Evidence & method
Evidence-based medicine, applied and examined
Forty-nine systematic reviews, six randomised programmes, and the discipline of publishing the answer that does not help.
Evidence-based medicine was not only the method here; it was also a subject of study. The programme produced systematic reviews by the dozen, built economic evaluation into trial protocols rather than bolting it on, and turned the same critical apparatus on the specialty’s own literature.
PubMed exact-author search, 17 August 2026.
The sequence
Every one of the six randomised programmes followed the same four steps, in the same order.
-
Review first
A systematic review to establish how thin the existing evidence actually is — not to justify the trial afterwards, but to decide whether it is worth running. Forty-nine are indexed.
-
Design to the exposed gap
The trial answers the question the review showed was unanswered, in a setting close enough to practice that the answer transfers.
-
Attach the outcomes that matter from the protocol
Patient-reported outcomes and economic evaluation written into the protocol, not added once the audiometry looked good.
-
Follow long enough for the answer to change
Two-year and five-year readouts are routine; the single-sided-deafness randomisation was still reporting new outcomes in 2025.
EBM as a subject in its own right
Two strands of work treat evidence itself as the object of study rather than the tool.
- Doctorate, 2012
- M. C. J. Aarts — evidence and its use in otorhinolaryngology A thesis about how the specialty reasons from evidence, supervised inside a programme that was simultaneously producing it.
- Methodology, 2018
- Quality of reporting and risk of bias in therapeutic otolaryngology publications Kaper NM, Swart KMA, Grolman W, van der Heijden GJMG. J Laryngol Otol 2018;132(1):22–28. The critical apparatus turned on the specialty’s own literature.
- Volume
- 49 systematic reviews indexed on PubMed Spanning cochlear implantation, single-sided deafness, bone-conduction devices, day-case otologic surgery, noise protection and stapes surgery.
Testing in the conditions the thing will meet
A sound booth will tell you what an earplug does in a sound booth.
The earplug trial ran at an actual outdoor music festival. The bilateral implantation trial tested in spatially separated listening rather than through a single loudspeaker. The day-case trials measured what happened on a real night at home, including the patients who crossed over and were admitted anyway. The setting is part of the design, not a concession.
Publishing what fails
Four results that complicated the preferred narrative, and were published as they stood.
No saving
Day-case surgery
Outcomes were equivalent, but the cost sub-analyses of both trials failed to find the total saving day-case surgery is assumed to deliver, and crossover to admission was substantial.
15 of 48
Intratympanic gentamicin
The Ménière trial was stopped early on safety and recruitment grounds, well short of its planned enrolment. Reported as such.
Artefacts
7 Tesla inner-ear MRI
Feasible in principle, and the paper says plainly that the artefacts limiting it have not gone away.
Advised against
Thulium laser stapedotomy
After clinical complications, the group recommended against a technology it had itself been measuring — on the strength of its own measurements.
Health economics, inside the trial
Cost-utility analysis is often a separate study done later by other people. Here it was written into the protocols and published alongside the clinical results — including when it contradicted them.
- 2011
- The cost-utility of bilateral cochlear implantation: a systematic review Lammers, Grolman, Smulders, Rovers. Laryngoscope. The review that came first.
- 2015
- CINGLE protocol — randomised trial and economic evaluation Economic evaluation named in the protocol title, not appended later.
- 2016
- Cost-utility of bilateral versus unilateral cochlear implantation in adults Smulders et al. Otology & Neurotology 37(1):38–45.
- 2024 & 2025
- Cost evaluations of day-case cochlear implantation and stapes surgery Derks et al. The two that found no total saving.
Prediction, validated
Revision stapes surgery has a success rate that varies widely between patients, and surgeons counsel from experience. An internally validated prognostic model put numbers on it — built and validated on the same discipline the trials used, and published with its limitations.
Wegner I, Vincent R, Derks LSM, Rauh SP, Heymans MW, Stegeman I, Grolman W. An internally validated prognostic model for success in revision stapes surgery for otosclerosis. Laryngoscope 2018;128(10):2390–2396.
What evidence-based does not mean
It does not mean every trial produces a clean answer. A randomised trial that stops early is evidence about feasibility. A systematic review that finds nothing usable is a finding about the field. An imaging study that reports its own artefacts is more useful than one that does not. The portfolio contains all three, and they are on this page for the same reason they are in the journals.