Patient-reported outcome measures as screening tools

Early trials of PROMs feedback in the 1970s and 1980s focused on the use of PROMs as screening tools to aid the detection of mental health and functional problems in primary care and later in hospital outpatient departments.^274–277 The majority of work in this area has focused on the value of PROMs in screening for depression. Screening for depression involves ‘the use of self-administered questionnaires or small sets of questions to identify patients who may have depression but who are not already diagnosed or being treated for depression’.²⁷⁸ In these trials, patients who had not been diagnosed with depression were asked to complete a depression screening questionnaire, such as the General Health Questionnaire or the Zung Self-rating Depression Scale, and, for patients randomised to the ‘intervention’ arm, the score on the questionnaire was then fed back to their GP, along with information on the score, above or below which the patient might be considered ‘at greater probability of’ suffering from depression. The use of PROMs as screening tools was premised on the idea that GPs were underdiagnosing depression in their patients and, consequently, were not treating or referring onwards this population of patients. The underlying, implicit assumption was that GPs were not aware that their patients were suffering from depression, and providing them with a score which represented the patient’s likelihood of having depression would increase GPs’ ability to detect depression in their patients and, consequently, they would take appropriate action to manage the condition.

Between 2003 and 2013, the UK QOF financially rewarded GPs for using a standardised depression measures. In 2010, NICE published a guideline for the management and treatment of depression in adults that did not recommend routine screening for depression. Rather, this guideline advised GPs to be vigilant about the possibility that the patient may have depression, especially for patients with a previous history of the condition or for patients with a chronic condition, and recommended that GPs ask patients about symptoms of depression if they suspected patients may be at risk.

Patient-reported outcome measures as clinical monitoring tools

Here PROMs are envisaged as tool to support the clinical management of the patient. The underlying programme theory is that PROMs offer a more comprehensive assessment of patients’ problems than clinical questioning alone, and that regular, ongoing feedback of PROMs to clinicians will enable clinicians to reflect on whether or not the treatment was working for this patient and, if not, to change the treatment accordingly. Long and Fairfield¹⁶² argued that monitoring whether or not the desired outcomes were being achieved for individual patients was an essential component of evidence-based medicine:

At an individual patient level outcomes data are, or should be, used by clinicians to monitor how well a treatment plan is working and how the desired outcomes are being achieved, with a view to modifying treatment as appropriate.¹⁶²

This approach has underpinned the use of PROMs to monitor patients’ progress in psychotherapy, where it is known as ‘patient-focused research’. For example, Lambert et al.²⁷⁹ explain that patient-focused research:

. . . is aimed at monitoring an individual patients’ progress over the course of therapy. This . . . information can serve as valuable feedback to the practitioner . . . who can make attendant treatment modifications in real time.²⁷⁹

The feedback of patient progress in psychotherapy is assumed to raise clinicians’ awareness of the gap between the patient’s desired progress in therapy and their actual progress, creating cognitive dissonance and thus motivating clinicians to modify their approach to helping their clients to improve progress, in line with audit and feedback theories. Sapyta et al.²⁸⁰ drew on Bickman’s²⁸¹ contextual feedback intervention theory to explain how the feedback of client progress in therapy would prompt clinicians to change their therapeutic practices with a client in order to improve the client’s progress:

for clinicians to engage in self-regulated change they need . . . knowledge (feedback) about a discrepancy between that goal and the current status . . . this contradiction between what one wants to accomplish and what one has actually accomplished creates dissonance, which is psychologically uncomfortable (Aronson, 1999). This dissonance is what motivates change.²⁸⁰

Here, the assumption is that client progress need only be shared with the clinician and does not need to be discussed with the patient for changes in clinician behaviour to occur. It is assumed that the client’s progress in therapy reflects the skill or practices of the clinician and they are the focus of the feedback.

Patient-reported outcome measures have also been envisaged as a tool to monitor treatment impact of other chronic conditions to enable clinicians to make changes to a patient’s treatment. Here, PROMs are assumed to act like a test result, similar to biomedical indicators such as blood pressure or HbA_1c, and provide another piece of information on which clinicians can base their decision-making. For example, Wagner et al.²⁸² evaluated the feedback of the SF-36 to neurologists caring for people with epilepsy. Patients completed the measure prior to their consultation, and the patient’s current and previous scores along with norms for the US population were placed in the patient’s notes used by the neurologists during the consultation on at least two occasions. Wagner et al.²⁸² hypothesised that:

Health status information would reveal a patient’s decline or improvement in functioning and well-being and provide additional information to the clinician . . . [this] would help the clinician to uncover problems and monitor treatment response and . . . assist the clinician in the management of . . . the patient.²⁸²

Similarly, Søreide and Søreide²⁸³ anticipated the following function of PROMs feedback in the care of patients with gastrointestinal cancer:

. . . surgeons . . . can use the results of PROMs instruments to track patient’s functional status and QoL [quality of life] changes through treatment . . . Obtaining formalized and ‘objective’ results (although based on patients ‘subjective’ report) might help surgeons . . . better communicate with patients during their treatment.²⁸³

The second of these quotations suggests that PROMs offer a more systematic approach to the collection of information about patients functioning than clinicians’ standard methods of history taking and treatment monitoring. Although not identical, this shares many of the ideas and assumptions underlying the use of disease templates and indicators, introduced as part of the QOF in primary care,²⁸⁴ which aim to provide a structure to the ways in which GPs monitor patients’ lifestyle behaviours and prompt them to offer lifestyle advice to patients or to use structured needs assessments in social care and health visiting to ensure that certain topics are addressed with clients.^241,285

In summary, PROMs feedback is assumed to help clinicians monitor the impact of patients’ treatment on their health, which in turn may lead to changes in treatment, referrals or further tests to explore the problem. It is thought that this in turn will lead to improved patient outcomes. Greenhalgh et al.⁶² developed a model to depict the implementation chain from PROMs feedback to improvements in patient outcomes, reproduced in Figure 18.

FIGURE 18

Greenhalgh et al.’s model of PROMs feedback in the care of individual patients. Reprinted from Social Science and Medicine, vol. 60, Greenhalgh J, Long AF, Flynn R, The use of patient reported outcome measures in clinical practice: lacking an (more...)

Patient-reported outcome measures as a tool to support personalised care planning and self-management

The idea of ‘collaborative personalised care planning’ underpins much of the recent policy and think-tank literature on the management of long-term conditions. Coulter et al.²⁸⁶ in a report for The King’s Fund described this as follows:

Collaborative personalised care planning aims to ensure that individuals’ values and concerns shape the way in which they are supported to live and self-manage their long term condition(s). Instead of focusing on a standard set of disease management processes, this approach encourages people with long term conditions to work with clinicians to determine their specific needs and express informed preferences for treatment, lifestyle change and self-management support. Then, using a decision coaching process, they agree goals and action plans for implementing them, as well as a timetable for reviewing progress.

p. 7. Reproduced with permission from The King’s Fund²⁸⁶

Its model recognises that patients engage in most of their self-management activities during their day-to-day lives, away from the GP surgery or clinic. Therefore, the time they spend in consultation with clinicians presents an opportunity to better support a patient’s own self-management. However, Coulter et al.²⁸⁶ also noted that:

It is acknowledged that having better conversations between clinicians and patients is not something that can be achieved without additional effort. Clinicians already have a structure for consultations ‘hardwired’ into their daily practice . . . The biggest change for clinicians involves recognising that the information about the lived experience and personal assets that the patient brings to the care planning process is as important as the clinical information in the medical record; processes also need to be in place to help the clinician identify and include the patient’s contribution.

p. 7. Reproduced with permission from The King’s Fund²⁸⁶

Thus, it acknowledges that clinicians may need to change the process of the consultation in order to recognise and incorporate the patient’s perspective into the ways in which they discuss the condition with the patient and make decisions. Coulter et al.²⁸⁶ presented a model of the consultation (Figure 19) that involves the integration of the clinician’s expertise and assessment with the patient’s knowledge of their condition to inform the management of the condition.

FIGURE 19

Model of the consultation in collaborative personalised care planning. Reproduced from Coulter et al., with permission from The King’s Fund.

Patient-reported outcome measures have been envisaged as one tool that might support collaborative, patient-centred care and patient self-management.²³ For example, Santana and Feeny²⁸⁷ presented a model (Figure 20) of how the completion of PROMs may support the management of people with long-term conditions. The first step in their model is that PROMs may facilitate communication between patients and clinicians (but also between patients and their relatives and among different clinicians). They noted:

FIGURE 20

How PROMs completion can support the management of people with long-term conditions. Reproduced from Quality of Life Research, Framework to assess the effects of using patient-reported outcome measures in chronic care management, vol. 25, 2014, pp. 1505–13, (more...)

Our framework theorizes that the completion of PROMs could affect communication among patient–clinician . . . by raising patient’s awareness of his/her condition and facilitating the description of his/her symptoms to clinicians. Simultaneously, the provision of the information from PROMs to clinicians could trigger discussion of issues about which the patient is aware and concerned.²⁸⁷

Similarly, Feldman-Stewart and Brundage²⁸⁸ envisaged that PROMs information can enable the patient to better describe their symptoms to clinicians in a language that clinicians can understand:

. . . filing out the form . . . improves patients’ skills at describing their symptoms, such as . . . identifying and classifying their symptoms . . . [this] resulted in the patient being more effective at conveying messages about his/her health state, which in turn were interpreted by the doctor and improved the doctors’ beliefs about the patient’s health states.²⁸⁸

They also theorise that PROMs information may debunk the myth that if a patient does not mention a symptom during the consultation, it means that this symptom is not important to them:

Providing PROs to physicians can help overcome a physician’s belief that symptoms not mentioned by patients do not bother the patients.²⁸⁸

Snyder et al.¹¹ argued that providing PROMs information to clinicians may improve the efficiency of the consultation by highlighting issues that require addressing:

The PRO results may be used to help prioritize the issues that require addressing in the clinic visit and promote efficiency.¹¹

Santana and Feeny²⁸⁷ went on to envisage that improved communication stimulated by the use of PROMs may result in patients feeling more involved in their own care. Furthermore, clinicians may use PROMs to educate patients about their condition, which could further enhance patient engagement:

We think that PROMs could improve communication between patient and clinicians involving patients in their own care. In situations in which clinicians use the PROMs data to discuss and educate patients, the use of PROMs data could have the potential to enhance patient engagement and activation.²⁸⁷

Armed with increased knowledge about the patient’s perspective, this, in turn can influence how clinicians make decisions about patient care management:

A potential effect of completing the PROMs may be that patients more frequently talk about the issues with the clinician and the clinician gains insight about patients’ perspectives. Consequently, once clinicians recognize the issues as clinically important, they could initiate changes (ordering new tests, changing medications and dosages, and referring patients to other specialists) and monitor patients’ progression at the clinic visits as well as between visits. Potentially, this process could improve patient management. We assumed that the routine use of PROMs in chronic care management provides useful information to engage patients and their relatives more effectively and efficiently.

Reproduced from Quality of Life Research, Framework to assess the effects of using patient-reported outcome measures in chronic care management, vol. 25, 2014, pp. 1505–13, Santana M, Feeny D, © Springer Science+Business Media Dordrecht 2013, with permission of Springer²⁸⁷

However, the above description envisages the clinician as leading the decision-making on patient care. Santana and Feeny²⁸⁷ acknowledged that PROMs feedback could also support a shared model of decision-making:

PROMs data provide information about patient experiences and patient own preferences for health outcomes and the processes of treatment. Such information is not known by clinicians but is nonetheless important in choosing a specific treatment plan. The discussion between the clinician and the patient about the optimal treatment is of great importance given the availability of treatment options and the uncertainty of medical treatment outcomes.

Reproduced from Quality of Life Research, Framework to assess the effects of using patient-reported outcome measures in chronic care management, vol. 25, 2014, pp. 1505–13, Santana M, Feeny D, © Springer Science+Business Media Dordrecht 2013, with permission of Springer²⁸⁷

Thus, the overall programme theory is that PROMs may provide a vehicle through which the patients’ concerns and perspectives about their condition and its management are integrated into the process of information sharing, goal setting and action planning to support both clinical management and the patient’s own self-management of long-term conditions. In turn, clinicians and patients engage in a process of shared decision-making about care and treatment that reflects the patient’s experiences and preferences.

Coulter et al.’s²⁸⁶ model implied that, for personalised care planning to occur, clinicians need to view the patient’s perspective of their condition and its treatment as equally important as a biomedical perspective. It also suggests that a process of integrating the patient’s perspective (offered through PROMs feedback and the patient’s own descriptions) with clinicians’ own clinical judgement and information from biomedical information (e.g. through tests and scans) is required. Furthermore, PROMs are envisaged as offering a systematic way of ensuring that patients’ concerns are addressed.

Etkind et al.⁶⁰ presented a useful logic map (Figure 21) to summarise the steps through which PROMs feedback can improve the process and outcomes of care in the context of palliative care.

FIGURE 21

Etkind et al.’s logic model of PROMs feedback in palliative care. Reproduced from Etkind et al. under the Creative Commons Attribution license (CC BY-NC-ND 4.0). ESAS, Edmonton Symptom Assessment Scale; HADS, Hospital Anxiety and Depression Scale; (more...)

The models outlined above have varied in whether they conceptualise improvements in clinician–patient communication can have direct positive effects on the patient’s well-being and/or whether this is mediated through improvements to the process of care, which in turn improve patient outcomes. Street et al.²⁸⁹ developed a theoretical model to explain how improvements in communication may have both direct benefits on psychological well-being and indirect benefits on health outcomes through changes to the process of care (Figure 22). For example, improvements to communication may have a direct influence on psychological well-being, as patients may gain therapeutic benefits when a clinician validates their perspective or expresses empathy towards them. However, improved communication can also have indirect benefits through involving patients in decision-making, which in turn increases the likelihood that the decisions made address patients’ needs. They also explained that ‘a clinicians clear explanations and expression of support could lead to greater patient trust and understanding of treatment options [proximal outcome] . . . this . . . may facilitate patient follow-through with a recommended therapy [intermediate outcome] which in turn improves . . . health outcome’.²⁸⁹

FIGURE 22

Street et al.’s model of the direct and indirect pathways from communication to outcomes. Reprinted from Patient Education and Counselling, vol. 74, Street RL, Makoul G, Arora NK, Epstein RM, How does communication heal? Pathways linking clinician-patient (more...)

Finally, the electronic collection of PROMs data is hypothesised to facilitate the use of PROM in supporting patient-centred care and self-management. ePROs that are integrated into a patient’s electronic health record can support patient self-care by enabling patients to monitor their own symptoms both during and after treatment and direct patients to self-care advice. For example, the Psychosocial Oncology and Clinical Practice Group at the University of Leeds have developed an online tool, eRAPID, to enable the remote collection of PROMs data that are linked to the patient’s electronic health record.²⁹⁰ Warrington et al.²⁹¹ noted that:

Support for self-management can be provided by offering specific semi-automated advice, based on PROMs scores linked to validated clinical algorithms. Such algorithms can direct survivors with persistent low-level problems to available self-help or community services.²⁹¹

It is also thought to enable more efficient monitoring of patients at the end of treatment, to determine the level of follow-up and support they may require:

. . . can be used for the initial assessment of cancer survivors at the end of their treatment to inform the individualised care plan, help with risk stratification and allocation to the appropriate care pathway (self-management, shared care or complex case management).²⁹¹

Thus, it is theorised that patients may use ePROs to monitor and manage their own health independently from their contact with clinicians and to inform their decisions about whether or not and when to contact health professionals.

Patient-reported outcome measures feedback is not practical or feasible

A number of counter-theories focus on the practicalities of collecting and feeding back PROMs during the clinical consultation. It is argued that a significant barrier to the collection and use of PROMs in the care of individual patients is the ‘impracticality and burden of real-time administration and scoring of paper forms in the clinic’.²⁹² However, developments in IT now mean that patients can complete PROMs electronically through tablets and mobile phones and via the internet: so-called ePROs. It has been hypothesised that the electronic collection of PROMs data can facilitate their use in routine clinical practice by providing real-time data to both clinicians and patients. In support of this idea, Jensen et al.²⁹³ argued that:

PRO use is . . . facilitated by . . . real time electronic platforms that collect, store and report PRO data to inform clinical care. Electronic (e-PRO) assessment systems allow efficient standardised assessments, decreased respondent burden . . . improved ease of use.²⁹³

For example, Snyder et al.²⁹⁴ developed a website called Patient Viewpoint, which ‘allows clinicians to assign PROs for the patient to complete, just as they would any laboratory test’.²⁹⁵ Snyder et al.²⁹⁴ indicated that the website was designed to:

allow both patients and clinicians to track changes in status . . . so that the results of the PRO data can be evaluated in the context of the patient’s other clinical information, the website links to an organization’s EMR [electronic medical record].²⁹⁴

Thus, the electronic collection of PROMs data is assumed to facilitate their collection and rapid feedback, which in turn is expected to enhance the use of data by clinicians. Furthermore, their integration with patients’ electronic health records is theorised to enable clinicians to better integrate PROMs data with information from other biomedical or physiological tests into their decision-making.

Psychometric properties of current patient-reported outcome measures do not support their use in clinical practice

A number of counter-theories focused on the idea that the psychometric properties of current PROMs do not support their use in the care of individual patients. For example, critics of the use of depression screening questionnaires have argued that the majority of the original studies assessing the diagnostic accuracy of depression screening tools included patients who had already been diagnosed with depression, which may have exaggerated the accuracy of these tools.²⁹⁶ As such, it has been estimated that the number of patients currently not diagnosed with depression who would have been identified by screening tools ‘may be less than half the number predicted by existing studies’.²⁹⁷ This may also increase the number of false positives (i.e. the number of people who are categorised by the screening tool as having depression but who do not have an underlying depressive disorder). These people may be prescribed unnecessary antidepressive medications, which are not without side effects. Thombs et al.²⁹⁷ also drew attention to the possibility of the ‘nocebo effect’, defined as ‘the opposite to the placebo effect, whereby expectation of a negative outcome may lead to the worsening of a symptom’.²⁹⁸ Thombs et al.²⁹⁷ suggested that telling a patient that they have depression when they do not could trigger this effect and worsen outcomes for patients.

Goldberg,²⁹⁹ who developed one of the most commonly used depression screening questionnaires for research studies, the General Health Questionnaire, criticised ‘the temptation for clinicians . . . to use screening questionnaires in too-simplistic a way, assuming that those with scores above some arbitrary threshold are psychiatric cases and those below are not’. To reduce the risk of false positives and false negatives in primary care, Goldberg advocated that, for patients who have a high score on the General Health Questionnaire, GPs should ‘look at the questionnaire with the patient and ask additional probe questions suggested by particular symptoms’ in order to ascertain whether the patient is likely to have a transient disorder or a more enduring problem that requires treatment.²⁹⁹ In other words, the screening tool should be used not as a definitive diagnostic tool but as a stimulus for further exploration by the clinician. However, this requires time and effort on the part of the practitioner that they may not be able or willing to give. Furthermore, Gilbody et al.⁵¹ note that clinicians may ‘intuitively recognise’ that, in general practice, where the prevalence of depression is around 15%, only 50% of those with a positive screening result will actually have depression and, therefore, may be ‘unwilling to act on positive test results’.⁵¹ Thus, rather than move to treat people who do not have depression, GPs may ignore the results of depression screening questionnaires completely as they distrust the results.

Similarly, it has also been argued that currently available measures were not sufficiently precise to permit their use in monitoring change in individual patients in clinical practice. For example, McHorney and Tarlov³⁰⁰ conducted a literature review to compare the psychometric properties of five commonly used generic PROMs and concluded that ‘across all scales, reliability standards for individual assessment and monitoring were not satisfied’. However, in response, Hahn et al.³⁰¹ compared the reliability and measurement error of the same five generic PROMs with common clinical measures (e.g. blood glucose screening, forced expiratory volume measurements and systolic blood pressure monitoring). They concluded that ‘by offering a juxtaposition of common medical measurements and their associated error with HRQoL measurement error, we note that HRQoL instruments are comparable with clinical data’.³⁰¹ They observed that HRQoL instruments are no more or less precise in monitoring individual patients than commonly used biomedical indicators, but continue to be perceived as less reliable, and are therefore rarely used by clinicians in their care of individual patients. Hahn et al.³⁰¹ argued that this is because clinicians are unfamiliar ‘with the interpretation and potential utility of the data’; in other words, because clinicians rarely use such measures in their daily clinical work, they have not developed the so-called ‘tacit knowledge’ required to interpret and understand such data, as they have with commonly used biomedical indicators.

Patient-reported outcome measures may constrain rather than support the clinician–patient relationship

A number of counter-theories have focused on the idea that PROMs may constrain, rather than support, the clinician–patient relationship. As Ong et al.³⁰² argued, the clinician–patient consultation serves a number of functions, which include (1) creating a good interpersonal relationship, (2) exchanging information and (3) making treatment-related decisions. The underlying assumptions about how PROMs are intended to improve patient care anticipate improvements in all these functions. However, counter-theories suggest that the use of PROMs during the consultation may threaten the doctor–patient relationship because they do not sit easily with clinicians, who prefer to talk directly to the patient. Lohr and Zebrack³⁰³ explained that:

Practicing physicians tend to be both skeptical of and possibly irritated by pressures to use HRQOL instruments in daily practice. This skepticism pertains . . . to whether formulaic and standardized instruments provide any added value in eliciting information about their patients . . . the annoyance stems from the perceptions that researchers thought physicians were not doing enough or not doing right by their patients.

Quality of Life Research, Using patient-reported outcomes in clinical practice: challenges and opportunities, vol. 18, 2009, pp. 99–107, Lohr KL, Zebrack B, © Springer Science+Business Media B.V. 2008, with permission of Springer³⁰³

Similarly, Wright²⁶² noted that ‘Clinicians usually do not rely on health status questionnaires in routine practice to judge the success of therapy’ but instead ‘feel more comfortable with actually asking them if they are better’. These quotations contain some implicit ideas about why clinicians do not use PROMs in the routine care of their patients. The first is that clinicians resent the implication that their current methods of history taking and talking to their patients are not sufficient to gather the appropriate and necessary information from patients to judge the success or otherwise of treatment. The second is that they question if PROMs, owing to their structured nature, are able to capture the individual concerns of patients. The authors also suggest that clinicians may feel uncomfortable using PROMs in their discussions with patients.

Other counter-theories assert that, rather than facilitating communication between clinicians and patients, PROMs feedback may damage the clinician–patient relationship and hinder communication. For example, Lohr and Zebrack³⁰³ expressed concern that clinicians may view PROMs as ‘short cuts to an appropriately complete and nuanced patient history’ and that patients may view PROMs as ‘offputting – a lesser substitute for true conversation and sharing’. They also warned that, rather than supporting the patient’s agenda, the use of PROMs in the care of individual patients could ‘detract from improving outcomes because they divert attention away from problems uppermost on the patient’s agenda and toward clinician-centred issues’.

Linked to this idea that standardised PROMs may constrain the clinician–patient relationship, others have argued that individualised measures may offer a solution to this problem. They note that standardised PROMs assume that all items are equally important to all patients and do not allow patients to indicate how important each item is to them and thus may not reflect the views of individual patients.³⁰⁴ It has been argued that individualised PROMs may be more appropriate than standardised PROMs for use in routine clinical practice, as they allow patients to nominate what is important to them and indicate how important that domain is to their HRQoL.^305,306 Macduff³⁰⁷ argued that the process of completing individualised instruments such as the SEIQoL could provide the ‘therapeutic foundation’ for goal setting and developing the clinician–patient relationship, while the numerical data produced as a result of completion might be used as a measure of the effectiveness of clinician interventions. In other words, individualised measures can have value as a ‘conversation opener’ or vehicle for building the clinic–patient relationship, which is different from their use as a measure of the outcome of interventions. However, Macduff³⁰⁷ noted that the areas nominated by patients may change over time, which presents challenges in using individualised PROMs as indicators of the outcome of interventions. For example, it is not clear whether patients should be asked to rate the domains they originally nominated or whether it is acceptable to rate new areas identified by patients as important.

Other counter-theories question the broader idea that PROMs capture the patient’s perspective and thus offer a vehicle through which this can be more effectively communicated to clinicians and give primacy to the patient’s agenda during consultations. It is thus assumed that the balance of power between doctor and patient during the consultation will shift, with greater power being afforded to patients. However, Pilnick and Dingwall³⁰⁸ have questioned the basic premise of this argument. They note that, although interventions that attempt to increase the ‘patient centredness’ of consultations have changed the communication styles of clinicians and increased patient satisfaction, according to quantitative measures, observations of talk during consultations over many decades continue to show ‘the remarkable persistence of asymmetry’ in the doctor–patient relationship. They argue that this ‘asymmetry may have roots that are inaccessible to training programmes in talking practices’ and in fact ‘lies at the heart of the medical enterprise: it is founded in what doctors are there for’.³⁰⁸ In other words, the asymmetry is functional in the context of the wider social order; it enables doctors not just to care for patients but to maintain that social order by adjudicating on who has the right to be deemed ‘sick’. As such, it is not going to be changed by interventions that focus on changing communication practices during consultations.

Other counter-theories have focused on the challenges and costs that may be incurred by patients in the process of using ePROs to self-monitor and self-manage their own health. Lupton³⁰⁹ noted that it is assumed that through the practices of digital self-monitoring and self-care patients can have ‘control over one’s recalcitrant body and its ills’ and thus be empowered to take greater control of their health. However, she notes that digital self-monitoring may require patients to engage in self-monitoring at particular times of the day; thus, ‘empowerment’ becomes a ‘set of obligations’. She also observes that not all patients have the necessary economic or cultural capital to enact the ‘empowered consumer role’ that is envisaged by discourses of digital self-monitoring and self-care and that such patients may ‘find it difficult to challenge medical authority or simply may not wish to do so’.³⁰⁹ Finally, she argues that the very process of engaging in self-monitoring can be ‘too confronting, tiring or depressing for people who are chronically or acutely ill’.³⁰⁹

The challenges of using patient-reported outcome measures for multiple purposes

Another set of counter-theories focuses on the potentially unintended consequences of using PROMs for multiple purposes. Some of these theories consider how this may affect the behaviour of patients, while others consider how this might influence the behaviour of clinicians. Some have expressed concerns that when the results of PROMs are used by clinicians to determine access or continued use of treatments or therapies, patients may manipulate their answers to PROMs, which may misrepresent their true feelings but ensure that access to the desired treatment is maintained. For example, Lohr and Zebrack³⁰³ warned that:

Will patients, deliberately or inadvertently, give misleading information on PRO instruments that might prompt unease, if not actual mistrust of patients by their doctors? . . . for instance, in using pain measures when patients are seeking narcotics or other prescription drugs for reasons other than pain per se . . .³⁰³

Others have highlighted the challenges of using PROMs both for performance management purposes and in the care of individual patients.^16,269 For example, Wolpert¹⁶ noted that PROMs data are often mandated for routine collection as measures of service quality without considering how such measures can be integrated with ‘clinical conversations or clinical care’. She argued that the value of using PROMs data for audit purposes is often disconnected from the challenges faced by those tasked with implementing the measures on the ground, where they may undermine the clinical encounter. Wolpert explained that ‘the standard questions may seem irrelevant to a given patient and can be experienced as a potential burden for clinicians and patients alike’.¹⁶ Thus, PROMs that may be useful as measures of service quality may not support clinicians in their care of individual patients; however, it is clinicians who are expected to collect these data.