Table 2 Commonly described selection processes at the population-level in the epidemiological literature classified according to the proposed framework
From: A novel framework for classification of selection processes in epidemiological research
Level | Type of mechanism | Timing of the selection | ||
|---|---|---|---|---|
At exposure entry | During exposure | Post-outcome | ||
(1) Population- level | (1.1) Selection in exposure | (1.1.1) Self-selection Confounding by indication Healthy worker hire effect | (1.1.2) Side effects causing exposure changes Healthy worker survivor effect | (1.1.3) Subclinical symptoms causing exposure changes |
(1.2) Selection in population composition | (1.2.1) Index event Self-selection | (1.2.2) Competing event Depletion of susceptibles | (1.2.3) Berkson’s fallacy | |
At study entry | During follow-up | Post-outcome | ||
(2) Study-specific | (2.1) Selection in exposure | (2.1.1) Self-selection | (2.1.2) Non-compliance1 | (2.1.3) Non-compliance1 |
(2.2) Selection in population composition | (2.2.1) Restriction of source population Study base definition Self-selection Healthy volunteer effect | (2.2.2) Loss to follow-up | (2.2.3) Loss to follow-up Selective participation in a case-control study |