I would be willing to wager that the topics tackled in this issue will provoke robust opinions. I would counsel a pause before we launch into the necessary discussions. Psychology has provided plenty of evidence over the years that we are not as rationale as we assume. It is all too easy to become polarised in any debate. Ezra Klein explores this in his book, Why We’re Polarized, and we are quick to deploy mechanisms that preserve our sense of self and our immediate relationships. We resist factual information that threatens these and having a higher IQ or an ability to parse evidence is no defence. Indeed, some research suggests that rather than being fair and even-handed, people who are more intelligent ‘invest their IQ in buttressing their own case.’ 1
So, here it is: is it time for GPs to give the care of children to primary care paediatricians? Turn to Debate & Analysis this month for the head-to-head. The authors marshal the facts, but it’s impossible to miss the emotive undercurrents on both sides. It is an important discussion, as we need to have some difficult conversations about how to deliver primary care when the GP workforce simply doesn’t exist in the required numbers.
The prospect of AI and algorithmic medicine can be divisive. We have research on the Centor and McIsaac scores in this issue flagging their potential uses as well as their undoubted limitations. Many would shout that algorithms can’t cope with the complexity of humans, which rather ignores the complexity, if not fallibility, of the human who is trying to make decisions. That aside, in their editorial, Dambha-Miller, Everitt, and Little take clinical scoring systems to task.
Christopher Wylie’s exposé of Cambridge Analytica emphasises the awesome power of algorithms and how they were stunningly exploited in the Trump presidential campaign.2 But they had 5000 data points per person across 250 million users and served millions upon millions of ads on Facebook.
The power at a population level is mind-boggling. The major flaw with clinical scoring systems is the paltry number of data points when faced with a single patient, and, as is highlighted in the editorial, the limited datasets used to generate them in the first place.
Parenthetically, I can’t help noticing that the ludicrous 10-minute consultation casts its long shadow in many of these discussions. The emphasis on the individual consultation rings hollow when it is conducted at a pace and superficiality that would make speed-daters blush. Is there a phrase that more comprehensively damns general practice than ‘one problem, one appointment’? It seems that this is a nettle our leaders will not grasp. We will never secure the long-term future of the GP workforce without working patterns that nurture sustainable careers. Inadequate consultation length is a structural flaw that shapes every discussion, whether it’s the care of children or the incorporation of technology. But I digress.
I encourage you to engage with the debate, but more than that, I’d encourage us all to come to any discussion in full recognition that we all carry our own notional Gladstone bags jammed with biases, just like every other human being. Klein puts it elegantly:
‘Our brains reflect deep evolutionary time, while our lives, for better and for worse, are lived right now, in this moment.’ 1
I am just keeping the Editor’s Briefing seat warm for this month as we transition between editors. I will take a moment to offer a final note to our outgoing editor. Roger, we will miss you. You have been wise, compassionate, and generous with your time and expertise. I regard myself as extraordinarily fortunate to have had your guidance for these past 7 years and the BJGP owes you a great debt. Thank you.
- © British Journal of General Practice 2020
