A combined in vivo and in vitro testing method for surgical masks is presented. The mask to be tested was placed over the nose and mouth of a manikin which was accurately constructed according to human measurements. The manikin was placed in a special counting chamber and a known number of bacteria were blown through the mask by a nebulizer and an automatic respirator, closely simulating a human sneeze or cough. The chamber was then evacuated completely through an Andersen sampler so that all bacteria in the chamber which had not been retained by the mask could be collected and counted within this sampler. The efficiency of the different masks was expressed as the percentage of bacteria retained by the masks as compared to a control when the manikin did not wear a mask. The set-up closely resembled that of a surgeon bent over a patient in the operating room when he coughs or sneezes while wearing a mask.

The relative efficiency of the different masks was found to be in this descending order: polypropylene fibers (Filtron), polyester-rayon fibers (Aseptex), glass fibers (Deseret), and paper.