Etminan et al (See C 27490 fo) found that elderly people taking lithium had approximately double the rate of motor vehicle crashes compared with controls. They imply that lithium is responsible and say that patients must be told of the increased risk. However, both the inference and the advice are unwarranted and unhelpful since lithium is simply a proxy for having bipolar disorder. As no information is given about the relative risk of having a crash in the presence of bipolar disorder, it cannot be justified to warn patients against taking lithium if they have this condition. It is quite possible — indeed clinical judgement would suggest — that a patient with bipolar disorder may be a much safer driver when taking a mood stabilising agent than when he or she is not. The authors include a comparison with carbamazepine, presumably to show that the increased risk is restricted to lithium. This comparison is also likely to be spurious, as most older people taking carbamazepine are probably being treated for other conditions, not for bipolar disorder. I am surprised that this paper passed statistical review when there is such an obvious confounding variable. As a result, patients will be subjected to probably unnecessary anxiety about their treatment and their fitness to drive. (Author/publisher)
Abstract