“…Two studies were excluded because spinal cord infarction was induced experimentally. Twenty-seven papers describing a total of 322 cases remained for analysis (Griffiths 1973, Zaki and others 1974, Griffiths and others 1975, De Lahunta and Alexander 1976, Greene and Higgins 1976, Zaki and Prata 1976, Hayes and others 1978, Chick 1979, Gill 1979, Kornegay 1980, Schubert 1980, Doige and Parent 1983, Gilmore and De Lahunta 1987, Penwick 1989, Dyce and Houlton 1993, Oliver 1993, Cauzinille and Kornegay 1996, Junker and others 2000, Hawthorne and others 2001, Gandini and others 2003, Axlund and others 2004,Grunenfelder and others 2005, Ueno and others 2005, De Risio and others 2007, De Risio and others 2008, Nakamoto and others 2009, Chung and others 2013). Two of these 27 papers included the same cases but presented different information about them; these cases were only counted once but data from both studies were utilised (De Risio and others 2007, 2008).…”