Seite - 200 - in Applied Interdisciplinary Theory in Health Informatics - Knowledge Base for Practitioners

pressure, blood oxygen levels and heart rate and rhythm [32]. When aggregated over time, these data could alert clinicians to impending deterioration, prompting a phone call, an invitation to clinic or a home visit. The technology was adopted patchily (even by participating sites in a randomised controlled trial, all of whom had initially agreed to participate), and the service model which it supported was not straightforward enough to implement in community or hospital-based heart failure clinics. Below, we consider the different kinds of complexity in the SUPPORT-HF study and how the different NASSS domains can help in analysing this complex case. The condition: Heart failure affects 1–4% of the adult population; it is commoner in ethnic minorities and people from socio-economically poor backgrounds, and its prevalence increases with age (the average age of first diagnosis of heart failure patient is 76) [33]. It has multiple causes and complex pathophysiology; heart failure that results from an isolated defect in a part of the heart (e.g. a leaky valve) is now much less common than heart failure linked to general deconditioning in an obese person who also has high blood pressure and diabetes. On average, four to five comorbidities add to symptom and treatment burden and influence prognosis. Co-existing frailty, depression and cognitive impairment are common. The course of the condition is highly variable but it can lead to rapid deterioration and/or sudden death. Heart failure frequently causes extreme fatigue and may cause confusion; patients typically describe themselves as bewildered and frightened. The technology: The SUPPORT-HF technology consisted of standardised instruments for biomarker monitoring (weight, blood pressure, heart rate) along with a tablet computer (which had been developed using a co-design methodology) into which patients entered data for remote transfer to a monitoring centre. Participants in both arms of the trial received the technology and automated feedback messages (e.g. if results went outside pre-set parameters). In the intervention arm, the patient’s family physician was alerted to out-of-range results and offered suggestions for changes in therapy, whereas in the control arm, results were made available on a Web portal for the patient’s physician to access if they chose to. The value proposition: Because the technology had been developed as a research initiative, the value chain was somewhat speculative. The assumption was that using telehealth would enable the hard-pressed community heart failure nurses to take on a higher case load (from 35-50 patients per nurse to an estimated 200 patients per nurse). This was thought to be possible because of reduced travel time for nurses (who did a lot of their work by home visiting) and the assumption that processing remote data (blood pressure, heart rate and rhythm, body weight) would be a quicker way to monitor the course of a patient’s heart failure than undertaking regular clinical examinations of the patient. However, for various reasons, the trial was slow to recruit (in some but not all sites) and many patients were either not entered into the study (because the care package was considered clinically inappropriate) or because they were unable or unwilling to undertake the monitoring, or because broadband was unavailable. Thus, at the time of writing, the anticipated economies of scale in relation to nurse caseload have not yet been realised (and may have been over-optimistic). Another as-yet unknown transaction cost of the telehealth model is the cost of supporting and maintaining the technology in patients’ homes. The intended adopters: Staff at the different SUPPORT-HF sites engaged variably with the study, sometimes leading to slower than predicted recruitment. Some heart failure nurses were extremely keen but others engaged only superficially with the trial T.GreenhalghandS.Abimbola /TheNASSSFramework–ASynthesisofMultipleTheories200