Can health promotion also do harm?

In Austria there is a saying Hilft’s nix, schadt’s nix (literally translated ‘Doesn’t help, doesn’t harm’)—meaning, even if something does not help, it may be considered a well meant try, which is not going to do any harm either. This saying gets widely applied to all sorts of contexts (you could say that Austrians are rather relaxed): ‘I’ve started doing yoga—doesn't help, doesn’t harm, eh?’, ‘I’ve changed the batteries—doesn't help, doesn’t harm [shoulder shrug]’. It seems to me that this mindset is often applied to health promotion interventions—they might not always help (everyone), but they surely will not do any harm, will they? Health promotion interventions are meant to promote the health of individuals, communities, organisations etc., to empower, to build capacities, to be sustainable, so simply put: to help and do good. But what if some interventions do not only do good, what if they—unintentionally—cause harm? Unintended effects of health promotion interventions are usually not foreseen, can be difficult to observe, but occur none the less. In fields like medicine and biology, the study of, and accountability for, unintended effects is an integral part of assessment processes—there is awareness that an intervention may have effects that were not planned and researchers aim at documenting and preventing them. In the field of health promotion so far there is only little awareness. Health promotion interventions are by definition complex interventions: they should be holistic, multi-strategic and set on several levels (Hawe et al., 2009; Mittelmark, 2014). They are often settings-based, thereby understanding the specific settings (e.g. schools, hospitals and cities) as dynamic open systems where an intervention does not have a clear foreseeable outcome but can cause a series of effects (Dooris, 2006; Rosas, 2017). There is awareness that effects and outcomes of health promotion interventions are generally difficult to measure and that there are not enough evaluations of health promotion interventions, thereby also failing to give some pointers to unintended effects (Nutbeam, 1998; Kliche et al., 2007; Borghi and Jan, 2008). Therefore, while the definition of unintended effects is rather simple—any effect of a health promotion intervention that was not planned and intended—they are difficult to anticipate, identify, and observe within the field of health promotion. Literature on unintended effects of health promotion is rather scarce and fairly difficult to find. It is likely that several researchers and practitioners have observed unintended effects of their health promotion interventions, but few emphasize on them in a publication. Furthermore, even if (some) unintended effects are reported, the terminology is not yet convergent to an agreed language. To give some examples, I have found: ‘unintended effects’ (Mittlemark, 2014), ‘non-health outcomes and spill-over effects’ (Goebbels et al., 2012), ‘adverse effects’ (Krølner et al., 2012), ‘spin-off effects’ (Agarwal and Brydges, 2018) and my colleagues and I have previously written about ‘side effects’ (Gugglberger et al., 2017). Fortunately, not all of the reported unintended effects are harmful. There are unintended effects of health promotion interventions that are actually desirable—in the context of the health of individuals, as well as non-health related effects. Within a lifestyle behaviour change intervention (directed at changing smoking, physical activity and dietary behaviours) Goebbels and colleagues identified several non-health outcomes from the perspective of participants, like increased awareness of health risks, increased body satisfaction, stress reduction and relaxation and increased social interaction (Goebbels et al., 2012). A study of a school health promotion network showed that gaining new skills and new relationships through their health promotion work as well as benefitting from improved infrastructure were some positive unintended effects for the teachers involved in the implementation work (Gugglberger et al., 2017). In a community health promotion intervention that aimed at assessing cardiovascular, metabolic and fall risks of an older adult population as well as offering lifestyle education, relationships between older adults were built and the target group felt more connected (Agarwal and Brydges, 2018). Such positive effects, even if unintended, are most likely welcome and do not have to be prevented. Some of them could probably have been foreseen with more careful planning and mapping of the intervention. In any case, they should be observed and documented once they occur within the context of health promotion interventions, at the very least to give credit to the intervention. Much more problematic are unintended effects that are undesirable; cases where interventions have negative consequences for the target group or people involved in the implementation. In this context, stigmatization is probably the most widely discussed unintended effect of health promotion interventions in the literature. For example, a workplace health promotion intervention observed that the programme had negative effects on judgements of overweight and obese employees (Powroznik, 2017). Riley and colleagues discuss the stigmatization of lung cancer as an unintended consequence of anti-tobacco campaigns (Riley et al., 2017). McIsaac et al. (2018) conclude that school food programmes can contribute to stigmatization and social exclusion. Besides stigmatization, there are some reports of other undesirable unintended effects. Within the study of the school health promotion network it was found that teachers implementing health promotion interventions were often stressed, frustrated and suffered from the additional workload (Gugglberger et al., 2017). A study that investigated the awareness of risk of osteoporosis warned that making women uncertain and worried could be a ‘serious side effect of health promotion’ (Hvas et al., 2005). There have surely been many other unintended effects. There must have been cases of health promotion interventions that have backfired because the target group did not want their health promoted; interventions that were too much, tried too hard and ended up either boring the target group or asking too much of them; and interventions that caused physical or psychological harm where individuals ended up needing further treatment instead of having their health promoted. Following ‘the good’ and ‘the bad’ there are also ‘the ugly’ unintended effects. Some examples from the (public health) international aid industry can show how severe and even life-threatening unintended effects can be: Daniels et al. (1999) documented that oral rehydration solutions, used to treat cholera patients in Guinea-Bissau, were easily contaminated when prepared in the field, allowing the growth of many types of bacteria. Daniel examined unintended harm of humanitarian aid targeting children affected by HIV/AIDS in Tanzania (Daniel, 2014). She found several examples, including conflict and division between those receiving and those not receiving aid, dependency and decreased self-reliance, as well as humiliation of those receiving aid as a result that donors determined both the nature of the aid and the method of its distribution. Jobe describes unintended consequences of humanitarian volunteerism in post-earthquake Haiti, where new structures, such as free water and food distribution, had huge economic implications and random and uncoordinated donations of medical supplies overwhelmed local resources (Jobe, 2011). In addition to these examples, two papers from related fields can be helpful resources to understand and anticipate unintended effects in health promotion. The first is a scoping review by Allen-Scott et al. (2014) from the field of public health. The authors have created a typology of public health intervention-associated unintended harm outcomes (including: physical, psychosocial, economic, cultural and environmental harm). They also summarize the underlying factors that have led to the unintended harm: (i) limited and/or poor quality evidence of the intervention, (ii) boomerang effect, where the prevention of one extreme led to another extreme, (iii) lack of community engagement (participation and knowledge translation), (iv) ignoring root causes (such as living and social conditions) and (v) the implementation of higher-income country public health intervention in a lower- or middle income country (Allen-Scott et al., 2014). The second publication presents a typology of unintended effects of health communications campaigns (Cho and Salmon, 2007). The authors distinguish between short- and long-term effects, individual- and societal-level effects, effects on intended and unintended audiences, content-specific and content-diffusive effects, as well as desirable and undesirable effects (all of which can be unintended). They also name several unintended effects, including ‘obfuscation’—creation of confusion and misunderstanding, ‘desensitization’—when repeated messages about a health risk desensitize the public, and ‘social reproduction’—when existing social distributions of knowledge, attitudes and behaviours are reinforced. Dealing with complex interventions in complex settings, we have to be aware that not all unintended effects can be foreseeable and therefore not all can be prevented. However, careful planning and mapping of interventions can prevent many mistakes. Breton and de Leeuw (2010) and Van den Broucke (2012) have argued for the deliberate, conscientious and well-argued application of theory to health promotion intervention and research. Theories with appropriate heuristic pathways should enable the researcher and practitioner to identify core and peripheral effects and their pathways, be they positive or negative. In regards to harmful effects of health promotion the primary aim should therefore be to anticipate and prevent or, at least, to monitor them and find strategies to contain or control them. In this context, Mittelmark suggests to enhance our intervention frameworks and models to include the concept of unintended harm (Mittelmark, 2014). One such example is the work of Krølner et al. (2012) that has mapped out ‘unintended adverse effects as well as anticipated potential positive side effects’ in a research model, prior to implementation of the programme. Agarwar and Brydges have chosen an ethnographic methodological approach, arguing that this way they could gain a holistic view of the intervention programme, which also incorporates the observation of, what they call, ‘spin-off’ effects’ (Agarwar and Brydges, 2018). Some will say that a focus on implementing interventions that have already been tested and evaluated and where effects of the intervention should therefore be predictable could be an answer; others will say that it is important to adapt previously implemented health promotion interventions to the specific needs of a setting in order to prevent harmful effects (Bopp et al., 2013; van Daele et al., 2014). In the end, there is not just one way to approach the issue of unintended effects of health promotion—both conceptually and methodologically—but rather several ways. We should embrace the interdisciplinarity of our field and use several approaches and perspectives for the complex issue of unintended effects. Let us look at how other research fields deal with this issue, but let us also incorporate the idea of unintended effects into existing models of health promotion. We need to ask new questions in our research and evaluations; in fact, maybe we should focus evaluations primarily on the observation and documentation of unintended effects. Focussing on unintended effects is in line with the principles of evaluation in health promotion, especially holism, equity and sustainability (Rootman et al., 2001). Such an approach could provide more insight into a programme, including its benefits; it could produce more knowledge and create more accountability than is otherwise possible. Again, it is worth looking at resources from other fields, where in the context of social development programmes, Jabeen has developed the ‘unintended outcomes evaluation approach’ (Jabeen, 2018). The first crucial step, however, will be to increase awareness of unintended effects—within the theorizing, planning, implementation and evaluation of a health promotion intervention, as well as by documenting and writing about them. The more unintended effects of health promotion are published, the more aware we will become of them, thus making them easier to anticipate. Lisa Gugglberger Institute for Health Promotion and Disease Prevention, Vienna lisa.gugglberger@ifgp.at