Obstetric Outcome in Women With Threatened Miscarriage in the First Trimester

Wijesiriwardana, Ajith MRCOG1; Bhattacharya, Sohinee MBBS, MSc2; Shetty, Ashalatha MD1; Smith, Norman MD3; Bhattacharya, Siladitya MD4

Obstetrics & Gynecology:
doi: 10.1097/01.AOG.0000199952.82151.de
Original Research

OBJECTIVE: To assess pregnancy outcomes in women with threatened miscarriage in the first trimester.

METHODS: This was a retrospective cohort study based on data extracted from the Aberdeen Maternity and Neonatal Databank. Cases included all primigravid women with first-trimester vaginal bleeding who delivered after 24 weeks of gestation between 1976 and 2004. The control group comprised all other women who had first pregnancies during the same period. Data were analyzed by univariate and multivariate statistical methods.

RESULTS: Compared with the control group (n = 31,633), women with threatened miscarriage (n = 7,627) were more likely to have antepartum hemorrhage of unknown origin (odds ratio [OR] 1.83, 95% confidence interval [CI] 1.73–2.01). Elective cesarean (OR 1.30, 95% CI 1.14–1.48) and manual removal of placenta (OR 1.40, 95% CI 1.21–1.62) were performed more frequently in these women, who also had a higher risk of preterm delivery (OR 1.56, 95% CI 1.43–1.71) and malpresentation (OR 1.26, 95% CI 1.13–1.40). Threatened miscarriage in the first trimester is required in 112, 112, 17, 85, 32 patients, respectively, for each additional case of manual removal of placenta, elective cesarean, antepartum hemorrhage of unknown origin, malpresentation, and preterm delivery.

CONCLUSION: Pregnancies complicated by threatened miscarriage are at a slightly higher risk of obstetric complications and interventions.


In Brief

Threatened miscarriage is associated with a slight increase in obstetric complications.

Author Information

From the 1Department of Obstetrics and Gynaecology, Aberdeen Maternity Hospital; 2Dugald Baird Centre for Research on Women’s Health, Aberdeen Maternity Hospital; 3Ultrasound Department, Aberdeen Maternity Hospital; and 4Department of Obstetrics and Gynaecology, University of Aberdeen, Aberdeen, United Kingdom.

The authors thank Dr. Doris Campbell and Ms. Linda Murdoch for facilitating the extraction of data from the Aberdeen Maternity and Neonatal Databank.

Corresponding author: Dr. Ajith Wijesiriwardana, Consultant Obstetrician and Gynaecologist, Hon. Senior Clinical Lecturer, Department of Obstetrics & Gynaecology, Cumberland Infirmary, Carlisle, Cumbria, CA2 7HY United Kingdom; e-mail: ajithwijesiriwardana@yahoo.co.uk.

Article Outline

The presumptive diagnosis of threatened miscarriage is based on a history of vaginal bleeding in early pregnancy in the presence of a closed cervix. The diagnosis is confirmed by ultrasonic evidence of an intrauterine gestational sac with a positive fetal heart.1 First-trimester vaginal bleeding affects up to 25% of all pregnancies1 and has been reported to lead to spontaneous miscarriage in 50% of affected women before fetal viability is assessed by ultrasonography.2,3 Although few studies have evaluated outcomes other than viability at term, most agree that adverse pregnancy outcome is associated with first-trimester vaginal bleeding.2–4 The outcome of ongoing pregnancies after first-trimester bleeding is of relevance to women and obstetricians for planning antenatal care and clinical interventions in pregnancy. The literature on this subject is characterized by relatively small sample sizes, arbitrary categorization of cases into groups according to perception of severity of vaginal bleeding,2–4 and inconsistency in terms of selected outcomes. A meta-analysis5 of this subject has assessed only perinatal outcomes after threatened miscarriage.

The geographical setting in the Grampian Region of Scotland is unique in that Aberdeen Maternity Hospital is the only hospital offering comprehensive early pregnancy services as well as maternity services. Maternity-related data in all women in the region are routinely collected and recorded in the Aberdeen Maternity and Neonatal Databank. Stringent and consistent coding criteria, which are essential for studies of secular change, have been used for the coding of gestational length, birth weight, pregnancy complications, and the classification of perinatal events within the Databank. These data were therefore considered to be appropriate for examining obstetric outcomes in women with threatened miscarriage.6 The objective of this study was to compare pregnancy and neonatal outcomes in women with and those without a history of hospital attendance with threatened first-trimester miscarriage.

Back to Top | Article Outline


This was a retrospective cohort study based on data extracted from the Aberdeen Maternity and Neonatal Databank. Cases included women who sought hospital assessment for vaginal bleeding before 12 weeks of gestation between 1976 and 2004, but who subsequently continued the pregnancy to a minimum gestational period of 24 weeks. Controls were all women delivering after 24 weeks within the same period of time, but without history of first-trimester bleeding. The sample population was limited to primigravid women to minimize the potential confounding effect of parity. Women presenting with complete, incomplete, or missed miscarriages and women opting for termination were excluded, as were women with multiple pregnancies and pregnancies with fetal malformations or hydatidiform moles. Women who had a second-trimester miscarriage were also excluded.

Our objective was to investigate the risks associated with viable intrauterine pregnancies complicated by early pregnancy bleeding. The following outcome measures were selected:

1. Pregnancy complications: antepartum hemorrhage including placenta previa, placental abruption (separation of normally situated placenta), and antepartum hemorrhage of unknown origin; preeclampsia (blood pressure > 140/90 with proteinuria), eclampsia, preterm prelabor rupture of membranes (rupture of membranes at < 37 weeks of gestation).

2. Delivery complications: induction of labor, instrumental delivery by forceps or vacuum extraction, malpresentation, cesarean delivery (elective or emergency), manual removal of placenta, and postpartum hemorrhage.

3. Neonatal complications and parameters: preterm delivery (delivery before 37 completed weeks but after 24 weeks), stillbirth (antepartum or intrapartum death), early neonatal death (death within the first 7 days of life), low birth weight (birth weight less than 2,500 g), standardized birth weight score (observed birth weight − mean birth weight in that cell/standard deviation of birth weight in that cell),7 an Apgar score less than 7 at 5 minutes and admission to the neonatal unit.

Data on women with ongoing pregnancies were obtained from the Aberdeen Maternity and Neonatal Databank. The extracted data were transferred to SPSS 13.1 for Windows (SPSS Inc, Chicago, IL) to facilitate data analysis.

A probability value < .01 was considered statistically significant. The Student t test was used to compare differences between mean values of the 2 groups. The χ2 test was applied to compare differences between categorical variables. Univariate and multivariate logistic regression analyses were performed where appropriate, and the results expressed as odds ratios (OR) with 95% confidence intervals (CI).

This study was performed with an anonymized data set generated from the Aberdeen Maternity and Neonatal Databank and did not involve any direct contact with patients. The Grampian Research Ethics Committee did not feel that the project was one which required to be ethically reviewed under the Governance Arrangements for Research Ethics Committees in the United Kingdom.

Back to Top | Article Outline


A total of 39,260 women in their first pregnancies, who delivered after 24 weeks of gestation between 1976 and 2004, were identified from the Aberdeen Maternity and Neonatal Databank. Of these, 7,627 (19.4%) women experienced vaginal bleeding before 12 weeks of gestation, and these women made up the study cohort, while 31,633 (80.6%) women who had no vaginal bleeding in the first trimester formed the control group. The demographic characteristics of the 2 groups of women are summarized in Table 1. Mean age, height, and weight were comparable in the 2 groups. More women with threatened miscarriage were either married or cohabiting (82.5% versus 80.6%) and had a partner or husband whose social class was I or II (23.1% versus 21.6%). Social class data are in accordance with the Registrar General’s occupation-based classification system.8 This system defines social classes I to V according to the individual’s occupation. For example, doctors and lawyers fall into social class I, while unskilled manual workers are classified into social class V. Smoking was more common in the control group (37.1% versus 30.4%).

The obstetric outcomes for women with and those without threatened miscarriage are shown in Table 2. Women with threatened miscarriage were more likely to have antepartum hemorrhage, (OR 1.77, 95% CI 1.09–2.87) and other antepartum hemorrhage of unknown origin (OR 1.83, 95% CI 1.73–2.01). The incidence of placenta previa was higher in the threatened miscarriage group, but the difference did not attain statistical significance at the 1% level. These differences remained statistically significant after adjusting for marital status, social class, and smoking. There was also a significant difference in the incidence of postpartum hemorrhage between the 2 groups, which was not apparent after adjusting for antepartum hemorrhage, cesarean delivery, and manual removal of placenta.

After adjusting for potential confounders, the risks of 2 obstetric procedures were found to be higher in women with threatened miscarriage—elective cesarean delivery (OR 1.30, 95% CI 1.14–1.48) and manual removal of placenta (OR 1.40, 95% CI 1.21–1.62). There were no differences in the incidence of preeclampsia or eclampsia, preterm prelabor rupture of membranes, induction of labor, forceps or vacuum delivery, placental abruption, or emergency cesarean delivery between the 2 groups of women.

Perinatal outcomes for women with and those without threatened miscarriage are presented in Table 3. We found no significant differences in the stillbirth rate, Apgar scores at 5 minutes less than 7, and standardized birth weight score between the 2 groups. Threatened miscarriage was associated with preterm delivery (OR 1.56, 95% CI 1.43–1.71), fetal malpresentation (OR 1.26, 95% CI 1.13–1.40), and low birth weight (OR 1.22, 95% CI 1.09–1.37). These perinatal outcomes remained significantly higher in the threatened miscarriage group after adjusting for potential confounders (Table 3). Univariate analysis showed an increased incidence of neonatal death and admission to the neonatal unit. This association was not borne out by subsequent analysis, which adjusted for confounders (Table 3).

Numbers needed to harm (NNH) were calculated for those outcomes where the adjusted odds ratios demonstrated statistically significant differences between cases and controls. The formula used was NNH = ([PEER × {OR − 1}] +1)/(PEER × [OR − 1] × [1 − PEER]), where PEER is patients’ expected event rate.9 Exposure to threatened miscarriage in the first trimester is required in 112, 112, 17, 85, 32, and women, respectively, for each additional case of manual removal of placenta, elective cesarean delivery, antepartum hemorrhage of unknown origin, malpresentation, and preterm delivery.

Back to Top | Article Outline


The results of this study indicate that antepartum hemorrhage of unknown origin, manual removal of placenta, elective cesarean delivery, preterm delivery, and malpresentation are more common after threatened miscarriage. To our knowledge this is the largest epidemiological study of obstetric outcome in women with threatened miscarriage. The presence of the Aberdeen Maternity and Neonatal Databank provided a unique opportunity to study obstetric events in a large cohort of women in a defined geographical area. Recall bias is minimized because data are collected as events occur and entered by dedicated trained staff. There are a number of mechanisms which ensure the consistency and validity of data entry into the database, including strict definition of clinical conditions, use of valid ranges, and inbuilt consistency checks. These have been described previously.10

One of the weaknesses of the study is its retrospective observational nature. It could be argued that some of the differences between cases and controls may be incidental and only appear to be statistically significant because of the large numbers involved. Presenting the data as numbers needed to harm puts some of these risks into perspective by identifying the actual numbers of women who need to be exposed to the risk of threatened miscarriage for each additional complication. Conversely overadjustment for multiple confounders might have diluted some of the genuine differences between the groups. Another limitation of the study is that data on the amount of bleeding were not collected. Consequently, although it is true to assume that the amount of bleeding was sufficient to warrant evaluation in a hospital setting, we were unable to report on the association between severity of bleeding and obstetric outcome. Moreover, women who had very light bleeding who did not seek hospital consultation may have been missed from this data set. In this context, it is important to realize that subjective assessment of blood loss is often erroneous in these situations and objective assessment often impractical. Except in extreme situations, the management of women with first trimester bleeding is more likely to be dictated by ultrasound confirmation of viability rather than volume of blood loss. We therefore believe that these findings are relevant to the general population of women with threatened miscarriage.

In this sample, women with threatened miscarriage were more likely to be married (or cohabiting) and to have a husband/partner with a social class of I or II. This may reflect a lower threshold for seeking medical care promptly in this group. Smokers were less likely to present with first-trimester threatened miscarriage. The relationship between smoking and adverse reproductive outcomes is complex. Smoking has been linked with miscarriage in some studies.11,12 In a recent paper, Yang et al13 were unable to prove an association between cigarette smoking and overall vaginal bleeding in pregnancy but found a correlation between passive smoking and second-trimester bleeding. This finding was supported by Mulik et al.14 Other data showing a reduced risk of preeclampsia in smokers suggest that smoking may affect trophoblastic invasion of uterine vessels within the developing placenta.15 The association between smoking and absence of threatened miscarriage detected in the present study may be related to the difference in social class between cases and controls. It is difficult to rule out the possibility that this might be an incidental finding that has gained statistical significance because of the large sample size.

Our study shows an association between threatened miscarriage and preterm delivery (< 37 weeks), in keeping with evidence from previous studies.2,15,16 Batzofin et al16 and Williams et al17 reported that threatened miscarriage doubled the risk of delivery before 37 weeks. Strobino and Pantel-Silverman18 did not find an association between preterm deliveries and light vaginal bleeding in the first or second trimester, whereas Weiss et al19 found that the risk of delivery before 37 weeks was significantly increased in women irrespective of whether they had light or heavy first-trimester bleeding. In our study the association between first-trimester bleeding and preterm delivery persisted when deliveries before 34 weeks were analyzed separately. This is an important observation because data from most previous studies have not distinguished between preterm birth before 34 and 37 completed weeks of gestation, a difference that could have a significant impact on neonatal care and outcomes. It could be speculated that the presence of blood after threatened miscarriage, with the disruption of the chorioamniotic space and the resultant chronic inflammatory reaction, might precipitate preterm labor. A hematoma could also form a nidus for intrauterine infection, which, in turn, could stimulate uterine contractions.

Our results also suggest a trend toward an increased prevalence of placenta previa in the study population. The location of the chorion frondosum within the uterine cavity in early pregnancy may explain this association, with an inferior position more likely to cause first-trimester bleeding, as well as a higher risk of placenta previa later on in pregnancy. Das et al20 reported an increased risk for a low-lying placenta among women with threatened miscarriage but found no difference in placental location compared with control subjects by 36 weeks of gestation. Weiss et al19 found a similar association that was not statistically significant. Mulik et al14 found a significantly higher risk of placenta previa at 37 weeks in women who experienced a first-trimester vaginal bleed.

Data linking cesarean delivery to threatened miscarriage are very limited. Our study showed a higher incidence of elective cesarean delivery among the study group, but, in contrast to the findings of Weiss et al,19 there was no evidence of an association with emergency cesareans. Thus, the difference may be due to the cumulative effect of placenta previa and malpresentation.

Unlike previous reports suggesting that pregnancy induced hypertension was significantly more common in subjects with threatened miscarriage, we found no such association.4,19 Again, in contrast to previous data from relatively small studies,14,19,21 no link between threatened miscarriage and preterm premature rupture of membranes could be confirmed. The risk of placental abruption was found to be no higher in women with first-trimester bleeding. This association had been highlighted by Mulik et al,14 whereas Weiss et al19 found an increased incidence of placental abruption only with heavy first-trimester bleeding.

The incidence of manual removal of placenta was found to be higher among women with threatened miscarriage. Hertz and Heisterberg21 reported that retention of placenta was associated with threatened miscarriage, and the rate of manual removal was 14%. They postulated that adhesive scarring between the uterine wall and the placenta at the site of bleeding might be responsible for the increased incidence of retention of placentae in women with threatened miscarriage.

Low birth weight was associated with first-trimester bleeding after adjusting for preterm delivery. The proportion of preterm deliveries in this study cohort (9.6%) was comparable to the figures from previous work.21,22 The standardized birth weight scores were, however, not significantly different between women with and those without first-trimester bleeding. The advantage of standardized birth weight score over crude birth weight is that the former is adjusted for gestational age, sex, and parity. Of the two, this is a better indication of growth restriction. It is thus a more accurate marker of growth restriction, and we have chosen not to put undue emphasis on birth weight alone in our interpretation of these results. Although it has been suggested that growth restriction may be due to some degree of placental insufficiency secondary to scarring at the site of placental bleeding,21 we did not find a significant correlation between the two in this study.

Williams et al17 found a 2.5-fold increase in the risk of neonatal death (95% CI 1.1–5.5) in women with threatened miscarriage, an association that was initially supported by results of this study. But the significance of this association was lost when adjusted for antepartum hemorrhage, preterm delivery, malpresentation, and low Apgar scores. Mulik et al14 showed that preterm delivery, placental abruption, and low birth weight were independently responsible for the early neonatal death rates in their study population. They also showed that threatened miscarriage, which was not independently responsible for early neonatal death rates, might therefore have had an indirect effect on subsequent poor perinatal outcome.

Although it is reassuring that the majority of women with first-trimester bleeding have pregnancy outcomes comparable to those without such bleeding, it is evident that they face a higher relative risk of some adverse obstetric and neonatal outcomes. This information is helpful in terms of counseling women and alerting clinicians to early signs of these complications.

Back to Top | Article Outline


1. Cunningham FG, Gant NF, Leveno KJ, Gilstrap LC, Hauth JC, Wenstrom KD, editors. William’s obstetrics. 21st ed. New York (NY): McGraw-Hill; 2001.
2. Farrell T, Owen P. The significance of extrachorionic membrane separation in threatened miscarriage. Br J Obstet Gynaecol 1996;103:926–8.
3. Chung TK, Sahota DS, Lau TK, Mongelli JM, Spencer JA, Haines CJ. Threatened abortion: prediction of viability based on signs and symptoms. Aust N Z J Obstet Gynaecol 1999;39:443–7.
4. Verma SK, Premi HK, Gupta TV, Thakur S, Gupta KB, Randhawa I. Perinatal outcome of pregnancies complicated by threatened abortion. J Indian Med Assoc 1994;92:364–5.
5. Ananth CV, Savitz DA. Vaginal bleeding and adverse reproductive outcomes: a meta-analysis. Paediatr Perinat Epidemiol 1994;8:62–78.
6. Samphier M, Thompson B. The Aberdeen Maternity and Neonatal Databank. In: Mednick SA, Baert AE, editors. Prospective longitudinal research: an empirical basis for the primary prevention of psychosocial disorders. 1st ed. Oxford (UK): Oxford University Press; 1981. p. 61–5.
7. Campbell D, Hall M, Lemon J, Carr-Hill R, Pritchard C, Samphier M. Clinical birthweight standards for a total population in the 1980s. Br J Obstet Gynaecol 1993;100:436–45.
8. General Register Office. Standard occupational classification. London (UK): Her Majesty’s Stationery Office; 1990.
9. Centre for Evidence-Based Medicine. NNT and NNH. Available at: http://www.cebm.net/nnts.asp. Retrieved December 14, 2005.
10. Bhattacharya S, Campbell DM, The incidence of severe complications of preeclampsia. Hypertens Pregnancy 2005;24:181–90.
11. Klonoff-Cohen H. Female and male lifestyle habits and IVF: what is known and unknown. Hum Reprod Update 2005;11:179–203.
12. Lintsen AM, Pasker-de Jong PC, de Boer, EJ, Burger CW, Jansen CA, Braat DD, et al. Effects of subfertility cause, smoking and body weight on the success rate of IVF. Human Reprod 2005;20:1867–75.
13. Yang J, Savitz DA, Dole N, Hartmann KE, Herring AH, Olshan AF, et al. Predictors of vaginal bleeding during the first two trimesters of pregnancy. Paediatr Perinat Epidemiol 2005;19:276–83.
14. Mulik V, Bethel J, Bhal K. A retrospective population-based study of primigravid women on the potential effect of threatened miscarriage on obstetric outcome. J Obstet Gynaecol 2004;24:249–53.
15. Wisborg K, Kesmodel U, Henriksen TB, Hedegaard M, Secher NJ. A prospective study of maternal smoking and spontaneous abortion. Acta Obstet Gynecol Scand 2003;82:936–41.
16. Batzofin JH, Fielding WL, Friedman EA. Effect of vaginal bleeding in early pregnancy on outcome. Obstet Gynecol 1984;63:515–8.
17. Williams MA, Mittendorf R, Lieberman E, Monson RR. Adverse infant outcomes associated with first-trimester vaginal bleeding. Obstet Gynecol 1991;78:14–8.
18. Strobino B, Pantel-Silverman J. Gestational vaginal bleeding and pregnancy outcome. Am J Epidemiol 1989;129:806–15.
19. Weiss JL, Malone FD, Vidaver J, Ball RH, Nyberg DA, Comstock CH, et al. Threatened abortion: a risk factor for poor pregnancy outcome, a population-based screening study. Am J Obstet Gynecol 2004;190:745–50.
20. Das AG, Gopalan S, Dhaliwal LK. Fetal growth and perinatal outcome of pregnancies continuing after threatened abortion. Aust N Z J Obstet Gynaecol 1996;36:135–9.
21. Hertz JB, Heisterberg L. The outcome of pregnancy after threatened abortion. Acta Obstet Gynecol Scand 1985;64:151–6.
22. Uerpairojkit B, Charoenvidhya D, Tannirandorn Y, Wacharaprechanont T, Manotaya S, Samritpradit P, et al. Sonographic findings in clinically diagnosed threatened abortion. J Med Assoc Thai 2001;84:661–5.

Cited By:

This article has been cited 1 time(s).

Obstetrics & Gynecology
First-Trimester Vaginal Bleeding and Complications Later in Pregnancy
Lykke, J; Dideriksen, K; Lidegaard, Ø; Langhoff-Roos, J
Obstetrics & Gynecology, 115(5): 935-944.
PDF (474) | CrossRef
Back to Top | Article Outline
© 2006 The American College of Obstetricians and Gynecologists