Listening to bodies and watching machines: Developing health information skills, tools and services for people living with chronic kidney disease

Abstract

When patients need information, they may visit a doctor, ask a nurse, or look online. But these are not the only sources of information used by patients. This paper examines discussion threads in online renal support groups to describe how people living with kidney failure conceive of help, information and support. I use Actor Network Theory to locate unexpected sources of information and authority such as information obtained from lived experiences, from bodies or from the machinery used to maintain health. By virtue of being embedded in networks of interrelating elements, the patient emerged as a potential information locus: potentially able to confirm information such as medical measurements, make measurements themselves, generate information related to their own sensations, and summarise information about the trajectory of their illness. Implications for practice include an expanded conception of health information literacy for patients, and expanded possibilities for health information service provision.

Keywords:

• information behaviour
• information needs
• information literacy
• health information
• actor network theory
• online discussion groups

Introduction

A person with kidney failure visits a hospital to see a kidney specialist. They talk. The specialist orders a blood sample. Later, the patient is called back to the specialist who changes their prescription of treatments and medications. On the surface it is a simple arrangement, but one that required interactions between complex elements: the knowledgeable specialist, the skilled nurse who drew the blood and the technicians who analysed it; also the machines used for analysis, which are sophisticated and expensive. They produced a set of results, which were added to the patient's record, and compared to previous results for that patient. And of course, the patient came to the specialist and told the story of their recent symptoms. The process is a network of interacting elements: the doctors, nurses and technicians, the machines and equipment, the test results and the medical record, were all involved in producing a new understanding of the patient's health, and contributed to modifying the prescribed treatment. In this article I am going to unpack this network further into different constituent parts, to examine the ways that people with chronic diseases like kidney failure understand their health and manage their illness for themselves, including how they locate and evaluate information relevant to their care.

The analysis draws on data from online discussion forums for people with kidney failure. I explore sources of authoritative information for patients, some of which have turned out to be objects or experiences. I first identify a range of significant informational elements mentioned in the online interactions, and then demonstrate how informational elements appeared in discussions, entwined with and inter-relating with each other. Lastly I explore how information professionals could support renal patients and their families who wish to take a more active role in perceiving, understanding and using this expanded range of information related to their health and treatments.

Chronic kidney disease as a setting for meaning making

At the end of 2009, one in nine Australians over the age of 25 years had signs of chronic kidney disease, and over 18,000 Australians were receiving treatment for end-stage renal disease (ESRD) (Kidney Health Australia 2010, 2012b). Worldwide in the same year 38,000 Canadians and 871,000 US citizens were being treated for ESRD (Canadian Institute for Health Information 2012; National Kidney and Urologic Diseases Information Clearinghouse 2012). In Australia the costs of treating end-stage kidney disease from 2009 to 2020 are estimated to be around $12 billion to the Australian government (Kidney Health Australia 2012a). A report published by Kidney Health Australia in 2010 estimated that increasing the use of home dialysis over 10 years might lead to net savings of between $378 and $430 million for the Australian health system (Cass et al. 2010). Policies and programmes that improve patients' ability to self-care are important in this context.

Kidney failure and dialysis are intense, on-going physical experiences (Tong et al. 2009; Polaschek 2003), with symptoms that must be managed with patient-specific care (El Nahas and El Kossi 2009; Davies and Engel 2009). When the kidneys inadequately filter the blood, toxins build up in the body. The renal patient may feel lethargic, nauseous, itchy and forgetful. When the kidneys produce insufficient urine, excess fluid gathers in the tissues. The body swells, and fluid on the lungs makes breathing difficult, also putting strain on the heart (Kidney Health Australia 2012b; Marcovitch 2010). People sometimes report trying to sleep sitting up to cope with the weight of fluid on their lungs. Dialysis removes fluid and some toxins, and commonly requires either large needles in the arm, or a permanent tube in the belly. For those who dialyse via needles, treatments are commonly three times a week, for four to five hours at a time. Side effects can include pain and bruising from needles, or cramping if dialysis is too fast or too much fluid is removed (Lee et al. 2007; O'Sullivan and McCarthy 2009; Lindsay Waters 2008). On-going adjustments to treatments are made to avoid these kinds of symptoms, and also to avoid complications such as fluid accumulation in the lungs or chronic heart failure (Purcell et al. 2004). The ‘right amount of dialysis’ is maintained by watching for adverse symptoms. Without treatment, however, death commonly takes less than a fortnight (Renal Resource Centre 2008).

Kidney patients describe themselves as isolated in everyday life, because friends and family members may not understand their illness (Tong et al. 2009). However, I found active communities of people sharing their experiences of kidney failure online, helping each other to make sense of situations. An online discussion starts with a comment, question or story written by a member and posted to the group. Other members write posts in reply, which are archived in the order they are posted creating discussions called ‘threads’. As the wife of a kidney transplant recipient I was positioned to be accepted as an appropriate member of these groups, and for my PhD I obtained consent to investigate social sense-making processes in three renal discussion groups.

In parallel with the need for patient-specific care described above, the idea of discovering the needs and constraints of one's own body also appeared frequently in discussions online. Common advice there was that everyone is different. Health professionals must estimate some treatment parameters such as the amount of fluid to be removed from a particular patient in a haemodialysis session. This is an “imprecise science” (Purcell et al. 2004, 635) described in online discussions as “trial and error” (Godbold 2013). In online discussions, renal patients described how they learnt by experience to recognise symptoms and bodily changes and to estimate their body's treatment needs. Information related to changes was of interest to patients as well as health professionals.

Both clinical and sociological medical research often evoke a passive patient and authoritative clinician (Parsons 1951a, 1951b) in which the medical professional is the central source of information and locus of decision-making. When patients exhibit symptoms, nurses may adjust treatment (Purcell et al. 2004). This may be appropriate for people who can be cured of their illness and return to normal life, but chronically ill people face lifelong management of illness and are encouraged to undertake elements of self-care. Some renal literature suggests active involvement of chronically ill patients in monitoring changed symptoms (Jirkovská 2008; Robinson et al. 2011). In the patient-empowerment literature, the doctor-patient relationship is reimagined to an active co-production of healthcare involving both patients and clinicians (Department of Health 2001; Hartkopf Smith 2009; Commonwealth of Australia 2010). Rather than suggesting that patients should be invited to become involved in their care, in this article I describe situations in which patients are already involved. This tends to be a cautious involvement on the part of patients: discussions online suggest that patients are aware of power differentials between patients and clinicians, and are strategic in their interactions (Curtin and Mapes 2001) to avoid becoming known as ‘difficult’ (Stockwell 1984; Werner and Malterud 2003; Hor et al. 2013).

Being ‘informed’ by things and situations

To balance our accounts of society, we simply have to turn our exclusive attention away from humans and look also at nonhumans. (Latour 1992, 227)

According to the French theorist Bruno Latour, sociologists have trouble explaining how society holds together, and so they invent forces such as ‘culture’ to explain why practices continue as they do (Latour 2005, 1992). He suggests that attention be paid to what he calls non-human actors like tools, buildings and machines. I suggest that information researchers should make a similar move. For example, research into information needs, seeking and use often refers to the importance of ‘context’. Context is viewed as a setting for action, a kind of fixed scene within which information seeking occurs (Courtright 2007). It is a background that is not of the same ‘stuff’ as information and is rarely itself informative. Perhaps we are so fascinated by the information people can get from each other, especially via documents, that we ignore the information that surrounds us all the time. Because this study was located in medicine and considered the perspectives of patients, the ‘informativeness’ of bodies, objects and situations became unmistakeable. When Olsson (2010) explored the primacy of the body as a sense-making site, he described an entwined relationship between bodily experience and people's incorporation of understanding. This article takes up that perspective of interrelated context and understandings – of information as necessarily contextualised and evolving (Dervin 2003). I do so by talking about the elements that circulate in a situation (Latour 1996) or elements of interest, to understand how elements that might be called ‘contextual’ interact in situations in informative ways.

Earlier, I described a network of elements involved in the provision of healthcare to a kidney patient: the doctor, the patient, the blood sample and so on. Usually doctors are imagined to be at the centre of such networks because patients come to doctors, doctors order tests, the results return to doctors, and doctors prescribe treatments. It is of course possible to describe the situation differently, for instance positioning blood or the kidneys at the centre of an emerging illness. I am going to put the patient at the centre. I do this to show that it is a very productive perspective. I reposition patients as a central source of information because they ‘are’ the body in question. I also position patients as central in the practices of medical care, for three reasons: because they move in and with the ill body, in ways which slow or assist healing processes; because they self-care; and because they decide when to bring changes to the attention of medical professionals (Mol and Law 2004; Mol 2002a). I use this perspective to explore the range of informational processes potentially available to patients due to their position of simultaneously knowing about and being the body in question (Mol and Law 2004), and I consider how information professionals might help patients develop their understandings of medical situations and their ability to communicate with health professionals.

Research design

A goal for my PhD was to study understandings as they were contingently co-created between people in social settings (Garfinkel [1967] 1984). Therefore data from ‘live’ interactions were suitable, such as the discussions in online groups, while retrospective interviews were not (Rawls 2008; Fox 2008). To understand how meanings developed online I spent almost two years observing and contributing to discussions in three online groups. Although I asked no research questions outright, my identities as researcher and renal wife remained visible to members via my signature and profile. Consent to join the groups was negotiated with moderators, while consent to use posts was requested by showing proposed writing to contributors, obtaining feedback and permission on a case by case basis. Drafts were then provided to the groups for further comments. Ethics approval for the project was obtained from the University of Technology, Sydney.

There were hundreds of posts a week in these discussion boards, with archives going back to 2005. To create a sample of data I copied all posts made during a week in May 2011 into a database and analysed those posts which contained content written by contributors about renal failure. This created a convenience sample of 787 posts from 147 people. I first used the sample to describe the population of contributing members in the three groups. In that week 74% of contributors were renal patients, 18% were people living with patients and 2% were health professionals. Most contributors were from first-world English-speaking countries: 60% from the US, with others from Australia, the UK, Canada and New Zealand and the Caribbean. Of the contributors, 58% were female and the average age was 45 years. I also used that sample of 787 posts to quantify the incidences of themes in discussions. I produced a thematic analysis to identify the themes that appeared to me to be common in the data, and then a content analysis in which I quantified the frequency of particular themes of interest to me. This discussion is based on part of that content analysis. Here, I describe how I analysed sources of help, such as information sources, and how I then explored attributions of authority. The rest of the PhD involved other analyses that do not form part of the discussion in this article.

Enlisting the help of others

I first explored what people named as sources of information, help and support. This followed Dervin who identified help as an important element of sense-making processes (Dervin and Frenette 2003, 242–243). She investigated sources of help by asking people: Did this help? What might help? As my research did not involve interviewing people I instead analysed existing interactions between peers, noting what they reported to each other as helpful, informative or supportive.

Following Latour (2005, 1992) I included both human actors and non-human actors as potential information sources. Human actors were health professionals and patients and each discussion group as a whole, referred to as “here” in posts—“you'll get lots of help here”. Non-human actors included objects and measurements. I also noted references to contributors' experience including physical and non-physical experiences. It could be argued that experience is both human and non-human, an outcome of interactions between human(s) and other actor(s). Regardless of that ambiguity, people specifically referred to having experience as significant, so I included experiences as an element of interest in the analysis. I coded both direct references to ‘having experience’ and examples of particular experiences in interaction with other themes, such as when a post described obtaining information through experiences: “I got a terrible cramp and realised I was dialysing too fast”.

I was uncertain of how to tease help, information and support apart, so I allowed contributors to define them and noted when they said they found support and when they said information could be or was obtained. Sources of help included when people said they found that something helped, but also when people described having obtained help, such as “the nurse fixed my dressing” or “calcium binders help me control my phosphate levels”. In this way the distinction between information, help and support came from the words used in posts. Help seemed to relate to practical needs while support seemed to be used when emotional needs were concerned, but this distinction was not always clear in posts, nor is it necessarily useful to try and separate them.

Figure 1 shows that information, help and support came from a wide range of sources rather than people relying on professionals as the sole or most important source. For instance, practical help most frequently involved objects such as medications and tools. This suggests that the objects created to help people live with this chronic illness are working and are helping people outside of consultation or clinic hours (Latour 1992). Measurements and the discussion boards were frequently cited sources of information. Though they were rarely attributed as sources of support, professionals provided practical help and information. It could be argued that objects provided by health professionals are an indirect measure of professionals' help, but it was not always evident from where particular tools or equipment were obtained. I coded posts according to what was said, counting instances of professional help only when clinicians were mentioned, which brought forward differences between objects and the clinicians who might have provided them. For instance, being available out of hours and having the potential to be used in creative ways are two ways in which objects were characterised differently from clinicians.

Figure 1 Sources of information/help/support in posts.

Deciding who you can rely on

From the earliest analysis, I noticed frequent references to control and authority, and I knew that relations between actors can create temporary sites of authority or power (Law 1992; Latour 2005). So I undertook a further analysis linking a range of actors to attributions of authority, and attributions of power. When actors were described as able to do something or to stop something from happening, I coded it as an attribution of power. Examples included the surgeon who chose where a dialysis tube would be located in the belly; the machine alarms that kept someone awake all night; and the headache that stopped someone from working. Attributions of authority were instances where a particular actor was confirmed as ‘knowing’ or where they had the last word—for example consulting measurements such as blood levels to decide if you are well, or being reassured about a situation because a doctor said it was alright. I also noted when people cast doubt on the authority of an actor.

Looking only at the human agents (clinicians, patients and the discussion groups) on the left of Figure 2, you might see that health professionals were often seen as powerful and authoritative.

Figure 2 Attributions of power/authority in posts.

However once non-human actors were taken into account, the landscape changed. Objects emerged as powerful—able to make things happen—far more frequently than clinicians. This is unsurprising since they included the unwell bodies of patients as well as needles and medications. Again, if a needle damaged a vein I coded it to the needle unless the patient blamed the nurse using it. Meanwhile, an impressive range of other actors emerged as authoritative. Experiences were noted as authoritative more often than health professionals. These were situations where contributors reported that they know something worked because they tried it, or that they knew there was a problem because of a sensation they noticed.

Practitioners were doubted in posts more often than any other actor, for complex reasons which did not eclipse their continued authority (discussed in Godbold 2013). Doubting experience came about, for example, when people suggested that they might need dialysis less frequently because they skipped dialysis and ‘felt fine’. At these times, others posted to suggest that the good feeling could not be trusted because other, unfelt changes were happening in the body. These were times when the authority of measurements might come into play. Blood tests could be relied on to demonstrate dangerous chemical changes in the body requiring dialysis to restore relatively healthy balance. Measurements were rarely doubted in posts—perhaps because they have the objective aura of science (Latour 1993). Apparently patients both could and could not trust the experience of their body. Compliance with biomedical approaches to treatment, represented by measurements, had the final word in discussions online. I also noted times when patients used measurements to argue with clinicians, demonstrating the distinction between these two elements by suggesting fallibility in clinicians even when biomedical paradigms were unquestioned.

In summary, Figure 2 shows that power and authority were attributed to a wide range of actors, rather than being only located with health professionals.

Taking notice of detail

People described how they learned, like health professionals, to ‘see’ into the body with measurements such as blood test results and body weight (Foucault [1963] 2003). They also taught each other to understand the body via sensations, like a needle going correctly into a vein, or the feelings associated with low blood pressure just before one faints. By paying attention to sensations and differences, people became more aware of them, and better able to respond to or articulate them (Latour 2004). They used measurements and sensations as informative elements that they brought together to justify or question how they understood situations. As Figures 1 and 2 gloss over details, I now explain how people interacted with objects and measurements. An example comes from the following quotes from a person learning to put in her own needles:

i have to learn what different pains mean! it's ok to get pain, cos it means in one way that something isn't right. (Patient, Australian Dialysis Buddies)

Later she commented:

when i pushed back, got a bit of a swelling and it hurt a bit. so i figured i possibly had put the tip of the needle [right through the vein], so pulled the needle back a little bit, and then there was no probs. (Patient, Australian Dialysis Buddies, 6 days later)

This is an example both of ‘experiential authority’ and ‘informational experiences’ because here the sensations experienced by the patient, connected to her memories of previous attempts, tell her whether the needle is well placed. This example brings forward what is valuable about the informational elements available to the patient. I could say that information arose out of this context, but it is closer to say that the context was informational, and closer still to say that the interactions between related elements were informative, ending distinctions between information and context. The experiential information was immediate and specific, and allowed flexible responses and flexible care. This matters because in medical scenarios, timely recognition of change matters—and, from her central position in the context of care, this patient was in a key position to be able to recognise relevant changes.

Terms popular in information studies research, like ‘information needs’, ‘information sources’ and ‘information seeking’, have a tendency to isolate parts of what is going on, while silencing others (Law 2004). This creates an impression of separate activities. Information needs might be imagined to occur at one point, leaving information seeking until it takes place later. But in the heat of situations such as the one described above, what we might call information evolves, responds and changes in real time (Dervin 2003; Garfinkel [1967] 1984; Sundin and Johannisson 2005). As this patient learned to insert her needles, information needs, seeking and sources were interrelated and emerged from the range of elements in the situation. Concentrating only on information belies the richness of situations, where remarks, pains, machine alarms, memories, determination, and clumsy fingers all interact. Sometimes, patients were not so much information-seekers with information needs arising out of a context, as people dealing with meanings that were contingent within situations as they unfolded.

The advantages of expanding the network of care

I began this paper by describing some networked elements involved in healthcare, taking a traditional perspective with the doctor at the centre. I was inspired by Latour and other social theorists who have explored how human and non-human elements become interrelated, creating actor-networks; their work is referred to as Actor Network Theory (ANT). They suggest paying attention to how these “networks of heterogeneous materials” (Law 1992, 381), including humans and non-humans, are enrolled in the production of things like ‘knowledge’, social institutions, and healthcare. Despite on-going tensions between elements, networks manage to hold a recognisable form and be relocatable from one location to another: Latour called them immutable mobiles (Latour 1987). He examined the tensions between elements, and the continuous work required to keep networks together and make them look durable. I became interested in the tensions between two related actor networks (as a naïve shorthand I call them the doctor-network-of-care and the patient-network-of-care) and noticed how the elements involved seemed to enjoy fluid as well as durable relationships. In this I am inspired by descriptions of fluidity put forward by Annemarie Mol and John Law (Mol 2002a; Mol and Law 1994). Mol and Law described networks as “fluid” when they are adaptable enough to make up for changes or missed elements without apparent discontinuity. In some ways care of dialysis patients has this kind of fluidity. For example, the doctor-network-of-care I described could have contained different elements, such as the results of a biopsy instead of a blood test, involving different machinery, skills and equipment. A decision to change medications could be based on all the elements I listed or it could be based on only some of them. Care was fluid in another way: the patient might change their medication schedule, change what they eat, or even (for home dialysis patients) change their dialysis schedule without first consulting clinicians. Sometimes (probably more often than we suspect) such small changes go unnoticed by clinicians, meaning that the network has exhibited the adaptability described as fluid-like by Mol and Law.

But an element that is essential in all configurations of care is the patient. We know already that patients' sensations are a key element of care, and that they are generally only accessible if the patient describes them. I presented a patient-centric network of interconnections between their bodies, their sensations, their ways of doing things like self-treatment and compliance, and their memories of their history of illness and care. Paying attention to the perspective of the patient brings new elements to light that patients can access easily and others cannot. These are elements that are located in the patient-network-of-care. Figure 3 below, displays a sketch combining the two networks I have described.

Figure 3 Elements involved in patient safety (foregrounded and backgrounded).

The left-hand side of Figure 3 shows elements of the doctor-network-of-care, with doctors in the centre. On the right and in the shaded area are other elements often referenced by patients in the online discussions, the patient-network-of-care. It is significant to note that two fundamental elements of care are located in the patient network: the patients' sensations, which may begin and modify care, and the patients' ways of self-treating, which affect how care unfolds in everyday life and in the long term.

ANT theorists try to develop understandings of how and why networks hold together, and are less concerned with trying to manipulate networks. Nevertheless, Figure 3 suggests an intriguing lack of certain relations; for example, the separation between patients and their medical records, between patients and equipment, and between health practitioners and patients' companions. Perhaps not all elements benefit by being related. Renal medical knowledge is complicated. Patients who question their treatments can annoy doctors (Broom 2005) and can be wrong (Kopelson 2009) and learning about medicine frightens some patients (Tuominen 2004). Yet there are recognised benefits to patients and health professionals when patients are involved in their own care. Advocates of ‘empowered’ or ‘active’ patients include national health departments in the UK, US and Australia (Department of Health 2001; Hartkopf Smith 2009; Commonwealth of Australia 2010). These government agencies have begun developing policy and programmes to support patients who are able to understand and have an active role in their treatments. They argue that such patients are safer (Hartkopf Smith 2009) and healthier (Department of Health 2001). Patient safety and good clinical outcomes are beneficial to both patients and health professionals. This means that both patients and health professionals gain by patients being willing and able to communicate information, particularly that information to which only patients have access. Patients and clinicians collaborate whenever patients describe their sensations to doctors, but this is a basic collaboration which sometimes necessitates that patients pretend they have nothing to contribute. Meanwhile active patients who are able to self-care are not ‘created’ by training so much as sustained by on-going, deliberately collaborative relations with practitioners (Mol 2002b). So becoming an active patient collaborator is not something that a patient can do alone: “the active patient is an emergent relationship [between patient and practitioners]” (Heldal and Tjora 2009, 1).

There is much more to be said about the issues raised by Figure 3, but the main questions with which I will conclude this paper concern the implications for patients and families, and how information professionals might support patients to develop the elements in the patient-network-of-care.

What this means for patients and families

Health practitioners see patients for only a fraction of the hours in the week. During the rest of the time, chronically ill people continue to deal with their illness. Hence, the importance to patients of information sources other than practitioners. Patients have a contingent, responsive relationship with real-time information relating to their treatments. They also have sole access to some of their body's experiences. The connection from experience to authority is valuable to renal patients and their companions because over time and with experience they can develop their own authority. This can be supported by encouraging contact with experienced peers, and particularly by encouraging patients and their companions to view their experiences as informational.

By virtue of being embedded in context, some patients can operate as an information locus with access not just to information about how they feel right now and how their illness and their lives are interacting. Some patients or their companions also collect information such as medical measurements, are able to summarise information about the trajectory of their illness, and are able to remember what medications or treatment settings they should receive. Given changing staffing and the segmented disciplines involved in renal care, it is both logical and valuable that patients and their partners keep track of the details and history of their own medical case. Contributors to the online discussions advise each other to be proactive in gathering and understanding such information. They perceive active information gathering as key to maintaining safe and coherent care.

One barrier against doing so is medical equipment because it is not designed to be user-friendly for patients, and is often inaccessible. For example, patients often cannot read their dialysis machine from their dialysis chair. Another barrier is health professionals (Godbold 2013; Hor et al. 2013). Some renal health practitioners do encourage patients in this way (Lidz, Meisel, and Munetz 1985) particularly in the training of home dialysis patients (Keeping and English 2001; McCarley 2009). But according to accounts online, other health professionals need careful management by patients if they are not to become offended or suspicious of people who want to ‘know’ too much about their care. Moreover, patients are not necessarily motivated to ‘be’ this kind of information locus, nor does everyone have the skills to do so. The need for skills, strategies and user-friendly equipment suggests areas of potential assistance that could be provided by information professionals.

Implications for information professionals

In the remaining paragraphs I explore how information professionals might help those patients who so desire, to capitalise on their central position and become more able to make use of the ‘information’ that comes their way. However I mainly approach this as a good thing. Ethical, legal and interpersonal problems raised by this rather neoliberal perspective on the patient's role also require exploration by information researchers and sociologists (see for example, Veitch 2010).

First are questions related to information provision. The traditional role of information professionals is to archive and provide access to information (Davis and Shaw 2011), in particular by developing and working with information systems or within “memory institutions [… such as libraries, archives and museums, which] organise and provide access to our cultural records” (Davis and Shaw 2011, 207, quoting Hjerppe 1994). We think of our role as providing access to information or knowledge rather than providing access to experience or skills. Meanwhile, Web 2.0-style self-publishing, including blogs and online discussion groups allow people to share experiences and consult about skills, sometimes posting videos in which they demonstrate techniques. Participants commonly turn to online discussion groups of various kinds for useful information and support (Armstrong and Powell 2009; Sharf 1997; Bacon, Condon, and Fernsler 2000; Bonniface and Green 2007). In contrast with fears that patients may go renegade with self-sourced information (Broom 2005), the online communities I studied maintained dominant biomedical perspectives on treatment (similarly noted by Copelton and Valle 2009) and the ultimate authority and power of trusted clinicians. Meanwhile, research exploring the quality of information and reliability of support provided online remains relatively rare (but see Savolainen 2010, 2011).

Information professionals possess a range of skills that could be used to help connect people with the experience of their peers. They could direct people to interactive resources such as online discussion groups, and could also develop archives of links to online groups, blogs and similar sources of patients' stories. Information professionals could help people to set up their own blogs or groups and help people to learn how to participate in them. Reference support could be of use to large discussion groups, for example by fact-checking or developing wikis of frequently asked questions, hot topics and popular resources. Existing interfaces could be improved, for instance by developing better search functions. Discussion group software could also be developed which can identify common questions and useful answers (Hartzler and Pratt 2011).

Second are implications for an expanded conception of information literacy, noting some patients' desire to understand their own health and measurements. Many libraries already provide literacy training with the goal of improving people's abilities to find and use information, developing their search skills and capacity to recognise authoritative information. Patient-liaison-librarians could provide training in how to find and interpret medical information and how to judge the quality of information in online discussions. Most renal departments provide orientation workshops and training in self-care to patients and their partners (Bergstrom, Barany, and Holm 1999; Fox and Kohn 2008); information professionals could provide complementary support and training (Lloyd 2010), such as how to interpret medical statistics and blood results (Kamal and Burkell 2011) and strategies for the management of medical paperwork and schedules (Olson 2009). Similar literacy training may be useful to people with other chronic diseases, such as diabetes. Since patients are already going online for information, some of this medical literacy training could be provided online, either as online tutorials or as online classes similar to Kidney School (Schatell 2002; Freier et al. 2010). Kidney School provides general knowledge about kidney failure as well as techniques for self-care; such resources could go further by moving beyond the provision of general information toward the development of patients' abilities to specify and tailor understanding.

Last is the potential for developing user-friendly medical equipment for use by patients who want to home test or monitor their treatments in-clinic. Examples of information management technologies which support the complex needs of people with chronic illnesses are the blood glucose meters carried by diabetes patients, or phone applications which keep track of complicated medication schedules and provide reminders to users (OnTimeRx 2012). The design and usability of equipment makes a difference because when patients cannot access or understand their equipment they are also excluded from conversations about it.

Conclusion

In medical scenarios, the timely recognition of changing ‘relevant information’ matters—and with their central position in the context of care, the patient can function as an attentive information locus. Paying attention to kidney patients' perspectives I have identified elements commonly brought together by patients when discussing their own care, recognising the information available to patients from sensations, bodies, measurements and machines. I suggest that it is worthwhile encouraging patients and their partners to develop their record-keeping skills, their recognition and descriptions of sensations and their capacity to communicate with healthcare professionals. It is similarly worthwhile to provide services and tools that facilitate their connection to peers and their routines of self-care. Such opportunities have recognised potential to improve clinical outcomes for people with chronic illnesses. They are worth exploring—both in terms of how to deliver support and in terms of equity issues for people who are not well placed for this kind of active role in their care.

Acknowledgements

The author gratefully acknowledges the support of IHateDialysis, Australian Dialysis Buddies and KidneyKorner, and the knowledge and generosity of the members of those forums, as well as the advice and support of her supervisors Michael Olsson and Rick Iedema. Thanks to the anonymous reviewers for their many useful suggestions and to Suyin Hor, Elizabeth Hanley and Jessie Lymn for their feedback on drafts of this paper. This research was supported by an APA Scholarship from the Australian Commonwealth Government.