Non-interventional studies - limitations, challenges, and opportunities

Randomized clinical trials (RCTs) are the gold standard for evaluating efficacy and safety of medicinal products, particularly for obtaining marketing authorization. The evidence from RCTs has high internal validity due to randomization and, most often, blinding. However, their external validity is limited, as they collect data form a homogeneous study population, in an experimental setting for a relative short time. Therefore, there is an increasing recognition that real world evidence (RWE) needs to complement RCTs. Large RCTs are challenging mainly due to high cost and difficulties in patient recruitment, thus RWE is mostly generated from observational studies, which are non-interventional (NIS). NISs have considerable external validity as they collect data from large number of participants from normal treatment settings and are useful to collect real life information on drug safety. Nevertheless, absence of randomization and control group limit their internal validity and there are concerns about data quality and hidden biases.

In this dissertation, we studied the value of NIS to determine factors associated with the initial dosing and up-titration of propiverine and study how dosing relative to other factors affects treatment outcome on overactive bladder (OAB) symptoms. Data from 2 prospective NIS of 1335 and 745 OAB patients, respectively, receiving propiverine, were analyzed post-hoc, using multivariate analysis. Furthermore, as an example of a NIS with retrospective data collection, we evaluated the potentials of routinely collected inpatient data as a complementary source for existing drug surveillance systems. We sampled 2326 cases using a standardized procedure and evaluated them regarding drug relation and preventability of event. Lastly, this thesis covers two common basic statistical reporting issues, as an implication for analysis and reporting.

This work showed the potential of NIS to complement the evidence from RCT to improve the effectiveness of the treatment by providing guidance for choosing the correct dose. Moreover, we found out that routinely collected inpatient data in the context of a NIS study design could be used as markers for non-preventable adverse drug effects and medication errors, thus providing a meaningful complement to existing drug surveillance system. Finally, we highlighted the importance of a proper statistical analysis and reporting in the interpretation and reporting of the study finding in evidence-based medicine.



Citation style:
Could not load citation form.


Use and reproduction:
All rights reserved