INTRODUCTION

To date, there is no standardized consensus definition of miscarriage or preterm birth. Definitions differ from country to country, and vary by upper gestational age and pregnancy viability.

Miscarriage, or early pregnancy loss, is most often defined as pregnancy loss prior to viability. However, many miscarriages occur prior to a formal diagnosis of pregnancy, and these are often not included in official statistics. Estimating incidence of miscarriage is limited by the lack of centralized databases across geographic regions. Viability is also a ‘moving target’ and differs from a legal and clinical perspective. The WHO defines miscarriage as prior to 28 weeks' gestation,¹ whereas in the US the American Society of Reproductive Medicine uses a lower viability threshold of 20 weeks.^2,3 In contrast, the Royal College of Obstetricians and Gynaecologists in the UK defines miscarriage as pregnancy loss under 24 completed weeks of gestation.⁴ This variation in the definition of viability is reflective of countries’ vital statistics and socioeconomic demographics.

The variation in the ‘end point’ of miscarriage follows through to the difficulty in defining preterm birth. Preterm birth is defined as delivery prior to completion of 37 weeks' gestation.⁵ However, without a consensus definition on what viability is, there can be no clear definition for the earliest gestation age that is considered preterm birth versus a miscarriage.

Throughout this chapter, the terms miscarriage and early pregnancy loss are used interchangeably, although there is variation on what early pregnancy loss refers to, with some sources using it specifically to refer to first-trimester losses.⁶

WHAT IS THE PROBLEM, AND WHY DOES IT MATTER?

These unclear and non-standardized definitions make it difficult to estimate true incidence of miscarriage worldwide.

Incidence of miscarriage

Miscarriage is difficult to assess accurately for a variety of reasons, not least of which is stigma within societies and the fact that many early miscarriages occur prior to establishment of formal pregnancy confirmation or care.

Early miscarriages occurring prior to the establishment of formal pregnancy care make it difficult to assess the true burden of disease. The LIFE study was a US cohort study of perinatal patients attempting conception, and utilizing home pregnancy tests.⁷ This dataset allowed estimates of pregnancy loss in a population prior to establishment of care. The miscarriage rate amongst this group was 28%. In developing countries, there have been few studies. A preconception cohort of female textile workers in Anhui, China, recruited 518 women and followed their conception attempts over 1 year, and reported a pregnancy loss rate of 32.5% (which included chemical pregnancy loss).⁸ The Rajarata Pregnancy Cohort in Sri Lanka was a pregnancy cohort, which recruited over 90% women in the first trimester of pregnancy in the defined area during the study period.⁹ They made specific note of the difficulty in accounting for pregnancy loss prior to the establishment of care, commenting that ‘3830 pregnancies were registered with the field staff, but 143 women had miscarriages within 2 weeks before the baseline assessments at the clinic. This number accounted for 31% of all miscarriages in the cohort and was missing from the public clinic records’.¹⁰ These numbers differ from the largest pooled estimate of miscarriage risk currently published, which quotes miscarriage risk at 15%, based on patient self-reported outcomes and record linkage.¹¹

The incidence of miscarriage is estimated differently in different countries, and may also vary depending on the population studied. Quenby et al. estimated the incidence of miscarriage at 15%, based on data from nine observational studies.¹¹ However, this is based on data gathered only in Europe and North America, which limits generalizability. Agampodi et al. again comment on this dearth of data, and its implications.¹⁰ Particularly in countries with poorer access to care, there may be delayed presentation for initial prenatal care and confirmation of pregnancy and/or loss. Furthermore, there are valid biological hypotheses to suggest that miscarriage rates could vary across populations. Maternal age at conception, environmental exposures (smoking, alcohol, caffeine), prenatal vitamin use, are all associated with miscarriage risk and vary between populations and countries. Data such as this is imperative if we are to truly estimate the burden of early pregnancy loss worldwide.

Finally, miscarriage stigma is significant and may lead to underreporting of miscarriage. In many communities, miscarriage is seen as the mother’s fault, or that she is to blame due to something she did or did not do during the pregnancy. Even when there is less blame surrounding miscarriage, there is an unwillingness to discuss miscarriage – people are often uncomfortable discussing loss, and may be told that it was common, early, or the effect of the loss may be minimized. This contributes to a ‘culture of silence’ surrounding miscarriage and early pregnancy loss.¹²

Incidence of preterm birth

Preterm birth is increasing throughout the world, and is now the second leading cause of death of children under 5, after pneumonia.¹³ However, we are still grappling with a consensus definition. Born Too Soon: The Global Action Report on Preterm Birth, published in 2012, was a report by over 50 national and international organizations that provided the first ever estimates of preterm birth rates by country, using country-collected statistics.¹⁴ The definition of preterm birth varies by country and geographic region. The ICD-10 criterion recommends inclusion of all live births prior to 37 weeks' gestation, with no lower boundary, which can complicate between-country comparisons.¹⁵ This variation often hinges on the concept of viability, or the point at which a fetus can survive ex-utero, a time point which has decreased as medical advances improve, and has decreased more in countries with more resources. The Born Too Soon report recommended subdividing preterm birth into three categories of gestational age: extremely preterm (<28 weeks), very preterm (28–32 weeks) and moderately preterm (32–37 weeks).¹⁴ This classification system allows for accurate and comparable statistics to be gathered in both the very and moderately preterm groups, while recognizing that differences between countries' resources makes the extremely preterm group a more heterogenous and difficult to capture population.

Consequences of inaccurate data

Lack of accurate data leads to disparities in representation and thus care. This has a myriad of effects.

Misclassification may affect management in future pregnancies if risk stratification is not recognized. For example, patients who have undergone recurrent pregnancy loss are at increased risk of a future pregnancy loss, but also a variety of adverse pregnancy outcomes. There is also further data associating risk of miscarriage and preterm birth with subsequent maternal health outcomes outside of pregnancy, specifically cardiovascular death.^16,17 If this is not recognized at an individual patient level, future health outcomes may be worse. The single greatest risk factor for preterm birth is a history of prior preterm birth.¹⁸ Decreasing the risk of subsequent preterm births thus involves identifying those patients at greatest risk prior to conception, and in early pregnancy. If this is not recognized, the problem will continue to increase.

Furthermore, at a population level, research into causes and potential treatments for pregnancy loss and preterm birth will be limited if the etiology is unknown or misclassified. It is well established that the majority of miscarriages are secondary to genetic causes.^3,11 However, further research into environmental effects on miscarriage could allow for risk reduction and amelioration for patients, or public health interventions to reduce risk.¹⁹ It is a common misconception that the only way to decrease preterm births is investments in NICU care, but ongoing research in the moderately and very preterm groups indicates that lower intensity investments such as kangaroo care can significantly improve outcomes.²⁰ More research is needed particularly in the extremely preterm groups worldwide both to identify factors that can decrease the incidence of extremely preterm birth, but also improve neonatal outcomes following extremely premature birth. Research funding will also be affected by inaccurate underestimates of incidence, leading to less resources being allocated if the burden of disease is underestimated.

Clinical resources will also be limited if the burden of disease is underestimated. For example, one of the established management strategies for recurrent pregnancy loss is high-intensity ‘TLC’ clinics with increased surveillance and visits, access to psychosocial resources, and regular appointments with ultrasound in early pregnancy.^21,22 There is increasing evidence that in-vitro fertilization in combination with preimplantation genetic testing for aneuploidy (PGT-A) may be a valid treatment for recurrent pregnancy loss.^23,24 These are resource-intensive options, and to economically justify them an accurate measure of burden of disease is important. Similarly, many resources, which are important to reduce the risk of preterm birth, involve comprehensive prenatal care, including identification and treatment of infectious diseases and smoking cessation incentives, all of which require large-scale public health measures to reach the intended communities.²⁵

As mentioned by Agampodi et al.,¹⁰ in communities and countries with less access to healthcare, miscarriage and preterm birth is likely to be underestimated. This will lead to a widening gap in care between lower- and higher-income countries and communities. Within the US, disparities between ethnic and racial groups exist in both access to fertility care and subsequent live birth rates, with African American women waiting significantly longer for referral to care compared to their White counterparts, and all non-White groups experiencing a disparate live birth rate.^26,27 One prospective cohort study in the US did assess pregnancy loss on the basis of race, and found an increased hazard of miscarriage in Black people compared to White.²⁶ Quenby et al. pooled 19 studies encompassing 504,000 women, which assessed disparities in pregnancy loss on the basis of race, and found an increase in miscarriage risk in Black women.¹¹ Worldwide, disparities between groups is less researched, and merits further research.

With regard to preterm birth, over 90% babies born extremely preterm in low-income countries die within the first few days of life, compared to just 10% of extremely preterm infants born in high-income countries.¹⁴ Gathering accurate data on these extremely preterm infants, and ensuring that these are classified accurately is imperative in identifying reasons for this and interventions that can decrease the risk of preterm birth. While advanced neonatal ICU technology contributes significantly to this differential, other low-tech interventions can also significantly impact survival. However, if countries automatically classify extremely preterm infants as stillbirths or inevitable deaths, this data will be missing or not recorded.

WHAT ARE THE POTENTIAL SOLUTIONS?

Standardized definitions

Ideally, standardized definitions would make quantifying and classifying pregnancy loss and preterm birth easier. It is important to recognize that variation in the clinical definition of viability is dependent on a countries’ health system and resources. In order to ensure that a realistic picture is captured, rather than a binary cutoff between miscarriage and stillbirth (following preterm birth), a continuous variable could be used (e.g., estimated gestational age at time of loss), with each country then deciding on the binary variable cutoff that best suits their current health system capabilities. This allows for accurate capturing of data, with the ability to subdivide data into different categories over time as the clinical scenario changes.

The LIFE study followed a cohort of periconceptional couples in the US, and found 28% couples who became pregnant suffered pregnancy loss, all before 22 weeks' gestation. The rate of pregnancy loss increased from time of conception and plateaued at approximately 72 days.⁷ The novel use of a periconceptional cohort highlighted the number of pregnancies that are lost prior to establishment of care and/or a ‘formal’ diagnosis of pregnancy.

Centralized registry

A centralized registry would enable greater capturing of data, and allow for within country comparison of trends in miscarriage and preterm birth over time.

A centralized database in each country would also allow for transnational comparisons, and using the continuous scale outlined above, the upper and lower gestational ages could be adjusted varying on the country in question.

A centralized registry linked to other health outcomes would allow for research into the association between miscarriage and preterm birth and other health outcomes. Regarding pregnancy outcomes, there is a high risk of recurrence for both preterm birth and recurrent miscarriage. Other adverse pregnancy outcomes associated with recurrent pregnancy loss include preterm birth, placental abruption, and hypertensive disorders of pregnancy.²⁸ Preterm birth is the single biggest risk factor for a subsequent preterm birth. It is also associated with maternal and neonatal adverse pregnancy outcomes. Outside of pregnancy outcomes, recent research has demonstrated a link between recurrent miscarriage and preterm birth and subsequent cardiovascular health outcomes.^16,17 A centralized database would allow for larger studies to explore this association further, and amongst different populations.

The question remains how this data is entered into the database once it is established. One proposal is to limit the data gathering to clinical miscarriages, i.e., those that occur once a pregnancy has been confirmed by ultrasound. This makes for ‘cleaner’ data, and more objective data, and helps with standardization. However, it means that people must have access to ultrasound early in pregnancy, which is not often the case in many LMICs. It also means that earlier pregnancy loss, e.g., chemical pregnancy loss, will not be included. Another option is to allow self-identification of miscarriage and open the database to pregnant people themselves to self-report. This could be done in conjunction with their healthcare provider, if they seek that care, but could also be done independently. Options to provide methods of pregnancy confirmation (urine pregnancy test, missed period, ultrasound) could be used to attempt to capture these earlier losses and how pregnant people deal with them. For preterm birth, options in the database may extend to time of delivery, signs of life at birth, and if resuscitative care was offered. This is particularly important at lower gestational ages where the limits of viability vary depending on resources and availability.

DISCUSSION/CONCLUSIONS

In conclusion, pregnancy loss and preterm birth are common and have a huge clinical impact on patients. To date, difficulty in defining pregnancy loss and preterm birth, and the geographical variation in definitions used, has made it harder to quantify the problem globally, identify etiologies, and develop potential solutions. This chapter outlines steps that can be taken to standardize definitions and improve data capture to improve maternal and neonatal pregnancy outcomes and future health.

PRACTICE RECOMMENDATIONS