The slide that was not in the abstract

Across 16 presentations at a congress in Hong Kong this month, one theme recurred with a persistence that the programme did not anticipate: duration. Not whether these drugs work, which nobody in the room disputed, but for how long they will be taken, by whom, and who pays for the twentieth year.

Of the 16 presentations the Journal attended, 8 addressed treatment duration or discontinuation directly. Two reported registry data on persistence at twelve and twenty-four months, both showing attrition substantially higher than trial discontinuation rates, and both attributing the majority of it to cost and coverage rather than tolerability.

The decile plot showed the familiar picture: a long right tail of very large responses, a substantial middle, and a group with minimal change. Questions from the floor concentrated on whether anything in the baseline dataset separated the groups. The presenter’s answer was that nothing did, at least not usefully.

Congresses in this field are heavily sponsored, and the Journal pays its own registration and travel. Where a session was industry-supported we say so, because the sponsorship shapes the programme even where it does not shape the data.

"I see patients with unlabelled vials every month and there is no guidance anywhere that tells me what to do," said Dr Beatriz Sarmiento, clinical pharmacologist, Universidad de Navarra. "That session was the first time I have heard it discussed at a podium."

"I see patients with unlabelled vials every month and there is no guidance anywhere that tells me what to do," said Dr Ingrid Haldorsen, clinical epidemiologist. "That session was the first time I have heard it discussed at a podium."

The Journal pays its own congress registration and travel and accepts no industry hospitality. Where a session was industry-supported, our reporting says so.