Skip to main content

Table 1 Key findings showing the characteristics of categorisation amongst the exposure variables in epidemiological studies

From: Assessing the reporting of categorised quantitative variables in observational epidemiological studies

Characteristics of categorisation % of articles & CI regions
Prevalence of categorisation 61% (CI = 39%, 80%)
Decision informing categorisation
Hypothesis-driven categories 7% (CI = 0%, 34%)
Unknown (reasons not provided in the articles) 93% (CI = 66%, 100%)
Criteria used for categorisation
Established external criteria (e.g., WHO standards) 14% (CI = 2%, 43%)
Arbitrary grouping 29% (CI = 8%, 58%)
Equally spaced interval grouping 36% (CI = 13%, 65)
Quantile grouping 21% (CI = 5%, 51%)
Number of categories used amongst grouped exposures
2 7% (CI = 0%, 34%)
3 7% (CI = 0%, 34%)
4 29% (CI = 8%, 58%)
5 29% (CI = 8%, 58%)
6 14% (CI = 2%, 34%)
10 14% (CI = 2%, 34%)
Proportion of trend testing 57% (CI = 29%, 82%)