Meta-analyses are considered the golden standard of evidence-based medicine,
giving us the highest quality evidence within the field. A meta-analysis is a transparent statistical analysis based on multiple
scientific studies on a topic,
leading to a pooled estimate of a given
effect across studies. The goal of a meta-analysis is to calculate the tendencies in scientific literature.
How can it help?
A meta-analysis is a great tool to get an overview of the effects and side effects of a drug or treatment. This is true in a wide range of fields: In pharma during drug development, in market access looking for the effect in current drugs on the market or testing one’s own drug against the current standard of treatment. These are all cases where a meta-analysis is the optimal choice. At Silvi.ai, we develop meta-analyses that live up to the highest standards, e.g. Cochrane and EUnetHTA guidelines.
To create a meta-analysis, Silvi automatically extracts data from the studies included in the screening phase. The Silvi document parser will parse the PDF of the study using OCR and detect which tables report effect measurements. Then Silvi uses machine learning to analyze the tables and extract the available outcomes such as endpoint, population, and efficacy. Currently Silvi can detect 85% of all outcomes. Silvi is directly connected to the statistical analysis tool “R”, ensuring a quick and integrated statistical processing of data.