ABSTRACT
Background:
The growing importance of collecting Broad Consent (BC) for research utilizing electronic health records in Germany has brought attention to the need for a deeper understanding of potential selection bias in the process. Since 2020, the BeoNet-Halle outpatient database has been collecting anonymous and pseudonymous patient data from primary care and specialty practices throughout Germany, with the practice being an integral part of this data collection effort. The primary objective of the pilot study is to explore potential socioeconomic discrepancies between patients who provided BC and the general practice population.
Method:
This is a single-center, cross-sectional study. The study was performed with patients from one Medical Care Center including eight GPs. We categorized patients with at least one interaction with a general practitioner from March 2021 to January 2023 into two sets: patients who approved BC versus a randomly chosen representative sample (RS) of non-BC inquirers. We mailed a sociodemographic survey to both groups.
Results:
A total of 561 patients were analyzed, with the BC group responding more actively (60.7%) than the RS group (29.7%). Age and gender were similar between the BC group and RS group. Being widowed, divorced, or unmarried and being neither open nor hostile toward research was associated with an increased likelihood of giving consent. Analysis of personality traits did not show any impact on giving consent.
Conclusions:
Overall, this study outlines that there is some bias between BC and RS. Possible associations in BC decisions that offer insights into complex decisions to participate in medical research are marital status, immigrant background, income, and age. Findings emphasize the potential of BC for outpatient research, warranting further investigation to optimize its application in the general practice setting.
Keywords: BeoNet-Halle, broad consent, family medicine, general practice, informed consent, primary care, socioeconomic differences
Introduction
In many countries, obtaining informed consent stands as a fundamental prerequisite for conducting research using electronic health records (EHRs).[1] In Germany, for instance, before an individual’s data can be used in research, they must be informed about the purpose, type of use, risks, and benefits associated with the data utilization. Subsequently, written consent is obtained for the data’s usage.[2] However, it is often the case that not all possible uses of the data are foreseeable at the time consent is given. For this reason, a more generally formulated and therefore more comprehensive wide consent is obtained. Various surveys showed that more than one in two Germans would be ready to make their data available to medical research for an unlimited period of time.[3,4]
The so called Broad Consent (BC) appears particularly advantageous as it enables the collection, storage, and secondary use of patient data for research purposes without the need for specific consent for each individual investigation.[5] BC allows researchers to access a wider range of data, facilitating comprehensive healthcare studies and ethical research practices.[6] Undoubtedly, a substantial milestone in the integration of BC for EHR research within the German context was reached in 2020 when the Medical Informatics Initiative (MII) successfully developed a nationally harmonized BC procedure.[7] Presently, this procedure is in active implementation across multiple sites, encompassing university hospitals and, with minor adaptations, also within community settings. However, literature on BC implementation in practice seems to be largely missing, also because BC is a fairly new concept introduced in Germany and some few other countries.[7,8]
Numerous studies have hints for selection biases arising from obtaining informed consent in general practice (GP) setting.[9,10,11,12,13] Some of these studies look at patients with a specific disease pattern. In Irish GP practices, for example, patients with ischemic heart disease were more likely to give consent, when they have undergone a percutaneous transluminal coronary angioplasty, with a blood pressure „<140/90” mmHg and a total cholesterol level „<5 mmol/l” or when they were an ex-smoker.[9] Patients diagnosed with iron-deficiency anemia in a GP setting in Birmingham were more likely to give consent if they were older, male, and not socioeconomically disadvantaged.[10] When not screening for specific diseases, data from two London medical practices showed no association between consent and demographic factors.[14] In a systematic review, differences were found between those who agreed and those who disagreed in terms of age, gender, race, education, income, and health status.[12] However, the results of the 17 studies examined were inconsistent with regard to the extent and direction of the effect. In another large-scale systematic review including 27 studies, men and older people were more likely to consent to the use of their data.[11]
In contrast, research addressing bias in BC acquisition within GP settings remains absent. The scrutiny of potential selection bias during the acquisition of BC within a GP setting holds considerable importance. This becomes particularly relevant when taking into account prior research that highlights the strong willingness of German GP patients to grant BC in this context, primarily due to the substantial trust that exists within the long-term doctor–patient relationship.[1] According to a systematic review, the majority of studies fail to delve into the socioeconomic rationales behind giving consent by not documenting on disparities based on factors such as age, gender, race, income, education, or health status.[12] However, estimating socioeconomic disparities related to BC may shed light on the inclusivity and diversity of the data used for research, which may influence the generalizability of study findings or which might have implications for health equity decisions.[15,16]
Being part of the outpatient database BeoNet-Halle,[6] the present investigation aims to explore socioeconomic differences between a group of patients that provided BC and a random sample (RS) of patients of a German GP setting.
Materials and Methods
For this study, we followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for cross-sectional studies.[17]
Database
BeoNet-Halle serves as a comprehensive outpatient database, housing anonymized and pseudonymized data derived from various GP and specialty practices. Administered by the Martin-Luther-University Halle-Wittenberg, this initiative aims to create a rich repository of healthcare information. Presently, the anonymized dataset encompasses records from 100,703 patients across nine GP practices, while pseudonymized data are available for 474 patients from a single single Medical Care Center. The inclusivity of BeoNet-Halle is a notable feature as any medical practice, irrespective of specialization, can participate. The overarching goal is to meticulously map individual care clusters, fostering a holistic understanding of patient care. Notably, participating practices vary in practice form (individual/group practice), employed physicians, practice size, and the number of patients treated annually. Given the pilot phase of the project, the recruitment of many participating doctors has been facilitated through university contacts. This underscores the collaborative nature of the initiative and the ongoing efforts to establish a robust foundation for outpatient care mapping.
Since 2020, the BeoNet-Halle collects anonymous and pseudonymous patient data from participating general practices. The nationally agreed consent documents of the MII served as the consent template, with their wording largely adopted.[7,8] Modifications in the BeoNet-Halle BC involve the exclusion of biosample collection and utilization. In addition, patients are given the option to grant BC for their patient data to be linked across participating practices and other institutions within the BeoNet-Halle framework.
The BC procedure within BeoNet-Halle follows these steps: In alignment with ¦630e of the German Civil Code regarding disclosure obligations, the BC disclosure is carried out verbally by either the GP or a staff member specially trained for this purpose.[18] For this study, a trained nurse at the reception provides patient information and clarifying any open questions. After sufficient time for consideration and consultation with the GP if desired, the patient is given the BC form for review and signature. Notably, during the study phase, a transition was made from the conventional paper-based BC form to electronic consent via a tablet. Once a patient provides their BC, the HL7 FHIR-compliant (Health Level 7® standard Fast Healthcare Interoperable Resources®) BC form is automatically transmitted to a trusted third party and processed fully electronically.[19] The consent or revocation status is entered manually into the patients’ EHR. From consented, the patient data are transmitted pseudonymously from the practice management system into the BeoNet-Halle database, where it is made available for researchers.
Sampling and study population
The current cross-sectional investigation was a subproject within the BeoNet Halle project. It was conducted with patients from one Medical Care Center including eight GPs. The General Practice was recruited at an early stage of the BeoNet project and is currently the only BeoNet-Halle practice to have collected pseudonymised patient data via BC.
As part of the recruiting process in this study, the responsible nurse asked for BC from patients who visited the practice for an appointment with the GP. The two sample groups were defined as follows:
First group (called BC): All patients who had already agreed to BC after they had been recruited by the nurse in charge. Second group called random sample (RS): a sample of patients, selected by using a random generator, which had at least one physician contact in the practice in the survey period but were not yet contacted to participate in BeoNet Halle at all.
In the survey period (February 2023 to May 2023), patients in both groups were contacted via mail with enclosed questionnaires (Annex 1). To increase the response rate, a reminder was sent after an interval of 1 month.
Questionnaire
A partially self-designed paper-based questionnaire was sent via mail to each patient (Annex 1). The questionnaire was developed and designed by an interdisciplinary team of general practitioners, social scientists, and health researchers. The questionnaire included a total of 31 items from the following categories: personal information, education and training, occupation, income, personal health status, openness to medical research, and personality questions. While the sociodemographic items were self-designed, the two items related to health-related quality of life (Item 5.1 and 5.2, compared to Annex 1) were used from the validated SF-36 instrument.[20] The personality dimensions (Item Block 7) were assessed using a validated short version of the Big Five Inventory 2 (BFI-2).[21] Prior to the main survey of the two patient groups, the questionnaire was subjected to a qualitative pretest using a Think-Aloud method involving 15 patients. This led to minor modifications.
Data analysis
Sociodemographic variables for BC and RS were presented using both absolute and relative frequencies to account for missing values in individual items; frequencies were reported as % (n/nvalid). For the analyses, we coded a dichotomised variable ‘BC Yes or Random Sample’ from the data set. We compared the gender and age in the BC with the general practice population using percentages. The items related to personal health status (SF-36) and the short version of the Big Five instrument were rated on a 5-point Likert scale (“excellent” as one pole and “poor” as the opposite pole). Regarding the Big Five Inventory 2 (BFI-2), there were calculated sum scores based on the coding of the items for the validated categories of the instrument (Neuroticism, Extraversion, Openness, Compatibility, Conscience).
For the analysis, categorical items were grouped together and recoded accordingly in order to increase the power of the cases in the individual categories (family status, education, employment, main occupation, nationality, age groups).
A multivariate logistic regression was conducted to explore potential associations with the willingness to consent. The outcome variable was dichotomized into two categories: consent BC and RS (0). Aligning with the research focus, the model incorporated socioeconomic variables, variables on openness to research, and Big Five personality traits as independent variables. Variables from the SF-36 questionnaire were excluded from the analysis as they were deemed irrelevant to addressing the research question. The possible associations between BC status and the items were measured with odds ratios (ORs), with their respective 95% confidence intervals (95% Cls) and significance tests according to Wald (p). To assess the model’s quality, the omnibus test of the model coefficients, Nagelkerke’s R-square, and the Hosmer-Lemeshow test as integral components of the model fitting process were employed.
Missing values were excluded from all analyses on a case-by-case basis. Statistical analysis was performed using a combination of IBM SPSS Statistics 25 for data analysis of sociodemographic data and logistical regression. Significance was assumed at a probability error of P < 0.05.
Ethical approval
The study obtained ethics approval from the Martin-Luther-University Halle-Wittenberg’s researcher ethics committee (reference number: 2023-010). Ethical approval allowed the researchers to collect pseudonymized sociodemographic and morbidity relevant data from patients who gave BC. From the RS, data were collected and analyzed and completely anonymized.
Results
Sample characteristics
Ultimately, 561 patients participated in the survey (BC n = 264/435, response 60.7%/RS n = 297/1000, response 29.7%). The age distribution exhibits a similar pattern in both groups. The majority of participants in both samples held German nationality. Most individuals in both groups were married. A considerable proportion of participants in both groups were parents. Moreover, the majority of respondents in both groups lived in a small town with a small number, indicating a history of migration in each group. In terms of educational attainment, the prevailing trend was the completion of an apprenticeship, with only a minority holding technical college or university degrees. Most respondents in both groups were retired, with only a smaller proportion employed. The responses related to average net income revealed that a notable portion fell within the 1000 to 2000 Euro (€) range. Similarly, respondents with an average net income below 1000 € were also well represented [Table 1].
Table 1.
Comparative analysis of demographic and socioeconomic characteristics between Broad Consent and Random Sample groups including valid responses (n/nvalid) and percentage (%) distribution
| Variable | Broad Consent Group (n/nvalid) | % | Random Sample (n/nvalid) | % |
|---|---|---|---|---|
| Sex | ||||
| Female | (140/262) | 53.0 | (154/296) | 52.0 |
| Age | ||||
| ≤ 19 | (5/264) | 1.9 | (0/297) | 0.0 |
| 20-39 | (30/264) | 11.4 | (28/297) | 9.4 |
| 40-59 | (65/264) | 24.6 | (83/297) | 28.0 |
| 60-79 | (118/264) | 44.7 | (131/297) | 44.1 |
| ≥ 80 | (46/264) | 17.4 | (55/297) | 18.5 |
| Nationality | ||||
| German | (255/262) | 97.3 | (284/296) | 95.9 |
| Other EU country | (3/262) | 1.1 | (8/296) | 2.7 |
| Non-EU country | (4/262) | 1.5 | (4/296) | 1.4 |
| Current life situation | ||||
| Single | (44/261) | 16.9 | (35/296) | 11.8 |
| Unmarried with partner | (28/261) | 10.7 | (37/296) | 12.5 |
| Married with partner | (152/261) | 58.2 | (149/296) | 50.3 |
| Divorced | (11/261) | 4.2 | (21/296) | 7.1 |
| Widowed | (22/261) | 8.4 | (44/296) | 14.9 |
| Living in a shared apartment | (4/261) | 1.5 | (10/296) | 3.4 |
| Having kids | ||||
| Yes | (183/215) | 85.1 | (220/257) | 85.6 |
| Residential area | ||||
| Big city | (10/259) | 3.9 | (17/293) | 5.8 |
| Small town | (154/259) | 59.5 | (176/293) | 60.1 |
| Village | (95/259) | 36.7 | (100/293) | 34.1 |
| Migration background | ||||
| Yes | (5/228) | 2.2 | (16/269) | 5.9 |
| Highest educational/professional degree | ||||
| None yet | (6/263) | 2.3 | (12/296) | 4.1 |
| Elementary school | (19/263) | 7.2 | (22/296) | 7.4 |
| Completion of the Polytechnic High School in the GDR | (39/263) | 14.8 | (35/296) | 11.8 |
| Secondary school (intermediate level) | (17/263) | 6.5 | (14/296) | 4.7 |
| High school diploma | (9/263) | 3.4 | (4/296) | 1.4 |
| Completed vocational training | (128/263) | 48.7 | (143/296) | 48.3 |
| University of Applied Sciences degree | (29/263) | 11.0 | (34/296) | 11.5 |
| University degree | (16/263) | 6.1 | (32/296) | 10.8 |
| Current job | ||||
| I am employed | (81/263) | 30.8 | (105/295) | 35.6 |
| I am a student at school | (3/263) | 1.1 | (0/295) | 0.0 |
| I am a college student or apprentice | (4/263) | 1.5 | (0/295) | 0.0 |
| I am retired | (144/263) | 54.8 | (158/295) | 53.6 |
| I am a homemaker or taking care of children and/or dependent persons | (4/263) | 1.5 | (10/295) | 3.4 |
| I am unemployed | (15/263) | 5.7 | (16/295) | 5.4 |
| Others (e.g., permanently occupationally disabled) | (12/263) | 4.6 | (6/295) | 2.0 |
| Employment status | ||||
| Employee | (57/251) | 22.7 | (81/283) | 28.6 |
| Worker | (28/251) | 11.2 | (25/283) | 8.8 |
| Self-employed | (6/251) | 2.4 | (6/283) | 2.1 |
| Civil servant | (2/251) | 0.8 | (0/283) | 0.0 |
| Trainee, intern, or paid internship | (0/251) | 0.0 | (0/283) | 0.0 |
| Other | (5/251) | 2.0 | (5/283) | 1.8 |
| Retired (full-time) | (153/251) | 61.0 | (166/283) | 58.7 |
| Net income in euros | ||||
| Under 1000 | (65/253) | 25.7 | (52/280) | 18.6 |
| 1001 to 2000 | (114/253) | 45.1 | (133/280) | 47.5 |
| 2001 to 3000 | (31/253) | 12.3 | (29/280) | 10.4 |
| Over 3000 | (3/253) | 1.2 | (11/280) | 3.9 |
| Don’t know | (0/253) | 0 | (1/280) | 0.4 |
| No information provided | (40/253) | 15.8 | (54/280) | 19.3 |
We compared gender and age distributions between the BC and the general practice population. The BC closely mirrored the practice population in gender distribution, with 47% men and 53% women (practice: 48% men, 51% women, 1% unknown gender). The mean age in the BC group was 60.0 ± 16.8 years), while the practice population’s mean age is 50.8 ± 24.8 years).
Possible correlations with BC
The logistic regression model in Table 2 presents possible correlations between sociodemographic characteristics, Big Five personality traits, openness to research, and providing BC compared to the RS. The omnibus test was significant, while the Hosmer-Lemeshow test showed no significance. The regression model is therefore admissible for the analyses. The Cox and Snell R-squared and Nagelkerke’s R-squared model clarification were in the medium range. Significant associations were observed for individuals who were living unmarried with a partner (OR = 3.03; P < 0.036) and for those who were widowed or divorced (OR = 2,85; P < 0.024), all showing a higher likelihood of giving BC compared to the RS. In terms of openness to new research, to be neither open nor hostile for research was associated with a higher likelihood of giving BC in comparison to RS (OR = 4,56; P < 0.014). No significant associations were identified for Big Five and other sociodemographic variables, such as education or employment.
Table 2.
Logistic binary regression model of the associations of sociodemographic variables, factors related to openness to research and Big Five Personality traits of giving BC
| A): Test of model quality with Omnibus test, Cox & Snell’s R-square and Nagelkerke’s R-square as well as Hosmer-Lemeshow test | ||||
|---|---|---|---|---|
| Omnibus-Test | ||||
|
| ||||
| Chi-Square | df | P | ||
| Step | 127.58 | 83 | 0.001** | |
| Block | 127.58 | 83 | 0.001** | |
| Model | 127.58 | 83 | 0.001** | |
|
| ||||
| Model Quality | ||||
|
| ||||
| Cox and Snell R-Square | Nagelkerkes R-Square | |||
|
| ||||
| 0.302 | 0.403 | |||
|
| ||||
| Hosmer-Lemeshow-Test | ||||
|
| ||||
| Chi-Square | df | P | ||
|
| ||||
| 9.989 | 8 | 0.266 | ||
|
| ||||
| B): Possible measures of association are tested for significance using odds ratios and their respective 95% confidence interval and Wald test. Included in Analyses after clean up missing values n=363 | ||||
|
| ||||
| Logistic regression | ||||
|
| ||||
| Variable | OR | [95%CL] | P | |
|
| ||||
| Sociodemographic | ||||
|
| ||||
| Sex: | ||||
| Male | Ref. | |||
| Female | 0.84 | [0.52; | 1.36] | 0.492 |
| Age: | ||||
| 20–39 | Ref. | |||
| 40–59 | 2.15 | [0.81 | 5.68] | 0.121 |
| 60–79 | 2.60 | [0.80 | 8.42] | 0.110 |
| Over 80 | 2.01 | [0.57 | 7.07] | 0.274 |
| Nationality: | ||||
| German | Ref. | |||
| Not German | 0.72 | [0.11 | 4.74] | 0.738 |
| Family status: | ||||
| Single | Ref. | |||
| Unmarried with partner | 3.03 | [1.07 | 8.59] | 0.036* |
| Married with partner | 1,47 | [0.68 | 3.20] | 0.324 |
| Divorced or widowed | 2.85 | [1.14 | 7.13] | 0.024* |
| Shared apartment | 4.36 | [0.82 | 23.1] | 0.083 |
| Having kids: | ||||
| Yes | Ref. | |||
| No | 1.11 | [0.49 | 2.51] | 0.791 |
| Residential area: | ||||
| City | Ref. | |||
| Village | 0.88 | [0.52 | 1.49] | 0.644 |
| Education and Training: | ||||
| none yet | Ref. | |||
| Lower Education | 0.18 | [0.01 | 1.74] | 0.140 |
| Middle school education | 0.18 | [0.02 | 1.58] | 0.124 |
| Higher school qualification | 0.27 | [0.03 | 2.44] | 0.246 |
| Current job: | ||||
| Employed | Ref. | |||
| Not yet/Not working | 0.37 | [0.06 | 2.05] | 0.257 |
| Retired | 0.20 | [0.03 | 1.15] | 0.073 |
| Employment status: | ||||
| Salaried Employee | Ref. | |||
| Worker | 0.74 | [0.28 | 1.93] | 0.547 |
| Self-employed | 0.50 | [0.10 | 2.54] | 0.408 |
| Full-Time Retiree | 2.16 | [0.44 | 10.66] | 0.341 |
| Net income in euros: | ||||
| Under 1000 | Ref. | |||
| 1001 to 2000 | 1.65 | [0.83 | 3.27] | 0.150 |
| 2001 to over 3000 | 1.68 | [0.67 | 4.19] | 0.266 |
| Don’t know | 1.12 | [0.47 | 2.64] | 0.787 |
|
| ||||
| Openess to new Research | ||||
|
| ||||
| Openness: | ||||
| 1. completely open minded | Ref. | |||
| 2. | 2.13 | [0.82 | 5.54] | 0.118 |
| 3. | 4.56 | [1.35 | 15.38] | 0.014* |
| 4. | 3.02 | [0.63 | 14.32] | 0.163 |
| 5. not open minded at all | 0.51 | [0.00 | 1.00] | 0.163 |
| Participation: | ||||
| 1. completely open minded | Ref. | |||
| 2. | 1.39 | [0.52 | 3.69] | 0.502 |
| 3. | 0.84 | [0.26 | 2.72] | 0.778 |
| 4. | 2.47 | [0.53 | 11.46] | 0.246 |
| 5. not open minded at all | 1.73 | [0.00 | 1.00] | 1.000 |
|
| ||||
| Big Five 2 | ||||
|
| ||||
| Neuroticism | 0.91 | [0.76 | 1.08] | 0.281 |
| Extraversion | 1.03 | [0.90 | 1.18] | 0.636 |
| Openness | 0.90 | [0.80 | 1.02] | 0.131 |
| Compatibility | 0.89 | [0.78 | 1.03] | 0.137 |
| Conscience | 0.89 | [0.76 | 1.04] | 0.148 |
Significance level: * <0,05; ** < 0,01; *** < 0,001
Discussion
Summary of the main findings
There were only a few differences found between the BC and RS groups when assessing socioeconomic variables and personality traits. The likelihood of giving BC compared to RS increased with being unmarried, divorced, or widowed and in terms of neither being open nor hostile for research.
Sample characteristics and representativeness
In this investigation, no significant associations between gender and giving consent was observed. Several existing studies, however, indicate that women exhibited lower willingness to consent in cases of specific consent.[10,11] Explanations and causes for gender-related differences in consent were not yet conclusively researched.[12]
The interaction between consent bias and nonresponse bias remains uncertain, as indicated in the existing literature.[22,23]
The response rate within the BC group exceeded previous research benchmarks focused on specific consent in a GP setting.[9,24] This is in line with our findings that the BC group was generally more open to medical research and research related activities compared to the RS group. It underscores the potential influence of the participants’ inherent motivation and high trust in the GP–patient relationship within this particular patient cohort.[1,24]
Possible associations in giving broad consent
The findings showed a possible impact of relationship status on consent decisions. Divorced, widowed, and unmarried individuals demonstrated higher chances of granting consent, whereas no significant differences were observed among married participants. These findings are in line with previous research which reported higher nonconsent odds among married individuals.[12,15]
Age did not play a role in giving BC. However, previous research reported a dose–response relationship with nonconsent and increasing age; other studies indicate consent odds at younger ages („(<39)” years) or in the age group 80–89 years.[22,24] Other research on consent likelihood in different age groups is inconclusive.[12,22] For example, the Canadian Stroke Registry Network found age-related differences in consent rates between study phases.[25]
Also observed in our study, existing research consistently indicates diminished participation rates among ethnic minority populations in health-related research investigations.[25] Studies conducted within a European context have further elucidated that researchers frequently neglect to establish recruitment targets for specific ethnic cohorts. This omission is attributed to factors such as cultural barriers, differences in perspectives on consent, or varying degrees of confidence in the research protocols.[26]
No notable associations in personality traits existed between the BC and the RS. While only one study reported lower consent odds for participants with elevated anxiety and depression levels, this divergence in traits warrants further exploration to uncover underlying factors contributing to this distinction.[14] Our findings open up the possibility of examining the interplay between personality traits and consent decisions in more detail.
Although the results of this study were analyzed in a German setting, they might be applicable to other regions and are potentially interesting wherever broad consent is to be used for research. Our results can complement previous studies, for example, for Indian Medical Research and Broad Consent.[26]
Strengths and limitations
This study addresses an important topic with high relevance for future research in the outpatient setting. Little research exists to date that specifically examines sociodemographic and socioeconomic variables as possible factors influencing decision-making in BC. The high response rate of BC respondents in particular supports the expressiveness of the results. The RS group as a general sample of GP practice patients allows for a comparison with the BC group.
However, our study has some limitations that must be considered when interpreting the results. First, we could only rely on older literature as there is hardly any recent literature on possible bias factors in BC samples and BC bias could rather describe a German problem. First, language barriers during the recruitment process between the medical assistants and the migrant patients might have resulted in mainly German-speaking patients being asked for BC consent, which could lead to sample bias. The composition of the RS group should be mentioned as a possible limitation. As these individuals might also be asked to participate in BC at a later stage, it is difficult to accurately separate the RS group from the BC group. Regarding the questionnaire, it must be mentioned that several parts were self-developed and that it was only qualitatively tested but was not subjected to any major piloting beforehand. Notably, a temporal bias exists in the BC group due to a survey delay for some patients of up to 1.5 years after recruitment.
Implications for further research
Future studies, especially longitudinal, should also examine the influence of trust in the doctor–patient relationship on consent processes in general practice. The impact of patient–GP trust on outpatient research and utilization of BC is still not well investigated, although trusting patient–GP relationships may favor long-term research and longitudinal designs. Even if biases were found between the RS and BC groups, no conclusions can be drawn about the comparability of the medical data. To harness the full potential of BC for outpatient research, especially in GP settings, potential influencing factors for consenting must be understood, and possible barriers identified and overcome. This requires further studies directly comparing BC participants with BC refusers.
Supplementary information
Annex 1) Questionnaire
Authors’ contributions
KM, FB, and TF contributed to the idea, conception, and design of the study; KM is responsible for the overall content. KM and FB developed the questionnaire. KM, FB, and MR were responsible for the process of recruitment and data collection.
KM, FB, CB, TF contributed to data acquisition, analysis, and interpretation and were involved in writing the manuscript with KM drafting the first version. CB, AB, MR and TF were consultants on data analysis and critically revised the manuscript. TD and SV critically revised the manuscript. All authors participated in reading and approving the final manuscript.
Availability of data and materials
All data generated and analyzed by this study are included in this published article.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
Acknowledgements
We particularly thank the GPs of the General Practice from which data was drawn, we thank trained medical assistant for her consistent consent management and support in the process of data collection and analysis. We also express our gratitude to the software provider for data extraction from the practice management system. For proofreading we acknowledge Dawn Bielawski.
Annex 1
Questionnaire “Health and Life”
Dear patient,
Thank you for taking the time to complete our short questionnaire!
This questionnaire pertains to your personal life situation and your overall health status as part of the registry study on the topic of “Health and Life.” Please answer all questions, as all information is crucial for the continuous improvement of our patients’ healthcare.
Answering all questions will take approximately ten minutes.
Warm regards,
Your medical team
Please tick as follows: ◻ ⊠ ◻ ◻ Please use a ballpoint pen or a not-too-heavy felt-tip pen
| 1. Personal details | |||||
|---|---|---|---|---|---|
| 1.1 What nationality/nationalities do you have? | |||||
| ○ German | ○ other EU country | ○ no EU country | ○ none | ○ not sure | |
| 1.2 Age in years | |||||
| ___ ___ ___ | |||||
| 1.3 Gender | |||||
| ○ female | ○ male | ○ diverse | |||
| 1.4 How do you currently live? | |||||
| ○ single | ○ unmarried with partner | ○ married with partner | |||
| ○ divorced | ○ widowed | ○ shared apartment | |||
| 1.5 Do you have children | ○ yes ○ no | ||||
| 1.6 Where do you currently live? | |||||
| ○ in a big city | ○ in a small city | ○ in a village | |||
| 1.7 Do you have a migration background? | ○ yes | ○ no | |||
|
| |||||
| 2. Education and training | |||||
|
| |||||
| 2.1 What is your highest educational/professional qualification? | |||||
| ○ none yet | ○ elementary or secondary school | ○ graduation from a GDR polytechnic secondary school | |||
| ○ intermediate school | ○ highschool | ○ completed vocational training | |||
| ○ technical college degree | ○ university degree | ||||
|
| |||||
| 3. Employment | |||||
|
| |||||
| 3.1 Which predominantly applies to you? | |||||
| ○ I am employed | ○ I am a pupil | ○ I am a student or trainee | |||
| ○ I am a pensioner, retiree | ○ I am a housewife/husband or care for children and/or persons in need of care | ○ I am unemployed | |||
| ○ Other (e.g. permanently disabled) | |||||
| 3.2 If you are primarily employed: What is your main activity? | |||||
| ○ Employee | ○ Blue-collar worker | ○ Self-employed | |||
| ○ Civil servant | ○ Trainee, traineeship or paid internship | ○ Other (e.g. mini-job) | |||
|
| |||||
| 4. Income | |||||
|
| |||||
| 4.1 What is your own average net income per month? | |||||
| ○ under 1000€ | ○ 1000 - 2000€ | ○ 2000-3000€ | |||
| ○ over 3000€ | ○ don’t know | ○ not specified | |||
|
| |||||
| 5. Personal health condition | |||||
|
| |||||
| 5.1 Think about the past four weeks: How would you describe your general state of health? | excellent | ○ ○ ○ ○ ○ | poor | ||
| 5.2 Compared to last year, how would you describe your current health. | currently much better | ○ ○ ○ ○ ○ | currently much worse | ||
| 5.3 How old do you feel compared to your actual age? | younger | ○ ○ ○ ○ ○ | older | ||
| 5.4 In the past 4 weeks, have you had any of the following problems with your work or other regular daily activities due to your health condition? | |||||
| Have you reduced the amount of time you spend working or doing other activities? | ○ yes | ○ no | ○ don’t know | ||
| Have you achieved less than you would like? | ○ yes | ○ no | ○ don’t know | ||
| Were you limited in work or other activities? | ○ yes | ○ no | ○ don’t know | ||
| Did you have difficulty performing the job or other activities (e.g., additional time required)? | ○ yes | ○ no | ○ don’t know | ||
| 5.5 How severe has your physical pain been in the last 4 weeks? | |||||
| ○ i had no pain | ○ very light | ○ light | ○ moderate | ○ strong | ○ very strong |
|
| |||||
| 6. Openness to medical research | |||||
|
| |||||
| 6.1 How open are you to participating in future research projects? | completely open minded | ○ ○ ○ ○ | not open minded at all | ||
| 6.2 How open are you to participating in innovation projects in medical research? | completely open minded | ○ ○ ○ ○ | not open minded at all | ||
|
| |||||
| 7. Questions about personality | |||||
|
| |||||
| I am rather reserved, reserved. | fully agree | ○ ○ ○ ○ ○ | disagree | ||
| I trust others easily, believe in the good in people. | fully agree | ○ ○ ○ ○ ○ | disagree | ||
| I am comfortable, prone to laziness. | fully agree | ○ ○ ○ ○ ○ | disagree | ||
| I’m relaxed, I don’t let stress get me down. | fully agree | ○ ○ ○ ○ ○ | disagree | ||
| I have little artistic interest. | fully agree | ○ ○ ○ ○ ○ | disagree | ||
| I get out of myself, I’m sociable. | fully agree | ○ ○ ○ ○ ○ | disagree | ||
| I tend to criticize others. | fully agree | ○ ○ ○ ○ ○ | disagree | ||
| I complete tasks thoroughly. | fully agree | ○ ○ ○ ○ ○ | disagree | ||
| I get nervous and insecure easily. | fully agree | ○ ○ ○ ○ ○ | disagree | ||
| I have an active imagination, I am imaginative | fully agree | ○ ○ ○ ○ ○ | disagree | ||
Many thanks for your help!
References
- 1.Kongeter A, Schickhardt C, Jungkunz M, Bergbold S, Mehlis K, Winkler EC. Patients'willingness to provide their clinical data for research purposes and acceptance of different consent models: Findings from a representative survey of patients with cancer. J Med Internet Res. 2022;24:e37665. doi: 10.2196/37665. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Orzechowski M, Woniak K, Timmermann C, Steger F. Normative framework of informed consent in clinical research in Germany, Poland, and Russia. BMC Med Ethics. 2021;22:53. doi: 10.1186/s12910-021-00622-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Lesch W, Richter G, Semler SC. Sharing data for research: Attitudes and perspectives on data donation in Germany. In: Richter G, Loh W, Buyx A, Graf von Kielmansegg S, editors. Data-Rich Medicine and the Problem of Consent: Ethical, Legal and Social Science Perspectives. Berlin, Heidelberg: Springer Berlin Heidelberg; 2022. pp. 211–26. [Google Scholar]
- 4.Richter G, Borzikowsky C, Hoyer BF, Laudes M, Krawczak M. Secondary research use of personal medical data: Patient attitudes towards data donation. BMC Med Ethics. 2021;22:164. doi: 10.1186/s12910-021-00728-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Zenker S, Strech D, Ihrig K, Jahns R, Muller G, Schickhardt C, et al. Data protection-compliant broad consent for secondary use of health care data and human biosamples for (bio)medical research: Towards a new German national standard. J Biomed Inform. 2022;131:104096. doi: 10.1016/j.jbi.2022.104096. [DOI] [PubMed] [Google Scholar]
- 6.Wiertz S. The temporal dimension of broad consent. Ethics Med. 2022;34:645–67. [Google Scholar]
- 7.Medizininformatikinitiative. Template text for patient consent forms 2023. Available from: https://www.medizininformatik-initiative.de/en/template-text-patient-consent-forms .
- 8.Moser K, Mikolajczyk R, Bauer A, Tiller D, Christoph J, Purschke O, et al. [BeoNet-Halle-development of a multifunctional database for the automated extraction of healthcare data from general practitioner and specialist practices. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz. 2023;66:569–77. doi: 10.1007/s00103-023-03691-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Buckley B, Murphy AW, Byrne M, Glynn L. Selection bias resulting from the requirement for prior consent in observational research: A community cohort of people with ischaemic heart disease Heart. 2007;93:1116–20. doi: 10.1136/hrt.2006.111591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Damery S, Ryan R, McManus RJ, Warmington S, Draper H, Wilson S. The effect of seeking consent on the representativeness of patient cohorts: Iron-deficiency anaemia and colorectal cancer. Colorectal Dis. 2011;13:e366–73. doi: 10.1111/j.1463-1318.2011.02724.x. [DOI] [PubMed] [Google Scholar]
- 11.Hill EM, Turner EL, Martin RM, Donovan JL. “Let's get the best quality research we can”: Public awareness and acceptance of consent to use existing data in health research: A systematic review and qualitative study. BMC Med Res Methodol. 2013;13:72. doi: 10.1186/1471-2288-13-72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kho ME, Duffett M, Willison DJ, Cook DJ, Brouwers MC. Written informed consent and selection bias in observational studies using medical records: Systematic review. BMJ. 2009;338:b866. doi: 10.1136/bmj.b866. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Macleod U, Watt GC. The impact of consent on observational research: A comparison of outcomes from consenters and non consenters to an observational study. BMC Med Res Methodol. 2008;8:15. doi: 10.1186/1471-2288-8-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Harris T, Cook DG, Victor C, Beighton C, Dewilde S, Carey I. Linking questionnaires to primary care records: Factors affecting consent in older people. J Epidemiol Community Health. 2005;59:336–8. doi: 10.1136/jech.2004.025296. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Huang N, Shih SF, Chang HY, Chou YJ. Record linkage research and informed consent: Who consents? BMC Health Serv Res. 2007;7:18. doi: 10.1186/1472-6963-7-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Demarest S, Van der Heyden J, Charafeddine R, Tafforeau J, Van Oyen H, Van Hal G. Socio-economic differences in participation of households in a Belgian national health survey. Eur J Public Health. 2013;23:981–5. doi: 10.1093/eurpub/cks158. [DOI] [PubMed] [Google Scholar]
- 17.Equator-network. STROBE Statement—Checklist of items that should be included in reports of cross-sectional studies 2023. Available from: https://www.equator-network.org/wp-content/uploads/2015/10/STROBE_checklist_v4_cross-sectional.pdf . [DOI] [PubMed]
- 18.Justiz Bd. German Civil Code (BGB): §Section 630e Duty to provide information 2023. Available from: https://www.gesetze-im-internet.de/bgb/__630e.html .
- 19.Bialke M, Geidel L, Hampf C, Blumentritt A, Penndorf P, Schuldt R, et al. A FHIR has been lit on gICS: Facilitating the standardised exchange of informed consent in a large network of university medicine. BMC Med Inform Decis Mak. 2022;22:335. doi: 10.1186/s12911-022-02081-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Coperation R. 36-Item Short Form Survey (SF-36) 2023. Available from: https://www.rand.org/health-care/surveys_tools/mos/36-item-short-form.html .
- 21.Gesis Zb. The German version of the Big Five Inventory 2 (BFI-2) 2016. Available from: https://zis.gesis.org/skala/Danner-Rammstedt-Bluemke-Lechner-Berres-Knopf-Soto-John-Die-deutsche-Version-des-Big-Five-Inventory-2-(BFI-2)
- 22.Dunn KM, Jordan K, Lacey RJ, Shapley M, Jinks C. Patterns of consent in epidemiologic research: Evidence from over 25,000 responders. Am J Epidemiol. 2004;159:1087–94. doi: 10.1093/aje/kwh141. [DOI] [PubMed] [Google Scholar]
- 23.Woolf SH, Rothemich SF, Johnson RE, Marsland DW. Selection bias from requiring patients to give consent to examine data for health services research. Arch Fam Med. 2000;9:1111–8. doi: 10.1001/archfami.9.10.1111. [DOI] [PubMed] [Google Scholar]
- 24.Zondag AGM, Groenhof TKJ, van der Graaf R, van Solinge WW, Bots ML, Haitjema S. Asking informed consent may lead to significant participation bias and suboptimal cardiovascular risk management in learning healthcare systems. BMC Med Res Methodol. 2023;23:98. doi: 10.1186/s12874-023-01924-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Tu JV, Willison DJ, Silver FL, Fang J, Richards JA, Laupacis A, et al. Impracticability of informed consent in the Registry of the Canadian Stroke Network. N Engl J Med. 2004;350:1414–21. doi: 10.1056/NEJMsa031697. [DOI] [PubMed] [Google Scholar]
- 26.Behera SK, Das S, Xavier AS, Selvarajan S, Anandabaskar N. Indian Council of Medical Research's National Ethical Guidelines for biomedical and health research involving human participants: The way forward from 2006 to 2017. Perspect Clin Res. 2019;10:108–14. doi: 10.4103/picr.PICR_10_18. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
All data generated and analyzed by this study are included in this published article.