Objective
We sought to compare Japanese nulliparous labor progression with Friedman’s classic 1955 curve and Zhang’s 2002 curve.
Study Design
We developed a labor curve using retrospective record reviews of 2369 Japanese nulliparas, at term, spontaneous labor onset and singleton vertex deliveries of normal birth weight infants.
Results
The new Japanese Suzuki-Horiuchi labor curve with slower cervical dilation in the active phase was like Zhang’s and differed from Friedman’s curve. Labor length was approximately 5 hours occurring between 4-10 cm compared with Friedman’s 2.5 hours and Zhang’s 5.5 hours. Even at 10-cm dilation, labor lasted >2 hours at the 95th percentile of time interval.
Conclusion
Similar to Zhang’s curve, the Suzuki-Horiuchi curve was smooth and more gradually sloped than Friedman’s curve. Appraise “arrested or protracted labor” with these slower labor curves in mind using Friedman’s curve cautiously.
Friedman’s curve continues to appear as the definitive index within industrialized countries and is found in midwifery and maternal-infant textbooks as part of labor management for gauging the advancement of normal labor. It is widely used as an assessment tool during clinical training of midwives to teach them how to determine the advancement of labor. Friedman’s classic graphic labor curves for labor progress prediction were established in 1955 based on a US population, but labor management methods have changed since the 1950s. Mid forceps and low forceps deliveries commonly performed in the 1950s are no longer practiced. Furthermore, the characteristics of pregnant women have changed, for example, there is an increase in the number of older nulliparas plus nutritional status has improved. Consequently research has commenced examining the extent to which the Friedman curve is still a valid way of determining advancing labor, given changes in women’s age, lifestyle, advances in medical technology, and other factors.
In 1999, Albers, in the United States, investigated 1473 Americans with varied ethnic origins: American Indian, Hispanic, and non-Hispanic primiparas. Albers found the active labor phase to be longer, approximately 7.7 hours compared with Friedman’s average of 4.9 hours. Albers presented only numerical data; he did not publish a labor curve. In the retrospective investigation of 1162 nulliparous client records from 1992 through 1996 by Zhang et al, the acceleration phase was slower than reported by Friedman. Although new curve variables were developed, such as the careful definition of start times by Hendricks et al, and extension of the work of Hendricks et al by Philpott and Castle, and the marked changes in fetal and maternal conditions in recent decades, no new valid curves have been developed to supplant the Friedman curve. Likewise, in Japan, there were no surveillance studies undertaken to reexamine the Friedman curve.
Despite the lack of specific evidence pertaining to Japanese nulliparas, midwifery clinicians question the usefulness of the Friedman curve. For example, the sample of 89 nurse managers from the North American continent by Cesario observed that there was a wider range of normal than traditionally accepted and that the second stage of labor could safely extend beyond the established 2-hour limit often imposed. Their conclusion was that the Friedman curve should be modified.
The creation of a valid and reliable labor curve based on the biological and demographic data of Japanese women is necessary for an evidence-based approach for a process that has, in the past, relied heavily on the expert clinical experience of Japanese midwives as they observed the limitations of the Friedman curve as a predictive tool. Therefore, the purpose of our research was to document the progress of nulliparous cervical dilation during normal delivery conditions and to create a cervical dilation model curve for comparison with both Friedman and Zhang curves.
Materials and Methods
Design and subjects
The study design was a retrospective review of medical records and development of a labor curve. From January 2001 through December 2005, data were collected at 4 primary medical institutions performing normal deliveries and that agreed to participate in the study. Selected were all 2369 cases, meeting the inclusion criteria, from the total of 16,423 deliveries. Subjects’ medical records were posted from the delivery ledger by a single researcher. A trained staff person checked for posting errors.
Inclusion criteria were Japanese, singleton nulliparous pregnancy, gestational age between 37 weeks 0 days–41 weeks 6 days, spontaneous onset of labor, vertex presentation at admission, cervical dilation <7 cm at admission, and duration of labor from admission to delivery >3 hours. We excluded those cases having cesarean delivery, labor induction, and epidural anesthesia.
Friedman investigated 500 primiparas and performed cervical dilation measurements at least once per hour, thereby establishing data collection consistency. However in this research, the midwife in 1 hospital and 1 clinic performed a cervical dilation measurement every hour from the time of admission as directed by the institution care standards. However, in the other 2 hospitals, cervical dilation measurements were guided only by events: admission, membrane rupture, intensive contractions, fetal heart rate change, and breathing changes or pushing.
To investigate the pattern of labor advance by seeing the relation between labor slowing or lapsed time and cervical dilation, a smoothing B-spline curve was created on the basis of the fully dilated cervix of 10 cm. We took the same approach as Zhang et al. Patients were admitted at various points of cervical dilation but all ended at 10 cm. Therefore a regression was carried out in a reverse approach, with 10 cm as the starting point and going backward to 0 cm. A 10th-order polynomial provided the best fit for the time and dilation values. The statistical analysis software JMP 6.0.0 (SAS Institute Inc, Cary, NC) was used. The significance level was set at 5%. The institutional research ethics screening committee granted approval for this study. Researchers maintained subject anonymity.
Results
Subjects’ characteristics
All subjects (n = 2369) were Japanese; the mean age at the time of delivery was 28.4 years (SD 4.5; range, 16–44). Average height was 158.7 cm (SD 5.2); average weight was 61.7 kg (SD 7.6). Before pregnancy the mean body mass index was 20.2 (SD 2.4) and the mean weight was 51.1 (SD 6.9). The mean number of gestational weeks at delivery was 39.8 (SD 0.9). The average amount of intrapartum blood loss was 322.4 mL (SD 199.8; median, 268; range, 42–1650). Newborn characteristics were average weight, 3036.3 g (SD 311.6; range, 2104–4210). Apgar score ≥8 was 98.5% >1 minute and 100% >5 minutes.
The cervical dilatation at the time of admission was 2.9 ± 1.6 cm. The first stage of labor was 12.3 (SD 7.7) hours. Midwives performed an average of 5 pelvic examinations from admission to the second stage of labor (range, 2–16; median, 4).
Comparison of demographic and medical variables
We differed from Friedman in relation to the delivery age composition ( Table 1 ). In our sample, only 56.1% were in the 20- to 30-year-old age range. In Friedman’s sample, 70.6% were 20-30 years old. In Japan, the trend is toward delaying the birth of the first child. In 2003, the average age of the mother in our study was 28.6 years compared to 24.4 years in the 1950s study (Health and Welfare Statistics Association).
| Variable | Current study | Friedman study | Zhang study |
|---|---|---|---|
| Year of data collection | 2001-2005 | Early 1950s | 1992-1996 |
| Candidate ethnic group | 2369 nulliparous Japanese | 500 primigravid Americans |
|
| Data-gathering process | Secondary analysis of medical records | Direct assessment of cervical dilation | Secondary analysis of medical records |
| Pelvic examination frequency | Irregular | >1 h | Irregular |
| No. of pelvic examinations | 5 (2–16) | na | 6 |
| Average age, y (range) | 28 (16–44) | na (13–42) | 23 (18–34) |
| Age 20-30 y, % | 56.1 | 70.6 | – |
| Average birthweight, g (range) | 3036 (2104–4210) | na (2080–4710) | na (2500–4000) |
| Birthweight 2500-4000 g, % | 97.8 | 84.5 | 100 |
| Labor induction, % | 0 | 4.4 | na |
| Oxytocin augmentation, % | 6.5 | 9.4 | 50 |
| Anesthesia delivery, % | 0 | 8.4 | 48 |
| Vacuum/low forceps, % | 3.5 | 51.2 | 13 |
| Mid forceps or cesarean delivery, % | 0 | 5.6 | na |
| Cervical dilation at admission, cm | 3.0 | na | 3.5 |
| Duration of first stage of labor, h, mean (SD) | 12.3 (7.7) | 13.3 (7.6) | 7.3 (3.3–13.7) a |
Rather large differences were also seen in delivery interventions across the decades. Vacuum and forceps deliveries were 3.5% in this research compared to Friedman’s results of 51.2% and Zhang’s results of 13%. Furthermore, oxytocin augmentation was given to 6.5% of the nulliparas in this research compared to 9.4% of the primigravida in Friedman’s sample and 50% of nulliparas in Zhang’s sample. Anesthesia was given to 8.4% in Friedman’s study and 48% in Zhang’s sample while this study excluded cases receiving anesthesia. We found the median cervical dilation at the time of hospitalization to be 3.0 cm while Zhang found the median to be 3.5 cm.
Comparison of curve construction
When Friedman created a curve, he did so by analyzing smaller sections of the labor graph, but since the width of these sections was unknown, calculating the inclination of a curve was impossible. For this reason, to compare our findings with the Friedman curve, we set up lambdas equivalent to this width, and created a smoothing spline curve that corrected for the roughness measure resulting from the action of delivery progress ( Figure 1 ). Since individual onset of cervical dilation varied widely, we used 10 cm of cervical dilation as the base. We considered it as 0 hour and created the smoothing spline curve. With 10 cm of cervical dilation set to 0, the X-axis was set as the minus time (hours), and the Y-axis was set as cervical dilation (cm), creating a scatter plot. As a result, the graph is reversed going left to right, compared with Friedman’s curve ( Figure 1 ).
