Arch Womens Ment Health (2006) 9: 303308

DOI 10.1007/s00737-006-0145-9

Original contribution

Patient choice of treatment for postpartum depression: a pilot study

T. B. Pearlstein1;2;3 , C. Zlotnick1;2;3 , C. L. Battle1;3 , S. Stuart4 , M. W. OHara5 ,
A. B. Price1;2 , M. A. Grause2 , and M. Howard1;2
1
Department of Psychiatry and Human Behavior, Brown Medical School, Providence, Rhode Island, U.S.A.
2
Women and Infants Hospital, Providence, Rhode Island, U.S.A.
3
Butler Hospital, Providence, Rhode Island, U.S.A.
4
Department of Psychiatry, University of Iowa, Iowa City, Iowa, U.S.A.
5
Deparment of Psychology, University of Iowa, Iowa City, Iowa, U.S.A.

Received March 7, 2006; accepted July 19, 2006

Published online August 21, 2006 # Springer-Verlag 2006

Summary for infant and child development (Grace et al, 2003;

Objective: The lack of systematic efficacy research makes the selec-
tion of optimal treatment for postpartum depression (PPD) difficult.
Moreover, the treatment decisions for women with PPD who are breast-
feeding are heavily influenced by their concerns about infant exposure
to antidepressant medication. The objective of this pilot trial was to ex-
amine the clinical characteristics of women with PPD associated with
treatment selection.
Method: This open pilot trial offered 23 women with PPD one of
3 treatment options: sertraline, interpersonal psychotherapy (IPT), or
their combination administered in an outpatient mental health setting
over 12 weeks. Baseline and treatment outcome measures included the
Hamilton Rating Scale for Depression (HRSD), the Beck Depression
Inventory (BDI) and the Edinburgh Postnatal Depression Scale (EPDS).
Results: Completers across all 3 treatment groups (n 18) experi-
enced significant clinical improvement with each of the 3 treatment mo-
dalities on the HRSD ( p < 0.001), BDI ( p < 0.001) and EPDS ( p < 0.001).
There were trends for women with a prior depression to more fre-
quently choose sertraline as a treatment (alone or with IPT, p 0.07),
and for women who were breastfeeding to choose sertraline (alone or
with IPT, p 0.10) less frequently.
Conclusion: In this small sample of women with PPD, most women
chose IPT with or without sertraline. A larger randomized study could
further confirm the suggested predictors of treatment selection identi-
fied in this study: previous depression and breastfeeding status.
Keywords: Postpartum depression; breastfeeding; sertraline; inter-
personal psychotherapy; antidepressant medication; treatment selection.
Introduction
Postpartum depression (PPD) is increasingly recognized
as causing substantial morbidity and functional impair-
ment in the mother and long-term negative consequences
Newport et al, 2001). Women with PPD, their families
and their clinicians face the dilemma of choosing treat-
ment with psychotherapy, antidepressant medication, or
their combination; with minimal systematic research com-
paring these modalities to guide them. The factors gov-
erning the treatment selected by a woman with PPD are
diverse and include knowledge about untreated maternal
illness and available treatments, risk perception and
individual beliefs (Sit & Wisner, 2005).
Among psychotherapies, interpersonal psychotherapy
(IPT) has been demonstrated to be efficacious in the
treatment of PPD (OHara et al, 2000), and is ideally
suited to the role transition and social role conflict issues
associated with being newly postpartum (Segre et al,
2004). Other psychotherapies have been shown to be
efficacious for mild to moderate PPD as well (Dennis,
2004; Stuart et al, 2003). One potential barrier to care
for women with PPD is the reduced availability of psy-
chotherapists trained in IPT or cognitive-behavior ther-
apy (CBT) in many areas (Abreu & Stuart, 2005).
It is generally assumed that the efficacy and tolerabil-
ity of antidepressant medication in PPD is equivalent to
that in nonpuerperal MDD, but this assumption has only
been empirically examined to a minimal degree (Abreu &
Stuart, 2005). Recent reviews of open trials, case series and
case reports of antidepressant medication in women with
PPD suggest efficacy (Abreu & Stuart, 2005; Dennis &
304
Stewart, 2004). However, many of these antidepressant
trials excluded women who were breastfeeding. The only
existent placebo-controlled antidepressant treatment study
for PPD in 87 women included 4 treatment arms com-
prising fluoxetine or placebo, each with one or 6 coun-
seling sessions based on cognitive therapy principles
(Appleby et al, 1997). Fluoxetine was superior to pla-
cebo regardless of whether women received one or 6
counseling sessions. Breastfeeding women were exclud-
ed in this study. A recent study compared paroxetine and
the combination of paroxetine=CBT in 35 women with
comorbid postpartum anxiety disorders and depression
(Misri et al, 2004). Approximately half of the total sam-
ple was breastfeeding. Both treatments led to significant
symptom improvements; however, there were no signif-
icant differences between the treatments.
Besides the lack of randomized, placebo-controlled
trials of the efficacy of antidepressant medication in
PPD to inform treatment selection, the treatment options
for postpartum depressed mothers are further compli-
cated by the lack of prospective, controlled data about
potential negative short-term and long-term effects in
breastfeeding infants exposed to antidepressant medica-
tion. A growing observational database suggests mini-
mal or no short-term adverse effects in infants (Burt et al,
2001; Misri & Kostaras, 2002; Weissman et al, 2004),
but many mothers are still hesitant to take medication
without definitive studies regarding long-term effects on
child development (Sit & Wisner, 2005).
A recent systematic review reported that patients with
major depressive disorder in primary care prefer psy-
chotherapy over antidepressant medication for treatment
if psychotherapy is available (van Schaik et al, 2004). A
recent study in 405 postpartum women without depres-
sion reported a significant preference for psychotherapy
over pharmacotherapy (Chabrol et al, 2004). This find-
ing was particularly significant in breastfeeding women,
even after psychoeducation about antidepressant medi-
cation and breastfeeding was provided. A chart review
study of women with PPD in a perinatal psychiatric
partial hospital program reported that 61% of 74 breast-
feeding mothers were willing to start antidepressant
medication, which was significantly less than the 86%
rate among 145 non-breastfeeding mothers (Battle et al,
2006). Consistent with this, a previous study suggested
that women with PPD prefer psychotherapy to pharma-
cotherapy for treatment (Whitton et al, 1996). Another
study reported that 31% of 13 breastfeeding women with
PPD declined antidepressant medication because they
were breastfeeding, and compliance was poor among
T. B. Pearlstein et al
those who did start due to their concerns about not hear-
ing the baby at night, problematic side effects, harmful
long-term effects, and the stigma associated with anti-
depressant medication (Boath et al, 2004).
In sum, the treatment decision of a woman with PPD
is complicated by frequent preference for psychotherapy
which may not be easily accessible, the lack of systema-
tic efficacy research for somatic treatments and by con-
cerns about infant exposure to antidepressant medication
with breastfeeding. This pilot study was designed to
evaluate the feasibility of recruitment and the identifi-
cation of factors governing treatment choice for a larg-
er randomized controlled trial of IPT, sertraline and a
sertraline=IPT combination. This small open trial that
we report is the first study, to our knowledge, to explore
the clinical characteristics associated with these 3 treat-
ment choices. This open trial also compared pre- and
post-treatment depression severity levels in order to
evaluate whether women responded overall to 12-week
treatments for PPD.
Methods
Women with PPD were recruited from a psychiatric Day
Hospital serving pregnant and postpartum women and a peri-
natal outpatient clinic in Providence, Rhode Island. Eligible
women were offered 3 treatments over 12 weeks: sertraline
alone, IPT alone, or combined sertraline=IPT. Enrolled subjects
participated from February 2003 through January 2005. All study
participants provided signed informed consent and the study was
approved by the institutional research board.
Inclusion criteria required a diagnosis of major depressive dis-
order made by clinical interview, a Beck Depression Inventory
(BDI) (Beck & Beamesderfer, 1974) score
25 or a 17-item
Hamilton Rating Scale for Depression (HRSD) (Hamilton,
1960) score
14, age between 18 and 50 years of age, delivery
of a healthy infant within the prior 6 months, and use of birth
control. Exclusion criteria included a current primary diagnosis
of an anxiety disorder, a current or past diagnosis of bipolar
disorder or psychotic disorder, a diagnosis of a substance abuse
disorder, anorexia nervosa or bulimia nervosa in the previous
year, acute suicidal or homicidal risk, current treatment with an
adequate dose of an antidepressant medication or other psycho-
tropic medication, current psychotherapy, or significant medical
problems suggesting a contraindication to sertraline. Sporadic
(not daily) use of zolpidem, zaleplon, low-dose trazodone, low
dose benzodiazepine or benadryl were allowed for insomnia.
Prospective subjects were recruited by one of two psycho-
therapists trained in IPT, one of whom was a nurse clinician in
the Day Hospital and the other who was a therapist in the peri-
natal outpatient clinic. Potential subjects were informed about
the negative consequences of untreated PPD and the potential
benefits of treatment. Women were specifically informed about
the risks and possible adverse effects from taking sertraline.
Breastfeeding mothers were informed about the current research
knowledge about short-term and long-term risks and adverse
Patient choice of treatment for postpartum depression 305

effects with infant exposure to sertraline. Women were given Table 1. Demographic characteristics of entire sample
information about IPT, its applicability for postpartum issues,
Range Total
and the existence of studies supporting its efficacy for PPD.
sample (n 23)
Women were informed about the possible benefit of combined
psychotherapy and pharmacotherapy treatment compared to Mean SD
either treatment alone for major depressive disorder. Alternative
treatments, such as other psychotherapies and antidepressant Age 1940 28.48 6.01
medications, were also discussed. Treatment choice was selected Number of weeks postpartum 325 8.52 5.64
Parity (live births) 13 1.61 0.66
by the women following the informed consent procedure there
Total number of pregnancies 14 2.00 0.93
was no random assignment of treatment condition. All breast-
feeding women who elected treatment with sertraline (either Race=Ethnicity n %
solely or in combination with IPT) agreed to have the study per- White, non-Hispanic 18 78.3
sonnel contact the infants pediatrician by letter to provide infor- Hispanic 2 8.7
mation that she was receiving sertraline and to request that the African-American 2 8.7
pediatrician inform study personnel if adverse effects were noted Native-American 1 4.3
in the infant. Education
The sertraline component was conducted in accordance with 11th grade or less 1 4.3
the medication model (Fawcett et al, 1987) and was managed 12th grade or GED 7 30.4
by one of the psychiatrists (T.B.P.), with 8 sessions held over Some college 5 21.7
12 weeks. Sertraline was titrated in a flexible-dose regimen up College degree 6 26.1
Graduate degree 4 17.4
to 150 mg daily based on clinical response (by interview) and
tolerability. Sertraline was started at 25 mg daily, could be Marital status
increased to 50 mg daily after 2 weeks, to 100 mg daily after Single 6 26.1
Married 16 69.6
4 weeks, and to 150 mg daily after 8 weeks. IPT was adminis-
Separated or divorced 1 4.3
tered in 12 individual 50-minute sessions by one of two psy-
chotherapists (A.B.P. or M.A.G.) who had received IPT training
by two of the co-authors (S.S. and M.W.O.) and both therapists
received regular IPT supervision (from S.S.) throughout the scores not obtained (n 1). Demographic characteristics
study. Treatment outcome was monitored with the clinician- of study participants are presented in Table 1. Eleven of
rated 17-item HRSD and the self-rated Edinburgh Postnatal
the 23 women selected IPT alone, 2 selected sertraline
Depression Scale (EPDS) (Cox et al, 1987) and BDI obtained
at baseline, 4 weeks, 8 weeks and 12 weeks. alone, and 10 selected combined sertraline=IPT. Both
Statistical tests were conducted to examine study data using of the women selecting sertraline alone cited time con-
SPSS version 12.0. Measures of central tendency were calculat- straints which prohibited participation in weekly psycho-
ed to describe the baseline characteristics of study participants. therapy sessions.
Differences in symptom levels across the 3 treatment groups at The women endorsed moderately severe depressive
baseline were tested using analysis of variance. Fishers exact
symptoms at baseline as determined by both self-report
test was used to compare the clinical characteristics of women
who selected treatment with medication (sertraline alone or (EPDS, BDI) as well as clinician-rated (HRSD) mea-
sertraline=IPT) versus those who opted for treatment without sures of depression (see Table 2). There were no signifi-
medication (IPT alone). T-tests were used to examine whether cant differences in mean baseline HRSD, EPDS, and
baseline symptom differences existed between completers and BDI scores among women in each of the three treatment
non-completers. Final sertraline dose among women who chose
conditions. In addition, no differences were detected
sertraline alone was compared with the final dose among women
in the sertraline=IPT group using t-tests. Finally, paired samples among the three treatment conditions in terms of parity,
t-tests were used to test for pre-post treatment differences in number of weeks postpartum, presence of concurrent
symptom levels, and analyses of covariance were conducted to Axis 1 disorder, and prior use of antidepressant medi-
compare outcomes across treatment groups. cation. At the time of enrollment, approximately half of
the study participants were breastfeeding (n 12).
Results To investigate whether differences existed between
postpartum women who selected a treatment that included
Demographics and pre-treatment group differences
medication and those who selected treatment without
Over the course of enrollment, 26 postpartum women medication, the sertraline alone and sertraline=IPT groups
were screened to determine their eligibility for study were combined, comparing a medication group (n 12)
participation. Of these women, 23 met inclusion criteria with a no-medication group (n 11). There was a trend
and were enrolled. Reasons for exclusion included al- for breastfeeding women to opt for treatment without
ready being on another antidepressant (n 1), baseline medication (66.7%) rather than treatment with medication
depression scores too low (n 1) and baseline depression (33.3%) (Fishers exact test, p 0.10; Fig. 1). In addition,
306 T. B. Pearlstein et al

Table 2. Clinical characteristics by treatment option

Baseline depression ratings IPT (n 11) Sertraline (n 2) Sertraline=IPT Total sample (n 23)
(n 10)

Mean SD Mean SD Mean SD Mean SD

HRSD 17.45 1.86 20.50 0.71 18.80 4.16 18.30 3.10

EPDS 16.73 4.88 15.00 5.66 18.70 3.97 17.44 4.50
BDI 20.09 8.08 20.00 7.07 25.30 5.68 22.35 7.22
Baseline characteristics n % n % n % n %
Breastfeeding 8 72.7 0 0 4 40.0 12 52.2
Co-morbid Axis I disorder 1 9.1 1 50.0 2 20.0 4 17.4
Previous Major Depression 1 9.1 2 100.0 4 40.0 7 30.4
Previous PPD 0 0 0 0 2 20.0 2 8.7
Previous antidepressant use 1 9.1 1 50.0 4 40.0 6 26.1
Adjunctive sleep aid 0 0 1 50.0 4 40.0 5 21.7

Treatment compliance n % n % n % n %
Dropped out of treatment 2 18.2 0 0 3 30.0 5 21.7
Completed treatment 9 81.8 2 100 7 70 18 78.3
Post-treatment depression Mean SD Mean SD Mean SD Mean SD
HRSD 5.56 4.88 10.00 2.83 3.86 3.45 5.39 4.41
EPDS 6.44 3.25 4.50 3.54 5.57 4.28 5.89 3.55
BDI 6.33 6.38 7.00 6.66 6.14 6.31 6.33 5.93

HRSD Hamilton Rating Scale for Depression; EPDS Edinburgh Postnatal Depression Scale; BDI Beck Depression Inventory; IPT interpersonal
psychotherapy; PPD postpartum depression.

Fig. 1. Treatment selected by breastfeeding vs. non-breastfeeding Fig. 2. Treatment selected by women with history of MDD vs. no
PPD patients. Fishers exact test, 2-sided, p 0.10; PPD postpartum history of MDD. Fishers exact test, 2-sided, p 0.07; MDD major
depression depressive disorder

there was a trend for women with previous histories of

depression to choose treatment that included medication
(85.7%) rather than a treatment approach without med-
ication (14.3%) (Fishers exact test, p 0.07; Fig. 2).
Compliance, treatment outcome, and adverse effects
Eighteen of the 23 enrolled women completed the
12-week protocol. Of the 5 women who terminated
early, 3 were noncompliant with protocol sessions, one
switched to couples counseling, and one withdrew after
8 IPT sessions reporting that treatment was no longer
necessary. Those remaining included 9 women in the
IPT condition, 2 in the sertraline condition, and 7 receiv-
ing the sertraline=IPT combination. Non-completers had
significantly higher mean BDI scores at baseline than
completers (29.40  3.30 vs. 20.34  1.41, t(21) 2.84,
p 0.01). They were not significantly different on mean
baseline HRSD or EPDS scores.
Patient choice of treatment for postpartum depression
Paired t-tests examining differences between pre- and
post-treatment depression scores on the HRSD, EPDS,
and BDI indicated that, on average, completers experi-
enced significant clinical improvement, irrespective of
treatment condition. In the overall sample, HRSD scores
decreased on average by 12.56  5.16 points, represent-
ing a 70% mean reduction in symptoms (t(17) 10.33,
p < 0.001); EPDS scores decreased on average by
10.94  5.86 points, representing a 65% mean reduction
(t(17) 7.97, p < 0.001); and BDI scores were decreased
by 14.05  9.10 points, representing a 69% mean reduc-
tion (t(17) 6.56, p < 0.001). Analyses of covariance
were conducted to compare treatment outcomes (HRSD,
EPDS, BDI) across treatment conditions, controlling for
pre-treatment depression severity. Because the sertraline
alone group was very small, consisting of only 2 parti-
cipants, this group was omitted from the analysis. No
significant differences in outcome were identified be-
tween the IPT alone group and the sertraline=IPT group.
Thus, in this small sample, there was no evidence for
differential efficacy of treatments in terms of depression
symptom reduction.
The two women who chose sertraline alone received
the full dose of 150 mg daily. The 10 women who se-
lected combined sertraline=IPT reached an average daily
dose that was significantly lower (95.00  55.03 mg,
t(9) 3.16, p 0.012). Approximately half of the women
in each of the two sertraline treatment options used
adjunctive as needed sleep aids during some por- traline, the administration of the treatments by research
tion of the 12 week trial. Side effects were generally
mild, and no participants discontinued sertraline due to
side effects. Side effects reported by the 12 women who
received sertraline included soft stools (n 4), sedation
(n 3), gastrointestinal distress (n 3), delayed orgasm
(n 3), nausea (n 2), headache (n 1), frequent stools
(n 1), decreased appetite (n 1) and dry mouth (n 1).
Breastfeeding women who were receiving sertraline
were asked during each medication management ses-
sion about the presence of adverse effects in their
infants; no adverse effects were reported. In addition,
no communication was received by any of the breast-
feeding infants pediatricians regarding concerns about
adverse effects.
Discussion
Most women in the study, when informed of the advan-
tages and concerns of both IPT and sertraline, selected
IPT with or without the addition of sertraline. There was
a trend for women with a history of depression to choose
sertraline as a treatment alone or with IPT. The trend for
307
breastfeeding women to be less likely to choose sertra-
line than psychotherapy is similar to the finding that
breastfeeding women were less likely to take a psycho-
tropic medication than non-breastfeeding women in a
chart review of a larger sample examined from the same
Day Hospital population (Battle et al, 2006). It is possi-
ble that a previous depression or previous antidepressant
treatment may be associated with pharmacotherapy use,
while breastfeeding may be associated with psycho-
therapy as a preferred treatment option. While the small
sample size must be noted, the suggested associations may
reflect the types of treatment choices made by women
with PPD.
The significant reductions in HRSD, EPDS and BDI
scores at the end of treatment suggest that all 3 treat-
ments (IPT, sertraline, IPT=sertraline) may be effica-
cious for completers, but the sample size was too small
to reliably distinguish differential efficacy among these
treatments. The benefit of IPT in this study, as noted by
the reduction of HRSD and BDI scores, was similar in
magnitude to the response noted in the large published
trial of IPT (OHara et al, 2000). The benefit of sertra-
line in this study was similar to the reductions noted on
HRSD and EPDS scores with paroxetine in a recent trial
(Misri et al, 2004) and a previous open sertraline trial
(Stowe et al, 1995).
Strengths of the study include relative uniformity in
the presentation of the risks and benefits of IPT and ser-
qualified clinicians, and systematic evaluation of clinical
symptoms at baseline and through treatment. Limita-
tions include limited generalizability given the recruit-
ment of women from specialized perinatal settings and
a largely educated, married and Caucasian sample. In
addition, the small sample size, lack of randomization,
and lack of a placebo condition precludes any conclu-
sion about the efficacy of the treatment options.
It is important to successfully and expeditiously treat
PPD in light of the deleterious effects on the mother,
child and other family members. Future studies need to
firmly establish the efficacy of antidepressant medication
for PPD under randomized, placebo-controlled condi-
tions. Psychotherapy, antidepressant medication and their
combination also need to be evaluated in randomized
comparator trials. Breastfeeding women should be in-
cluded, while remaining cognizant of concerns about in-
fant exposure to antidepressant medication and the lack
of definitive data about long-term effects on children.
If future studies establish differential efficacy of treat-
ments, or delineate clinical predictors of response to
308 T. B. Pearlstein et al: Patient choice of treatment for postpartum depression

differential treatments, women with PPD and their clin- Grace SL, Evindar A, Stewart DE (2003) The effect of postpartum
depression on child cognitive development and behavior: a review
icians can make more informed treatment choices. and critical analysis of the literature. Arch Womens Mens Health 6:
263274.
Hamilton M (1960) A rating scale for depression. J Neurol Neurosurg
Acknowledgment Psychiatry 23: 5662.
Misri S, Kostaras X (2002) Benefits and risks to mother and infant of
The authors thank Elizabeth Balkcom, R.N. for her assistance in drug treatment of postnatal depression. Drug Saf 25: 903911.
data collection and maintenance. Misri S, Reebye P, Corral M, Milis L (2004) The use of paroxetine and
cognitive-behavioral therapy in postpartum depression and anxiety: a
randomized controlled trial. J Clin Psychiatry 65: 12361241.
References Newport DJ, Wilcox MM, Stowe ZN (2001) Antidepressants during
pregnancy and lactation: defining exposure and treatment issues.
Abreu AC, Stuart S (2005) Pharmacologic and hormonal treatments for
Semin Perinatol 25: 177190.
postpartum depression. Psychiatr Ann 35: 568576.
OHara MW, Stuart S, Gorman LL, Wenzel A (2000) Efficacy of
Appleby L, Warner R, Whitton A, Faragher B (1997) A controlled study
interpersonal psychotherapy for postpartum depression. Arch Gen
of fluoxetine and cognitive-behavioural counselling in the treatment
Psychiatry 57: 10391045.
of postnatal depression. BMJ 314: 932936.
Segre LS, Stuart S, OHara MW (2004) Interpersonal psychotherapy for
Battle CL, Zlotnick C, Miller IW, Pearlstein T, Howard M (2006)
antenatal and postpartum depression. Primary Psychiatry 11: 5256,
Clinical characteristics of perinatal psychiatric patients: a chart review
66.
study. J Nerv Ment Dis 194: 369377.
Sit DK, Wisner KL (2005) Decision making for postpartum depression
Beck AT, Beamesderfer A (1974) Assessment of depression: the depres-
treatment. Psychiatr Ann 35: 577585.
sion inventory. Mod Probl Pharmacopsychiatry 7: 151169.
Stowe ZN, Casarella J, Landry J, Nemeroff CB (1995) Sertraline in the
Boath E, Bradley E, Henshaw C (2004) Womens views of antidepres-
treatment of women with postpartum major depression. Depression 3:
sants in the treatment of postnatal depression. J Psychosom Obstet
4955.
Gynaecol 25: 221233.
Stuart S, OHara MW, Gorman LL (2003) The prevention and psy-
Burt VK, Suri R, Altshuler L, Stowe Z, Hendrick VC, Muntean E (2001)
chotherapeutic treatment of postpartum depression. Arch Womens
The use of psychotropic medications during breast-feeding. Am J
Ment Health 6: S57S69.
Psychiatry 158: 10001009.
van Schaik DJ, Klijn AF, van Hout HP, van Marwijk HW, Beekman AT,
Chabrol H, Teissedre F, Armitage J, Danel M, Walburg V (2004)
de Haan M, van Dyck R (2004) Patients preferences in the treatment
Acceptability of psychotherapy and antidepressants for postnatal
of depressive disorder in primary care. Gen Hosp Psychiatry 26:
depression among newly delivered mothers. J Reprod Infant Psychol
184189.
22: 512.
Weissman AM, Levy BT, Hartz AJ, Bentler S, Donohue M, Ellingrod
Cox JL, Holden JM, Sagovsky R (1987) Detection of postnatal depres-
VL, Wisner KL (2004) Pooled analysis of antidepressant levels in
sion. Development of the 10-item Edinburgh Postnatal Depression
lactating mothers, breast milk, and nursing infants. Am J Psychiatry
Scale. Br J Psychiatry 150: 782786.
161: 10661078.
Dennis CL (2004) Treatment of postpartum depression, part 2: a crit-
Whitton A, Warner R, Appleby L (1996) The pathway to care in post-
ical review of nonbiological interventions. J Clin Psychiatry 65:
natal depression: womens attitudes to post-natal depression and its
12521265.
treatment. Br J Gen Pract 46: 427428.
Dennis CL, Stewart DE (2004) Treatment of postpartum depression,
part 1: a critical review of biological interventions. J Clin Psychiatry
65: 12421251.
Fawcett J, Epstein P, Fiester SJ, Elkin I, Autry JH (1987) Clinical Correspondence: Teri B. Pearlstein, MD, Director, Womens
management imipramine=placebo administration manual. NIMH Behavioral Health, Women and Infants Hospital, 101 Dudley
Treatment of Depression Collaborative Research Program. Psycho- Street, Providence, RI 02905, U.S.A.; e-mail: Teri_Pearlstein@
pharmacol Bull 23: 309324. brown.edu