The FINDER program sets out to Foster Innovation Networks in a Digital Era. Such an ambition cannot be limited to a single industry or a single user group, but is by definition aimed at broader applicability in which social relevance is paramount. In this context, research affiliated with the FINDER project also bridges boundaries to other domains, with recent work by some of the FINDER staff traversing into the field of social science and medicine.
Based on ideas form network theory, we looked into the social interactions as they took place over more than a decade, as we tracked a major English-speaking online health community. Interaction data for the study were collected in July 2017 and consisted of all contributions made this timespan, totaling to 147,561 user profiles and 197,980 discussions, which in turn contain 484,250 replies. Unlike some online health communities, which connect patients with healthcare professionals, the site, which is subject to our study, in intended for non-professionals. The rate of users, who consider themselves as healthcare professionals, is less than one percent. Although a survey conducted by the operators of the community showed that a third of participants claim they visit the website to learn about new and different treatments, the advice sought and given is often not only factually relevant but often emotionally supportive. Similar to a traditional support group, the social exchange amongst peers is encouraged.
The scoping is interesting also in the context of FINDER, that sets out to understand the role of new technology developments in the context of collaborative networks, and we pose, may well infuse notions of effective collaboration and social influencing into more FINTECH attuned online endeavors. A commonality to the ongoings of one of the FINDER project partners – Voleo – hence is rather straightforward, with online peer to peer investment advice banking also on voluntary, discretionary advisory actions by the community constituents.
So what did we find?
When seeking advice online about health concerns, forums dedicated to medical themes are increasingly becoming an appreciated source of information for many individuals. In online health communities, patients can ask questions or otherwise seek advice that is particularly relevant to them. While they may find some of the advice useful, other advice may be perceived as less valuable. By studying the advice-seeking, advice-giving, and advice-evaluation behaviours in one of the largest online health communities in Europe, we looked at what determines which advice is perceived as helpful, and why. Utilising zero-inflated negative binominal modelling, our results show that advice received from others, who have similar predominant interests, is valued more when reaching out for lay expertise. If this advice is given by peers, who can also draw on expertise from other health areas, allowing for a combination of diverse “lay” expertise, the advice is valued even more. Advice provided by those who are quick to obtain the latest knowledge available in the larger community further reinforces these effects. Our findings offer an original view to understand the influence of lay expertise exchanged via online health communities and hold implications for both policy-makers and medical practitioners regarding their approach to patient-initiated use of social media for health-related reasons.
Policy wise our study is interesting also. From various fields, researchers and policymakers have made efforts to understand the behavioural and social causes of human behaviour in health communities. With the advent of online communities, online health communities continue to challenge healthcare professionals in their health advice, being called upon to provide for an alternative of validating opinions. Indeed, also historically, individuals have probably always sought advice about their health status from peers, yet, facilitated by the pervasive adoption of digital technologies and digitally enabled infrastructures in the healthcare domain, patients can now do so much more readily and pervasively, contacting distant peers as well. As a consequence, policymakers and medical professionals should prepare for patients to possibly have already sought extensive advice when they request medical services. These insights may well transcend beyond the boundaries of the health community – fueling new ideas in other tech enabled domains of interaction, we pose.
For policymakers, these insights are of specific relevance. Being better informed about what determines which advice is found to be valuable by people, who seek advice, and why helps in making more accurate assessments on the conditions under which patients will follow through in their treatment trajectory. With studies repeatedly suggesting adherence to medical treatment plans to be a primary determinant of treatment success, being able to better predict individual-level perception of advice quality received holds value. Indeed, failure to adhere not only affects patients but also the healthcare system as it drives costs and constrains healthcare capacity. For policymakers in the health domain, adopting suitable techniques to link into peer advice in online health communities provides an opportunity to improve treatment plan adherence. Also in the initial, orientating phase on health advice, further insight into the relational conditions that drive or constrain patients from using medical advice, can be of assistance to health care professionals such as physicians, pharmacists and nurses. In their daily practice, they are increasingly required to interact with patients that second guess, or at the very least cross check their medical advice via online health communities. For instance, in relation to the worry medical professionals might have about the nature of the advice individuals receive online, our study suggests that the more medical terms are used in a reply, and the more serious or ‘risky’ the condition the advice seeker’s question is about, the less likely it is that the advice is perceived as helpful. However, slightly longer texts are liked more.
Additionally, and from the viewpoint of the medical advisee, the medical advice received through the formal healthcare channels may or may not be correct, or may fail to take account of the advisee’s complete medical situation, and may or may not be correctly interpreted. This view, predicated on the assumption that professional medical advice is to be contrasted with peer advice, is a substantial concern for many, and yet virtual advising on healthcare related themes is increasing rapidly and will most likely continue doing so. While we cannot ascertain what the medical quality of the advice obtained for the advisee is, we are able to determine how much they appreciated the advice. We find that when advisers are able to connect readily to advice-seekers because they have a similar knowledge background, this substantially increases the chances of the advice given being valued (‘liked’). When the adviser also has knowledge from other medical knowledge domains he or she can leverage as well, the advice is appreciated even more. This finding suggests that medical advice seekers value advice givers, who have a degree of empathy, which may be something professional medical advice givers may need to acknowledge somewhat more.
Additional qualitative research – as currently going on within the FINDER project amongst others – will need to confirm this, but our findings suggest that perhaps many individuals, even among those who are actively seeking advice, come to seek initial information and perhaps mostly consolation, and perhaps pointers of where to look for additional information. One speculation could be that advice seekers are preparing for a consultation with a medical professional – this could explain why advisees do not like medical terms in responses. In this latter interpretation, the peer advice individuals receive is not to be contrasted with the advice medical professionals extend. A broader or deeper understanding of what it means for an advisee to “like” a response to their question, and if, or how, peer and professional advice may be interpreted and valued using different metrics, would add to the quantitative study of the patterns of behaviour in the full network that we analysed. With recent advances in digital technologies such as blockchain, virtual and augmented realities, future work may also want to look into the type of online communities in further detail, advancing our understanding on how the advance of digital technology in the health care domain gives rise to new opportunities for patient-peer advice. By the very nature of our single case study design our study does not allow for such comparison, yet helps to trigger thoughts on how the use of digital technology may change the way value is created and extracted amongst patients and health care professionals as they engage in online health platforms.
Paper forthcoming with Social Science and Medicine, coauthored by Jasmina Rueger (Wageningen University), Wilfred Dolfsma (Wageningen University) and Rick Aalbers (Radboud University / FINDER), a prior version was presented at the FINDER Research Excellence Workshop at Atos Amstelveen, The Netherlands (February 10th and 11th 2020) – organized by this ITN project.