Time trend analysis of perinatal mortality, stillbirth, and early neonatal mortality of multiple pregnancies for each gestational week from the year 2000 to 2019: A population-based study in Japan.

Multiple pregnancies pose a high risk of morbidity and mortality in both mothers and infants; thus, obtaining reliable information based on a large population is essential to improve management. We used the maternal and child health statistics, which are published annually, from the database of the Ministry of Health, Labor, and Welfare. The data obtained were aggregated in 5-year intervals, and we used them to analyze the proportion of the number of births for each week of pregnancy to the total of each singleton and multiple pregnancy. For perinatal health indicators (perinatal mortality, stillbirth, and neonatal mortality), the obtained data were calculated and plotted on graphs for each week of pregnancy. Moreover, these indicators were calculated by dividing them into first twin and second twin fetuses. Stillbirth weights were aggregated in several groups, and a histogram was displayed. Between 2000 and 2019, there were 21,068,275 live births, 67,666 stillbirths, and 16,443 early neonatal deaths, excluding 7,148 (7,104 singletons, 44 multiple births) cases, in which the exact gestational weeks at birth were unknown. More than 95% of multiple pregnancies were twin births. Perinatal mortality, stillbirth, and early neonatal mortality rates in multiple pregnancies were the lowest at approximately 37 weeks of gestation and lower than those of single pregnancies at approximately 36 weeks of gestation. Perinatal mortality and stillbirth rates were higher during the delivery of the second twins than the first-born twins, but the early neonatal mortality rate remained approximately the same during the delivery of both twins. As the data in the government database are accumulated and published continuously, indicators can be calculated in the future using the method presented in this study. Further, our findings may be useful for policymaking related to managing multiple pregnancies.

Introduction

Multiple pregnancies have a higher risk of maternal and fetal morbidity and mortality than singleton pregnancies. In particular, complicated multiple pregnancies force obstetricians to decide the specific timing of delivery, considering both maternal and fetal risks. Obstetricians are often required to predict the ideal timing of delivery, balancing between avoiding maternal complications that increase with later gestational age (e.g., gestational hypertension, preeclampsia [1, 2], placental abruption [2], thromboembolism [2]) and preventing fetal immaturity because of preterm birth. Moreover, diseases specific to multiple pregnancies, such as twin-to-twin transfusion syndrome, usually worsen the prognosis of newborns. In general, many studies have recommended setting the timing of delivery in multiple pregnancies earlier than that in a singleton pregnancy [3-6].

The latest population-based study conducted in Japan, which determined the appropriate timing of delivery in multiple pregnancies to reduce complications owing to fetal factors, dates back to 1996 [7]. Since then, no other large-scale data analyses of perinatal health indicators (perinatal mortality, stillbirth, and early neonatal mortality rate) have been reported.

Considering the constant changes in perinatal medicine, it is essential to update perinatal health indicators by gestational weeks and time of delivery, and determine the temporal changes of indicators to improve the management of multiple pregnancies in the future. In this study, we investigated the perinatal health indicators for each gestational week of multiple pregnancies that occurred between 2000 and 2019, based on the information available on the government database, and analyzed the time trends. Additionally, we analyzed the trends of indicators in singleton pregnancies and compared them to those in multiple pregnancies.

Materials and methods

Data source

We used the maternal and child health statistics, which are published annually, from the database of the Ministry of Health, Labour and Welfare (MHLW) [8]. In Japan, the "Family Register Act" stipulates that all births, deaths, and stillbirths that occur in Japan should be reported to MHLW. MHLW aggregates data based on the notification and publishes a database annually. The data collection rate is 100%, but it is not possible to grasp the extent of illegal births, stillbirths, or abortions. Live birth and stillbirth are grouped into one category for <20 gestational weeks and ≥49 weeks, respectively, and data for 20–48 weeks are published for each gestational week. Birth weights of <500 g and ≥6,900 g are grouped into one category, and data for 500–6,900 g are published for every 100 g. Births with defects are also included. As this national database is used as an official basis for policymaking, it is considered to be the most reliable database for population studies in Japan. The online database provides the number of live births, stillbirths, and early neonatal deaths that occur in each gestational week of multiple pregnancies since 2000 [9]. Using this information, we conducted a time-lapse analysis using data of 5-year-periods to evaluate perinatal health indicators associated with singleton and multiple pregnancies that occurred between 2000 and 2019 in Japan (S1–S4 Files).

Perinatal health indicators

Perinatal mortality rate

Perinatal mortality rate is typically expressed as 1,000 × (number of stillbirths after 22 weeks of gestation + number of early neonatal deaths) / (total number of births and stillbirths). Early neonatal death is defined as death within 1 week (7 days) of live birth.

The formula for perinatal mortality rate used in this study is as follows:

Stillbirth rate

In general, the stillbirth rate is expressed as 1,000 × (number of stillbirths) / (number of live births and stillbirths).

The formula for the stillbirth rate used in this study, based on the number of gestational weeks is as follows:

Early neonatal mortality rate

Typically, the early neonatal mortality rate is expressed as 1,000 × (number of early neonatal deaths) / (number of live births). In this study, the early neonatal mortality rate according to the number of gestational weeks was defined as follows:

These indicators were calculated in the “birth-based approach” [10]. The “fetuses-at-risk approach” is considered more appropriate as an epidemiological model of fetal mortality in a population with some maternal risk factors (e.g., smoking and hypertension). If multiple pregnancies are considered to be a risk factor for the fetuses, it is more appropriate to use the “fetuses-at-risk approach.” However, in multiple pregnancies, specific issues should be considered. In cases where a fetus dies in the uterus, the gestational age at stillbirth of the deceased fetus is recorded as that at delivery of the other live infant. To use the “fetuses-at-risk approach,” it is necessary to include the gestational age at the time when the fetus died in the uterus. However, especially in cases of dichorionic twins, the standard management in twins is continuing pregnancy until the term period, even if one fetus dies in the uterus. Therefore, it is usually impossible to estimate the exact intrauterine survival time of the dead fetus from the database. From these points of view, we thought that it would be better to apply the “birth-based approach” at this stage.

Data analysis

The raw data obtained were aggregated every 5 years (2000–2004, 2005–2009, 2010–2014, and 2015–2019). We analyzed the proportion of the number of births for each week of pregnancy to the total of each singleton and multiple pregnancy. For perinatal health indicators (perinatal mortality, stillbirth, and neonatal mortality), the obtained data were plotted on graphs for each week of pregnancy without using a regression curve. Stillbirth weights were aggregated every 4 weeks until the extreme-moderate preterm (22–33 weeks), and late preterm (34–36 weeks) and term (after 37 weeks) were aggregated as one category each. A histogram was generated (S5 File).

Ethical consideration

This study only used data that were aggregated and published nationwide and were freely available online; no human participants or animals were involved. In Japan, an institutional review board approval is not required for this type of study [11]. Consequently, the requirement for ethical approval for the current study was waived.

Results

Time trends of births and perinatal health indicators for multiple pregnancies

Between 2000 and 2019, there were 21,068,275 live births, 67,666 stillbirths, and 16,443 early neonatal deaths, excluding 7,148 (7,104 singletons, 44 multiple births) cases, in which the number of gestational weeks at birth was <22 weeks or unknown. Table 1 shows the 5-year trends in the number, ratio, and perinatal health indicators of live births and multiple live births, respectively. There is a decreasing trend in the number of live births, which decreased by 18% (from 5,749,395 to 4,712,748) between 2000–2004 and 2015–2019. The number of multiple births also showed a decreasing trend, which decreased by 22% (from 121,359 to 94,018) between 2000–2004 and 2015–2019. The ratio of multiple pregnancy live births to the total live births remained unchanged at approximately 2%. The ratio of twin births to the total number of multiple births showed an increasing trend from 96.4% to 98.1%. Moreover, all perinatal health indicators showed decreasing trends in both singleton and multiple births.

Table 1

Time trends for multiple pregnancies and perinatal health indicators.

Year	2000–2004	2005–2009	2010–2014	2015–2019
Live birth (total)	5,749,395	5,406,214	5,192,770	4,712,748
Stillbirth (total)	22,345	17,672	15,287	12,362
Early neonatal death (total)	5,772	4,345	3,520	2,806
Live birth (multiple pregnancy)	121,359	116,003	99,383	94,018
Twins	117,023	113,058	97,240	92,258
Triplets or more	4,336	2,945	2,143	1,760
Multiple births in total births (‰)	2.1	2.1	1.9	2.0
Twins in multiple births (‰)	96.4	97.5	97.8	98.1
Perinatal mortality rate (‰)
Singleton	5.0	4.2	3.7	3.3
Multiple	23.6	19.2	16.8	15.0
Stillbirth rate (%)
Singleton	4.0	3.3	3.0	2.7
Multiple	16.7	14.2	13.2	11.8
Early neonatal mortality rate (%)
Singleton	1.0	0.8	0.7	0.6
Multiple	6.9	5.0	3.5	3.2

Time trends for the distribution of singleton and multiple pregnancies in each gestational week

Fig 1 shows the 5-year trend in the distribution of singleton and multiple pregnancies over each N gestational week. The highest proportion of gestational weeks in singleton pregnancies was 39 weeks, which has remained unchanged over the past 20 years. The number of deliveries at 38 weeks of gestation increased, and the proportion of deliveries after 40 weeks of gestation decreased. Conversely, the highest proportion of gestational weeks in multiple pregnancies was 37 weeks, which showed an increasing trend in the past 20 years. The next highest proportion of gestational weeks was 36 weeks, followed by 35 weeks. The proportion of deliveries after 38 weeks decreased.

Fig 1

Time trends for the distribution of singleton and multiple pregnancies based on the gestational age in weeks.

(a) Singleton, (b) Multiple.

Time trends for the perinatal mortality rate based on the gestational age

Fig 2 shows the 5-year trends of perinatal mortality rates of singleton and multiple pregnancies based on the gestational age. The same trend was observed in each 5-year-period. The perinatal mortality rates of both singleton and multiple pregnancies showed a decreasing trend between 22 and 37 weeks of gestation. In singleton pregnancies, the perinatal mortality rate was the lowest between 39 and 41 weeks of gestation, while in multiple pregnancies, the perinatal mortality rate showed an increasing trend after 38 weeks of gestation. However, the perinatal mortality rate was lower in multiple pregnancies between 22 and 36 weeks of gestation than in singleton pregnancies.

Fig 2

Perinatal mortality rates of singleton and multiple pregnancies in each gestational week.

(a) 2000–2004 years, (b) 2005–2009 years, (c) 2010–2014 years, (d) 2015–2019 years.

Time trends for stillbirth rate based on the gestational age

Fig 3 shows the 5-year trends of stillbirth rates of singleton and multiple pregnancies based on the gestational age. The same trend was observed in each 5-year-period. The stillbirth rates of both singleton and multiple pregnancies showed a decreasing trend between 22 and 37 weeks of gestation. The stillbirth rate in singleton pregnancies was the lowest between 39 and 41 weeks of gestation, while in multiple pregnancies, the stillbirth rate increased after 38 weeks of gestation. Additionally, the stillbirth rate was lower in multiple pregnancies between 22 and 36 weeks of gestation than in singleton pregnancies.

Fig 3

Stillbirth rates of singleton and multiple pregnancies in each gestational week.

(a) 2000–2004 years, (b) 2005–2009 years, (c) 2010–2014 years, (d) 2015–2019 years.

Time trends for early neonatal mortality rate based on the gestational age

Fig 4 demonstrates the 5-year trends of early neonatal mortality rates of singleton and multiple pregnancies based on the number of gestational weeks. The early neonatal mortality rates of both singleton and multiple pregnancies showed a decreasing trend between 22 and 37 weeks of gestation. In singleton pregnancies, the early neonatal mortality rate was lowest between 39 and 41 weeks of gestation, while in multiple pregnancies, the early neonatal mortality rate decreased after 38 weeks of gestation. Furthermore, the early neonatal mortality rate was similar between singleton and multiple pregnancies up to approximately 28 weeks of gestation and slightly lower in multiple pregnancies between 28 and 36 weeks of gestation.

Fig 4

Early neonatal mortality rate in each gestational week of singleton and multiple pregnancies.

(a) 2000–2004 years, (b) 2005–2009 years, (c) 2010–2014 years, (d) 2015–2019 years.

Comparison of the perinatal outcomes of the first- and second-born twins

Fig 5 shows the trends of perinatal mortality, stillbirth, and early neonatal mortality rates of twins by birth order. (S6 File) The perinatal mortality and stillbirth rates were approximately twice as high in the second-born twins compared with the first-born twins, while the difference in early neonatal mortality rates was negligible between the twins. Fig 6 shows the birth weight trends of stillborn infants born after 22 weeks of gestation. After 30 weeks of gestation, stillborn infants weighing <500 g were the most frequently observed.

Fig 5

Time trends in perinatal mortality, stillbirth, and early neonatal mortality rates of first and second twins.

(a) Perinatal mortality rate, (b) Stillbirth rate, (c) Early neonatal mortality rate.

Fig 6

Birth weight time trends of stillbirth infants.

Discussion

In this study, the perinatal health indicators of multiple pregnancies in each gestational week was calculated using data published by the government. First, the rate of multiple pregnancies has remained unchanged at approximately 2% in the past 20 years. The perinatal mortality, stillbirth, and early neonatal mortality rates were found to be the lowest at approximately 37 weeks of gestation in multiple pregnancies, and these trends were unchanged through 2000–2019. Second, the perinatal mortality and stillbirth rates were higher during the delivery of the second twins than the first-born twins, but the early neonatal mortality rate remained approximately the same in the delivery of both twins.

The strength of this study is that it is suitable for providing a temporal overview of the perinatal health indicators in Japan, as we have used the government database, which is considered reliable, albeit with a few exceptions. Furthermore, as the data will be accumulated and published in the future, the method used in this study can be used to calculate these indicators in the years to come, which also allows for the comparison of these indicators with other countries.

The limitations of this study are as follows: (1) we were unable to examine the background factors of the mothers and fetuses, gestational complications, and differences in perinatal health indicators caused by chorionicity and amnionicity during multiple pregnancies, and (2) we were unable to analyze the differences in perinatal outcomes of different modes of delivery, as well as the long-term prognoses of the newborns. Hence, the optimal timing of delivery for multiple pregnancies should not be determined by using this study’s results alone. Our results should be used only as a reference.

Trends in births and the perinatal health indicators due to multiple pregnancy

In Japan, the rate of multiple pregnancies has remained unchanged at approximately 2% in the past 20 years. This rate, which varies widely by race and nationality, was reported to be 0.3–1.0% worldwide [12-14]. From the 1980s to 2010s, the rate of multiple pregnancies increased, especially in developed countries [12-14]. It was also reported that the multiple pregnancy rate is particularly high in urban areas within the same country [15]. The increase in the number of multiple pregnancies has been attributed to two main factors: (1) the natural increase because of the increase in advanced maternal age pregnancies and (2) the artificial increase owing to the increased use of assisted reproduction technologies [16]. In the United States of America, the multiple pregnancy rate has been declining again since the 2010s [17, 18]. In Japan, the number of transferable embryos has been strictly limited by the guidelines of the Japanese Society for Reproductive Medicine in 2007, when the use of assisted reproduction technologies accounted for 1.8% of all newborns. This restriction is thought to be one of the reasons why an increase in multiple pregnancies has been prevented [19].

In this study, the perinatal mortality, stillbirth, and early neonatal mortality rates were found to be the lowest at approximately 37 weeks of gestation in multiple pregnancies. A large cohort study that was conducted in 2020 [20] reported that the short- and long-term prognoses of newborn twins were most favorable at approximately 37 weeks of gestation, which is consistent with the guidelines provided by the United States of America, United Kingdom, and Japan [3, 4, 6]. A randomized clinical trial was conducted to investigate the ideal timing of delivery of twin pregnancies, but was discontinued because of the insufficient sample size [21]. Moreover, because of the difficulty of conducting randomized control trials due to ethical concerns, we believe that population-based studies, such as the present study, will continue to be useful in the future.

In this study, the perinatal mortality, stillbirth, and early neonatal mortality rates in multiple pregnancies were lower than those in singleton pregnancies up to approximately 36 weeks of gestation. These indicators in all multiple pregnancies were approximately five times higher than those in singleton pregnancies; however, the short-term prognoses of preterm newborns were better in multiple pregnancies. Similar trends were observed in the United States of America, where the neonatal mortality rate of low-birth-weight newborns was reported to be lower in multiple than in singleton pregnancies [13]. One reason for these paradoxical mortality curves (Figs 2–4) in this study may be the fact that this study was conducted using the "birth-based approach" [10]. It has been argued this paradoxical appearance of gestational age-specific perinatal mortality curves in twins vs. singletons [22-24]. Yudkin et al. tried to explain this paradox and proposed the "fetuses-at-risk approach" formulation [25]. To use the aforementioned approach, it is necessary to include the gestational age at the time when the fetus died in the uterus. If the estimated period of death of the deceased twin fetus can be tracked on the database, it will be possible to apply the "fetuses-at-risk approach" to verify this paradox.

Another possible reason for these paradoxical mortality curves is assumed to be the fact that multiple pregnancies are often managed at large medical institutions consisting of numerous medical staff and advanced medical equipment, mainly in perinatal, maternal, and child medical centers in Japan. Approximately 50% of deliveries in Japan are handled in obstetric clinics, most of which are low-risk pregnancies. In contrast, approximately 80% of high-risk multiple pregnancies are managed in large hospitals, such as perinatal centers [26]. Therefore, early access to medical care during sudden obstetric complications, emergency illnesses and delivery of preterm infants may be one of the reasons why these indicators are low in multiple pregancies. For example, from 2015 to 2019, 79.7% (73,585/92,258) of children born from multiple pregnancies are occupied by the perinatal registration facility of the Japan Society of Obstetrics and Gynecology (JSOG), and the same tendency is observed in other age groups [26]. As all the JSOG-registered perinatal facilities (408 facilities in 2019) [26] belong to large hospitals, including perinatal medical centers, it can be assumed that the majority of multiple pregnancies were managed as high-risk pregnancies, which may have contributed to the improvement in perinatal outcomes of preterm newborns.

Perinatal outcome of the first and second twins

Perinatal mortality and stillbirth rates were higher during the delivery of the second-born twins than the first-born twins, but the early neonatal mortality rate remained approximately the same in the delivery of both twins. Some studies have reported that deliveries of the second twins had a higher perinatal mortality rate than those of the first-born twins [27, 28]; however, they did not report the discrepancy between the perinatal mortality rates of the two deliveries, which was found to be twice as much in our study. However, there was a little-to-no discrepancy in the outcomes of newborns (i.e., the early neonatal mortality rates) between singleton and multiple pregnancies. The majority of stillborn infants delivered after 34 full weeks of gestation had a birth weight of <500 g (Fig 6). In view of this, in some cases, one of the fetuses might have died before 22 weeks in the uterus, which might have been recorded as stillbirth during the delivery of both twins after 34 weeks of gestation. Such circumstances could have caused a discrepancy between stillbirth and perinatal mortality rates. In this study, the early neonatal mortality rate of second-born twins was 1.4 times higher than that of the first-born twins (3.8 vs. 2.7 per 1,000), which is close to the figures reported by a study in the United States of America [27]. Therefore, the differences in perinatal mortality and stillbirth rates between the two twins may not indicate the actual status of perinatal care in Japan.

Conclusions

In this study, the perinatal mortality, stillbirth, and early neonatal mortality rates were the lowest at approximately 37 weeks of gestation in multiple pregnancies. In twin pregnancies, the values of the perinatal health indicators were lower than those in singleton pregnancies until 36 weeks of gestation. As the analysis was conducted using data from the government database, this study is capable of overviewing the trends in the perinatal health indicators for the entire population of Japan. Furthermore, since these government data will be contiuously accumulated and published in the future, trends in these indicators can be demonstrated using the method that we presented in this study. We believe that the findings of this study may be useful as reference information for developing health policies related to the management of multiple pregnancies, which also allows for comparison of these indicators with other countries.

Live birth_multiple pregnancy.

(XLSX)

Click here for additional data file.

Distribution of gestational weeks of delivery.

(XLSX)

Click here for additional data file.

Indicators_singleton pregnancy.

(XLSX)

Click here for additional data file.

Indicators_multiple pregnancy.

(XLSX)

Click here for additional data file.

Birth weight of stillbirth_multiple pregnancy.

(XLSX)

Click here for additional data file.

Indicatotrs between first and second twins.

(XLSX)

Click here for additional data file.

7 Mar 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Alireza Abdollah Shamshirsaz

Academic Editor

PLOS ONE

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf.

2. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ.

3. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

Thanks a lot for submitting your manuscript to PLOS ONE. I am looking forward to seeing your final manuscript soon.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I read the manuscript entitled "Time trend analysis of maternal and child health indicators of multiple pregnancies

from year 2000 to 2019 for each gestational week: A population-based study in Japan" with great interest, there are, however, some concerns as follow:

Comment 1.

Lines 23-24. Since more than 95% of multiple pregnancies lead to twin births, the indicators obtained were considered to be representative of twin pregnancies.

This statement is not generalizable to all multiple pregnancies please rephrase it.

Comment 2.

Methodology of abstract is not clear or even in some parts is missing.

Comment 3.

Lines 27-28. This may be because most multiple pregnancies were managed in large-scale medical institutions.

Based on your results you cannot conclude this.

Comment 4.

Line 75. Authors have evaluated perinatal, stillbirth and early neonatal mortality. How did they cover maternal health index?

Comment 5.

There is no information on how time laps analysis was done.

Comment 6.

Start the discussion with your main findings.

Reviewer #2: PONE-D-21-26832

This population-based study examines stillbirth, neonatal and perinatal mortality rates among multiple births in Japan over the period 2000-2019. The study also examines perinatal mortality rates between first and second twin in twin births, as well as birthweight distribution by gestational age among stillbirths. The manuscript can be further strengthened by addressing the following comments.

Major Comments

1. The authors have used a ‘births-based’ approach to calculate gestational-age specific stillbirth and perinatal mortality rates, that is, the denominator equals total births (live births + stillbirths) at the specific gestational age. The authors should provide a strong justification for why this approach was used rather than the ‘fetus-at-risk’ approach. Use of the ‘births-based’ approach results in intersecting gestational-age specific mortality curves (such as those seen in Figures 2-4) in which the ’higher risk’ group appear to have lower mortality rates at earlier gestations compared to the ’lower risk’ group. It has been argued that the appropriate denominator should be those at risk of stillbirth which is delivered and undelivered fetuses surviving to the specific gestational age rather than births at the specific gestational age.

Refs:

Yudkin 1987 Lancet 1987;1:1192-4 https://doi.org/10.1016/S0140-6736(87)92154-4

Joseph & Kramer Acta Obstetricia et Gynecologica Scandinavica 2018 doi.org/10.1111/aogs.13194

Smith Am J Obstet Gynecol . 2005 Jan;192(1):17-22. doi: 10.1016/j.ajog.2004.08.014.

Other Comments

Title

2. Maternal and child health indicators is a very broad term and as this manuscript focuses exclusively on perinatal mortality, the authors should consider using more specific terms in the manuscript title.

Abstract

3. The authors should include a brief description of the study population, state the statistical methods used to analyse the data, and present some results within the abstract. For example, where there any major exclusion criteria? What was the gestational age threshold for the database? The authors should also consider presenting overall mortality rates before describing trends over time.

4. Lines 27-34 should be omitted. These are explanations which are more suitable for the discussion section than the abstract.

Introduction

5. The second sentence (lines 42-46) should be revised to improve clarity.

6. Please also include references for lines 42-46

Methods

7. For international audiences who are not familiar with the Ministry of Health, Labour and Welfare database, the authors should provide additional information about the database and the quality of the data collected. For example, what proportion of the birthing population in Japan is captured by the database? What are the criteria for inclusion in the database – what gestational age or birthweight thresholds are used? Are births with a congenital anomaly included or excluded? How valid and reliable are the data from the database? Are there any published validation studies using these data?

8. The authors should state which statistical methods were used. Have the authors considered formal tests such as linear regression for the trends over time presented?

Results

9. In the text of the results, the authors should first describe the study population before referring to Table 1. For example, present the total number of births over the study period, describe any births excluded from the analyses for whatever reasons.

10. In Table 1, please include the number of stillbirths and neonatal deaths

11. Mortality rates should be presented as per 1000 rather than per 100 (%).

12. Lines 124-128 – Perhaps this is referring to the highest proportion of births occurring at 39 weeks for singletons and 37 weeks for multiples, rather than maximum number of gestational weeks?

Discussion

13. The study limitations should be moved so that they follow on from the discussion of study strengths.

14. If the ‘fetus-at-risk’ approach is used, the discussion on intersecting mortality curves (lines 227-243) will likely need to be revised.

**********

Reviewer #1: Yes: Kamran Hessami

Reviewer #2: No

14 Apr 2022

Responses to journal requirements

We would like to thank you for your critical comments and insightful suggestions, which have helped us improve our manuscript. As indicated in the responses below, we have integrated all your comments and suggestions in the revised version of our manuscript.

[Comment]

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming.

[Response]

We have reedited the manuscript format according to the style stipulated by PLOS ONE.

[Comment]

2. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager.

[Response]

We have registered the ORCID iD of the corresponding author in the Editorial Manager.

The ID is: Masahiko Nakata 0000-0002-3621-1251

[Comment]

3. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly.

[Response]

We have added the supporting information with captions at the end of the manuscript.

Additionally, we have updated the in-text citations referring to the supporting information in the text.

Responses to Reviewer 1

We would like to thank you for your critical comments and insightful suggestions, which have helped us improve our manuscript. As indicated in the responses below, we have integrated all your comments and suggestions in the revised version of our manuscript.

[Comment]

1. Lines 23-24. Since more than 95% of multiple pregnancies lead to twin births, the indicators obtained were considered to be representative of twin pregnancies. This statement is not generalizable to all multiple pregnancies please rephrase it.

[Response]

As the reviewer pointed out, generalizing the data obtained in this study as “twin pregnancy data” is inaccurate. Therefore, in this paper, the terms are unified as “multiple pregnancies.”

[Comment]

2. Methodology of abstract is not clear or even in some parts is missing.

[Response]

Following the reviewer's advice, we have added a brief description of our methodology to the abstract. (Lines: 21–28)

[Comment]

3. Lines 27-28. This may be because most multiple pregnancies were managed in large-scale medical institutions. Based on your results you cannot conclude this.

[Response]

We acknowledge that the description in the Discussion section was insufficient. Approximately 50% of all deliveries in Japan are handled in obstetric clinics, most of which are low-risk pregnancies. On the other hand, about 80% of high-risk multiple pregnancies are managed in large hospitals such as perinatal centers. Therefore, early access to sudden obstetric complications, emergency illnesses and preterm infants may be one of the reasons why these indicators result in low values.

This point has been described in discussion (Lines: 297–309) and the concerned information from the abstract has been deleted.

[Comment]

4. Line 75. Authors have evaluated perinatal, stillbirth and early neonatal mortality. How did they cover maternal health index?

[Response]

The term we used was inaccurate. Therefore, the term "maternal and child health indicators" in original manuscript has been changed to "perinatal related health indicators (perinatal, stillbirth, and neonatal mortality)" in the revised manuscript.

[Comment]

5. There is no information on how time laps analysis was done.

[Response]

The raw data obtained were aggregated for every 5 years (2000–2004, 2005–2009, 2010–2014, and 2015–2019). We analyzed the proportion of the number of births for each week of pregnancy to the total for each singleton and multiple pregnancy. For perinatal related health indicators (perinatal mortality, stillbirth, and neonatal mortality), the obtained data were plotted on graphs for each week of pregnancy without using a regression curve. Stillbirth weights were aggregated every 4 weeks, with the exception of extreme-moderate preterm (22–33 weeks), late preterm (34–36 weeks) and term (after 37 weeks), which were aggregated as one category each. The stillbirth weights were displayed as a histogram.

We have added a "data analysis" subsection in "materials and methods" section. (Lines: 117–125)

[Comment]

6. Start the discussion with your main findings.

[Response]

In accordance with your suggestion, we have started the Discussions with our main findings.

We would like to again thank you for your constructive comments, which have highly enriched our manuscript.

Responses to Reviewer 2

We would like to thank you for your critical comments and insightful suggestions, which have helped us improve our manuscript. As indicated in the responses below, we have integrated all your comments and suggestions in the revised version of our manuscript.

[Comment]

1. The authors have used a ‘births-based’ approach to calculate gestational-age specific stillbirth and perinatal mortality rates, that is, the denominator equals total births (live births + stillbirths) at the specific gestational age. The authors should provide a strong justification for why this approach was used rather than the ‘fetus-at-risk’ approach. Use of the ‘births-based’ approach results in intersecting gestational-age specific mortality curves (such as those seen in Figures 2-4) in which the ’higher risk’ group appear to have lower mortality rates at earlier gestations compared to the ’lower risk’ group. It has been argued that the appropriate denominator should be those at risk of stillbirth which is delivered and undelivered fetuses surviving to the specific gestational age rather than births at the specific gestational age.

[Response]

We thank the reviewer for their thought-provoking advice.

As the reviewer pointed out, the paradoxical mortality curves in this study can be explained by the point that this study was conducted using the "birth-based approach." The "fetuses-at-risk approach" is considered more appropriate as an epidemiological model of fetal mortality in a population with some maternal risk factors (e.g., smoking, hypertension). If multiple pregnancies are considered to be a risk factor for the fetuses, it is more appropriate to use the "fetuses-at-risk approach." However, in the case of multiple pregnancies, specific issues should be considered. In the case where one fetus died in utero, the gestational age at stillbirth of the deceased fetus is recorded as the gestational age at delivery with the other live infant. To use the reliable "fetuses-at-risk approach," it is necessary to include the gestational age at the time the fetus died in utero. In singleton pregnancy, delivery is attempted promptly in the event of intrauterine fetal death. Therefore, a relatively accurate fetal survival time can be estimated. However, for multiple pregnancies, especially in the case of dichorionic twins, the standard management is continuing pregnancy until term period even if the one fetus dies in utero. Therefore, it is sometimes impossible to estimate the exact intrauterine survival time of dead fetus. From this point of view, we considered the "birth-based approach" better for this study.

This point was added to discussion along with the references suggested by the reviewers. (Lines: 274–292)

[Comment]

2. Maternal and child health indicators is a very broad term and as this manuscript focuses exclusively on perinatal mortality, the authors should consider using more specific terms in the manuscript title.

[Response]

Per the reviewer’s suggestion, we have changed the manuscript title to “Time trend analysis of perinatal mortality, stillbirth, and early neonatal mortality of multiple pregnancies for each gestational week from the year 2000 to 2019: A population-based study in Japan.”

[Comment]

3. The authors should include a brief description of the study population, state the statistical methods used to analyse the data, and present some results within the abstract. For example, where there any major exclusion criteria? What was the gestational age threshold for the database? The authors should also consider presenting overall mortality rates before describing trends over time.

[Response]

Following the reviewers' suggestions, we have added the brief methodology and results to the abstract. (Lines 21–37)

[Comment]

4. Lines 27-34 should be omitted. These are explanations which are more suitable for the discussion section than the abstract.

[Response]

As the reviewers suggested, we have deleted this part.

[Comment]

5. The second sentence (lines 42-46) should be revised to improve clarity.

6. Please also include references for lines 42-46

[Response]

Following the reviewers' suggestions, the description has been revised as follows, along with the corresponding references;

In particular, complicated multiple pregnancies force obstetricians to decide the specific timing of delivery, with consideration of both maternal and fetal risks. Obstetricians are often required to predict the ideal timing of delivery, balancing between avoiding maternal complications that increase with later gestational age (e.g. gestational hypertension, preeclampsia [1,2], placental abruption [2], thromboembolism [2]) and preventing fetal immaturity due to preterm birth. Moreover, diseases specific to multiple pregnancies, such as twin-to-twin transfusion syndrome, usually worsen the prognosis of newborns. (Lines: 45–52)

[Comment]

7. For international audiences who are not familiar with the Ministry of Health, Labour and Welfare database, the authors should provide additional information about the database and the quality of the data collected. For example, what proportion of the birthing population in Japan is captured by the database? What are the criteria for inclusion in the database – what gestational age or birthweight thresholds are used? Are births with a congenital anomaly included or excluded? How valid and reliable are the data from the database? Are there any published validation studies using these data?

[Response]

In Japan, the "Family Register Act" stipulates that all births, deaths, and stillbirths that occur in Japan should be reported to the Ministry of Health, Labor and Welfare (MHLW). MHLW aggregates data based on the notification and publishes a database every year. The data collection rate is 100%, but it is not possible to include illegal births, stillbirths, or abortions. Live birth and stillbirth are grouped into one category for less than 20 gestational weeks and 49 weeks or more, respectively, and data for 20 to 48 weeks are published for each gestational week. Birth weights of less than 500 g and 6,900 g or more are grouped into one category, and data for birth weights 500 to 6,900 g are published every 100 g. Births with defects are also included. Since this database is used as a national official basis for policy making, it is considered to be the most reliable database for population studies in Japan.

The study of multiple pregnancies using this database is taken as an example from JAMA 1996 by Minkami et al. (reference No.7) These points have been added to materials and methods, and the introduction. (Lines: 74–84)

[Comment]

8. The authors should state which statistical methods were used. Have the authors considered formal tests such as linear regression for the trends over time presented?

[Response]

We have added a "data analysis" subsection to the "Materials and Methods" section. (Lines: 117–125)

The raw data obtained were aggregated every 5 years (2000–2004, 2005–2009, 2010–2014, and 2015–2019). We analyzed the proportion of the number of births for each week of pregnancy to the total of each singleton and multiple pregnancy. For perinatal related health indicators (perinatal mortality, stillbirth, and neonatal mortality), the obtained data were plotted on graphs for each week of pregnancy without using a regression curve. Stillbirth weights were aggregated every 4 weeks with the exception of extreme-moderate preterm (22–33 weeks), late preterm (34–36 weeks), and term (after 37 weeks), each of which was aggregated as an individual category. This stillbirth weights were displayed as a histogram.

[Comment]

9. In the text of the results, the authors should first describe the study population before referring to Table 1. For example, present the total number of births over the study period, describe any births excluded from the analyses for whatever reasons.

[Response]

In accordance with the reviewer’s suggestion, the following description has been added at the beginning of the Results section.

Between 2000 and 2019, there were 21,068,275 live births, 67,666 stillbirths, and 16,443 early neonatal deaths, excluding 7,148 (7,104 singletons, 44 multiple births) whose number of gestational weeks at birth was less than 22 weeks or unknown. (Lines: 135-137)

[Comment]

10. In Table 1, please include the number of stillbirths and neonatal deaths.

[Response]

As the reviewers pointed out, we have added stillbirths and neonatal deaths to the table.

[Comment]

11. Mortality rates should be presented as per 1000 rather than per 100 (%).

[Response]

Per the reviewer’s comment, the text and table description have been revised.

[Comment]

12. Lines 124-128 – Perhaps this is referring to the highest proportion of births occurring at 39 weeks for singletons and 37 weeks for multiples, rather than maximum number of gestational weeks?

[Response]

In accordance with the reviewer’s suggestion, the description in the text has been revised. (Lines: 119–120)

[Comment]

13. The study limitations should be moved so that they follow on from the discussion of study strengths.

[Response]

As the reviewer pointed out, the description of limitations have been moved. (Lines:232-239)

[Comment]

14. If the ‘fetus-at-risk’ approach is used, the discussion on intersecting mortality curves (lines 227-243) will likely need to be revised.

[Response]

In this study, we have used the "birth-based approach." Additionally, we have added a description of the rationale behind our choice. (Lines: 274–292)

The paradoxical mortality curves in this study can be explained by the point that this study was conducted using the "birth-based approach." The "fetuses-at-risk approach" is considered more appropriate as an epidemiological model of fetal mortality in a population with some maternal risk factors (e.g., smoking, hypertension). If multiple pregnancies are considered to be a risk factor for the fetuses, it is more appropriate to use the "fetuses-at-risk approach." However, in the case of multiple pregnancies, specific issues should be considered. In the case where one fetus died in utero, the gestational age at stillbirth of the deceased fetus is recorded as the gestational age at delivery with the other live infant. To use the reliable "fetuses-at-risk approach," it is necessary to include the gestational age at the time the fetus died in utero. In singleton pregnancy, delivery is attempted promptly in the event of intrauterine fetal death. Therefore, a relatively accurate fetal survival time can be estimated. However, for multiple pregnancies, especially in the case of dichorionic twins, the standard management is continuing pregnancy until term period even if the one fetus dies in utero. Therefore, it is sometimes impossible to estimate the exact intrauterine survival time of dead fetus. From this point of view, we considered the "birth-based approach" better for this study.

We would like to again thank you for your constructive comments, which have greatly enriched our manuscript.

Submitted filename: Responses to journal requirements.docx

Click here for additional data file.

5 Jul 2022

Dear authors, thanks for choosing PLOS ONE to publish this important work. I also thank you for addressing the comments of the reviewers. In order to improve readability and clarity of the manuscript and thereby its scientific and practical contribution to the filed, I invite you to address the following points.

The method section needs revision. Rationale for the selection of the analytical and methodological approaches should be described in the methods section. For instance, it would increase the merit of the article if you include brief introduction of birth-based approach and the "fetuses-at-risk approach” in the methods section. Note that, what has been included in the discussion is not enough and there should be some introduction of the method in the methods section.

==============================

Please submit your revised manuscript by Aug 19 2022 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

11 Jul 2022

Responses to Academic Editor:

The authors would like to thank the Academic Editor for the constructive critique to improve the manuscript. We have made every effort to address the issues raised and to respond to all comments. The revisions are indicated in red font in the revised manuscript. We hope that our revisions meet the Academic Editor’s expectations.

[Comment]

The method section needs revision. Rationale for the selection of the analytical and methodological approaches should be described in the methods section. For instance, it would increase the merit of the article if you include brief introduction of birth-based approach and the "fetuses-at-risk approach” in the methods section. Note that, what has been included in the discussion is not enough and there should be some introduction of the method in the methods section.

[Response]

We would like to thank the Editor for the comment. As the Editor suggested, we have moved the methodology description from the Discussion to the Methods section and compressed this description in the Discussion section. The revised parts are as follows:

“These indicators were calculated in the "birth-based approach" [10]. The "fetuses-at-risk approach" is considered more appropriate as an epidemiological model of fetal mortality in a population with some maternal risk factors (e.g., smoking and hypertension). If multiple pregnancies are considered to be a risk factor for the fetuses, it is more appropriate to use the "fetuses-at-risk approach." However, in multiple pregnancies, specific issues should be considered. In cases where a fetus dies in the uterus, the gestational age at stillbirth of the deceased fetus is recorded as that at delivery of the other live infant. To use the "fetuses-at-risk approach," it is necessary to include the gestational age at the time when the fetus died in the uterus. However, especially in cases of dichorionic twins, the standard management in twins is continuing pregnancy until the term period even if one fetus dies in the uterus. Therefore, it is usually impossible to estimate the exact intrauterine survival time of the dead fetus from the database. From these points of view, we thought that it would be better to apply the "birth-based approach" at this stage.” (Lines 114–127)

“One reason for these paradoxical mortality curves (Figs 2–4) in this study may be the fact that this study was conducted using the "birth-based approach" [10]. It has been argued this paradoxical appearance of gestational age-specific perinatal mortality curves in twins vs. singletons [22-24]. Yudkin et al. tried to explain this paradox and proposed the "fetuses-at-risk approach" formulation [25]. To use the aforementioned approach, it is necessary to include the gestational age at the time when the fetus died in the uterus. If the estimated period of death of the deceased twin fetus can be tracked on the database, it will be possible to apply the "fetuses-at-risk approach" to verify this paradox.” (Lines 285–293)

We would like to thank again the Editor for the constructive comments, which have helped us significantly improve the quality of our work.

13 Jul 2022

Time trend analysis of perinatal mortality, stillbirth, and early neonatal mortality of multiple pregnancies for each gestational week from the year 2000 to 2019: A population-based study in Japan

PONE-D-21-26832R2

Dear Dr. Nasahiko Nakata,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Garumma Tolu Feyissa, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

15 Jul 2022

PONE-D-21-26832R2

Time trend analysis of perinatal mortality, stillbirth, and early neonatal mortality of multiple pregnancies for each gestational week from the year 2000 to 2019: A population-based study in Japan

Dear Dr. Nakata:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Garumma Tolu Feyissa

Academic Editor

PLOS ONE