A new Slop University research poster audits a physiotherapy clinic's no-show risk score across four sites, finding it still predicts in-person non-attendance but collapses to chance once appointments move to video.
A physiotherapy clinic's no-show score still predicts who skips an in-person visit, but is no better than chance once the visit moves to video
The School of Continuous Improvement has built a career on catching an indicator after it has quietly stopped meaning what everyone still assumes it means, and a new Slop University research poster hands it fresh material: a physiotherapy clinic’s no-show risk score, reliable for years, that stopped working the day the clinic added video appointments and nobody noticed.
The score was trained on people who had to find a car park. Nobody told it some of them would only ever need to find a link.
— Dr Mirela Hanke, Postdoctoral Fellow, School of Continuous Improvement
Dr Hanke, working with Dr Torun Ezeigwe and Dr Solveig Adeyemi, tracked 9,540 appointments across four clinic sites over fourteen months, crossing the booking software’s own risk band against whether a visit happened in person or over video. For in-person appointments, the score’s gradient held: patients flagged high-risk missed appointments at more than six times the rate of patients flagged low-risk. For telehealth, the gradient flattened almost to nothing — a dropped video call and a genuine no-show share one outcome code, and the score can no longer tell them apart.
“The numbers don’t fail loudly,” said Dr Ezeigwe, Director of the Master of Applied Measurement. “They just stop meaning what the booking desk still assumes they mean, and the courtesy calls keep going out on schedule either way.”
The poster joins the School’s expanding audit of decision tools still issuing verdicts long after the population they learned from has moved on.
The full poster is available from the University’s research repository under an open licence, doi:10.5555/slop.451z4q.