Individual Critique of a Quantitative Analysis Report
“A Randomized Control Trial of Continuous Support in Labor by a Lay Doula:” a Critique of a Quantitative Analysis Report
The process of giving birth places exceptional strain on a mother and is associated with significant risks and complications. According to a recent report published by the Centers for Disease Control and Prevention, 32.3% of all births in 2009 were delivered via cesarean section, a record rate (Martin et al., 2011). There are substantial risks involved with such a procedure for the mother as well as the child. Prolonged labor can increase the likelihood of having to perform a cesarean section (Lopez-Zeno, Peaceman, Adashek & Socol, 1992), while labor analgesia is also associated with a number of serious complications (Norris et al., 1994). The benefits of supportive care for women during labor are well established in the literature but nonetheless present a substantial financial burden that most underinsured women are not able to carry. The authors of the study, A Randomized Control Trial of Continuous Support in Labor by a Lay Doula, conducted a doula training program for a female friend or family member of the mother and examined the effect of their presence on the labor outcomes in women. This could potentially improve the health of the mother and child without placing additional strain on the nursing staff. This paper presents a critique of the various elements of the quantitative research report.
The purpose of the study was to determine the effect of supportive care of women in labor by comparing labor outcomes in women attended by a trusted support person trained in doula care with outcomes in women who did not have additional doula support. The research question that the authors were trying to answer in the study was whether supportive care during labor, specifically by a female friend or family member trained in doula care, would have a clinical impact during the delivery. Does it reduce the anxiety and stress-associated complications such as length of labor, rate of Cesarean section and Apgar scores of the baby? The hypothesis of the authors is not explicitly stated but based on previous studies and current knowledge about the beneficial effects of supportive care, the likely hypothesis of the study was that an additional support person would lower the complications of the labor process in comparison to the control group. The only independent variable of this study was the presence or absence of an additional support person during labor. Dependent variables were length of labor from onset of contractions to delivery of neonate, type of delivery as either spontaneous vaginal, forceps, vacuum-assisted, or cesarean, type and timing of analgesia and Apgar scores of the baby at one and five minutes.
The conceptual framework of this study is based on the physiological effect of the catecholamine hormones that guide the labor process. According to past studies, an elevated level of these hormones during labor reduces blood flow to the uterus and placenta and reduces the contractility of the smooth muscles. This is associated with slower dilation rates, longer labors and therefore higher complication rates. The emotional stress and anxiety of the labor process induces the release of epinephrine, one of the catecholamines, and thereby exacerbates the effect on the uterus and placenta. The conceptual model guiding this study relies on the idea that emotional support during this stressful time by a familiar person who is trained in techniques to comfort the mother, can reduce the levels of catecholamines and thereby increase the contractility of the smooth muscles responsible for delivering the baby. Moreover, the support provided by the doula may also help release the hormone oxytocin, which further strengthens contractions. These assumptions by the authors are well documented in the literature and provide a valid theoretical framework for the study.
The extensive literature review conducted by the authors found several studies that report a significant benefit of supportive care during labor. According to one study, the support of an untrained close female relative resulted in less cesarean sections, less intrapartum analgesia, and less use of oxytocin (Campbell, Lake, Falk & Backstrand, 2006). The authors provide a comprehensive review of the literature to support their assumptions and their study design. They used previous studies to validate the impact that hormones have during labor and to emphasize their assumption that supportive techniques may affect the hormone levels. Further, the authors provide gaps in knowledge to justify the purpose of their study. Specifically, they reference a study that showed little impact of the continuous support during labor of a professional nurse and cite possible explanations for this. This gives credence to their hypothesis that a personal friend instead of a professional could potentially be more effective in helping a mother during labor.
The design of the study was appropriate to obtain the most statistically reliable results. The authors strictly controlled the study population by limiting eligibility to nulliparous, singleton-, low-risk pregnancies with women able to identify a female friend of family member. Only after obtaining consent, did the researchers reveal what group the subject had randomly been assigned to. The pregnant woman and her female friend attended two, 2-hour instructional sessions during which the women were taught coping strategies during labor and the doulas were taught how to provide guidance and comfort measures. The training content was consistent for every training session despite a variety of settings. Each subject received printed material for future reference. On the day of delivery, the personal doulas attended labor along with any other family members or friends of the mother. There were no restrictions placed on who attended the labor and professional medical personnel were not blinded to the group assignation of the subject. This may be seen as a weakness as discussed below.
The study population consisted of 600 women with 300 equally distributed between the two groups. The likelihood of a randomization producing two groups of exactly the same size is relatively unlikely and engenders some doubt about the distribution method. A power analysis was calculated prior to the study to determine the appropriate number of subjects needed to obtain statistically significant results. A total of 586 women completed the study for a 76% power value. The representativeness of the samples was weak as an overwhelming majority of women were white, non-Hispanics in their early twenties. However, the demographic distribution was consistent between the two study groups, which adds strength to the design and helps to control one of the extraneous variables associated with the study. The study was conducted at a women’s ambulatory care center located at a tertiary perinatal care hospital in New Jersey. The setting is appropriate in providing ready access to eligible subjects; however, a less advanced or lower resourced facility might have produced even more significant results in favor of the authors’ hypothesis.
There are a number of extraneous variables inherent in the study, which the researchers appropriately controlled for. The most significant potential contaminant is the nature of the women’s pregnancy. The complications of labor increase with age and number of pregnancies and despite randomization, a heterogeneous pool of pregnancy types would have influenced the labor outcomes. By strictly controlling for the type of pregnant woman eligible for enrollment, the authors reduced the likelihood of confounding their results with an additional variable. Another possible extraneous variable, though less impactful than the type of pregnancy, is the site of the doula training provided by the research assistant. The training was provided in all types of settings. The authors attempted to control for this variable by using the same research assistant, and providing precisely the same content and reference material.
Outcome data were collected from the intrapartum period by a retrospective hospital record review. The validity and reliability of this method is very high because the hospital records are based on objective data that is routinely collected as part of hospital policy. The authors do not explicitly discuss the reliability of the study instruments. This should not be considered an omission because the authors are not involved in the data collection process, which makes it more reliable by removing bias. Specific measures taken to protect the rights of the subjects were not mentioned. One potential but unlikely influence on the data is the subjective evaluation of the neonates by the delivering physician in form of the Apgar score. Knowing what group the subject belonged to may have influenced the physician in his evaluation.
The data were analyzed using standard statistical comparison tests. Two data analyses were performed: one on the entire population and another on only those subjects that met the eligibility criteria throughout the labor process. In this way, the authors removed any subjects that may negatively influence the data and be inappropriate for answering the research question. The data analysis procedures were statistically valid and thorough and the authors showed sensitivity towards removing subjects that in the process of the study disqualified themselves in some way from the original eligibility criteria. Two major strength of the study lie in the stringency of the inclusion criteria and the decision of the authors not to control for who was allowed to attend the birth. The former controlled for the number of possibly confounding variables and made the two study groups as homogenous as possible and the latter ensured that the control group did not have a negative “care omission” effect — that is the removal of a support person. One inevitable weakness of the study is that women in the control group are made cognizant of the benefit of a support person, which may prompt them to recruit someone for their own birth. Another weakness is the likely variation in the doula training provided by the one research assistant. Providing four hours of training to 300 women over the course of four years inevitably introduces a source of variation.
The study found that support from a female friend or family member produces significantly shorter length of labor, greater cervical dilation at the time of epidural anesthesia, and higher Apgar scores at 1 and 5 minutes. In short, supportive care improves labor outcomes. The major implication of this study is the cost effectiveness of providing basic doula training to a person that is close to the mother. This allows low-income, under-insured women to receive the benefits that doula techniques offer. It also improves birth outcomes without the additional costs for the hospital to provide doula services. The authors conclude that, “providing education to become a lay doula should be considered as the standard of care during the prenatal period” (Campbell, Lake, Falk, & Backstrand, p.462, 2006). This study provides the data to substantiate their claim.
Campbell, D.A., Lake, M.F., Falk, M., Backstrand, J.R. (2006). A Randomized Control Trial of Continuous Support in Labor by a Lay Doula. J Obstet Gynecol Neonatal Nurs, 35, 4, 456-64
Lopez-Zeno, J.A., Peaceman, A.M., Adashek, J.A., Socol, M.L. (1992). A Controlled Trial of a Program for the Active Management of Labor. N Engl J. Med, 326, 450-454
Martin, J.A., Hamilton, B.E., Ventura, S.J., Osterman, M.J.K., Kirmeyer, S., et al. (2011). Births: Final Data for 2009. National Vital Statistics Reports, 60, 1
Norris, M.C., Grieco, W.M., Borkowski, M., Leighton, B.L., Arkoosh, V.A., Huffnagle, H.J., Huffnagle, S. (1994). Complications of Labor Analgesia Epidural vs. Combined Spinal Epidural Techniques. Anaethesia & Analgesia, 79(3): 529-537