The occurrence of abruptio placentae in Canada:

1990 to 1997

Vol. 25 No. 2, 2004

Teresa Broers, Will D King, Tye E Arbuckle and Shiliang Liu

Abstract

Abruptio placentae is a serious obstetric condition associated with an increased incidence of perinatal mortality and morbidity. Despite this, there is little information on the occurrence of abruptio placentae in Canada. The Discharge Abstract Database from the Canadian Institute for Health Information was used to identify a cohort of women who had singleton live or stillbirth deliveries in Canada between 1990 and 1997 (n = 2,162,815). Rates of abruptio placentae and abruptio placentae ending in stillbirth were examined by calendar year, province, maternal age and urban/rural status. There is a trend towards an increasing rate of abruptio placentae by year, from 10.9 (95% confidence interval [CI] 10.5-11.3) cases/1,000 deliveries in 1990 to a high of 12.1 (95% CI 11.6-12.5) cases/1,000 deliveries in 1996, while the rate ending in stillbirth remained relatively constant. The abruptio placentae rate was highest in mothers over 40 years of age and the case-fatality rate highest in those under 20. These results provide a baseline reference for rates of abruptio placentae in Canada.

Key words: abruptio placentae; Canada; epidemiology; placenta diseases; pregnancy complications

Introduction

Abruptio placentae, the premature separation of a normally implanted placenta before delivery, can be a serious pregnancy complication to both mother and infant. It is associated with an increased incidence of preterm delivery as well as maternal and perinatal mortality, causing between 15% and 25% of all perinatal deaths.^1-3

The rate of abruptio placentae in North America is approximately 0.1-0.2 per 1,000 pregnancies,^4-7 but reported rates can range from 0.04-0.35/1,000.^1,3,8,9 This wide range in reported incidence rates may be explained partly by the differing criteria for diagnosing abruptio placentae as well as the increased recognition of milder forms of the event, i.e., the separation of the placenta from the uterine wall can be complete, partial, or marginal (involving only the placental margin). Complete detachment of the placenta from the uterus is more likely to result in a fetal death than partial or marginal separation,⁴ while a marginal abruption may go undetected.

The primary etiology for abruptio placentae is still unknown, but several risk factors have been identified, including pre-eclampsia, pre-pregnancy hypertension, previous history of placental abruption, increased maternal age, cigarette smoking, and cocaine use.^1,3,9-11 It has also been hypothesized that the etiology for a marginal or partial abruptio placentae may differ from that of a complete abruptio placentae.⁴

Despite the potential severity of abruptio placentae, particularly in the case of stillbirth and maternal death, no data have been reported on the geographic and temporal distribution of abruptio placentae cases in Canada. The Canadian Perinatal Health Report - 2000, prepared by Health Canada's Bureau of Reproductive and Child Health and the Canadian Perinatal Surveillance System Steering Committee, provides information on numerous reproductive indicators.¹² Abruptio placentae, and other placental conditions such as placentae previa, are not monitored, however. Descriptive information on abruptio placentae is important in order to gain a better understanding of the event across the country. This paper describes rates of abruptio placentae, abruptio placentae ending in stillbirth and case fatality by year, province, maternal age and urban/ rural residence in Canada between 1990 and 1997.

Methods

The Discharge Abstract Database, maintained by the Canadian Institute for Health Information (CIHI), was the source of abruptio placentae cases in this descriptive study. CIHI receives data from participating acute care hospitals on all inpatient separations in Canada (i.e., discharges, transfers, or death).13 Few hospitals in Quebec send discharge data to CIHI; therefore data from the Province of Quebec are not included in the present study. This study does, however, include data from Manitoba, which includes slightly more than one third of hospital deliveries over the study period, and Nova Scotia, which sent more than one third of all discharge data from 1990 to 1993 and complete data for 1994 to 1997.12,14 All deliveries were extracted from the complete set of hospital separations for the 1990 to 1997 period using Case Mix Group codes from the CIHI complexity grouping methodology for the mother's records.13,14 Abruptio placentae cases were coded in the database using the International Classification of Diseases -Ninth Revision (ICD-9: 641.2).15 Deliveries that are registered in the Discharge Abstract Database have a gestational age of 20 weeks or greater.

Births from the Northwest Territories and Yukon were combined into one category due to small numbers of cases. Urban and rural status was determined using the second digit of the Forward Sortation Area segment of the postal code, where ‘0' indicated a rural residence and all others indicated an urban residence. Observations with invalid postal codes (approximately 2% of total records in the database) were not included in the final cohort.

Rates are calculated for all abruptio placentae cases and for those abruptio placentae cases ending in stillbirth (abruptio-stillbirth). The occurrence of abruptio-stillbirth cases was analyzed separately as a proxy for a complete placental abruption, since a strong association has been shown between the degree of placental separation and fetal death.⁴ Case-fatality rates were expressed as number of the abruptio-stillbirth cases per 100 total abruptio placentae cases. All data were analyzed using the Statistical Analysis System software package (version 8.0, SAS Institute, Inc., Cary, NC) and exact binomial 95% CI were calculated for all rates.

Results

The cohort of births from which rates were calculated comprised 2,162,815 hospital deliveries with singleton live births or stillbirths among nine Canadian provinces (excluding Quebec) and two territories between 1990 and 1997. Table 1 presents the rate of abruptio placentae and abruptio-stillbirth for singleton deliveries by year. Rates of abruptio placentae increased over the first six years of the study period, from 10.90 (95% CI 10.51-11.30) per 1,000 deliveries in 1990 to a high of 12.05 (11.64-12.47) cases per 1,000 deliveries in 1996. There was a slight decrease in the rate for 1997. Over the entire period of observation, the abruptio placentae rate was 11.25 (95% CI 11.1-11.4) per 1,000 singleton births. The rate of abruptio-stillbirth cases was 0.78/1,000 (95% CI 0.75-0.82) over the period and was relatively consistent across the period of observation. The case-fatality rate was approximately 7% over this time period.

The rates of abruptio placentae varied across provinces (Table 2), with rates below 8/1,000 in Newfoundland and Prince Edward Island and rates above 13/1,000 in Nova Scotia, Saskatchewan, and Yukon/ North West Territories. This provincial pattern of rates was similar for abruptio-stillbirth, with the exception of Saskatchewan, which had a high rate of abruptio placentae and a moderate rate ending in stillbirth. A case-fatality rate statistically higher (p < 0.05) than the rest of Canada was observed in New Brunswick, and the case-fatality rate in Saskatchewan was significantly lower than that in the rest of Canada. Although the highest case-fatality rate was observed in Prince Edward Island, this was based on a small number of events and was not statistically different from the case-fatality rate for the rest of Canada (p=0.17).

The rate of abruptio placentae is highest for mothers in the two older age groups, aged 35-39 years and aged 40 years and over (Table 3). However, the case-fatality rate is highest in women under 20 years of age (11%), and it decreases with age to 5% in those aged 40 years and over.

Table 4 presents rates by urban and rural status for abruptio placentae cases and abruptio-stillbirth. Rates of abruptio placentae are similar for urban (11.33/1,000, 95% CI 11.17-11.49) and rural (10.95/1,000, 95% CI 10.65-11.25) residence. However, case-fatality rates are higher for rural residence (7.89%) compared to urban residence (6.72%) (p=0.004).

Discussion

This descriptive analysis of over two million singleton deliveries recorded in the Discharge Abstract Database indicates that the incidence of abruptio placentae in Canada was approximately 1% over the 1990-1997 period. This abruptio placentae rate is comparable to rates found in other population-based studies. A large cohort study in the US found an overall rate of 11.5/1,000 over the 1979-1987 period,⁷ while in Norway, a rate of 6.6/1,000 was seen over a longer study period, from 1967-1991.¹⁶

The increase over time has also been observed in an earlier study. In Norway, the abruptio placentae rate in 1971 was 5.3/1,000, while in 1990, this rate rose to 9.1/1,000.¹⁶ A study by Saftlas et al. in the US saw a similar increase, from 8.2/1,000 in 1979 to 11.5/1,000 in 1985.⁷ Saftlas speculated that a change of rates such as that observed in our study may reflect a true increase in the rate of abruptio placentae, or it may result from changes in detecting abruptio placentae cases, i.e., an increase in partial or marginal abruptions, due to better ultrasound technology or more sensitive case definition. Similar detection changes may account for the rise in rates in Canada. However, the increase over time may also be real. The prevalence report in previous studies^{1,4,9,10,17-19} of a number of the risk factors associated with abruptio placentae have been on the decline. For example, smoking rates have decreased in Canada over the past decade.²⁰ Live birth rates have also declined among women of reproductive age in Canada, from 61.1 per 1,000 in 1981 to 51.1 per 1,000 in 1997.¹² Although these factors were not available for analysis in our data, we speculate that the observed increased rates of abruptio placentae are attributable to other factors, including changes in ascertainment and reporting.

Our reported case-fatality rates are comparable to those reported in the US (7.1%) between 1979 and 1987;7 however, the US study did not examine trends in this rate by year, region, or maternal age.

Factors that could account for associations between abruptio placentae rates and increased maternal age may include increased parity and prior abruptio placentae.⁹ Pre-pregnancy hypertension, another risk factor for abruptio placentae,¹⁰ also increases with age.¹⁹ The high proportion of cases ending in a stillbirth among the youngest maternal age groups could be related to underutilization of prenatal health care services in this age group.²¹

Findings on the association between maternal age and abruptio placentae have been inconsistent in the literature. A case-control study in the US with 884 cases found no association,³ while an Italian case-control study and a US cohort study both found significant associations, either with increasing age¹¹ or with maternal age greater than 35.¹⁰ The Saftlas study on incidence of abruptio placentae in the US examined the rates of abruptio placentae by age for two specific time periods, 1979-1982 and 1983-1991, for white versus black women.⁷ Rates for the first period remained relatively consistent among white women, but in the more recent period, abruptio placentae rates were highest among teenagers, at approximately 13/1,000, then dropped to 8/1,000 for women between 20-24, and finally increased to 10/1,000 and 11/1,000 for the 25-29 and 30+ age groups, respectively. A similar pattern, but at slightly lower rates, was seen among black women for the 1979-1982 period, while in the second study period, rates fell from 13/1,000 for teenagers, 12/1,000 for the 20-24 group, 11/1,000 for the 25-29 age group, and then increased to 13/1,000 for women aged 30 or older.⁷ Our study reported a similar U-shaped relationship with age, with even higher rates for women over 40.

No literature was found that reported regional variations of abruptio placentae rates within a population. The variations in rates observed in several provinces in Canada could be attributable to differences in unmeasured risk factors (e.g., some environmental contaminant) or due to differences in ascertainment and reporting.

The accurate identification of abruptio placentae cases is the primary limitation in this study. The number of abruptio placentae cases is influenced by potential under-reporting of mild cases that may be less likely to be recorded in the administrative database utilized in this study. In addition, the definitive diagnosis of abruptio placentae may occur after delivery when the placenta is sent to the pathology laboratory, so information on the final diagnosis may not be available when data abstracting is conducted. This would underestimate the abruptio placentae rate reported in our study.

However, a quality check of the data had positive findings. Wen et al. examined the CIHI database used in this study and found that the number of illogical and out-of-range values were few.¹⁴ Furthermore, for most adverse pregnancy conditions and outcomes, including abruptio placentae, Wen et al. found the prevalence to be within a reasonable range of that reported in the literature.¹⁴ An additional strength of the study is the very large number of births that were analyzed (n = 2,162,815) over the 1990-1997 time period, using quality data.

Thus, these study results provide a baseline reference for rates of abruptio placentae in Canada as well as those that specifically end in stillbirth. The significant variation that was identified in abruptio placentae rates and case-fatality rates according to time and maternal age underline the need for further investigation into this condition. Because abruptio placentae is related not only to stillbirth but also to perinatal death and sequelae from perinatal asphyxia, inclusion of abruptio placentae as an indicator in the Canadian Perinatal Surveillance System should be considered to better understand the distribution of this event and to facilitate future observational investigations on the condition.

Acknowledgements

We wish to thank the Health Surveillance and Epidemiology Division of Health Canada for providing the access to CIHI's Discharge Abstract Database.

References

Author References

Teresa Broers, Department of Community Health and Epidemiology, Queen's University, Kingston, Ontario, Canadaand Health Surveillance and Epidemiology Division, Population and Public Health Branch, Health Canada, Ottawa, Ontario, Canada Will D King, Department of Community Health and Epidemiology, Queen's University, Kingston, Ontario, Canada Tye E Arbuckle, Department of Community Health and Epidemiology, Queen's University, Kingston, Ontario, Canada,and Biostatistics and Epidemiology Division, Healthy Environments and Consumer Safety Branch, Health Canada, Ottawa, Ontario, Canada Shiliang Liu, Health Surveillance and Epidemiology Division, Population and Public Health Branch, Health Canada, Ottawa, Ontario, Canada Correspondence: Will D King, Department of Community Health and Epidemiology, Abramsky Hall, Queen's University, Kingston, Ontario, Canada K7L 3N6; Fax: (613) 533-6686; E-mail: kingw@post.queensu.ca