Identifying patients from a large number of people by a small number of tests

An approach is proposed for screening patients from more people with less tests. It can be used for various diseases, and one of the direct application is the suppression of pandemic (e.g., coronavirus pneumonia). In this approach, people are divided into several groups, and each group has multiple people. For each group, the samples (e.g., respiratory secretions) of all the people are mixed to generate a merged sample. All the people can be excluded if the merged sample indicates no illness. If the merged sample indicates illness, the group will be divided into smaller groups, and then handled recursively in the same way. By doing so, negative groups are excluded gradually, and the group size becomes smaller and smaller. Finally, each group has only one person. This approach is especially suitable for the case when only small percentage of people are ill (e.g., infected).