AmbuFlex: tele-patient-reported outcomes (telePRO) as the basis for follow-up in chronic and malignant diseases

The use of patient-reported outcomes (PRO) in clinical practice is becoming increasingly common, and several studies have reported improved patient–clinician communication, early recognition of important symptoms, more effective self-management, and better use of resources [ 1– 7]. A PRO measure is the patient’s own report of health status, e.g. symptoms, health-related quality of life, and functional status. The American Food and Drug Agency defines PRO as “A measurement based on a report that comes directly from the patient about the status of a patient’s health condition without interpretation of the patient’s response by a clinician or anyone else” [ 8]. This definition focuses on the source of information and emphasises the importance of the patient perspective.

Experiences of use of electronic PRO (ePRO) systems have been extensively reported [ 1– 7, 9– 12]. The primary goal is to improve quality of care by better support of clinical activities, e.g. prospectively symptom monitoring used as consultation support. Haverman et al. [ 13] describe an ePRO system in daily paediatric clinical practice developed to systematically monitor health-related quality of life (HRQOL) in children with a chronic arthritic disease, and Snyder et al. [ 14] describe the development of the website PatientViewpoint, designed to collect PRO in outpatient clinical oncology. Both systems use PRO before a scheduled consultation, but Snyder et al. [ 14] emphasise the potential of using PRO between visits.

Integrating PRO into clinical practice has great potential when PRO becomes a central part of the patient pathway and is fully incorporated into daily clinical practice [ 15]. Follow-up for patients with chronic diseases is traditionally lifelong and managed by regular pre-scheduled visits. These visits may occur when the patient is well, and neither the patient nor the clinician finds the visit necessary [ 16]. Unnecessary outpatient visits place an increasing burden on already overstretched healthcare services, making it difficult to respond rapidly to a patient’s acute requests for attention [ 16, 17]. A PRO assessment can be used to evaluate the need for a clinical visit, thereby managing resources better [ 18]. Outpatient clinics could potentially minimise large numbers of routine visits if a PRO assessment is obtained when the patient is still at home. The benefit comes from making PRO the basis for outpatient follow-up instead of now—the patient visit [ 15]. However, few attempts, if any, have been made to make PRO the basis for outpatient follow-up [ 15].

In ePRO systems, PRO data may be collected at home or in the waiting room using computers, tablets or a patient kiosk. If PRO is used as the basis for outpatient follow-up—and even a replacement of unnecessary visits, PRO data must obviously be collected at a distance (e.g. from home), and we will refer to the latter as telePRO [ 18].

AmbuFlex is a generic clinical telePRO system for mixed-mode (web and paper) PRO data collection for use in clinical practice. The overall goal is to use PRO for clinical decision support to improve quality of care, promote patient-centred care, optimise the use of resources in the healthcare system, and use data for research purpose (Table  1).

The aim of this paper was to describe experiences with implementing telePRO as the basis for follow-up in chronic and malignant diseases using the generic PRO system AmbuFlex, where the patients define the need of an outpatient consultation by delivering PRO.

AmbuFlex is the frontend of the WestChronic system, used for research purposes in clinical epidemiological studies since 2004 [ 18]. AmbuFlex consists of three generic, configurable elements (Table  2) [ 18]: PRO data collection, PRO-based automated decision algorithm, and PRO-based graphical overview for clinical decision support. We describe our experiences with respect to these three elements.

As of December 2015, AmbuFlex has been implemented in nine diagnostic groups in Denmark. This paper included seven clinical projects with internal project management by AmbuFlex funded by Central Denmark Region, one randomised controlled trial with external project management funded by Aarhus University Hospital, and one clinical project with external project management funded by the Danish Cancer Society. The characteristics of all projects are presented in Table  3. In addition to the general aims in Table  1, each implementation had additional aims according to the specific diagnostic group. In epilepsy (b), narcolepsy (c), sleep apnoea (e), prostate cancer (f), asthma (h), and renal failure (i), the aim was to facilitate greater flexibility in the provision of care and thereby increase patient self-management, improve the quality of care, and achieve a better utilisation of resources. In rheumatoid arthritis (d), the primary aim was to examine the effect of a PRO-based telemedicine intervention to assess flare-ups in disease activity using a validated PRO instrument [ 23, 24] combined with a blood test. In colorectal cancer (g), one of the aims was to use PRO data to assess a patient’s health status before chemotherapy treatment in order to prescribe the chemotherapy in advance. In coronary heart disease (a), the primary aim was to screen patients for depression and anxiety.

So far, the generic PRO system AmbuFlex has been implemented in nine diagnostic groups at 15 outpatient clinics in Denmark. In these cases, telePRO was used as the basis for the contact between the patient and the clinic. In epilepsy clinics, up to 70 % of outpatients were referred to telePRO-based follow-up, and up to 48 % of the incoming PRO questionnaires could be handled with no other contact than the PRO.

Although results from other large-scale PRO implementations have been reported [ 40– 42], we have not been able to identify other examples in which PRO was used instead of scheduled visits and as the basis for the contact. In nearly half of the cases, telePRO was the only contact between the patient and the clinic. This is important for two reasons. First, when telePRO is the basis for the contact, it is no longer an optional or added task for patients and clinicians, but fully integrated into the patient care. This being achieved, the other potential benefits of PRO measures in clinical practice may be obtained such as improved quality of care, better symptom assessment, more patient-centred care, and more efficient use of resources [ 7, 42, 43]. Second, when potentially up to half of the visits may be replaced by a less resource demanding activity, there may be an economic argument for the shift to telePRO, because the savings made could cover the expenses associated with implementation of telePRO. A recent national analysis of the clinical use of telePRO initiated by the Danish government and Danish regions based solely on experiences with AmbuFlex concluded that there was an overall economical potential related to (1) cancelled consultations, (2) reduced reimbursement of patient transportation costs, and (3) reduced need of destruction of medicine in cancer treatment. A national strategy for dissemination of clinical telePRO was therefore recommended [ 44].

Implementation of telePRO in clinical practice involves several issues related to the specific patient group, questionnaire, technology, and organisation [ 19, 41]. Unfortunately, a PRO instrument with documented psychometric properties relevant for the actual clinical decision is often not available, especially in non-malignant diseases. In other cases, a relevant PRO was available, but no relevant cut-off values were documented. In these cases, the content and cut-off values were negotiated based on iterative inputs from clinicians, review of the literature, and interviews with patients [ 18]. Reliability and other validity tests are of great value in improving data quality. Reliability studies of AmbuFlex questionnaire including developed ad hoc items will be conducted in the future. The key issues in all AmbuFlex implementations are involvement of patients as well as support from frontline clinicians and administrative leaders [ 18]. Since telePRO is used for clinical decisions, even automatic decisions, strict attention to the sensitivity should be given, and the algorithm should also reflect this. To ensure high sensitivity, clinicians assigned a green, yellow, or red colour to each item response in the questionnaires. More than about 15 % automatic green responses can rarely be reached without compromising the security in relation to detect patients in need of attention. All outpatients are instructed to contact the clinic in case of sudden worsening.

High response rates are needed to achieve satisfactory rates of completion [ 41]. In the Dutch ePRO system KLIK, an average of 70 % of patients have completed questionnaires prior to the consultation [ 45, 46], but in other systems only half of the patients completed the assessment before a visit [ 42]. Patients’ health literacy can be a barrier when completing PROs [ 41] and should be taken into account by the clinician before referring a patient to telePRO follow-up. If high response rates are crucial, a mixed-mode survey must be considered. Findings from randomised studies in other patient populations support this [ 47, 48]. We use a dual-mode system in which patients can choose between web and paper questionnaires that results in response rates beyond 90 %. This is discussed in more detail elsewhere [ 18].

Health professionals value PRO data when they are useful for the clinical decision-making process, whereas potential barriers may arise when the use of PRO appears to be disruptive to normal work flow [ 5, 49]. From the clinicians’ perspective, telePRO information should be of major importance in the clinical assessment of the disease, and a physical examination should not be central for evaluation of the patients’ clinical status [ 18]. The clinicians have to reallocate resources to handle the incoming questionnaires and have available time slots for consultations when an appointment is needed. A potential barrier in telePRO could be clinicians’ reluctance to an open-access strategy, as clinicians may believe that most patients would want to be seen in the outpatient clinic. This is, however, not the case in outpatients with epilepsy, as only 23 % ask for a consultation. Another barrier related to successful implementation is clinicians’ lack of knowledge on how to effectively utilise PRO data in their clinical practice [ 45, 49]. Santana et al. [ 45] recommend learning programmes teaching clinicians how to use and react to PRO in clinical practice. We work closely with frontline clinicians who are motivated; however, it can be a challenging process to convince a medical staff who is not enthusiastic about the use of PRO in clinical practice. Training programmes could be useful in telePRO implementations increasing involvement and motivation in the entire healthcare team.

Based on our experiences so far, we can suggest some characteristics that should be fulfilled when considering a patient group for telePRO. First, patients should need regular and systematic disease monitoring with several outpatient follow-up. Second, the disease activity and thus the need of medical attention should vary over time. Third, PRO should be essential for the clinical evaluation, if necessary, together with laboratory data in the health record, and finally, the evaluation of the health status should obviously not depend on a physical examination, e.g. auscultation (Table  5). We believe that a substantial proportion of diagnostic groups in outpatient follow-up fulfil these criteria.

There is a great potential in engaging patients more extensively in PRO data collection and implementation. For outpatients with epilepsy, we have now designed a website, “My Epilepsy”, where patients have access to their own questionnaire responses, and contacts to the outpatient clinic can be initiated by the patient based on a PRO assessment. The website is linked with the Danish National Health Website (‘Sundhed.dk’). A randomised controlled study comparing the actual AmbuFlex with this new open-access version is in progress. Based on the results, this patient-centred follow-up approach will probably be used extensively in future telePRO implementations.

To our knowledge, AmbuFlex is the first generic PRO system that has transferred follow-up of entire diagnostic groups to a PRO platform for outpatient care. The AmbuFlex system is generic and not limited to specific diagnostic groups, organisations, or electronic medical records. The system has been standard practice since 2012 in epilepsy outpatient clinics and subsequently in eight other diagnostic groups. Experiences from the nine telePRO implementations have shown an impact on the organisation of patient care, since 48 % of the epilepsy respondents did not need further contact with the clinic other than the PRO itself. This could indicate the need for a reorganisation of conventional care in the healthcare system. Finally, based on our experiences, we recommend use of telePRO for patients with chronic diseases with many consecutive contacts, where PRO is essential for clinical evaluation. In the implementation process for new patient groups, involvement of patients as well as frontline clinicians and administrative leaders is essential.