Research Regarding Labor Progress and Adverse Outcomes
As labor progresses, there is a myriad of wonderful changes and fluxes taking place. Some of the more noticeable physical changes are the baby repositioning and moving down; the contractions changing in intensity and frequency; and the cervix softening, thinning, moving forward, and dilating (opening). In hospitals, labor progress is primarily measured by the changes in cervical dilation over time. For many decades, the Friedman labor dilation curve was the tool used to determine the adequacy of cervical change; thereby, guiding the clinician’s care of the laboring patient. Friedman decided that the normal rate of dilation should be based on a statistical custom; hence, he designated the 95th percentile as the demarcation of the outer edge of normal. To gave credence to this designation, Friedman noted that the ranges of “ideal” labors matched the normal limits that he ascertained from his experiments. Subsequently, because the 95th percentile of Friedman’s patients dilated at a rate of 1.2cm/hour once active labor was established, it was thought that those who dilated at a slower rate had labor aberrations (Friedman 1955). Labor units adapted the Friedman curve and began holding laboring people to the rate of dilation of no less than 1cm/hour.
More recent data suggest that in the modern birthing population the statistically determined “normal” rate of cervical dilation is slower than Friedman described. Additionally, the rate of dilation is not uniform from one centimeter increment to the next (Zhang 2010). The table below depicts the findings of the Zhang 2010 study. The 95th percentile rate of dilation for nulliparas (people giving birth for the first time), suggests that it is quite normal for more than six hours to pass between 4cm and 5cm dilation, and for more than 3 hours to lapse between 5cm and 6cm dilation.
The new research prompted debate about study design and mathematical analysis, and whether the new data can be applied to individual labors (Cohen 2015). However, this debate ignores the fact that determining a normal rate of dilation by the 95th percentile cut off is a statistical standard, and has not been shown to correlate to a temporal labor threshold beyond which adverse outcomes begin to add up. Essentially, applying the bell curve to dilation rates, is not necessarily relevant to determining a safe length of time in labor. Subsequently, several researchers (Albers 1999; Cheng 2010; Harper 2014) carried out trials attempting to identify a point in labor where the negative outcomes begin to notably increase.
The Albers 1999 trial expressly examined low-risk births, both nulliparas (first-time birther) and multiparas (birthed before), in the care of midwifes. The study evaluated a moderate number of individuals who experienced a spontaneous onset of labor (labor that began without interventions) and presented to the hospital at 4cm dilation or less. The study excluded individuals once their membranes have been ruptured (broken bag of waters) for more than 24 hours, if they received Pitocin, epidural, or had an operative (assisted by vacuum/forceps) or cesarean birth. Methodologically, this was not the best design since people who choose to birth naturally may still end up with an operative or cesarean birth despite best efforts. Hence, it would have been meaningful to see what the outcomes would have been with that population included. Therefore, the results of this trial are applicable to people who birth without interventions and deliver via the birth canal unassisted by vacuum or forceps, which is not something one can know in advance, when trying to make decisions in labor. However, it is worth noting that the people who were excluded due to an operative or cesarean birth comprised only 12% of the original group. Additionally, excluding individuals with ruptured membranes beyond 24 hours also limits applicability of the findings to people who do not experience a prolonged rupture of membranes.
Considering the limitations, this study suggests that among people who avoid interventions, do not have prolonged rupture of membranes, and go on to birth via the birth canal unassisted by vacuum/forceps, there is no increase in adverse outcomes even among the slowest 25% of birthers when compared to faster labors. Hence, those with slower labors did not have more instances of postpartum hemorrhage (heavy blood loss after delivery), elevated adult temperature after birth, 5-minute Apgar score (a measure of the baby’s vitality) less than 7, or newborn resuscitation. The study design did not include assessment of intrapartum (during labor) adult fever in the outcomes measured; hence, whether this parameter was more frequent among this trial’s slower birthers is unknown.
In terms of labor duration, this study measured the length of time required to dilate from 4cm to 10cm. This value for the 95th percentile of first-time birthers was 17.5 hours. If this was averaged to give an approximate hourly rate of dilation, it would be 0.3cm/hour, which is four times slower than the value derived from Friedman’s experiments. Though the Albers trial did not report the 75th percentile rate of dilation, at which the cut-off for comparison was made, the rates of dilation of the participants in this trial are some of the slowest published.
The Cheng 2010 trial reports on a very large group of term, first-time birthers, who experienced a spontaneous labor onset (not induced). The authors excluded individuals whose babies displayed heartrate changes indicative of serious oxygen interruptions that required delivery by cesarean. This exclusion is odd since labor may influence the fetal oxygen reserves and result in the said heartrate changes with subsequent cesareans. However, only 2% of birthers were excluded due to cesarean births for significant changes in the fetal heartrate. Otherwise, people who delivered by cesarean for other reasons were included in the study.
Individuals with Pitocin augmentation and epidurals were included in the study. The trial evaluated specific outcomes stratified by duration of the first stage of labor, which was defined as starting with the onset of painful, regular contractions that accomplish cervical change and completing with 10cm dilation or cesarean. The outcomes were compared across three groups, one group was the fastest 5% of birthers, the next group was comprised of those whose first stage duration was in the 5% to 95% interval (the definition of normal by statistical standards), and the third group were birthers whose labor duration exceeded the 95th percentile.
The labor duration for each of the three comparison groups is as follows: among the fastest 5%, the first stage completed in less than 2.8 hours; among those with labor duration that’s considered “normal” (5th to 95th percentile), the first stage ranged from 2.8 to 30 hours; while in the group that was slower than the 95th percentile, labors lasted longer than 30 hours. The group that had the longest labors, had more frequent cesareans than those with shorter labors. Meanwhile, 75% still delivered via the birth canal with no increase in use of vacuum/forceps when compared to the 5th to 95th percentile group. Prior to adjusting for variables that may have skewed the results, the authors noted that among those with first stage longer than 30 hours, there were more cases of postpartum hemorrhage (heavy blood loss after delivery), chorioamnionitis (fever of at least 100.4˚F during birth plus an elevated fetal heartrate), endomyometritis (uterine infection after birth), neonates with 5-minute Apgar scores (measure of the baby’s vitality) less than 7, and admissions to the NICU (Neonatal Intensive Care Unit, a unit where newborns receive a higher level of care). However, once the extraneous effects of specified variables have been removed, the only outcomes that remained elevated among those with longer labors were chorioamnionitis and NICU admissions.
The authors also examined how the risk of adverse outcomes increases before the 95th percentile threshold. When the reference group was comprised of individuals whose first stage was between 2.8 and 23.8 hours (the 5th to 90th percentile), those with labors between 23.8 and 30 hours (90th to 95th percentile) had roughly 60% more cases of chorioamnionitis. When those with labors longer than 30 hours (greater than 95th percentile) were compared to the 5th to 90th percentile group, there were roughly 80% more cases of chorioamnionitis and NICU admissions.
The findings of this trial suggest that labors beyond 30 hours from the onset of painful contraction that change the cervix to complete dilation, do not result in more deliveries by vacuum/forceps; 3rd or 4th degree perineal lacerations (deep tears); postpartum hemorrhage (heavy blood loss after birth); endomyometritis (uterine infection); 5-minute Apgar scores less than 7; umbilical artery pH less than 7.0 or base excess of 12 or greater (parameters that indicate a buildup of acid in the baby’s blood due to inadequate oxygen supply); meconium aspiration (baby poop in the baby’s airways); neonatal sepsis (severe newborn infection); shoulder dystocia (baby’s shoulder getting caught on the pubic bone after the head emerges); or physical newborn birth trauma.
Conversely, there were more cesarean births with longer labors, more cases of clinical chorioamnionitis (adult fever plus fetal heartrate elevation), and babies admitted to the NICU. However, NICU admissions without a documented reason, is not a useful measure. This is because it introduces a redundancy. Until recently, infants born in the presence of clinical chorioamnionitis would be automatically admitted to the NICU for observation and tests to rule out infection, even if the newborns did not have symptoms of infection (Cheng 2010; Cotten 2015). Very few babies in this trial developed an infection, known as neonatal sepsis, even among those born when clinical chorioamnionitis was present during labor. Therefore, since the majority of babies born in the presence of clinical chorioamnionitis do not develop an infection, an automatic admission to the NICU for infants who do not have symptoms of infection is no longer recommended (Higgins 2016; Jan 2017; Jefferies 2017; Towers 2017). It is also worthwhile to note that increased chorioamnionitis among those with longer labors was in part the result of more frequent epidural use among those participants, since epidural anesthesia promotes inflammation, it increases the likelihood of developing a fever in labor (Abramovici 2014; Greenwell 2012). Additionally, the group that dilated the slowest, most likely had more numerous cervical exams, which also increase the risk of developing a fever in labor (Hastings-Tolsma 2013; Kovo 2012). Unfortunately, this trial did not report the proportion of people with prolonged rupture of membranes among the groups, which is an independent risk factor for chorioamnionitis.
The Harper 2014 study analyzed a large sample of term nulliparas (first-time birther) and multiparas (birthed before); including, low-risk and high-risk pregnancies and spontaneous and induced labors. The authors excluded anyone who delivered by cesarean prior to reaching 10cm dilation. The trial reported on the association between length of time required to dilate from 4cm to 10cm and specified adverse outcomes. The comparison groups were as follows: the reference group was comprised of the fastest 90% of birthers, with the slower-labor groups broken up into 90th to 94th percentile, 95th to 96th percentile, and those beyond the 96th percentile. For nulliparas in spontaneous labor, the fastest 90% reached complete dilation by 12 hours and 10 minutes; those in the 95th percentile required 14 hours and 25 minutes to reach complete dilation; while in the 97th percentile the duration was 16 hours and 30 minutes. If we were to calculate an approximate average hourly rate of cervical dilation, the 90th percentile was 0.5cm/hour, while the 97th percentile was 0.4cm/hour.
Those with longer labors were more likely to be induced, be obese, have diabetes, and macrosomic infants (more than 4kg or 8lbs 13oz). Those who dilated slower, were more likely to experience prolonged pushing times, develop a fever during labor, experience shoulder dystocia at delivery (baby’s shoulder getting caught on the pubic bone after the emergence of the head), and have their baby be admitted to a higher level of care. These risks increased slightly as the labor duration got longer. Conversely, cord blood pH <7 or base excess -12 or less (buildup of acidity indicating fetal oxygen deprivation during labor), 5-minute Apgar score <3, operative delivery, and heavy blood loss after delivery were unaffected by labor duration.
Some of the limitations of this study were that independent risk factors for development of fever during labor, such as epidural and prolonged rupture of membranes, were not reported. Additionally, these findings only capture the effect of longer labors in people who reached 10cm dilation; therefore, the outcomes of longer labors that result in cesarean births during the first stage are not reflected. Finally, inclusion of newborn admissions to a higher level of care without a documented reason, introduces a redundant outcome. Babies would be admitted to a higher level of care simply due to the presence of adult fever and not for symptoms of infection (Cordero 2015 ; Cotten 2015; Jenner 2018). In fact, the rate of nursery admissions mirror the percentage of adult fevers almost exactly.
Moreover, it is not surprising that shoulder dystocia was observed more frequently with longer labors in this trial. There were more individuals among the slower 10% of birthers with diabetes, obesity, and very large babies, all of which are risk factors for shoulder dystocia (Cordero 2015; Jenner 2018; Stotland 2004). Basically, longer labors do not lead to shoulder dystocia, it is just that babies who are at higher risk for shoulder dystocia are also more likely to contribute to longer labors (Said 2016). Obesity is also a risk factor for longer labors, as is induction (Harper 2014; Hautakangas 2018), both of which are overrepresented in the slowest 10% of birthers.
The authors of Cheng 2010 trial touched upon an important point. When it comes to clinical chorioamnionitis or adult fever during labor, the causal relationship between these findings and labor duration can be evaluated from a different perspective. There is evidence that puts into question the assumption that longer labors lead to chorioamnionitis. This evidence suggests that the excessive inflammation inherent in chorioamnionitis begins to develop very early in labor. Essentially, this abnormal early inflammatory response eventually results in altered uterine function; thereby, leading to longer labors (Cierny 2014; Goetzl 2010; Mark 2000).
In summation, modern evidence regarding labors that are longer than the currently accepted hospital norms, suggests that as labor length increases there is less likelihood of delivering via the birth canal and greater likelihood of developing clinical chorioamnionitis or fever during labor. Although the latter is questionable because independent risk factors like epidural, prolonged rupture of membranes, abnormal early inflammatory response, and routine cervical exams have not been ruled out. A very small percentage of term infants born in the presence of clinical chorioamnionitis will have a serious infection or neurologic injury (Conti 2015; Shi 2017), neither of which was reported at a higher rate among slower birthers. In groups of slower birthers with a higher proportion of induced labors, obesity, diabetes, and very large babies, the risk for shoulder dystocia is higher. Additionally, those with a longer first stage of labor are also more likely to experience prolonged pushing times. Essentially, there is no obvious threshold beyond which labors become undoubtedly unsafe. Moreover, it is unlikely that interventions to increase the frequency of contractions actually increase chances of delivering via the birth canal or decrease chances of a shoulder dystocia (Bugg, 2013; Hill 2016). Finally, it is not clear how effective Pitocin augmentation is at reducing instances of clinical chorioamnionitis, as there is evidence that excessive inflammation precedes protracted labors (Cierny 2014; Goetzl 2010; Mark 2000).