Characteristics of Included Studies

MethodsParticipantsOutcomesInterventionsNotes
HARRIS2002
n= 446
Age: Mean 29
Sex: all females

Diagnosis:
 100% Thyroid-antibody-positive by Various endocrinology assessments
 Treatment Timing: Postpartum (6 weeks to 6 months)
 At-Risk Population?: Yes: early gestation thyroid-antibody-positive
Exclusions: At screening: existing thyroid disease, premature delivery. No other exclusion criteria described.

 Baseline: % EPDS >13: treatment group: 18.3%; placebo group 15%
 % RDC diagnosis: treatment group: 17.4%; placebo group 20.1%
 NOTE: “EPDS score was significantly one point higher in the active group than in the placebo group” at baseline

Data Used
 N's not compliant
 RDC(Research Diagnostic Criteria)any dep.diagnosis
 RDC major depressn (definite & probable combined)
 EPDS >=13
Notes: OUTCOMES TAKEN AT: 12, 16, 20, 24 weeks
DROPOUTS: (<80% compliant, LSE) treatment group = 22% (49/223); placebo group = 25% (56/223)
ASSESSOR BLINDNESS: Yes

Group 1 N= 223
 Thyroxine. Mean dose 100μg - DESCRIPTION: From 6 wks PP partpts recieved tablet supply (daily dose)
 COMPLIANCE: Tablet count @ subsequent visits (12, 16, 20, 24 weeks) check compliance & re-prescribe
 ASSESMENTS: Psychiatric (see outcomes) & medical assessments (fT4, fT3, TSH, etc)
Group 2 N= 223
 Placebo - DESCRIPTION: From 6 wks PP partpts received tablet supply (daily dose)
 COMPLIANCE: Tablet count @ subsequent visits (12, 16, 20, 24 weeks) check compliance & re-prescribe
 ASSESSMENTS: Psychiatric (see outcomes) & medical assessments (fT4, fT3, TSH, etc)

Outcome at 24 weeks only extracted for analysis (endpoint)
Study Type: RCT
Study Description: Prevention of depression (physical intervention)
SIGN Grade: 1++
Type of Analysis: Completer (>80% compliant) (prevention study)
Blindness: Double blind
Duration (days): Mean 210

Followup: No follow-up
Setting: UK
Notes: RANDOMISATION: Computer-generated sequence of numbers. All investigors and participants blind to allocation throughout treatment.
Info on Screening Process: Approx. 7500 women screened for thyroid antibodies; approx. 700 positive. 446 entered into study. Common reasons for non-involvement: premature delivery; pre-existing thyroid disease; refusal (requirement to attend outpatient facility regularly)
Results from this paper:
Data extracted into REVMAN

KEY METHODOLOGICAL CONCERNS
Unclear numbers ranomised into each condition (assumed equal numbers into each at randomisation).
Treatment group significantly higher EPDS score at baseline (confounding factor?).
No information given regarding numbers not completing the study (drop-outs), or on adverse reactions.
Several method points unclear: not clear what is meant by probable depression; “cut-off of 13 on EPDS”' is not more strictly defined (ie. >13 or >=13); if assuming “RDC: any” refers to both minor and major depression, numbers for each were not reported.

No response from authors on these queries (incorrect email address).
HARRISONHOHNER2001
n= 374
Age: Mean 22
Sex: all females

Diagnosis:
 No clinical diagnosis sought
 Treatment Timing: Antenatal (11–21 weeks, through to delivery)
 At-Risk Population?: No risk-related entry criteria were set
Exclusions: Inability to read English; infant with serious health problems (& from Levine et al.1997: taking medication; obstetric conditions; pre-existing diseases (eg. renal); frequent use of calcium supplements; <75% compliant on single-blind compliance test).
 Notes: Being a follow-up study to an RCT, only participants who had completed RCT were considered for inclusion in this study.
 Baseline: None (MH) given

Data Used
 EPDS >= 14
 EPDS (mean)

Notes: OUTCOMES TAKEN AT: 6wks PP (EPDS>=14) & 12wks PP (EPDS>=14 & EPDS mean)
LOST TO FLLW-UP:6wk fllw-up 48% (377/779) did not return survey; 12wk flw-up 68% (532/779) not sampled. Total sample only: nos each group NR.
ASSESSOR BLINDNESS: N/A (self-report)

Group 1 N= 192
 Calcium. Mean dose 2000 mg - DESCRIPTION: Elemental calcium (calcium carbonate tablets)
 WHEN: Taken in split dose (morning and evening meals)
 DURATION: From week 17 gestation (mean, range 11–21weeks) through to delivery
Group 2 N= 182
 Placebo - DESCRIPTION: Tablets identical to calcium tablets
 WHEN: Taken in split dose (morning and evening meals)
 DURATION: From week 17 (mean, range 11–21 weeks) through to delivery

Participants recruited from cohort from previous ongoing trial (Levine et al. (1997) Trial of calcium to prevent preeclampsia, New England Journal of Medicine, 337, 69–76). 779 surveys sent; 377 returned survey; 3 excluded from analysis (coding error).
Study Type: Follow-up study
Study Description: Prevention of postnatal depression (physical intervention)
SIGN Grade: 1++
Type of Analysis: Completer (prevention study)
Blindness: Double blind
Duration (days): Mean 133

Followup: 12 weeks
Setting: USA, Portland (Oregon) and Albuquerque (New Mexico). At 6-week follow-up both regions analysed; only Portland followed up at 12 weeks.
Notes: RANDOMISATION: Tablet packages numbrd by pharm. manufacturer (computer-generated random sequence). Particpts assigned next numbrd package @ enrolment.
Info on Screening Process: No information given; participants recruited as part of previous, ongoing trial (see Levine et al.1997).
Results from this paper:
Data extracted into REVMAN
KEY METHODOLOGICAL CONCERNS
Data gathered from two regions of America (Portland, Oregon and Albuquerque, New Mexico). At 6-week follow-up both regions included in analysis; at 12-week follow-up only Portland included. Outcomes: EPDS @ 6 weeks admin via questionnaire; @ 12 weeks via face-to-face visit (Portland site only).
At 6 weeks follow-up only Portland group showed trend towards difference between intervention and control groups on mental health outcomes. Possible regional effect? Confounding factor?
LAWRIE1998
n= 180
Age: Mean 32
Sex: all females

Diagnosis:
 No clinical diagnosis sought
 Treatment Timing: Immediate postpartum (during first few days)
 At-Risk Population?: No risk-related entry criteria were set
Exclusions: At initial recruitment: <19 years old, planning to use hormonal contraception, current antidepressant medication/psychotherapy.
There were no exclusions post-randomisation.
 Baseline: EPDS mean (SD): treatment group 13.3 (5.8), placebo group 12.6 (5.4)
 MADRS mean (SD): treatment group 6.2 (6.6), placebo group 6.4 (7.3)

Data Used
 EPDS >11
 MADRS >9
 MADRS >18
 MADRS (mean)
 EPDS (mean)
Notes: OUTCOMES TAKEN AT: Baseline, 1 wk (no data given for 1 wk), 6 wks, 12 wks postpartum
LOST TO FOLLOW UP: @ 6 wks into trial: 14% (13/90) placebo; 4% (4/90) trtmnt grp. @ 3 mnth follow-up: 10% (9/90) placebo; 3% (3/90) trtmnt grp.
ASSESSOR BLINDNESS: NR

Group 1 N= 90
 Noresthisterone enanthate. Mean dose 200 mg - DESCRIPTION: Noresthisterone enanthate (synthetic progestogen)
 DOSE: Single dose within 48 hours of delivery
 ADMIN: Via intramuscular injection over 2 mins to prevent guessing of contents (different viscosities of placebo and test solution)
Group 2 N= 90
 Placebo. Mean dose 1 mL -
 DESCRIPTION: Normal saline solution
 DOSE: Single dose within 48 hours of delivery
 ADMIN: Via intramuscular injection over 2 mins to prevent guessing of contents (different viscosities of placebo and test solution)
Study Type: RCT
Study Description: Prevention of postnatal depression (physical intervention)
SIGN Grade: 1+
Type of Analysis: Completer (prevention study)
Blindness: Double blind
Duration (days): Mean 1

Followup: 3 months
Setting: South Africa, Johannesburg
Notes: RANDOMISATION: Done in blocks of 4 using random number table (no further information given).
Info on Screening Process: Approx 720 women approached immediately postpartum (study required that they were using non-hormonal contraception) to volunteer, 180 accepted. Reasons for refusal: dislike of injections, unable to return for follow-up, generally not interested, etc
Results from this paper:
Data extracted into REVMAN

NOTE: Authors note that MADRS/EPDS are not diagnostic tools: their cut-off scores indicate “risk of depression”.
LLORENTE2003
n= 138
Age: Mean 31
Sex: all females

Diagnosis:
 No clinical diagnosis sought
 Treatment Timing: Immediate/early postpartum
 At-Risk Population?: No risk-related entry criteria were set
Exclusions:
Chronic medical condition, taking dietary supplements (excepting vitamins), smoker, been pregnant >5 times
 Baseline: BDI mean (SD): treatment group 7.1 (4.7); placebo group 6.5 (4.2)
 Participants assessed with BDI, EPDS and SCID-CV at baseline but only BDI reported - no indication of “% with 'diagnosis' “.

Data Used
 BDI-II (mean) (self-report)
 EPDS (mean)
 Depression diagnosis (DSM or similar)
 Leaving study early for any reason
Data Not Used
 BDI >=19 - N's not clearly reported
 BDI >= 9 - N's not clearly reported

Notes: OUTCOMES TAKEN: BDI (EPDS, SCID-CV to subsample also) @ baseline (shortly before delivery),3wks,2mths,4mths post-delivery
DROPOUTS: 27% (none due to adverse effects; nos by grp NR). LOST TO FLLW-UP: 14% DHA grp; 10% placebo grp
ASSESSOR BLINDNESS: NR

Group 1 N= 69
 DHA (docosahexaenoic acid). Mean dose 200 mg - DESCRIPTION: Algae-derived triglyceride capsules
 DOSE: 200 mg DHA/day
 ADMIN: start <1 week post-delivery
 COMPLIANCE: Dispensed more capsules than required -- instructed to return unused (was always within 5% of expected returned)
Group 2 N= 69
 Placebo - DESCRIPTION: Capsules identical to DHA capsules
 ADMIN: start <1 week post-delivery
 COMPLIANCE: Dispensed more capsules than required -- instructed to return unused (was always within 5% of expected returned)

All women planned to breastfeed infants for at least 4 months as part of larger study on effects of DHA
Study Type: RCT
Study Description: Prevention of postpartum depression (physical intervention)
SIGN Grade: 1++
Type of Analysis: Completer (prevention study)
Blindness: Double blind
Duration (days): Mean 120

Followup: 14 months
Setting: USA, Texas
Notes: RANDOMISATION: Double-masked fashion according to a computer-generated randomisation scheme- no other information given
Info on Screening Process: 138 women randomised; no other information given. All participants planned to breastfeed their infants for at least 4 months as part of larger study of effects of DHA on breastfeeding mothers and their infants (no reference given for the larger study).
Results from this paper:
Data extracted into REVMAN
KEY METHODOLOGICAL CONCERNS
No information on numbers screened given. No numbers of those randomised to each group (total randomised 138; therefore assume 69 in each group). “Opportunity sample” -- no risk factors/ no depression at entry.
BDI was the only outcome measure used at every assessment point, for whole sample. BDI dichotomous data not extracted: not clear if these numbers overlap (i.e. Are the people who display “moderate” symptoms (BDI >20) also represented in the “mild” numbers (BDI >10)?). Data reported, but not extracted: BDI >10: DHA group 9/44, placebo group 11/45; BDI >20: DHA group 4/44, placebo group 2/45.
EPDS and SCID-CV admin to sub-sample of population only, and only post-trial data reported in paper for these measures.

From: APPENDIX 18, CHARACTERISTICS OF REVIEWED STUDIES

Cover of Antenatal and Postnatal Mental Health
Antenatal and Postnatal Mental Health: The NICE Guideline on Clinical Management and Service Guidance.
NICE Clinical Guidelines, No. 45.
National Collaborating Centre for Mental Health (UK).
Leicester (UK): British Psychological Society; 2007.
© NCCMH. All rights reserved.

All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Enquiries in this regard should be directed to the British Psychological Society.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.