Table 2

Characteristics of articles selecting socioeconomic status (SES) constructs from the census to estimate socioeconomic differences in injury risk/outcome

CharacteristicPre cent (N)
Frequency of statistical association/gradient with an injury cause
 Cultural domain80 (5)*
 Demographics domain100 (6)*
 Education domain78 (9)*
 Ethnicity domain86 (5)*
 Housing domain50 (13)*
 Income domain80 (21)*
 Occupation domain98 (6)*
 Population domain78 (5)*
Rationale as to why the SES construct was chosen
 Specific reference to previous injury outcome study30 (10)†
 General reference to other health outcome study27 (9)†
 No reference to its use in previous research42 (14)†
Description of SES construct steps
 Complete description45 (15)†
 Incomplete description48 (16)†
 Minimal description6 (2)†
Methodological approach
 Pairwise comparisons of multiple variables52 (17)†
 Composite indicator (eg, principal component analysis)15 (5)†
 Only one SES variable assessed33 (11)†
Injury causes
 All cause (morbidity/mortality)25 (64)‡
 Burn/fire8 (21)‡
 Falls4 (10)‡
 Intentional (unspecified)6 (14)‡
 Intentional self-harm4 (9)‡
 Intentional third party18 (45)‡
 Motor vehicle collisions3 (7)‡
 Other0 (1)‡
 Pedestrian3 (7)‡
 Unintentional (unspecified)11 (27)‡
 Weapons related—intentional9 (23)‡
 Weapons related—unintentional9 (23)‡
Interaction between patient race and area SES assessed30 (10)†
Social gradient assessed21 (7)†
Geocoding error/matching discussed9 (3)†
Prevention recommendations discussed role of social class and health33 (11)†

  • Percentages may not equal 100% due to rounding.

  • *Percentages drawn from the number of instances effect was observed from all SES indicators classified within its domain.

  • †Percentages drawn from the number of instances the criterion was observed from all manuscripts reviewed (n=33).

  • ‡Percentages drawn from the number of instances the injury cause was evaluated against a measure of SES.