What a Simulated Replication of Adam et al.'s Patient-Preference Study Must Show
A health-services team testing a treatment experience needs to distinguish what patients prefer about the consultation from whether a treatment is clinically effective. A discrete choice experiment can address consultation tradeoffs. A simulation of that experiment needs evidence against the same patient task before informing a new patient-experience program.
The human preference study
Adam and colleagues studied treatment-process preferences and willingness to pay among patients using acupuncture, homeopathy or general medicine in Germany. The published choice experiment found that active listening and time were especially relevant to the studied patients. This is evidence about treatment-process preferences, rather than proof that any treatment improves a clinical outcome. Original study.
A new consultation design may involve different costs, appointment lengths and patient needs. Preserve the human study's full attribute levels and population when testing replication. A selected list of listening, time and cost does not establish that two complete questionnaires match.
The comparison requires a public run record
This page has no public Subconscious artifact containing matched human and synthetic estimates, model configuration, run date and uncertainty. It therefore makes no claim that a simulated ranking reproduced the original ranking. The human paper can substantiate its own results, but cannot verify a later simulation's result.
| Proposed use | Requirement |
|---|---|
| Develop a questionnaire | Patient input and realistic treatment-process alternatives |
| Check a simulation | Matched design, population and estimated attribute effects |
| Choose a new consultation format | Evidence for the intended patient group and feasible service |
| Claim clinical benefit | Appropriate clinical outcome evidence |
Design around the service decision
Suppose a clinic is considering a longer appointment at a higher fee. Clearly separate time with the practitioner from attention, waiting time and price. A longer appointment may also change which patients can access the service. The alternatives should reflect the clinic's actual constraints, and the analysis should inspect relevant patient groups separately.
A simulation can help locate ambiguous tradeoffs and identify estimates worth checking. It cannot demonstrate that current patients will choose the format, that clinicians will deliver it consistently, or that clinical outcomes will improve. Those questions require their respective patient, operational and clinical evidence.
Set a validation threshold
Define what difference in patient preference would change the service decision. Compare effects and uncertainty at that level, rather than treating an overall rank agreement as sufficient evidence. Where the simulation diverges, retain the disagreement in the report and investigate the task and population assumptions.
The public method evidence describes the validation measure available today. A decision review can scope the treatment-process question and a separately specified human study. Human recruitment, study ownership and evidence requirements should be agreed explicitly before relying on a fielding plan.