Characteristics of Included Studies

MethodsParticipantsOutcomesInterventionsNotes
APPLEBY1997
n= 87
Age: Mean 25
Sex: all females
:
Diagnosis:
 100% Major or minor depression disorder by Revised Clinical Interview Schedule (CIS-R) =/>12
 Treatment Timing: Postpartum (baseline 6–8 weeks)
 At-Risk Population?: Not applicable -- treatment study
Exclusions: English not adequate; living outside district; EPDS score <10; chronic depression (>2 years) or resistant depression; current drug/alcohol misuse; severe illness requiring close monitoring /hospital admission; breastfeeding.
 Notes: PHARM STATUS (re. psychological intervention comparison): placebo was taken by all participants (fluoxetine was given to 2 other arms in the trial).

Data Used
 EPDS (mean)
 HRSD (mean)
Data Not Used
 Revised Clinical Interview Schedule (CIS-R) - Not compatible with other measures
Notes: OUTCOMES TAKEN AT: Baseline (6–8 weeks postpartum) & 1,4 & 12 weeks (approx 20 weeks postpartum) ITT LOCF/SDs calculated
DROPOUTS: Group 1: 29%; group 2: 26%; group 3: 38%; group 4: 29%
ASSESSOR BLINDNESS: Yes

Group 1 N= 21
 Counselling (six sessions + placebo) with - a.k.a. “Cognitive behavioural counselling”
 CONTENT: Pract & emot concerns and support; CBT-derived
 WHERE: Clinic
 DELIVERED BY: Psychols. (designed for e.g. HV after brief training)
 DURATION: 1 x 1hr then 5 x 30 min sessions (6 sessions over 12 weeks)
Group 2 N= 23
 Counselling (single session + placebo) with - a.k.a. “Cognitive behavioural counselling”
 CONTENT: Practical & emotional concerns and support; CBT-derived
 WHERE: Clinic
 DELIVERED BY: Psychologists (designed for e.g. HV after brief training)
 DURATION: 1 x 1-hr session
Group 3 N= 21
 Counselling (six sessions + fluoxetine).
 Mean dose ?NR - CONTENT:
 Counselling as in counselling + placebo groups
 No further information provided about drug administration/dose/etc.
Group 4 N= 22
 Counselling (single session + fluoxetine).
 Mean dose ?NR - CONTENT:
 Counselling as in counselling + placebo groups
 No further information provided about drug administration/dose/etc.

DATA QUERIES: Emailed author for dichotomous data. Reply: no additional data available.
Study Type: RCT
Study Description: Treatment of postpartum depression (TA) (both psychological and pharmacological interventions)
SIGN Grade: 1+
Type of Analysis: ITT - LOCF
Blindness: Double blind
Duration (days): Mean 90

Setting: UK
Notes: RANDOMISATION: By computer-generated random numbers (no further detail reported on procedure).
Info on Screening Process: 20-mth recruitment period. 2978 new mothers in a maternity ward were eligible. 2395 consented to screening 6–8 wks later. Of 503 with EPDS =>10, 406 agreed to further assessment. 188 participts met depression criteria; 101 declined and 87 were randomised.
Results from this paper:
KEY METHODOLOGICAL CONCERNS
No “no treatment” group - lack of pure control.
No information on drug administration/dose/side effects reported.
At initial screening, high number of women refused to enter trial because of possibility of randomisation to antidepressant treatment.
NB: 4-arm study comparing 1 or 6 sessions' counselling either with fluoxetine or placebo. Placebo groups used for comparisons in psychology review.
All available outcomes collapsed by factor (i.e. all fluoxetine; all placebo; all 1 session; all 6 sessions).
ARMSTRONG2003
n= 20
Age: Range 21–30
Sex: all females

Diagnosis:
 100% Depressive symptomatology by EPDS >=12
 Treatment Timing: Postpartum (any time between 6 weeks & 18 months)
 At-Risk Population?: Not applicable -- treatment study
Exclusions: Participation was subject to a fitness test/ GP approval. No other exceptional exclusions.
 Notes: PHARM STATUS: Half of participants taking medication; some receiving counselling.

Data Used
 EPDS (mean)
 EPDS >= 12
 GHQ-12 [self-report depression] mean
Data Not Used
 SSI (Social Support Interview) - Not being extracted
 DASS (Depression Anxiety Stress Scale) - Not being extracted
Notes: OUTCOMES TAKEN AT: Baseline; mid-treatment at 6 weeks; endpoint (after 12 weeks of treatment)
DROPOUTS: None reported
ASSESSOR BLINDNESS: N/A (self-report measures)

Group 1 N= 10
 Standard care - a.k.a. “Control”
 WHERE: Australia
 CONTENT: Researchers were available by phone contact
Group 2 N= 10
 Group exercise with social support -
 CONTENT: Mainly group pram-pushing with some group social sppt
 WHERE: Community-based DELIVERED BY: unclear (nurse?)
 DURATION: 1.5–2hrs exercise p/wk + estim. 1hr social sppt/wk [estim. total = 18–34.5hrs over 12 wks]
 COMPLIANCE: Mean 66% of sessions

DATA QUERIES: Author supplied dichotomous depression data and drop-out clarification.
Study Type: RCT
Study Description: Treatment (TA)
SIGN Grade: 1+
Type of Analysis: ITT
Blindness:
Duration (days): Mean 84

Setting: Australia
Notes: RANDOMISATION: By participant selection of 1 of 2 sealed envelopes, each allocation to a different group.
Info on Screening Process: Recruitment was from referral or advertisement. No reporting of numbers excluded during recruitment/ screening stage.
Results from this paper:
KEY METHODOLOGICAL CONCERNS
Small sample; unclear attrition across the study (author email reply: no dropouts from either group); 30%–40% recruited via advertising; 50% taking medication for postnatal depression and some receiving counselling.
ARMSTRONG2004
n= 24
Age: Mean 30
Sex: all females

Diagnosis:
 100% Depressive symptomatology by EPDS >=12
 Treatment Timing: Postpartum (any time between 6 weeks & 18 months)
 At-Risk Population?: Not applicable -- treatment study
Exclusions: Those physically unable to undertake the exercise were exluded. No other exceptional exclusions.
 Notes: PHARM STATUS: 55% taking antidepressants and/or counselling.
 Baseline: EPDS mean (SD): pram-walking group (n=9) 17.25 (4.00); support group (n=10) 17.17 (4.45)

Data Used
 EPDS (mean)
Data Not Used
 SSI (Social Support Interview)
Notes: OUTCOMES TAKEN AT: Baseline; mid-treatment at 6 wks; endpoint (after 12 wks of treatment); 3 mth follow-up (described but not reported)
DROPOUTS: 3 exercise vs 2 social support withdrew (various reasons)
ASSESSOR BLINDNESS: N/A (self-report measures)

Group 1 N= 12
 Group exercise - a.k.a. “Pram-walking exercise programme”
 CONTENT: Pram-walking towards target heart rate; muscle stretches
 WHERE: Community-based
 DELIVERED BY: “Facilitators” (nurse/SW)
 DURATION: 12 wks of 2 x 40min sesns per wk (plus 1 solo sesn)
 ATTENDANCE: 75%
Group 2 N= 12
 Social support (group) - a.k.a. “Social support group”
 CONTENT: Unstructured discussion for social & emotional but not practical support. Baby/child welcome.
 WHERE: Community centre
 DELIVERED BY: Nurse/SW
 DURATION: 1 x 90 min weekly session over 12 weeks
 ATTENDANCE: 73%
Study Type: RCT
Study Description: Treatment (TA)
SIGN Grade: 1+
Type of Analysis: Completer
Blindness:
Duration (days): Mean 84

Setting: Australia
Notes: RANDOMISATION: “Four-block randomised sequence” by sealed envelopes containing assignment, opened in a sequential manner.
Info on Screening Process: Over a 6-wk period, n=? women with depressive symptomatology responded to publicity or referral. 66 of these were EPDS =>12, of whom 42 decided not to participate. The 24 who decided to participate underwent a fitness test and were then randomised.
CHEN2000
n= 64
Age: Mean 29
Sex: 60 females

Diagnosis:
 100% “Postnatal distress” by BDI >=10
 Treatment Timing: Postpartum (approx. baseline after 3 weeks PP)
 At-Risk Population?: Not applicable -- treatment study
Exclusions: <18 years old; non-survival of infant; < junior high-school education; BDI < 10.
 Notes: PHARM STATUS: NR

Data Used
 BDI >= 10
 BDI Change Scores (continuous)
Data Not Used
 Perceived Stress Scale
 BDI (dichotomous) completer - Unclear reporting of post-randomisation attrition.
Notes:
OUTCOMES TAKEN AT: Baseline; endpoint (60/64 assessed)
DROPOUTS: 4 participants randomised to intervention never attended (moved away/lost interest/too busy)
ASSESSOR BLINDNESS: N/A self-report measures

Group 1 N= 34
 Social support (group) - a.k.a. “Support group”
 CONTENT: Discussion of: transition, postnatal stress, communication, life planning
 WHERE: Clinic
 DELIVERED BY: Nurse to groups of 5/6 mothers & infants
 DURATION: 4 x weekly sessions of 1.5–2hrs
 ATTENDANCE: 92% overall
Group 2 N= 30
 Standard care - a.k.a. “Routine care”
 CONTENT: NR
 WHERE: Taiwan

DATA QUERIES: Wrote to author for randomisation method detail, and clarification of attrition between randomisation (n=115) and “enrolment” (n=64). Replied 03/12/05.
Study Type: RCT
Study Description: Treatment (TA)
SIGN Grade: 1−
Type of Analysis: ITT (dichotomous) & completer (continuous)
Blindness:
Duration (days): Mean 28

Setting: Taiwan
Notes: RANDOMISATION: “Randomly assigned”. No further detail - author email reply suggests allocation not sufficiently random.
Info on Screening Process: 1107 mothers approached 2/3 days PP & 941 consented (completing & BDI at 3 wks PP). 414 returned BDI, of whom 115 met BDI criteria, of whom 64 agreed to study enrolment. 4 LSE intervention grp; 51 did not consent to begin treatment - no reasons recorded.
Results from this paper:
KEY METHODOLOGICAL CONCERNS
Unclear randomisation procedure and post-randomisation attrition
COOPER2003
n= 193
Age: Mean 28 Range 17–42
Sex: all females

Diagnosis:
 100% DSM-III-R major depression disorder by SCID for DSM-III-R
 Treatment Timing: Postpartum (from 8 weeks)
 At-Risk Population?: Not applicable -- treatment study
Exclusions: Premature delivery, infant abnormality, non-singleton
 Notes: PHARM STATUS: NR
 Baseline: “Social adversity” thought to be “unbalanced” between the groups

Data Used
 EPDS (mean)
 Depression diagnosis (DSM or similar)
Data Not Used
 Infant Cognitive/Dev Outcomes (see Murray 2003)
 Maternal-Infant Interaction (see Murray 2003)
Notes: OUTCOMES TAKEN AT: endpoint 4.5 MPP (88% assessed); 9 MPP (93%); 18 MPP (92%); 5 yrs PP (72%) not ITT
DROPOUTS: 1 rel/att & 2 psych refused allocation. 6 lis V; 1 rel/att; 8 CBT & 4 std care d/o (=<4/10 sesns) or moved away
ASSESSOR BLINDNESS: Yes

Group 1 N= 43
 Relationship/attachment-based therapy - a.k.a. “CBT
 CONTENT: Modified interaction guidance. Primary treatment aim not depr but infant interaction/management
 WHERE: Home
 DELIVERED BY: Therapists (extra training)
 DURATION: Sesn length NR for 10 wks (8–18 wks postptm)
 ATTENDANCE: 95% >4/10
Group 2 N= 52
 Standard care - a.k.a. “Routine primary care” control
 DELIVERED BY: GPs and HVs
Group 3 N= 48
 Listening visit - a.k.a. “Non-directive counselling”
 CONTENT: Airing of feelings on any current concerns as per Holden1989
 WHERE: Home
 DELIVERED BY: Therapists (extra training)
 DURATION: Session length NR for 10 wks (8–18 wks postpartum)
 ATTENDANCE: 85% attended >4/10
Group 4 N= 50
 Psychotherapy - a.k.a. “Psychodynamic therapy
 CONTENT: Understanding infant relatnship through mother's own early attachment history
 WHERE: Home
 DELIVERED BY: Therapists (extra training)
 DURATION: Sessn length NR for 10 wks (8–18 wks postpm)
 ATTENDANCE: 80% >4/10

DATA QUERIES: None. Psychotherapy arm used in general psychological therapy comparisons.
Study Type: RCT
Study Description: Treatment (TA)
SIGN Grade: 1+
Type of Analysis: Completer
Blindness:
Duration (days):

Setting: UK
Notes: RANDOMISATION: By drawing coloured balls. (No further details on allocation concealment.)
Info on Screening Process: 3222 primiparous women were screened early postpartum using EPDS by post. >=12 scorers were given further assesssment. 206 DSM-III-R depressed eligible participants were invited to the study; 13 refused, 193 were randomised.
Results from this paper:
HOLDEN1989
n= 55
Age: Mean 26
Sex: 60 females

Diagnosis:
 100% Depression by Goldberg's Standardised Psychiatric Interview
 Treatment Timing: Postpartum (approx. 12–18 weeks)
 At-Risk Population?: Not applicable -- treatment study
Exclusions: Not clearly reported
 Notes: PHARM STATUS: 12 women took antidepressants, although only 3 per group took therapeutic dose
 Baseline: More delivery complications in the intervention group

Data Used
 Depression diagnosis (DSM or similar)
Data Not Used
 EPDS (self-report) - median scores only reported
Notes: OUTCOMES TAKEN AT: Baseline (12 wks postpartum) and approx 13 wks later after treatment (approx 9 mths postpartum)
DROPOUTS: 5/55 d/o post-randomisation (allocation not reported). Completer data only reported n=50.
ASSESSOR BLINDNESS: Yes

Group 1 N= 26
 Listening visit - a.k.a. “Counselling”
 CONTENT: Rogerian, non-directive, non-judgemental
 WHERE: Home
 DELIVERED BY: HV brief training; unaware diagnosis
 DURATION: At least 30min non-infant-care-focused discussion x 8wks
 ATTENDANCE: Mean 8.8 visits; no further detail
Group 2 N= 24
 Standard care - a.k.a. “Control group”
 CONTENT: NR
 WHERE: UK, Scotland
 DELIVERED BY: HV

DATA QUERIES: Dropout allocation not reported (n=5); for continuous severity data, median scores only reported; no contact with author attempted (> 5-year-old study).
Study Type: RCT
Study Description: Treatment (TA)
SIGN Grade: 1+
Type of Analysis: Completer (allocation of dropouts not reported)
Blindness:
Duration (days): Mean 56

Setting: UK, Scotland
Notes: RANDOMISATION: “Allocated using random numbers”. No further detail on allocation procedure concealment.
Info on Screening Process: 734 women screened using EPDS at approx 6wks postpartum. >=12 scorers (total no. not reported) who were available attended a diagnostic interview at 12wks postpartum. 60 met research diagnostic criteria, of whom 5 declined and 55 agreed to be randomised.
HONEY2002
n= 45
Age: Mean 28
Sex: all females

Diagnosis:
 100% “Probable postnatal depression” by EPDS >12
 Treatment Timing: Postpartum up to 12 months
 At-Risk Population?: Not applicable -- treatment study
Exclusions: No psychotic symptoms; most recent baby over 12 months of age. No other exceptional exclusions.
 Notes: PHARM STATUS: Some antidepressant use - no details.
 Mean baby age at baseline = 5.4 months.

Data Used
 EPDS (mean)
 EPDS >= 12
Data Not Used
 Ways of Coping Checklist - Revised
 Duke-UNC Social Support Questionnaire
 Dyadic Adjustment Scale (DAS)

Notes:
OUTCOMES TAKEN AT: Baseline, post-treatment and 6-month follow-up (87% followed up; missing 13% data substituted with group means)
DROPOUTS: 4 dropped out of intervention (reason NR) but were assessed post-treatment
ASSESSOR BLINDNESS: NR

Group 1 N= 22
 Standard care - a.k.a. “Routine primary care”
 WHERE: Wales
 BY: HVs etc.

Group 2 N= 23
 CBT (group) - a.k.a.” Psychoeducational group”
 CONTENT: Education on practical and emotional coping; cognitive behav. management of thoughts and anxieties about motherhood; relaxation
 WHERE: NR
 DELIVERED BY: HVs in groups of 4–6
 DURATION: 2h x 8wks
 ATTENDANCE: NR

DATA QUERIES: None
Study Type: RCT
Study Description: Treatment (TA)
SIGN Grade: 1+
Type of Analysis: ITT
Blindness:
Duration (days): Mean 56

Setting: UK, Wales
Notes: RANDOMISATION: “Block randomisation”. (No further details on allocation concealment.)
Info on Screening Process: Women referred into study if attending mother and baby clinic and EPDS >12. Numbers screened not reported; 45 participants met criteria and were entered into study.
Results from this paper:
KEY METHODOLOGICAL CONCERNS
No diagnostic interview: EPDS only. Missing data at 6-month follow-up for 3 participants in each condition were replaced by group mean, not LOCF. Some post-treatment antidepressant use but not clear how many women involved, although used as a covariate in statistical analyses and not found to be important.
MANBER2004
n= 61
Age: Mean 33
Sex: all females

Diagnosis:
 100% DSM-IV major depression disorder plus HRSD >=14 by HRSD
 Treatment Timing: Antepartum (average 20 weeks gest. at baseline)
 At-Risk Population?: Not applicable -- treatment study
Exclusions: <18 years; outside 11–28 weeks' gestation at screening; receiving care outside the area; non-DSM-IV current non-psychotic major depressive episode criteria; <14 HRSD; diagnostic for >2yrs; current mood medication; various comorbid psychiatric and physical conditions.
 Notes: PHARM STATUS: None; women taking medication for mood were excluded.
 Baseline: High education and income levels

Data Used
 Clin. Interview: SCID-MDE (major depression)
 BDI (mean)
 HRSD (mean)
Notes: OUTCOMES TAKEN AT: Baseline, mid-trtmt, 8wk endpoint (89% assessed) & (responders only) 10wks PP
DROPOUTS: 3 active vs 2 ctrl vs 1 msg discontinued before endpt; 8wk response = no criteria for MDD & HRSD=<14 & =>50% HRSD reduction
ASSESSOR BLINDNESS: NR

Group 1 N= 20
 Acupuncture for depression - a.k.a.”Active acupuncture”
 CONTENT: Treatment individually tailored, specific for depression, adapted for pregnancy
 WHERE: NR
 DURATION: “Acute phase” only: 12 sessions of approx 30 mins for 8 weeks
 ATTENDANCE: NR
Group 2 N= 21
 Acupuncture (non-depression) - a.k.a. “Active control acupuncture”
 CONTENT: Treatment individually tailored, not specific for depression, adapted for pregnancy
 WHERE: NR
 DURATION: As per other acupuncture
 ATTENDANCE: NR
Group 3 N= 20
 Massage - CONTENT: Standardised, minimal verbal contact. Aimed to act as attention control with physical contact, relaxation and respite from stress
 WHERE: NR
 DURATION: As per acupuncture conditions
 ATTENDANCE: NR
Study Type: RCT
Study Description: Treatment (TA)
SIGN Grade: 1+
Type of Analysis: ITT (dichotomous) & completer (continuous)
Blindness:
Duration (days): Mean 56

Setting: US, California
Notes: RANDOMISATION: “Randomly assigned”; no further detail.
Info on Screening Process: Of a total of 88 women screened, 61 eligible pregnant women were recruited from obstetric clinics and from advertisements, and were randomised.
Results from this paper:
KEY METHODOLOGICAL CONCERNS: Limited randomisation description. Study has continuation phase but for responders only.
MILGROM2005
n= 192
Age: Mean 30
Sex: all females

Diagnosis:
 100% DSM-IV major or minor depression by DSM-IV-TR clinical interview
 Treatment Timing: Postpartum (mean age 18 weeks)
Exclusions: Premature birth; infant birthweight <2.5 kg, congenital abnormality, major abnormality; concurrent major psychiatric disorder; depression affecting competence to give informed consent (e.g. psychotic depression); risk requiring crisis management; participation in other psychological programmes; significant difficulty with English.
 Baseline: Mean EPDS score at screening (SD) 16.6 (4.3)

Data Used
 BAI (Beck Anxiety Inventory) (continuous)
 BDI (continuous)
Data Not Used
 Social Provisions Scale - not being extracted

Group 1 N= 46
 CBT (group) - CONTENT: Focus - life events, maternal mood, practical issues; relaxation deferred - earlier introduction of pleasant activities and time management
 WHERE: Clinic
 DELIVERED BY: Senior therapist
 DURATION: 12 weekly 90-min sessions
 ATTENDANCE: 67%
Group 2 N= 47
 Counselling (group) - CONTENT:Focus - life events, maternal mood, practical issues, supportive listening, history taking, problem clarification, goal formation, problem solving
 WHERE: Clinic
 DELIVERED BY: Senior therapist
 DURATION: 12 weekly 90-min sessions
 ATTENDANCE: 72%
Group 3 N= 66
 Counselling (individual) -
 CONTENT:Focus - life events, maternal mood, practical issues, supportive listening, history taking, problem clarification, goal formation, problem solving
 WHERE: Clinic
 DELIVERED BY: Senior therapist
 DURATION: 12 weekly 90-min sessions
 ATTENDANCE: 58%
Group 4 N= 33
 Standard care - CONTENT: Case management by maternal and child health nurse, referred to other agencies/services as necessary
 WHERE: Australia

Mean and SD data supplied by authors.
Study Type: RCT
Study Description: SIGN Grade: 1−
Type of Analysis: ITT
Blindness: Single blind
Duration (days): Mean 84
Followup: 12 months

Setting: Australia, clinical setting
Notes: Randomised: coded slips of paper drawn from a bag; individual randomisation unsuitable - recruitment randomised in cycles.
Info on Screening Process: 4148 screened (EPDS); 533 EPDS >= 12; 177 declined assessment; of 356, 192 eligible (DSM-VI diagnosis depression, agreed to participate, etc) and randomised.
Results from this paper:
KEY METHODOLOGICAL CONCERNS
Method of randomisation unclear, but clearly stated participants NOT randomised on individual basis; high attrition rates.
MISRI2000
n= 29
Age: Mean 33 Range 23–46
Sex: all females

Diagnosis:
 100% DSM-IV Major depres disorder with postpartum onset by Mini-Neuropsychiatric Interview (MINI)
 Treatment Timing: Postpartum (mean baby age 5 months at baseline)
 At-Risk Population?: Not applicable -- treatment study
Exclusions: Non-married/cohabiting
 Notes:
 PHARM STATUS: Use of medication is reported (no detail)

Data Used
 Mini-Neuropsychiatric Interview (MINI)
 EPDS (mean)
Data Not Used
 Kellner Symptom Questionnaire (self-report mood)
 Measures of Father's Wellbeing/Dyadic Adjustment
 Parental Bonding Instrument (PBI)
 GHQ (scaled version) [self-report depression]
 Dyadic Adjustment Scale (DAS)
Notes:
OUTCOMES TAKEN: Full psychiatric assessment was performed at baseline and the 7th visit (100% assessed)
DROPOUTS: None
ASSESSOR BLINDNESS: Not blinded

Group 1 N= 13
 Psychoeducation (with partner) - a.k.a. “Psychoed. visit for patients with partners”
 CONTENT: Partners attend 4/7 visits.
 Positive interaction encouraged.
 WHERE: Clinic (?)
 DELIVERED BY: Therapist
 DURATION: 6 weekly visits + 1 follow-up session 1 month later
 ATTENDANCE: 100%
Group 2 N= 16
 Psychoeducation (mother only) - a.k.a. “Psychoeducation visit for patients only”
 CONTENT: Not reported
 WHERE: Clinic (?)
 DELIVERED BY: Therapist
 DURATION: 6 weekly visits + 1 follow-up session 1 month later
 ATTENDANCE: 100%
Study Type: RCT
Study Description: Psychosocial treatment for postnatal depression (TA)
SIGN Grade: 1+
Type of Analysis: ITT
Blindness:
Duration (days):

Setting: Canada, reproductive mental health clinic
Notes: RANDOMISATION: Each new eligible participant was “randomly assigned” (no detail reported on procedure).
Info on Screening Process: Of total (n=?) women referred to a reproductive mental health programme for postpartum illness during the study period, 29 consenting women with DSM-IV and EPDS =>12 with consenting partners were enrolled in the study.
Results from this paper:
KEY METHODOLOGICAL CONCERNS
Small sample, unclear randomisation
OHARA2000
n= 120
Age: Mean 30
Sex: all females

Diagnosis:
 100% DSM-IV major depression by SCID for DSM-IV & HRSD >=12
 Treatment Timing: Postpartum (average 6 months at baseline)
 At-Risk Population?: Not applicable -- treatment study
Exclusions: Women not married or cohabiting <6 months; lifetime diagnosis (using SIDP-IV or SCID) of organic brain syndrome, schizophrenia/ antisocial personality/ bipolar disorder, psychotic depression; current SU; serious OCD; panic or eating disorders.
 Notes: PHARM STATUS: NR
 Baseline: Average depressive episode length at baseline = 7 months

Data Used
 HRSD (mean)
 Depression diagnosis (DSM or similar)
 HRSD > 6
 BDI > 9
 BDI (mean)
Data Not Used
 PPAQ (Postpartum Adjustment Questionnaire
 Social Adjustment Scale - Self-report (SAS-SR)
 Dyadic Adjustment Scale (DAS)
Notes:
OUTCOMES TAKEN AT: Baseline; mid-therapy (4 weeks & 8 weeks); endpoint (12 weeks; 83% assessed)
DROPOUTS: 12 IPT vs 9 waitlist. No reasons reported.
ASSESSOR BLINDNESS: Non-blind

Group1 N= 60
 Waitlist - a.k.a. Waitlist condition
 CONTENT: Fortnightly contact either for outcome assessment or via phone to check suicide risk/ability to wait for treatment
 WHERE: USA, Iowa
 ATTENDANCE: NR
Group 2 N= 60
 Interpersonal psychotherapy (IPT) - aka: IPT
 CONTENT: Manualised; goal-focussed; depression, roles, conflict & competence in interpers & postpart. contexts
 WHERE: NR (Clinic?)
 DELIVERED BY: Doctoral psychotherapist with 40hrs IPT for PPD training
 DURATION: 1hr x12 wks
 ATTENDANCE: NR
Study Type: RCT
Study Description: Treatment (TA)
SIGN Grade: 1+
Type of Analysis: ITT (dichotomous) and completer (continuous)
Blindness:
Duration (days):

Setting: USA, Iowa
Notes: RANDOMISATION: By random no tables, blocked by depress. history. Re-randomised after 77th & 108th particpt to achieve equal group ns. No more details.
Info on Screening Process: 20620 letters were sent to all recently delivered women. N=? consented to multistage screen for DSM-IV depression. Of 345 meeting criteria, 77 were ineligible, 132 declined and 16 were treated as training cases. 120 remaining were randomised.
Results from this paper:
KEY METHODOLOGICAL CONCERNS
Dyadic adjustment data: possible significant baseline differences. (IPT = 93.36 WLC = 87.60); high attrition rate.
ONOZAWA2001
n= 34
Age: Mean 32 Range 18–45
Sex: all females

Diagnosis:
 100% Major depression by EPDS >12
 Treatment Timing: Postpartum (approx. baseline 9 weeks)
 At-Risk Population?: Not applicable -- treatment study
Exclusions: Non-primiparous; non-singleton birth; birth outside of 37–42 weeks' gestation; no infant health complications
 Notes: PHARM STATUS: 1 massage participant taking antidepressants
 Baseline: Mean EPDS (SD): treatment group: 16.1 (3.09); control group: 16.7 (3.20)

Data Used
 EPDS (self-report)
Data Not Used
 Assessment of mother-child interaction on video
Notes:
OUTCOMES TAKEN AT: Baseline (approx 9 wks postpartum) and endpoint (5 wks later)
DROPOUTS: 7 (massage) vs 2 dropped out during the intervention phase
ASSESSOR BLINDNESS: N/A (self-report measures)

Group 1 N= 19
 Infant massage (group) - a.k.a. “Massage group”
 CONTENT: Infnt massage instructn for new parents. Observing & responding to infnt body lang. Participants also attended support group.
 WHERE: Hospital
 DELIVERED BY: Trained instructor
 DURATION: 1x 1hr pw for 5wks
 ATTENDANCE: NR
Group 2 N= 15
 Social support (group) - a.k.a. “Control group”
 CONTENT: Informal group discussion of practical parenting problems and coping strategies. Particpts in both conditions attended.
 WHERE: Hospital
 DELIVERED BY: Researcher
 DURATION: 1 x 30min sesn per wk for 5 wks.
 ATTENDANCE: NR

DATA QUERIES: 25 of those randomised withdrew before intervention, but are excluded from ITT as allocation is not reported. Authors supplied mean and SD for endpoint EPDS.
Study Type: RCT
Study Description: Treatment (TA)
SIGN Grade: 1+
Type of Analysis: Completer (continuous data only)
Blindness:
Duration (days):

Setting: UK
Notes: RANDOMISATION: “Randomly allocated”. No further detail on procedure or allocation concealment.
Info on Screening Process: 830 eligible women sent EPDS at 4 wks postpartm, of whom 581 completed and returned it, with 91 scored >12. Of these, 32 declined and 59 agreed to participate and were randomised, of whom 25 (allocation NR) did not participate due to inconvenient timings.
Results from this paper:
KEY METHODOLOGICAL CONCERNS
Attrition between randomisation and intervention (25/59; due mainly to inconvenient timings of the study) not counted in the endpoint analysis.
PRENDERGAST2001
n= 37
Age: Mean 32
Sex: all females

Diagnosis:
 100% DSM-IV major or minor depression by Clinical interview (NR) plus EPDS, MADRS etc.
 Treatment Timing: Postpartum up to 12 months
 At-Risk Population?: Not applicable -- treatment study
Exclusions: No exceptional exclusions
 Notes: PHARM STATUS: 1 in each group on some medication.
 Mean weeks postpartum at baseline: 14.5
 Baseline: 32% major, 68% minor depression. EPDS mean (SD): CBT 15.9 (2.8); control 13.7 (2.3). N.B. Significant difference.

Data Used
 EPDS (mean)
 EPDS >= 10
 MADRS (mean)
 MADRS >= 10
Data Not Used
 Parenting Stress Index (PSI)
 DASS (Depression Anxiety Stress Scale)
Notes: OUTCOMES TAKEN AT: Baseline; endpoint; then 6-month follow-up by postal report (90% followed up). NB: CBT group statistically significantly higher EPDS at baseline
DROP-OUTS: None during treatment phase
ASSESSOR BLINDNESS: Unclear

Group 1 N= 17
 CBT - a.k.a. “Modified CBT for PND”
 CONTENT: Psychoeducation, negative thought and emotion management, assertiveness training, coping skills
 WHERE: Home
 DELIVERED BY: Early childhood nurses with special training
 DURATION: 1hr x 6wks
 ATTENDANCE: 100% = 100%
Group 2 N= 20
 Standard care - a.k.a.”Ideal standard care”
 CONTENT: Non-specific emotional support, mothercraft advice
 WHERE: Clinic, Australia
 DELIVERED BY: Early childhood nurse with no additional training
 DURATION: 20–60 min x 6 wks
 ATTENDANCE: 45% >6 incl. 30% >3

DATA QUERIES: Emailed author for EPDS change scores between time 1 & 2 & time 1 & 3. Reply received Jan 06: no change scores available, but MADRS time 2 & 3 & diagnostic data @ time 2 (MADRS) provided and entered.
Study Type: RCT
Study Description: Treatment (TA)
SIGN Grade: 1+
Type of Analysis: ITT (dichot & continuous) & completer (continuous)
Blindness:
Duration (days): Mean 42

Setting: Australia
Notes: RANDOMISATION: Randomisation tables used
Info on Screening Process: Participants screened into study using EPDS >=12. Diagnoses established by clinical interview.
Results from this paper:
KEY METHODOLOGICAL CONCERNS
Control condition was virtually an active condition. Significant baseline differences in EPDS. Wrote to author for change scores; meanwhile used mean endpoint scores. Reply received Jan 06: no change scores available, but MADRS time 2 & 3 & diagnostic data @ time 2 (MADRS) provided and entered.
SPINELLI2003
n= 50
Age: Mean 29
Sex: all females

Diagnosis:
 100% DSM-IV major depression disorder by SCID for DSM-IV
 Treatment Timing: Antenatal (recruited between 6 & 36 weeks gest.)
 At-Risk Population?: Not applicable -- treatment study
Exclusions: Drug/alchohol misuse <6 months previously; acute suicide risk; comorbid Axis I disorder; current antidepressant medication.
 Notes: PHARM STATUS: No antidepressants 79% of Ps were Latina; study and therapy were conducted in English and Spanish. Majority of Ps = low soc/ec stat, poor soc support. Unclear which Ps delivered during treatment and which were treated only during pregnancy.
 Baseline: No significant baseline demographic differences.
 Mean 21 weeks' gestation at baseline.
 Mental health history: 73% of participants reported past history of major depression.
 Baseline continuous data: BDI= IPT: 17.94 (6.68); PSYED: 13.94 (6.11)
 EPDS= IPT: 20.09 (17.11); PSYED: 17.11 (5.92). HRSD-21= IPT: 24.63 (4.62); PSYED: 23.18 (5.70)

Data Used
 EPDS (self-report)
 BDI (self-report)
 Hamilton Rating Scale for Depression - HRSD (clin)
 Clinical Global Impression (CGI)
Data Not Used
 Maudsley Mother-Infant Interaction Scale
Notes:
OUTCOMES TAKEN AT: Weekly for 16 wks (% assessed unclear). BDI somatic cluster not taken.
DROPOUTS: 4 IPT vs 8 psychoed never attended. Completers = participts at 1 session at least (n=38); data is ITT for this grp using LOCF.
ASSESSOR BLINDNESS: NR

Group 1 N= 25
 Interpersonal psychotherapy (IPT) - a.k.a. “IPT for antepartum depression”
 CONTENT: IP and pregnancy issues incl. grief, roles, IP deficits
 WHERE: NR. Clinic?
 DELIVERED BY: Therapists with 5 days' training & manuals for IPT for pregnancy
 DURATION: 45 mins x 16 weeks
 ATTENDANCE: NR
Group 2 N= 25
 Psychoeducation (mother only) - a.k.a. “Parenting education control programme”
 CONTENT: Practical, non-emotional advice and active support. IPT-type intervention avoided.
 WHERE: New York, USA. Clinic?
 DELIVERED BY: Therapist
 DURATION: 45 mins x 16 weeks
 ATTENDANCE: NR

DATA QUERIES: Attendance unclear. No “improvement” threshold set (assumption: where scores show no positive change). Data reported as across 16-wk period rather than at endpoint. Author supplied mean & SD severity etc.
Study Type: RCT
Study Description: Treatment (TA)
SIGN Grade: 1+
Type of Analysis: Completer
Blindness:
Duration (days): Mean 112

Setting: US, New York outpatient clinic
Notes: RANDOMISATION: By random number tables (No further details on allocation concealment).
Info on Screening Process: 200 approached (recruited from outpatient clinics), of whom 150 excluded. Other patients self-referred via info in the press.No further details.
Results from this paper:
KEY METHODOLOGICAL CONCERNS
No attendance data given other than numbers not attending any sessions. Unclear definition of “improvement” (no threshold set, assume it is simply where scores did not show any positive change). Timepoints for data slightly unclear: e.g. recovery is assessed as occurring any time over 16-wk period, rather than at endpoint. Unclear which participants delivered during treatment and which were treated only during pregnancy. Included self-referred patients.
WICKBERG1996
n= 48
Age: Mean 28
Sex: all females

Diagnosis:
 25% Probable depression by EPDS >=12 & MADRS >=10
 Treatment Timing: Postpartum between 3 & 4 months (approx.)
 At-Risk Population?: Not applicable -- treatment study
 75% Major depression by DSM-III-R
 Treatment Timing: Postpartum between 3 & 4 months (approx.)
 At-Risk Population?: No risk-related entry criteria were set
Exclusions: Referral to psychiatric/psychological treament elsewhere; assessed as seriously ill and requiring specialist help.
 Notes: PHARM STATUS: None on antidepressants
 Baseline: No significant baseline differences on demographics, pregnancy or psychiatric history

Data Used
 DSM-III-R criteria for major depression
MADRS >= 10
Notes:
OUTCOMES TAKEN AT: Baseline (approx 3.5 mths PP) and endpoint 6 wks later
DROPOUTS: Total = 7: 2 list vis vs 2 std care in need of specialst treatmnt excl. Plus 2 LV vs 1
SC (allocatn NR; assume equal groups at start point)
ASSESSOR BLINDNESS: Yes

Group 1 N= 24
 Standard care - a.k.a. “Ordinary routine care”
 CONTENT: Possibility of visiting child health clinic whenever needed
 WHERE: Sweden
 DELIVERED BY: e.g. Child health nurses
Group 2 N= 24
 Listening visit - a.k.a. “Counselling visits”
 CONTENT: Non-judgemental, empathic, supportive. Mother as focus (not infant)
 WHERE: Sweden, at home
 DELIVERED BY: Child health nurse with 2 days' training
 DURATION: 1 hr x 6 weeks
 ATTENDANCE: NR

DATA QUERIES: To calculate dropouts, equal groups at start point assumed (because allocation unclear for 3 dropouts). Endpoint diagnosis data entered in Revman counts only baseline-diagnosis particps plus specialist referall dropouts. Continuous MADRS NR.
Study Type: RCT
Study Description: Treatment (TA)
SIGN Grade: 1+
Type of Analysis: ITT
Blindness:
Duration (days):

Setting: Sweden
Notes: RANDOMISATION: “Randomly allocated”. No further detail; nor for allocation concealment.
Info on Screening Process: 1874 aproached to complete EPDS twice, at 2 and 3 months postpartum. 1655 complied; those >=12 on both occasions (n=94) were given MADRS and DSM assessments. 57 scored >=10, of whom 16 were lost, ineligible or declining, leaving 48 to be randomised.
Results from this paper:
KEY METHODOLOGICAL CONCERNS
Unclear randomisation procedure and blinding. Allocation of dropouts not explicit. Continuous data (MADRS) reported in insufficient detail. Not all sample diagnosed with depression, just had MADRS above threshold.

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.

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.