS Central Register.
Consent was significantly higher among people who live in:
• England (42.4%)
• Participants aged 16–24 (45.9%)
• Participants who considered their ethnicity to be British/Irish White (42.4%).
Consent by education (two highest, two lowest):
• 49.2%, higher degree.
• 45.6% A level.
• 36.8%, none/still at school.
• 21.7%, commercial qualification, no O-level/GCSE.
Consent was not affected by socio-economic or health characteristics.
Patients with cancer consented 46.2% of the time.
Difference by sex: Male, 42%; Female, 40.7%.
Recent users of GP services were underrepresented among consenters.
Self-reported health was generally not associated with consent. There are two exceptions; diabetes and obesity which were associated with rates of consent.
|, USA, not reported||Survey, percentage of respondents||General medicine/patients with DM and congestive heart failure||4647||
Participants at sites where no advanced permission required reported (n = 1174)|
587 (50), congestive heart failure
Type of review required:|
• 10 sites required full IRB review prior to patient contact.
• 1 site had an expedited IRB review.
• 4 sites deferred to the IRB at RAND Health.
Type of consent required:
• 8 sites required no advance permission to contact the potential participants for a telephone survey.
• 5 sites required oral permission.
• 2 sites required written advance permission.
Overall, the contact rates and eligibility rates were similar across different sites.
Response rate varied based on type of consent process; this was highest for sites requiring no advanced permission to contact potential study participants.
Sites without advance permission requirements
• 85% of eligible participants consented to the telephone survey.
• Participation rates were highest for sites not requiring advanced permission.
• 58% of participants at these sites completed the telephone survey compared with 39% at sites requiring oral advanced permission and 27% from sites requiring advanced written permission.
Sites requiring permission
• The sites with written advanced permission had the lowest overall cooperation rate of 39%, with only 43% of potential participants providing permission for researchers to access their contact details.
Patients with congestive heart failure were slightly more likely to complete the survey than individuals with diabetes.
|, Hong Kong, not reported||RCT nested within a cohort, chi-square test, multivariable logistic regression, likelihood ratio test||General medicine/subsample of the FAMILY cohort||1200||
94 (7.8), 18–29
197 (16.4), 30–44
423 (35.3), 45–59
307 (25.6), 60–74
179 (14.9), ≥ 75
436 (36.3), primary
587 (48.9), secondary
177 (14.8), tertiary
33.3% of respondents returned signed consent forms.|
Subgroup analyses found requesting HKID significantly reduced consent among adults aged 18–44 years of age (OR 0.53, 95% CI 0.30–0.94, compared with no request).
Souvenir incentives increased consent among women (OR1.55, 95%CI 1.13–2.11, compared with no souvenirs), but no overall effects were noted. Younger people also had increased rates of consent with a souvenir.
Higher income and older age were associated with health record linkage.
The request for a unique personal identifier did not substantially reduce consent proportions.
|, Canada, November 2003||Survey, descriptive statistics, multiple logistic regression.||AIDS, MS, mental health/outpatients||235||
68 (28.9), 20–39
129 (54.9), 40–59
34 (14.5), ≥ 60
4 (1.7), unknown
39 (16.6), < grade 12
35 (14.9), grade 12
156 (66.4), attended/finished post-secondary
5 (2.1), no answer
Previous experience with medical research, yes
Respondents believed that the physician involved in their care should be the only ones who should be able to access their health data without consent.|
• Over 78% of respondents believed that consent was required when the data is identifiable.
• Over 84% of respondents believed that consent was required when the data is of a sensitive nature.
• 17% were not sure about seeking consent when it was not feasible.
• 63% believed that consent was not necessary when the information was anonymous.
Access to health information without consent:
• 33.2%, medical researchers.
• 4.3%, drug companies.
• 4.3%, employer.
• 8.1%, insurance company.
• 6%, the government.
Overall, 78.3% were advocates for consent. The difference in patient groups was not significant.
• Sex and employment status were the only factors to predict consent. Women (OR = 1.96, 95% CI: 1.04–3.71) and those who were employed (OR = 2.29, 95% CI: 1.00–5.25) were both advocates for consent.
• 77.4% agreed/strongly agreed that researchers should get consent when the data was identifiable.
• 22.5% agreed/strongly agreed with one of consent for all future medical research.
• 38.3% agreed/strongly agreed that people should be informed about their health information being used (no additional consent required).
• 77.2% agreed/strongly agreed that consent is required for each new project.
• 34% agreed/strongly agreed that researchers should be able to use un-identifiable information without consent.
|, Northern Ireland, September to December 2015||NILT survey, univariate and multivariate analyses||General medicine/general public||1202||
Results of weighted demographics|
144 (12), 18–24
175 (14.6), 25–34
172 (14.3), 35–44
214 (17.8), 45–54
180 (15.0), 55–64
310 (25.8), ≥ 65
7 (0.6), not answered/refused
224 (18.6), no qualification
555 (46.2), school level
369 (30.7), graduate level
54 (4.5), not answered/refused
Respondents believed that academic researchers should be allowed to use data that has been linked by a third party.|
• When the data is linked by the NHS—88% (males) and 86% (females) responded defiantly or probably should be allowed.
• 10% had problems with NHS linking data.
• When the linking was undertaken by researchers and the data included postcodes—63% (males) and 58% (females) responded defiantly or probably should be allowed.
• When the linking of data is undertaken by people who keep hospital records and linked to school records—74% (males) and 71% (females) responded yes (probably or defiantly).
Those with a long-term health condition were more likely to be against data sharing (24–25%) compared with those with no long-term health condition (20%).
Respondents were divided on the issue of consent for linked data analysis.
• 30% believed that it was not necessary to ask for consent to link data if individuals will not be identified.
• 34% believed that consent should always be sought but ‘if the difficulties are too great, important research should not have to be abandoned for this reason’.
• 31% of respondents believed that data should only be used with individual patient consent before linking the data with anything else. If consent was not possible this may mean not undertaking the research.
• By sex, 34% of males and 29% of females agree with this statement.
• Those with no qualifications were more likely not to allow research without consent (38%, no qualification to 26% with graduate=level qualifications).
• 30% of respondents with no health condition, 15% of respondents with a health condition (ADLs not affected), and 42% (ADLs reduced a little) and 40% (ADLs reduced a lot) would require consent.
• 5% did not know if consent was needed to link data.
Connection between the requirement to get consent to use data and trust in the organisation to secure the data.
• Those who felt that research should not be done without consent, were more likely to state that they did not trust the NHS (20% vs 9%), the government (35% vs 20%), GP practices (11% vs 5%), commercial organisations (63% vs 54%) and academics (29% vs 19%).
|, UK, 2009||BHPS Wave 18, multivariate bivariate probit models||General medicine/BHPS Wave 18 participants||6433||Not reported||
In the UK consent must be obtained to link administrative data.|
41% of respondents gave consent for health data linkage; 32% consented to the linkage of benefit (economic) records; 39% agreed to link to education records.
Consent to data linkage was related to the respondent’s views on privacy and community-mindedness.
• Males were more likely to consent to health and benefit data linkage.
• Respondents aged above 24 years of age were less likely to consent; however, this was not statistically significant.
• Socio-economic variables did not influence consent.
• Respondents with higher levels of education were more likely to consent.
• Refusal to answer questions on income from investment was a strong predictor of not giving consent.
• Undertaking voluntary work was associated with giving consent for health records access.
The probability to consent is associated with the number of household members who have already consented to health and benefit data linkage.
|, UK, not reported||Postal survey, randomised factorial design||General medicine/general public||245||
Between 65 and 74
86.9% of respondents gave consent to access the medical records.|
The inclusion of an income question or seeking to access medical records did not reduce the response rates in older people.
• 75.8% of respondents answered the income question.
|, Australia, 1999||Survey (mail), chi squared with Yates correction||General medicine/university workers/those with upper body and neck disorders||292 (200—women employed by a university; 92—women with upper body and neck disorders)||
184 (63), ≤ 45
158 (54), secondary education
Participants were randomised 1:1 to receive a postal questionnaire plus or minus an authorisation form (the other group received the authorisation form later).|
38% (university workers) and 24% (patients) provided authorisation after receiving the authorisation form later, compared with 31% and 17% in those who received the form at the same time.
• Differences in authorisation rates may be due to the sensitive nature of data for the patient group.
A delay in seeking approval to link data was associated with improved survey response rates while not impacting on authorisation rates.
|, Germany, 2011 and 2014||Computer assisted personal interviews, binary logistic regression||General medicine/lidA study participants||4148||
Year of birth#|
1000 (24.1), low level
1767 (42.6), medium level
1381 (33.3), high level
Health status, subjective
2240 (54), very good to good
1336 (32.2), satisfactory
572 (13.8), less good to poor
4015 (96.8), born in Germany
718 (17.3), no disease
983 (23.7), 1 disease
917 (22.1), 2 diseases
1535 (37), ≥ 3 diseases
93.8% of respondents gave consent to link primary data with at least one other data source (administrative or claims data).|
• 0.8% (n = 30) SHI data only.
• 22.1% (n = 867) IAB data only.
• 77.1% (n = 3021) IAB and SHI data.
Respondents from the former West Germany (excluding West Berlin) were less likely to have their study data linked with both data sources or with IAB data only compared with respondents from the former East Germany and Berlin.
Those with technical or advanced technical school were less likely to agree to link their primary data with IAB data only than those with apprenticeship or school-based vocational training.
Those with no apprenticeship or with a non-regular apprenticeship were more likely to refuse consent completely.
Respondents who gave no information on income were more likely to refuse consent to both IAB and SHI data linkage. If they did give consent it was predominantly for IAB data only.
Respondents’ health status influenced levels of consent. Respondents with two or three or more diseases agreed more frequently to both data sources.
• Subjective health (SF-12v2) did not impact the levels of consent.
|, New Zealand, not reported||Survey, chi-square tests||General medicine/general public||203||
106 (56), 18–34
69 (37), 31–-60
14 (7), ≥ 61
146 (72), New Zealand Europeans
55 (27), Maori
Respondents were more willing to share their information if it was de-identified.|
• 60% of respondents had some concerns about sharing anonymous data with people other than HCPs.
Responses were influenced by the data recipient and the nature of the information.
Respondents were willing to consider data sharing all their health information with HCPs if consulted. Few refused to share their information.
• Over 50% of respondents did not wish to share data with government agencies or health insurers.
Respondents were less likely to share data of a sensitive nature, with those least involved in their care.
Implied consent was not always well-informed consent; patients should be made aware of the current data-sharing practices.
Use of a hybrid model of consent—general consent with specific denial may be appropriate to access data within the clinical setting, and general denial with specific consent to access data for other purposes.
• It is unclear if this consent would be for a single episode or more general access.
• The method for obtaining consent varies based on the proposed use of data.
|, Canada, March to April 2005||Survey, response frequencies||General medicine/general public||1230||
480 (39), 18–39
504 (41), 40–59
246 (20), ≥ 60
406 (33), HS or less
172 (14), some postsecondary
492 (40), completed postsecondary
123 (10), postgraduate or professional degree
Access to personal information for health research.|
• 4% of respondents did not believe that information in their paper medical records should be used for medical research.
• 32% believed that permission should be obtained for each use.
• 29% supported broad consent.
• 24% supported notification and opt-out consent.
• 11% believed that no need for notification or consent.
80% of those willing to give general permission wanted to be able to periodically review this decision.
• 36% wanted no or minimal involvement.
• 24% satisfied with notification
• Opportunity to opt-out 43% very important and 46% somewhat important.
• 12% acceptable to use with no notification or permission.
Respondents preferred data to be extracted by a nurse at the doctor’s office, but not the secretary. A research assistant from the university was also acceptable.
• 20% were not happy with this option.
70% of respondents supported the introduction of a common EMR.
• 9% believed that this data should not be used at all; this is higher than that for paper records (4%).
• 27% believed that EMR data could be used without permission (compared with 12% for paper records).
27% of respondents were against linking income data with EMR records.
• 16% supported notification
• 17% supported use without permission or notification.
|, USA, not reported||Survey, t test and multivariate logistic regression||General medicine/general public||1106||
(55), HS or higher education
67% of respondents provided consent to access data for health services research; 8% did not answer (passive non-consenter).|
• 25% actively refused.
Those who consented were older and included fewer women and African Americans than those who consented.
Difference between non-consenters and those who passively non-consenters were noted.
• Non-responders were older, less educated, lower income, and included more African Americans compared with those who actively refused consent.
Respondents who actively consented were younger, included more women and were more educated compared with those who consented.
Respondents who consented had a significantly lower PCS compared with those who did not consent.
• Those with the highest physical functioning scores were more likely to refuse consent (significant).
Non-consent rates were highest in respondents seeking: contraception (46.7%), treatment of urinary disorders (39.4%), uncomplicated DM (37.5%), headache (37%), and female genital disorders (35.4%).
Severity of illness did not impact the rates of consent between groups.
Respondents were less likely to give consent if they did not answer questions on smoking status, income, or functional status.
|, UK, not reported||Interviews, thematic analysis was undertaken using the Framework approach||General medicine/individuals included in the ALSPAC birth cohort study||55||
12 (21.8), 17^
35 (63.6), 18
8 (14.5), 19
7 (12.7), at university
25 (45.5), A-levels
8 (14.5), GCSE’s
12 (12.8), other
3 (5.45), none
51 (92.7), white British
3 (5.5), other
1 (1.8), refused
Health status, self-reported
9 (16.4), disability/long term illness
46 (83.6), no disability/long term illness
Some respondents were unsure of the effectiveness of data anonymisation and therefore did not believe that consent was unnecessary.|
Participants equated consent with opt-in and being asked if their data could be used for a specific study.
No consensus was reached for any of the scenarios for consent.
1. Linking teenage pregnancy data with state benefits:
• One suggested that this study not take place (n = 1).
• Unclear/unsure (n = 11).
• Request consent (n = 34).
• No consent required (n = 9).
2. Linking birthweight to future health outcomes:
• Unclear/unsure (n = 8).
• Request consent (n = 36).
• No consent required (n = 11).
3. Linking mental health records and criminal records:
• Unclear/unsure (n = 15).
• Request consent (n = 20).
• No consent required (n = 11).
4. Linking asthma and postcodes:
• Unclear/unsure (n = 15).
• Request consent (n = 26).
• No consent required (n = 14).
|, England, Wales and Scotland, March to April 2008||Face to face interviews, adjusted proportions||National cancer database/general public||2872||
1315 (46), 16–44^
997 (35), 45–64
564 (20), ≥ 65
542 (19), Degree or higher
1496 (52), Below degree
837 (29), No qualifications
Ever had cancer? No
Confusion by medical practitioners about the need for consent to use medical data for research. The Data Protection Act (1998) allows for the use of data for medical research without consent.|
82% had not heard of the registry, but 95% believed that the data it collects was important.
Conflict between the current Act and organisational policies and procedures which state that assumptions cannot be made about patients being willing to have their health data shared for purposes other than for direct patient care.
|, Belgium, February 2017||Interviews, deductive analysis using QUAGOL||Reuse of clinical trial samples and data/clinical trial participants||16||
35–79, mean 62, median 64
10 (62.5), higher education
6 (37.5), college or university
15 (93.5), Belgium
1 (6.25) Polish
4 (25), colorectal
3 (18.75), ovarian
1 (6.25), gastric and lung
1 (6.25), colorectal and lung
2 (12.5), pancreatic
2 (12.5), gastric
1 (6.25), cholangiocarcinoma
1 (6.25), unreported
Only results about data sharing are reported|
Data was seen by participants to be a similar resource to tissue samples; however, this position is not supported legally where the samples are not considered the same.
Respondent views varied on the need for re-consent prior to data access; a stratified approach may allow individual preferences to be met.
• Where data is re-used by the original research team, no re-consent was needed.
Respondents were supportive of an interactive consent tool where preferences could be individualised. This allows for greater control of their data.
Where data is shared with an ‘unknown’ group of researchers, some wanted to be involved by re-consenting. These respondents did not object to the idea of data sharing, rather they were concerned about data security and a lack of trust.
|, Scotland, May to June 2009||Focus groups, thematic analysis||General medicine/general public||19||
1 (5), < 60
15 (79), 60–74
3 (16), ≥ 75
Numbers taken part in medical research
Numbers with chronic health condition
Numbers with loyalty cards
Respondents expressed a positive attitude towards medical research.|
• The nature of the research and who was accessing the data were important.
• Respondents were surprised that anonymised data could be used in Scotland without consent.
Attitudes varied on the use of anonymised health data from medical records.
• A small number believed that consent was needed; this was related to the possibility that the data could never really be fully anonymised.
• This is even in the setting of a one-off consent for all future use.
• Most participants did not indicate the need to be informed of the data use.
The process of anonymisation does not necessarily exclude the need for consent.
Respondents recognised that re-consenting patients would be logistically difficult.
|, UK, not reported||Focus group, thematic analysis||General medicine/general public||19||
Age, mean (range)|
11 (58), employed
5 (26), retired
1 (5), unemployed seeking work
2 (11), unemployed due to illness or disability
Results of the qualitative focus group presented|
Respondents were positive about research; a few were surprised that this research using existing data is currently undertaken.
• All respondents would consent to a review of their medical records if asked.
Participants were divided equally when asked about the use of medical records without prior informed consent.
• Great good and public benefit versus best practice to ask.
By providing information on research bias and research processes respondents became more accepting of using medical data for research without consent.
• Participants were aware of difficulties of re-consent and potential low response rate which may bias findings.
• Some still believed that informed consent was necessary.
• Some curiosity about what research they are contributing to.
Those who wished to be informed about the research, opt-out consent was acceptable. This gives the individual the right to refuse but also informs them how the data is being used.
Anonymisation of data and data encryption were seen as safeguards to research with no prior consent.
• A minority suggested that there are no safeguards that made them happy with a no consent model.
• The role of ethics committees in patient protection was not included in respondents’ comments.
The use of data and who was accessing it was important to researchers.
|, USA, not reported||Focus groups, emergent content analysis||General medicine/general health||30||
4 (13, 31–40
4 (13), 41–50
8 (27), 51–60
4 (13), 61–70
6 (20), 71–80
1 (3), ≥ 80
11 (37), some HS
7 (23), HS
7 (23), some college
3 (10), college
4 (13), white
5 (17), black
20 (67), Latino
2 (7), other
|If no consent is sought, some believed that it would be an invasion of privacy to access health data. This extended to the secondary use of this data.||1|
|, USA, 2006 and 2008||Interview (telephone and enhanced face to face), multilevel random effects logistic regression||General medicine/health and retirement study||6384||
Age, by birth cohort|
747 (11.7), < 1923
428 (6.7), 1923–1930
3543 (55.5), 1931–1941
792 (12.4), 1942–1947
875 (13.7), 1948–1953
5235 (82.0), white
875 (13.7), black
275 (4.3), other
1481 (23.2), 0–11 years
2190 (34.3), 12 years
1334 (20.9), 13–15 years
1379 (21.6), ≥ 16 years
Overall consent rates were 67.8%.|
• Males (69.1%) were more likely to consent to data linkage than females (66.9%).
• White respondents were more likely to consent than other racial groups (69.1% versus 65.8%).
• Married respondents were more likely to consent than those separated/divorced (69.5% versus 64.9%).
• No differences by age cohort or level of education.
After controlling for variables only level of education was associated with consent. Respondents with a college qualification were more likely to consent compared with those with lower levels of education.
Privacy and confidentiality concerns impacted the likelihood to consent.
• Respondents who did not answer financial questions were less likely to provide consent.
Those who are resistant to interview were also less likely to refuse consent.
|, Australia, not reported||Interviews, framework approach||General medicine/general public||26||
Between 24 and 41
3 (12), ≤ Year 12
6 (23), TAFE
16 (62), tertiary
1 (4), post-graduate
Respondents were supportive of data linkage for health research, particularly when it will benefit society.|
• Participants recognised that people have the right to refrain from participating in research using their data.
Misconceptions about data linking were noted, with some believing that it is sharing of personal and health information within a healthcare system.
The assumption that opt-in and specific consent was not supported. Most participants believed that data linkage research could be undertaken without consent.
• Researchers should only be able to access de-identified data.
• Current protections are sufficient.
• Most preferred no consent over being informed of the intent to use data.
The assumption that participants would prefer to provide consent for both identifiable and non-identifiable data was not supported.
• Generally, most participants believed that consent is not required for data linkage.
• De-identified data should not be treated the same as identifiable data.
• Many believed that once identifiers were removed, the information became detached from the individual and was ‘just information’.
• Not all supported this belief with some still requiring consent.
• For most respondents, the focus was on data analysis, not the linking process.
In all scenarios presented, respondent s chose ‘no consent’ required.
• The linking of de-identified health data and criminal records by experts, while acceptable without consent, while acceptable, the respondents believe that patients with mental health issues should be allowed to provide consent.
• The linking of health, WorkCover and employment data linked by researchers was one scenario where consent and notification of research was required by the majority (n = 15). This was due to the researchers undertaking the data linkage, not an independent linking organisation.
Respondents shifted their views on consent based on the scenarios presented; moving from consent required to no consent.
Reasons for requiring no consent included: acceptable due to benefit, large dataset serves as protection, practical considerations, audit activities do not need consent, use of de-identified data does not breach privacy.
|, England, September 2015 to December 2017||Interviews and online survey, thematic analysis||Human Fertilisation and Embryology Authority registry/fertility clinic attendees||
40, online survey)
36 median, 30–46 range
16 (80), British white
14 (70), managerial or professional
2 (10), intermediate
3 (15) routine or manual
1 (5), student
14 (70) agreed to share data
2 (10) refused to share their data
3 (15) were unsure about sharing data
1 (5) agreed and disagreed with data sharing at different times
32 (80), agreed to share data
4 (10), refused to share their data
2 (5), were unsure about sharing data
2 (5), agreed and disagreed with data sharing at different times
• Consent forms are completed while patients are under stress, and consent for use of data is seen as a low priority.
• Perceived benefits and risks were associated with different levels of consent.
• Increased awareness of research using routine data may increase consent rates.
• Consent processes and rates vary by clinic; some patients when presented with lots of forms just tick ‘no’ to all questions.
• Language was seen as a barrier to consent in patients who need interpreters.
• Refusing consent if retrospective opt-out would be difficult.
• Providing education and information is important to increase rates of c consent.
• Information needs to clearly state the type (anonymous or identifiable) of data to be used.
• Respondents (n = 20) in the higher socio-economic groups (n = 15) 3 refused consent, 2 were unsure about what they had agreed.
• Respondents in lower socio-economic groups all consented to data use.
Influences on consent included: perceived benefits and harms, trust, understanding (forms, research, anonymity), state of anxiety/being overwhelmed.
|, USA, November 2003 to June 2004||
Deliberative sessions and surveys.|
Quantitative-Chi squared for categorical data and ANOVA for continuous variables
Qualitative- content analysis
Characteristics of deliberators provided|
Age, years—mean (SD)
80 (37), BS or BA or higher
Respondents’ willingness to share data changed based on who requested access to their data.|
• 34% of respondents believed that the current method of consent was OK (ethics boards decide when additional consent is required to access medical records).
• 66% believed that patients should be able to tell the review board whether or not they want to share their data.
• 26% would want opt-out consent.
• Those who wanted opt-in consent: 35% wanted blanket authorisation; 39% wanted to be asked for consent every time.
|, Australia, February to December 2006||
Focus groups, thematic analysis|
Survey, chi square test of independence
|General medicine/general public||
(23, focus group
Age, years (survey population)|
35 (5), 18–19
138 (19.7), 20–34
141 (20), 35–44
208 (29.7), 45–59
178 (25.4), ≥60 y
Education (survey population)
66 (9.4), not finished HS
159 (22.7), finished HS but no HSC
131 (18.7), finished HS and HSC
17 (2.4), some technical or commercial/TAFE
24 (3.4), finished technical or commercial/TAFE
13 (1.8), some university/C.A.E.
85 (12.1), tertiary diploma
15 (2.1), now at University/C.A.E.
145 (20.7), university/C.A.E. degree
40 (5.7), post-graduate degree
Concerns relating to consent and privacy are not always connected.|
Opinions on obtaining consent for using health information for research varied from not important to the provision of study information.
73% would share their health data for medical research.
• 12% would not and 14% were unsure.
• The majority would share sensitive data if it were not identifiable.
• 92% of respondents wanted to be asked permission to use data
• 83% want to know what organisation and the type of research being conducted prior to allowing access.
• 5% were opposed to the idea that an individual’s permission should be sought prior to its use for a purpose other than medical treatment.
• 40% of those wanting to be asked for permission were not concerned about privacy.
• Therefore, 89% of those not concerned about privacy believe that permission should be sought when details used for a purpose other than medical care.
The findings were similar between the focus groups and survey.
|, England, June to July 2016||Surveys and interviews, not described||Cancer registry/Cancer patients and non-cancer patients, cancer||2033 (1033 with cancer, 1000 general public)||
Age, years, cancer group|
31 (3), 18–34
155 (15), 35–54
847 (82), ≥ 55
Age, years, general public
290 (29), 18–34
350 (35), 35–54
350 (35), ≥ 55
Cancer status, cancer group only
186 (18), localised/stable
31 (3), advanced
671 (65), remission/cancer free
Cancer type, cancer group only
52 (5), bladder
93 (9), bowel/colorectal
258 (25), breast
134 (13), prostate
62 (6), cervical/womb
155 (15), skin
300 (29), all others
1002 (97), cancer group
920 (92), general public
Family or friend who has/had cancer, general public group only
640 (64), yes
475 (46), cancer group
490 (49), general public
65% of PLWC support the current data collection process compared with 52% in the general public (automatic inclusion in cancer registry at diagnosis).|
• 23% (PLWC) neither oppose nor support; 2% did not know.
• 28% (general public) neither oppose nor support; 7% do not know.
Awareness of registry changed respondents’ perceptions of the registry and data collection methods.
Knowledge of opt-out option was linked with support in PLWC. 73% of respondents who think that their information can be withdrawn also supported the current method of data collection.
9% of PLWC and 12% of general public respondents oppose the current system. Of those who oppose the current data collection practices:
• 50% (PLWC) and 43% (general public) of respondents believed that permission should be obtained before inclusion in the registry.
• 9% (PLWC) and 6% (general public) concern about not being told about the registry.
• 6% (PLWC) and 17% (general public) registry should be opt-in not opt-out.
• 6% (PLWC) and 2% (general public) believed identifiable information should not be included.
• 6% (PLWC) and 2% (general public) concern regarding third-party access.
• 2% (PLWC) and 7% (general public) believed that the diagnosis/treatment is enough to deal with.
• 10% (PLWC) and 7% (general public) did not know.
Across both PLWC and general public, respondents’ concern were similar.
• 50% (PLWC) and 43% (general public) were concerned about permission not being sought prior to registry inclusion.
• 6% (PLWC) and 17% (general public) believed that the registry should be opt-in.
PLWC who were opposed to the current data collection methods, but who supported cancer data collection more broadly, were concerned about the lack of a consent process.
|, Australia, not reported||Focus groups and semi-structured interviews, open coding and NVivo analysis||Epidemiological research/general public and expert stakeholders||
45 (calculated based on the below)|
4 focus groups with general public (4 to 8 persons per group)
2 focus groups with Aboriginal and Torres Strait Islander peoples (4 to 8 persons per group)
5 people from diverse cultural backgrounds
20 expert stakeholders
The process of obtaining informed consent or refusal should be regarded as the cornerstone of contemporary research ethics.|
Some expressed a sense of obligation about making routinely collected data available for research if government health services have been used; however, consent must be given. In contrast, this same position was used to argue that consent should not always be sought.
• The individual may not recognise their obligation to society to share their data for research.
|, Great Britain, November to December 2015||Deliberative workshops and face to face interview, not reported||Commercial access to health data/general public, doctors, individuals with chronic or rare disease||2263 (246 focus groups 2017 interviews)||Not reported||
Knowing that there are safeguards in place changed the acceptability of data sharing.|
• 49% no safeguard
• 56 to 64% where a safeguard is in place.
54% of respondents would like to be asked permission before data is shared with a commercial organisation, even if this means that the research may not proceed if permission is not granted.
53% wanted strict rules about data not being passed to third parties.
Many participants changed their views on consent after discussion, moving from consent being needed every time to seeing that this may be impractical.
• By increasing education around research and data use, respondents may not wish to opt-out.
Participants wanted more information on aggregation and anonymisation. They noted a lack of consistency more generally about opting-in and opting-out of things generally in society.
If consent cannot be sought for all access to data, there should be clear communication why this is.
Consent can help reassure the participant that they have control on their data.
|, Canada, not reported||Public dialogues and survey, ANOVA and MANOVA||General medicine/general public||98||
37 (38), 20–39
35 (36), 40–59
26 (27), ≥ 60
26 (26), HS or less 18 (18.7), some post-secondary
42 (42.7), completed post-secondary
12 (12.5), post graduate or professional degree
Broad opt-in consent for the use of personal information was the favoured option in the presented scenarios.|
• Consent for each use and assumed consent were equally represented across the spectrum.
Where profit was part of the scenario, respondents moved to consent processes with greater control.
• Personal control was the commonly cited approach for improving the individual’s confidence in the use of their information for research.
No one method of consent was acceptable to all respondents and different types should be developed for different types of research.
• Individuals should be able to see who has accessed their record for non-clinical care.
Aspects for each consent method were liked by respondents.
• Consent for each use was an opportunity for education, increased communication, and a better understanding of the individual’s role in research.
• This research was seen to respect the individual’s privacy.
• Respondents acknowledge this process is burdensome for the research team and may lead to sampling biases.
• Assumed consent was seen as the most efficient approach and the least subject to bias. Reduced burden on the individual needing to give consent.
• Places onus on the individual to opt-out if they do not want to participate.
• Lack of individual education and control with this approach.
• Broad consent was seen as a compromise between consent for each use and assumed consent.
• Less burdensome than consent for each use, but as an opt-in approach, offered individual control.
Ability to control access was a key theme; respondents wanted the opportunity to opt-out.
Concern about control over data once released to researchers was expressed.
1. Using prescribing information for quality improvement: 21% happy for data to be used without notification; 27% notification with opt-out option. One-fifth opted for broad consent and 25% wanted to be asked for permission for each use.
2. Using prescribing information for market research: ~60% believed information should not be used for this purpose. 14% believed permission should be sought each time; only 5% opted for passive use without notification; 13% notification with opt-out.
3. Linking work, education, or income with health information: 10% believed this information should not be linked; ~ 25% preferred that permission be sought each use; 17% use of data with data; 17% use of data with notification and opt-out process.
4. Linking of health information with leftover tissue: non-commercial access—similar response to scenario 3.
5. Linking of health information with leftover tissue: commercial access (for profit): 45% permission required for each use; 18% no linkage should occur at all.
The use of data safeguards made respondents more receptive to research using data.
|, Canada, November 2006 to July 2007 and September 2007||Survey and focus groups, regression analysis using generalised estimating equations||General medicine/DM, HT, chronic depression, alcoholism, HIV, BC, LC and general public.||1780||
(n = 1137)|
33), HS or less
33), poor or very poor
27), good or excellent
Completed survey n = 403|
Consent preferences were similar between health conditions.
Research involving profit or linking of health information to education, income or occupation records required more restrictive consent processes.
• Respondents were more willing to link biological samples with health information that with education, income, or occupation records.
Consent preferences ranged from ‘Just use it’ to ‘information should not be shared’. The majority believed the following, rates were similar between the general public and target health condition populations.
• Just use it was seen as appropriate for quality improvement linking of data and biological sample where no profit is made.
• Information used for marketing and for profit were not widely supported.
• Permission to use data before each use was required when information was linked to biological samples for profit or when it was liked with information on education, income, or occupation.
Those more concerned with privacy required more restrictive consent processes.
Use of data for profit was conditional on more restrictive consent processes.
No consent option was supported by a majority of respondents making it difficult to propose a single method for the secondary use of health data.
|, USA, 1997||Letter requesting consent, percentage of respondents, logistic regression||General medicine/attendees to medical centre||2463||
618 (25.1), 20–39
624 (25.3), 40–59
612 (24.8), 61–79
609 (24.7), ≥ 80
3.2% declined authorisation (95% CI, 2.4% to 4.0%) access to medical records for research.|
• If those who did not respond to request are also considered to have refused, 20.7% (95% CI, 18.5% to 22.9%).
Women were more likely to refuse authorisation compared with males (4% and 2.4% respectively, p=0.067).
Patients under the age of 60 years were also more likely to refuse compared with older patients (5.4% compared with 1.2% respectively, p < 0.001).
Respondents over 120 miles (193 kilometres) from the study centre were more likely to authorise access compared with local respondents (2.1% versus 5.8% respectively, p=0.001).
Respondents with a more sensitive diagnoses such as mental health conditions, infectious diseases or parasitic disease, and reproductive issues) were less likely authorise access.
|, New Zealand, not reported||Citizens jury, not applicable||Pharmacoepidemiology research/general public||13||
18–65 (7 (54) were 45)
The jury agreed that researchers contracted by a public body should be permitted to access medical records with identifiable information only in some circumstances.|
• Routine analysis to identify potential adverse effects from newly introduced medicines.
• Investigating emergency concerns about the adverse effects of medicines currently being used.
Several safeguards were established to protect the use of the data.
Most jury members were happy for their medical information (identified solely by their medical record number) to be used for research and to improve the health of New Zealanders.
|, UK, not reported||Citizens jury, not applicable||General medicine/general public||34||
8 (23.5), 18–29
10 (29.4), 30–44
10 (29.4), 45–59
6 (17.6), ≥ 60
13 (38.2), no qualification
11 (32.4), apprenticeship or other qualification
10 (29.4), degree level or above
33 (97.1) were in favour of secondary use of data for research.|
Methods of consent
• 24 (70.6), opt-out.
• 6 (17.6), opt-in.
• 3 (8.8), all records should be available
• 1 (2.9), no change.
Public benefit was a justification for access.
Views on data sharing changed over the jury process.
17(50), became more willing to share for public benefit.
2 (5.9), opted for more control.
|, Ireland, 2007||Letter requesting consent, not described||General medicine/general public||1178||
50.9 (SD 20.8) (mean), 47 (median)
14.5% (n = 171) of respondents explicitly opted out of the research.|
• 142 by letter.
• 15 by website.
• 2 by phone.
• 12 returned letters without indication.
Those who opted out via the website were younger compared with those who responded via letter (nonparametric, 53.5 years compared with 38.7 years, p < 0.05).
Patients who opted out were slightly older (52.8 years versus 50.4 years; not statistically significant) and were more likely to be female (83% versus 79.5%; not statistically significant).
1.9% (n = 22) opted into the research.
2 patients expressed concern regarding the opt-out methodology and how their contact details were obtained.
83.4% participated indicating that opt-out consent was and acceptable method of recruitment.
|, USA, January 1997 or February 1997||Verbal request to sign consent form, descriptive statistics, Mantel-Haenszel test, and multivariate logistic regression||General medicine/general public||15,997||
4384 (27), ≤ 16^
5892 (37), 17 to 40^
4059 (26), 41 to 64
1662 (10), ≥ 65
90.6% of participants granted authorisation to use data from their medical records for research.|
• 3.6% refused authorisation.
• 4.5% were undecided.
• 1.3% were not asked (for example those receiving emergency care).
Refusal rates were higher in respondents who received care for mental health concerns, trauma, or eye care.
Those aged 39 years and older were also more likely to refuse.
The highest rates of undecided were in women presenting for pregnancy care.
|, Australia, March to December 1997||Letter requesting consent, not reported||General medicine/participants in the ALSWH||39,883||
14,228 (36), 18–23
13,338 (33.4), 45–50
12,317 (31), 70–75
49.4% provided consent for record linkage: 37% of the young women, 59% (n = 7898), of the mid-age women and 53% of the older women.|
Consenters in all age groups had higher levels of education and were more likely to have private health insurance.
Similar proportions of consenters and non-consenters in the young and mid-age groups experienced major personal illness and hospitalisation in the previous year.