Systematic Reviews Workshop presented at Trillium 2003

1
Systematic Reviews Workshoppresented at Trillium
2003

• Postpartum Research Group
• Department of Family Medicine
• McMaster University
• Hamilton, Ontario, Canada

2
The Evidence-Based Postpartum Research Group

• Cheryl Levitt, Liz Shaw, Janusz Kaczorowski,
• Sharon Wong, Russ Springate, Murray Enkin,
• John Sellors

3
Systematic Reviews

• What are they?
• Why are they important?
• What are their benefits?
• How do you do them?

4
What are they?

• a summary of the medical literature that uses
  explicit methods to
  
• systematically search
• critically appraise, and
• synthesize
• the world literature on a specific issue
• Sackett et al, 2000

5
Levels of evidence
6
Why are they important?

• Where is the knowledge we have lost in
  information?
  
•  
• -T.S. Elliot, The Rock
•  

7
Why are they important?

• health care providers, researchers, and policy
  makers are inundated with unmanageable amounts of
  information
  
• e.g., 5,000 journals (8,000 citations) are
  entered weekly in MEDLINE

8
Why are they important?

• efficiently integrate existing information
• provide data for rational decision making
• separate the insignificant, unsound, redundant
  studies from the salient and critical studies
  
• establish whether scientific findings are
  consistent and generalizable

9
What are their benefits?

• minimizes biases (e.g., limiting to only RCTs)
• minimizes random error (e.g, by accumulating data
  from very large numbers of individuals from
  multiple smaller studies)
  
• improves reliability and increases accuracy of
  conclusions
  

10
What are their benefits?

• To establish generalizability of studies
  (multiple studies considered together provide
  interpretive context for appraising
  generalizability across populations, settings,
  treatments)
• To establish consistency of interventions and
  their effects

11
What are their benefits?

• some systematic reviews use statistical methods
  for combining results of individuals studies
  
• called meta-analyses
• can increase power and precision of estimates of
  treatment effects
  

12
What are their benefits?

• Can be used to formulate guidelines and
  legislation on use of various treatments (or
  diagnostic tests, etc.)

13
How do you do them?
14
Step 1 Write a clear, primary research question

• Include
• Intervention
• Population
• Outcomes
• Some time reference (intervention
  timing/duration, follow-up)
  
• (a clear question should be searchable!)

15
Step 2 Develop a priori inclusion/exclusion cr
iteria

• Consider
• Type of intervention
• Population
• Time reference(s)
• Types of study (e.g., RCTs)
• Language, country restrictions
• (be explicit to allow reproducible, duplicate
  assessment!)

16
Step 3 Develop Literature Search Strategy

• Data sources
• MEDLINE, CINAHL, PsycLit, PsychInfo,
  EMBASE/Excerpta Medica, AMED, Cochrane Library,
  HeatlhSTAR, etc., etc
  
• MEDLINE
• OVID (institutional access) vs PubMED (widely
  accessible)
  

17
Other Searching Considerations

• PubMeds Clinical Queries
• PubMeds Related Articles
• Evidence-based products (Best Evidence, Clinical
  Evidence, ACPJC online)
  
• Contacting authors, experts, drug manufacturers
• Grey literature (published on www)
• Secondary references (hand searches)

18
Step 4 Examine Initial Search Results

• How relevant are your search results?
• Do you need to further search the literature
  base?
  
• If your research question is broad, you may wish
  to re-search using more focused search terms

19
Step 5 Assess potential studies against a prio
ri
selection criteria

• duplicate assessment (reduces assessor biases)
• e.g., we used duplicate assessment by clinician
  and methodologist pairs, followed by consensus
  process
  
• review abstracts where available
• retrieve full papers where necessary

20
Step 6 Data Abstraction

• Assess methodologic quality of studies
• Appraise in duplicate, then consensus
• Appropriateness of randomization
• Blinding
• Allocation concealment
• Follow-up rates
• Description of drop-outs/withdrawals
• Intention-to-treat analysis
• A priori power calculations
• Outcome collection (objective vs subjective)

21
Step 7 Data synthesis/summary

• If heterogeneity of studies exist (varying
  patient populations, different interventions,
  diverse outcomes, etc), statistical pooling of
  data (meta-analysis) may not be possible
• Therefore, summarize data qualitatively in
  systematic fashion

22
Step 7 Data synthesis/summary

• How clinically relevant are outcomes?
• How valid are individual study results?
• Are results consistent from study to study?
• In individual studies, what is the magnitude of
  the treatment effects, and, how precise are these
  effects?

23
Step 7 Data synthesis/summary

• How different were patient populations from study
  to study?
  
• Are treatments feasible in your setting?
  (consider your reference population, e.g.,
  primary care)
  
• What are the potential benefits and harms of the
  treatments?

24
Conclusions

• Because of the vast amount of available
  information, systematic reviews are invaluable.
  
• Systematic reviews efficiently integrate existing
  information to enable rational decision making

25
Conclusions

• Systematic reviews ascertain the consistency and
  validity of findings, and whether findings can be
  generalized across populations, settings, and
  treatment variations.

26
Conclusions

• Systematic reviews use explicit methods to limit
  bias and random error, improving the reliability
  and accuracy of conclusions.

27
Conclusions

• Because of the many advantages of systematic
  reviews, they are often used to formulate
  evidence-based guidelines and legislation around
  treatment practices.

28
Assessing potential studies by selection
criteria (Step 5)

• Mock-up exercise
• Abstracts review

29
Selection Criteria Postpartum Example
30
Inclusion criteria 1. Intervention pertaining
directly to therapy or prevention in postnatal
women. 2.      Intervention initiated within the
first year of birth, beginning after the birth
process (following the third stage of birth).
3.      Measured at least one outcome in
postnatal women. 4.  Fully-published in the Engl
ish language. 5.      Design must be a randomize
d trial. 6.      Conducted in Canada, the United
States, Europe, Australia, or New Zealand.
31
Exclusion criteria 1.Interventions for lacta
tion suppression, endometritis, or
hypertensive disorders (e.g., the HELLP
syndrome). 2.Intrapartum interventions (Inter
ventions conducted between the time of labour
and birth) 3.Prenatal interventions that m
ight impact on postpartum outcomes.
32
1 J Dairy Res 2003 Feb70(1)9-17
  Effects of dietary supplements of zinc-methioni
ne on milk production, udder health and zinc meta
bolism in dairy goats.   Salama AA, Caja G, Alba
nell E, Such X, Casals R, Plaixats J.
  Unitat de Produccio Animal, Departament de Cien
cia Animal dels Aliments, Universitat Autonoma de
Barcelona, 08193 Bellaterra, Spain.
  Twenty-two Murciano-Granadina dairy goats were
used to investigate the effects
of organic Zn supplementation of a diet
containing a high level of inorganic Zn.
Goats were kept in pens, machine milked once a
day throughout lactation and fed
a diet based on a dehydrated mixture of
whole-plant maize and alfalfa ad
libitum, alfalfa pellets, barley grain and a
concentrate mixture. Treatments
were (1) control, and (2) supplemented with 1
g/d Zn-Methionine (Zn-Met) included in the concen
trate mixture. After parturition, goats were
blocked in week 3 and dietary treatments were app
lied until week 23. From weeks 3-20, feed
intake, milk yield, milk composition, milk
somatic cell count (SCC), and udder
health were measured. In week 21, all goats were
injected intraperitoneally with
1 g/d DL-methionine for 5 d to establish the
effects of methionine under the
conditions of udder stress induced by hand
milking on the second day. During
weeks 22 and 23, diet digestibility, and N and Zn
balance were determined. Dry matter intake, milk
yield, and milk contents of total solids, fat,
total and true protein, and casein did not differ
between treatments, but whey protein and
non-protein nitrogen contents were significantly
lower for the Zn-Met group. Milk SCC tended to de
crease as a result of Zn-Met supplementation but
differences between treatments were not
significant when halves with persistent
infection were excluded. Hand milking increased
SCC in both groups, but udders
of supplemented goats showed a lower reaction.
Apparent absorption of N significantly increased
and Zn retention tended to increase in Zn-Met
supplemented goats. We conclude that Zn-Met
supplementation can enhance resistance to udder s
tress in dairy goats. Effects were attributed to
the organic Zn and not to the methionine componen
t. Zn retention and protein utilization were also
improved by the Zn-Met supplement.  
33
2 J Am Diet Assoc 1998 Feb98(2)143-8
  Counseling and motivational videotapes increase
duration of breast-feeding in
African-American WIC participants who initiate
breast-feeding.   Gross SM, Caulfield LE, Bentle
y ME, Bronner Y, Kessler L, Jensen J, Paige VM.
Department of Maternal and Child Health of The
Johns Hopkins University School
of Hygiene and Public Health, 615 N Wolfe St,
Room 2501, Baltimore, MD 21205,
USA.  OBJECTIVE To evaluate the relative effects
introducing motivational videotapes
and/or peer counseling in Special Supplemental
Nutrition Program for Women, Infants, and Childre
n (WIC) clinics serving African-American women
have on breast-feeding duration. DESIGN Experime
ntal intervention study. Pregnant women
were enrolled at or before 24 weeks gestation and
were followed up until postpartum week 16. Women
were interviewed at enrollment, 7 to 10 days, 8
weeks, and 16 weeks postpartum. SUNJECTS/SETTING
One hundred fifteen African-American
WIC participants who initiated breast-feeding and
who had been enrolled in 1 of 4 clinics. INTERVEN
TION Two-by-two factorial design, in which 4
clinics were randomly assigned to receive either
no intervention, a motivational video
package intervention, a peer-counseling
intervention, or both interventions.
MAIN OUTCOME MEASURES Breast-feeding duration in
weeks and relative risk ratios
for breast-feeding cessation before 16 weeks
postpartum. STATISTICAL ANALYSIS
PERFORMED Contingency table analysis, including
chi2 tests and log-rank tests
multivariate analysis using Cox proportional
hazards regression analysis. RESULTS A higher pr
oportion of women were breast-feeding at 8 and 16
weeks postpartum in the intervention clinics than
in the control clinic. The proportion of women r
eporting breast-feeding declined at 8 and 16
weeks postpartum, but the rate of decline was slo
wer in the 3 intervention clinics
than in the control clinic. Being younger than 19
years of age or older than 25 years of age, havin
g a male infant, and returning to work or school
all negatively affected breastfeeding duration, w
hereas previous breast-feeding
experience positively influenced breast-feeding
duration. APPLICATIONS/CONCLUSIONS WIC-based pee
r counselor support and motivational
videos can positively affect the duration of
breast-feeding among African-American women. WIC
nutritionists and other health professionals in
contact with this population should expand their
efforts toward promoting increased duration of br
east-feeding.
34
3 BJOG 2002 Oct109(10)1164-70
  A two-centred pragmatic randomised controlled t
rial of two interventions of postnatal support.
  Reid M, Glazener C, Murray GD, Taylor GS.   D
epartment of Public Health, Glasgow University,
UK.   OBJECTIVES To establish whether providing
additional postnatal support during
the early postnatal months influences women's
physical and psychological health
and to identify health service benefits. DESIGN
Pragmatic randomised controlled
trial with a 2 x 2 factorial design with two
interventions. SETTING Community
centres, Ayrshire and Grampian, Scotland.
POPULATION One thousand and four
primiparous women, 83 completed the baseline
questionnaire, 71 at six months.
METHODS (1) An invitation to a local postnatal
support group run weekly with a
facilitator, starting two weeks postpartum. (2) A
postnatal support manual, posted two weeks postpa
rtum. MAIN OUTCOME MEASURES Data regarding
primary outcome postnatal depression (Edinburgh P
ostnatal Depression Scale, EPDS),
secondary outcomes, general health measures
(SF-36), social support (SSQ6), use
of health services and women's views of
interventions were collected at two
weeks postpartum and at three and six months.
RESULTS There were no significant
differences in EPDS scores between the control
and trial arms at three and six
months, nor were there differences in the SF-36
and the SSQ6 scores. The 95 CI
for the difference in EPDS effectively excluded a
change in mean score of more than 10 with either
intervention. There were no differences in
health service attendances in primary or secondar
y care between the control and trial arms. Of
those women who attended the groups, 40 attended
six or more. Women reported favourably on the 'pa
ck' with the majority reading it a few times and
feeling that it was aimed at them. CONCLUSIONS W
ide-scale provision by the National
Health Service of either support groups or
self-help manuals is not appropriate
if the aim is to improve measurable health
outcomes.  
35
4 Am J Psychiatry 2002 Jan159(1)43-7
  Comment in Am J Psychiatry. 2002 Aug159
(8)1437-8 discussion 1438.   Gender, poverty,
and postnatal depression a study of mothers in
Goa, India.   Patel V, Rodrigues M, DeSouza N.
  Sangath Centre for Child Development Family G
uidance, Goa, India. vikpat_at_goatelecom.com   OB
JECTIVE This study described the natural history
of depression in mothers who recently gave birth
in a low-income country and to investigate the
effect of risk factors, particularly related to i
nfant gender bias, on the occurrence and
outcome of depression. METHOD The authors
studied a group of pregnant mothers
recruited during their third trimester of
pregnancy from a district hospital in
Goa, India. The mothers were interviewed at
recruitment, 6-8 weeks, and 6 months
after childbirth. Interview data included
presence of antenatal and postnatal
depression, obstetric history, economic and
demographic characteristics, and
gender-based variables (preference for male
infant, presence of marital violence). RESULTS D
epressive disorder was detected in 59 (23) of
the mothers at 6-8 weeks after childbirth 78 of
these patients had had clinically
substantial psychological morbidity during the
antenatal period. More than one-half of the patie
nts remained ill at 6 months after delivery.
Economic deprivation and poor marital relationshi
ps were important risk factors for the
occurrence and chronicity of depression. The
gender of the infant was a determinant of postnat
al depression it modified the effect of other
risk factors, such as marital violence and hunger
. Depressed mothers were more disabled and were m
ore likely to use health services than
nondepressed mothers. CONCLUSIONS Maternal and i
nfant health policies, a priority in low-income
countries, must integrate maternal depression as
a disorder of public health significance. Interve
ntions should target mothers in the antenatal
period and incorporate a strong gender-based comp
onent.  
36
5 Lancet 1983 Oct 292(8357)990-2
  Post-partum rubella immunisation a controlled
trial of two vaccines.   Black NA, Parsons A, Ku
rtz JB, McWhinney N, Lacey A, Mayon-White RT.
  The effectiveness of two rubella vaccines (RA 2
7/3 and Cendehill) in women vaccinated post partu
m was compared. Significantly more women who
received RA 27/3 gave satisfactory seroconversion
responses than women who received
Cendehill (97.6 vs 82.2). This difference was
reflected in the geometric mean
titres (43.6 vs 17.0). More women who received RA
27/3 had minor side-effects, but the difference w
as not significant. The serological response to
both vaccines was not affected by the concurrent
administration of anti-D immunoglobulin. In view
of these findings, the replacement of Cendehill
vaccine with RA 27/3 vaccine for women vaccinated
post partum should be considered.
37
6 Br J Obstet Gynaecol 1997 Mar104(3)340-6
  Randomised trial comparing a policy of early wi
th selective amniotomy in uncomplicated labour at
term.   Johnson N, Lilford R, Guthrie K, Thornt
on J, Barker M, Kelly M.   Department of Obstetr
ics, St James's University Hospital, Leeds
University, UK.   OBJECTIVE To compare two mana
gement policies rupture of the fetal membranes
when women are in normal labour or leave them
intact as long as feasible. SETTING The labour w
ard of a city university hospital. DESIGN
Automated randomised clinical trial. PARTICIPANTS
1540 women in uncomplicated term
labour. Data on labour duration, blood loss,
oxytocin use and fetal condition
were collected from 1132 women. Some data from
nulliparous women has been presented earlier by t
he UK Amniotomy Group. MAIN OUTCOME MEASURES
Duration of labour, Apgar score, fetal morbidity
and maternal morbidity including perineal
injury, mode of delivery, epidural rates and the
total number of vaginal examinations in the first
stage of labour after amniotomy. INTERVENTIONS
Amniotomy at the next vaginal examination or
amniotomy only if indicated. The
median cervical dilatation at membrane rupture
was 2 cm greater in the first group. RESULTS A p
olicy of routine amniotomy in labour had no
measurable advantage over selective amniotomy for
parous women (difference 4 min) but
shortened labour in nulliparous women by 1 h
(Mann-Whitney U test P There was a suggestion of a higher caesarean
section rate (OR 1.9 95 CI 0.9-3.5), and there
were more vaginal examinations after membrane
rupture in the group allocated routine amniotomy.
There were no measurable differences in
oxytocin use, fetal condition at birth, retained
placenta rates, blood loss, pain or analgesia req
uirements. CONCLUSION Routine amniotomy may
shorten the first labour but not subsequent ones.
There is a suggestion that routine
surgical interference may be harmful by
increasing the risk of caesarean
section, and this agrees with data from other
trials (common odds ratio 1.2 95
CI 0.92-1.6).
38
7 Lancet 1996 Aug 10348(9024)364-9
  Randomised controlled trial of a reduced-visits
programme of antenatal care in
Harare, Zimbabwe.  Munjanja SP, Lindmark G, Nystr
om L. Department of Obstetrics and Gynaecology,
University of Zimbabwe, Avondale,Harare.
  BACKGROUND Many of the individual components o
f antenatal care have been studied in randomised
controlled trials, but few studies have compared
whole programmes of antenatal care. Our aim was t
o test the hypothesis that a new
programme of antenatal care with fewer
goal-oriented visits would give an
equivalent or better result in the outcomes
associated with pregnancy and delivery. METHODS
In a randomised clinical trial in Harare,
Zimbabwe, we compared a new programme of antenata
l care with the standard programme. The new
programme consisted of fewer but more objectively
oriented visits and fewer procedures per visit. S
even primary care clinics were randomly assigned
to the two programmes-three to the standard progr
amme and four to the new programme.
FINDINGS Over a 2-year period, 15,994 women were
recruited into the study at the time they booked
antenatal care. 97 of the women were followed
up, 9,394 who had followed the new programme, and
6,138 from clinics with the standard
one. Women allocated to the new programme made,
as planned, fewer visits than those in the standa
rd programme (median 4 vs 6 visits,
respectively). The proportion of antenatal referr
als was also lower (13.6 vs 15.3 odds ratio
0.87 95 CI 0.79-0.95) because of significantly
fewer referrals for pregnancy-induced hypertensi
on (2.5 vs 3.8 0.66 0.55-0.79).
Nevertheless, there were significantly fewer labo
ur referrals for severe hypertension or
eclampsia (2.1 vs 2.6 0.81 0.66-1.00). The
risk for preterm (icantly lower for women on the new programme
(10.1 vs 11.5 0.86 0.78-0.96). There were no
other significant differences between the
programmes in other major indices of pregnancy
outcome, including antenatal referrals for other
causes, labour referrals, obstetric
interventions, low birthweight, and perinatal and
maternal mortality and morbidity.
INTERPRETATION An antenatal care programme with
fewer more objectively oriented visits can be
introduced without adverse effects on the main
intermediate outcome pregnancy variables.
39
8 Indian Pediatr 1996 Apr33(4)279-86
  Effect of vitamin A supplementation to mother a
nd infant on morbidity in infancy.   Venkatarao
T, Ramakrishnan R, Nair NG, Radhakrishnan S,
Sundaramoorthy L, Koya PK, Kumar SK.   Institut
e for Research in Medical Statistics, Madras.
  OBJECTIVES To assess the impact of Vitamin A s
upplementation to the mother soon
after delivery and to the infant at six months on
morbidity in infancy. DESIGN Randomized double b
lind placebo controlled field trial. SETTING 51
villages in two contiguous Primary Health Centers
in Villupuram Health Unit District of
Tamil Nadu, South India. SUBJECTS 909 newly
delivered mother-and-infant pairs.
INTERVENTIONS Both mother and infant received
Vitamin A (300,000 IU for mothers
and 200,000 IU for children) in 311 instances
(AA) mother received Vitamin A
but infant received Placebo in 301 instances
(AP) and both mother and infant
received Placebo in the remaining 297 instances
(PP). MAIN OUTCOME MEASURES Incidence of diarrhe
a and Acute Respiratory Infection (ARI)
distributions of infants by frequency of episodes
and number of infected days. RESULTS 233 in
the AA Group and 228 each in the AP and PP Groups
were followed up regularly. The incidence of diar
rhea in these infants was 97.4, 96.9 and 94.7
in the three groups, mean number of diarrheal epi
sodes was 4.4, 4.6 and 4.2 and median
number of days in infancy with diarrhea was 26,
26 and 22 days, respectively. For ARI, the incide
nces were 96.6, 95.6 and 96.1, means were 4.8,
5.1 and 4.8 episodes, and the medians were 32, 34
and 34 days, respectively. CONCLUSIONS
Prophylactic administration of mega doses of
Vitamin A to the mother soon after
delivery and to the infant at six months do not
have any beneficial impact on the incidence of di
arrhea and ARI in infancy.
40
9 Int J Group Psychother 1993 Apr43(2)191-203
  A group psychotherapy approach to postpartum
depression.   Gruen DS.   In this article, the
current knowledge of postpartum depression is
reviewed, leading to an understanding of this for
m of depression as an expression of
psychosocial factors compounded by the biological
changes of the puerperium. A rationale for treati
ng women in the context of the postpartum period
is described. A novel approach to treatment is pr
esented. The treatment uses the
medium of group psychotherapy to administer
interventions. The interventions are
introduced in three phases across six months in
weekly group sessions. The systematic interventio
n program focuses on decreasing four symptoms
common to postpartum depression depressed feelin
gs, anxiety, distress, and low
self-esteem. Spouses are involved in the
treatment plan from the onset,
participating in assessment sessions and in a
couples group.  
41
10 Eur J Obstet Gynecol Reprod Biol 1993
Jan48(1)19-22   Management of the third stage
of labour in women at low risk of postpartum
haemorrhage.   Thilaganathan B, Cutner A, Latime
r J, Beard R.   Department of Obstetrics and Gyn
aecology, Royal Sussex County Hospital,
Brighton, UK.   OBJECTIVE To compare active man
agement with physiological management of the
third stage of labour in women at low risk of
postpartum haemorrhage. DESIGN
Randomised allocation of women in labour at low
risk of postpartum haemorrhage
to either physiological or active management of
the third stage. SETTING Labour
ward in a district general hospital. PATIENTS
193 Women with spontaneous vaginal delivery at te
rm completed the study. Exclusion criteria were
induction or augmentation of labour, antepartum o
r previous postpartum haemorrhage,
premature rupture of membranes, previous
caesarean section, raised blood
pressure, cervical lacerations and third degree
tears. INTERVENTIONS Active management with synt
ometrine and controlled cord traction or
physiological management, where the cord was not
clamped and the placenta was delivered by
maternal effort. MEASUREMENTS Blood loss was
measured subjectively at delivery
and estimated objectively by comparing the
haemoglobin in labour with that on
the third postpartum day. The duration of the
third stage was also measured as
was the incidence of retained placenta and blood
transfusion. RESULTS There was
no significant difference in the estimated blood
loss or haemoglobin drop between the two groups (
P 0.5). In addition the duration of the third
stage was significantly longer in the physiologic
al group (P having physiological management, 7 received
oxytocics for presumed postpartum
haemorrhage. Only one case in the active group
required further oxytocics and
one other case in this group required a manual
removal of placenta. CONCLUSIONS
This preliminary study confirms that active
management results in a reduction in
the length of the third stage of labour. However,
it suggests that active management does not reduc
e blood loss when compared to physiological
management in the woman at low risk of postpartum
haemorrhage.  
42
11 Lancet 1997 May 3349(9061)1277-81
  Randomised double-blind placebo-controlled stud
y on adverse effects of rubella
immunisation in seronegative women. 
Tingle AJ, Mitchell LA, Grace M, Middleton P,
Mathias R, MacWilliam L, Chalmers
A. Department of Paediatrics, University of
British Columbia, Vancouver, Canada.
  BACKGROUND The objective of our study was to i
nvestigate the association of adverse clinical mu
sculoskeletal and neurological events in healthy
postpartum women with live attenuated (RA27/3 str
ain) rubella-virus vaccine, and to assess
the frequency of acute and recurrent arthralgia
and arthritis and associations
with acute and recurrent muscle pain (myalgia)
and neurological manifestations
(paraesthesias). METHODS We used a randomised
placebo-controlled, double-blind
design in a community setting. 636 women were
enrolled and, after 90 women dropped out, 546 hea
lthy women aged 18-41 years, who were rubella
seronegative on routine screening were immunised
parenterally with either monovalent live
attenuated (RA27/3 strain) rubella vaccine (n
270) or saline placebo (n 276)
in the postpartum period. Outcome measures were
the occurrence of acute and persistent or recurre
nt joint manifestations (arthralgia or arthritis)
at 1, 3, 6, 9, and 12 months after immunisation.
Occurrence of muscle pain (myalgia), and
neurological symptoms (paraesthesia) was also
assessed at the same times. FINDINGS 543 women c
ompleted 1-month follow-up. 456 women completed
the 12-month assessment. There were no difference
s at the time of immunisation between rubella vac
cine and placebo groups in distribution of age,
ethnic origin, parity, time between delivery and
immunisation, breastfeeding history,
or histories of earlier rubella vaccination or
joint complaints. Results indicated a significant
ly higher incidence (p 0.006 odds ratio 1.73
95 CI 1.17-2.57) of acute joint manifestatio
ns in rubella-vaccine recipients (30)
than in placebo recipients (20). Frequency of
chronic (recurrent) arthralgia or
arthritis was only marginally significant (p
0.042 1.58 1.01-2.45). INTERPRETATION RA27/3
rubella vaccine given to seronegative women
during the postpartum period was significantly as
sociated with development of acute
arthralgia or arthritis. Although the numbers of
women assessed and length of follow-up revealed o
nly marginally significant differences in
persistent or recurrent joint manifestations betw
een rubella vaccine and placebo recipients,
it is possible that susceptible women who are
given rubella vaccination may experience this out
come.
43
12 Obstet Gynecol 1994 May83(5 Pt 1)761-5
  Timing of the postpartum Papanicolaou smear.  
Rarick TL, Tchabo JG.   Georgetown University
Medical Center, Washington, DC.
  OBJECTIVE To study the relation between the ti
ming of the postpartum examination and the postpa
rtum Papanicolaou smear. METHODS One hundred
eighty-four women in labor at Arlington Community
Hospital were randomized to receive their postpar
tum Papanicolaou smear at 4, 6, or 8 weeks after
delivery. Samples from the exocervix and endocerv
ix were obtained and submitted to the
pathology department, which was unaware of the
study. RESULTS Twenty-three women were lost to f
ollow-up. Among 161 women who had a postpartum
examination, 48 (30) had it at 4 weeks, 61 (38)
at 6 weeks, and 52 (32) at 8 weeks. The
groups were similar with respect to age,
ethnicity, gravidity, parity, marital
status, number of sexual partners, age at first
intercourse, history of cervical
infections, previous history of abnormal
Papanicolaou smears, and cervical
surgery. Among the 139 women who had a normal
prenatal smear, 26 (59) had an
abnormal smear at 4 weeks (24 92 had
inflammation, one 4 had inflammation
with nuclear atypia, and one 4 had cervical
intraepithelial neoplasia CIN
III), 17 (32) were abnormal at 6 weeks (15 88
had inflammation, one 6 had
inflammation with nuclear atypia, and one 6
had CIN III), and 12 (28) were
abnormal at 8 weeks (11 92 had inflammation
and one 8 had inflammation with nuclear atypia
). There were no differences in the distribution
of abnormal Papanicolaou smears at the repeat sme
ar done 3 months after the postpartum
examination. CONCLUSION The incidence of
abnormal Papanicolaou smears increased
as the postpartum interval decreased from 8 to 4
weeks.