Corbett & Alda raise the interesting question of how long a treatment trial should last before it can be established whether lithium is effective for a specific individual. As they note, existing experimental studies are not necessarily designed to address that particular question, which raises significant conceptual and analytic challenges.

Their interesting suggestion of assessing maintenance treatments through comparison of cumulative morbidity over long periods may be becoming a more feasible prospect through the combination of electronic health records analysis^{Reference Hayes, Pitman, Marston, Walters, Geddes and King1} with the increased availability of longitudinal mood monitoring outside experimental studies.^{Reference McKnight, Bilderbeck, Miklowitz, Hinds, Goodwin and Geddes2}

Pending these new data, the available evidence indicates that lithium is likely to reduce the risk of manic relapse rapidly, whereas full effects against depressive relapse probably develop over a longer period.^{Reference Taylor3}