Justus Hofmeyr, James Neilson, Zarko Alfirevic, Caroline Crowther, Lelia Duley, Metin Gulmezoglu, Gillian Gyte, Ellen Hodnett
The Cochrane Collaboration and John Wiley & Sons, 2008
418 pp, $60, pb
A Cochrane Pocketbook is a selection of abstracts from Cochrane reviews relevant to the care of women and their babies before, during, and after pregnancy. It includes brief descriptions of the nature of evidence, and the Cochrane Pregnancy and Childbirth Group’s review methodology, followed by chapters designed to make it easy for users to locate a topic and browse the available evidence: Context of Care; Antenatal Care; Medical Problems in Pregnancy; Disorders Affecting the Unborn Baby; Pregnancy Complications; Induction of Labour; Care During Childbirth; Care After Childbirth; and Rogues’ Gallery.
Chapters contain short abstracts summarizing the relevant Cochrane Reviews and Protocols, along with Cochrane Database (CD) numbers to facilitate an online search of the full reviews. References are not provided. Each topic is headed by its Cochrane Review title and a brief conclusion statement, and then divided into background, objectives, methods, main results, and author’s conclusion. The 425 entries are equally divided among reviews with good evidence, reviews with poor evidence, and topics without randomized trials or Cochrane review.
Strengths of the book include its scope. Many useful topics are summarized in a small space. The book’s stated guiding principle is “to advise no interference in the process of pregnancy and childbirth unless there is compelling evidence that the intervention has worthwhile benefits for the mother and/or her baby—only then is there a good chance that benefits will outweigh both known adverse effects and those which may not have been thought of” (p xiii). For the most part, authors are careful about the limitations of current evidence, qualifying recommendations with comments about external validity and women’s values. Where evidence is thin or lacking, this is stated up front. A full 142 entries lack current evidence to guide clinical care—focusing attention on future research directions, but using a lot of space to tell us what is not known. In contrast to A Guide to Effective Care in Pregnancy and Childbirth, the reviews contain more evidence than guidance and are more sterile as a result.
In light of new research, some evidence would be expected to soon be out of date, but there are some striking examples. Reviews from 1998 on the timing of antibiotics at cesarean section and from 2001 on amnioinfusion for meconium-stained liquor have been superseded by new large randomized trials: antibiotics before surgery are superior to those after cord clamping, and amnioinfusion does not decrease the risk of meconium aspiration syndrome. Some of the listed protocol questions have also since been answered (i.e., steroids for elective cesarean section at term; method of placental removal at cesarean section). Although citing these references would have opened the book to criticism for not following standard Cochrane methodology, sticking solely to existing Cochrane Reviews, even when clearly out of date, detracts from the book’s utility and user-friendliness.
The book lacks a consistent voice, yielding occasional redundant and contradictory reviews. Magnesium sulfate is neuroprotective for preterm infants (CD004661; n= 3,701), yet increases perinatal mortality when used for preterm labor (CD001060; n= 727). These reviews have common authors, one of whom also co-authors the handbook, yet readers are left to decide which review to believe. In women with a prior history of preterm birth, treatment of bacterial vaginosis in pregnancy is said to both reduce (CD002250) and not reduce (CD00026) preterm birth. Duplicate reviews on Joel-Cohen cesarean section are also unnecessary (CD004662 and CD004453).
Although Cochrane is ostensibly solely based on randomized trials, “large observational studies” are appended to support the safety of some practices (external cephalic version), but not others (breech birth). Since the book is based entirely on meta-analysis, there are also expected deficiencies in external validity (epidural analgesia, breech birth) and the drawbacks of reporting relative rather than absolute risks.
Continuous fetal monitoring versus intermittent auscultation in labor increases cesarean and operative vaginal births but decreases the risk of neonatal seizures. A relative risk of neonatal seizures of 0.5 looks serious until, with some effort from the full-length online review, one calculates the absolute risk difference to be 1.5/1,000 and finds that no difference in long-term outcome or cerebral palsy has been shown. The author’s conclusion that “The real challenge (with CEFM) is how best to convey this uncertainty to women to enable them to make an informed choice without compromising the normality of labour” (p 304) is absurd and contrary to the stated guiding principle. Thankfully, most guideline authors recommend intermittent auscultation, sparing women the “choice” between a higher cesarean section and operative vaginal birth rate and a remote chance of neonatal seizures without long-term sequelae.
Finally, the book intermittently uses a simplistic system of happy, neutral, and sad faces to grade interventions, intended to give the reader a quick and dirty impression of the review’s conclusion. Unfortunately, these faces are sometimes incorrect. Examples include selections from the Rogues’ Gallery, a collection of eight unhappy faces representing interventions ostensibly proved to cause harm. The smaller aforementioned magnesium sulfate review gets an unhappy face for increased perinatal mortality, even though it may be outweighed by neuroprotective benefit shown in the larger review. Peritoneal closure at cesarean section also gets an unhappy face in the Rogue’s Gallery due to a clinically insignificant increase in analgesia requirement and a 5 percent absolute increase in the likelihood of postoperative fever, both possibly related to the use of chromic suture. The author concludes: “Long-term studies following cesarean section are limited; there is therefore no overall evidence for non-closure until long-term data become available” (p 346)—perhaps a wise caution given longitudinal cohort data suggesting increased surgical adhesions at repeat cesarean section after peritoneal nonclosure. Again, the jury is out and the unhappy face and place in the Rogues’ Gallery are unjustified.
It is seductive to try to simplify evidence. A Cochrane Pocketbook: Pregnancy and Childbirth contains concise, useful reviews of many topics and lists a large number of clinical situations where no randomized trials are available to guide health practitioners. However, much of maternity care is complex. Variations in trial conditions affect both the integrity of meta-analyses and their external validity to individual parturients. Large studies now enable us to quantify very small levels of risk that may not justify large-scale intervention or be clinically significant to many informed women. Understanding the limitations of external validity inherent in many trials and meta-analyses, quantifying and contextualizing small levels of risk, and placing parturient autonomy first are key to providing women with truly informed choice. With those efforts, this book helps to a limited extent.