Table 3 Identified strengths and limitations of data linkage

Useful for hypothesis generating [36,38,38, 48]

Ecological fallacy [4, 13, 14, 18, 22, 23, 25, 33, 37, 47, 49, 52, 53]

Study questions touching on several fields [11, 19]

No access to individual-level data [12, 16, 27]

Large sample size and greater statistical power [9, 23]

Large data granularity [21, 41]

No extra costs and time [11, 23]

Unobserved confounding [17, 18, 23,18,33,23, 33, 34, 38, 39, 46]

Explore complex relations [17]

Data of different AST methods is combined [9, 15]

Possible to assess impact of nationwide interventions [27]

Cautiousness needed with external validity [5]

Adjustment for other factors possible [35]

Completeness of data [4, 35]

Not possible to link all entities [34, 40]