Events per person-time are not the percentage of people who quit

Name both parts of the measure

Find what counts as an event and whether one participant can contribute more than one. Then identify the observation time: person-days, person-months, or another stated unit. The denominator sums eligible time across participants; it is not simply the number of people enrolled or the length of the study’s calendar.

Cochrane’s effect-measures chapter distinguishes count and rate data from binary participant outcomes. This article provides a fictional arithmetic example, not a nicotine-exposure calculation or a treatment evaluation. A rate can describe observed events without establishing how many distinct participants experienced them.

Cochrane: Cochrane Handbook, Chapter 6: Choosing effect measures

Keep a simple invented calculation in its units

Suppose a hypothetical dataset has 12 defined recurring events over 60 person-months of observation. The rate is 0.2 events per person-month. It is not “20 percent of people had the event,” because the denominator is time and some events may belong to the same person.

The fictional numbers illustrate units only. They do not describe a real quitting study or a safe amount of vaping. If the report instead counts each person once for whether an event occurred, it needs a participant denominator and a defined follow-up. Do not substitute one calculation for the other because their numerical values happen to look similar.

Check how observation time was accumulated

Participants may contribute different follow-up lengths. Read when time begins and ends, whether time after an event remains eligible, and how missing follow-up is handled. Without those definitions, a rate comparison may be difficult to interpret. A longer total observation time does not automatically mean more participants or better complete coverage.

If a report uses a rate ratio, keep that measure separate from a risk ratio or hazard ratio even if abbreviations resemble one another. Locate the full definition and uncertainty. This introductory guide does not prescribe a model for recurrent events or show how to derive participant probabilities from aggregate rates. Those tasks require the relevant data and assumptions.

Retain the unit in the final sentence

A useful summary names the event and says “per person-month” or the actual reported unit. If the headline instead gives a percentage of quitters, verify that the source also contains a participant-based quitting outcome. An event-rate result alone cannot supply that percentage or identify which adults achieved abstinence.

For a personal decision, ask an appropriate professional about any clinical implication. Everyday tracking can preserve your own count and its coverage, but it cannot recreate a study’s person-time comparison or establish a causal effect. The reading task is complete when the unit remains visible and a count per observed time is no longer mistaken for a fraction of people who quit.