Table 13.

Summary of Lithium Discontinuation Guidance

Guideline	Discontinuation guidance
APA	If tolerated and effective, treatment should continue for at least an acute phase (typically 4–8 wk) and potentially beyond for relapse prevention.
BAP	Lithium should be maintained in combination with an antidepressant for at least 1 y to prevent relapse, though little available evidence in continuation phase. Routine use of lithium monotherapy in continuation phase not recommended. Supported their recommendations with conflicting evidence demonstrating that lithium had nonsignificant benefit over placebo or antidepressants for prophylaxis (Burgess et al., 2001; Cipriani et al., 2006), but lithium plus an antidepressant more effective than antidepressant monotherapy in preventing relapse in TRD patients who responded to lithium augmentation or ECT (Bauer et al., 2000; Sackeim et al., 2001).
CANMAT	–
CPG-S	–
ICSI	Care should be taken when discontinuing lithium as abrupt withdrawal associated with higher relapse rates
MPG	Advised against lithium monotherapy for prophylaxis. At least 2 y continuation in patients with ≥2 recent episodes with significant functional impairment, then reevaluate. If first episode, continue for 6–9 mo following remission. Following noncompliance, recommence at previous dose if adherence and tolerance good (Abou-Saleh et al., 2017a). Treatment should be indefinite if adherence good and treatment well tolerated, particularly if suspected bipolarity. Few data relating to lithium discontinuation in unipolar depression. Should be reduced slowly over at least 1 mo, avoiding incremental reductions >0.3 mmol/L in recurrent depression. Referred to work that proposed lithium should be used for prophylaxis in depression if had been 2 depressive episodes in 5 y or following 1 severe episode with strong suicide risk (Abou-Saleh et al., 2017b).
NICE	Responders with multiple historical relapses should remain on medication regardless of length of treatment pre-response. Insufficient evidence to determine effect beyond 2 y or if relevant to first-episode patients. If a medication to be stopped, it should be the augmenter not AD. Relapse likelihood lower if lithium continued (Prien et al., 1984). Not enough evidence to determine clinically important difference between continuing lithium as monotherapy or discontinuing both lithium and AD (i.e., placebo).
RANZCP	Care should be taken when discontinuing lithium as abrupt withdrawal associated with higher relapse rates (Bschor et al., 1999). Lithium monotherapy in maintenance phase could be considered if ADs not well tolerated (Cipriani et al., 2006).
TMAP	–
WFSBP	Continuation for at least 1 y if patient responds. Gradual withdrawal tapered over at least 3 mo if treatment has been >6 mo. If symptoms reoccur, maintenance dose of antidepressant and lithium should be resumed. Recommended administration for 2–4 wk and then monitor patient response (Bschor et al., 2003). Efficacy of lithium maintenance treatment well established (Coppen et al., 1990; Schou, 1997; Davis et al., 1999; Bauer et al., 2000; Paykel, 2001; Bschor et al., 2002). Evidence for lithium monotherapy in prophylaxis not sufficient (Souza and Goodwin, 1991; Burgess et al., 2001), but more recent meta-analysis showed efficacy of lithium when used for preventative purposes, though relative efficacy was not (Cipriani et al., 2006). Lithium + AD in continuation phase more beneficial than antidepressant + placebo, AD monotherapy, or lithium monotherapy (Kim et al., 1990; Bauer et al., 2000; Bschor et al., 2002). Stopping lithium can lead to relapse or recurrence.

Abbreviations: –, not reported by guideline; AD, antidepressant; AAPs, atypical antipsychotics; APA, American Psychiatric Association; BAP, British Association of Psychopharmacology; CANMAT, Canadian Network for Mood and Anxiety Disorders; CPG-S, Clinical Practice Guidelines in the Spanish NHS; ECT, electroconvulsive therapy; ICSI, Institute for Clinical Systems Improvement; MPG, Maudsley Prescribing Guidelines; NICE, National Institute for Health and Care Excellence; RANZCP, Royal Australian and New Zealand College of Psychiatrists; TMAP, Texas Medication Algorithm Project; WFSBP, World Federation of Societies of Biological Psychiatry.