Does adherence in a quit-smoking study mean that people quit?

01

Read one participant's record without filling its blanks

All records in this example are fictional. A study plans four telephone consultations during weeks 1–4 and defines adherence in advance as completing at least three. Participant A completes three calls but reports smoking during the week before the six-month assessment. Participant B completes two calls and reports no smoking during that same assessed week. A meets the attendance rule without meeting the smoking endpoint; B meets that endpoint without meeting the attendance rule.

Participant C completes all four calls but cannot be reached at six months. Attendance is known; smoking status at that visit is not. The analysis may apply a stated missing-data rule, but that does not turn four completed calls into an observed smoking assessment. None of these invented people establishes a programme effect.

02

What exactly was someone supposed to do?

TIDieR distinguishes planned from actual delivery; CONSORT asks how participant adherence and provider fidelity were assessed. These reporting questions help locate the intervention schedule and its use definition rather than treating ‘completed’ as self-explanatory.

Extract the action, threshold, period, data source and denominator from the report. Attending three scheduled calls, returning a worksheet and self-reporting use of a resource are different records. Who kept the attendance log? Does a returned blank sheet qualify? Did a missed call mean the participant declined it or that the provider never offered it? If the paper does not explain, leave that item unresolved. Do not silently impose your preferred threshold.

[1][2]

03

Keep the two clocks and two missing records apart

In the example, call completion covers weeks 1–4, while the smoking endpoint concerns the week before a six-month visit. Neither clock can replace the other. Ask whether smoking was assessed in all attendance categories, how it was measured and which missing-record rules were used. A missing call log is not the same missing fact as an unanswered smoking assessment.

You can write a precise summary without computing a new quit rate: ‘A met the call-completion rule but not the assessed smoking endpoint; B had the opposite pattern; C's six-month smoking assessment is unavailable.’ This retains what was observed and what remains unknown. It is more informative than calling all completers successful or all non-completers failures.

04

An association is not an instruction to do more

Suppose the invented report later says people completing more calls more often met its smoking endpoint. That comparison does not tell us why. Work schedules or access could affect attendance and outcomes; early smoking experiences could also affect whether someone keeps attending. The observation alone cannot distinguish these possible explanations.

Random assignment to an offered programme would not, by itself, make later completion random. Read the report's separate analysis and limitations before giving the association a causal interpretation. Do not turn it into a moral score, a promise that extra attendance will produce a result, or advice to stop, extend or change personal care.

05

Use the distinction to ask a better question

For a paper, request the definitions and separately reported records rather than asking only for ‘the success rate’. For your own support, discuss practical barriers and questions with a qualified professional; research attendance thresholds do not decide your care. NHS information gives UK local-service routes. Confirm current availability and arrangements where you live.

This page reads public research and collects no attendance, smoking, symptom or medicine information. You can keep questions privately and share necessary facts directly with your chosen professional under that service's privacy arrangements.

[3]

06

What to keep in mind

• Intervention participation and smoking outcome are separate observations.
• Read the threshold, window and measurement, not just ‘completed’.
• Unknown follow-up stays unknown, even when attendance is complete.