Secondary Logo

Journal Logo

Association of Obstetric Intervention With Temporal Patterns of Childbirth

Clark, Steven L. MD; Perlin, Jonathan B. MD, PhD; Fraker, Sarah MS; Bush, Jamee MBC; Meyers, Janet A. RN; Frye, Donna R. RN; Garthwaite, Thomas L. MD

doi: 10.1097/AOG.0000000000000485
Contents: Original Research
Free

OBJECTIVE: To examine the gestational age at and day and time of delivery in current U.S. obstetric practice.

METHODS: We examined electronic records from 72 hospitals in 16 states during a 4-month period during 2013. Day of week of delivery, time of day of delivery, and route of delivery were examined in various subpopulations delivering both with and without obstetric intervention.

RESULTS: Records of 21,381 women were examined. The distribution curve for gestational age at delivery peaked at 39 weeks both for the entire group and for the subgroup entering labor spontaneously and delivering vaginally without augmentation. Statistical modeling suggests that medical intervention accounts for only a fraction of the shift in distribution peak from 40 to 39 weeks of gestation. Three temporal peaks of total and primary cesarean delivery were seen on weekdays, corresponding to immediate preclinic, lunch time, and immediate postclinic timeframes. These peaks were not seen on weekend days. The risk of nonelective primary cesarean delivery during a weekday was approximately one third higher than on a weekend (relative rate 1.36, confidence interval 1.24–1.49).

CONCLUSION: The recently described shift in peak distribution of U.S. gestational age at delivery from 40 to 39 weeks of gestation may reflect an underlying physiologic change in the U.S. population and is not exclusively related to obstetric intervention. During the work week, factors other than medical necessity appear to have a marked association with both timing of delivery and rate of cesarean delivery and may affect up to one third of primary cesarean deliveries.

Factors other than medical necessity appear to significantly affect the temporal pattern of childbirth in the United States.

Hospital Corporation of America, Nashville, Tennessee.

Corresponding author: Steven L. Clark, MD, PO Box 404, Twin Bridges, MT 59754; e-mail: slclark@bcm.edu.

Financial Disclosure The authors did not report any potential conflicts of interest.

The timing of spontaneous labor and delivery in humans has long been the subject of both medical and anthropologic inquiry.1–3 In recent decades, such data have been increasingly obscured by a progressive trend toward obstetric intervention in the forms of labor induction, labor augmentation, and operative delivery.4,5 Such interventions arise both from concerns for the well-being of mother and child and from less well-defined social factors, including convenience for both mother and clinician.

We undertook a study examining the day and time of delivery in current obstetric practice. Our goal was to investigate these factors both in spontaneous labor followed by spontaneous delivery and in deliveries potentially affected by medical or surgical intervention. Such data provide insight into the degree to which factors other than underlying physiology and medical necessity affect the timing and route of delivery.

Back to Top | Article Outline

MATERIALS AND METHODS

We examined birth data from 72 nonuniversity hospitals affiliated with the Hospital Corporation of America over a 4-month period, May 1 to August 31, 2013. These facilities had annual delivery volumes ranging from 168 to 4,568 and are located in 16 states. From geographic, demographic, and socioeconomic standpoints, our population is representative of the U.S. population as a whole.6–8 Data were extracted from a system-specific modification of the General Electric Centricity Perinatal electronic nursing record, previously validated both by the developer and the Hospital Corporation of America. Data were originally entered by the nurse caring for the patient. Our analysis included all term (37 weeks of gestation or greater) live-born neonates. Gestational age was determined from the physician-generated final estimated date of confinement field in the prenatal record.

Statistical analysis of our data was conducted using the χ2 test with examination of standardized residuals and a two-sample t test using a threshold of P<.05 for significance (statistical software R). In analyzing the peaks in the graphs of delivery times, a threshold of P<.01 was used.

To further evaluate the effects of the practice of elective delivery beyond 39 weeks of gestation on the distribution of births in the population, we constructed a hypothetical model in which a normally distributed population of similar size with a peak at 40 weeks of gestation is depleted by random elimination of 25% or 50% of data points beyond 39 weeks of gestation. The distribution of the simulated deliveries was generated from a skew normal distribution with a location parameter of 41, a scale parameter of 2, and a shape parameter of −3. Observations were then removed if they were under 37 weeks of gestation to only include term births.

Because this project involved only examination of deidentified aggregate data for quality improvement purposes, it was exempt from institutional review board approval based on 45CFR46.101(b)2 and 46.102(f) as well as 45CFR164.514(a)–(c) of the Health Insurance Portability and Accountability Act.

Back to Top | Article Outline

RESULTS

During this 4-month period, 21,381 women delivered in study facilities. A total of 14,237 (66.6%) had a vaginal delivery, and 7,095 (33.2%) had a cesarean delivery. Data regarding mode of delivery were not available for 49 women (0.2%.). Table 1 represents parity distribution of this population before the index delivery. A total of 5,460 (25.5%) women entered labor spontaneously and had a vaginal delivery without augmentation and are referenced as “spontaneous deliveries” in the text and figures.

Table 1

Table 1

Figure 1 represents gestational age at delivery both for the population as a whole and for the subset of women who entered labor spontaneously and delivered vaginally without augmentation. The distribution curve for gestational age at delivery peaked at 39 weeks both for the entire group and for the subgroup. These distributions were significantly different from one another (P<.001). Figures 2–5 represent day and time of birth for various types of deliveries and the relationship between the two. Figure 6 represents the time of birth for women who entered labor spontaneously and had a vaginal delivery without augmentation. During the weekdays, the distribution of gestational ages for all deliveries and for spontaneous vaginal deliveries was significantly different (P<.001). Three temporal peaks of total and primary cesarean delivery were seen on weekdays, corresponding to immediate preclinic, lunch time, and immediate postclinic timeframes. However, there was no such difference in delivery time distribution on weekends (P=.63). Figure 7 represents the relationship between the day of week and primary cesarean delivery rate in patients who were nulliparous before the index pregnancy. There was a significantly higher rate of primary cesarean delivery in nulliparous women delivering on weekdays as opposed to weekends (33.6% compared with 24.7%, relative rate 1.36, 95% confidence interval 1.24–1.49).

Fig. 1

Fig. 1

Fig. 2

Fig. 2

Fig. 3

Fig. 3

Fig. 4

Fig. 4

Fig. 5

Fig. 5

Fig. 6

Fig. 6

Fig. 7

Fig. 7

Table 2 and Figure 8 demonstrate the effects on an initially normally distributed population of planned delivery of 25% and 50% of women beyond 39 weeks of gestation.

Table 2

Table 2

Fig. 8

Fig. 8

Back to Top | Article Outline

DISCUSSION

Available data suggest a role for the fetal hypothalamic–pituitary–adrenal system in the initiation of spontaneous labor as well as a physiologic response to melatonin secretion in the regulation of natural labor onset.1,9–13 Older studies suggested that natural labor onset peaked around midnight and that nocturnal labor is significantly shorter than diurnal labor.1,14 This predominantly nocturnal pattern of birth in humans and other primates has been seen as an advantageous evolutionary adaptation to reduced predation and the availability of group social support at night.1 However, more recent data suggest that the historic nocturnal pattern of birth has largely been reversed with most deliveries occurring during the day; unfortunately, such data are often confounded by the temporal effects of medical intervention.1,2

In a similar manner, recent data document a shift in the peak of the distribution curve of human birth from 40 to 39 weeks of gestation, a change generally attributed to an increased tendency for artificial intervention in the birthing process.4,5

Although such intervention certainly plays a role, our data (Fig. 1) document a similar shift in women who enter labor spontaneously. This shift could be the result of depletion of the population of women destined to enter spontaneous labor after 39 weeks of gestation by the common practices of earlier labor induction or repeat cesarean delivery. An alternative explanation would involve an actual physiologic shift in the mean gestational age at which women enter labor spontaneously; such changes are well described with respect to other physiologic parameters such as age at menarche, birth weight, and mean adult height and are primarily attributed to enhanced nutrition.15,16 Indeed, the observed increase in mean birth weight of term neonates and the known effect of uterine distention on labor onset would provide physiologic plausibility to this interpretation.17

To further investigate these possibilities, we examined a hypothetical model in which an initially normally distributed population of similar size with a peak at 40 weeks of gestation is depleted by random elimination of data points beyond 39 weeks of gestation.

As demonstrated in Table 2 and Figure 8, selective depletion of more than 50% of women who were destined to enter labor spontaneously beyond 39 weeks of gestation would be necessary to achieve a distribution similar to that seen in our actual population of women who entered labor and delivered spontaneously. From a clinical standpoint, this would require planned delivery at or before 39 weeks of gestation in more than 50% of all women who were originally destined to enter labor beyond 39 weeks of gestation. However, in our population, a total of 40.7% of women actually delivered at or before 39 weeks of gestation and only a fraction of these women would be in the “artificial delivery” group with the majority entering labor spontaneously at 37–39 weeks of gestation. These data as well as differing degrees of skewness between this theoretical population and our observed population (0.14 compared with 0.03; Table 2; Fig. 8) suggest that medical intervention accounts for only a portion of this shift and that the physiologic mean length of gestation in this U.S. population today may actually be closer to 39 rather than 40 weeks of gestation. It is also interesting to note a growing body of data documenting a nadir in adverse pregnancy outcomes at 39 rather than 40 weeks of gestation; from an evolutionary perspective, it is difficult to imagine a disconnect between the most common gestational age for the onset of spontaneous labor and the optimal gestational age for delivery. It is also possible that this shift is related to better ultrasound dating and that the time-honored physiologic distribution curve for delivery never actually did peak at 40 weeks of gestation. Unfortunately, direct definitive confirmation or refutation of this interpretation will be difficult, because virtually any population in which sufficiently accurate medical records exist to examine this question in detail will also have a significant rate of artificial intervention. In any event, these data support a contention that in a well-dated U.S. population, “full term” pregnancy may today be better represented by 39 rather than 40 weeks of gestation.18–20Figure 1 also suggests that the concept of a 42-week postterm pregnancy is today primarily of historical interest.

As demonstrated in Figure 6, the historic predominantly nocturnal birth pattern also no longer obtains even in women entering labor spontaneously and delivering vaginally without augmentation. Like with changes in age at menarche and mean adult height, environmental factors such as nutrition, workforce participation, maternal stress, and disruption of the traditional cycles of day and night by differential physical activity and artificial light may have altered the environmental triggers necessary for evolutionary expression of a genetic preference for nighttime birth.21Figure 2 demonstrates a spike in births corresponding to immediate preclinic hours, lunch time, and immediate postclinic hours, Monday through Friday; no such pattern is seen on weekends. Figures 4 and 5 demonstrate these spikes to be principally related to cesarean delivery. These spikes were statistically significant both for all births and for cesarean deliveries (P<.01). Importantly, none of these patterns persist on weekends.

The preoffice hours spike in cesarean deliveries is, to a significant degree, related to scheduled repeat cesarean deliveries and as such is not of particular concern from a quality standpoint. However, the lunch time and immediate postoffice hours spikes in overall cesarean deliveries and the spikes seen at all three times for primary cesarean deliveries are of concern because labor arrest disorders are the most common indication for primary cesarean delivery and recent data suggest that a primary driver of such indications is failure to allow sufficient time before making the diagnosis of first- and second-stage labor arrest.22–24 The conclusion that decisions regarding primary cesarean delivery may be influenced by nonmedical factors is confirmed in Figure 8, which demonstrates a statistically significant increase in the likelihood of primary cesarean delivery on weekdays as compared with weekends. Only 1.5% of primary cesarean deliveries in our system are elective (patient choice).25 If one assumes that the rate of weekend primary cesarean deliveries more accurately reflects decisions based on medical necessity without the influence of weekday practice considerations, Figure 7 suggests that approximately one third of primary cesarean deliveries in nulliparous women might be avoided without compromising patient safety simply by altering the current system of intrapartum care to make everyday a weekend day in terms of physician convenience.

Figure 6 demonstrates a time-of-birth curve for spontaneous labor similar to that seen in Figure 2 for all labors on weekend days. Deviations from the pattern of birth seen in Figure 6 and on weekends in Figure 2 represent the effects of obstetric intervention; it is instructive regarding the extent of such intervention that only approximately one fourth of all women today enter labor spontaneously and progress without augmentation to vaginal delivery.

The well-documented representative nature of both demographics and practice within our hospital system and a system-wide cesarean delivery rate at the national average suggest that these observations are not unique to our system but apply generally to obstetric practice in the United States.6–8,22,25 Indeed, recent data from our system demonstrate that cesarean delivery rates may be reduced with a hospitalist model of obstetric care in which delivery decisions are unencumbered by personal, sleep, or office practice considerations.26,27 In contrast, the current approach to care in the United States generally reflects a system in which the need to perform a delivery almost always represents a distraction from some other personal or professional activity. In this respect, the observation that every system is perfectly designed to achieve exactly the results it is achieving would seem to apply.28 Attempts to reduce the cesarean delivery rate by changing engrained modes of individual physician practice are important. These include proper interpretation of fetal heart rate tracings and appropriate management of labor arrest. However, without the type of systems changes recommended by the Institute of Medicine, such changes are by themselves likely to prove as ineffective in the future as they have in the past.29,30

Back to Top | Article Outline

REFERENCES

1. Bernis C, Varea C. Hour of birth and birth assistance: from a primate to a medicalized pattern? Am J Hum Biol 2012;24:14–21.
2. Mancuso PJ, Alexander JM, McInitre DD, Davis E, Burke G, Leveno KJ. Timing of birth after spontaneous onset of labor. Obstet Gynecol 2004;103:653–6.
3. Rosenberg KR. The evolution of modern human childbirth. Am J Phys Anthropol 1992;35:89–124.
4. Centers for Disease Control and Prevention, Data and Statistics. Birth by gestational age. Available at: http://www.cdc.gov/DataStatistics/2007/births/. Retrieved May 27, 2014.
5. Center for HEALTH STATISTICS, Minnesota Department of Health. Gestational age at birth, 1994 and 2004. Minnesota VitalSigns 2006;2:1–6. Available at: http://www.health.state.mn.us/divs/chs/vitalsigns/052006GestationalAge.pdf. Retrieved May 27, 2014.
6. Clark SL, Belfort MA, Hankins GD, Meyers JA, Houser FM. Variation in the rates of operative delivery in the United States. Am J Obstet Gynecol 2007;196:526.e1–5.
7. Foley MR. Reestablishing trust in the medical profession: making a significant impact on maternal mortality in the United States. Am J Obstet Gynecol 2014;211:1–2.
8. Clark SL, Christmas JT, Frye DR, Meyers JA, Perlin JB. Maternal mortality in the U.S.—predictability and the impact of protocols on fatal post-cesarean pulmonary embolism and hypertension-related intracranial hemorrhage. Am J Obstet Gynecol 2014;211:32.e1–9.
9. Ankel-Simons F, Rasmussen DT. Diurnality, nocturnality, and the evolution of primate visual systems. Am J Phys Anthropol 2008;(suppl 47):100–17.
10. Baake BA. Circadian variation in the observed duration of labor. Possible causes and implications. Acta Obstet Gynecol Scand 1991;70:465–8.
11. Glattre E, Bjerkedal T. The 24-hour rhythmicity of birth. A population study. Acta Obstet Gynecol Scand 1983;62:31–6.
12. Honnebier MB, Nathanielsz PW. Primate parturition and the role of maternal circadian system. Eur J Obstet Gynecol Reprod Biol 1994;55:193–203.
13. Sharkey JT, Puttaramu R, Word RA, Olcese J. Melatonin synergizes with oxytocin to enhance contractility of human myometrial smooth muscle cells. J Clin Endocrinol Metab 2009;94:421–2.
14. Jolly A. Hour of births in primates and man. Folia Primatol (Basel) 1972;18:108–21.
15. Chumlee WC, Shubert CM, Roche AF, Kulin HE, Lee PA, Himes JH, et al.. Age at menarche and racial comparisons in US girls. Pediatrics 2003;111:110–3.
16. QuickStats: mean weight and height among adults aged 20–74 years, by sex and survey period—United States, 1960–2002. MMWR 2005;54:7714.
17. Manabe Y, Sakaguchi M, Mori T. Distention of the uterus activates its multiple pacemakers and induces their coordination Gynecol Obstet Invest 1994;38:163–8.
18. Bailit JL, Grobman W, Zhao Y, Wapner RJ, Reddy UM, Varner MW, et al.. Non-medically indicated inductions vs expectant management in term nulliparas. Am J Obstet Gynecol 2014 [Epub ahead of print].
19. Gibson KS, Waters TP, Bailit JL. Maternal and neonatal outcomes in electively induced low-risk term pregnancies. Am J Obstet Gynecol 2014 [Epub ahead of print].
20. Rosenstein MG, Snowden JM, Cheng YW, Caughey AB. The mortality risk of expectant management compared with delivery stratified by gestational age and race and ethnicity. Am J Obstet Gynecol 2014 [Epub ahead of print].
21. York TP, Lindon JE, Neale MC, Strauss JF III. The contribution of genetic and environmental factors to the duration of pregnancy. Am J Obstet Gynecol 2014;210:398–405.
22. Rouse DJ, Owen J, Hauth JC. Active phase arrest: oxytocin augmentation for at least 4 hours. Obstet Gynecol 1999;93:323–8.
23. Rouse DJ, Owen J, Savage KG, Hauth JC. Active phase labor arrest: revisiting the 2-hour minimum. Obstet Gynecol 2001;98:550–4.
24. American College of Obstetricians and Gynecologists; Society for Maternal-Fetal Medicine; Caughey AB, Cahill AG, Guise JM, Rouse DJ. Safe prevention of the primary cesarean delivery. Am J Obstet Gynecol 2014;210:179–93.
25. Clark SL, Miller DD, Belfort MA, Dildy GA, Frye DK, Meyers JA. Neonatal and maternal outcomes associated with elective term delivery. Am J Obstet Gynecol 2009;200:156.e1–4.
26. Iriye BK, Huang WH, Condon J, Hancock L, Hancock JK, Ghamsary M, et al.. Implementation of a laborist program and evaluation of the effect upon cesarean delivery. Am J Obstet Gynecol 2013;209:251.e1–6.
27. Clark SL. Sleep deprivation: implications for obstetric practice in the United States. Am J Obstet Gynecol 2009;201:136.e1–4.
28. Carr S. A quotation with a life of its own. Available at: http://www.psqh.com/julaug08/editor.html. Retrieved May 15, 2014.
29. Institute of Medicine. To err is human—building a safer health system. Washington, DC: National Academies Press; 2000.
30. Institute of Medicine. Crossing the quality chasm—a new health care system for the 21st century. Washington, DC: National Academies Press; 2001.
Figure

Figure

© 2014 by The American College of Obstetricians and Gynecologists.