Student Projects
Quantitative Research Proposal

Eating and Drinking during Labor and Perinatal Outcomes

Kaye Sanford
Chapter 2: Review of the Literature and Conceptual Framework
Restatement of the Problem
Midwives support those women who choose to exercise the option to eat and drink during labor. The policy of giving nothing by mouth (NPO) to laboring women continues to be practiced today, long after the reasons for that directive have disappeared. Further education of health professionals and pregnant women regarding intake in labor is now required to encourage better collaboration in the development and institution of appropriate policies that are more closely in keeping with the available evidence for best practice regarding this important issue.

With this argument in mind, this study seeks to explore the effects of eating and drinking on maternal and fetal wellbeing during labor. As the practice and the ongoing debate regarding it does relates to maternal wellbeing, it is useful to explore both the physiological and psychological effects of withholding food and liquids during labor. In terms of the wellbeing of the fetus, it is important to understand the physiological effects on the fetus of the NPO practice. With the goal of directing discussion toward a constructive alliance and a more effective policy choice, this inquiry explores both the risks and the benefits associated restricting with eating and drinking on the birthing process.

Methods Used to Identify and Retrieve the Relevant Literature

A search for quantitative and qualitative articles and available literature related to the issues of eating and drinking during labor and perinatal outcomes was conducted
by using electronic journals and databases such as Medline, the library at Philadelphia University, Pubmed (the National Library of Medicine), Google Scholar, EBSCOHost, ProQuest, and online resources. Search terms included: eating, drinking, labor, general anesthesia, regional anesthesia, latent labor, active labor, ketosis, accelerated starvation, pregnancy, and maternal deaths. The research literature was then selected based on the criteria of recentness and appropriate applicability to the topic. In addition, the current issues of midwifery and medical journals were examined for appropriate articles that related to the issue examined by this inquiry.

A Review of the Current Literature
The following literature review closely explores articles that pertain to eating and drinking during labor and the physiological and psychological effects of that practice on the mother.

Eating and Drinking in Labor: The Influence of Caregiver Advice on Women’s Behavior

Scheepers, Thans, DeJong, Essed, LeCessie & Kanhai (2001) conducted a study to determine whether caregiver advice affected eating and drinking behavior during labor. The study, conducted in the Netherlands, invited 135 randomly selected midwives and obstetricians to participate. Of these, 32 midwives and 34 obstetricians agreed to participate, resulting in 330 nulliparous women for the study. Each midwife and obstetrician selected 5 nulliparous women under their care at the start of delivery. Each woman answered open-ended questions concerning the advice they had been given related to eating and drinking during labor. Midwives and obstetricians both provided data that included maternal age, duration of pregnancy, duration of labor and medical interventions used during the delivery.

A total of 211 women completed questionnaires that were accompanied by sufficient clinical data; 66 percent of the women were not given any advice on eating and drinking during labor. Women who were given advice usually followed that advice. Of the total group, 37 percent of the women had intake other than water. Of the total group, 75 percent ate solid food. After adjusting for other prognostic factors, the incidence of an instrumental delivery due to a nonprogressing second stage of labor was lower in women with caloric intake (13% versus 24%, p=0.04). The results showed that of those women who did receive advice regarding ingestion of food or drink during labor, they usually followed it. The statistical analysis included descriptive statistics, odds ratios, and 95% confidence intervals, a chi-square test and a multivariate logistic regression. The authors of the study concluded that the women did follow the advice of their health provider, suggesting that if further studies should lead to such similar clear conclusions that women would be thus motivated to change their behavior.

Two positive results from this study did emerge: (1) Women are inclined to heed the suggestions of their midwives and/or physicians and (2) no correlation between eating during labor and adverse outcomes. The main weakness of the survey research was its superficial approach regarding the cause and effect between intake and progression of labor despite hints at the importance and nature of the relationship between the two. While this may not be the intent of this type of survey research, the authors would have done better to keep their speculation commentary in line with their research purpose.

The study demonstrated that nulliparous patients tend to follow the advice of their midwives or obstetricians because these prospective mothers have no previous experience on which to rely, except their social network of family and friends. The degree to which nulliparous patients follow obstetric advice, however, implies a belief in the provider’s authority, at least to the extent that their provider has the expertise and knowledge to provide them with the healthiest and safes practices. These conclusions seem sound; care providers do indeed have a strong influence on women’s choices regarding in labor and birth. This study is important, as it provides precise information on how health care practitioners, both midwives and physicians, are dealing with the current issue of oral intake during labor. Self-reporting data inherently involve a number of fundamental biases, all of which are made more complicated by a relatively low average knowledge regarding terms of medical specificity. That being said, the value here is in allowing women in their own terms and language to explain the rationale for their prenatal decisions.

Natural Eating Behavior in Latent Labor and Its Effects on Outcomes in Active Labor

Parsons, Bidewell, & Nagy’s (2006) study conducted a quantitative analysis to determine whether eating and drinking during the latent phase of labor, defined as “regular painful contractions 5 minutes apart until the woman’s cervix was 3cm dilated” (p. 2) had any effect on the duration of labor and birth outcomes for mother or infant. Over a 7-month period, four hospitals in Sydney, Australia, were contacted for this prospective, comparative study. Three hundred thirty-one, low-risk, nulliparous women were initially selected, and of these, 103 were excluded because of induction of labor, Cesarean section, or both. An additional 11 women were excluded from the study because information regarding the commencement of labor conflicted with the data recorded by the practitioners who were managing their care, such that the actual labor phase could not be definitively identified. Another 41 women were excluded because they ate food during the active phase of labor (10 women ate during the active phase of labor only, and 31 women ate during both the latent and the active phases of their labor).

The final sample size for analysis contained 176 low-risks, nulliparous women. Eighty-two women elected to consume food, and of this group, 23% consumed full meals (e.g., meat, vegetable, pasta with sauce, fish, and fast foods) and 77% consumed light meals (e.g., toast, soup, cereal). The other 94 women elected to have an intake of clear fluids only. The women were enrolled in the study at 36 weeks’ gestation and given a survey form to complete when their labor commenced and all foods and fluids had been voluntarily consumed. The investigators were interested in the eating and non-eating group rate of medical interventions, adverse birth outcomes, or vomiting.

Data analysis was conducted by using STATISTICA software. Eating and non-eating group demographics were compared by using t tests (maternal age and gestation) and a x test (ethnicity) to establish the groups that differed based on relevant background variables. The effect of eating during latent labor on the incidence of adverse labor and birth outcomes was examined via x tests and discriminant analyses. The x tests compared the percentages of the eating and non-eating groups experiencing adverse outcomes. A multivariate discriminant analysis tested for differences between each labor and birth intervention and outcome for the two groups while controlling for the incidence of other outcomes and background variables (Parsons, Bidewell, & Nagy, 2006).

Both groups were similar in ethnic composition (P=.62). Maternal age was higher in the eating group (P=.04). The mean gestation was the same for both groups (P=.56). Twenty-three percent (23%) of the eating group and 30% of the non-eating group vomited during labor (any phase). The vomiting was unrelated to whether a woman ate or did not eat during the latent phase of labor (P=.80). The latent phase of labor for the women that ate was twice as long as those that did not eat (P< .001).

The researchers concluded that no difference existed between the two groups in the need for intervention by medical professionals, vomiting, or adverse outcomes at birth; however, the group who consumed food at will during the latent phase of labor recorded longer labor duration, a finding that suggests that eating during the latent phase of labor may be responsible for increasing the length of a woman’s labor. The most likely explanation for this was one of limited resources: Energy that might have been used to progress in labor may have instead been used for metabolizing the consumed food. The lack of controlled scientific studies here means that there was no evidence to confirm or deny the opinion on either side of the debate. Controlled randomized clinical trials cannot investigate every phenomenon and this study does provide evidence for us to consider.

The strength of the study was the use of nulliparous women and the use of naturalistic methodology that allowed the women to be free in their choices. The weaknesses of this study appeared to be the limited size of the population studied, and the inability of the study to control for the type and amount of food being consumed. While naturalist in its approach, this research study may have missed out on essential nuances in terms of particular metabolic reactions to certain food types. It might have been beneficial if the study had included more detail about the particular pain control methods that were used for any participants. The information from the study is interesting and unique. While there are other studies involving eating and drinking during labor, none address a breakdown of the diet the women were ingesting. This study provides further information on the debate on whether eating and drinking during labor is safe. As a naturalist study, its results can be helpful in predicting the natural give and take of food intake desires during labor and their consequence for labor. However, these results cannot serve as a conclusive foundation for predicting who will have improved or worsened outcomes (eating versus non-eating). It also remains unclear as to whether or not the food intake led to a better experience with labor. It should also be noted that while labor was shorter with those women who consumed no food, temporal duration does not in and of itself constitute a better experience with labor from the perspective of the birthing mother.

An Evaluation of Isotonic “Sport Drinks” During Labor

This study evaluated the use of isotonic drinks to prevent ketosis and the risk of aspiration (Kubli, Scrutton, Seed, & O’Sullivan, 2001). After receiving approval from the ethics committee at St. Thomas Hospital, London, the study was initiated. Upon admission to the hospital, 91 women in early labor were asked to join the study; 31 declined consent. The 60 remaining women joined the study and were randomly assigned to assure that the treatment groups were similar to each other; 30 women were to receive water only, and 30 women were to receive an isotonic sport drink and water throughout their labor. Isotonic drinks are known as “sport drinks” (e.g. Gatorade, PowerAde). Women requesting IV pain medications were excluded on the grounds that analgesics delay gastric emptying. Women requesting epidural analgesia were given a small-dose infusion. The isotonic drink used in the study was Lucozade Sport, which is a mixture of dextrose, maltodextrin, sodium, potassium, calcium, and glucose and comes in orange and lemon flavors.

All 60 women were at least 37 weeks’ gestation and presented without any known complications. The 30 women receiving the Lucozada Sport were asked to consume up to 500 milliliters (ml) the first hour and then 500ml every 3 to 4 hours during their labors. The women consuming water only did not have limit on consumption. A metabolic profile was obtained by blood sample at admission and again at the end of the first stage of labor. Real-time ultrasonography was used within 45 minutes of delivery and by the same investigator to measure gastric residual between the groups. Additional data recorded included incidence and volume of vomiting, stages of labor, mode of delivery, Apgar scores, and blood gases. Metabolic, gastric volume and vomiting data were analyzed by parametric methods using linear regression analysis. Nonnormality and unequal variances correction was found by using a Huber-white sandwich estimator.

All data were analyzed by the intent-to-treat principle, which means that individuals who did not strictly adhere to the particular managerial guidelines of the study were nevertheless included in the study results, so that the results would accurately reflect the “real” experience of subsequent strategies predicated on the results. Intent-to-treat works because often patients deviate from adherence due to a prognosis or some misunderstanding, conditions that repeat in the same situation outside the clinical trials. Analysis was then completed using statistics software package stata. All subjects of the study were similar in age, parity, induction, and cervical dilation. Metabolic blood work was similar at admission, but by the end of the labor, plasma -hydroxybutyrate (a key ketone producer) and non-esterified fatty acids (NEFA) increased, and glucose significantly decreased in the water group. The plasma -hydroxybutyrate was significantly reduced, but glucose remained stable in the sport drinks group. Gastric residual measurement was comparable between the two groups. The total sport drink consumed was 925ml compared to the water consumed at 478ml. The mean calorific intake of the sport drink was 47 calories/hour and no calorie/hour for the water only group. The results of the study indicated that the sport drinks reduced the volume of ketone production without increasing the residual gastric volume when compared to the water only group. This finding suggests that the isotonic drinks may be a safe alternative to water for women in labor. In the study, secondary results were minimal and did not deviate significantly from the results from those, who did not consume the isotonic drinks, even in the instance of aggregate gastric volume. That being said, the authors emphasize that the isotonic drink was selected for its relatively low carbohydrate concentration and its faster gastric emptying; thus carbohydrate calories were lower than might be the case with other commercial isotonic drinks.

This article would be classified as quantitative because it presented and analyzed exact measurements of the participants and the fluids used in the study. The strength of the study was the considerable detail of different aspects of the samples, including age, BMI, cervical dilation, analgesia, plasma levels, duration of labor, mode of delivery, neonate, and amounts of the fluids consumed throughout labor. The study does demonstrate that isotonic drinks are beneficial during labor, evidencing the researchers’ hypothesis. In addition, there was a significant sample size, which aided in the accuracy of the results. Another positive finding was that no one backed out of the study, which aids in its accuracy, as well.

The main limitation of the study was that it only compared sport drinks against water. Many women in labor are offered clear liquids, such as apple juice, clear soda or tea. The study would have been more useful if measured against a sample of women who were drinking those types of liquids, as well. Whether or not these particular sport drinks in this study are beneficial during pregnancy would be the next logical step for the research team. This study assists this inquiry by providing specific data that shows that there are other options to feeding women during labor and besides the risk of aspiration. Isotonic drinks may provide an alternative for those women who self-regulate to only liquids and those who desire supplemental caloric intake during labor. Although there is a need to study isotonic drinks in contrast with other caloric beverages, this study suggests that options other than food prevent the production of ketones during labor.

The information gathered from this study contributes to an understanding of whether laboring women should eat or not. Other areas that need to be studied include a comparison between the effect of sports drinks and other caloric beverages on the production of ketones during labor and a comparison of sport drinks against food given during labor.

A Comparative Study of the Effect of Food Consumption on Labour and Birth Outcomes in Australia

In a prospective, comparative design study that used concurrent controls, Parsons, Bidewell, and Griffiths (2005) examined the effects of women’s food consumption during labor on the progress of their labor, its duration, and the birth outcomes. Subject to methodological limitations, the goal of this study was to determine whether eating during the first stage of labor increases or decreases the duration of labor and the need for medical intervention. The authors used a naturalistic passive approach which allowed the participants to eat whatever and whenever rather than to eat a prescribed diet. At four public hospitals in Sydney, Australia, a volunteer sample totaled 217 nulliparous women (123 women ate food during labor and 94 women consumed clear fluids only); these nulliparous women were considered low-risk when studied. After 36 weeks’ gestation, each woman was given a survey form on which she documented the time that regular contractions started and the foods consumed during this time.

The women were divided into four groups: (1) 82 women who ate food only during early labor; (2) 10 who ate while in labor; (3) 31 who ate during both early and established labor; and (4) 94 who consumed only clear liquids. Demographic differences between the four groups were compared using one-way analysis of variance (ANOVA) tests (maternal age and gestation) and a x test (ethnicity [racial origin]). An x test was used to determine the time that labor started. The association between labor and admission to the hospital was determined by cervical assessment; and its association with eating behavior was examined by using a Bonferroni adjusted one-way ANOVA with a follow up Tukey HSD test for the comparison between the four groups (Parsons et al., 2005).

Of the 123 women who chose to eat during labor, 82 were admitted to a hospital that allowed women to eat during labor; the other 41 were admitted to a hospital that had stricter policies on eating during labor. The same policies were in place for the 94 women who consumed only clear liquids. Of these 123 women, diets consisted of all types of food groups, but results from the study indicated that as labor progressed eating decreased. Results of the data analysis revealed that eating during the first stage of labor appeared to prolong labor by 2.16 hours, and eating during early and established phases of labor prolonged labor by 3.5 hours. Other variables, such as vomiting, medical interventions, and birth outcomes were unaffected by the intake of food or fluid.

The strengths of the study are that the sample size was ample and the methods appeared to be sound; however, allowing free access to any foods may have been one disadvantage. The weakness of the study was that the hospitals chosen for the study did not have the same policies in place for allowing women to eat during labor. This article supports the current inquiry by providing data that medical interventions, vomiting, or birth outcomes are unaffected by eating during labor.

This study confirms data that was discussed in the research study noted above, albeit with different author information. There does exist some important differences in the reporting of the data. The previous naturalist study discussion centered on the relative absence of scholarship regarding the actual implications of oral intake on adverse labor outcomes. This second study target more specifically the issue of labor duration and the debate between those who believe food consumption helps and those who believe it hinders labor progression. The data set still suffers the discredit of duplicate publication (indeed, one wonders how the authors could cite a “paucity” of research in their 2006 study while seeing the publication of this 2005 study on the same subject), but the different contextual markups are telling, if only because they emphasize the two different and somewhat oppositional consequences of the research: Adverse consequences (vomiting, gastric volume) are insignificant, but can slow labor progression. Neither study examines the psychophysiological effects of food intake during labor from those actually experiencing the labor; it is not known whether a slower labor is perceived as a negative outcome.

The Effect of Unrestricted Oral Carbohydrate Intake on Labor Progress

Tranmer, Hodnett, Hannah, and Stevens (2005) explore the effects of unrestricted carbohydrate intake on the occurrence of dystocia in low-risk nulliparous women. The study was interesting to explore in terms of whether or not unrestricted oral carbohydrate intake while in labor would reduce the high number of Cesarean sections in North American hospitals by ensuring a steady energy supply for laboring women. The authors note that “available studies show that the energy requirements of labor may be similar to the energy needs of continuous moderate aerobic exercise (p. 324).” With this in mind, the standard nothing by mouth (NPO) practice that is prevalent in North American hospitals and the concurrent increase in Cesarean sections compared with the relatively low c-section rates of birthing centers, may be the result of inadequate maternal nutrition during labor.

To study whether encouraging an increased oral intake of carbohydrates during labor had an effect on labor progress, the researchers here observed the labors of 328 nulliparous women at a Southeastern Ontario teaching hospital. The study divided the women into two groups: The “usual care group” (165 women) and the “intervention group” (163 women). The usual care group received no instruction regarding nutrition during labor and was restricted to ice chips and popsicles during labor. The intervention group received literature about nutrition during labor. These women were permitted unrestricted access to whatever they wanted to eat and drink during labor. Both groups were asked to record their intake both at home and once they entered the hospital.

The study found that 56% of the women in the intervention group reported that they ate or drank some source of carbohydrates compared to only 13% of women in the control group. Using the standard criteria of a mean rate of dilatation of < 0.5 cm/hr. for a period of at least 4 hrs after a cervical dilatation of 3 cm to define dystocia, the study found that there was no significant difference between the usual care group and the intervention group in regard to obstetrical interventions, ketone readings, intrapartum complications (maternal fever, abnormal bleeding, or meconium-stained amniotic fluid) or newborn complications. Using a less stringent definition of dystocia, namely 1.0 cm/hr as opposed to 0.5 cm/hr, a more positive effect was also noted. Even though this positive effect was not statistically significant; at 0.5 cm/hr, the rates were 36% dystocia with the intervention group and 44% of the control, while at 1.0cm/hr, the rates moved from 50% in the intervention group to 61% in the control.

The researchers hypothesized results that confirmed that unrestricted access to carbohydrates would have a noticeable effect on labor progression and, subsequently, the number of Cesarean sections performed. A finding of even a small difference may indicate the need for further study. The sample size was ample, and the methods appear to have been sound, but allowing free access to any types of foods may have been one disadvantage, in that the study could account for any subsequent variations in metabolic rate associated with different types of food. Relating labor to aerobic exercise and attempting to explore how the body reacts to oral nutrition during the physical exertion of labor indicates that some guidance over what women eat may be advisable. In essence, providing women with nutritious foods that have been demonstrated to sustain aerobic activity adequately may be better than allowing women in labor to eat anything they want.

The study does appear to indicate that, given the option, women do prefer to exercise their personal right to eat and drink during labor. The fact that the study did not find any negative differences between the groups supports the premise that there is no need for the practice of NPO during labor.

This quantitative study produced surprising results for the researchers, but remains significant in that it indicates that there appears to be little, if any, reason to restrict the intake of food and liquids during labor. The authors do indicate that “the findings from this trial do not provide convincing evidence to change practice”; however, the fact that the intervention group did not experience any adverse effects after from being encouraged to eat and drink during labor and their willingness to do so does indicate that the practice of NPO during labor certainly does need to be questioned and more directly. The study concentrated on the benefits of unrestricted oral intake during labor, but the fact that there was no harm that accrued may be far more telling than the authors expected and, therefore, cannot be understated.

Future studies regarding what women in labor eat and drink should also consider if and at what percentage of women elect to receive epidural analgesia, and how obstetric medicine intervenes in labor and delivery, and use of non-medical support in combination with hydration and nutrition. In essence, although the study did not affirmatively suggest that women should be permitted unlimited access to food and drink during labor, it simultaneously suggested that preclusion of food intake lacks the scientific support needed to justify it.

Dysfunctional Labor and Myometrial Lactic Acidosis

In a study by Quenby, Pierce, Brigham, and Wray (2004), the authors proposed that abnormal levels of lactic acid in myometrial capillary pH influenced the force and frequency of uterus contraction. In a cross-section study, women undergoing a Cesarean section between November 2001 and November 2002 at Liverpool Women’s Hospital were participants in the study. Women were divided into five (5) groups: 1) twenty-three (23) were women non-laboring, elective Cesarean sections; 2) four (4) were women with failed inductions (never reached 2cm or the active phase of labor); 3) seven (7)were women with regular contractions, dilating but eventually underwent Cesareans due to fetal distress or late diagnoses of breech presentation; 4) eight (8) were women in active labor with a diagnosis of cephalopelvic disproportion (CPD); 5) ten (10) were women with dysfunctional labor patterns that failed to progress even with adequate oxytocin management.

The criteria for admission into the study were determined by two of the authors who were working at the facility at the time of the study. Each woman had to be at least 37 weeks’ gestation and able to give informed consent. The authors noted that all women were fed according to hospital policy, although this policy goes unspecified in the results, an unfortunate omission. Women were excluded if they received general anesthesia. During the Cesarean section, a sample of blood was taken from the first incision into the uterus. The blood was tested for pH, O2 saturation, and lactate levels.

Data was analyzed by using Arcus software for personal computers. Shapiro-Wilks test found that the data were not normally distributed. The Kruskal-Wallis test was used to detect variance among the groups. The results showed that women having dysfunctional labor (group 5) had higher myometrial capillary lactate concentrations and lower pH levels (7.35) than the other groups (7.46-7.49). These results supported the authors’ position that myometrial acid-base balance is associated with dysfunctional labor patterns. The authors theorized that the uterine muscle itself may contribute to the production of lactic acid and that the increase in acidity could adversely affect uterine contraction. While such a claim is now disputed regarding general muscle fatigue, this study shows that similar adverse consequences have been demonstrated in both rat and human uterus contractions.

The weakness of this study is that the women ate or drank while laboring versus the NPO status of those women scheduled for Cesarean section. The study does not identify the types of intravenous fluids that were infused into the groups. Although the sample size was small, the study gains strength from the fact that the participants were all at one hospital and two of the researchers selected the participants. The study supports this inquiry by documenting the safe outcomes with Cesarean section after oral intake in labor, against the overwhelming presumption that oral intake will adversely complicated Cesarean deliveries and the health/safety of the mother. At the same time, the small sample size and a physiologic study limit its application in most areas. The strongest claim the authors can make is their explanation for the failure of oxytocin to stimulate labor (the increased uterine contractions may counteract themselves through their subsequent production of lactic acid, which in turn inhibits uterine contraction). Nonetheless, for the purposes of this study, it should be noted that the culprit for dysfunctional labor is not food intake but underlying physiological outcomes—outcomes that may serve to confound other studies that do claim a connection between oral intake and adverse labor outcomes.

A Series of Anesthesia-related Maternal Deaths in Michigan, 1985-2003

The authors of this study, Mhyre, Riesner, Polley, and Naughton (2007) discuss their findings from a study of anesthesia related maternal deaths. The authors gathered information on the administration of anesthetic interventions related to maternal deaths. Presumably given from anesthesiologists point of view, the study was conducted with two purposes in mind: First, to determine if anesthesia was or was not the cause of the maternal deaths, and second to explore how the administration of anesthesia was related, or may have contributed to the deaths of the women for whom anesthesia had been used during delivery.

The study centered on identifying which pregnancy-associated deaths documented during the years 1985-2003 were anesthesia related or anesthesia contributing. A criteria developed by the State of Michigan was used to determine what constituted a pregnancy-associated death, defined as “the death of a woman while pregnant or within one year of the termination of pregnancy, irrespective of cause (p. 1097)” or a pregnancy-related death, “the death of a woman while pregnant or within one year of termination of pregnancy from any cause related to or aggravated by her pregnancy or its management, but not from accidental or incidental causes (p. 1097).”

The authors found a surprisingly high 855 pregnancy-associated deaths, eight which were anesthesia- related and seven which were anesthesia-contributing. None of the deaths occurred during the induction of general anesthesia, and no deaths were reported during labor or delivery. All of the deaths occurred either during post-operative recovery, when women are released with little but their own medical insights to guide them after birthing, or later, with some of the deaths occurring as late as 26 days’ postpartum. The standard time span for identifying a maternal death as either pregnancy-related or pregnancy-associated increased from 42 days in 1986 to a full year by 1992, making comparisons of maternal deaths more difficult across those years.

To determine which maternal deaths were anesthesia-related and which were anesthesia-contributing, the authors used the criteria used by the State of Michigan during its own reviews namely that the “primary cause was defined as a death attributable to anesthesia, either as the result of the medications used, method chosen, or the technical maneuvers performed, whether iatrogenic in origin or resulting from an abnormal patient response. A contributory factor was defined as death to which anesthesia contributed, but one that would likely have occurred even in the absence of an anesthetic intervention (p. 1099).” The study found that the majority of deaths that could be considered either anesthesia related or anesthesia contributing were reported for women who were obese, morbidly obese, and/or African-American. The study also indicated that “more than half of the deaths” could be attributed to “lapses in postoperative monitoring and inadequate supervision by an anesthesiologist (p.1096).”

The authors conducted their study in a manner consistent with that used by the State of Michigan in its own regular review of maternal mortality known as Maternal Mortality Surveillance. The authors indicate that the State of Michigan has conducted such reviews since 1950 and most recently completed a State review of maternal deaths for the years 1972-1984. The authors did not indicate how often the State of Michigan conducts these reviews or whether or not the State regularly reviews a dozen or more years of records at one time. The authors conducted their own review for the years 1985-2003. They do not indicate why they chose to conduct their study on records from this eighteen- year period or why they elected to do their review before the State of Michigan conducted its own review of the records from these years.

The study sample was extremely small. However, this study is undeniably an indication of how anesthetic intervention and anesthesia has also progressed during the past several decades. Yet, the study does not leave researchers with sample sizes that are large enough to be able to render results truly indicative of any specific factors. Research on uncommon occurrences does benefit, by necessity, by applying different standards for significance. Yet in this case it remains unclear whether or not the results of this Michigan study can be extrapolated outside of this particular data set, given the changing standards and the difficulty in confirming and decoding most state records.

The fact that of the fifteen anesthesia-related deaths, the majority were also associated with women who were obese, morbidly obese, or African-American, indicates that anesthesia may not have been the decisive factor. Indeed, the fact that the hospital staff failed to monitor and/or respond properly to issues that arise during the mother’s post-operative recovery indicates that caretaker vigilance may need to be increased at least for those women who fall into the identified danger categories of being either obese, morbidly obese, and/or African-American.

The study could be classified as quantitative, in that the authors are examining the numbers involved, though the numbers do not lend themselves to the generation of further analytical methods. Descriptive statistics promote understanding of particular pieces of the puzzle, but the lack of inferential statistics limits applicability. Put differently, the limitation here is less methodological than it is operational, in that aggregation and content analysis make the most sense given the confines of the study.

Perhaps the most interesting result of this study is that anesthesia-related deaths can be traced conclusively to the consumption of food or liquid. Of the eight anesthesia-related maternal deaths, one woman had content in her stomach as a complicating factor, according to the medical record, there was “aspiration in PACU; stomach contents noted during laryngoscope and intubation.” Her Cesarean, however, was the result of chronic placental abruption with a stillborn fetus, and she died 21 days after her Cesarean and accompanying hysterectomy and the cause of death was Klebsiella, a potent form of pneumonia, and was thus unrelated to the complication of anesthesia caused by her eating. The woman who aspirated her stomach contents did so, not during delivery, but during the recovery period. This woman’s admission was far more complicated than a “normal” vaginal labor and delivery during which a woman might be permitted to eat and drink.
The authors note that, even though the anesthesiologist is the most qualified individual to handle intubation, the administration of anesthesia, and any post-operative issues that may arise as a result of anesthesia, “an anesthesiologist may not always be immediately available (p. 1101).” These shortages produce the likelihood of human error (the authors identify 19 errors present in the eight anesthesia-related maternal deaths). In addition, the study may also indicate that the individuals who are performing the duties of the anesthesiologist might require either more intense supervision or additional training in order to avoid the lack of monitoring or response that was obviously a contributing factor in far too many of the maternal deaths identified in this study (particularly in the case of the obese, morbidly obese, and African-American populations).

Eating in Labour: A Randomized Controlled Trial Assessing the Risks and Benefits

The study by Scrutton, Metcalfe, Lowy, Seed, and O’Sullivan (1999) was designed to determine whether permitting women to eat during labor would increase gastric residual, alter results of the metabolic profile, and prolong labor and influence outcomes. In a randomized controlled trial (RCT), 94 women presenting in early labor at St. Thomas Hospital were assigned into an eating or starved group. Each woman was required to be at least 37 weeks’ gestation without any preexisting health complication. Any woman that requested intramuscular analgesia was excluded due to the effect of slowing gastric emptying. Women assigned to the eating group were allowed to select from a low-residue diet (consisting of a variety of meal options: Cornflakes cereal, toast, butter and jam for breakfast, semi-sweet biscuits for a snack, toast with butter or jam and low fat cheese for lunch, and tea, coffee, hot chocolate, fruit juice and water allowed throughout labor) throughout labor. Women in the starvation group were allowed water only. Metabolic profiles were obtained early in labor and again at the end of the first stage. Values from both groups were recorded, and the data was compared. Ultrasound readings were performed on both groups using a high resolution scanner, consistent investigator, and within one hour after delivery. All variables related to the labor were obtained and recorded: Duration of each stage of labor, oxytocin augmentation, mode of delivery, Apgar scores, and blood gas samples. Gastric volume, metabolic results, and vomiting data were analyzed using linear regression analysis with dummy variables for treatment, adjusting for baseline metabolic indices. Demographic data and maternal and fetal outcomes were analyzed using t-test and chi-squared analysis. Data were analyzed using the statistics software Stata (Scrutton et al., 1999).

Of the 94 women, 48 were placed in the eating group and 46 in the starvation group: all women were similar in age, parity, gestation, and dilation at time of admission to the hospital. The results of the data showed no difference between the two groups in duration of labor, mode of delivery, oxytocin augmentation, or Apgar scores. All blood gas samples were similar in results. The study did note that as labor progressed, the women chose not to eat, indicating a natural reduction in the desire to consume calories. However, after delivery, in the starvation group, the plasma -hydroxybutyrate was higher, and the glucose and insulin lower. Gastric residual measured by antral cross-sectional area showed the eating group to have higher volume to their vomit and a greater frequency (17/38 versus 8/19, with a volume vomited of 309 vs. 104, a P of .001). The authors conclude, somewhat predictably, that eating in labor prevents ketosis, but increases gastric volume. The two strengths of this study are: first, the study provides firm, scientific data about changes in the metabolic process, given the introduction of food during the labor process. This data supports the idea that food can help provide energy to allow the birth to precede without medical interventions thanks to a more robust reserve of maternal energy. It also highlights the relative lack of adverse consequences of ingestion, for while vomit frequency and volume may increase, no other negative sequelae presented themselves. Second, the study offers an alternative to the consumption of food – isotonic drinks – that runs a middle ground between two medical extremes. However, this suggestion is one of the study’s weaknesses, as the study neither tests nor predicates this suggestion on evidence regarding the efficacy of isotonic beverages. The suggestion comes as something of a surprise, given the lack of clear negative outcomes from ingestion. Vomiting was deemed unpleasant, but starvation would be deemed unpleasant by the women who have their desire for food intake tightly curtailed.

In addition, given that there is no substantive discussion of birth outcomes, and given that nearly the same number of women accepted epidurals in each group, it seems as if any statement about variations in Apgar, fetus weight, birthing outcome, would need to attend more specifically to the role, timing, and reasons for the onset of the epidural, which likely complicates the results by smoothing out potential differences and encouraging medically-timed monitoring or Cesarean section. It is not clear from the study if both groups had the same number of epidural using ladies, and no evidence is provided about the timing of the epidural and what implications that might have on the study outcomes. Since epidurals necessarily limit mobility and birthing positions, they may play an as yet under acknowledged role in outcomes as diverse as lactic acid production and aspiration. This article provides support to the need for maternal energy during labor, though it does emphasize that only a small amount of caloric intake is needed to stave off ketosis.

Frequency of Eating During Pregnancy and Its Effect on Preterm Delivery

The purposes of Siega-Riz, Herrmann, Savitz, and Thorp’s study (2001) were to determine if meal patterns throughout pregnancy could affect early preterm delivery as well as hospital admission for prematurity. Participants were recruited from a cohort study in North Carolina called the Pregnancy, Infection, and Nutrition (PIN) study. Between August 1995 and December 1998, 4,160 women ranging from 24 and 29 weeks gestation were eligible for the study. Prenatal labs and specimens (genital tract, blood, urine) were collected at the time of the recruitment. Self-administered questionnaires related to dietary intake were given which examined meals and snacks consumed during the day. In addition, phone interviews were also conducted to obtain data regarding sociodemographics, health status including vomiting during pregnancy, and previous health history. Medical records from hospitals where participants received care were also reviewed for gestational age at delivery, dating for pregnancy, spontaneous or induced, and premature rupture of membranes. Out of the 4,160, 2247 were successfully eligible for the PIN study. Out of successful recruited women, 90% completed the required dietary questionnaires and 2,081 had provided data through telephone communications to sociodemographics and prenatal and health history questions. Of the 2,081, 33 women did not have gestational age at delivery and were excluded from the multivariate analysis investigating the risks between dietary frequency and preterm birth.

Descriptive statistics were generated and a t test or a chi-square test was used to determine the statistical significance. Meal questionnaires were reviewed and the majority of the participants consumed three meals plus two or more snacks per day. The results of the study showed women who ate snacks and meals less often were slightly at risk for giving birth prematurely: 4% of the population ate 3 meals with no snacks, with a preterm birth rate of 16.6%, .5% ate 2 meals and no snacks, with a preterm birthrate of 9%, 14.5% ate 2 meals plus 1 snack, with 12.4% preterm birth rate, 7.8% ate 2 meals and 2 snacks, with a 13.8% preterm rate, and finally, a single percentage of the women had only 1 meal, regardless of the number of snacks, resulting in a preterm birth rate of 19%. The strength of the article is that it states that metabolic effects of eating frequency in general have demonstrated an increase in positive physiologic adaptations such as improved blood glucose control. The sample size is adequate. The weakness of the article is that the diet of the women was not addressed. The participant may have been eating three meals per day, but the type of meals should have been presented (the self-classification of foods as a meal was left up to the respondents).

In order to understand how women are at risk for preterm delivery due to poor eating habits, the field needs further research to address what type of risks the laboring women may encounter if prohibited to eat or drink during labor. Nevertheless, a few lessons can be discerned from the results. The correlation between lower food intake and higher risk of preterm birth suggests that a certain minimal level of caloric intake is necessary for healthy fetal development, as well as healthy progression to term. In addition, the fact that so many of the women clustered at 3 meals a day plus 2 snacks (71.5% of the populations) indicates a desire to consume food to meet the demands placed upon the body. It remains unclear how late into the pregnancy this desire continues, but that the compelling desire for food consumption mirrors the healthiest course for food intake is worth noting, as it supports those who would site the birthing mother’s interests as being reasonably informed by their own perception of their physiological needs.

Carbohydrate Solution Intake during Labour just before the Start of the Second Stage: A Double-blind Study on Metabolic Effects and Clinical Outcome

This randomized, double-blind, placebo-controlled quantitative study by Scheepers, DeJong, Essed, and Kanhai (2004) was done to determine if carbohydrates given at the start of the second stage of labor might reduce the need for medical interventions such as instrumental assistance. Also they wanted to determine if carbohydrates consumed before the second stage of labor could cause acidosis due to increased lactate levels in mother and infant. Between October 1999 and February 2001, at the Leyenburg Hospital, The Hague, and the Netherlands, 202 nulliparous women without complicated pregnancies were asked to participate in the study either before or during labor. The experimental group was given the solution during the second stage of the birthing process when dilation of the cervix was 8-10 cm, while the control group was given a placebo (water with artificial sweeteners and aromas). The researchers measured indicators such as: time until reaching full dilation, the duration of the second stage, mode of delivery (natural or instrumental), Apgar scores and arterial umbilical cord pH. Ultimately, the results showed no statistically significant difference in instrumental deliveries between the two groups, for while the placebo group involved more instrumental deliveries, the numbers were not conclusive of differences and may have been due to other factors (obstructed labor). This result contradicted an earlier study by the authors, where the data pointed to an increased incident of instrumental deliveries accompanying carbohydrate intake. That study, however, let women consumer carbohydrate-rich drinks whenever they wanted during the course of their first stage of labor. To explain the disparity, the researchers suggest that “not all carbohydrate regimens have proven beneficial” for athletes and something similar may be at work here. If the women were used to low activity levels, they might have experienced the carbohydrate intake as hypoglycemia, which could then provoke an insulin response that would lower energies, a response that would be counter-responded in more strenuous physical activities. This remains a hypothesis, however.

One of the strengths of this study was the careful selection of subjects. Only those in good health with relatively low-risk pregnancies were included, in part to reduce the complexity, but mostly to control for extraneous variables that could affect outcomes independent of treatment. The team was quite thorough in exploring each outcome and making sure all relevant indicators were covered.

This study was unable to demonstrate that consuming food during labor was either beneficial or harmful. Metabolic indicators did not reach dangerous levels in either group, and neo-natal outcomes were very similar. Ultimately, this means that it is up to the discretion of the woman and her midwife to decide whether eating or drinking is prudent. Some women might crave sustenance during labor while others lose their appetite completely. Unfortunately, the authors’ conclusion caves to convention in the face of their own thorough research. They write: “At present, it is difficult to establish whether carbohydrate supplementation during labor is beneficial or not.” The question begged by the study is the one that is at this point so routine that it is awarded an undue presumption; instead of asking if ingestion is beneficial, the question should be whether or not the data helps to determine whether exclusion of carbohydrate ingestion is beneficial or not. Given that exclusion of food intake is the norm, one expects the data to confirm it; that it does not implies that the data in this study supports safety with oral intake in the second stage of labor.


It is common practice to restrict food and the majority of liquids from women in labor in most medical facilities (Tranmer, Hodnett, Hannah, & Stevens, 2005). The rationale for the practice has consistently been that should the woman need general anesthetic, the stomach contents may then aspirate into the lungs, resulting in complications such as choking, pulmonary edema, partial lung collapse, and even death. There is some evidence that, when the administration of general anesthesia was common during childbirth, aspiration was associated with a substantial number of maternal deaths (Parsons, Bidewell, & Nagy, 2006). Therefore, the initiation of the practice of withholding food and liquids from a laboring mother was medically warranted. However, advances in anesthetic administration indicate a clear need for a reexamination of this practice.

Digestions slows during labor, which means that anything ingested during labor may remain in the stomach longer than normal, and these stomach contents may pose difficulties during intubation and administration of general anesthesia. That being said, the majority of laboring women no longer require general anesthesia during labor and delivery (it may be worth wondering whether the requirement was ever as extensive as practitioners once believed), and those who do need some form of pain control fare very well with the administration of regional anesthesia instead of a general. According to O’Sullivan, Liu, & Shennan (2007) deaths during anesthesia were due to aspiration or difficult airway management during resuscitation, not actual eating or drinking during labor. Additionally, studies reviewed for this inquiry indicate that although aspiration may once have been a cause of maternal death, medical advances have reduced the former complications of aspiration primarily to pneumonia. Indeed, in practice areas less concordant with the model here in the United States, the data contraindicates a danger from oral intake. According to Scheepers, DeJong, Essed, and Kanhai (2004) in the Netherlands where women are allowed to eat during labor, the incidence of mortality due to aspiration is comparable to countries with a restriction policy. Deaths were reported though the mid 1970’s as a result of aspiration: what once may have been a significant danger for laboring women has became responsible for less than 2% of maternal deaths. It was reasoned that the majority of these deaths could have also been prevented by the use of regional anesthesia which was accepted well by the laboring mother and the fetus. Obstetric deaths associated with regional anesthesia declined in the mid-1980s (Mhyre, Riesner, Polley, & Naughton, 2007). Still anesthesiologists, as well as obstetricians, nurses, and midwives, continue to restrict oral intake in laboring women, a practice that owes its existence more to routine than to exhaustive research.

The woman who labors with no food or drink may build up gastric juices that are more acidic and which may be more harmful to the lungs if aspirated than the actual food and liquids she might be permitted during oral intake. Additionally, women who receive general anesthesia are likely to be lying flat on their back which further increases the danger of aspiration: thus any build-up of gastric juices would be just as dangerous.
Klassen (1999) points out that a 1946 groundbreaking study entitled The Aspiration of Stomach Contents into the Lungs During Obstetric Anesthesia, by C. L. Mendelson, was the source of the practice of ensuring an empty stomach for anesthesiologists by restricting the intake of food and liquids for laboring women. Interestingly however, Klassen (1999) also points out that Mendelson reported the use of an opaque mask for the administration of anesthesia, the attendance of novice anesthetists and evidence of delayed gastric emptying during labor. Since this landmark study was published, treatment of obstetrical patients has changed. Present-day obstetric practice has evolved from the routine use of general anesthesia to the frequent use of regional nerve blocks when and if anesthesia is required. Modern-day practice for a general anesthetic has also changed and use of a rapid sequence induction technique with reoxygenation followed by endotracheal intubation is standard practice. Furthermore, general anesthesia in obstetrics is no longer considered the field of the novice as it was in Mendelson’s time; instead, anesthesia is considered a specialty practice and anesthesiologists assigned to administer obstetric anesthesia often have additional training above the specialty requirements. Finally, and as a direct result of this early study it is always assumed that the woman in labor has a full stomach, and she is managed accordingly during the induction of a general anesthetic (p. 9).

Klassen’s research presented important implications for the subject at hand, namely that as a short-term solution, NPO, may have been warranted until medical techniques for the administration of anesthesia and the management of pain during childbirth advanced to the point that oral intake by laboring mothers was no longer a pertinent issue. However, since it appears that the short-term solution may have well outlived the problem, a reconsideration of the practice of NPO for laboring mothers has not been reviewed and is currently long overdue.

The Conceptual Framework of the Study

The purpose of a theoretical and conceptual framework is to make research findings meaningful and generalizable (Polit & Beck, 2004). A conceptual framework for this study will supply organization and logic which will direct the development of the inquiry about eating and drinking during labor and the physiological and psychological benefits. Through this inquiry hopefully a change in current practices and policies will follow. The conceptual framework of any study identifies the methods the proposed research intends to use to explore an issue. Smyth (2004) stated that “the framework is a research tool intended to assist a researcher to develop awareness and understanding of the situation under scrutiny and to communicate this.” With this goal in mind, this study seeks to explore the effects of eating and drinking on maternal and fetal well-being during labor and delivery. As the study relates to maternal well-being, it is useful to explore the physiological and psychological effects of this common practice as well. In terms of the well-being of the fetus, it is important to understand the physiological effects that the practice produces on that fetus. This general framework for a research study is supported by existing literature that details both prior studies and opinions on the merits and/or drawbacks of withholding what one study asserts is the natural desire of the laboring mother for oral intake (Parsons, Bidewell, & Griffiths, 2005).

The framework of this inquiry is that of patient-centered care. For many women, the act of laboring and giving birth is an intensely spiritual and life-changing event. As a conceptual framework, patient-centered care requires an emphasis on maternal preference and perception, physiological and psychological effects of NPO, and eating and drinking in labor as accepted practice.

The debate regarding the benefits of Eating and Drinking During Labor

Obstetricians became concerned that restriction of food and fluids during labor could lead to levels of ketosis that were potentially damaging to both the progress and outcome of labor. It is believed by some researchers that intravenous dextrose can prevent ketosis while other researchers believe that it can produce yet more problems which are more difficult to solve. Because of these beliefs, the use of intravenous dextrose has been largely abandoned. Many institutions are continuing to use NPO policies because, their active labor management policies and a lack of factual scientific evidence that mild degrees of ketosis slow labor are not enough to change procedures. The study by Metcalfe, Lowy, Seed, and O’Sullivan (1999) demonstrated that permitting women to eat during labor reduced the chances of developing ketosis, but increased the chances of vomiting during labor due to an increase in gastric contents.

The debate over allowing food during labor has become more involved, and there is a need to re-evaluate our current attitudes toward the practice. According to Tamar Windsor, C.N.M. (personal communication, July 1, 2007) NPO policies can do the mother and baby more harm than good. There is evidence to suggest that starvation policies are distressing to the laboring woman. According to Dr. J. Hood MD, an anesthesiologist (personal communication, July 1, 2007), the emptying time of the stomach for a solid is slower than liquids, so consumption of solids should be avoided during labor. Like solids, semi-solids empty from the stomach at varying rates depending on volume, fat content, pH, osmotic pressure and temperature. In general, fluids exit the stomach more quickly; however, that process can be delayed by low pH, variations in osmolarity, or lower temperatures.

It has been suggested that the best approach is to identify low risk laboring women and allow them a light diet during labor. Unfortunately, this approach fails to address a large part of the problem. Labor is an incredibly exhausting bit of physical exertion, with the body undergoing tremendous changes, most of which are induced through a series of regulated muscle contractions. These contractions alter the location and position of the baby as well as preparing and reshaping the birth canal. The process involves a massive amount of caloric expenditure (Blackburn & Loper, 1992), and at the end of the intrapartum changes, the mother uses her remaining strength to push the baby through the birthing canal. Following the success in that, she may then start nursing the child. For mothers without sufficient caloric reserves, continuing labor progression and birthing may be more difficult.

The Physiological Effects of NPO During Labor

Athletes experiences a delay in digestion much like the woman in labor. However, sports medicine does not recommend that the athlete refrain from an oral intake of nutritive substances. According to Scheepers et al. (2004) the intake of carbohydrates during exercise has been demonstrated not only to enhance performance, but also to delay fatigue. In fact, the athlete is advised to remain hydrated and well nourished for proper performance of the muscles and the cardiovascular system. It can be presumed that any individual attempting a strenuous activity should pay special attention to nutrition. In the case of pregnant women, medical providers advise the intake of additional calories and increased nutrition from the moment of conception through the following forty weeks (Sleutel & Golden, 1999), that is, until the woman reaches the point of labor. Once a woman has begun labor, the majority of medical providers advise that she restrict her oral intake to clear fluids. Some institutions continue to advise the extreme recommendation of allowing only ice chips or sucking on a damp rag, both intended just to keep the woman’s mouth from drying out. Intravenous fluids can be provided for additional hydration. Yet, few would argue that giving birth is an intense physical process, similar to the energy output of an athlete in competition, and often taking a considerable amount of time. It stands to reason, then, that the laboring body may be in as much need of nutrition as the athlete’s body.

It has been argued that the laboring woman should no more go without oral nutrition than the long-distance runner should run a 24-hour marathon without food or drink. The laboring woman experiences increased cardiac output and muscle activity that must be maintained with proper nutrition. Withholding nourishing food and liquids from the laboring woman can result in weakened muscles and undesirable changes in her blood chemistry. Once the laboring mother’s natural store of glycogen has been depleted, her body will use stored fat for energy resulting in the buildup of ketones in the bloodstream (American College of Nurse Midwives, 2000). Siega-Riz, Herrmann, Savitz, and Thorp (2001) agree that “prolonged periods of time without food can cause hypoglycemia, which stimulates a cascade of neuroendocrine events that may ultimately affect the health of the fetus.” Siega-Riz et al. stress the need for future research into “when meal patterns may have the most detrimental effect on birth outcomes.” In the mother, low glycogen stores can result in muscle weakness that affects the function of the uterus to be less efficient causing labor to slow and even stall. Should this occur the medical staff attending the woman may begin interventions such as pitocin, forceps and vacuum delivery, and of course a Cesarean section (Beggs & Stainton, 2002). The body’s muscles cannot sustain labor intensive activity for an extended period of time without ongoing proper nutrition. A body cannot function optimally, let alone properly, once forced into starvation mode.

On the other side of the argument, there are a number of studies that indicate that the consumption of food and liquids during early and established labor produces labors up to 3.5 hours longer than average. These same studies note, however, that the increased length of labor did not adversely affect the labor itself or the final outcome of the birth. Additionally, the authors of one study asserted that a longer labor should not be an indication for medical intervention as long as labor is progressing and both mother and child are doing well (Parsons et al., 2005). Latent labor in particular is often mild in its effects, despite duration.

The Psychological Effects of NPO during Labor

The Royal College of Midwives has devoted an entire chapter of its manual to maintaining nutrition during labor. Their Midwifery Practice Guideline gives “evidence-based guidelines for midwifery-led care in labor,” and lists the following statements regarding the practice of NPO during labor (Spiby, 2005), for instance:
There is insufficient evidence to support the practice of starving women in labor in order to lessen the risk of gastric acid aspiration.
Fasting may result in dehydration and acidosis which, combined with starvation and fatigue, can increase the need for active management and instrumental delivery.
Eating and drinking can allow a woman to feel normal and healthy. Denial of food can be seen as authoritarian and intimidating, and increase feelings of apprehension.
The majorities of sources agree that mild maternal ketosis is a physiological part of normal labor and might even be beneficial.
Narcotics appear to be the major factor in delaying stomach emptying. If these are used, then women should stop eating, and drinking be reduced to sips of water.
While there are no risk factors suggesting the need for general anesthesia, women who wish to eat and drink in labor should be offered a light, nutritious and easily absorbable diet.
The desire to eat, however, would appear to be most common in early labor. As women do not usually wish to eat in active labor, and it is inappropriate to be encouraging them to do so against their natural instincts.

The Guideline further discusses the importance of maintaining a woman’s psychological health during labor and delivery. Spiby (2005) points out that NPO is a restriction that exists only in hospitals. Perhaps the most important evidence provided by this document is that a woman’s body may be the best indicator of whether or not eating and drinking during labor is warranted. Assuming that the human body functions independent of medical thought, that a woman’s body may be indicating a natural physical desire for nutrition at the point during labor at which such nutrition may be most important, i.e., just as the hard work is beginning.

Spiby (2005) calls attention to the psychological interpretation of a laboring woman, namely, that the withholding of food and liquids during labor may increase maternal apprehension because control is being taken away by medical authoritarian. At the one time when a woman should feel in control of her bodily functions, that control is taken from her.

Further studies regarding the safety of ending NPO must be conducted. However, as Spiby (2005) points out “Death or significant morbidity from pulmonary aspiration is so rare that the use of randomized controlled trials to examine this outcome is impossible.” Still, productive studies can be conducted that precisely define how NPO affects the laboring woman and fetus physiologically and psychologically. It would be beneficial to examine closely the other ways that this practice affects mother, child, and birth outcomes, and those can eventually be determined. Chapter 3: Methods

Research Design

Identify your design and discuss its strengths and weaknesses
This quantitative research will collect data from three groups of women randomly assigned to one of three dietary regimens groups and perinatal outcomes will be compared. This study will compare general and clear liquid diets versus NPO except for ice chips during labor. A convenience sampling design of nine hundred pregnant nulliparous women between the gestational ages of 37 to 40 weeks will be randomly assigned into the three dietary regimen groups. A randomized clinical trial will be used for this study to determine if eating and drinking during labor causes adverse perinatal outcomes. Clinical trials involve the testing of a treatment; seeking evidence of the effectiveness of the treatment; random assignment of participants; collection of data; and involves a large heterogeneous sample of subjects from multiple dispersed sites to ensure that findings are not unique to a single setting (Polit & Beck, 2004).
Participants will be randomly divided into three equivalent groups; 1) a diet of protein, starches, fruits, vegetables, 2) clear liquids only, 3) nothing by mouth (NPO) except for ice chips.
The first group will be allowed a general diet consisting of meat (patient’s choice according to tastes, religion, and allergies), starch (in the form of pasta) and fruits or vegetables.
The second group will be allowed a clear liquid diet consisting of broth, tea, Jell-O, sodas, and juices. Patients will be given the choice of whether or not they choose to consume the diet and will be permitted to do so to their personal satisfaction.
The final group, the control group, will be allowed the standard protocol of nothing by mouth (NPO), except water and ice throughout the labor process.
This research design notation is expressed in the following form: X is designated as treatment and 0 is designated as observation.
Group 1: X1(General diet) -----01(Perinatal Outcomes)
Group 2: X2(Clear Liquid) -----02(Perinatal Outcomes)
Group 3: X3(NPO x ice chips)—03(Perinatal Outcomes)

Carefully and accurately administering and tracking what each participant consumes throughout the study provides consistent data that can be measured and quantified to show possible relationships between eating and outcomes: however, issuing the correct food regimen to the correct patient to guard against food allergies and personal preference may be a challenge. Also, there is the possibility that a patient may opt for a whole food meal and then chose not to eat once labor has begun and discomfort is experienced, resulting in a lack of appetite, termed intent-to-treat. Intent-to-treat analysis includes all randomized patients in the groups to which they were randomly assigned, regardless of their adherence with the entry criteria, regardless of the treatment they actually received, and regardless of subsequent withdrawal from treatment or deviation from the protocol (LaValley, 2003).

Internal and external validity
Internal validity refers to the extent to which it is possible to make an inference that the independent variable is truly causing or influencing the dependent variable (Polit & Beck, 2004). External validity refers to the generalizability of the research findings to other settings or samples (Polit & Beck, 2004).
Controls used for threats to internal and external validity in this inquiry are:
1) Pre-screening survey was developed to gather pertinent data regarding preexisting health problems or special diets that may interfere with data collected (Appendix D). Data collected will be analyzed statistically via ANCOVA for their effects on outcomes.
2) Participants will be assigned to dietary regimens with completion of pre-screening survey on a random basis. If patients are randomized into equivalent groups, bias is decreased or eliminated (Polit & Beck, 2004).
3) All subjects will be low-risk nulliparous, 37 to 40 weeks gestation, not planning a Cesarean section, and able to comply with study’s instruction regarding intake during labor. Participants are expected to represent a diverse economic, cultural, and racial background.
4) Dietary regimens will be available at clinical facilities.
5) Participants will have prenatal record on file to determine health status, course of pregnancy, gestational dating, and any other pertinent information.
6) The principle researcher will do all data entry with Statistical Package for the Social Sciences (SPSS) program.

Name the measures you will use to control validity threats
To insure that a diverse group of participants is selected who reflect the population as a whole; the study will use three independent testing locations in socioeconomically and culturally diverse areas of the country. Participants will be screened through the researchers utilizing a standardized and approved questionnaire (Appendix D) and will be selected based on this questionnaire and telephone interviews or face to face meetings with the researchers. To account for participant attrition during the study, due to unforeseen circumstances, an additional ten percent of the target sample size will be recruited to participate in the study.
Additionally, two research assistants (obstetricians or midwives) on the project will receive approved and controlled questionnaires to complete with all required information. The data will be analyzed using Statistical Package for the Social Sciences (SPSS) software for accuracy and consistency by the principle researcher. Any questionnaires or medical records with incomplete or missing information will be eliminated from the final analysis.

Sampling Methods

Nine hundred participants in this multi-center trial will be randomly assigned. Three sites will be used in this study; 300 participants from each site will be assigned to random groups at that site. The following sites will be used in the study.
The first site will be Florida Hospital located in Orlando, Florida. Florida Hospital is one of the premiere maternity hospitals in Florida and is centrally located in one of the most diverse populations in America, allowing for the potential of great differences in income, race, education, and personal beliefs.
The second site chosen will be California Pacific Medical Center in San Francisco, California. California Pacific is one of the largest, private, not-for-profit, academic medical centers in California and is renowned for its excellent services; it provides care to 20,000 first-time mothers and draws upon scientific evidence of the best practices for healthy pregnancy and safe delivery.
The final site of study will be Oklahoma University Medical Center (OUMC) in Oklahoma City, Oklahoma. Located in a lower income and racially, diverse area of the country, OUMC specializes in caring for minority populations and those who may not have access to the full-time medical care, classes, and training given to first time mothers in more affluent areas.

Population of interest
The selection of participants must be nulliparous, be at a low risk for birth complications, be between 37 and 40 weeks gestation, do not plan to have a Cesarean section, come from a diverse economic, cultural and racial background, and be willing to comply with the study’s instructions regarding food and drink intake prior to and during labor.

Sample Size
Using convenience sample in three predetermined locations, nine hundred participants between 37 to 40 weeks gestation will be selected based on the below inclusion and exclusion criteria. Each participant will complete a pre-screening survey (Appendix D) to be eligible for the study. Convenience sampling uses the most conveniently available people for the study; it may, however, pose a problem in that the participants available might be atypical of the population (Polit & Beck, 2004) generating statistics based on prejudice and invalid results.
A sample size of 300 participants from three clinical sites will be recruited for this study. Power analysis is a procedure for estimating sample size requirements (Polit & Beck, 2004). When using 900 participants, the study will have a more representation sample of the population. Smaller samples tend to produce less accurate estimates than larger one, so in other words, the larger the sample, the smaller the sampling error (Polit & Beck, 2004).

Inclusion criteriaExclusion criteria
1. Read/write English
2. Nulliparous
3. Low risk pregnancy
4. Between 37 and 40 weeks gestation
5. Any race/age
6. Early labor
7. Cervix <5cm
1. Scheduled cesarean
2. High risk pregnancy(multiples)
3. Preterm/Postterm
4. Multiparity
5. Poor/No prenatal care
6. Preexisting health conditions
(diabetes, GI conditions)
7. Cervix >5cm
8. Active labor

Entry to site / access to population
The research group will be working with OBGYN.NET, American Society of Anesthesiologists (ASA), and the Society for Obstetric Anesthesia and Perinatology (SOAP).
OBGYN.NET caters to the specific needs of professionals interested in obstetrics and gynecology, the medical industry, and women who have a goal to improve service to aid in the delivery of women’s healthcare. The site was designed by OBGYNs and is monitored by an advisory board of qualified physicians. OBGYN.NET is involved in the continual improvement in women’s health and proper labor procedures and continually strives to find new research and new methods. OBGYN.NET contains vast resources of information and research, has a large membership of doctors and hospitals, and works with various groups and organizations to further research and information gathering.
The American Society of Anesthesiologists (ASA) is dedicated to the educational, research and scientific association of physicians organized to raise current standards of the medical practice of anesthesiology and improve patient care. Doctors are located in clinics and hospitals nationwide and share knowledge and resources throughout their network of members and affiliates. Always striving to improve on the safety and effectiveness of their chosen field, the ASA is continually involved in leading edge research to develop their basis of knowledge and training.
The Society for Obstetric Anesthesia and Perinatology (SOAP) was founded in 1968 to provide a forum for discussion of problems unique to the peripartum period. SOAP comprises anesthesiologists, obstetricians, pediatricians, and basic scientists who share and interest in the care of the pregnant patient and the newborn. Covering all fields relating to childbirth, SOAP members have a vast array of resources, clinics, hospitals, and research at their disposal to further their understanding and practices.
By working directly with members of these highly respected groups, the researcher will have direct access and contact with both the patients required to complete the study and the institutions that have been chosen as testing sites. Using their leverage and resources will greatly improve the efficiency and outcome of this project, while adding additional validity. These three groups were chosen for their forward-thinking philosophy with regards to eating and drinking during labor and their dedication to new and groundbreaking research. Access to the population will be achieved through correspondence with the Ob/GYN Department Unit Manager of each institution describing study and requesting permission and information on how to achieve study approval. Also, the principle researcher is a Student Nurse-midwife affiliated with OUMC participating in the study, and therefore will have access to the study population.

Ethical Consideration

Consenting process
This entire study will be reviewed by the Philadelphia University Institutional Review Board (IRB) for approval. The IRB evaluates the project for risk versus anticipated benefits, informed consent, participants’ safety, privacy and confidentiality, and their rights and welfare (Polit & Beck, 2004). To insure the integrity of the study and to protect all participants, researchers and clinical facilities involved in the study, a universal consent form (Appendix B) and a release form (Appendix C) will be required of each individual, group, or institution involved. These consent and release forms will be submitted to OBGYN.NET, ASA, and SOAP for review and approval prior to being distributed to individual doctors, patients, volunteers, or institutions. All participants will be given information relating to the objective and projected benefits of eating and drinking during labor. All women who meet criteria will be invited to voluntarily participate and their confidentiality will remain intact. Each participant will be given a recruitment letter (Appendix F) and an informed consent form (Appendix B) which they are required to sign at the time of assignment into the study. Personal contact information of researchers will provided to each participant in case they have any questions or concerns. Participants may leave the study at any time.

Site review and authorization

Each site has been chosen due to its location, population served, accessibility, and reputation. If the researcher deems a site unable or unwilling to participate in the study, then a comparable alternative site that reflects the same criteria as the original site with be sought. The researcher will review for feasibility, diversity, appropriateness, and partner site participation by members of OBGYN.NET, ASA, or SOAP. After reviewing each site the researcher will obtain authorization from the sites’ ethics committees and all relevant governing authorities prior to the onset of the study.

Timeframe and Procedure
Once the principle researcher has obtained approval for the study, the first task will be to recruit one obstetrician and certified nurse-midwife to work directly with the chosen research sites. As soon as appropriate researchers are selected and approved and have consented to the study, a thorough site visit and evaluation will be completed. When approval from the clinical sites has been received, the researchers will begin to distribute the purpose forms (Appendix E), the pre-screening questionnaires (Appendix C), and the consent forms (Appendix B) to prospective participants in their care. All participants who meet the inclusion criteria and have signed consent form will be enrolled into the study and a copy of the consent form will be sent to the clinical site.
On admission, each woman will be randomly assigned base on the final digit of their medical record number; number one digit will be assigned to dietary regimen one, number two digit will be assigned to dietary regimen two, and number three digit will be assigned to dietary regimen three. Labor and delivery staff at the three clinical sites will be educated about the research study and will be trained to record data in the participant’s record. Data collected during labor and delivery will include: start of labor, active labor, nausea, vomiting, cervical exam, dietary regimen (time and amount taken), whether labor was induced, use of epidural or intravenous pain medication, method of delivery (vaginal, Cesarean), and infant apgars. If Cesarean delivery was performed, cervical dilatation and reason for Cesarean should be documented.
Once the woman has completed her labor, the researcher meets with the participant before discharge to assist with the completion of post-labor questionnaire (Appendix D) and for follow up questions. The researcher will do a chart review, evaluate all test results, view all fetal monitor tracings, and record all relevant and pertinent information on all participants who delivered at their clinical site.
At the conclusion of the institutional portion of the study, the researchers will test the correlation between what a laboring mother ingests during labor and perinatal outcomes. The principle researcher anticipates twenty-two months for completion of gathering and analyzing the date considering the sample size of nine hundred women in three different clinical sites.

Data analysis plan
The research question this study is asking is: Is there a relationship between eating and drinking during labor and perinatal outcome, including patient control and satisfaction? Data will be compiled, organized, and entered by the researchers, with the aid of a data transcriber using the Statistical Package for Social Studies software over a four-month timeframe. The study’s variables and outcomes (specific dietary regimens, perinatal outcomes, measures of control or satisfaction) will be examined by using descriptive statistics. Descriptive statistics enable researchers to summarize large amounts of data and to present quantitative descriptions in a manageable format using frequency distributions, averages, and percentages (Polit & Beck, 2004). The researchers will examine the data in terms of measures of central tendency (mean, median, and mode), distribution and standard deviations, and variance. Using descriptive statistics will provide the researchers a synopsis or summary of the data and help them determine if comparisons can be made between the study’s variables. Upon review of the data, researchers will proceed to inferential statistics if there are enough data to draw a conclusion.

Plan for disseminating findings
A prepared manuscript of disseminating findings will be published in the Journal of Midwifery & Women’s Health. The findings will also be presented to the OB/GYN board and Staff at Oklahoma University Medical Center at one of their monthly meetings.

Name one conference at which you will present your findings
To further relate the findings and broaden the scope of the research into this area, the researcher will present the manuscript at the annual meeting of the American College of Obstetrics and Gynecologists. This will allow for open discussion of the topic of women eating and drinking during labor.

Philadelphia University - IRB No.123

Appendix A:

Feasibility: Budget & Timetable

In the event that further information is needed from me (circle one response) I do/I do not consent to further contact by the researchers.

Participant Signature _________________________________ Date__________

This research was approved by the Philadelphia University Institutional Review Board on ____________.

Researcher's Signature _______________________________ Date__________
Appendix: C

Pre-Screening Survey for “Eating and Drinking during Labor”
Research study

7. What is your due date? _________________ 10. Who is your obstetrician or midwife? ______________________________ 20. Any additional information regarding your health or medical history we need to consider? ____________________________________________________________
21. Do you feel you can participate in the study “eating and drinking during labor?
1. Your Name: ________________________________________________________________________ 2. Obstetrician/Midwife Name: 8. Any food allergies?:_________________
10. When did you begin second stage of labor?: ___________________________ 14. How long from admission until the baby was born: ___________________
16. Did you have a Cesarean Section: _______________________
Appendix E
Purpose of the study & Contact Information
Dear Future Participant,

You will be participating in a formal study to examine eating and drinking during labor and perinatal outcomes. This study, which is supported by the facility, will examine issues related to patient satisfaction and the benefits of eating and drinking during labor. It will help healthcare providers provide better care for all women in labor. You may ask questions related to the study before you consent or at any time during the study.

Your participation in this study is completely voluntary and you have the right to withdraw from the study at any time before its completion without any compromise to the care you will receive. There will be no financial costs for you nor will you receive any financial compensation for your participation in this study.

Thank you very much for your time and assistance in this endeavor.

Researcher’s contact information
Kaye Sanford, SNM
580-421-1100 Home
580-421-1115 Office


American College of Nurse-Midwives. (2007). Core Competencies for Basic Midwifery Practice. Retrieved September 1, 2007 from

Blackburn, S.T. (2003). Gastrointestinal and Hepatic Systems and Perinatal Nutrition. Maternal, Fetal, & Neonatal Physiology: A Clinical Perspective (pp 426-427). St.Louis, MO: Saunders.

Chervenak, F. A., McCullough, L.B., & Birnbach, D.J. (2003). Ethics: An Essential Dimension of Clinical Obstetric Anesthesia. Anesth Analg, 96, 1480-5.

Douglas, J. (2004). General anesthesia for obstetrics: a deadly or a winning combination. Canadian Journal of Anesthesiologists, 51(6) R1-R4.

LaValley, M. (2003). Intent-to-Treat Analysis of Randomized Clinical Trials. ACR/ARHP Annual Meeting, Orlando Florida. Retrieved October 1, 2007 from

O’Sullivan, G., Liu, B., & Shennan, A. (2007). Oral Intake During Labor. Int Anesthesiol Clin, 45(1):133-147.

O’Sullivan, G. & Scrutton, M. (2003). NPO during labor. Is there any scientific validation? Anesthesiol Clin North America, 21:87-98.

Parsons, M. (2004a). Midwifery Dilemma: to fast or feed the labouring woman. Part 1: The Case for Restricting Oral Intake during Labour. Australian Journal of Midwifery, 16(4)7-13.

Parsons, M. (2004b). Midwifery Dilemma: to fast or feed the labouring woman. Part 2: The Case Supporting Oral Intake during Labour. Australian Journal of Midwifery, 17(1)5-9.

Parsons, M., Bidewell, J., & Nagy, S. (2006). Natural Eating Behavior in Latent Labor and Its Effect on Outcomes in Active Labor. J Midwifery & Women's Health, 51, e1- e6.

Parsons, M., Bidewell, J., & Griffiths, R. (2007). A comparative study of the effect of food consumption on labour and birth outcomes in Australia. Midwifery, 23, 131-138.

Polit, D. F., Beck, C. T. (2004). Nursing Research: Principles and methods (7th ed.). Philadelphia: Lippincott Williams & Wilkins.

Quenby, S., Pierce, S., Brigham, S. & Wray, S. (2004). Dysfunctional Labor and Myometrial Lactic Acidosis. Obstetric and Gynecology, 103(4), 718-23.

Scheepers, H., Thans, M., Pieter A., Gerard G.M., LeCessie, Saskia, & Kanhai, H. (2001). Eating and Drinking in Labor: The Influence of Caregiver Advice on Women’s Behavior. BIRTH, 28:2.

U.S. Department of Health and Human Services (2000). Healthy People 2010. Retrieved July 8, 2007 from

Weber, R. (2006). Comprehensive Reliability. CSE Magazine, 5, 52.
. . . . . . .