Special Report on Maternal Mortality and Severe Morbidity in Canada Enhanced Surveillance: The Path to Prevention

Contributors

Principal Authors

I. D. Rusen, MD, MSc, FRCP
Robert Liston, MD, ChB, FRCSC, FRCOG

Contributing Authors

Shi Wu Wen, MB, PhD
Sharon Bartholomew, MHSc

Expert Review Committee

Karen Acheson, MD, CCFP
Maureen Heaman, RN, MN, PhD
Robert Liston
Dana Paquette, MSc
I.D. Rusen Reg Sauve, MD, MPH, FRCPC

Maternal and Infant Health Section Staff

Sharon Bartholomew
Ernesto Delgado
Patric Fournier-Savard, MSW
Ling Huang, MD, MSc
Adèle Lemay-Jones
Gina Marin
Catherine McCourt, MD, MHA, FRCPC
Louie McLeod
Hajnal Molnar-Szakács, MD, MSc
Dana Paquette
Jocelyn Rouleau

Maternal Health Study Group,
Canadian Prenatal Surveillance System
(External Members)

Thomas F. Baskett, MB, FRCS
Margaret Cyr
Kinga David, MHSc
William Fraser, MD, MSc, FRCSC
Maureen Heaman K.S. Joseph, MD, PhD
Robert Kinch, MD
Michael Kramer, MD
Robert Liston
Beverley O'Brien, RM, DNSc
Karen Roberts
Reg Sauve
Susan Taylor-Clapp, RN, MSc

Acknowledgements

The Canadian Perinatal Surveillance System wishes to acknowledge the participation of the Offices of the Provincial and Territorial Chief Coroners and Chief Medical Examiners, and the four established Provincial and Territorial Maternal Death Review Committees. Their collaboration was essential for this project and it is greatly appreciated.

We would like to thank Dr. Alison Macarthur, Associate Professor, Department of Anesthesia, University of Toronto, for her constructive comments on an earlier draft of this report.

We are also grateful to Dr. William Fraser and Dr. Alexandre Dumont of the Department of Obstetrics and Gynaecology, University of Montreal, for their careful review of the French language version of this document.

Introduction

Maternal mortality has been used as a measure of the status of women, their access to health care and the capacity of the health care system to respond to their needs.¹ An estimated 529,000 maternal deaths occurred worldwide in 2000. Levels of maternal mortality show wide disparities between countries and regions, with 13 countries accounting for 67% of all maternal deaths. The highest estimated maternal mortality ratios (MMRs) in the world in 2000 were from Afghanistan and Sierra Leone at 1,900 and 2,000 maternal deaths per 100,000 live births, respectively.²

In response to this staggering burden of maternal deaths, the reduction of maternal mortality has been highlighted as an important goal for several initiatives, including the United Nations' Millennium Development Goals (MDGs). The MDGs aim for the global reduction of maternal mortality by 75% from 1990 levels (global estimate of 430 per 100,000 live births¹) by 2015.³

In many developed countries, including Canada, maternal mortality has already decreased markedly during the 20th century.^4,5

The most recent MMR estimates for selected countries with good death registration and good attribution of cause are presented in Figure 1. It is evident that Canada's MMR (based on national vital statistics) is one of the lowest in the world. Nonetheless, we cannot be complacent about our levels of maternal mortality. Pregnancy-related mortality surveillance in other countries has demonstrated increased mortality among disadvantaged groups in the population.^5,6 Overall MMRs may mask important elevated risks for particular sub-groups of the Canadian population. Despite Canada's relatively low MMR, opportunities to further reduce the number of women dying as a result of their pregnancy must be explored.

Specific strategies to reduce maternal mortality are dependent on the baseline burden of maternal deaths. In the worst affected regions of the world, basic safe motherhood interventions are urgently needed and are likely to have a profound impact on the number of women dying of pregnancy-related causes.

These interventions are described in detail elsewhere.⁷ In developed countries, with significantly lower levels of maternal mortality, enhanced surveillance efforts are a necessary starting point to identify prevention opportunities to further reduce mortality.

Limitations of vital registry systems to is a accurately track maternal deaths have been well documented.^2,4 Maternal deaths may be missed and misclassified, even in countries with complete vital registration coverage and universal medical certification of death. Published estimates of MMRs often adjust officially reported rates of maternal death to account for the problem of under-reporting.²However, to better understand the persistent causes of maternal mortality and identify preventive opportunities, more countries are now seeking to enhance statistical, quantitative data with in-depth qualitative information.²

In-depth case investigations and reports can increase awareness of the occurrence and preventability of maternal deaths, recommend specific actions to improve quality of care and encourage the development and enhancement of maternal death reporting systems.^2,6,8

Models of enhanced surveillance and in-depth case investigation and reporting are available from several countries:

• In the United Kingdom (U.K.), the Confidential Enquiry into Maternal Deaths in the United Kingdom was initiated in England and Wales in 1952. Now covering the U.K. as a whole, the aim of the enquiry is "to form part of the Government's agenda for clinical governance to help ensure that all pregnant women and recently delivered women receive the best possible care delivered in appropriate settings and taking account of individual needs."⁶ The report of the enquiry is produced every three years and provides detailed case information, as well as key recommendations based on the findings. Of critical value to the enquiry is a government requirement that all maternal deaths should be subject to the confidential enquiry and that all health professionals have a duty to provide the information required.⁶ The U.K. enquiry and report are viewed as the "gold standard" for maternal mortality surveillance.⁸
• The Report on Maternal Deaths in Australia has been produced triennially since 1964. The report aims to provide quality assurance of maternal care during pregnancy and the puerperium, to provide public health surveillance of maternal deaths and to refine a standard for the reporting of maternal deaths.⁸ The Australian report also provides individual case-level detail, as well as comments on preventive opportunities. In contrast to the centrally implemented U.K. enquiry, the Australian report is the product of various state and territorial Confidential Death Enquiries, a National Hospital Morbidity Database and a National Mortality Database.⁸ Consequently, the data incorporated into the report are variable both detail and quality.⁸
• In the United States, the Pregnancy Mortality Surveillance System (PMSS) was established in 1987 and provides ongoing surveillance of all pregnancy-related deaths reported through individual state health departments, maternal mortality review committees, media and individual providers.⁹ In addition to careful review of the immediate and underlying cause of death, the PMSS obtains additional information primarily from maternal death certificates and matched birth and fetal death certificates. The published reports comment on the available socio-demographic factors and their relation to the risk of maternal death.⁵ However, unlike the U.K. and Australian reports, the PMSS does not conduct individual reviews of the medical circumstances surrounding death and no individual case-level detail is provided.

In Canada, at the local level, deaths that occur in hospitals are usually the subject of thorough investigation by hospital review committees.⁴ However, the scope and outcome of these internal reviews are not routinely reported outside the institution. At the provincial and territorial levels, the conduct of maternal death investigations is likely variable. Furthermore, these provincial and territorial practices have not been previously summarized. Finally, at the national level, there is currently no systematic mechanism in place to synthesize and report on maternal deaths in Canada.

An issue closely related to maternal death investigation and reporting is the surveillance of severe maternal morbidity or "near miss" - i.e., women who were at serious risk of death but survived.⁶ As maternal deaths have become increasingly rare in industrialized countries, severe maternal morbidity surveillance has been proposed as a supplementary indicator to maternal mortality to monitor the quality of maternity care.^6,10-12

Severe maternal morbidity surveillance provides greater numbers of cases to review and to form the basis of conclusions and recommendations.⁶ Several of the existing national maternal death review committees have acknowledged the importance of enhancing severe maternal morbidity surveillance efforts.^6,8 The most recent Canadian Perinatal Health Report has examined rates of amniotic fluid embolism (AFE) and post-partum hemorrhage. ¹³ However, there are no published reports on the nation-wide occurrence of overall severe maternal morbidity in Canada.

The Canadian Perinatal Surveillance System (CPSS) was developed by Health Canada in 1995, with a mandate to contribute to improved health for pregnant women, mothers and infants in Canada through ongoing monitoring and reporting on perinatal health determinants and outcomes.¹³ One of CPSS's three study groups, the Maternal Health Study Group (MHSG), focuses on key behaviours, health services and outcomes related to maternal health. The MHSG identified enhanced surveillance of maternal mortality and severe maternal morbidity as priority areas for maternal health surveillance in Canada. In this regard, the MHSG has undertaken studies on under-reporting of maternal mortality in Canada,⁴ cause-specific mortality during and after pregnancy,¹⁴ maternal readmission,¹⁵ and a comparison of maternal mortality and morbidity between trial of labour and elective cesarean section among women with previous cesarean delivery.¹⁶ The MHSG, with the support of the Society of Obstetricians and Gynaecologists of Canada (SOGC), initiated this special report on maternal mortality and severe maternal morbidity in Canada with the following four main objectives:

• to describe the current provincial and territorial activities to investigate and report on maternal deaths
• to review and summarize available individual case-level reports of maternal deaths occurring in Canada between 1997 and 2000
• to describe the occurrence and trends of severe maternal morbidity in Canada
• to make recommendations to enhance the surveillance of maternal mortality and severe maternal morbidity in Canada

Methods

This project was undertaken in five distinct stages

Provincial/Territorial-Level Maternal Death Review Activities Determined

In each province and territory, the chief coroner/medical examiner and the college of medicine or college of physicians and surgeons were contacted and asked whether or not a formal provincial or territorial maternal death review committee exists. If the answer was yes, representatives from these committees were contacted and details regarding the committees' activities and willingness to participate in the project were obtained. Where no formal review committee existed, the role of the provincial and territorial coroner(s)/ medical examiner(s) in maternal death investigations was explored, as well as their willingness to participate in this project.

List of Deaths Constructed from National Databases

As the first step in compiling maternal death case reports, a list of deaths during 1997-2000 was obtained from three administrative databases.

The first source was Statistics Canada's Canadian Vital Statistics System, which contains national databases of non-nominal data on live births, stillbirths and deaths.¹³ The process of death certification in Canada with specific reference to pregnancy has been described in detail elsewhere.⁴ Maternal deaths are those deaths that have been assigned an underlying cause of death code in chapter 11 ("Complications of pregnancy, childbirth and the puerperium") of the 9th Revision of the International Classification of Diseases(ICD-9)¹⁷ or chapter XV ("Pregnancy, childbirth and puerperium") of the 10th Revision of the International Classification of Diseases (ICD-10).¹⁸ From the Canadian Vital Statistics System, a preliminary list of deaths was constructed, containing the following variables: province of residence, age, date of death and single underlying cause of death.

The second data source utilized was the Discharge Abstract Database (DAD) from the Canadian Institute for Health Information (CIHI). The DAD captures hospital separation information -- transfer, discharge or death -- from the majority of Canada's acute care hospitals. Hospital separations from the province of Quebec and from parts of Manitoba were not included in the DAD during the study period for maternal deaths. The DAD contains non-nominal data on each hospitalization, including demographic and residence information, length of stay, most responsible diagnosis, secondary and co-morbid diagnoses, and procedures performed during the hospitalization.¹³ A combination of diagnostic and procedure codes was utilized to identify all deliveries, as well as abortions and ectopic pregnancies. Any record of these events combined with a hospital separation of "death" was used to capture hospital-based deaths.

Third, hospital separations from the province of Quebec were examined utilizing the Système de maintenance et d'exploitation des données pour l'étude de la clientèle hospitalière (MED-ÉCHO).

Finally, deaths obtained from the two hospitalization databases were combined with the deaths from Statistics Canada's Canadian Vital Statistics System to obtain the list of database deaths for the period 1997-2000.

Individual Case-Level Data Obtained from Provincial/Territorial Death Review Committees or Coroners/Medical Examiners

Using the list of deaths obtained from the database searches, provinces and territories were approached for information on cases occurring in their jurisdiction during 1997-2000. In provinces and territories with established maternal death review committees, these bodies were contacted. Where such committees did not exist, the chief coroner/medical examiner or designate was approached. (In the Northwest Territories, information regarding maternal deaths was obtained from the Chief Coroner. Details regarding the existing maternal death review committee were not obtained until after the case review was completed.)

Provincial and territorial contacts were asked to confirm individual cases and provide information for these cases. Provinces and territories were also asked to identify and provide information on maternal deaths that were not ascertained from the databases. Provinces and territories with no deaths recorded in the databases during 1997-2000 were still contacted to identify any maternal deaths that may have been missed. Provincial and territorial contacts were asked to provide actual maternal death investigation reports or to complete a data collection tool (see Appendix A) which was constructed to capture key case information. No nominal information was requested.

Expert Review and Analysis of Maternal Deaths

An expert review committee was convened with representation from obstetrics, obstetric nursing, coroner/medical examiner's system, provincial/ territorial maternal death review committees and national surveillance and epidemiology. Only cases with available coroner/medical examiner or provincial/territorial death review committee reports, or completed data collection forms were included in the formal review process. Information for each case was reviewed and discussed in detail. Classification of maternal death, as well as designation of principal cause, was carried out for each case. Definitions used to classify deaths are presented in Table 1.

Table 1 Definitions of maternal deaths

Maternal deaths:a deaths of women while pregnant or within 42 days of the termination of the pregnancy, irrespective of the duration and site of the pregnancy, from any cause related to or aggravated by the pregnancy or its management but not from accidental or incidental causes. Direct obstetric deaths:a maternal deaths resulting from obstetric complications of the pregnant state (pregnancy, labour and puerperium); interventions, omissions or incorrect treatment; or a chain of events resulting from any of the above. Indirect obstetric deaths:a maternal deaths resulting from previous existing disease or disease that developed during pregnancy, which was not due to direct obstetric causes but which was aggravated by the physiologic effects of pregnancy. Incidental deaths:b deaths due to conditions occurring during pregnancy, where the pregnancy is unlikely to have contributed significantly to the death, although it is possible to postulate a distant association. Late maternal deaths:c deaths of women from direct or indirect obstetric causes occurring between 42 days and one year after termination of pregnancy.

a Definitions used in both ICD-9 and ICD-10. b Previously referred to as fortuitous deaths;"incidental" as defined in the Report on Maternal Deaths in Australia, 1994-96.8 c New ICD-10 category.

In addition to the new category of late maternal death, several other differences in the coding and classification of maternal deaths exist between ICD-9 and ICD-10 and have been outlined elsewhere.⁴The study period for this report includes years of both ICD-9 (1997-1999) and ICD-10 (2000) classification of deaths in Canada. As a result, the classification of death for this report was determined by consensus of the expert review committee and not with reference to one specific disease classification system. For example, deaths from cerebrovascular disorders during pregnancy, or within 42 days of termination of pregnancy, are classified as direct maternal deaths under ICD-9 but as indirect maternal deaths under ICD-10.⁴ The expert review committee classified cererbrovascular deaths based on individual case details and usually classified these deaths as direct maternal deaths for the purposes of this report.

Consensus was also reached as to the principal cause of death for each case. For example, deaths from intra-cranial hemorrhage (ICH) that occurred in conjunction with a diagnosis of pre-eclampsia were attributed to pre-eclampsia, while cases of ICH with no indication of pre-eclampsia were categorized as ICH deaths. Where such determinations were not possible based on the available information, additional details were sought from the respective coroner/medical examiner's office or provincial/territorial death review committee.

Following the classification and designation of cause for all reviewed cases, data were entered into EpiData (version 3.0, the EpiData Association, Odense, Denmark, 2003) and analyzed using Epi Info (version 6.04d, United States Centers for Disease Control and Prevention and World Health Organization, 2001). In keeping with international approaches, the overall maternal mortality ratio (MMR) was calculated as the combined number of direct and indirect maternal deaths per 100,000 live births.¹⁷ In addition, the direct, indirect, age-specific and cause-specific MMRs were calculated. Live birth counts for MMR calculations were obtained from Statistics Canada's Canadian Vital Statistics System.

Occurrence of Severe Maternal Morbidity

Data for women admitted to hospital for deliveries for fiscal years 1991-1992 to 2000-2001 were abstracted from CIHI's DAD using a combination of codes as discussed above. Hospital separations from the province of Quebec and from parts of Manitoba and Nova Scotia were not included in the DAD during this period of study. Therefore, these provinces are not included in the severe maternal morbidity analysis. In contrast to the methods for ascertaining deaths, the severe maternal morbidity analysis excluded early pregnancy events, i.e., ectopic pregnancies and abortions that did not result in a live birth or stillbirth outcome.

The conditions and procedures suggestive of severe maternal morbidity were selected by an expert panel (obstetrics, obstetric nursing, national surveillance and perinatal epidemiology) of the CPSS's Maternal Health Study Group. These include: amniotic fluid embolism (AFE); obstetrical pulmonary embolism (not including AFE); eclampsia; shock (obstetrical, septic and other); pulmonary, cardiac and central nervous system (CNS) complications of anaesthesia; cerebrovascular disorders in the puerperium (including intra-cranial venous sinus thrombosis); uterine rupture; adult respiratory distress syndrome; pulmonary edema; myocardial infarction; acute renal failure following labour and delivery; cardiac arrest/failure or cerebral anoxia following obstetrical surgery; severe post-partum hemorrhage requiring hysterectomy or transfusion; and assisted ventilation.

To measure the overall burden of severe maternal morbidity, a rate per 1,000 deliveries was calculated, with the numerator being the number of patients who had any of the above conditions or procedures. Preliminary analysis of the data identified a coding error for myocardial infarction in hospital separation records for deliveries in Saskatchewan.

Therefore, Saskatchewan records were excluded from the calculation of myocardial infarction rates. For the calculation of total severe maternal morbidity rates, Saskatchewan deliveries were included, but Saskatchewan myocardial infarction cases were treated as non-cases. To ensure hemorrhage cases were truly severe cases, only records with post-partum hemorrhage associated with either hysterectomy or transfusion were counted. The ICD-9 and the Canadian Classification of Diagnostic, Therapeutic and Surgical Procedures (CCP) codes for all relevant conditions and procedures are listed in Appendix B.

Overall rates of severe maternal morbidity in Canada were calculated for the decade of study. Also, rates of severe maternal morbidity in 1991-1993 were compared to rates in 1998-2000. Relative risks and 95% confidence intervals were calculated as a measure of secular trends, using 1991-1993 data as the reference.

[ Previous ] [ Next ]