Title: 18th Annual Obstetrics Update for Family Physicians, Vancouver, Nov. 2001
118th Annual Obstetrics Update for Family
Physicians, Vancouver, Nov. 2001 What do we
really know what works in postpartum
care? Cheryl Levitt, Professor and
Chair, Department of Family Medicine McMaster
University.
2(No Transcript)
3Credit to my colleagues The Evidence-Based
Postpartum Research Group Cheryl Levitt, Liz
Shaw, Janusz Kaczorowski, Sharon Brownlee, Russ
Springate, Murray Enkin, John Sellors, Barbara
Stroskopf, Maria Patriquin Supported by a grant
from the Bureau for Reproductive and Child
Health-Health Canada
4Project began as a resident research project in
1998-9 We wanted to develop evidence-based
guidelines for postpartum care
5- So, how do we develop
- evidence-based guidelines?
-
- Literature review
- Summary of the literature
- Guidelines
6- Usually guidelines are based on
-
- RCTs - level 1 (A)
- Cohort studies level 2 (B)
- Expert opinions level 3 (C)
-
7We started with RCTs Research question From
review of RCTs what is the effectiveness of
routine postpartum interventions after birth to 1
year in women?
8- Computer searches
- Medline (1966-1999)
- Cochrane (issue 3, 1999)
- CINAHL (1982-1999)
- PsychInfo (1967-1999)
- Reference lists from review articles
-
9Key words puerperium, postnatal care, maternal
welfare, puerperal disorders, breastfeeding,
lactation disorders, random, single and double
blind
10- Inclusion criteria ?
- Intervention clinically relevant
- 3rd stage - one year
- English language
- Human studies
- Randomized controlled trial
- Original study or review of original
studies
11- Exclusion Criteria ?
- Intrapartum interventions
- Outside of North America, Europe, Australia
and New Zealand - Lactation suppression, endometritis and
hypertensive disorders
12- Abstract Reviews
- Entire paper reviewed if there was inadequate
information - Identified abstracts reviewed by 2 clinicians
- Clinical relevance was assessed by 2 clinicians
13- Consensus
- Consensus by 2 clinicians/ or I of 3
clinician/methodologist pair - - abstract review
- - clinical relevance
- - data abstraction
- Disagreements were resolved by discussion
with the entire group
14RESULTS
CINAHL 32 Articles
Cochrane 356 Articles
Psycho Info 46 Articles
Medline 237 Articles
671 Articles
103 Duplicates Removed
568 Articles
140 Articles Included
428 Articles Excluded
15 RESULTS
140 Articles
Sorted by Topic
66 Articles Other
33 Articles Perineal Pain
41 Articles Breastfeeding
Consensus Reached for Clinical Relevance
63 Articles
16Data extraction
Jadad Scale to Assess Methodologic Quality of
Randomized Clinical Trials Citation (Study ID)
__________________________________________________
____________ 1. Was the study randomized? ?
yes ? no Add 1 point 0
points If yes, was randomization described
and appropriate? ? yes Add 1
point OR was randomization described and
inappropriate? ? yes Deduct 1
point 2. Was the study double-blind? ?
yes ? no Add 1 point 0 points If yes,
was blinding described and appropriate? ?
yes Add 1 point OR was blinding
described and inappropriate? ?
yes Deduct 1 point 3. Were
withdrawals/dropouts described? (numbers of
patients and reasons must be stated) ? yes ?
no Add 1 point 0 points
17- Huge time commitment
- Each article took at least one hour to review
- Each reviewed by two people
- MainPro C credits
18(No Transcript)
19- Challenges
- Huge number of studies total about 140
- other 63 articles analyzed first
- Obvious heterogeneity of studies
- Traditional methods of analysis (i.e. pooling of
data) not possible - Poor RCTs
- RCTs might not be best study for practice
20- contraception
- rubella immunization
- postpartum medical disorders UTIs, haemorrhoids,
constipation, weight loss. anemia - postpartum support
- early discharge
- single topics pap smear, smoking cessation
- nutritional supplements (other than BF)
- effects of exercise
- postpartum depression
21We tried to keep our eye on the ball
What are the clinically relevant questions to
guide us in this systematic review?
22- ? Contraception
- ? Rubella
- ? Postpartum medical disorders - Pap smears
23Contraception
- BCP
- IUDs
- Rhythm method
- Barrier methods
- Exclusive breastfeeding
24Contraception
- Clinically relevant Questions?
- Choice of contraception
- Timing
- Early vs. late insertion of IUD
- Efficacy, side effects, complications etc
- Ethical issues
25Contraception
3 RCTS meet our inclusion criteria 2 on IUDs
1 on progesterone vaginal ring NONE on
BCP NONE on barrier methods NONE on rhythm
26Contraception
- Methodological quality
- 2/5 (Van Kets)
- 2/5 (Hëikkla)
- 3/5 (Landgren)
-
- The randomization method was unclear in the
first two trials and none of the studies were
double-blinded.
27IUD
Clinical question What is the efficacy and side
effect rate of the various IUDs inserted
postpartum?
28IUD
- Result
- The 2 IUD studies looked at the use of the
Nova-T. Heikklä et al compared the Nova-T with
two Levo-Norgestrel releasing devices. Van Kets
et al compared the Nova-T with the Nova-T PP. - Main outcomes were expulsion, removal for
bleeding and pain and pregnancy rates in both
studies. Results are shown in Table 1.
29Comparison between the two IUD studies that
looked at the expulsion, removal and pregnancy
rates
NR not reported
30IUD
There were 40 women randomized to the Nova-T in
the Heikklä study (Finland) and 203 women in the
Van Kets trial (Belgium). Both of the
Levo-Norgestrel containing IUDs had significantly
reduced days of menstrual flow after 12 months
use compared with the Nova-T (plt.001).
31IUD
Clinical question Compared with delayed
insertion, is immediate postpartum insertion of
an IUD associated with decreased efficacy and
increased side effects?
32IUD
- Result
- Heikklä et al inserted all IUDs in the study
at 6 weeks postpartum, Van Kets et al inserted
all IUDs in their study immediately postpartum. - There are no studies that compare the two
insertion methods within the study.
33IUD
Discussion All 4 IUDs reported here
effectively prevented pregnancy and had low
expulsion rates, and removal for pain and
bleeding. Although it appears that immediate
postpartum insertion of the Nova-T IUD was
associated with higher expulsion rates and
removal for pain and bleeding than the 6 week
insertion, there was no direct within-study
comparison of these two insertion times,
therefore, no conclusions can be drawn. The
Cochrane Review came to similar conclusions. The
progesterone containing IUDs reduce menstrual
flow more after 12 months use compared with the
Nova-T.
34IUD
Conclusion All of the IUDs studied in these
RCTs were effective in preventing pregnancy, and
had acceptable side effect rates. Expulsion
rates were up to 6.6 at 18 months and removal
for pain and bleeding was less than 10. Many
questions including infection rates and
appropriate timing of insertion have not been
addressed.
35Pap Smears
Clinical question Does the timing of the
postpartum PAP smear influence the proportion of
dysplastic and non-dysplastic abnormal results
requiring further follow-up?
36Pap Smears
Quality 2/5 Result The study randomized 184
women into postpartum PAP smear screening at 4, 6
and 8 weeks using a spatula and cytobrush.
37Pap Smears
Conclusions Delaying the PAP smear until 8 weeks
postpartum reduces the number of inflammatory
smears. It is not clear that this translates
into any benefit for the women in terms of a
reduction in the number of follow-up smears.
Further RCT studies should examine the important
question when to do the postpartum pap smear and
follow-up of inflammatory smears.
38Rubella
Clinical question What is the efficacy and
safety of the rubella vaccines given immediately
postpartum?
39Rubella
Efficacy? Quality 2/5 Discussion RA27/3 vaccine
has a significantly higher rate of
sero-conversion compared to the Candehill
vaccine. Verification of this result in a good
quality study in different populations would
provide further evidence of the efficacy of the
vaccine, particularly as there was no arthralgia
with the Candehill.
40Rubella
- Safety? Quality 4/5
- Discussion
- There is a small but significant increased
risk of persistent arthralgia /arthritis at 12
months postpartum in women receiving the RA27/3
compared with placebo (22 vs. 15). However, up
to 30 of women infected with wild rubella virus
experience joint symptoms for up to 2 years.
Other adverse affects might be significant with
larger sample sizes.
41(No Transcript)
42(No Transcript)