Table 1. Aspects of databases that reflect the quality of data contained within.
Quality Domain | Explanation |
---|---|
Coverage | |
Representativeness | How well does the data source represent the population that it intends to? |
Completeness of recruitment | This feature measures the extent to which all eligible individuals have been included in the data collection scheme |
Variables included | What is the extent of the data collected on each individual? Are demographic, exposure, outcome, and confounding variables present? |
Completeness of variables | What is the extent of the missing data? |
Accuracy | |
Collection of raw data | Is the raw data collected or are aggregate averages collected? |
Explicit definitions | Are the variables explicitly defined? |
Explicit rules | Are there explicit rules for deciding how variables are recorded? For example, the timing of physiologic variables. |
Reliability of coding | Was the reliability of coded conditions and interventions tested? |
Independence of observations | Was the data recorder blinded to patient outcome at the time the data were collected? |
Data validation | Were data validated using outside sources? Were there consistency checks? |
Based upon Black, et al.(42)