The feasibility of using a double blind experimental cross-over design to study interventions for sick building syndrome.

The study of the causes of Sick Building Syndrome and its possible solutions have been bedevilled by methodological problems. This pilot study assessed the viability of using an expermental double blind cross-over study to overcome such difficulties. The experiment involved varying the rate of supply of outdoor air from 10 cubic feet per minute per person (cfmpp) to 20 cfmpp or 50 cfmpp by controlling the building's heating, ventilation and air-conditioning (HVAC) systems. Over 6 weeks, two trials of the rates were undertaken in a random order blinded or kept secret from both subjects and investigators. Each week unblinding, rating of the office environment and occurrence of SBS symptoms were measured. 254 of the 305 workers co-operated in the experiment. Leakage of air from the building led to problems in delivering the lowest rate of ventilation. The reporting of symptoms reduced during the trial, a trend which could be controlled by recommended design modifications. Secrecy about the order was maintained during the experiment. Weekly non-response did not lead to a response bias but reduced the number of subjects available for by a third for each trial. The report concludes that a modified version of this method could be used to evaluate many proposed responses to Sick Building Syndrome.