The Complexity - and Struggle - and Realities of MDTs
Chapter 2
Imagine you are referred onto a cancer pathway by your GP. You are seen at a hospital within two weeks of the referral, have two diagnostic tests during your first attendance and a follow-up consultation a week later in which you are informed you have cancer. This is a crushing blow, but you are told that your case will be discussed at the regional MDT this coming week, reassurance, at least, that you are receiving the gold standard in collaborative care. Unbeknownst to you, however, yours is but one of over 90 cases that will be discussed across 3-4 hours, each case receiving a mean average discussion time of around 2.5 minutes. Amongst the list of cases, there will be complex ones requiring an involved discussion with input from a range of clinical professionals. To make matter worse, in every MDT meeting there are cases for which a recommendation cannot be made because crucial information is unavailable. Perhaps the CT Chest on which the recommendation turns has not yet been performed or reported. Without a system in place to flag up in automated fashion the information missing, these cases remain on the list for discussion only for the MDT to reach the conclusion they cannot be meaningfully discussed.
There will be simpler cases too, decisions requiring mere ratification, reports read out verbatim by a diagnostician so that a protocol which indicates the best-practice next step in a pathway. These could have been safely applied outside of an expensively assembled room of experts, verbally confirmed and then recorded. Despite deriving next to no benefit from a multi-disciplinary group dialogue – it is not uncommon for only two out of a thirty strong group to contribute to the discussion of such cases – protocolised cases appear because too often there is no system in place, nor sufficient time, to safely manage them outside of the meeting. NHSE is now pushing the agenda again, set out just before COVID struck of ensuring that cases are what is called ‘streamlined.’
What this means in practice is a way of dividing the lengthy MDT meetings into cases that need discussion and those that do not, with the chair making recommendations to the referring clinician usually based on standard practice. Many MDTs are not set up to do simple streamlining, hence a problem. Most patients are referred using completed Word forms sent by email. Coordinators, as the service managers are known, create the lists by copying and pasting from one Word file to another, and then they need to be sorted into cases for streamlining.
How do you sort 60-90 cases that need streamlining as an MDT coordinator who is frequently not a clinician? You need good software designed by clinicians for clinicians who know the problems. This software must be easy to use and quickly sort cases into meaningful groups, and work on the fly in real time. Is there a role for machine learning or AI? Undoubtedly but AI or machine learning need good data to work effectively. Thus, a digital MDT platform that creates a database of 2000-3000 cases per year might be invaluable for future machine learning in the future but not currently when we still work with forms completed in word and sent by email. We should embrace AI to develop the algorithms for the future, but we cannot do this using data as currently deployed in the NHS.
Our next chapter will be on how we can do things better with a digital MDT referral platform that creates a live data repository accessible to clinicians for clinical research and can report COSD compliant data. Data that can be viewed in real time and compared hospital to hospital within an ICS or region to region. What is there to fear?
It is time to be Healthcareful®