Making Decisions or Enacting Shoulder Care? Challenging Assumptions About Matters of Concern in the Orthopaedic Shoulder Clinic

ABSTRACT

Because of increasing pressure on healthcare resources in many countries, public health policies are increasingly occupied with how to optimise decision‐making processes to ensure medical interventions are appropriate, necessary and aligned with the preferences and values of patients. One prominent example is the concept/model of shared decision‐making (SDM) which target clinician‐patient interactions in the clinic. Such public health goals raise renewed questions about what matters to people when seeking healthcare, which models such as SDM may not be able to fully account for. In this article, we draw on ethnographic fieldwork in two Danish shoulder clinics. We identify specific uncertainties as they manifest in guidelines and the clinic, and draw on pragmatist‐inspired theory, to argue that decision‐making emerges as interactional work in what has been conceptualised as the subjunctive mood, to make treatment proposals acceptable. In this process doubt and hope are inherent elements to be acknowledged and managed, rather than avoided. Thus, we argue that such pragmatist ideals are fundamental to shoulder care. Our findings call for an approach to clinical decision‐making in research and policy that considers the everyday concerns of patients and clinicians as they seek to navigate highly uncertain terrains.

1. Introduction

Has the term ‘shared decision making’ outlived its usefulness? […] Indeed the very use of the word ‘decision’ may be where we have gone wrong in this field. People shy away from decisions about complex issues, and words such as options, preferences, and values are from a lexicon that is unfamiliar to many, and frankly, tough to scope.

(Elwyn et al. 2016, 2–3)

The need for efficiency and prioritisation in healthcare increasingly demands certainty about what works and which treatments to choose. How clinical decisions are made is increasingly of interest to policy‐makers and others trying to ensure the right patients get the right treatment at the right time attempting to counter unnecessary medicalisation, avoiding needless risks and preventing healthcare that offers little value to patients (Pedersen 2015; Raft et al. 2021). Rational clinical decision‐making based on the best available evidence has long been seen as the core of effective biomedicine and care. At the same time, over the last four decades, clinical and public health literature has advocated for greater patient involvement in decision‐making to align medical decisions with patients' individual situations (Borders et al. 2005). Central to these efforts is the model of shared decision‐making (SDM), a concept first used in the U.S. during the 1980s. It has since been incorporated into policy‐making, insurance systems and healthcare organisations, and is gaining increasing momentum worldwide. Specifically, in Denmark, SDM manifests in a large research centre devoted to implementing conceptual and political decisions in all five administrative regions of the healthcare system (Dahl Steffensen et al. 2022). SDM can take many forms. It is sometimes explicated as an approach to care (Montori et al. 2023) and other times as a model for decision‐making that can be evidence‐based (Dahl Steffensen et al. 2022). Despite various interpretations, there is general consensus that SDM remains difficult to implement, and its achievements are ambiguous (Elwyn et al. 2016). The problem has been suggested to be one of organisation, with a need for more robust conceptual frameworks. However, as Elwyn and colleagues suggest in the introductory quote, the fundamental problem might be with the terminology. In this article we demonstrate how SDM fails to fully capture the unfolding meaning of interactions. Drawing on ethnographic material from shoulder clinics, we explore what matters to those involved in clinical surgical consultations. In line with bioethicist Jonathan Ives, we suggest that SDM and its related concepts may not be totally honest about, or helpful, in understanding what is happening in the clinic, and what matters to both patients and clinicians (Ives et al. 2018).

When described as a model, SDM frames clinical consultations as a transaction between healthcare professionals (HCP's) and patients that draws on a vocabulary of evidence, preferences, facts and values (Charles et al. 1999; Dahl Steffensen et al. 2017). Such an approach isolates the clinical encounter and risks eschewing the interlinking of what goes on there, with larger political and organisational phenomena (Kazimierczak 2018; May 2007; Rapley 2008). Anthropological and sociological studies have shown the lived complexity of decision‐making in the clinic as well as the importance of attending to the situated relationality of building trust and performing care (Birchley et al. 2024; Boholm et al. 2013). Sociologist Tim Rapley suggests that, in the context of illness, decisions‐in‐action are distributed among multiple actors, over time and across different spaces such as the clinic and the home (Rapley 2008). Decision‐making is further shown to be empirically difficult to pin down, always observer‐dependent and always about something in particular, usually the thing to be decided (Langley et al. 1995). A study focusing specifically on interactions with SDM in the clinic shows how rather than an idealised transactional process, healthcare professionals engage with multiple strategies and different ways of dealing with tensions and conflicts in specific situations (Cordey 2021).

Drawing on fieldwork from Danish shoulder clinics, this article unfolds matters of concern to clinicians and patients when making decisions about surgery, and how they speaks to ways of enacting care in the context of uncertain evidence. We begin by illustrating how specific uncertainties related to shoulder surgery manifest in research and the clinic. Then we show, through empirical examples, how decision‐making emerges as interactional work where doubt and hope are inherent elements to be accepted and managed, what Whyte calls the subjunctive mood (2005). Surgery then becomes a conditional certainty—a proposal that needs to be made acceptable. We show how decision‐making in everyday interactions in the shoulder clinics emerge as an ongoing and situated establishment of a particular treatment (e.g., surgery) as a plausible idea through management of specific uncertainties, where surgery is entangled with physiotherapy and other treatments in ways which makes it difficult to talk about them as separate options to choose between. Our argument challenges notions of decision‐making in models like SDM and calls for a need to address questions of how decisions are made about health and body, through concrete and lived examples, to reach more precise and nuanced understandings.

2. The Setting: Orthopaedic Shoulder Care in Denmark

This article is based on a research project that conducted both empirical and conceptual studies on decision‐making related to musculoskeletal symptoms and orthopaedic surgery in the Central Denmark Region. Whereas orthopaedics once sought to prevent and correct musculoskeletal deformities through varied treatment forms more closely allied with physiotherapy, in modern medicine, it has narrowed into a surgical specialty focused on invasive procedures, leaving the ‘medical’ diseases of the musculoskeletal system to rheumatologists (Gehrig 2011). Musculoskeletal symptoms such as pain and reduced function are the most common reasons for contact with the healthcare system (Moth et al. 2012). With increasing specialisation in the healthcare sector, GPs in Denmark are increasingly given a gate‐keeping role, following guidelines for when to send the patient for further examination or treatment in a specialised unit. For non‐traumatic shoulder disorders, clinical guidelines suggest a minimum of 3 months of supervised physiotherapy with exercise before considering further referrals (Chaudhury et al. 2010; Danish Health Authority 2021). Along with exercise, treatment typically includes pain medication and steroid injections. These approaches are often utilised before referring a patient to a specialist clinic in the hospital, but they can also serve as intermediate steps or as alternatives to surgery after a referral (Christiansen et al. 2016). In many Danish hospitals, including those where we conducted fieldwork, patients can get a consultation either with an orthopaedic surgeon or a physiotherapist (both specialising in shoulder disease), and diagnostic consultations follow the same structure. If the consultation with the physiotherapist ends with a suggestion of surgery, the patient is further referred to talk with a surgeon. Shoulder disease is commonly divided into traumatic or non‐traumatic, depending on whether a known injury was the onset of pain (Danish Health Authority 2017). In many cases of pain and reduced shoulder function the symptoms may disappear over time without medical intervention. For some underlying causal factors, such as progressive osteoarthritis, there is greater consensus about the usefulness of surgery; however, the medical community still disputes the futility of certain treatments for shoulder disease. Both diagnostics and treatment in the shoulder area are often described as difficult to standardise (Brabston and Bigliani 2014; Leape et al. 2023). At the same time, patients living with chronic shoulder pain suffer from life‐limiting effects, such as sleep deprivation, problems with daily activities, loss of concentration, depression and anxiety (Page et al. 2019). In the consultations we observed during fieldwork, most patients had lived with pain for several years, having explored physiotherapy, exercises at home, steroid injections and sometimes alternative therapies, with varying effects. They all had in common that the pain was still a presence in their lives and affected them to a degree that they found unacceptable.

3. Theorising the Relation Between Decisions and Care: Logics, Uncertainty and Pragmatism

Dutch philosopher Annemarie Mol has argued that a focus on decision‐making and choosing in the clinic, what she calls ‘the logic of choice’, misses the complexities involved in finding ways of living with injured and ill bodies (Mol 2008). She contrasts this logic to a ‘logic of care’. She writes:

It is as if making a decision were like making a calculation. Pros and cons, one side versus the other. In the logic of care this is different. […] After all, addition and subtraction require fixed variables, but in the logic of care no variable is ever fixed.

(Mol 2008, 62, our underlining and bolding)

The quote captures the tension between dominant logics in healthcare systems, between counting and calculating optimal choices and caring for the dependency and contingency that chronic illness and crisis produce. By now, this is a well‐documented theme in the social science literature (Manderson and Wahlberg 2020; Mol 2008; Vigh 2008; Whyte 2012). Science and technology studies (STS) scholars in the field of medicine (Lindén and Lydahl 2021; Mol 2008; Mol et al. 2010; Pols 2015; Struhkamp et al. 2009) have pointed out the importance of attending to the questions, ways of working and moods of attention in the clinic. Here, people and things often relate differently to each other than in policy and research. They have termed these ways of working ‘doctoring’, not to determine a specific activity of the medical doctor (as opposed to, e.g., a physiotherapist, nurse or dietician) but to name the mode of knowing and acting in the clinic that attunes to the contingent, ongoing, creative and uncertain realities of patients, bodies, technologies and diseases on a daily basis (Struhkamp et al. 2009). We draw on this line of thinking to explore the relationship between making decisions and enacting care in shoulder clinics.

The uncertainty of medicine as a double‐edged sword, which entails the potential to both help and create new problems, is well‐known (Whyte 2005; Whyte et al. 2002). Even in treatments that have been thoroughly tested and documented for years with ‘satisfied’ patients, and where evidence is regarded as certain due to successful research, there is always a distinction between the population in the study and the individual in the consultation room (Mol 2002). In the case of shoulder surgery, the uncertainty of evidence and heterogeneity in approaches within the medical community, makes these individual or case‐specific considerations paramount. Therefore, we argue that rather than a dyadic transaction of evidence and values, surgery decisions are subtle interactions, involving multiple variables in flux, on the often implicit question ‘is it worth it?’ (Ecks 2022).

To further conceptualise what matters to clinicians and patients in these interactions, we draw on the writings of medical anthropologist Susan Whyte about the management of illness and uncertainty in Uganda (2005). Whyte suggests that we study how uncertainty is managed locally to understand the situated concerns and available resources for controlling it (2005). Following her lead, we show how attention to lived treatment decision‐making and its related uncertainties in shoulder clinics is an important aspect of acknowledging the tensions and trade‐offs in medicine. Whyte draws on pragmatist philosophers and introduces the ‘subjunctive mood’ as a way of talking and acting that is not based on certainty but on conditionality as expressions of doubt and hope for improving the situation. She writes about ‘practicing healthcare in the subjunctive mood’ as a situated concern for the consequences of specific actions rather than a quest for general certainty or antecedent truth (Whyte 2005). She draws on William James, who writes: ‘The truth of an idea is not a stagnant property inherent in it. Truth happens to an idea. It becomes true, is made true by events’ (James 1974 in Whyte 2005, 255). Whyte's approach allows us to envision some of the assumptions evoked in the notion of ‘evidence‐based decisions’ in the clinic. In a pragmatist perspective, truths and actions cannot be separated, and knowing does not precede doing. When we think with these ideas in exploring shoulder care, the question of who will benefit from a surgical procedure cannot be answered (or at least not fully or with certainty) in an indicative way (based on certain a priori knowledge) but can be thought of as an idea expressed in the subjunctive mood; a provisional suggestion that indicates hope for improvement and a future with less pain (better sleep, wider motion range etc.), but only time will tell if this become ‘true’ (Whyte 2005). In this article, we investigate how this unfolds in interactions in the clinic and argue that such a pragmatist ideal is fundamental to the enactment of shoulder care.

4. Methods: Fieldwork in Shoulder Clinics

The findings in this article are based on ethnographic fieldwork in two shoulder clinics at two medium‐sized hospitals in Central Denmark, conducted from May 2024 to January 2025, utilising participant observation in consultations over an extended period and interviews with patients. The aim of the study was to develop concepts and knowledge that would facilitate a nuanced discussion about decision models such as SDM, but also to study the impacts and possibilities of a generally increased public health conversation in Denmark concerning choice and decision‐making within healthcare. In the Central Denmark region, SDM is currently implemented through courses for clinicians on how to produce decision aids and evidence‐based models for how to make decisions together with patients.

The shoulder specialty was selected as a case due to a consensus in the literature and among local HCPs that it entails more complex decision‐making. This is because it relies on more signs and symptoms than, for example, knee or hip surgery, which typically depends on simple X‐ray imaging and distinct signs of osteoarthritis, such as pain and function.

The first author conducted the fieldwork, spending 1–2 days a week in the clinics from May—June and from September—January, observing approximately 50 consultations with three different orthopaedic surgeons in two separate hospitals. Apart from consultations, the first author spent time with surgeons and physiotherapists in morning conferences, and during lunch breaks, where informal discussions about everyday life in the clinic, decision‐making and treatment choices took place. Following participant observation, eight patients were recruited for interviews, inquiring into particular interactions related to making decisions about surgery or treatment. The interviews were conducted at the hospital, in the home of the patient or online and lasted between one and 2 hours. Formal interviews with two specialised physiotherapists were also conducted, whereas informal conversations with surgeons, nurses and secretaries occurred in between consultations and in coffee breaks. The first author also participated in a 3‐day course on how to make decision support tools, as well as a national seminar on the latest research in shoulder care, to gain insight into contemporary discussions in the field.

Detailed field notes were written after each interaction in the clinic. Quotes from clinicians in the article are from field notes where utterances were pinned down verbatim. Patient interviews explored illness trajectories, decision‐making, treatment reflections and care experiences during their shoulder pain, as well as how pain impacted daily life. Interviews were transcribed verbatim. The material was analysed following an abductive approach (Tavory and Timmermans 2014), where we oscillated between engaging with the empirical material and theoretical concepts. The first author coded the material and presented initial ideas and sensitising concepts to the author group, along with empirical cases. Theoretical ideas and empirical insights have been continuously discussed in the author group, which comprises two medical anthropologists, an orthopaedic surgeon and a physiotherapist.

4.1. Ethics

All participants were informed about the project and the researcher's presence in the clinic, and all interview participants provided written informed consent forms. To make sure everyone in the consultation room was informed about the study, the surgeon shortly informed them when entering the room and asked for oral consent for first author to attend. After the consultation, the first author followed the patient out and explained more about the study, requesting written consent. The research project was approved by the Danish Data Protection Agency (reference number 1‐16‐02‐226‐24). The project was reported to the Committee on Health Research Ethics in the Central Denmark Region. Following assessment of the study protocol, the committee determined that this type of study required no ethical approval according to Danish legislation on health research (reference number 1‐10‐72‐6‐24). All names in the article are pseudonyms, and non‐important details have been changed to ensure anonymity.

Fieldwork is a relational knowledge production with inherent ethical concerns that extend beyond principles and consent. Taking part in discussions during fieldwork on medical decision‐making can reveal neglected details that could alter the meaning and safety of decisions, potentially introducing uncertainty where there was certainty. Moments occurred during conversations with patients, when the increased attention to the decision‐making process introduced with this study seemed to induce uncertainty. Whereas this is informative of some of the potential effects of giving choice and decision‐making more attention in the clinic, the first author tried to mitigate uncertainty by referring to statements and conclusions from the consultation, rather than inquiring further with open questions.

4.2. Findings

In the following sections we unfold some of the ways SDM notions may not be helpful in understanding lived experience of shoulder treatment decisions. We start by contrasting formulations of uncertainty in shoulder medicine from medical reviews and those expressed by clinicians during fieldwork, to challenge the notions of evidence inherent to SDM policy. In the next section, we think through the subjunctive mood as a way of practicing health care, that better capture the relationality of clinical decision‐making than the SDM notion of preference. Finally, we challenge the idea in SDM‐policy of treatments as delineated options to choose between, by showing the entanglement of surgery and physiotherapy in shoulder patient trajectories.

4.3. Unfolding Some Uncertainties of Shoulder Treatment

Prescriptive models of decision‐making want physicians and patients to engage in ‘option talk’, where the available treatments and their related evidence are discussed (Dahl Steffensen et al. 2022). This presupposes that what counts as an option is a treatment where the evidence can be presented in a clear and comprehensible way. It also presumes that two (or more) treatments can be delineated where one has clear advantages over the other. But how can the benefits and effects of a particular treatment for a particular patient be known? For many medical interventions, including orthopaedic surgery, evidence as results presented in research literature and guidelines is either ambiguous or not trusted by the professionals because it relies on too rigid research methods or poor studies (Grove et al. 2015). Shoulder pain and reduced function of the shoulder are common in the Danish population, but the effect of either surgery or physiotherapy (the two common treatments) is generally considered uncertain. Large studies and guidelines exist, but they are often perceived among surgeons and specialised physiotherapists as targeting a too narrow population, poorly performed or only ‘telling us something about who we should not operate on, but nothing about who we should operate on’ as one orthopaedic surgeon put it during a conversation.

Medical reviews also display the uncertainties at stake, exemplified below with the diagnosis of ‘frozen shoulder’ (Capsulitis Humeroscapularis):

Whatever treatment is chosen, pain relief plays a key role, however, the evidence for this is not strong and most patients improve over time regardless of treatment. The following is a review of the scientific evidence for different treatment options. What they all have in common is that the evidence is weak, often comparing single treatment modalities, whereas the clinical reality is that patients often use multiple concurrent treatments to try to alleviate symptoms and shorten the course of treatment.

(Bak and Isaksson 2019, 3, our translation and bolding)

The excerpt displays how time might be just as effective as surgery. But when has enough time passed to warrant action? The review further suggests:

The general consensus is to initially treat non‐operatively for 4–6 months [12], but in practice, if a treatment is ineffective, it should not be continued for that long. A study by Levine et al. [13] found that 89.5% of patients were cured with non‐operative treatment, but the success rate is highly dependent on the stage at which you meet the patient.

(Bak and Isaksson 2019, 3, our translation and bolding)

The medical review indicates the difficulties of knowing when to say that enough time has been tried but also questions the usefulness of comparing surgery and time through traditional methods of controlled trials, because, in reality, ‘time’ involves other activities and patients arrive at the clinic at very different stages of their shoulder disease. This indicates that deciding what to do in the clinic requires other activities than reading research literature results.

Yet, lacking controlled empirical evidence that a particular surgical treatment works (e.g., ‘subacromial decompression’ in cases of ‘subacromial pain syndrome’) also shapes recommendations, as we see in the national guideline for treatment of ‘non‐traumatic shoulder pain’:

It may be tempting to offer subacromial decompression to these patients, despite the fact that late subacromial decompression is not supported by controlled empirical evidence. The wording of the weak recommendation against subacromial decompression leaves open the possibility that surgery may be offered after careful consideration. It should be emphasised that the evidence base for this treatment option (late decompression) is poor. Research into the treatment of this patient group is ongoing.

(Danish Health Authority 2021, 5, our translation and bolding)

The guideline leaves the question of surgery open, which calls for careful consideration, and in this way, it recognises that uncertain evidence is not the same as knowing that a treatment cannot be effective. The reluctance to recommend surgery is linked to the irreversibility of the treatment and the possibility of worsening pain and function that may result from invasive procedures. Once surgery is done, it cannot be undone, and there is always the risk of collateral damage.

The uncertainty about effect revealed in medical reviews, is not the same as clinicians being uncertain about what to do and recommend in consultations, but it problematises the idea that shared decision‐making is about discussing already existing evidence with patients. Rather, in shoulder clinics, decisions about subacromial decompression must be based on a careful weighing of case‐specific variables that emerge through interaction rather than the success of the operation in a general population. Such weighing involves questions such as: How much pain is the patient in? How does it affect them? This difference in research and the clinic is reflected in the two overall questions: ‘Is it worth it (in this case)?’ or ‘Does it work (in general)?’ (Mol 2002). The recognition of this difference in questions guiding the activities of ‘doctoring in the clinic’ and the activities involved in ‘producing evidence’ might also be why the national guideline does not exclude surgery as a possibility but suggests clinicians carefully consider it.

Furthermore, the guideline introduces the knowledge problem as one that will be solved if only successful research is performed. It indicates that there is not yet evidence that can rule out surgery. Such framing enacts surgery as a single bounded treatment with stable variables and deciding upon surgery as a matter of knowing these variables beforehand (Mol 2002).

During conversations in fieldwork, surgeons also discussed the case‐specific variables related to themselves. Experiences with different types of surgery mattered, and they talked about asking themselves, ‘Do I believe I can successfully operate on this patient?’ Such deliberation involved assessing the perceived ‘real’ issue or problem, with questions such as: Is the pain genuinely deriving from the shoulder? Will this patient perceive the operation as a success? Is the primary concern the pain itself, or is it related to other factors such as social deprivation or psychological issues? In case‐specific deliberations of this kind, guidelines have little to offer, and judging surgery becomes a matter of situated concerns.

Rather than just ‘decided upon’ through a transaction of already existing evidence and preferences, the idea of surgery must be established as an acceptable proposal that holds an element of hope but no certainty or guarantees. In the next section, we explore an example of a patient case to show how the situated concerns related to hope, doubt and trust shape this establishment and how surgery is discussed and enacted as a treatment in the subjunctive mood.

4.4. Attending to Situated Doubt: The Work of Making Surgery Acceptable

Several patients in our study reported that they were willing to take the leap of faith that surgery entails, hoping it would improve their often desperate situation. It is not a question of wanting surgery or knowing that surgery will help, but to take an action that holds an element of hope. A patient with long‐term shoulder pain who waited for a consultation to discuss surgery expressed it like this:

It's not that I want surgery at all, but if at the next consultation he says to me, ‘Listen. We don't know what's in there, but we still have the option of saying that we'll open the door and take a look. If we can't do anything, we'll close and that's it’. So I also feel a bit like I have to go with it. … Yes, it is, because it's a big deal for me, and if there's any chance that he might close the door and say, ‘Well, that's it right there. We can fix that quickly and then it'll get better’, then I'd like to try that.

(Patient interview, April 2024)

The quote indicates that deciding on surgery is not about wanting surgery (as a ‘preference’) but rather about needing to do something to potentially improve a difficult situation.

Even when evidence exists for a particular surgical treatment (such as prosthesis for progressive osteoarthritis), the idea has to be established with conditional certainty, and interactions in the consultation room plays an important role here, as seen in the following example. All text in italics is field notes written in January 2025.

Andrea, in her early 60s, has lived with fluctuating shoulder pain for 2–3 years. The pain ‘came creeping’ rather than appearing suddenly. Her case resembles other cases in our material, as she explains the significant limitations the pain has given. She cannot lift her grandchildren or cook, and she has trouble sleeping because she cannot rest on her side.

Andrea consulted her GP, who suspected a frozen shoulder—a condition rarely needing surgery, where time and non‐surgical treatment are recommended as the first choice. Because of long waiting times in the public hospitals, she got an X‐ray and a private hospital referral. The private hospital confirmed the diagnosis, prescribing corticosteroid injections and pain medication while she also began physiotherapy. When her condition did not improve, her GP referred her to a public specialist clinic.

This is where first author meets Andrea, a few days after her initial consultation with a specialised physiotherapist. There, she was informed that a new x‐ray revealed pronounced osteoarthritis, leading to a recommendation of a shoulder prosthesis. She was also given a corticosteroid injection, greatly relieving her pain. The diagnosis and recommendation for major surgery shocked Andrea.

During the observed consultation with the orthopaedic surgeon, Andrea says: ‘I had expected the physiotherapist to tell me that we could fix it with some exercises…but this…’. Because the x‐ray reveals progressive osteoarthritis, and neither exercise nor pain medication has helped, the clear recommendation from the surgeon is to have the shoulder joint replaced with a prosthesis. The treatment is routine to the surgeon but risky for Andrea.

Andrea tells the surgeon that because she does not have any pain in the moment (due to the injection) she has a hard time saying yes. “Had you asked me three months ago, I would have allowed you to take the whole arm”. She reasons that she knows the pain will come back when the injection wears of, and then she will probably see things differently. While she is talking, she looks intensively either at the surgeon and her husband. She continues: “I mean, I know what I have, but not what I get, and what if you put the thing in wrong? How much pain will I be in then??!”.

In the excerpt, Andrea introduces all her fears about the surgery going wrong and the irreversible damage that will have been done. She cannot go back to what she had once she has accepted surgery. In this moment, the pain she has is acceptable. However, she is certain that the pain she lived with before the injection is not acceptable. The absence of pain at the moment of the consultation makes it difficult to make the decision. The surgeon introduces one more element of uncertainty—the timing of the surgery:

The surgeon tells Andrea that there is a ‘window’ for when it is good to replace the joint, because “if you go beyond that, the result may be poorer”. Andrea asks how long that is, and the surgeon says that no one can really tell, it depends on how her osteoarthritis develops.

Again, the right timing of the surgery is also conditional depending on how osteoarthritis develops in Andrea's shoulder, a process that is unforeseeable. Andrea is quiet, obviously hesitant.

To move Andrea out of the indecisive space, the surgeon tries to mitigate the uncertainty and says:

“Even though this is not the most common form of orthopaedic surgery, you do not need to be afraid, it usually goes well. National registers tells us that”. To this, Andrea answers:

If you have the correct diagnoses, right?”, and continues in a rhetorical tone of voice “but I guess I do, otherwise I would not be sitting here, right?”. The orthopaedic surgeon responds with a friendly smile and says: “There is nothing strange about your thoughts” and continues “there needs to be a maturation process, because there is nothing worse than saying yes to something when you are not ready, so now you know it, and you need to think about it.

Again, Andrea throws all the uncertainties at the table. What kinds of changes to living will the treatment imply? How can she trust the diagnosis and ultimately the physician? Rather than trying to answer these questions, the surgeon acknowledges her thoughts, producing a space for irreducible uncertainty, before suggesting the need for the idea to mature.

The alternative treatment for Andrea is to continue with corticosteroid injections: “How long can I get the injections?” Andrea asks. “Everything has a price”, the surgeon answers, and continues “with time, the injections will damage your musculoskeletal system in other ways, and you need to consider this in relation to your residual life expectancy, which in your case is 20–25 years. But if we can limit it to 1–2 injections per year, it will be ok”.

The surgeon responds to Andrea's need to map out alternatives for surgery, not by excluding them with talk of evidence but by introducing other possible costs and uncertainties, such as the impact of the medication on the underlying musculoskeletal structures, which in the end might lead to more pain and immobility. Yet, he balances it with a statement that leaves open a window for going this way—if they can limit the number of injections.

Andrea sighs and looks weary. In a calm voice, the surgeon repeats “it is still a maturation process”. Then he says ‘time will tell’.

Andrea asks if there are any exercises she can do. “They will probably make it worse” the surgeon says and shows her again on the images how the osteoarthritis creates bone beaks that irritate the tissue, and that in another part of the shoulder it is ‘bone against bone’.

In this situation, the surgeon is not primarily occupied with ‘laying out the options’ or ‘informing’ Andrea about all the known risks and potentials of surgery, but is concerned with getting her to feel comfortable with going through what he considers to be the right treatment. This involves trying to understand the fears and hopes that motivate Andrea and then trying to adjust communication. In this situation, with the images the surgeon has of Andrea's shoulder, the clear signs of progressive deterioration of her joints, his experience with similar patients and how the disease develops and the possibilities he (as a surgeon) has for helping, by replacing the damaged joint with a mechanical one—he is convinced that her situation will be improved by surgery. Andrea, on the other hand, fears surgery and wants to find a different treatment. Seeing this, the surgeon patiently and repeatedly tries to establish surgery as a plausible idea and conditional certainty, while also acknowledging that this might take time.

In the consultation, Andrea does not express values or preferences that can tell us if surgery is the wrong choice for her, but she needs time and interaction to become safe in making the decision. Andrea and the surgeon interact in a leaky and unbounded space, from where to consider possible treatments and future anticipations, while both engaging with situated concerns for how to improve the situation. The important thing for the surgeon is to move Andrea towards her best possibilities and to take the leap of faith that surgery is because he believes this is her best option. For Andrea, the important thing is to end up with an acceptable life with her arm. However, the uncertainty about the exact outcomes of the paths to follow put her in an anxious and disorienting situation. The surgeon tries to help by pointing to statistics, acting calmly and acknowledging her need for time, but also by sticking to a conditional certainty that surgery is the right treatment. In this way, the establishment of surgery as a plausible idea and acceptable proposal is complex interactional work, which involves a clinician who cares beyond the choosing, including the outcome of the choice. What matters to Andrea and the surgeon is not the decision itself but how to simultaneously care for shoulder pain, fear of surgery and the irreducible uncertainty of medical interventions.

4.5. Making Choices or Clearing Pathways—The Entanglement of Treatments in Shoulder Care

SDM as a model for clinical decision‐making suggests treatment options are neatly separated entities to deliberate on and choose between in consultations, but in everyday life in the clinic, treatments are entangled in complex ways. In an interview, a physiotherapist explains it like this:

In the shoulder clinic there are often several approaches to it, because there's a lot more soft tissue surgery and we can affect the soft tissue in other ways too. So sometimes we find other ways, which means that we don't always have just one set of splints we lay, but we have some alternating tracks that we can follow along the way, but also at the start. So we say we'll start over here, but we might move in a different direction

(Interview with physiotherapist, November 2024)

During fieldwork, rather than a one‐off dyadic event, decision‐making about surgery was often distributed over several clinical encounters. The generation of a hypothesis about what might be the best thing to do could be continuous work. A patient might go several times between the physiotherapist and the surgeon, or between different surgeons, whereas the professionals visited each other to exchange ideas and hear each other's interpretations.

The following case illustrates this.

Victoria, a 38‐year‐old woman, visits her physiotherapist. She has suffered a work accident. Despite three shoulder surgeries, she feels no relief. During her last appointment, her surgeon recommended specialised physiotherapy. She experiences constant pain and heightened sensitivity in her arm, is unable to work and has trouble sleeping while caring for her two small children.

Victoria is troubled by her impression that the surgeon has lost faith in her pain. Disappointed by what she sees as mistrust, she shares her feelings with the physiotherapist. The physiotherapist, familiar with the surgeon, explains that this may stem from disappointment in the surgical outcome rather than doubt about her pain. After examining her, the physiotherapist offers a new perspective on her pain, linking it to her movements post‐injury, primarily due to pain avoidance, and how this shift in movement affects her pain. This explanation complements the surgical perspective, acknowledging that since surgery did not improve her situation, alternative approaches should be explored:

The physiotherapist examines the patient's shoulder using ultrasound, observation, and movement tests. She immediately notices asymmetry in the shoulder blades. Using “Tanya”, a plastic training skeleton, she explains how the shoulder blade, humerus, and joint interlink like a “domino effect.” She suggests that the patient's shoulder was fragile before the trauma, which triggered a chain reaction leading to nerve, muscle, and tendon damage. Compensatory movements have worsened the issue, creating an entrapment and pain cycle.

(Fieldnotes, November 2024)

Crucially, the updated explanation regarding Victoria's shoulder introduces hope and reveals new opportunities for action. Carefully, she and the physiotherapist start trying out exercises and other solutions (such as taping the shoulder). In this situation, surgery and physiotherapy are not alternatives to one another that patients (or clinicians) choose between but complementary elements in an ongoing effort to manage pain and immobility.

The initial injury presented with characteristics that convinced the surgeon that surgery would do something good for Victoria, but this turned out not to be the case. While it is easy to think that a suboptimal decision has been made from the beginning, this was not a concern of Victoria's. When asked about it, she said:

I regret it in the sense that I had hoped it would have done something better, that it would have had an effect, and I would have felt better in terms of pain. The doctor, the one who operated on me, well, he also said that it's an experiment. He never promised me that I would be pain‐free, but he promised that it was an attempt to do so, but of course I had to decide for myself, but he would advise me to do it because many people who had been operated on with similar symptoms as I have or had, got better, what can I say, got better pain‐wise, and I thought that for me, I think, 'Yes please'. If it means that I can have a life where there is less pain, then I'll take it, but I've regretted it because I haven't gotten better.

(Patient interview, Victoria, November 2024)

This quote is brimming with ‘subjunctive moods’, expressions of hope, doubts and open‐endedness. Victoria's regret is about the failure of truth to happen to the idea of surgery rather than the choice to undergo surgery. She underlines the importance of the surgeon not promising or guaranteeing improvement but establishing surgery as a plausible idea in the moment. It is not a matter of being insufficiently informed or having made a suboptimal decision, but that things did not turn out the way they hoped.

Victoria's case shows the importance in shoulder clinics of ‘plotting hope’ (Mattingly 2010), as a way of carefully balancing between not producing too high expectations on one hand, and not producing anxiety or resistance towards surgery on the other. The case also reveals the importance of continuous caring for the shoulder when surgery fails and how this might be best achieved through close interdisciplinary collaboration. The above quote from the physiotherapist highlights a shared central concern regarding how to affect soft tissue between surgeons and physiotherapists, and discovering ‘which splints to lay out’ as a collaborative task. This involves (at least momentarily) relinquishing the notion of separate disciplines with fundamentally different approaches to ‘fixing’ a shoulder. They provide unique treatments and crafts, yet the boundaries around them are ambiguous. ‘Laying out splints’ indicates that the goal is not always in sight or known but is about clearing a pathway to walk, in what could otherwise be experienced by the patient as a chaotic terrain of pain, disability and related problems.

Following the logic of choice suggested by Mol, treatments are clearly separated entities, alternatives to know and choose between before the doing (Mol 2008). In the logic of care, treatments and elements of treatments cannot be separated from the doing—the ongoing work of finding out how precisely situations will evolve and what might be done about problems.

The irreversibility of surgery and the potential damage it can cause make clinicians go to great lengths to try other options first. However, as we have seen in previous sections, there is no neutral space from where to determine when conservative treatment has been attempted enough to conclude that surgery is the right choice.

Victoria's case reveals how focussing on shoulder care as a choice between physiotherapy or surgery misses much about what occurs in the clinic and matters to clinicians and patients. In the clinic, the key query is not choosing between two distinct treatment options but to find ways to navigate through situated possibilities, doubts and hopes for an acceptable life with a problematic shoulder.

5. Concluding Discussion

This article adds to discussions on reaching decisions in the clinic by giving an example of how decision‐making, rather than an idealised transaction of ‘something’ (e.g.., information and preferences) to be elicited and then transformed into the ‘best option’ (as some forms of rational decision‐making models picture it), is complex interactional work to make certain ideas or proposals acceptable in specific situations. We have drawn on pragmatist thinking to suggest that treatment decision‐making in shoulder clinics is done in the subjunctive mood to counter situated expressions of doubt and hope. We began by unfolding the specific uncertainties of shoulder surgery, which suggest that research and guidelines might offer clinicians and patients little when discussing the idea of surgery. Instead, the proposal to undergo surgery must be made acceptable by interacting with situated concerns, fears and expectations. Finally, in clinical practice, shoulder treatments are not separate entities to choose between but entangled ways of steering a path to walk when living with chronic pain.

Whereas several versions of SDM present a more modest aim of promoting conversation, collaboration and critical reflection on treatments in the clinic (Montori et al. 2023), in Denmark the concept serves as a vehicle for producing evidence‐based scripts for communication and decision‐making (Dahl Steffensen et al. 2022). This work separates the knowing and the doing, where decision‐making becomes a separate process, evaluated by its own criteria (e.g., ‘decision regret’), rather than as part of the continuous work of producing hypothesis, acting, evaluating and starting over again that much illness demands. Implementing SDM in a particular clinical context through managerial decisions and the production of decision aids, forces both patients and clinicians to think about their care practices and treatments in terms of separate options to choose between. Although this risks overemphasising the role of choice in life with illness, it also has potential secondary effects in terms of simplifying the complex work of managing uncertainty, doubt and hope in the clinic.

Although models such as SDM break down clinical decisions and interactions, to portray linear steps towards consensus‐building and agreement, we add to the chorus of scholars arguing that everyday clinical practice is far messier. Models of choice and decision‐making in the clinic may be ‘good to think with’ but must also be continuously challenged for their inability to capture lived experience and situated matters of concern.

Whyte's notion of practicing healthcare in the subjunctive mood, draws attention to a mood of a verb—to act and interact without certainty, but with potential, possibility, doubt, desire and hope. As we saw in the case of Andrea, who expressed profound uncertainty and anxiety in choosing, the physician used multiple strategies for making her feel more comfortable. Andrea did not need more information, and she already knew the preference of the physician, but she needed to be reminded of the possibility of acting without certainty.

As one possible way of thinking about interacting on healthcare choices, the subjunctive mood involves taking a pragmatist approach and acknowledging the impossibility of certainty before the doing (the indicative). Importantly, this does not mean that nothing can be known or that any decision is equally acceptable or appropriate. It involves the realisation that any sign of illness (e.g., a test result or symptom description) or treatment effect (e.g., statistical evidence) must always be translated and made meaningful in concrete lived situations (Engebretsen et al. 2020). In these concrete situations, clinicians and patients may have very different ideas about what is going on or what the next step should be, and communication does not always end in consensus or agreement. Victoria's case illustrates the importance of how relations and interactions play out in concrete settings for how the situation is evaluated afterwards and when treatment does not help.

The central issue is not who is making which decision but how particular ways of making decisions are also particular ways of enacting care. Whyte writes: ‘Subjunctivity […] is about the specific uncertainty that particular actors experience as they try something that matters to them—as they undertake to deal with a problem’ (2005, p. 251). In our view, attending to subjunctivity means refraining from modelling decision‐making and attending to the specific uncertainties of bodies, diseases and knowledge practices as they play out in everyday situations in research and the clinic. Even if the difference can seem to be a matter of words, SDM as a model for decision‐making based on evidence, evokes the idea of decisions that are made on a priori, certain or stable truths about treatments on one hand, and the life and body of the patient on the other. As we saw in Victoria's case, in the end it might not be the decision itself that is most important but the situation that emerges from ongoing interaction with particular treatments and professionals, and the possibilities they present for living and moving forward from that situation. In general, patients value getting better, but the road to getting there is full of potential blocks and drawbacks. Rather than a transaction of abstract entities such as facts and values, care in the shoulder clinic is an ongoing material‐discursive balancing act of ‘moving together in a direction that might hold a potential for improvement’, as one clinician expressed it.

Some proponents of SDM argue that it represent a way of making consultations more caring, but we argue that it might also carry the risk of reducing care to communication alone on one hand, and protocolising communication on the other, which again risks undermining the subtle fine‐grained ways that clinicians are already engaged in plotting hope and managing uncertainty to enact care. As healthcare systems seek ways to optimise and produce effective models for making good decisions, decision‐making becomes increasingly abstract as something to be understood as separated from the lived everyday experiences of patients and clinicians. Debates and discussions could benefit from bringing decision‐making into the concrete, messy and mundane interactions of clinicians and patients in clinics, where specific contexts, emotions, affects and politics play into what it is possible to decide upon and how.

Author Contributions

Anna Louise Skovgaard: conceptualization, investigation, writing – original draft, project administration. Brian Elmengaard: writing – review and editing, validation, resources. David Høyrup Christiansen: writing – review and editing, validation. Mette Terp Høybye: writing – review and editing, supervision.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Research data are not shared.