Stratos Initiative

The validity and practical utility of observational medical research depends critically on good study design, excellent data quality, appropriate statistical methods and accurate interpretation of results. Statistical methodology has seen substantial development in recent times. Unfortunately, many of these methodological developments are ignored in practice. Consequently, design and analysis of observational studies often exhibit serious weaknesses. The lack of guidance on vital practical issues discourages many applied researchers from using more sophisticated and possibly more appropriate methods when analyzing observational studies. Furthermore, many analyses are conducted by researchers with a relatively weak statistical background and limited experience in using statistical methodology and software. Consequently, even ‘standard’ analyses reported in the medical literature are often flawed, casting doubt on their results and conclusions. An efficient way to help researchers to keep up with recent methodological developments is to develop guidance documents that are spread to the research community at large.

These observations led to the initiation of the STRATOS (STRengthening Analytical Thinking for Observational Studies) initiative, a large collaboration of experts in many different areas of biostatistical research. The objective of STRATOS is to provide accessible and accurate guidance in the design and analysis of observational studies. The guidance is intended for applied statisticians and other data analysts with varying levels of statistical education, experience and interests (click to enlarge).

Level 1: Low statistical knowledge

We have to assume that most analyses are done by analysts at that level. It is important to point out weaknesses of approaches which are often used despite of problems (eg categorizing continuous variables in the analysis; complete case analysis if some variables have missing values) and to propose methods which may not be optimal or state of the art, but which are easy to use and which are still acceptable from a methodological point of view. Required software should be generally available.

Level 2: Experienced knowledge

Here we should point to methodology which is perhaps slightly below state of the art, but doable by every experienced analyst. We should refer to advantages and disadvantages of competing approaches, point to the importance and implications of underlying assumptions, and stress the necessity of sensitivity analyses. If these issues are well understood it is most likely that a sensible analysis strategy is chosen for the specific question. Sufficient guidance about software plays a key role that this approach is also used in practise.

Level 3: Expert in specific area

To improve statistical models and to adapt them to complex problems in reality researches develop new and more complicated approaches. However, usually it is unclear whether the use of such an approach has relevant advantages in practise. Most often, advantages are presented in a small number of examples and in specific situations, but a more systematic comparison to the state-of-the-art is missing. Software requires specific knowledge and is not generally available.

This level would give an overview of recent research with statements about possible advantages and disadvantages of the approaches. It could help to identify important weaknesses when using level 2 proposals in more specific situations. It will certainly help to identify areas needing more methodological research and would trigger the development of software for more general use.

The Steering Group has decided to start with seven topics of general interest. Two topic groups were added later. Guidance documents will be developed by separate topic groups (TGs), each comprising experts in different area of statistical methodology, alongside applied researchers who may represent future end-users of the STRATOS documents. Each TG will start by developing guidance aimed primarily at level 2 statistical knowledge, which is perhaps slightly below state of the art. STRATOS structure and the initial road map (click to enlarge). Schema Stratos The STRATOS initiative is closely connected to the International Society of Clinical Biostatistics (ISCB) and was launched at a half-day Mini-Symposium on the last day of the ISCB meeting in Munich, in August 2013.




To co-ordinate the initiative, to share best research practices and to disseminate research tools and results from the work of the topic groups (TG), several cross-cutting panels have been started recently. They aim to develop recommendations (sometimes rather loose as for simulation studies, sometimes more strict as for STRATOS publications) and to provide the infrastructure for those aspects of the initiative that apply to all or most of the TGs, and to coordination of the efforts of the individual TGs. Recommendations aim to support, simplify and harmonize work within and across the TGs. They will also help increase transparency in deriving guidance documents in STRATOS.

The following Panels have been created to date:

Chairs: James Carpenter, Willi Sauerbrei
Find more information here
Chairs: Simon Day, Marianne Huebner, Jim Slattery
Find more information here
Literature Review
Chairs: Gary Collins, Carl Moons
Find more information here
Chair: Bianca De Stavola, Stephen Walter
Co-Chairs: Mitchell Gail, Petra Macaskill
Find more information here
Simulation Studies
Chairs: Michal Abrahamowicz, Harald Binder
Find more information here
Data Sets
Chairs: Hermann Huss, Saskia Le Cessie, Aris Perperoglou
Find more information here
Knowledge Translation
Chair: Suzanne Cadarette
Co-Chair: Catherine Quantin
Find more information here
Chairs: Joerg Rahnenfuehrer, Willi Sauerbrei
Find more information here
Contact Organizations
Chair: Doug Altman, Willi Sauerbrei
Find more information here