Intervention Review

You have free access to this content

Preventive interventions for postnatal psychosis

  1. Adib Essali1,*,
  2. Samer Alabed2,
  3. Aisha Guul3,
  4. Norah Essali4

Editorial Group: Cochrane Schizophrenia Group

Published Online: 6 JUN 2013

Assessed as up-to-date: 13 NOV 2012

DOI: 10.1002/14651858.CD009991.pub2

How to Cite

Essali A, Alabed S, Guul A, Essali N. Preventive interventions for postnatal psychosis. Cochrane Database of Systematic Reviews 2013, Issue 6. Art. No.: CD009991. DOI: 10.1002/14651858.CD009991.pub2.

Author Information

  1. 1

    Modern Psychiatry Hospital, Psychiatry Centre, Damascus, Syrian Arab Republic

  2. 2

    University of Oxford, Continuing Education, Oxford, UK

  3. 3

    Sunderland Royal Hospital, Sunderland, UK

  4. 4

    Kalamoon University, Damascus, Syrian Arab Republic

*Adib Essali, Psychiatry Centre, Modern Psychiatry Hospital, 27 Al Zahrawi Street, Rawda, Damascus, Syrian Arab Republic.

Publication History

  1. Publication Status: Edited (no change to conclusions)
  2. Published Online: 6 JUN 2013


Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion

Destounis 1965Allocation: not randomised, case series.

Characteristics of ongoing studies [ordered by study ID]
Nakigudde 2008

Trial name or titleNakigudde 2008 - The role of psychoeducation on perceived social support of postpartum others with a psychotic illness

MethodsAllocation: randomised.

ParticipantsPostpartum mothers with a psychiatric illness in the current postpartum period.

Must have been admitted to the mental health facility for the current episode.

Estimated N = 200.

Interventions1. 12 weekly sessions of family psychoeducation for a period of 3 months.

2. Encouraging drug compliance.

OutcomesPrimary outcome: perceived social support of mothers [3 months]

Secondary outcome: psychological distress of the caregivers [3 months]

Starting dateAugust 2008-October 2010.

Contact informationJanet Nakigudde


Makerere University College of Health Sciences

Kampala, Uganda, P.O.Box 7072

NotesUnclear if this planned study ever started. No reply to enquiries.

Paykel 2000

Trial name or titleProphylaxis of puerperal psychosis

MethodsAllocation: randomised.

Blinding: blinded - no further details.

Duration: 2 months.

ParticipantsDiagnosis: previous schizoaffective and atypical psychoses of the kind that characterise as early onset post partum psychosis.

InterventionsWithin 72 hours:

1. Lithium.

2. Haloperiodol.

3. Placebo.


Starting dateOctober 1992-1995

Contact informationProf Eugene S Paykel
Box No 189
University Dept of Psychiatry
Addenbrooke's NHS Trust
Telephone: 01223-741928

NotesUnclear if this planned study ever started. No reply to enquiries.

Summary of findings for the main comparison. Preventive interventions compared to no preventive interventions for women with or at risk of postnatal psychosis

Preventive interventions compared with no preventive interventions for prevention of postnatal psychosis

Patient or population: Women at risk of postnatal psychosis
Settings: Anywhere
Intervention: Preventive interventions
Comparison: no preventive interventions

OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No of Participants
Quality of the evidence

Assumed riskCorresponding risk

No preventive interventionsPreventive interventions

Recurrence of postnatal psychosisSee commentSee commentNot estimable0
See commentWe identified no trial-based data - these questions remain unanswered.



Service utilisation outcomes - e.g. admitted to hospital

Adverse effects/events

Table 1. Non-RCT reports - identified in post hoc PubMed search

ReportStudy typeParticipantsInterventionsOutcomes

Stewart 1988Case-series4 women with history of postnatal psychosisLithium (900–1200 mg/d) immediately after delivery.All women did not develop postnatal psychosis.

Stewart 1991Case-series21 high risk post-partum women

Lithium (750-1200 mg).
Started within 24 hours of delivery. N = 16.

Started at 34 weeks. N = 4.

Throughout pregnancy. N = 1.
Two (10%) relapsed with postnatal psychosis.

Austin 1992Controlled-trial17 women with bipolar disorder or puerperal affective psychosis.Lithium, during pregnancy or within 48 hours after birth, for 3 months with serum levels of 0.4 mmol/L or more. N = 9.

Non-treatment group. N = 8.
Two (22%) in the lithium group relapsed with postnatal psychosis 10 days postpartum.

Six (75%) of the non-treatment group relapsed with postnatal psychosis.

Cohen 1995Retrospective
27 women with bipolar disorder who had experienced at >1 episode of bipolar disorder prior to index pregnancy. Prophylactic mood stabilisers (lithium). N = 14.

Did not receive any prophylactic treatment. N = 13.
No women taking a mood stabiliser developed postpartum psychosis. 1 developed affective instability 3 months postpartum. 

8 who received no treatment experienced manic or depressive relapse within the first 3 months postpartum. No information re postnatal psychosis available.

Sichel 1995Case-series11 women given oestrogen following delivery. 7 had history of puerperal psychosis and 4 of puerperal major depression.Oral premarin (10 mg/ day) administered following delivery in decreasing dosages over 4 weeks.

1 experienced psychosis relapse during first postpartum week.

Kumar 2003Case-series29 pregnant women with diagnosis of hypomania (bipolar II), mania (bipolar I), or schizoaffective disorder.Transdermal oestrogen:
1. 200 µg/day. N = 13.

2. 400 µg/day. N = 3.
3. 800 µg/day. N = 13.

All within 48 hours after delivery and reduced by half every 4 days for 12 days. 
12 relapsed in total.

Women taking oestrogen 800 µg/day needed less subsequent psychotropic medication and were discharged sooner than those taking lower doses.

Wisner 2004Controlled-trial26 pregnant women diagnoses with bipolar disorder.Intervention:
Divalproex sodium (VLP) 250 mg twice a day immediately after birth. N = 15.

Postpartum monitoring alone without medications. N = 11.
No women developed postnatal psychosis.

In the intervention group, 10 (67%) developed postpartum hypomania/mania, depression, or mixed-state diagnoses vs 8 (73%) of non-intervention group.

Sharma 2006Controlled-trial25 women with diagnosis of hypomania (bipolar II), mania (bipolar I).Olanzapine 5–10 mg alone or in combination with an antidepressant or mood stabilizer. N = 11.

Either antidepressants, mood stabilizers, or no medication for minimum of 4 weeks after delivery. N = 14.
No postnatal psychosis in olanzapine group.

2 women (lithium and no medication) experienced postnatal psychosis in non-olanzapine group.

Bilszta 2010Controlled-trial23 women with history of bipolar disorder or postpartum psychosis.

Controls: 15 healthy pregnant women.
1. Mood-stabilisers. N = 7.
2. Antidepressant or antipsychotic. N = 8.

3. No medication. N = 4.

All through pregnancy and postnatal period.
No women taking a mood stabiliser relapsed during study.

Bergink 2012Case-series29 women with history of postpartum psychosis without any manic or psychotic symptoms outside postpartum period. Lithium prophylaxis (0.8 mmol/L minimum plasma level) immediately postpartum.4 of 9 women without postpartum prophylaxis and history of postpartum psychosis relapsed (44%).
No relapse reported among women who initiated postpartum prophylaxis.

 Ordered chronologically
Table 2. Suggested design for a study

MethodsAllocation: randomised.

Blinding: none.

Duration: conception to 6 months after delivery

ParticipantsPregnant women with past history of postnatal psychosis.

N = any.*

Interventions1. Package of care, tailored to needs and likes of the woman and her carers, focusing on support of tailored use of medication, and ease of access to services.

2. Standard package of care.

OutcomesPostpartum psychosis.

Service utilisation outcomes.

Adverse effect/event.

Notes* Any size of trial would be proof-of-concept - although small trials (> 300) will be underpowered to find outcomes of interest.