Hi Jaap,

We feel that large-scale implementation is just a matter of "putting all the pieces together in the right time, right place". In our experience there are five factors of importance, we call our model I-PACT:

1. user-Interface: users (patients and professionals) must like the feel of the software plus must be able to use the software without much training, just because it is what people expect from software in 2013. For complex medical processes that is not easy.

2. Price: since the government is not going to pay extra for healthcare for an individual, we need relatively cheep solutions.

3. process Adaptation: professionals must learn to incorporate eHealth into daily practice.

4. Content: should adress all needed for a specific indication, be compliant with the status of medical knowledge and must be updated from time to time.

5.Technical issues: software must by secure, reliable, fast, adaptable

I think the answer depends upon the service model that you are using. If you are using a substitutive model, for example if home visits are replaced by telehealth visits or home monitoring, then the appropriate criteria is non-inferiority. If you have demonstrated non-inferiority, then one can go on to do a cost-comparison study; can we deliver the demonstrably same service with greater staff efficiency or lower costs? But if you are using an additive model, whereby the patients get their usual care plus telehealth then the appropriate research paradigm is cost-effectiveness. So, can we achieve better patient outcomes (however measured, ie physiological measures, service utilisation etc) by adding this extra service?

Large-scale implementation has its own particular problems. I have just been looking at this as part of my PhD. I'm not quite ready to be definitive, but one of my preliminary conclusions is that where it has worked the implementation model has been flexible, such as the Veterans Health Administration in the US, and where it is deeply problematic the implementation has been attempted via a large cluster RCT, such as the Whole System Demonstrator project in the UK. Look at the work of Carl May and Penny Hawe; they have both pointed out that large RCTs on health technology interventions with no flexibility often, years after commencement, produce weak, unconvincing findings, which do not convince decision makers or practicing clinicians. I think the WSD has fallen into precisely this trap.

Thats an important research question but the more important questions are: what infrastructure and training is required to give clinicans ( and patients) a chance of making a difference. What are the crucial things which need to be measured and how often, and what are the responses needed and when are they needed.

Most work does not take into account the very important need to integrate telehealth into existing solutions rather than bolting them on as extras.

This is an important question which has not been adequately explored in the current literature. The problems that are related to telehealthcare management in COPD patients are complex and may include the lack of decent infra-structure, the willingness of the healthcare providers to provide resources for the contunity of the service etc. There is always an issue in terms of the quality and quantity of support provided by the healthcare professionals to COPD patients to adhere to the programme. Techonology is the way forward for the future, if we are able to address some of the barriers related to this promising technology.

Whilst we have many small studies completed for people with various long term conditions I question whether we can just consider diagnosis the key to appropriateness for telehealth intervention. Having just completed a pilot study of telehealth technologies for people with early stage COPD (defined by our health care provider as someone diagnosed with COPD and who has less than three hospital admissions for this condition in the prior year) we found that many patients were too sick to use the technology and consequently declined participation in our trial. It would seem that we cannot aggregate populations in the way many studies have done to date and must consider how disease stage impacts upon willingness to participate and effectiveness of the intervention.

Hi,

@Victoria: Our daily experience with the COPD programme we are running is exactly the same. Implementation works well, when it is done in a flexible way. Patients and HCP's must be allowed to use the programme according to their needs. For example, some patients use it on a weekly base, some once a months, some even less. Some only use the eContact module, some only the diaries, some use the knowledge base etc etc. On the other hand some HCPs really look into the diaries regularly, others just answer the eContact questions and look into the diary when a patient sits before them for a regular consultation.

We are analysing the data of the first 400 patients from 4 hospitals in the Netherlands. Just looking into to CCQ and use of the system long term. We started in January 2011, we published some data at the ERS in Vienna las year, and are really curiuos what we will find this time...This is NOT a RCT, just real life data.

I think that using diagnosis alone as a selection criterion for telehealth is very problematic. Firstly, multimorbidity is the norm, with most patients who have one chronic condition actually having at least one other, and the number of conditions increases with age. Selecting by diagnosis leads to including some people who are possibly too well to benefit from telehealth (this was said to be the problem with some patients in the Whole System Demonstrator trial in the UK), and others being too ill, as described above. This is why using home telehealth to assist with palliative care, for example, has often failed in projects or research studies, due to low patient numbers; the patients go from reasonably well to quite ill with only a short window of time when telehealth to the home is really useful.

The alternative approach, of considering utilising a broader chronic disease management approach, and selecting the patients by their potential to benefit could be considered. We operate a medication management program by home telehealth in this way.