Opt‐in or opt‐out health‐care communication? A cross‐sectional study

Abstract Background Patients need medication and medical condition‐related information to better self‐manage their health. Health‐care professionals (HCPs) should be able to actively provide information outside of one‐on‐one consultations; however, patient consent may be required. Objective To investigate the Australian public's preferences, and factors that may influence their preferences, towards an opt‐in versus an opt‐out approach to health communication. Design A cross‐sectional study using a structured questionnaire administered via Computer‐Assisted Telephone Interviewing. Setting and participants Participants across Australia who were adults, English‐speaking and had a long‐term medical condition. Main outcome measures Preferences for opt‐in vs opt‐out approach to receiving follow‐up tailored information. Results A total of 8683 calls were made to achieve the required sample size of 589 completed surveys. Many (346/589; 58.7%) indicated that they were interested in receiving tailored, ongoing follow‐up information from their HCP. Nearly half (n = 281; 47.7%) preferred an opt‐in service and 293/589 (49.7%) an opt‐out service for receiving follow‐up information. Reasons for preferring an opt‐in service were being in control of the information received (n = 254); able to make a decision that is best for them (n = 245); opt‐in service would save time for HCPs (n = 217); they may not want or need the information (n = 240). Many (n = 255) felt that an opt‐out service should be part of the normal duty of care of their HCP and believed (n = 267) that this approach would ensure that everyone has access to information. Conclusions Respondents were interested in receiving tailored information outside of consultation times. However, preferences for an opt‐in or opt‐out approach were divided.


| INTRODUC TI ON
Information relating to medical conditions and medicines is important to help enable patients to better self-manage their health. 1 Health and medicines information may be accessed by patients in a number of ways including, but not limited to, spoken information delivered by health-care professionals (HCPs) either during a consultation, or over the phone; through online platforms such as websites, apps and social networking sites; and written forms such as patient information leaflets, newsletters and booklets. However, since patients' preferences and needs for information are diverse and vary throughout their patient journey, a 'one-size-fits-all' approach is unlikely to cater for everyone. 2 Patients desire tailored information 2,3 and are entitled to receive relevant health and medicines information that is appropriate for their needs. 4 It is therefore incumbent upon HCPs to identify and deliver such targeted health information. However, research has shown that patients' information needs have not always been well met at different points along the health-care continuum. 5 HCPs typically communicate health and medicines information at face-to-face consultations. However, consultation times can vary 6 which may then impact the degree to which complete, necessary information is received by patients in a timely manner. Furthermore, information that is provided during consultations may not be remembered, 7 or easily understood and enacted. 8,9 Although more patients are seeking information themselves via the Internet and social networking sites, the quality, accuracy and relevance of the information sourced can vary. 10 Furthermore, information-seeking behaviour may differ between patients, where certain factors such as medical condition, health literacy level and health locus of control can influence the likelihood of patients reading and/or sourcing information, such as written medicines information. 11 One approach to ensuring that patients receive appropriate, relevant and quality information, tailored to their individual needs throughout their patient journey, is for HCPs to actively provide information, not only during face-to-face consultations, but regularly using multi-modal communication channels as part of patient follow-up. 12 However, patient consent may be needed for provision of ongoing health information by HCPs outside of consultation times. For example, in Australia, according to the Privacy Act 1988-Australian Privacy Principles, patients must provide consent for a registered HCP to use their personal information to provide them with proactive support and advice on an ongoing basis 13 ; however, depending on the context and a case-by-case basis, implied consent may be acceptable for certain communications. 14 Seeking consent upfront is regarded as an opt-in approach and may limit the type and nature of support and information received by patients.
However, in countries such as the United States, where prior consent from patients is not required, HCPs may send additional information related to treatment/health care. 15 Patients not wishing to receive such information would need to 'opt-out' based on the Standards for Privacy of Individually Identifiable Health Information. 16 Healthcare systems with an 'opt-out' approach may therefore differ in how they facilitate the provision of tailored and timely information by HCPs outside consultation times.
Both opt-in and opt-out approaches have their merits, and it is important to consider patients' needs as well as preferences in receiving information when determining an approach for a country's health-care system. In exploring patient preferences for follow-up information, previous research has found great diversity, suggesting that a tailored approach may lead to increased patient satisfaction. 17 For example, the findings of a qualitative study conducted with a sample of Australian patients showed that while participants were receptive to receiving tailored information about their medicines and/or medical conditions from their HCP on an ongoing basis, there was variation in expressed preferences for the method of information delivery, frequency and type of content. 18 The concept of timely and tailored information provision that meets patients' needs on an ongoing basis underpins the rationale for this research. This study aimed to investigate the Australian public's preferences and factors that may influence their preferences towards an opt-in versus an opt-out approach to the provision of information by HCPs on an ongoing basis. Sydney, Australia, with adult, English-speaking participants who had at least one long-term medical condition. A semi-structured focus group protocol addressed discussion topics relating to participants' health and/or medicine information sources; and their perspectives on receiving information from their HCP(s), and systems and consenting process to facilitate receiving such information. Participants emphasized the importance of tailored information and held mixed views on how consent should be obtained for HCP-initiated information provision.

| ME THODS
Based on the themes and subthemes derived from the thematic analysis of the verbatim focus group transcripts, the items in the CATI structured questionnaire covered: • Previous experiences, opinions and preferences about receiving follow-up tailored information about medical conditions and/or treatments from HCPs outside of consultation times on a regular basis; • Patient perspectives regarding whether the above should be an opt-in or opt-out process; • Patient preferences for information in response to scenarios representing different time points in the patient medication-taking continuum; and • Demographics.
The structured questionnaire was piloted to establish its clarity, face and content validity, and feasibility of administration.
Two HCPs (a doctor and a pharmacist/researcher) who were external to the research team initially pilot-tested the questionnaire. Subsequent pilot tests were conducted with 4 non-HCPs/consumers. All pilot tests were conducted sequentially by telephone to simulate conditions under which the questionnaire would be administered (CATI). Piloting and subsequent revisions were conducted iteratively until no further improvements to the questionnaire were deemed necessary by the research team and the questionnaire could be administered in less than 15 minutes in consideration of the impact of survey length on respondent fatigue and response rates. 20 The final survey consisted of 4 sections (Table 1).

| Sample size and sample stratification
The sampling strategy involved several steps. The initial step involved the calculation of the sample size required to measure an estimated prevalence of the preference for receiving opt-out health information among members of the Australian public, estimated as 50% from a previous study. 18 For a simple random sample 23 with a confidence level of 95%, population percentage as 50%, and degree of precision of 5%, the minimum sample size required was 385 participants. This was inflated to 500 to account for sample stratification to ensure a nationally representative sample with adequate representation from subpopulations.
The stratification strategy used to generate interlocked quotas was based on data from the Australian Bureau of Statistics (ABS).
The four key parameters taken into consideration when calculating the sample quotas required for stratification were: 1. State/Territory population sizes (to obtain a nationally representative sample of the Australian population) 24 ; 2. Remoteness Areas (population proportions residing in Major Cities, Regional (Inner and Outer), and Remote (Remote and Very Remote) areas in Australia) 25 ; 3. Index of Relative Socio-economic Disadvantage (IRSD) (sampling was intended to encompass the full spectrum of the IRSD; IRSD deciles linked to Postal Areas were stratified into low, middle, and high IRSD groups for the sampling strategy) 26 ; and 4. Age (in light of the prevalence of long-term medical conditions in the Australian population by age). 27 In addition, an equal distribution of male and female respondents was also included as part of the overall target quota, as per ABS data.
However, due to the study inclusion criteria and CATI survey methods utilized, a skew towards a larger number of respondents aged 55 years and older was anticipated.
As the generated sample yielded a low proportion of the population residing in Remote Australia (n = 11 of a total of 500), oversampling of respondents within this group was undertaken to ensure an adequate sample size to detect a difference in preferences for an opt-in or opt-out approach between groups of participants by Remoteness Area (Major Cities, Regional Australia, and Remote Australia). Based on the calculations to determine the required sample size sufficient to detect a difference in proportions, 28 with a confidence level of 95%, 80% power, and the sample proportions for each group estimated to be 50% and 70%, respectively, the minimum sample size required would be 91. Thus, in order to meet this minimum, the required number of respondents residing in Remote Australia (inclusive of Remote and Very Remote) was oversampled and increased to n = 100. Thus, this constituted the need for a further 89 respondents to be recruited in addition to the initial proposed sample size of 500 respondents.
The final sample size, determined a priori, was 589 respondents, with all sampling quotas adjusted where necessary to reflect this sample size. It should be noted that there was some interlocking of these quotas, which was accounted for in the recruitment process to ensure that all stratification quotas would be met once all respondents were recruited.

| Nationwide administration of the survey
The CATI survey was administered nationally across all major Australian States and Territories, between May and August 2017.
Participation in the study was completely voluntary and respondents received no financial reimbursements.
Individuals were eligible to participate in the study if they were adults (18 years and above); English-speaking (ie did not require the assistance of an interpreter to participate); and had a long-term medical condition, that is an ongoing medical condition that had been experienced for at least 6 months, which required treatment with at least one medicine.
A market research company, contracted by the research team, contacted potential participants by telephone. Phone numbers were generated using a predictive dialler system via a random digit dialling process, which automatically removed wrong numbers (eg. numbers with no pulse tone) and only delivered live phone numbers to the interviewers. Postcodes linked to these generated phone numbers were used to help ensure nationwide administration and that stratified sample quotas were met. Once consent had been obtained and eligibility verified, the survey was administered by the interviewer and participant responses were recorded.

| Data analysis
Survey responses were analyzed using IBM ® SPSS ® Statistics (Version 24). Initially, the data file was examined, and variables were recoded where necessary in preparation for further analysis.
Due to the CATI process, there were no missing data as a relevant response had to be recorded by the interviewer in relation to each question during the survey administration process. In the case of a participant refusing to provide a response, the response was classified as a missing value.
Descriptive statistics were compiled to examine the demographics and other questionnaire items. The Mann-Whitney U test was used to compare continuous variables (which were not normally distributed) between two groups (eg. those who favored an 'opt-in' approach versus those who favored an 'opt-out' approach).
Chi-squared tests were conducted for categorical variables of interest. McNemar's test was used to compare proportions for paired categorical data, for example broad opt-in/opt-out preferences and subsequent preferences for each individual information topic.
The significance level in the statistical analyses was set a priori at P = 0.05.

| RE SULTS
A total of 8683 telephone calls were made to achieve the required sample size of 589 individuals who completed the survey, yielding an overall response rate of 6.8%. However, the survey completion

| Respondent demographics and subjective health literacy
Respondent demographics are shown in Table 3. The median number of ongoing medical conditions that respondents had that had lasted for more than 6 months and required treatment with medicine(s)

| Previous receipt of information outside of consultation times and level of interest in receiving follow-up information
A total of 530 (90%) respondents indicated that they had not received follow-up information from their HCPs in the 6 months preceding their study participation. However, the majority (346/589; 58.7%) indicated that they were either somewhat or very interested in receiving tailored, ongoing follow-up information from their HCP. Of those who expressed interest, the most common reason was that they wanted their HCP to adopt a more proactive approach with respect to information provision. Interestingly, only one in three respondents reported having received insufficient information from their HCP, as a reason for their interest in follow-up information.
Of those interested in receiving follow-up, the majority ex- A lack of interest in receiving such information was expressed by 34.3% of respondents (not at all or not very interested). The following reasons were provided: they received all necessary information during the consultation, they were currently adherent to their prescribed medicine(s), and/or they asked their HCP if they wanted information. Those who reported needing help to read written health and/or medicine information (P < 0.001) and those who reported having had problems learning about their medical condition or medicines due to difficulty reading written information (P < 0.001) were more likely to be interested in receiving follow-up information.

| Preference for an opt-in or opt-out approach to receiving ongoing follow-up information
When asked if an opt-in or opt-out service would be preferred for the provision of ongoing follow-up information outside of consultation times, 281/589 respondents (47.7%) indicated that they would prefer an opt-in service and 293/589 respondents (49.7%) preferred an opt-out service. Fifteen respondents refused to provide a response (2.5%).

| Opt-in
The majority who preferred an opt-in service indicated that all the proposed reasons for preferring an opt-in service listed in the

| Opt-out
Of those who preferred ongoing follow-up information provision to be an opt-out service (n = 293), the majority agreed that all the proposed reasons for preferring an opt-out service applied to

| Associations between demographic variables and broad opt-in/opt-out preferences for receiving follow-up information
Those who were interested in receiving information were more likely to prefer an opt-out approach for the receipt of follow-up information (P = 0.001) ( Table 4). No other associations were found.

| Preferences for receiving specific follow-up information
Respondents were requested to indicate their preferences regarding 7 specific topics of information (Table 1) proposed to be provided as part of follow-up (     For most information topics, respondents generally preferred follow-up to be provided every 3 to 6 months, with the exception of information regarding new and/or serious side effects. About a quarter (28.2%, n = 161) preferred new and/or serious side effects to be provided as soon as possible or once the information was available.
In general, respondents were also receptive to receiving follow-up information via email, with a higher proportion of respondents preferring this method of communication over telephone or SMS for all information topics (Table 5).

| Broad preference for an opt-in/opt-out approach compared to preferences related to specific information topics
When comparing the proportions preferring an opt-out approach when first asked and their preference in relation to a particular topic itself (Table 6)

| D ISCUSS I ON
The findings of this study suggest that there is some interest among the Australian public in receiving information from their HCPs about their medicines and medical conditions outside of consultation times. Opinions, however, were divided regarding an opt-out or opt-in approach. Factors that may have influenced preference for an opt-in or opt-out approach appeared to be specific to the type of information being received.
Despite only a small proportion having received follow-up information previously, most respondents expressed interest in receiving ongoing follow-up information tailored to their needs.
Notably, respondents who had received previous follow-up information were more likely to be interested in receiving follow-up information outside of consultations in the future. Those who selfreported needing help reading information, and difficulty learning Some respondents indicated that they were not interested in receiving the information; their non-interest was noted and the remaining sub-questions relating to their preference(s) for provision of follow-up for that specific information topic were therefore not relevant.

TA B L E 6
Comparison of broad opt-in and opt-out preferences in relation to preferences for specific information topics a a. about their medical conditions and/or medicines due to reading difficulties, were more interested in receiving follow-up information.
These findings highlight the need for information, expressed by patients, which has been noted in past research, [29][30][31][32] and the fact that HCPs are not meeting the needs of patients by providing tailored information targeted to their perceived needs. 29,33,34 Importantly, the results show that the groups who would be more interested could be targeted for opt-out information provision.
General practice standards in Australia have addressed the engagement of telephone and/or electronic communications under appropriate circumstances and within clearly defined parameters between the practice and their patients. 35 Thus, although an opt-out approach to provision of follow-up is not currently embedded within the Australian health-care system, there is opportunity for GP practices to proactively engage in the provision of tailored information outside of consultation times to help improve existing information exchanges. This also extends to other international contexts where opt-in approaches to further information provision are presently implemented.
On the other hand, respondents not interested in receiving follow-up tailored information believed that they already received the information they needed during consultations, and/or that they would not benefit from further information since they were adherent to their medicines. This is consistent with previous studies relating to health and/or medicines information seeking. 36,37 Similarly, preference for follow-up information provision by their GP by a majority of those who were interested in receiving follow-up is also consistent with the literature regarding patients' key sources of health and/or medicines information. 37,38 Respondents in the present study were almost equally divided in their preference for the receipt of ongoing follow-up information outside of consultation times and provided similar rationale for their choice, as observed in previous focus group discussions. 18 As would be expected, those participants who were interested in receiving follow-up information were more likely to prefer an opt-out process, since this would ensure that they received the desired follow-up information.
Preferences for an opt-in versus opt-out service differed in relation to specific topics of information. A higher proportion of re- Although current Australian legislation requires consent to be given by patients to receive follow-up information from HCPs (ie an opt-in system), the survey results suggest that a move to an opt-out process may be well received by the Australian public. In a patient-centered health-care system, patients have a right to receive information to assist them in optimal self-management and care. Moreover, patients should take responsibility and exercise autonomy when deciding whether they want or need the information being offered, rather than rely on HCPs alone to make that decision for them. An opt-out process may encourage more informed and shared decision making in health, in partnership with patients and their HCPs. Although GPs were consistently nominated as preferred follow-up information providers, previous studies involving cancer patients/survivors found that the specialist was the preferred HCP to provide follow-up. 17,[40][41][42] Therefore, preferences for providers of information may also be condition dependent. This, together with receptivity for an opt-out approach to follow-up information provision, should be further evaluated as part of future research.
This study had several limitations. The overall original sample stratification target quotas were met, apart from the age quotas where a lower proportion of persons aged 18-34 years participated in the survey. As the survey did not screen for the level of health literacy, and therefore did not actively recruit people across a range of health literacy levels, the study findings will not be generalizable to certain subpopulations such as people with low to poor health literacy. This survey did not intend to collect data on the participants' specific medical conditions. Thus, it is not possible to determine the influence of medical conditions on the level of interest and preference for an opt-in versus opt-out approach.

| CON CLUS IONS
Members of the Australian public are interested in receiving information outside of consultation times, with this information being tailored to their needs. However, preferences for an opt-in or opt-out approach were clearly divided.
Previous lack of receiving follow-up information coupled with interest in receiving follow-up information from their HCPs indicates that there are unmet needs among patients for further information.
However, further exploration as to whether there should be a transition towards a blanket opt-out approach is required prior to implementation. This exploration should involve all key stakeholders, such as HCPs and policy makers, and should also consider resource needs.

ACK N OWLED G EM ENTS
The research team wish to thank all study participants who responded to the surveys.

CO N FLI C T O F I NTE R E S T
At the time of the study, Stephen Robson was employed by the Healthy Thinking Group P/L. Vivien Tong, Ines Krass, and Parisa Aslani have no conflicts of interest to declare. Administration.

DATA AVA I L A B I L I T Y S TAT E M E N T
The study data are not available as the respondents to the survey have not provided consent for data to be made available beyond the research team. Furthermore, the research team only has the respondents' consent to publish de-identified group data.