webGP eConsult GPFV page suggests a rush to sign up to a plan by 23 December.
Claiming 300 practices using the product, it is notable to see the extensive quotes from their 2013 pilot study of 20 practices, 14 of which are their own in Hurley Group. What about the other 286 are not Hurley Group and would provide recent independent testimony? Doesn’t the absence of any more recent case studies seem odd?
“Self-help and signposting options attempt to reduce demand on the practice” – this is a true statement. It is not a claim that it does reduce demand.
Productivity effects are discussed on page 9. The time to process is stated as 2.9 mins per e-consultation, but does that include a phone call to the patient? 60% solved remotely is plausible, but it’s not clear whether the 2.9 mins includes a phone call. The other 40% requires face to face, and if the average duration is 10 minutes, that’s 40% x 10 minutes, a further 4 minutes. The best possible assumption means an average of 6.9 minutes to complete a demand arriving as an e-consultation, 31% saving on 10 minutes. But how much demand has shifted channel?
Speaking at the EMIS NUG on 23/9/16, CEO Dr Murray Ellender showed this chart of demand shift reaching 7%. This is over 2 years, in a young adult metropolitan demographic. Whether this was 7% of demand we don’t know, because if supply is limited, it may be 7% of supply (easily done by restricting numbers of phone or face to face appointments).
The best case is to take it as read, such that 7% x 31% gives an overall saving of 2.17% of consulting time.
We wonder whether sufficiently sensitive measures are in place to detect such a change.
Several reasons lead us to doubt the quality of this evidence, and therefore whether a “best case” assumption is reasonable, beyond the fact that this is a tied practice. This is not a time series chart, although the x-axis may suggest this. It is three data points, with a straight line between them. It contains no actual numbers of episodes, only a % of channel shift, hiding the units. For a comparison of how real data appears, please see our Greenway Belfast case study, a practice which runs a demand led system meaning very rapid access equally from phone or online demand. This is significant as there is no need for patients to use online access to jump the queue, yet the data shows 20% channel shift within 2 weeks of launch.
The lack of independent practice case studies, and the 0.7% demand shift in the original Hurley study, tells us that in the vast majority of practices the channel shift is an order of magnitude lower.
Note: we have requested a fair copy of the original case study from Dr Ellender, but no response was given. The presentation in full is available here. In a November 2016 presentation in Sweden, Dr Omar Hashmi claimed many were at 7% and some up to 30% – I wonder where is the data?