Between Friday evening and Monday morning nineteen messages arrived in my portal inbox. Four were results questions. Three were medication questions that were genuinely clinical. Two were requests to complete a form. One was a photograph of a rash. The rest were people telling me things I was glad to know.
None of those nineteen is a visit. Nothing about them appears in any record of what I did that week, and answering them took a little over two hours, most of it on Sunday.
The portal is one of the better things that has happened to my patients in twenty years and I want that at the top rather than buried as a concession.
Before it, the same nineteen questions either went unasked or went through a phone tree to a nurse who wrote a message that reached me a day later with the specifics smoothed off. A patient who thinks of something at nine at night can now ask it. A patient who is embarrassed can write it instead of saying it, and I have learned things through that channel that I would not have been told in a room. A patient who does not speak much English can compose in their own time. Every one of those is a real gain and none of them is nostalgia-proof — nobody who used the old system wants it back.
What happened alongside the gain is that a category of clinical work moved out of the visit and into a place where it is not counted, not scheduled, and not paid.
The mechanics are not mysterious. A visit is a unit. It has a length, a slot, a code, and a payment attached. A message has none of those. So the work exists, it is real clinical work with real liability attached — I am giving advice, in writing, that is in the chart forever — and it happens in the interstices of a day that was already fully allocated to the units that count.
The volume is not stable either, which is the part people outside this miss. Message volume rose steeply and has not stopped, for entirely benign reasons: patients discovered it works. When a channel is free at the point of use and reliably produces a thoughtful answer within a day, use expands until something limits it. The only limiting mechanism currently in the system is the clinician's evening.
I am aware that this is the shape of every complaint about overwork ever made and that it invites an unsympathetic response, so let me put the cost where it actually lands, which is not on me.
The two hours on Sunday are elastic and I am well paid. What is not elastic is attention, and the nineteenth message gets less of it than the first. Somewhere in a Sunday inbox is a message where the person is describing something that matters and describing it badly, in the way frightened people do, and finding that one requires reading all nineteen properly. That is the thing that degrades, and it degrades invisibly, and neither the patient nor I would be able to identify the instance afterwards.
Some systems have begun billing for message responses that constitute clinical decision-making. I have watched patients receive those bills and I understand entirely why it feels like a betrayal — the message was free for six years and then it was not — and I also cannot construct an argument for why this particular category of medical advice should be the one that is free. Both of those are true and I have not reconciled them.
What I would want first is much duller: that the time be counted. Not billed, necessarily. Counted, in the same system that counts my visits, so that when a schedule is built for me it is built against the whole of what the week contains. At the moment my Sunday is not in any denominator anywhere, and a thing that is in no denominator can always take a little more.
The photograph of the rash was fine. She had been going to come in on Tuesday and now did not have to, which is two hours of her day and forty dollars she kept, and I want that counted too, and there is no field for it either.
