Pleural drainage was differently performed in two similar neighboring hospitals (32.0 % vs. 58.2 %, p < 0.001), but the length of stay was shorter in the hospital using a more conservative approach (median 12 days vs. 18 days, p < 0.001). This result seemed unrelated to severity but associated with the shorter duration of intravenous treatment. This study adds to previous reports suggesting that pleural drainage is unnecessary in many cases; controlled studies are needed to determine which patients may actually benefit from its use.