Skip to content
No Vape

Another event happened first: read competing events in a quitting study

Distinguish an observed competing event from observation ending, and keep cumulative incidence separate from a cause-specific hazard.

No Vape · Updated

Open book on a cafe table
Illustrative image · Lucille Emi Oh / Unsplash

Read the event definition all the way through

Time to an event can mean time to a particular first event, not whether it ever happens. Write the event, starting point and any alternatives that preclude it under the study’s definition. Do not classify every change of routine or every missed response as competing. The actual research question determines which events are mutually exclusive.

Austin, Lee and Fine’s original 2016 methods tutorial distinguishes competing events, censoring, cause-specific hazards and cumulative incidence. It also explains competing first-event outcomes. We use those concepts without importing its clinical example or numerical results. This page does not estimate health risk, recommend a model or interpret a named quitting intervention.

Austin, Lee and Fine — Circulation, original methods tutorial via PMC: Introduction to the Analysis of Survival Data in the Presence of Competing Risks (2016)

Use an ordinary first-contact sketch

Imagine fictional research timing the first support contact through channel A or channel B. Once B is the first contact, A cannot also be the first, even if A is used later. Under that specific definition, the channels compete for first-event status. They would not necessarily compete in a different question asking whether A was ever used.

Now draw a person with no observed contact before follow-up ends. That observation ending is not evidence that B occurred. Keep separate marks for A, B and observation end. The invented sketch concerns event definitions, not a recommendation about support channels or an actual study finding. It shows why the meaning of first belongs beside the outcome.

Keep the statistical target attached

When a report gives cumulative incidence, ask whether the estimate accounts for the competing events described. A cause-specific hazard instead concerns the event rate among those still free of the relevant events. Preserve the published measure and its wording. Do not translate one directly into the other, add hazard ratios, or make an alternative estimate from a chart.

Treating a competing event as a censoring event can have different meanings depending on the analytical question. Read the method and stated interpretation rather than declaring every such step a mistake. A technical judgment requires the complete analysis and appropriate expertise. What a reader can do is notice when a probability claim no longer matches the measure being quoted.

State what could no longer happen

A concise reading note names the main event, the competing event, why one precludes the other as defined, and the reported target measure. Keep observation loss on a separate line. If the summary never explains alternative events, leave their handling unknown. Do not infer that people leaving follow-up permanently quit or experienced another specific outcome.

For a personal health question, seek qualified interpretation. This exercise makes no clinical forecast and supplies no new survival analysis. Its useful result is a clear explanation of which outcome could occur first and how the published estimate describes that possibility. Your own practical progress can be recorded without placing yourself into an unverified research probability.

Sources and scope

Sources inform the guidance. The worksheets and examples are No Vape editorial suggestions, not tested treatments.

  1. Austin, Lee and Fine — Circulation, original methods tutorial via PMC — Introduction to the Analysis of Survival Data in the Presence of Competing Risks (2016)
  2. No Vape — No Vape — app features and store links

Written with AI assistance and checked against the cited sources. No clinical review is claimed. This guide covers practical planning, not individual medical treatment. How we research and write