Seeking and avoiding information: a qualitative study of the information behaviour of cancer patients

Introduction

During serious illness many patients and their relatives seek information regarding the illness and the impacts it has on their lives (for recent examples see e.g., Treadgold et al., 2020; Ebel et al. 2017). This is also the case for cancer patients seeking to fulfil a wide variety of information needs (Abi Nader et al., 2016, Abu Sharour et al., 2020, Ancel, 2012, Arroyo and Tillinghast, 2009, Germeni and Schulz, 2014). Specifically for cancer patients, Rutten et al. (2005) defines the following information needs: cancer-specific, treatment related, prognosis, rehabilitation, surveillance and health, coping, interpersonal/social, financial/legal, medical system and body image/sexuality information.

Information seeking activities are very important for the patients and may have long-term impacts on their lives. The results of several studies show that information seeking affects health outcomes (Katavic et al., 2016, Kostagiolas et al., 2020) and can influence illness fatalism (Paige et al., 2020). Information seeking activities during critical illness have lifelong impacts (Huttunen and Kortelainen, 2021) and both lifestyle as well as health behaviour can be affected (Zhang, 2012). Many patients feel a loss of control, and information seeking can play a role in helping the patients gain an increased sense of control (Faller et al., 2016, Persakis and Kostagiolas, 2020; Petersen, Jensen and Frandsen, 2021). Adding further nuances to the role of information seeking when facing uncertainty, Miller (2014) argues that information seeking can both facilitate as well as complicate the individuals’ management of cancer-related uncertainty. Cancer patients have access to an abundancy of information related to their illness which makes it difficult to navigate the vast amount of information (Perrault et al. 2020, Chua et al. 2020a) and it may even lead to cancer information overload (Jensen et al. 2014, Jensen et al. 2020).

When cancer patients engage in information seeking activities to fulfil their information needs, they use a wide variety of approaches. There is a spectrum of information seeking behaviour ranging from some patients preferring to seek intensively whereas other avoid seeking information. Information avoidance has received little attention in the existing literature (Song et al., 2021), however, information avoidance is generally used often to actively avoid challenging information (Karim et al., 2019). About 30% of cancer patients do not seek information actively regarding their illness (Carlsson, 2000). Many cancer patients are uninterested or even deliberately avoid information related to their illness. In a large cohort of cancer patients Loiselle (2019) finds that almost 40% are uninterested or even deliberately avoid illness-related information.

Building on the work of Jensen, Petersen and Frandsen (2021) this study aims to explore the information behaviour of cancer patients and how avoidance can be an approach used as part of seeking activities and not as an alternative approach to seeking. Using qualitative methods, we explore how cancer patients actively seek information and, in some cases, cautiously and consciously avoid information that may be relevant, but does not meet their needs.

Related literature

The approaches or patterns of information seeking can be used to characterise behaviour, but they have also in many cases been used on an individual level (Germeni and Schulz, 2014). A recent example is the study by Loiselle (2019) who assesses cancer information seeking preferences by asking participants to characterise their own cancer information seeking preferences. Germeni and Schulz (2014) question the use of these models to characterise individuals in their meta-synthesis including 18 studies. They argue that individuals cannot be organised in groups based on their information seeking behaviour as information seeking and avoidance are not two distinct modes of behaviour. Chen (2021) studies the interplay of affect, cognition, and behaviour over a larger time scale of fibromyalgia patients and finds that that affect and cognition had different effects on information behaviour and this could change during the course of their illness. McCaughan and McKenna (2007), Germeni et al. (2015) and Shim et al. (2016) also stress that information seeking behaviour can evolve across different stages of the patient journey. Furthermore, Germeni and Schulz (2014, p 1371) argue that information seeking and avoidance are two sides of the same coin. Considering seeking and avoidance as closely related although they seem very different goes back a long way. Festinger (1957) argues that information can be avoided while seeking if particular sources are disregarded. In a more modern context, Kwanya (2016) uses the term echo chamber to characterise this shortcut to satisfying an information need. Other conceptualisations are available. Malinen et al. (2018) analyse social media and label the behaviour in which a person actively seeks information that supports their views and avoids other information as selective exposure and selective avoidance. Case et al. (2005) argue that behaviour aimed at preventing confrontation with specific information, also termed selective exposure, is an early characterisation of avoidance.

The concepts information avoidance and selective exposure are closely related. Sweeney, however, argues that the literature on information avoidance seek to understand why some avoid information under conditions of uncertainty or ignorance about the content whereas the literature on selective exposure seek to understand the preference for information known to be consistent or inconsistent with their attitudes, beliefs, or decisions.

Sweeny et al. define information avoidance as follows:

We define information avoidance as any behaviour intended to prevent or delay the acquisition of available but potentially unwanted information. Information avoidance can entail asking someone not to reveal information, physically leaving a situation to avoid learning information, or simply failing to take the necessary steps to reveal the content of information. That is, information avoidance can be active (e.g., by asking someone not to reveal information) or passive (e.g., by failing to ask someone a question that would reveal the information). Sweeny et al. (2010, p. 341)

Information avoidance is thus an alternative to information seeking. Barbour et al. (2012) argue that information avoidance is a communicative response to uncertainty, and it is influenced by situational factors and individual characteristics. It is not necessarily unhealthy, and models of information behaviour should account for avoidance. Nelissen et al. (2017) find that fear of cancer is positively associated with cancer information acquisition and avoiding behaviour.

Sweeny et al. (2010) argues that people have a choice between avoiding and seeking information that is not yet known and potentially unwanted. This distinction supports the recent categorisation of approaches to health information presented by Costello and Veinot (2020) also stressing that they are describing a spectrum of approaches to health information interaction ranging from avoidance to verification

• information avoiders, who close themselves off from health information
• information receivers, who encounter information in the dialysis clinic but do not seek it out
• information askers, who only pose questions about health to their healthcare providers but otherwise do not seek
• information seekers, who actively look for health information both in and out of the clinic
• information verifiers, who seek information and triangulate it among multiple sources

In this model information avoidance is described as a non-seeking behaviour although Song et al. (2021) argue that information avoidance and non-seeking are fundamentally two distinctive types of behaviour that are triggered by different motives and occur through different psychological mechanisms. Information avoidance is also conceptually different from information disinterest according to Nelissen et al. (2017). The spectrum of information seeking behaviour ranging from intensive seeking to avoiding information is also described by Lambert et al. (2009a,b) who identify five patterns of cancer information seeking:

• intense information seekers, who look for as much information as possible
• complementary information seekers, who add to what they already know
• fortuitous information seekers, who consult other patients for health information
• minimal information seekers, who do not look for information on their own
• guarded information seekers, who avoid some cancer information.

The five patterns represent different approaches to addressing the needs of the individual (Sheridan et al., 2020, Germeni and Schulz, 2014). In this model, information avoidance is described as guarded information seekers who avoid some cancer information but not necessarily all.

Methods

We conducted interviews with nine current and former cancer patients to gain a deeper insight into the information-seeking behaviour of cancer patients. Three men and 6 women were interwieved of which 2 are currently undergoing treatment and 7 are former cancer patients.

Byström and Järvelin (1995) argue that interviews enable thorough analyses of information needs and seeking. When interviewing current and former patients, one must take several factors into consideration, as the interviewees can be both fragile and vulnerable. Before the interview, all participants signed a consent document in which they consented to their conversation being recorded and used for research purposes. The document stressed that the transcripts would be treated anonymously. In the data, the interviewees' names were changed to IP1-IP9 and personal information was excluded.

The interviewees were found through the Danish Cancer Society, a non-profit organisation in which one of the authors is actively involved. Neither material nor financial compensation was provided to the interviewees for their participation in the study. Each interview was conducted by a member of the author team without prior relations to the interviewee so that any potential bias would be reduced. The interviewees were between 25 to 59 years old. Even though Danish was not the native language for one of the interviewees, all spoke Danish fluently and lived in Denmark. Therefore, the interviews were conducted in Danish and translated for this publication. Four of the interviewees had previously recovered from cancer whereas two of them were undergoing treatment at the time of the interview. Participants in the study were two men and four women.

Interviews began with an introduction that was not recorded, in which interviewees were introduced to the study and informed they can pause or interrupt the interview at any time. Before the interviewer started recording, the interviewee had the chance to ask questions. Interviews are conducted in a semi-structured interview format for approximately an hour, at which point any comments or remarks are added by the interviewee. Due to the lockdown caused by the COVID-19 pandemic during the time of the interviews and because it was easier to organise telephone interviews were conducted.

In the study, interviews were audiotaped, transcribed, and analysed. Transcripts were read repeatedly to identify the key themes and categories. Transcripts were coded for themes and concepts. Two authors coded the transcripts independently, then compared the results and reached an agreement. The coding focused on people's avoidance of information and their search for information during serious illness. The study used a hybrid process of inductive and deductive thematic analysis where data driven codes were integrated with theory-driven codes. The first round of coding was inductive using no theoretical framework to guide the coding. The next round of coding used the spectrum of information seeking behaviour of cancer patients presented by Lambert et al. (2009a,b). The spectrum by Lambert is specifically developed for cancer patients and the findings are organised using this framework. Other spectrums could have been used and in the discussion, we return to the spectrum by Costello and Veinot (2020) to allow for further perspectives on the data.

Findings

In the study, we identified different patterns of information seeking behaviour among the cancer patients interviewed. First, the patterns of information behaviour are tied to different stages in the illness. Some of the interviewees in this study are aware of their information behaviour being tied to their cancer continuum:

Well, [my information seeking behaviour] was split into two different time periods. […] Before being diagnosed I searched for symptoms and my searches were imprecise and the information I found was useless. I searched to find comfort. […] After being diagnosed I had something to search for. (IP7)

[Selecting information resources more carefully] came after my treatment was completed. I couldn’t cope with people being ill or dying. I just had to trust that it would not happen to me. (IP8)

Information seeking behaviour can thus vary depending on the stage of the patient journey. The information seeking behaviour is a response to a need which can change over time. Therefore, how individuals respond to their needs over the course of the cancer continuum can change and they may use several different approaches to information seeking.

We identified intense information seeking as one pattern. In some of the interviews, participants express a desire to know as much as possible and they seek out information intensely.

My primary strategy was to make sure I was completely covered with information, that I got help from my network I could, that I had googled all I could. (IP1)

I prefer knowing the facts rather than hiding them. I want to know what I am dealing with instead of being spared the merciless information. (IP7)

The thing about closing your eyes to something.... and prognoses, it does not matter at all, because it does not mean that it is something that will affect me. So no, the more information, the better. (IP6)

We identified complementary information seeking where the amount of information obtained is restricted. It is a process focused on getting good enough cancer information. Among the interviewees, some did not feel the necessity to seek information on the illness and primarily relied on the health professionals.

No, not so much. Now, of course, it is a few years ago, but it was all very overwhelming, and I was only 21-22 [years old] at the time ... so in the beginning it went very fast and you already get a lot of information. So in the first 2-3 weeks I could not keep up. Not mentally either, because it went so fast. But when I have had questions, I have just asked [the health professionals]. But otherwise, I think they were reasonably good to inform (IP2).

One of the interviewees actively chose early in the process to avoid searching on his/her own, but to rely on the health professionals for any questions:

I chose from the beginning not to [...] read online. That was probably to avoid.... I'm generally critical regarding online information, of course I googled some things. But it was minimal. Otherwise, I have actually relied heavily on the healthcare staff. Both regarding surgery and radiation therapy, yes... (IP3).

Fortuitous information seeking relies on consulting other patients for health information. In some cases the information seeker may look for personal experiences and discuss aspects that they feel the health professionals may not be able to guide them with.

Well, [communicating with other cancer patients] could give me insights into the experiences from like-minded patients. What they felt and what they had done to get better. […] the health staff cannot give me the same [information] as those who have felt it on their own body. (IP3)

I have been seeking the experiences of other cancer patients. I have ended up in some obscure websites where you cannot really rely on the information. (IP7)

So, I've used [the internet] primarily […] to relate to others. Even if you are very different, you can understand how it feels to be a cancer patient. (IP2).

The interviewees in this study seeking information fortuitously tend to use this approach as a supplement to complementary information seeking or intense information seeking. None of the interviewees relied primarily on consulting other patients for health information, but it was used together with other approaches:

It is much easier to search for information yourself and one can also find like- minded. (IP4)

We also identified minimal information seeking, where the patient do not look for information on their own. The interviewees in this study that did not seek information themselves argued that the need for seeking information on your own is tied to the relation with the health professionals.

So, I just related to what the doctors said. And they seemed calm so that way I did not need to read everything more about it, as it would just make me more worried (IP5).

I tried seeking information myself once, but it made me feel more ill. […]. [The doctors and nurses] were fantastic. They know how much you need to know, and they took the time to talk to me and explain everything in detail when needed [...] (IP9)

Consequently, being satisfied with the treatment and information offered by the health professionals makes the patient feel safe and thus seeking more information may not be necessary. Some of the patients seeking intensely expressed distrust with the health professionals.

And because they do not tell the whole truth in the hospital. There are many things they cannot inform about, or are allowed to, or willing to... I am not sure. (IP6)

I actually think I still believe more in [information] on the internet or I trust my network. It's hard when you already distrust your doctor when this person at some point has said something not entirely correct. (IP1)

Finally, we identified guarded information seeking where some cancer information is avoided. Cancer patients may not be interested in knowing everything they can about their disease. We find examples of guarded information seeking e.g. choice of sources. Lambert et al. (2009a,b) argues that choice of sources characterises the information seeking approach. Cancer patients can select a source because they expect it to give them an answer to their question, thus meeting their need for information. The sources may not necessarily include authoritative or quality-approved information. One interviewee was concerned that the treatment would have serious and permanent side effects in his or her case, and when the personnel at the health facility were unavailable to offer any guarantees, the interviewee sought information elsewhere.

It was worse when I googled the prognoses for facial paralysis, so I actually completely stopped googling it. Then I went to a healer instead, and I do not believe in [healers] at all. But just to give [the interviewer] a picture of how far out you can get. […] Well, it did [give positivity]. That's the worst part. I wish I could say no (IP1).

While the interviewee has no trust in healers generally, the patient nevertheless asked a healer about a crucial aspect of the treatment. It is essential that the patient gets a positive answer to the question about prognosis and if internet searches cannot provide that then an alternative source is identified. To avoid upsetting information, sources of information may thus be selected carefully. One interviewee reflects on the choice of resources when seeking information with particular focus on online support groups:

There was a lot of guess work… Many asked "now, I have these symptoms, what could it be?" and then there were hundreds of suggestions, and I think that could worsen my situation [….] I would get more upset. […] then I would rather ask a doctor who knows my specific situation (IP3).

Several interviewees said they chose the resources carefully to avoid getting upset. However, the resources relied upon vary with the information need, and are therefore not consistent. Discussing with others to learn how they have experienced the illness or treatment may be useful at different stages.

Well, Wikipedia and something like that I avoid ... but otherwise I searched for [information from] former cancer patients and their experiences […] - it was not the disease I searched for, because it is different from one person to another. I'm probably also critical, as no one has the same case as me (IP3)

I joined a Facebook group to meet other people going through the same as me, but after I finished my treatment, I had to leave the group. It was too much for me to see how people I had talked to were dying or relapsing… I needed to believe in my recovery, and I had to be very selective with the information I looked for (IP8)

One aspect not mentioned by Lambert et al. (2009a,b) is the use of an intermediary. Guarded information seeking could perhaps involve the use of an intermediary, someone who can filter the information to avoid the unwanted information. Several of the interviewees are protected from distressing information by a relative who filters the information:

It's a lot about getting something that made sense to me. So, I could cope with being in [a situation] which was so difficult. My relative passed the information on to me in a gentle way because I could not bear to have in-depth information about the illness at all (IP5).

I did not [actively search for information] until later. Not in the first 12 months. But my relatives did. I thought that I would probably get too much information. I needed it in smaller pieces. (IP9)

Consequently, there are many ways to engage in guarded information seeking. Information seeking approaches can be characterised in terms of type, amount and source as argued by Lambert et al. (2009a,b). However, guarded information seeking behaviour can also be seen in combination with other information seeking approaches. It is possible to seek information intensively and still be a guarded information seeker. IP1 is an example of intensive information seeking behaviour while selecting resources carefully and avoiding seeking information on certain aspects of the disease. IP3 is an example of information behaviour consisting of different approaches in combination with guarded information seeking. IP3 is not intensively seeking but when seeking information, although minimally, the patient is very consciously choosing the sources to rely on to match the information need.

Discussion

This study finds examples of different approaches to cancer information seeking. Information seeking by seriously ill patients are typically described as a spectrum ranging from intensive seeking to avoiding information. However, the findings of this study indicates that the approaches can overlap. Some patients seek intensively while consciously avoiding otherwise relevant information and thus seeking guarded. Before turning to the discussion possible limitations need to be considered. First, we interviewed patients at only one point in time and thus some interviewees were asked to recall their information behaviour where others described their current behaviour. There may have been cases of inaccurate recollections, and their view of events may also have changed over time. Second, the sample size was limited and therefore the phenomena observed may not reflect those experienced by all cancer patients. There is a need for more research to understand information behaviour in a more diverse sample of seriously ill patients.

In this study the spectrum by Lambert (2009a,b) was used to organise the findings. However, using the spectrum by Costello and Veinot (2020) would allow us to draw attention to other aspects of cancer-related information behaviour. Both spectrums are developed using qualitative study designs, although involving different groups of patients. They both range from avoidance to intense seeking, although they define the approaches differently and they overlap in different ways. Information avoidance can serve as an example of the different definitions. Information avoidance described by Lambert (2009a,b) involves that only some information is avoided whereas Costello and Veinot (2020) describe an information avoider as closed off from health information. Costello and Veinot (2020) describe two kinds of information seekers that seek little or no information on their own. In this study we see several of the interviewees describe their information seeking activities being minimal and this may add further nuances to their behaviour. However, in this study all patients were offered and thus encountered information at the hospital when undergoing treatment. Costello and Veinot (2020) also describe two kinds of information seekers that actively look for information as well as triangulating information among multiple sources. Again, in this study many of the interviewees describe how they seek information and understanding how they triangulate information among sources can add further nuances. The spectrums are thus different which may be tied to the specific group of patients underlying the development of the two spectrums and therefore in this study we rely on the spectrum developed for cancer patients. However, they are also very similar in describing information behaviour as a spectrum and our findings have implications for the understanding of information behaviour as a spectrum.

The findings in this study have practical as well as theoretical implications. First, the findings confirm that the information behaviour of cancer patients needs to be explored and characterised as behaviour, not people. The approaches should not be used to describe individuals as they may use a different approach at another stage in their cancer continuum. The approaches may also be combined in different ways and therefore, we should not characterise seekers but rather seeking behaviour. Second, the findings indicate that the information seeking approaches do not exclude each other which may indicate that a revision of the information seeking spectrum may be valuable. The spectrum characterises the information seeking behaviour ranging from intense seeking to avoidance. However, adding more nuances could allow for more complex understandings of information seeking approaches. Information seekers can avoid information while seeking intensively and therefore considering the approaches a spectrum may not allow for the complexities in the behaviour.

Furthermore, we find that information can be filtered by an intermediary, e.g., a family member to allow for more protection from overwhelming amounts of information or distressing information. This is also called “by proxy” which refers to an information seeker interacting with information sources through the initiative of another agent (McKenzie, 2003). MacKenzie describes three different kinds of proxy connections of which one is an intermediary or gatekeeper and argues that accounts of interaction by proxy are extremely varied in their characteristics and complex. In this study there are interviewees expressing how the intermediary protects them in different ways and thereby helps them cope. The use of intermediaries to filter information is a promising area for information behaviour research as it would allow us to understand this complex interaction further.

Finally, the characterisation of information seeking approaches addresses type, amount and source as argued by Lambert et al. (2009a,b). Information need is thus not explicitly mentioned. However, as argued by Wilson (2020) it is evident that information behaviour is prompted by an underlying cause and the subsequent information seeking behaviour can be understood through the information need. In the existing literature on information seeking models information need plays a key role (Case and Given, 2016). Wilson (1999, p. 251) states that “information-seeking behaviour arises as a consequence of a need perceived by an information user, who, in order to satisfy that need, makes demands upon formal or informal information sources or services”. Consequently, exploring information seeking approaches may allow for new insights if information need is addressed.

Conclusion

Patients suffering from serious illnesses have many different information needs and fulfilling those needs can improve their health-related quality of life and optimise their health care. Cancer patients use a variety of approaches to satisfy their information needs. During serious illness, information seeking can be a coping strategy. Different ways of seeking information can be identified in an individual, and over time, they may use several different strategies.

Seeking and avoiding have long been considered side by side, and avoidance has been seen as an alternative to seeking. In this study we have explored if they can complement each other. We have found that the two approaches may overlap and the use of one does not necessarily exclude the other. Thus, cancer patients can actively seek information while consciously avoiding information that is relevant but doesn't meet their needs.

Our findings call for further research to explore if our findings are tied to cancer patients or if they can be generalised to some extent. We also need further studies to explain how avoiding information can be part of active information-seeking during periods of significant life change such as serious illness.

About the author(s)

Jesper Gabs Jensen holds an MSc in IT - Web Communication from University of Southern Denmark. His research interests are in information behaviour, cancer information seeking and coping.
Emil Petersen holds an MSc in IT - Web Communication from University of Southern Denmark. His research interests are in information behaviour, cancer information seeking and coping.
Tove Faber Frandsen is professor with special responsibilities at the department of Design and communication, University of Southern Denmark. She received her Ph.D. from The Royal School of Library and Information Science, Denmark. Her research interests are in information behaviour, bibliometrics and systematic reviews. She can be contacted at t.faber@sdu.dk

How to cite this paper