Background
Improvement of patients’ health is the overriding goal of healthcare. Until recently, the pathophysiological disease understanding has been key to modern Western medicine’s effort to reach this goal [
1]. However, recognition of the importance of taking into account the patient’s perspective has gained prominence in recent years; in particular in the context of efforts to improve quality and effectiveness of healthcare [
1,
2]. Since patients’ perspective on their disease can only be given by themselves, this paradigm shift towards patient involvement has led to global proliferation and application of patient-reported outcome (PRO) measures throughout the healthcare system [
1]. Patient involvement refers particularly to patients’ right to have a central position in their own healthcare, and the benefits of this are expected to be a better patient outcome as a result of an improved interaction between the patient and the clinician [
3].Thus, to day patients are more empowered, call for more involvement and their contribution is valued by healthcare authorities [
4,
5].
In this study, we explored nurses’ and physicians’ experiences with the routine use of PRO measures in clinical practice in three outpatient clinics in the Central Denmark Region. The study reported here is part of a larger study of PRO-based follow-up in epilepsy outpatient care.
A patient-reported outcome measure is defined as “a measurement based on a report that comes directly from the patient about the status of a patient’s health condition without amendment or interpretation of the patient’s response by a clinician or anyone else” [
6]. In broad terms, PRO measures are questionnaires that measure patients’ perception of the impact of a condition and its treatment on their health [
7] Thus, PRO measures are useful when measuring aspects that are best known to patients or best measured from the patients’ perspectives [
1]. Such aspects include, e.g., symptoms not obvious to observers, psychological symptoms, the frequency and severity of symptoms, the impact of disease on daily life, social wellbeing, cognitive functioning, role activities, and health-related quality of life [
8]. Traditionally, PRO measures have been used at a group level in research and quality improvement [
2]. However, focus on use of PRO measures at the individual patient level has increased during recent decades [
9]. The use of PRO measures in clinical practice has several documented positive effects, including optimal monitoring of treatment response, facilitating patient-centred communication, support of the clinical decision-making process, optimal efficient use of healthcare resources, reporting of adverse drug reactions and as a tool for hospital performance assessment [
10‐
15]. Thus, PRO measures represent an idea to which is ascribed a range of positive effect on both patient and organisational level.
In 2016, Danish Regions decided to expand the positive experiences with PRO measures as a general idea. The aim was to improve the quality of health care by spreading knowledge about PRO measures and to standardise questionnaires on a national basis [
16]. Thus, PRO systems are now being implemented at scale, based on national initiatives [
17‐
19]. However, in previous research, it was recognised that PRO-measures are not a self-acting mechanism that automatically enhance quality in clinical practice [
20,
21]. In order for the potential of PRO measures to be realised, the idea of using PRO measures must be adopted by all relevant members of the organisation.
We know from institutional theory that new ideas are not automatically and uncritically adopted and incorporated into organisations [
22]. They are translated, shaped and changed during implementation processes, and formed by both supporting and inhibiting powers and mechanisms [
23,
24]. Thus, organisations can enact a range of strategies in response to pressures toward adopting new ideas [
22].
In routine outpatient practice, nurses along with physicians are the front-line clinicians responsible for the practical application of PRO-based follow-up [
25]. Among clinicians, there seems to be diverging attitudes towards the use of PRO measures in clinical practice [
9,
26,
27]. Some strongly advocate their use in clinical practice, as they believe that PRO measures provide data that make a valuable contribution to clinical practice; others, however, are sceptical [
2,
9,
27]. These diverging attitudes led us to consider that the perspectives of front-line clinicians as end-users of PRO measures can contribute to identifying organisational mechanisms that influence the success or failure of PRO-based follow-up. We also find it valuable to explore if these mechanisms have lasting potential when PRO measures are implemented at scale in the future.
The aim of this study is to explore what happens when PRO-based follow-up is implemented in routine clinical practice. We aimed to identify organisational mechanisms in relation to PRO-based follow-up seen from the clinicians’ perspective.
Methods
Design
The methodological approach used in this interview study was interpretive description (ID) [
28], informed by a perspective of critical realism [
29].
ID is an applied, inductive research strategy emphasising the significance of performing research arising from clinical practice with the aim of improving this practice [
28]. As required for interpretive description studies, we gathered and analysed data concurrently, as we allowed the preliminary analysis to guide the subsequent focus in the data collection phase. This iterative process of data collection and analysis allowed us to compare, reflect upon and explore data elements throughout the process. [
28].
We chose to integrate tenets of critical realism into our ID strategy as we aimed to frame, identify and comprehend complex mechanisms of action in relation to clinicians’ experiences with PRO-based follow-up. Critical realism rests on a belief in the existence of three different layers of reality. These three ontological domains encompass the “real”, the “actual” and the “empirical” [
29,
30]. Accordingly, reality is not transparent; however, it has powers and mechanisms that are not directly observable but can be experienced only indirectly by their ability to cause events in the empirical domain [
29].
Both ID and critical realism influenced the development of the interview guides and the analyses of the empirical data. We found that this integration of ID and a critical realist ontology gave us a sufficient strategy for understanding the complex mechanisms underlying the organisational implementation of the idea of PRO measures in clinical practice.
Setting
AmbuFlex is a generic web-based PRO system that supports demand-driven outpatient follow-up as opposed to follow-up with regular fixed consultation. AmbuFlex was developed by the third author and is the frontend of the WestChronic system, used for research purposes in clinical epidemiological studies science 2004 [
10]. AmbuFlex is not limited to specific patient groups and as per April 2018, AmbuFlex was implemented in 21 diagnosis groups in ten hospitals in Denmark [
19]. AmbuFlex’s overall aims are to improve the quality of care, enhance the patient-centredness of care and reallocate healthcare resources by using PRO measures as the basis for follow-up [
10,
25]. The method is termed PRO-based follow-up and differs from the traditional clinical use of PRO measures, where PRO measures simply compose a supplement to the consultation at each follow-up visit. PRO-based follow-up represents a new model of service delivery where the patient’s PRO measures are used as the very basis for outpatient follow-up.
In 2012, AmbuFlex/Epilepsy was implemented in three different outpatient clinics in the Central Denmark Region, and it is now the standard follow-up for 54% of patients with epilepsy in these clinics [
25] (Table
1).
Table 1
Outpatients with epilepsy in PRO-based follow-up
Epilepsy patients in total January 2018, n | 3958a | 1305a | 899a | 6162 |
Current AmbuFlex/epilepsy patients January 2018, n (%) | 2499 (63) | 579 (44) | 264 (29) | 3342 (54) |
Prior to implementing AmbuFlex, follow-up for patients with epilepsy was managed by regular pre-scheduled visits, typical every 6th or 12th month. These visits may have occurred when the patient was well, and neither the patient nor the clinician regarded the visit necessary [
25,
31]. In AmbuFlex/Epilepsy, regularly scheduled follow-ups may be substituted by regular questionnaires filled out by patients at home. The follow-up activity is determined by a clinician and patients receive the PRO questionnaire at fixed intervals (3, 6, or 12 months). Patients who are not capable or willing to fill in the PRO questionnaire on the Web have the opportunity to fill in the questionnaire on paper [
10].
Clinicians use the patients’ self-reported PRO data as a decision aid to identify those who need clinical attention [
31]. Nurses or physicians refer patients to PRO-based follow-up after having assessed their health status and their ability to fill in PRO questionnaires.
The questionnaire includes information about frequency of seizures, wellbeing, symptoms, health-related quality of life and information specific to aspects of daily life with epilepsy (Additional file
1). The PRO questionnaire was developed in close cooperation with clinicians. After a pilot test, PRO-based follow-up was implemented, and experiences with the system are continuously evaluated. Items are revised in an ongoing iterative process in which clinicians have influence in improving the system. Clinicians receive no formal training in practicing PRO-based follow-up [
25].
The patients’ PRO responses are automatically processed according to a specific algorithm and given a “green”, “yellow” or “red” status. A red status indicates that the patient needs or wishes contact, a yellow status indicates that the patient may need contact, and a green status indicates that the patient has no current need or wish of attention. A green status is automatically handled by the server software and a new PRO assessment is scheduled, whereas a clinician has to decide whether a yellow status patient needs contact or not.
In all cases, patients can request a contact, either a phone call or a consultation in the outpatient clinic, which will automatically overrule any automated decision that no visit is needed. Non-responders get three reminders and are contacted if they do not respond. Clinicians keep track of incoming yellow and red responses, and of non-responders [
25]. The response rate is estimated to be 92% for the initial PRO questionnaire and 95% for the subsequent ones [
25].
The PRO overview [
25] is presented graphically to the clinician within the electronic health record system. The PRO algorithm and the PRO overview are used as decision aids together with other available health record information to inform the decision whether the patient needs a contact or not [
25]. The nurses in the epilepsy outpatient clinics are responsible for handling the patients’ red and yellow PRO responses. In case of a red response, the nurse assesses whether the patient should be called in for consultation in the outpatient clinic or be contacted by phone. For yellow responses, the nurse assesses whether the specific patient needs a contact or not.
If the nurse assesses the patient’s PRO data to be without health issues that need clinical attention, the patient will receive a new questionnaire at a scheduled point in time. In addition, the nurse also decides whether the contact or consultation should be with a nurse or a physician. When handling PRO responses, nurses can refer tasks to physicians electronically, e.g. checking blood test, making phone calls to patients and authorising medication changes. Table
2 presents an overview of the distribution of green, yellow and red PRO response in the three epilepsy outpatient clinics. A total of 13,995 follow-up contacts in the 7-year period were PRO-based.
Table 2
PRO-based contacts in outpatients with epilepsy in Aarhus University Hospital, Holstebro Regional Hospital and Viborg Regional Hospital 2011–2018.
PRO questionnaire responses in total from 2011 to 2018, n (%) | 11,588 (100) | 1893 (100) | 514 (100) | 13,995 (100) |
PRO-algorithm, n (%) |
Green | 1501 (13) | 413 (22) | 94 (18) | 2008 (14) |
Yellow | 7307 (63) | 1114 (59) | 344 (67) | 8765 (63) |
Red | 2780 (24) | 366 (19) | 76 (15) | 3222 (23) |
No Contacta, n (%) | 5456 (47) | 1290 (68) | 343 (67) | 7089 (51) |
Contactb, n (%) | 6071 (52) | 569 (30) | 156 (30) | 6796 (49) |
Pendingc, n (%) | 61 (0.5) | 34 (2) | 15 (3) | 110 (1) |
The inclusion criteria for interviews were physicians and nurses working with AmbuFlex/epilepsy in the Central Denmark Region who had at least 6 months of experience with AmbuFlex. All of the 13 clinicians who were invited to participate in the study agreed to be interviewed. Thus, they were sampled entirely by convenience, as the 13 clinicians were the only clinicians who met the inclusions criteria in our data collection period (Table
3).
Table 3
Clinician participant profile
Profession | Nurse | 8 (62) |
Physician | 5 (38) |
Gender | Female | 10 (77) |
Male | 3 (23) |
Hospital | Holstebro | 2 (15) |
Viborg | 4 (31) |
Aarhus | 7 (54) |
Experience with PRO-based follow-up (months) | 6–12 | 2 (15) |
13–24 | 1 (8) |
25–36 | 3 (23) |
> 36 | 7 (54) |
The interviews were conducted by the first author from May 2016 to June 2017 and were carried out in offices of the outpatient clinics. Based on the research question and an initial literature review, an interview guide was developed. Preliminary analysis of data from the first interviews were used to develop the questions for the subsequent data collection. During the interviews, the clinicians were first invited to share their personal experiences of PRO-based follow-up. Subsequently, five main themes were raised during the interviews: PRO-based follow-up’s influence on: a) work procedures in the outpatient clinic b) intra- and interdisciplinary cooperation, c) the quality of the outpatient follow-up d) the patient-clinician relation and e) professional competencies and identity.
Data analysis
All interviews were audio-recorded with the clinicians’ permission. They were transcribed verbatim by the first author, and each participant was given a code number. The first and last author collaborated on the analysis, supported by discussion with co-authors. In the analysis, we first immersed into the data by reading all of the transcripts in order to develop a sense of the whole beyond our immediate impression of the data material. Thereafter, we arranged the data in terms of patterns that seemed to reflect similar properties. This was an ongoing iterative process in which patterns were gathered and disassembled repeatedly.
As we organised the data in thematic groups, specific dimensions of clinician experiences were increasingly gathered in recurring themes. These tentative groupings led us to consider the patterns and variety within those groups across the whole material. As the analysis process developed, possible relationships between the groups of data became more apparent, and we conceptualised the findings by extracting thematic patterns that represent clinician experiences with PRO-based follow-up. Data management was facilitated by the qualitative software programme NVivo™ [
32].
In the presentation of our findings, the term clinicians is used when the theme concerns both nurses and physicians, whereas for themes that apply only to one of the professions, the specific profession will be stated explicitly.
Discussion
The findings of this study provide insight into front-line clinicians’ perspectives on PRO-based follow-up in clinical practice. PRO-based follow-up gave rise to ambivalence in clinicians. Seen from the clinicians’ perspective, PRO-based follow-up could both increase and decrease the quality of follow-up. Moreover, PRO-based follow-up both enhanced and impaired clinicians’ work experiences. Additionally, the clinicians used strategies to ease some of the perceived disadvantages.
Some of the organisational mechanisms identified mirror those reported in other studies on clinicians’ experiences with PRO measures in clinical practice. A systematic review of clinicians’ use of PRO measures to improve the quality of healthcare documented that an increased workload was a barrier [
26]. A survey of clinicians’ preferences and perceived barriers for routine assessment of PRO measures in paediatric oncology practice illustrated that clinicians strongly value the routine use of PRO measures in clinical practice [
27]. However, it was found that the integration of PRO measures in the organisation was limited due to barriers such as lack of time, lack of financial resources and PRO measures not fitting within the existing clinical workflows [
27]. These barriers are in line with the clinicians’ experiences of organisational challenges and being overburdened by PRO-based follow-up. Nevertheless, none of these existing reports on clinicians’ experiences with PRO measures addresses clinicians’ experiences with PRO-based follow-up. Some of the clinicians’ perceived barriers in our study related to the distance between them and the patient embedded in PRO-based follow-up. PRO-based follow-up represents substantially changed outpatient care that encompasses components related to the practice of “telecare”. Especially nurses were ambivalent toward providing the care at a distance embedded in PRO-based follow-up. On the one hand, nurses found that their nursing practice improved when the PRO questionnaire guided the dialog with the patients during telephone conversations. On the other hand, nurses found that the distance induced by PRO-based follow-up deteriorated the quality of patient care. Thus, for nurses, providing care at a distance was PRO measures in outpatient follow-up often experienced as problematic. Such concerns are in line with common clinical concerns in relation to telecare [
33‐
35] Nurses are said to have particular concerns about telecare practices, because telecare will impede their relationship with patients and potentially prevent nurses from noticing important signs of trouble [
36]. Another important finding in our study was that the clinicians used strategies to ease some of the perceived disadvantages. The clinicians did extra tasks and worked around the scope of PRO-based follow-up. Thus, clinicians constituted a professional buffer as they deflected some of the negative mechanisms associated with PRO-based follow-up. As Table
2 illustrates, the clinicians’ individual assessments and contacts are extensive and therefore serve as a significant supplement to the automated server algorithm build into PRO-based follow-up. Clinicians serving as a professional buffer seems to counteract the built-in standardisation in PRO measures in order to make good the lack of resources in the outpatient clinics. These buffer actions were motivated and guided by the clinicians’ professional ethics and standards, as the nurses and physicians experienced a discordance between the freedom of responsibility that the PRO-system offered and their ethical professional standards. This finding is in line with ideas from institutional research [
22‐
24,
37]. In a theoretical institutional framework, it is stated that institutional ideas must be customised rather than simply adopted by the organisation, as a local customisation of an idea can be decisive for the ability of the idea to function as intended [
37]. From a critical realistic viewpoint, the professional buffer serves as an intuitional power making the idea of PRO measures work in routine patient follow-up. Thus, the main mechanism is PRO-based follow-up; yet, it is the professional ethics and standards that make the PRO function as intended. Mechanisms such as workload fluctuations and resource reductions challenge the clinicians’ intentions to provide a more customised and better outpatient care.
Based on our findings and the above discussion, we point to three critical areas when implementing PRO-based follow-up in routine clinical practice. The areas relate to a) the organisation of PRO-based follow-up, b) clinicians, and c) visitation of patients.
Organisational implications
Our findings documented that clinicians experienced increased workload associated with PRO-based follow-up and being overburdened. We documented that assignments in relation to PRO-based follow-up were given a lower priority in periods of busyness, which resulted in a build-up of PRO responses and a backlog of assignments. An organisational implication may therefore be to adapt clinical settings and by introducing an organisation of work procedures that underpins the prioritisation of PRO-based follow-up assignments on a par with other patient contact tasks. In addition, to comply with the increased workload, resources should be allocated to manage the PRO responses and to adequately handle the increased complexity of the visiting patients’ medical problems. Moreover, we found that both intra- and interdisciplinary support and collaboration in PRO-based follow-up were regarded as very important by clinicians. Therefore, organisation of PRO-based follow-up allowing for collegial feedback and discussions seems important in securing the success of PRO-based follow-up.
We documented diverging attitudes among clinicians towards the value of patients’ self-assessments. When clinicians do not appreciate the information provided by the patients’ PRO data and distrust the very value of these assessments, considerations regarding the clinicians’ readiness to work with PRO-based follow-up may be raised. Therefore, it is suggested that an awareness of clinical value of patients’ self-assessments could be part of the clinicians’ introduction to and training in working with PRO-based follow-up. Ideally, all clinicians involved in PRO-based follow-up would trust in the value of PRO measures.
Our study documented that the clinicians had concerns regarding some patients’ ability to fill in the PRO questionnaire, and they questioned the relevance of some patients’ self-assessments. This raises the question if the present referral approach is inexpedient. We suggest that a strong focus needs to be placed on the quality of the patient’s introduction to PRO-based follow-up. Ideally, decisions regarding the individual patient’s enrolment in PRO-based follow-up should be a shared decision between patient and clinician.
There are some limitations to our study. Concerning the transferability of the study’s findings, we acknowledge that the three involved outpatient clinics were “early adopters” whose self-imposed implementation of PRO-based follow-up may have influenced the identified mechanisms in relation to the ability of PRO measures to improve outpatient care [
37].
This study has focused on the clinicians’ perspective on PRO measures in clinical practice. Given that organisational mechanisms unfold at several levels in the healthcare organisations, it would be beneficial to also explore the perspectives of, e.g., patients, clinical leaders and other decision-makers in clinical practice. Thus, further research should include other organisational settings that have not implemented PRO-based follow-up as a self-imposed measure, and the perspectives patients and of other stakeholders should also be explored. Despite its limitations, the study certainly adds to our understanding of the complex organisational mechanisms relating to PRO-based follow-up in clinical practice. Thus, we expect our findings to be of relevance for other contexts in healthcare organisations were PRO measures are being implemented as they shed light on the mechanisms that underpin PRO measures as an instrument to facilitate quality outpatient follow-up.