Self‐responsibility, rationing and treatment decision making – managing moral narratives alongside fiscal reality in the obesity surgery clinic

Abstract Background Addressing the prevalence of severe obesity and its concomitant morbidities is widely acknowledged as one of the most pressing global health priorities. Nevertheless, a paucity of effective interventions and universal pressure on health‐care budgets means that access to obesity treatments is often limited. Although health‐care rationing can be conceived as a socially constructed process, little is known about how decisions emerge within the context of face‐to‐face doctor–patient interactions. Methods In this study, we used in‐depth interviews and clinic observations to investigate clinicians’ (n = 11) and patients’ (n = 22) experiences of the rationing of obesity surgery and to examine how broader cultural assumptions around personal responsibility for health emerged in the context of clinical interactions. Results Patients and clinicians worked within similar frameworks when it came to self‐responsibility for health and the appropriateness of providing publicly‐funded weight loss surgery. Issues around personal accountability dominated consultations, and patients were expected to provide narratives of the development of their obesity and to account for the failure of previous interventions. Clinicians faced the added pressure of having to prioritise a limited number of patients for surgery, which was predominantly managed through mandating pre‐referral weight loss targets. Discussion Although clinicians sought to maintain an empathic attitude towards individual patients, in practice they were conflicted by their responsibility to ration health‐care resources and tended to rely on entrenched models of behaviour change to allocate treatment. As a result, the content of consultations was mostly focused on issues of personal responsibility, reflecting wider stigmatized attitudes towards extreme obesity.


| INTRODUCTION
Clinically severe, or morbid, obesity is commonly defined as a body mass index (BMI) in excess of 40 (type III obesity), or in excess of 35 in the presence of significant co-morbidity (type II obesity). [1][2][3] Living with severe obesity is associated with a host of physical and psychiatric morbidities, most notably including diabetes, cardiovascular disorders, musculo-skeletal complaints, depression and several cancers, and the mortality rate for this population is approximately double that for those with a healthy BMI. 4 Effective interventions to treat severe obesity are sparse and many sets of national health-care guidelines now recommend the use of bariatric surgery as the intervention of choice where other treatments (such as diet and exercise programmes and anti-obesity drugs) have failed, subject to appropriate medical and psychological assessment. 1,2 Nevertheless, at least in state-funded health-care systems, it is widely acknowledged that access to bariatric surgery is significantly rationed, and in the United Kingdom, less than 1% of the theoretically eligible population (according to guidance from the National Institute of Health and Clinical Excellence [NICE]) are treated on the National Health Service (NHS) each year. 5 NICE guidance states that surgery should be available to those with a BMI>40 (or>35 in the presence of significant co-morbidity) where dieting and pharmaceutical interventions have failed to yield lasting weight loss. 1 Rationing is frequently argued to be a ubiquitous element of health-care provision. 6 However, the characterization of rationing as an economic inevitability has come under question in the sociological literature, which instead highlights the social construction of the concepts of "need" and "effectiveness," and underlines the importance of socio-political relations in decision making. 7 Considerations around the social construction of rationing are relevant at all levels of health-care decision making. Even the so-called "technical" costeffectiveness calculations used by organisations such as NICE in the United Kingdom are under-pinned by political and social value judgements (such as the accord given to Quality Adjusted Life Years [QALYs]), and at the meso-level observational studies show an ongoing "tussle" between clinicians and managerial staff for control over scarce resources with scant regard to explicit decision-making frameworks. [8][9][10] When it comes to decision making about individual patients evidence is sparser, although observations of multidisciplinary decision making have shown that moral and clinical discourses mingle as clinicians try to identify which patients have the most legitimate claim on resources. [11][12][13] Interview studies with clinicians have demonstrated the stress endemic in rationing at this level, with fears of prompting distress (or anger) in patients often resulting in clinicians retreating to implicit rationing strategies whereby patients are simply not told about all the available options for treatment. 14,15 The relevance of personal responsibility as a criterion in the allocation of scarce health-care resources has long been debated. Some ethicists have argued that the extent to which a health condition arises as a consequence of a freely chosen behaviour should impact on the degree of priority awarded to treating that individual, albeit in a rather limited way (ie the illness should not threaten life or fundamental capabilities and treatment should still be made available through a degree of copayment). 16 There is certainly some support for this from the public with all surveys that posit the question finding majority agreement for inclusion of self-responsibility in prioritysetting frameworks. 17 However, more in-depth studies with public representatives have shown that these views tend to be revised once participants are told more about the complexities of the situation and the multitude of difficulties in proving cause and effect. 18,19 Particular problems here include defining how much of an individual's health condition is caused by their behaviour or just "luck" (such as genetic make-up or in-utero exposures), and the amount of agency any one person might be expected to exercise over their behaviour, which is presumed to be impacted by both internal and external stressors. 20 It is for these reasons that policymakers and clinicians alike have largely rejected the adoption of self-responsibility into formal priority-setting frameworks other than at a very general level. 16 However, the extent to which personal responsibility is taken into account by clinicians in micro-level resource allocation decisions it is unknown. In this study, we aimed to use the example of decision making in the obesity clinic to investigate the operation of rationing at the level of the doctorpatient consultation. In this article, we report how considerations around self-responsibility for health impacted on prioritization for surgical treatment so that more can be understood about the implicit rationing mechanisms that occur when there are insufficient resources to treat all those who could potentially benefit from a treatment.

| METHODS
A qualitative approach was taken to the investigation, which involved the use of interview and observations (via audio-recording) so that both the individual views of patients and clinicians and naturally occurring interactional data could be accessed. The study took place in a UK NHS centre for obesity surgery, which saw a patient throughput of 400-500 patients per year, all of whom were referred by primary care physicians and met NICE criteria for surgical treatment.
However, only in the region of 50-60 surgeries were undertaken each year due to funding restrictions imposed by the local NHS funding body, which could not afford to implement NICE recommendations that obesity surgery should be more widely available. A longitudinal approach was taken to research and, where possible, patients were followed up for a period of 3 years following their initial consultation (Table 1). Clinicians were interviewed once prior to any contact being made with patients.
A purposeful approach was taken to sampling, and all those professionals involved in the prioritisation of patients for treatment were invited to participate in a research interview. In addition, consultant physicians, who emerged as the key decision-makers, were invited to take part in the consultation study. Patients were sampled purposefully, and anonymized records were used to identify those across a broad spectrum of ages, of both genders, and with a range of different BMIs within the morbidly obese category. Sampling and data analysis were undertaken iteratively, and initial research findings were used to guide ongoing sampling techniques. Data collection was undertaken by AOS using in-depth interviews and audio-recordings of consultations. Interviews with clinicians were undertaken first, and key topics explored included general attitudes to severe obesity, the legitimacy of funding obesity surgery on the NHS and approaches taken to the prioritisation of patients for treatment.
Where possible, interviews with patients were undertaken during the 2 weeks prior to their initial consultation at the obesity clinic and focussed on topics including individual illness trajectories, general attitudes to the provision of obesity treatment on the NHS, and expectations of forthcoming consultations. All interviews were audiorecorded on a digital device, and data were transferred as soon as possible to a secure University server, usually on the same day. Where patients were in agreement, their subsequent 6-monthly follow-up consultations at the weight management clinic were also recorded, and further in-depth interviews conducted (with patients only), usually within 2-3 weeks of the consultation date. Sampling of patients continued until we had recruited patients with a reasonable spread of age, sex and BMI within the morbidly obese range. However, the emphasis was on gaining an in-depth understanding of individual experiences, and all patients were invited to participate in each stage of data collection unless they had either specifically requested to withdraw from the research or had failed to respond to three contacts, which was interpreted as a preference not to be contacted further. It was not possible to blind the clinicians to which consultations were being recorded, as we were unable to get ethics approval for recording entire clinics.
All interviews and recordings of clinic observations were transcribed using a professional transcribing service. Initial interviews were then independently coded by all three authors and preliminary coding structures established jointly. These coding structures were then further developed by AOS, although these were frequently reviewed by JC and JD. Clinician interviews were all undertaken at the start of the study to build rapport and gain permission to observe clinics, and this data was used in the development of topic guides for patient interviews, which were coded separately. Atlas.ti was used to facilitate data management. Consultation data were coded thematically in the first instance, to establish the dominant structure of consultations and identify major themes arising therein, of which the most marked was personal responsibility for health. Individual consultations were then subjected to further scrutiny to specifically interrogate the theme of personal responsibility and link this to individual views and experiences of health-care rationing expressed in individual interviews.
Although we focused on interactional elements in analysis (such as turn-taking, interruptions), a full Conversation Analysis approach was not undertaken.

| RESULTS
Eleven clinicians and 22 patients were recruited to the research, and in total, 78 interviews were undertaken and 22 consultations recorded over a period of 3 years. Clinician participants were primarily consultant physicians, but also included the lead dietician, the service psychologist and a specialist nurse (referred to as Allied Health

| The context for decision making
Two distinct but related themes arose when participants were discussing the overall appropriateness of obesity surgery in individual interviews, which related to (i) the excessive consumption of food, and (ii) the excessive consumption of NHS resources.

| The excessive consumption of food
Reflections on the aetiology of severe obesity were a common occurrence in interviews, and both clinicians and patients discussed this within an overall framework whereby individuals were expected to take responsibility for the health consequences of individual Over-eating was discussed as the main causative factor of severe obesity by all professional informants. However, this was almost always discussed with reference to a wider "obesogenic" environment and often with reference to a psychological basis for eating difficulties that made it difficult for particular individuals to exercise full agency over their behaviour. Professional accounts were clearly under-written by a conviction that patients had to "earn" the right to surgery, and all clinicians said patients were expected to lose ten per cent of their weight at referral before they would even be considered for surgery.

| The interactional basis of health-care deliverydecision making in clinic
The need to prioritise patients for care and the requirement that they should lose ten per cent of their bodyweight prior to being considered for surgical treatment tended to dominate initial clinic consultations.
Clinicians first needed to negotiate patients' expectations of their referral, which were usually that they had been referred for surgical assessment, and explain the clinical service on offer, which for most patients was limited to a dietetic (and sometimes psychological) intervention. Although several clinicians openly acknowledged to patients that not all those referred into the service would eventually be treated surgically, only one (C8) related this directly to resource constraints and explained that they were trying to achieve the best overall outcomes from limited health-care resources. This approach seemed broadly acceptable to patients, who did not contest this basis for decision making either in consultations, or in post-consultation interviews. Other clinicians focused on the criteria that would secure a surgery slot for the patient at hand, which related to the attainment of a 10% weight loss target alongside satisfactory engagement with the dietetic team over a period of at least 12 months. All clinicians related this to both taking personal responsibility for health and getting the best possible health-care outcomes for the individual patient.

If you don't control yourself with your diet and you don't have a good pattern, you still gain weight after an operation and for that reason alone we don't want to put people through a useless procedure. (C5 to P10, follow-up consultation)
Despite many patients revealing in post-consultation interviews that they felt distressed and frustrated at being given a weight loss target to achieve, this was never disclosed in clinic, and any contestations that were raised were related instead to the requirement that this should take a minimum length of time, so extending the wait for surgery.

| Accounting for weight gain and recommendations for change
The majority of clinic consultation time was devoted to eliciting patients' stories of weight gain and making recommendations for behaviour change, both of which were usually explicitly related to personal responsibility for health ( Table 2).
Although the majority of clinicians clearly tried to maintain a non-judgemental stance at the outset, patients' contributions often became apologetic and confessional as they sought to account for their outsize bodies. Responses to these "confessions" varied with clinicians appearing at times to endorse patients' negative selfassessments and at others to seek to limit the emphasis on personal culpability put forward by patients. This is demonstrated by the two extracts presented in Box 1. In the first extract, the doctor attempts | 611

OWEN-SMITH ET al.
to comfort the patient in a number of ways. Initially, he tries to limit the culpability she accepts for her health state by alluding to "central stuff" which implies a physiological basis to her weight problem that she cannot control. He then goes on to elicit their commonalities by emphasizing their similar age and socialization experiences, which the patient engages with enthusiastically, leading to further disclosures about the aetiology of her condition. Lastly, the doctor assists her to construct a moral separation between herself and other patients, whom he denigrated as "creatures" and "less complex", further limiting the patient's personal accountability. In extract 2, by contrast, the clinician appears to endorse patient culpability by not engaging with her initial disclosure of poor mental health and focusing instead on the behavioural aspects of her narrative. In addition, her disclosure of previous weight loss is met with little acknowledgement of success ("not bad") and subsequently used as a platform to disclose the mandated weight target prior to consideration for surgery.
Although some patients welcomed their clinician's advice around healthy eating, for most it repeated information they already knew, and for some, it was burdensome and patronizing, contributing to the distress they already felt around their eating patterns and body size.

| Assessing progress and treatment outcomes
Nearly, the entirety of follow-up appointments was dedicated to the discussion of patients' success (or failure) to lose weight since their last appointment. Where patients had been successful in losing weight, this was generally met with praise and approval by clinicians, although interview data showed that this was not always interpreted in a positive light by patients. Follow-up interviews showed that those who continually failed to meet their weight loss targets experienced frustration at the perceived lack of support available, and some despaired as they began to regard the target, and therefore the surgery, as unattainable.

| DISCUSSION
Patients and clinicians worked within broadly similar frameworks when it came to the discussion of responsibility for health and the availability of publicly-funded weight loss surgery. Both groups accepted that at least some level of personal culpability was important in the development of the condition, and this common attitude framed clinic interactions and decision-making processes. Appointments tended to be dominated by discussions around individual narratives of weight gain and previous attempts to lose weight, which forced patients to reflect on previous deficits in their health-related behaviour. This led to a predominance of guilt-ridden, "confessional" narratives from patients where self-stigmatization was rife and clinicians seemed to be accorded additional status through their habitation of healthier bodies. This gulf in status seemed to make clinicians' other task in the consultation-to prioritise between patients for surgical treatment-easier, and patients were all told they needed to lose weight to "earn" their right to surgery.
Although post-consultation interviews showed that most patients were uncomfortable with this prioritisation technique, concerns were very rarely raised in clinic, where there was evidence of collusion between clinicians and patients to focus on behavioural aspects of obesity and ignore the evidence of wider system-level limitations on care.
These findings offer some support to earlier observational work in obesity clinics which showed how stigmatised attitudes towards extreme overweight (on the part of both patients and doctors) shape clinical interactions and accentuate the power imbalance in the consultation. 21 to publicly-funded elective surgery is a relatively commonly reported phenomenon in both the professional and lay press. 28,29 The research reported here shows that although most patients were able to engage in lifestyle conversations when discussing their eligibility for treatment, they became frustrated if the targets appeared too challenging or they were repeatedly unable to achieve them. This information will be useful to clinicians and policymakers when considering how notions of self-responsibility for health can be further integrated into explicit policy frameworks. 20 Of particular importance, here is the conservation of the therapeutic relationship between doctors and patients and the need to avoid the stigmatization of marginalized groups; 16 were investigated and used as triggers for deeper and more nuanced analysis. The study is limited by its focus on one clinical area and restriction to one geographic location in South West England and future research would benefit from a larger and more varied sample, alongside follow-up interviews with clinicians to assess any changes in attitude over the study period. In addition, a key element of the data presented relates to a social acceptance of clinicians acting as gatekeepers to care, which is common in state-funded health services. However, in largely private insurance systems, such as the US, or in state-funded insurance systems, this expectation may be less clear-cut, with more rationing responsibility resting with managerial staff or insurance bodies.
The research presented in this study contributes to the argument that health-care rationing is not simply an economic, rational process, but rather a socially constructed process that emerges through a dialogue embedded in existing social relations. 7 The broader social discourse demanding personal responsibility for health explicitly influenced decision making in clinic, but financial factors were rarely mentioned, meaning that rationing remained largely implicit and nonaccountable. The data presented here provide further insight into the operation of clinical rationing in practice and will be of use to clinicians and policymakers considering how the global rhetoric around selfresponsibility for health can be integrated into future frameworks for health-care priority setting.