Table 2.

Summary table

VariableRCTObservational with Existing DatabaseaObservational with Primary Data Collection
Timeliness of datab+++++++
Data quality+++++++
Potential for bias+++++
Logistical complexity++++++
  • A scale of + to +++, where + is the lowest and +++ is the highest, or a scale of $ to $$$, where $ is the cheapest and $$$ is the most expensive, is used. RCT, randomized clinical trial.

  • a Dialysis Outcomes, Practice and Patterns Survey (DOPPS) data may be considered an existing database in most respects.

  • b Timeliness of data may depend on when a postmarketing study is commenced. If a study is initiated immediately after drug/intervention reaches the market, it may take several years until a sufficient experience with that intervention is captured within large registries, such as the US Renal Data System, and available for analysis. In this case, dedicated data capture using a prospective approach (e.g., an RCT) may be quicker.