Stillbirth is one of the most devastating, as well as common, obstetric complications, affecting over 3 million pregnancies per year throughout the world. Definitions of stillbirth (and thus stillbirth rates) vary in different countries, based on gestational age. In the United States, stillbirth (defined as fetal death at ≥20 weeks’ gestation) affects about 1 in 160 pregnancies (6-7 per 1000 births). The rate is similar in most high-income countries (3-5 per 1000 births), but is considerably higher (20-100 per 1000 births) in low- and middle-income countries. A recent series published in the Lancet underscored the relative lack of attention and research funding stillbirth has received, as well as suggestions for reducing the rate of stillbirth.

A major unanswered question is the optimal diagnostic evaluation for cases of stillbirth. Indeed, the optimal evaluation for stillbirth remains controversial and there are few data with which to guide clinicians. There are several reasons why it can be difficult to ascertain a “cause” of stillbirth. First, the cause of stillbirth may be complex and multifactorial. Several conditions simultaneously occurring may contribute to a given stillbirth and it may not be possible to determine a single proximate cause. Second, many conditions are “risk factors” rather than causes of stillbirth. These conditions are often present in pregnancies with live births. Examples include smoking, obesity, or well-controlled diabetes. Third, a cause of death may not be identified, even after a thorough evaluation of stillbirth. Finally, there are likely causes of stillbirth that have yet to be discovered.

There are also several nonmedical issues that influence the diagnostic “workup” of stillbirths. Cost is a major concern and in many countries, patients and health care systems cannot or will not pay for all of the tests that may potentially be useful in the evaluation of stillbirths. In addition, emotional and cultural factors may influence a family’s willingness to permit a stillbirth investigation. Parents may believe that a “workup” will not change anything or they may be angry at the medical system as the bearer of bad news. Also, families may have reservations about autopsy because of religious beliefs or misconceptions about what actually takes place.

In this issue of the American Journal of Obstetrics and Gynecology , Korteweg and colleagues present data suggesting that autopsy, placental examination, cytogenetic analysis, and testing for maternal-fetal hemorrhage should be the “basic” workup for stillbirth. They conducted a multicenter prospective cohort study in the Netherlands including 1025 patients with stillbirth. Cause of death was assigned by a panel using the TULIP classification system. The “value” of each diagnostic test for determination of cause of death was determined by adding the percentage of time that a “positive test” established a cause and a “negative test” excluded a cause in the TULIP system. Placental examination was the most valuable test, helping to determine a cause of death in 95.7% of cases. The next most helpful test was autopsy, which was valuable 72.6% of the time, followed by cytogenetic analysis, which was valuable in 29.0% of cases.

It is also noteworthy that many other tests that are routinely recommended in the evaluation of stillbirth were not particularly valuable in the overall cohort. These included tests such as glucose testing, anticardiolipin antibodies, and viral serology in the mother. Rather than screen every case with all these tests, the authors propose a paradigm for a “staged” workup of stillbirth. Such an evaluation would include an autopsy, placental histologic examination, and fetal karyotype (the 3 most valuable tests), as well as a test for fetal-maternal hemorrhage (which cannot be accomplished at a later date) in all cases. Tests such as maternal serology, cultures, toxicology screen, assessment of hormones, and assays for antiphospholipid syndrome would only be performed when suspicion for a particular condition exists based on the clinical history or results of the first-stage evaluation. For example, assessment of glycated hemoglobin is appropriate if the fetus is macrosomic.

This protocol is more cost effective than a “shotgun” approach that includes tests for all conditions associated with stillbirth. It makes sense to perform tests in the first-stage evaluation that are likely to identify the most common causes of stillbirth. Ascertaining potentially treatable conditions also would be worthwhile.

A major problem with following these recommendations, especially in the United States, is underutilization of perinatal autopsy. For example, a study of stillbirths in Utah noted an autopsy rate of only 35% in tertiary care centers and 13.3% in community hospitals. In some cases, this is due to reluctance on the part of families to allow an autopsy. It is important for clinicians to explore the specific concerns or misconceptions that families may have about autopsy. Also, family members should be educated regarding the importance of autopsy with regard to making a diagnosis, facilitating the grieving process, providing emotional closure, delineating a precise recurrence risk, and in some cases, aiding the prevention of a recurrent stillbirth. If parents refuse complete autopsy, they will often agree to partial autopsy, external examination, and/or postmortem MRI as outlined in the Dutch protocol.

Another major issue with regard to perinatal autopsy is a relative lack of pathologists with expertise or fellowship training in perinatal pathology. This certainly is an issue in the United States. Consequently, the quality of perinatal autopsies may be suboptimal. The same is true for histologic examination of the placenta. In most centers, autopsies are not remunerated, which provides a disincentive for the training of more perinatal pathologists. A concerted effort should be made to make these services more widely available. Meanwhile, agreements can be made among hospital systems to “transport” stillborn infants and their placentas to centers with expertise in periantal pathology.

The recommendations of Korteweg and colleagues may not be applicable to all populations and would be difficult to accomplish in low-resource settings. The Dutch cohort may differ from other populations with regard to race, gestational age, and the exclusion of intrapartum stillbirths. Also, the value of each test will vary with the classification system used to ascertain causes of stillbirth. There are currently over 35 systems and none is uniformly agreed on as being superior to the others.

Nonetheless, most high-income countries should consider adopting the protocol put forth by Korteweg and coworkers in this issue of the American Journal of Obstetrics and Gynecology . These are the first guidelines developed based on evidence rather than empiric recommendations. The cohort studied was relatively large and unbiased. We are indebted to Drs Korteweg and Erwich and their colleagues for their numerous important contributions to our current knowledge about stillbirth. Widespread adoption of these guidelines would be another step forward in reducing the rate of stillbirth.