Curtin, William M. MD1; Krauss, Sarah MD1; Metlay, Leon A. MD2; Katzman, Philip J. MD2
OBJECTIVE: To estimate the percentage of deliveries eligible for pathologic examination of the placenta and compare with observed practice using the College of American Pathologists’ (CAP) 1997 guidelines for examination of the placenta.
METHODS: Records were reviewed from all live-birth deliveries 20 weeks or more of gestation in 2001 at Strong Memorial Hospital. The expected number of deliveries with CAP recommended indications was determined and compared with the observed number of deliveries in which the placenta was actually examined. Descriptive statistics, independent t tests, χ2 tests, difference between two population proportions test, odds ratios, 95% confidence intervals, and multiple logistic regression were used to analyze the data.
RESULTS: The observed number and percentage of deliveries with CAP recommended indications that had pathologic placental examination, 575 and 18.2% (95% confidence interval 16.9–19.6), was significantly lower (P<.001) than expected, 1,185 and 37.5% (95% confidence interval 35.8–39.2). The placenta was examined less frequently than expected in 9 of 14 categories. Independent predictors of examination of the placenta were gross placental abnormalities, multiple gestation, prematurity, peripartum fever, neonatal intensive care unit care of infant, cesarean delivery, and delivery by a maternal–fetal medicine specialist.
CONCLUSION: Using the CAP guidelines for submission of the placenta would result in pathologic examination in 37.5% of all deliveries. Less than one half of all deliveries in which the placenta was eligible for submission were actually examined. Current advances in our understanding of pathologic conditions of the placenta and their relation to infant outcomes may warrant reevaluating policy on placental examination at institutional and national levels.
LEVEL OF EVIDENCE: II
There is wide variation in clinical practice as to when the placenta is submitted for pathologic examination. When asked in a national survey what percentage of births the placenta was pathologically examined, pathologists gave the following responses: 1–10% (34.7%), 11–25% (36.4%), 26–50% (15.4%), 51–99% (7.6%), 100% (5.9%).1 In 1991, the Working Group on Indications for Placental Examination, sponsored by the College of American Pathologists (CAP), reported a list of maternal, fetal and neonatal, and placental conditions for which a gross and microscopic placental examination was recommended.2 Using these guidelines, one case–control study of infants with newborn encephalopathy found that only 11.2% of placentas from these births were examined; in the control group only 0.7% of placentas were examined despite 43.3% fulfilling one or more criteria for examination.3 It is probable that in most hospitals the placenta is pathologically examined much less frequently than expected.
The American College of Obstetricians and Gynecologists declined recommending any individual indications for placental examination in live births, expressing concerns that there were insufficient data to support the use of such a policy.4 Other experts have recommended that pathologists examine all placentas.5 In 1997, CAP published revised and more specific guidelines for examination of the placenta.6 The guidelines arose from a consensus conference that included representatives from perinatal pathology, neonatology, and perinatology, among others. The committee recommended that all placentas be examined by the clinician and triaged in the delivery room. Placentas that had clinical indications for examination or were grossly abnormal were recommended to be submitted for detailed gross and light microscopic study; placentas grossly normal and without clinical indications for examination were recommended to be refrigerated for three days and submitted only if maternal or neonatal complications arose in the interim. The committee agreed upon 25 clinical indications encompassing maternal, fetal and neonatal, and placental conditions for which gross and microscopic placental examination was recommended; there were 13 other clinical indications listed for which no agreement was reached (see box).
At our institution the decision to submit the placenta to pathology for gross and microscopic examination is at the discretion of the physician; there are no formal guidelines recommending specific clinical indications. There are instances when one physician would choose to submit the placenta, eg, fever in labor, whereas another would not. It was our hypothesis that placentas were frequently not submitted for pathologic examination that had appropriate indications. Using the 1997 CAP guidelines on clinical indications for submission of the placenta for examination, the goal of this study of all live births in 1 year was to compare the expected with the observed number of pathologic examinations of the placenta.
MATERIALS AND METHODS
This was an institutional review board–approved review of the computerized medical records of all mothers and their liveborn infants delivered at 20 weeks of gestation or greater in 2001 at Strong Memorial Hospital, the designated regional perinatal center for the New York State Finger Lakes Region. There was no formal policy in place regarding the submission of placentas for pathologic examination. A list of all deliveries and demographic information was obtained from the hospital birth certificate registry. We reviewed the summary discharge coding sheet for both mother and newborn, and any discharge summary or operative report in the hospital’s clinical information system. We recorded all CAP-recommended and other indications for placental examination as given in the box. All deliveries were categorized as to whether pathologic examination of the placenta was performed. An expected pathologic examination was defined as a delivery that had one or more recommended maternal, fetal or neonatal, or placental indications. An observed pathologic examination was defined as a delivery in which the placenta was actually examined pathologically. Some of the recommended clinical indications were modified to facilitate data collection. Any diagnosis of oligohydramnios was considered in the severe category, with no attempt to define the magnitude of the condition. The category severe diabetes was interpreted as any pregestational diabetes, and impaired glucose tolerance was interpreted to mean gestational diabetes. All multiple gestations, regardless of chorionicity or growth discordance, were considered in the recommended category. The clinical information provided to the pathologist, as reflected on the requisition forms of the submitted placentas, was assessed in three categories: gestational age, indication, and Apgar scores.
Data were expressed in percentages with 95% confidence intervals. Demographic data were compared using the independent t test and χ2 tests as appropriate. The observed and expected percentages of placentas examined pathologically were compared using the difference between two population proportions test.7 We hypothesized that the expected percentage of deliveries with indications for pathologic examination of the placenta would be 43% but the observed percentage of placentas examined would be only 30%. A sample size of approximately 1,500 deliveries would be required to demonstrate this difference with a power of 99% and significance of .001. The observed number of placentas examined pathologically in individual maternal, fetal or neonatal, and placental categories as a percentage of total deliveries was compared with those expected to be examined. A sample size of 2,964 deliveries would be required to demonstrate a 50% difference between the observed and expected placental examinations in individual categories at a prevalence of 2% with a power of 80% and significance of .001. There were a total of 14 categories that were indications for pathologic examination in at least 50 deliveries. A Bonferronni correction was made for the 14 individual comparisons, such that a statistically significant difference was set at P<.003. Data for suspected placental abruption, while accounting for only a small number of cases, were included because of the high submission rate for this category.
It became clear during the study that pregnancy, delivery, and birth information on the physician requisition forms for placental examination was usually incomplete. To understand which clinical factors predicted placental examination among the clinicians at our institution, the odds ratios (ORs) and 95% confidence intervals for placental examination in the individual CAP recommended indications (present in at least 50 deliveries except placental abruption) were calculated. In addition, ORs for the following variables believed to be factors in requesting placental examination were also calculated: maternal–fetal medicine specialist as the obstetrician, cesarean delivery, maternal age 20 years or younger, rupture of membranes more than 24 hours, delivery at 35–36 weeks, and all deliveries less than 37 weeks. For gestational age, all preterm births less than 37 weeks was used. These variables were then entered into a forward stepwise multiple logistic regression with placental examination as the dependent variable to develop a model of independent predictors for placental examination.8 Statistical analysis was performed using SPSS 9.0 for Windows (SPSS Inc., Chicago, IL).
During the year 2001 there were 3,160 deliveries at Strong Memorial Hospital between 20 and 42 weeks of gestation, including 70 sets of twins and 8 sets of triplets for a total of 3,246 infants. The demographic data are compared by placental examination status in Table 1. Not surprisingly, the group with placental examination had infants of lower gestational age and birth weight and a higher proportion of deliveries with multiple gestation and cesarean section. The summary of the placental examination status of these deliveries is given in Table 2. Overall, 616 (19.5%) deliveries had a pathologic examination of the placenta. Only 41 deliveries (1.3 %) in which the placenta was examined did not have CAP recommended indications; however, at least one half fulfilled CAP “other” indications for examination. There were then 575 (18.2%) deliveries in which one or more CAP recommended indications were present, and pathologic examination of the placenta was performed. However, 1,185 (37.5%) deliveries were identified that were expected to be pathologically examined. The difference between the expected and observed percentages was significant (P<.001). Less than half (48.5%) of those deliveries with clinical indications, according to CAP guidelines, actually had the placenta examined.
Evaluation of the completeness of information on the requisition form accompanying the placenta showed that 96% gave an indication, 60% included a gestational age, and only 3.3% gave the Apgar scores of the infant. Only 2.8% of requisition forms from submitted placentas gave all three. Comparisons of individual CAP-recommended indications are given in Table 3. In five categories (delivery 34 weeks or less, multiple gestation, Apgar 6 or less, placental indications, suspected abruption), the difference between the expected and observed number of placental examinations was not significant. Significantly fewer placentas than expected (P<.003) were examined in the remaining categories.
The odds ratios for clinical variables associated with placental examination are given in Table 4. The results of the multiple logistic regression demonstrated that the best model included the following variables as independent predictors of placental examination: placental indications (documentation of a gross placental abnormality by the delivering physician), multiple gestation, gestational age less than 37 weeks, peripartum fever or infection, admission to other than a level 1 nursery, cesarean delivery, and maternal–fetal medicine specialist as the delivering physician (Table 5). Although the overall model is statistically significant and the model’s estimate fit the data, only 56% of the variation in placental examination can be explained by the independent variables selected.
Despite 37.5% of all deliveries having indications for pathologic examination of the placenta, using the 1997 CAP criteria, only 18.2% of all deliveries had a placental examination. There was no evidence of oversubmission of the placenta, because 93% of all submitted placentas had appropriate CAP indications. In reviewing individual indications for placental examination (Table 3), the observed and expected rates of placental examination were not statistically different for the categories of preterm delivery (less than 34 weeks), Apgar 6 or less at 5 minutes, multiple gestations, suspected placental abruption, and placental indications (a gross placental abnormality noted by the delivering physician). It would seem then that physicians are in agreement that these are important indications to submit the placenta for examination. In all the other categories, significantly fewer than expected placentas were examined. This would indicate less agreement among physicians in these categories.
Limitations of this study are that it was retrospective and reviewed only the discharge coding summaries, dictated discharge summaries, and operative notes, all of which were available in the hospital’s clinical information system. It did not include chart review. It is possible that some indications were underreported whereas others were overreported. The category of birth weight less than 10th percentile illustrates this point. Although it is expected that this would be coded in approximately 10% of all deliveries, only 2.8% of infants were coded as such. It is possible that the small for gestational age infant is coded only at very low birth weights. The CAP guidelines are very broad, and it is possible that clinicians submit placentas only in the more severe cases, eg, a woman with severe preeclampsia but not mild preeclampsia.
Because the observed number of placentas examined in this study differed from the expected number, it is unlikely that most clinicians are aware of the 1997 CAP guidelines for placental examination, or alternatively, do not agree that all these placentas need to be submitted. The results of the multiple logistic regression suggest that physicians make decisions on whether to submit the placenta by considering just a few factors: gestational age, mode of delivery, infant admission to the neonatal intensive care unit, maternal fever, and gross placental abnormalities. In addition, maternal–fetal medicine specialists were more likely to submit the placenta for examination than were other obstetricians. Although placentas from cesarean deliveries were more frequently examined than those from vaginal deliveries, there was no evidence that physicians submitted the placenta to pathology simply because it was a cesarean delivery. A likely explanation is that the physician more carefully considered all of the indications and the logistical setting facilitated the submission. A limitation of the model in Table 5 is that it explains only 56% of the variation in placental examination.
The finding that 37.5% of all deliveries in our population had indications for pathologic placental examination is in agreement with two prior case–control studies that used the earlier 1991 College of American Pathologists guidelines and found that 43% of deliveries had one or more indications.3,9 Spencer and Khong,10 using the 1997 CAP guidelines, found that in their institution 50% of deliveries should have had the placenta examined, but only 20% were actually examined. They had high rates of placental examination in the categories gestational age 34 weeks or less and multiple gestation, similar to our study. They also found, similar to our study, that the amount of information included on pathology request forms was frequently incomplete. The pathologist needs to have enough clinical information both to search more thoroughly for key histopathologic features and to interpret findings in light of the clinical scenario. In the infant that is small for gestational age, for example, findings such as infarcts and increased circulating nucleated red cells would support the idea that uteroplacental insufficiency was the cause of suboptimal growth. The interest and expertise of the pathologist in placental pathology are also important factors in obtaining clinically useful information.
Altshuler2 gave the following reasons for placental examination: diagnostic (mother or neonate), prognostic (to predict the outcome of future pregnancies), investigative, and legal. Parvovirus, cytomegalovirus and Listeria are examples in which placental examination can be diagnostic. Maternal floor infarction, a placental lesion associated with intrauterine growth restriction and stillbirth, is associated with an increased risk of recurrence in subsequent pregnancies.11 Knowledge of this information may be used prognostically in future pregnancies to possibly prevent subsequent fetal deaths.12 Research (investigative) uses for placental examination have been important in preterm birth and cerebral palsy. Clinically silent antepartum infection is believed to be a factor in a significant proportion of preterm births as evidenced by the findings of histologic chorioamnionitis.13 Histologic chorioamnionitis has been associated with cerebral palsy.14 Cerebral palsy not infrequently results in obstetric litigation. The study by Booth et al9 notes that 75% of children with cerebral palsy born weighing 2,500 g or more had one or more maternal or neonatal indications for placental submission by the 1991 College of American Pathologist’s guidelines. If guidelines such as these were to be instituted nationally, this could represent a significant research opportunity to gain insight into pathologic processes associated with cerebral palsy. Placental examination may provide legal defense against allegations of malpractice when abnormal findings indicate that fetal or placental disease contributed to adverse pregnancy outcomes independent of the clinical care rendered during the pregnancy.15 Some medical liability insurers have included versions of the College of American Pathologists guidelines in their risk management literature to their physicians. Other experts have listed similar versions of maternal or fetal conditions that would suggest the need for gross and microscopic placental examination in infants with risk factors for neonatal encephalopathy and cerebral palsy.16
The use of the CAP 1997 guidelines would result in a reasonable number (37.5%) of prompt gross and microscopic placental examinations. Ideally, placentas from all other deliveries should be refrigerated for at least 72 hours and submitted if a new maternal or neonatal indication arises. Adopting these guidelines would maximize submission of the placenta for appropriate clinical indications while minimizing unnecessary examinations. It is ultimately up to individual institutions to develop their own policies regarding when it is appropriate to submit the placenta for pathologic examination. Although more research relating specific gross and histopathologic abnormalities to short- and long-term infant outcomes is desirable, our understanding has advanced to a much higher level over the last 10–20 years. This may warrant reevaluating policy on placental examination at institutional and national levels.
1. Gersell DJ. ASCP survey on placental examination. American Society of Clinical Pathologists. Am J Clin Pathol 1998;109: 127–43.
2. Altshuler G, Deppisch LM. College of American Pathologists Conference XIX on the Examination of the Placenta: report of the Working Group on Indications for Placental Examination. Arch Pathol Lab Med 1991;115:701–3.
3. Badawi N, Kurinczuk JJ, Keogh JM, Chambers HM, Stanley FJ. Why is the placenta being ignored? Aust N Z J Obstet Gynaecol 2000;40:343–6.
4. American College of Obstetricians & Gynecologists. Placental Pathology. ACOG Committee Opinion 125. Washington, DC: ACOG; 1993.
5. Salafia CM, Vintzileos AM. Why all placentas should be examined by a pathologist in 1990. Am J Obstet Gynecol 1990;163:1282–93.
6. Langston C, Kaplan C, Macpherson T, Manci E, Peevy K, Clark B, et al. Practice guideline for examination of the placenta developed by the placental pathology practice guideline task force of the College of American Pathologists. Arch Pathol Lab Med 1997;121:449–76.
7. Daniel W.. Hypothesis testing. In: Biostatistics: a foundation for analysis in the health sciences. 7th ed. New York (NY): John Wiley & Sons, Inc.; 1999. p. 204–94.
8. Hosmer DW, Lemeshow S. Applied logistic regression. 2nd ed. New York (NY): John Wiley & Sons, Inc.; 2000.
9. Booth VJ, Nelson KB, Dambrosia JM, Grether JK. What factors influence whether placentas are submitted for pathologic examination? Am J Obstet Gynecol 1997;176:567–71.
10. Spencer MK, Khong TY. Conformity to guidelines for pathologic examination of the placenta. Arch Pathol Lab Med 2003;127:205–7.
11. Andres RL, Kuyper W, Resnik R, Piacquadio KM, Benirschke K. The association of maternal floor infarction of the placenta with adverse perinatal outcome. Am J Obstet Gynecol 1990;163:935–8.
12. Clewell WH, Manchester DK. Recurrent maternal floor infarction: a preventable cause of fetal death. Am J Obstet Gynecol 1983;147:346–7.
13. Guzick DS, Winn K. The association of chorioamnionitis with preterm delivery. Obstet Gynecol 1985;65:11–6.
14. Redline RW, O’Riordan MA. Placental lesions associated with cerebral palsy and neurologic impairment following birth. Arch Pathol Lab Med 2000;124:1785–91.
15. Altshuler G. Placenta within the medicolegal imperative. Arch Pathol Lab Med 1991;115:688–95.
16. Gibbs RS, Rosenberg AR, Warren CJ, Galan HL, Rumack CM. Suggestions for practice to accompany Neonatal Encephalopathy and Cerebral Palsy. Obstet Gynecol 2004;103:778–9.