Advertisement
JOGC

Severe Maternal Morbidity Associated With Maternal Birthplace: A Population-Based Register Study

  • Marcelo L. Urquia
    Correspondence
    Corresponding Author: Dr. Marcelo L. Urquia, Manitoba Centre for Health Policy, University of Manitoba, Winnipeg, MB
    Affiliations
    Manitoba Centre for Health Policy, Department of Community Health Sciences, Rady Faculty of Health Sciences, University of Manitoba, Winnipeg, MB

    Centre for Urban Health Solutions, Li Ka Shing Knowledge Institute, St. Michael's Hospital, Toronto, ON

    Institute for Clinical Evaluative Sciences, Sunnybrook Hospital, Toronto, ON
    Search for articles by this author
  • Susitha Wanigaratne
    Affiliations
    Centre for Urban Health Solutions, Li Ka Shing Knowledge Institute, St. Michael's Hospital, Toronto, ON

    Institute for Clinical Evaluative Sciences, Sunnybrook Hospital, Toronto, ON
    Search for articles by this author
  • Joel G. Ray
    Affiliations
    Institute for Clinical Evaluative Sciences, Sunnybrook Hospital, Toronto, ON

    Department of Medicine, St. Michael's Hospital, Toronto, ON
    Search for articles by this author
  • K.S. Joseph
    Affiliations
    Department of Obstetrics and Gynaecology, Faculty of Medicine, University of British Columbia, Vancouver, BC

    School of Population and Public Health, Faculty of Medicine, University of British Columbia, Vancouver, BC
    Search for articles by this author
Open AccessPublished:July 18, 2017DOI:https://doi.org/10.1016/j.jogc.2017.05.012

      Abstract

      Objective

      This study sought to quantify the risk of severe maternal morbidity (SMM) according to maternal country of birth in Canada.

      Methods

      The study analyzed 1 252 543 in-hospital deliveries of Ontario residents discharged between April 1, 2002, and March 31, 2012. The main outcome measure was a composite indicator of SMM used for surveillance. The top 10 most common component conditions were also evaluated. Maternal country of birth and other immigration characteristics were obtained through linkage with official immigration records. We used modified Poisson regression with generalized estimating equations to assess associations according to maternal country of birth.

      Results

      Overall, immigrant women (N = 335 544) did not differ from Canadian-born women (n = 916 999) in SMM rates (12.1 vs. 12.0 cases per 1000 deliveries, respectively). However, SMM varied substantially according to maternal region of birth, from 9.2 cases per 1000 deliveries among immigrants from Western countries to 23.0 cases per 1000 deliveries among immigrants from Sub-Saharan Africa. Even larger variations were found when immigrants were categorized by their specific countries of birth. The top 10 contributing conditions to SMM among Canadian-born women were also the main contributors among immigrant subgroups. The notable exception was HIV infection, the top contributor among immigrants from Sub-Saharan Africa, whose rate of HIV infection was 43 times that of Canadian-born women (95% CI 34.39–55.23). After excluding HIV cases, disparities in SMM were largely reduced among Sub-Saharan African women but did not disappear.

      Conclusion

      There is large heterogeneity in SMM and its component conditions among Canadian immigrants depending on country of origin.

      Résumé

      Objectif

      Cette étude avait pour but de quantifier le risque de morbidité maternelle sévère (MMS) au Canada, selon le pays d'origine de la mère.

      Méthodologie

      Nous avons analysé 1 252 543 accouchements en milieu hospitalier, dont les mères étaient des résidentes ontariennes ayant obtenu leur congé entre le 1er avril 2002 et le 31 mars 2010. L'indicateur de résultat principal était un indicateur composite de MMS utilisé à des fins de surveillance. Les 10 facteurs contributifs les plus fréquents ont également été évalués. Des données relatives aux pays d'origine des femmes et à l'immigration ont été obtenues à partir des documents d'immigration officiels. Nous nous sommes servis d'un modèle de régression de Poisson et d'équations d'estimation généralisées pour évaluer les résultats obtenus en fonction du pays d'origine des femmes.

      Résultats

      Dans l'ensemble, les immigrantes (n = 335 544) et les Canadiennes de naissance (n = 916 999) présentaient un taux de MMS semblable (12,1 cas c. 12,0 pour 1 000 accouchements, respectivement). Or, la MMS variait considérablement selon la région de naissance de la mère, allant de 9,2 cas pour 1 000 accouchements chez les immigrantes venant de pays occidentaux à 23,0 cas pour 1 000 accouchements chez celles venant d'Afrique subsaharienne. Des variations encore plus grandes ont été observées lorsque les immigrantes ont été classées selon leur pays de naissance. Les 10 principaux facteurs contributifs de la MMS chez les Canadiennes de naissance étaient les mêmes que chez les sous-groupes d'immigrantes. L'infection au VIH, principal facteur contributif chez les immigrantes d'Afrique subsaharienne, était cependant l'exception importante : le taux d'infection au VIH était 43 fois plus élevé chez ces femmes que chez les Canadiennes de naissance (IC à 95 % : 34,39–55,23). Une fois les cas de VIH exclus, l'écart de MMS observé était beaucoup moins grand dans le groupe des femmes d'Afrique subsaharienne, mais toujours présent.

      Conclusion

      La MMS et ses facteurs contributifs sont très hétérogènes chez les immigrantes canadiennes et varient selon le pays d'origine des femmes.

      Key Words

      Maternal mortality varies more between countries than within countries. Non-industrialized countries account for approximately 99% of the global burden of maternal deaths. Increasingly, immigrants originate from non-industrialized countries where maternal mortality is roughly 20 times higher than in immigrant-receiving industrialized countries.
      • Boyd M.
      • Vickers M.
      100 years of immigration in Canada.
      WHO, UNICEF, UNFPA, World Bank Group and the United Nations Population Division
      Trends in maternal mortality: 1990 to 2015. Estimates by WHO, UNICEF, UNFPA, World Bank Group and the United Nations Population Division.
      However, a recent meta-analysis reported that migrant women in Western European countries have twice the risk of maternal mortality than native-born women.
      • Pedersen G.S.
      • Grontved A.
      • Mortensen L.H.
      • et al.
      Maternal mortality among migrants in Western Europe: a meta-analysis.
      The shrinkage in the disparity among immigrants from the destination to the receiving country may reflect several factors, including self-selection of healthier women for migration, selective immigration admission policies,
      • Urquia M.L.
      • Vang Z.M.
      • Bolumar F.
      Birth outcomes of Latin Americans in two countries with contrasting immigration admission policies: Canada and Spain.
      • Wanigaratne S.
      • Cole D.C.
      • Bassil K.
      • et al.
      Contribution of HIV to maternal morbidity among refugee women in Canada.
      greater material resources for everyday life, and access to better primary and obstetric care in receiving societies,
      • Urquia M.L.
      • Glazier R.H.
      • Mortensen L.
      • et al.
      Severe maternal morbidity associated with maternal birthplace in three high-immigration settings.
      despite cultural barriers that may potentially preclude immigrants from fully benefiting from available health care services.
      • van den Akker T.
      • van Roosmalen J.
      Maternal mortality and severe morbidity in a migration perspective.
      As maternal mortality has become a rare event in industrialized countries in recent decades, research and surveillance have increasingly focused on various indices of severe maternal morbidity (SMM) or “near miss” to capture more accurately the serious and life-threatening conditions affecting pregnancy, labour, and the postpartum period that could potentially result in maternal death.
      • Joseph K.S.
      • Liu S.
      • Rouleau J.
      • et al.
      Severe maternal morbidity in Canada, 2003 to 2007: surveillance using routine hospitalization data and ICD-10CA codes.
      • Roberts C.L.
      • Cameron C.A.
      • Bell J.C.
      • et al.
      Measuring maternal morbidity in routinely collected health data: development and validation of a maternal morbidity outcome indicator.
      • Sousa M.H.
      • Cecatti J.G.
      • Hardy E.E.
      • et al.
      Severe maternal morbidity (near miss) as a sentinel event of maternal death: an attempt to use routine data for surveillance.
      • Kayem G.
      • Kurinczuk J.
      • Lewis G.
      • et al.
      Risk factors for progression from severe maternal morbidity to death: a national cohort study.
      Some of these SMM indices have been created using multiple diagnoses and procedures recorded in population-based administrative databases.
      • Joseph K.S.
      • Liu S.
      • Rouleau J.
      • et al.
      Severe maternal morbidity in Canada, 2003 to 2007: surveillance using routine hospitalization data and ICD-10CA codes.
      • Roberts C.L.
      • Cameron C.A.
      • Bell J.C.
      • et al.
      Measuring maternal morbidity in routinely collected health data: development and validation of a maternal morbidity outcome indicator.
      • Sousa M.H.
      • Cecatti J.G.
      • Hardy E.E.
      • et al.
      Severe maternal morbidity (near miss) as a sentinel event of maternal death: an attempt to use routine data for surveillance.
      • Wahlberg A.
      • Roost M.
      • Haglund B.
      • et al.
      Increased risk of severe maternal morbidity (near-miss) among immigrant women in Sweden: a population register-based study.
      Such comprehensive definitions of composite SMM are relatively frequent (relative to maternal death), thereby making it possible to study variations in SMM within population subgroups. Despite this advantage, previous studies did not have large enough study sizes to examine SMM in immigrant women from specific countries of origin (women could be aggregated only by world regions) or to study specific components of the composite indicator, thus compromising specificity.
      • Urquia M.L.
      • Glazier R.H.
      • Mortensen L.
      • et al.
      Severe maternal morbidity associated with maternal birthplace in three high-immigration settings.
      • Wahlberg A.
      • Roost M.
      • Haglund B.
      • et al.
      Increased risk of severe maternal morbidity (near-miss) among immigrant women in Sweden: a population register-based study.
      • Knight M.
      • Kurinczuk J.J.
      • Spark P.
      • et al.
      Inequalities in maternal health: national cohort study of ethnic variation in severe maternal morbidities.
      • Zwart J.J.
      • Jonkers M.D.
      • Richters A.
      • et al.
      Ethnic disparity in severe acute maternal morbidity: a nationwide cohort study in the Netherlands.
      The objectives of this study were to describe variations in SMM according to regions and specific countries of origin and to note the top component conditions contributing to SMM among immigrants to Ontario. Two features make the Province of Ontario an excellent setting for studying variations in immigrant health according to country of birth. Ontario has the largest concentration of immigrants in Canada and receives approximately 50% of the 250 000 immigrants arriving in Canada each year from most parts of the world. In addition, antenatal, hospital, and obstetric care services are provided free of charge to all legal permanent residents eligible under the provincial health insurance plan.

      Methods

      Study Design

      We used a retrospective cohort study design that linked population-based administrative databases at the Institute for Clinical and Evaluative Sciences in Toronto by using unique encoded identifiers.

      Study Population

      We included all in-hospital delivery episodes of women discharged between April 1, 2002, and March 31, 2012, who were Ontario residents with a valid health card number at the time of delivery.

      Data Sources

      We obtained records for women admitted to an Ontario hospital for childbirth from the Discharge Abstract Database of the Canadian Institute of Health Information. We used diagnosis and procedure codes (from the ICD-10-Canada
      Canadian Institute for Health Information
      Final report. The Canadian enhancement of ICD-10 (International Statistical Classification of Diseases and Related Health Problems, tenth revision).
      and the Canadian Classification of Health Interventions
      Canadian Institute for Health Information
      International Statistical Classification of Diseases and Related Health Problems, tenth revision. CCICD-10. 2009.
      ) to identify women who had any SMM.
      • Joseph K.S.
      • Liu S.
      • Rouleau J.
      • et al.
      Severe maternal morbidity in Canada, 2003 to 2007: surveillance using routine hospitalization data and ICD-10CA codes.
      • Liu S.
      • Joseph K.S.
      • Bartholomew S.
      • et al.
      Temporal trends and regional variations in severe maternal morbidity in Canada, 2003 to 2007.
      The labour and delivery information contained in this database has been shown to be accurate.
      • Joseph K.S.
      • Fahey J.
      Validation of perinatal data in the Discharge Abstract Database of the Canadian Institute for Health Information.
      The database also contains information on maternal age at delivery, number of previous live births, and birth plurality. Close to 99% of births in Ontario take place in hospitals.
      • Hutton E.K.
      • Cappelletti A.
      • Reitsma A.H.
      • et al.
      Outcomes associated with planned place of birth among women with low-risk pregnancies.
      Immigration characteristics, such as country of birth and other sociodemographic characteristics at the time of immigration, were obtained from the Ontario portion of the federal Immigration, Refugees and Citizenship Canada Permanent Resident Database, which is considered virtually complete and of high accuracy in most fields because of its administrative and legal use. Less than 1% of records in the database contained missing values. Approximately 90% of individuals in the database were matched to an Ontario resident with a valid provincial health care card. Unmatched individuals may have either moved out of the province (i.e., to other Canadian provinces or other countries or returned to their countries of origin) or may be misclassified as non-immigrants if they stayed in Ontario. To minimize misclassification in immigration status, we excluded women who first ever registered for the provincial health insurance plan after December 2010, which is the last month of arrivals included in the immigration database.

      Measures

      We used an SMM indicator developed for surveillance of SMM in Canada that has been described in detail elsewhere.
      • Joseph K.S.
      • Liu S.
      • Rouleau J.
      • et al.
      Severe maternal morbidity in Canada, 2003 to 2007: surveillance using routine hospitalization data and ICD-10CA codes.
      • Liu S.
      • Joseph K.S.
      • Bartholomew S.
      • et al.
      Temporal trends and regional variations in severe maternal morbidity in Canada, 2003 to 2007.
      In brief, this SMM composite includes a comprehensive set of maternal illnesses, from pre-existing medical conditions (e.g., heart disease, renal disease, and HIV infection) to pregnancy complications during labour and delivery (e.g., eclampsia and postpartum hemorrhage). Although some women may have more than one condition, SMM was considered present if a woman had at least one condition.
      Maternal country of birth was objectively assessed on the basis of legal documentation provided by immigrants during the immigration process. Other maternal characteristics included year of delivery, maternal age at delivery (15–19, 20–24, 25–29, 30–34, 35–39, 40–49), parity (0, 1, 2–3, 4+), multifetal pregnancy (yes vs. no), rural residence (yes vs. no), and neighbourhood income quintile (Q1 to Q5).

      Data Analysis

      The SMM rate was expressed as the number of cases of SMM per 1000 deliveries. To understand more clearly the contribution of each component condition to the frequency of SMM, we calculated proportional morbidity, which was expressed as the proportion of SMM cases affected by a specific component condition, and ranked conditions in descending order within each world region. In addition, we compared the rates of the top 10 component conditions in each world region with that of the Canadian-born population by using rate ratios (RRs) with 95% CIs obtained with modified Poisson regression models.
      • Zou G.
      A modified Poisson regression approach to prospective studies with binary data.
      Repeat pregnancies in the same woman were accounted for by including an additional maternal unique identification number (ID) cluster in the generalized estimating equations. We finally compared the rate of SMM for specific maternal countries of origin relative to that of Canadian-born women, before and after adjusting for maternal characteristics, and after excluding women with HIV infection (it is debatable whether this condition constitutes an SMM).
      The study was approved by the Institutional Review Board at Sunnybrook Health Sciences Centre and St. Michael's Hospital (#12-087) in Toronto.

      Results

      There were 1 323 947 deliveries during the study period, of which 71 404 (5.4%) were excluded. Among those excluded, 58 950 deliveries (82.6%) were to women whose eligibility date for Ontario health coverage was after December 2010, 101 deliveries were to immigrants whose country of birth could not be accurately classified, and 12 353 deliveries had missing data (including 8631 records with unknown income quintile, 469 deliveries with a maternal age <15 or ≥50, 1643 with unknown parity, 69 with unknown marital status, 6 with unknown language ability in English or French, 1585 with unknown GA, and 3137 with a GA <20 or >45 weeks). The final study population included 1 252 543 deliveries. Excluded women did not substantially differ in the risk of SMM (12.4 per 1000 deliveries) from women included in the study.
      Table 1 shows the characteristics of deliveries according to maternal birthplace. Overall, immigrants did not differ from non-immigrants in SMM rates (12.1 vs. 12.0 per 1000 deliveries, respectively). However, when immigrants were subdivided by world region of birth, variations increased, with SMM rates ranging from 9.2 per 1000 deliveries among immigrants from Europe and Western nations to 23.0 per 1000 deliveries among women from Sub-Saharan Africa. Other delivery characteristics also differed according to maternal birthplace. Teenage pregnancies were most common among women from the Caribbean and Latin America and least common among Asian groups. Women from Sub-Saharan Africa were the most likely to have had four or more previous live births. Preterm delivery was most common among women from the Caribbean. Most immigrant subgroups tended to reside in low-income neighbourhoods, particularly women from Sub-Saharan Africa, who were also more likely to be refugees. Although Caribbean women exhibited the highest proportion among all immigrants in their ability to communicate in English or French, they had the lowest educational attainment.
      Table 1Characteristics of women delivering in Ontario 2002–2012, by world region
      Data are presented as N (%) unless otherwise stated.
      CanadaAll immigrantsImmigrants by world region
      Europe and Western nationsSouth AsiaEast Asia, Southeast Asia, and PacificMiddle East and North AfricaSub Saharan AfricaCaribbeanLatin America
      Number of deliveries916 999335 54452 425104 45874 97033 68624 57526 43718 993
       SMM, all women10 964 (12.0)4054 (12.1)484 (9.2)1066 (10.2)964 (12.9)362 (10.7)566 (23.0)387 (14.6)225 (11.8)
      Singleton pregnancy899 183 (98.06)330 546 (98.51)51 454 (98.15)103 021 (98.62)74 163 (98.92)33 055 (98.13)24 133 (98.20)26 020 (98.42)18 700 (98.46)
       SMM10 214 (11.4)3848 (11.6)461 (9.0)1000 (9.7)926 (12.5)343 (10.4)538 (22.3)368 (14.1)212 (11.3)
      Multiple pregnancy17 816 (1.94)4998 (1.49)971 (1.85)1437 (1.38)807 (1.08)631 (1.87)442 (1.80)417 (1.58)293 (1.54)
       SMM,750 (42.1)206 (41.2)23 (23.7)66 (45.9)38 (47.1)19 (30.1)28 (63.3)19 (45.6)13 (44.4)
      Previous caesarean section116 541 (12.71)47 814 (14.25)5595 (10.67)16 569 (15.86)9513 (12.69)4836 (14.36)4756 (19.35)3470 (13.13)3075 (16.19)
      Current caesarean section251 265 (27.40)94 061 (28.03)12 871 (24.55)29 743 (28.47)21 108 (28.16)8864 (26.31)7997 (32.54)7581 (28.68)5897 (31.05)
      Maternal age
       <2042 217 (4.60)4316 (1.29)666 (1.27)589 (0.56)506 (0.67)468 (1.39)333 (1.36)1156 (4.37)598 (3.15)
       20 to 34693 640 (75.64)253 625 (75.59)38 609 (73.65)89 021 (85.22)49 965 (66.65)25 420 (75.46)17 484 (71.15)19 201 (72.63)13 925 (73.32)
       ≥35181 142 (19.75)77 603 (23.13)13 150 (25.08)14 848 (14.21)24 499 (32.68)7798 (23.15)6758 (27.50)6080 (23.00)4470 (23.53)
      Number of previous live births
       0421 016 (45.91)143 536 (42.78)24 582 (46.89)42 823 (41.00)36 555 (48.76)12 986 (38.55)7776 (31.64)10 590 (40.06)8224 (43.30)
       1324 584 (35.40)123 659 (36.85)19 864 (37.55)40 094 (38.38)29 636 (39.53)11 281 (33.49)7121 (28.98)9004 (34.06)6839 (36.01)
       2–3153 418 (16.73)60 159 (17.93)7289 (13.90)19 876 (19.03)8403 (11.21)8091 (24.02)7147 (29.08)5994 (22.67)3359 (17.69)
       4+17 981 (1.96)8190 (2.44)870 (1.66)1665 (1.59)376 (0.50)1328 (3.94)2531 (10.30)849 (3.21)571 (3.01)
      Preterm delivery (20 to 36 wk)70 370 (7.67)23 497 (7.00)3273 (6.24)7323 (7.01)4806 (6.41)2027 (6.02)1854 (7.54)2947 (11.12)1273 (6.70)
       Multiple pregnancy17 816 (1.94)4998 (1.49)971 (1.85)1437 (1.38)807 (1.08)631 (1.87)442 (1.80)417 (1.58)293 (1.54)
      Income Q1 quintile166 596 (18.17)112 466 (33.52)10 543 (20.11)39 183 (37.51)21 983 (29.32)11 578 (34.37)13 034 (53.04)10 581 (40.02)5564 (29.30)
      Income Q5 quintile176 642 (19.26)30 247 (9.01)9442 (18.01)5298 (5.07)7088 (9.45)3525 (10.46)1403 (5.71)1337 (5.06)2154 (11.34)
      Rural residence125 161 (13.65)3446 (1.03)1959 (3.74)258 (0.25)397 (0.53)135 (0.40)92 (0.37)71 (0.27)534 (2.81)
      Immigrant characteristics
       RefugeeN/A41 827 (12.47)6381 (12.17)10 562 (10.11)3805 (5.08)6159 (18.28)9531 (38.78)606 (2.29)4783 (25.18)
       Single at arrivalN/A136 547 (40.69)24 644 (47.01)28 577 (27.36)31 213 (41.63)10 614 (31.51)12 660 (51.52)7218 (27.30)9 620 (50.65)
       Maternal education at arrival
      Less than high schoolN/A125 471 (37.39)19 616 (37.42)33 275 (31.85)22 539 (30.06)11 386 (33.80)11 372 (46.27)18 266 (69.09)9017 (47.48)
      High school or/post-secondary diplomaN/A109 707 (32.70)17 447 (33.28)31 273 (29.94)26 441 (35.27)11 392 (33.82)10 341 (42.08)7253 (27.44)5560 (29.27)
      University diplomaN/A100 366 (29.91)15 362 (29.30)39 910 (38.21)25 990 (34.67)10 908 (32.38)2862 (11.65)918 (3.47)4416 (23.25)
       Knowledge of English or FrenchN/A206 382 (61.51)20 674 (61.37)56 416 (54.01)40 525 (54.05)20 674 (61.37)18 442 (75.04)26 272 (99.38)10 866 (57.21)
       Duration of residence, mean (SD)N/A6.55 (5.67)8.27 (6.50)5.00 (4.52)6.10 (5.30)5.70 (5.34)6.78 (5.47)10.40 (6.04)7.90 (6.57)
      a Data are presented as N (%) unless otherwise stated.
      Table 2 shows the rates of the component conditions of SMM that ranked among the top 10 in proportional morbidity in at least one immigrant subgroup. Proportional morbidity is shown in Supplemental Table S1. Overall, 13 conditions ranked in the top 10 in at least one immigrant subgroup. The top 10 conditions observed among the Canadian-born women were also important contributors to SMM among immigrants. Blood transfusion was the most common component among all women, with the only exception being women from Sub-Saharan Africa, for whom HIV infection ranked first, with a rate of 77.7 cases per 10 000 deliveries, the highest condition-specific rate among all groups. Obstetric embolism ranked 10th among women from Western countries, the Caribbean, and Latin America and 11th among non-immigrants. Sickle cell anemia with crisis ranked ninth and 18th among women from the Caribbean and Sub-Saharan Africa, respectively, but was rare among other groups.
      Table 2Rate (per 10 000 deliveries) and ranking of the top 10 conditions contributing to SMM, by maternal region of birth
      SMM componentCanadian-born (N = 916 999)All immigrants (N = 335 544)Europe and Western countries (N = 52 425)South Asia (N = 104 458)East Asia, Southeast Asia, and Pacific (N = 74 970)Middle East and North Africa (N = 33 686)Sub Saharan Africa (N = 24 575)Caribbean (N = 26 437)Latin America (N = 18 993)
      Blood transfusion (whole blood: platelet plasma, and red cell)56.6 (#1)55.3 (#1)39.1 (#1)47.8 (#1)67.6 (#1)48.4 (#1)76.9 (#2)65.8 (#1)63.2 (#1)
      PPH and blood transfusion32.9 (#2)30.7 (#2)23.5 (#2)26.1 (#2)40.0 (#2)28.8 (#2)38.7 (#3)28.0 (#2)36.3 (#2)
      Eclampsia10.2 (#3)7.7 (#6)4.8 (#8)7.8 (#5)6.7 (#8)4.7 (#9)11.4 (#7)16.3 (#3)9.0 (#3)
      Cardiac arrest or failure, myocardial infarction, or pulmonary edema9.4 (#4)10.4 (#3)8.8 (#4)9.4 (#4)12.1 (#3)9.2 (#4)16.3 (#4)11.7 (#4)6.3 (#6)
      Rupture of uterus before onset of labor9.4 (#5)7.1 (#8)5.7 (#7)7.5 (#6)5.9 (#10)8.0 (#7)15.1 (#5)5.3 (#12)3.7 (#11)
      Puerperal sepsis8.7 (#6)10.2 (#4)6.5 (#6)10.4 (#3)11.5 (#4)11.0 (#3)14.6 (#6)9.5 (#6)7.9 (#4)
      Repair of bladder, urethra, or intestine7.1 (#7)7.0 (#9)11.1 (#3)5.6 (#7)6.8 (#7)8.6 (#6)5.7 (#11)4.9 (#12)5.8 (#7)
      Embolization or ligation of pelvic vessels and PPH5.2 (#8)7.6 (#7)7.2 (#5)4.7 (#8)11.5 (#5)8.9 (#5)7.3 (#9)7.6 (#8)7.9 (#5)
      Assisted ventilation through tracheostomy or endotracheal tube4.0 (#9)5.5 (#10)3.4 (#9)3.7 (#9)8.1 (#6)5.0 (#8)9.8 (#8)7.6 (#7)(#16)
      Data were suppressed because of disclosure rules (case count ≤6).
      Placenta previa with hemorrhage and blood transfusion3.7 (#10)4.2 (#11)2.7 (#11)3.0 (#10)6.7 (#9)2.7 (#14)7.3 (#10)3.4 (#14)4.7 (#9)
      Obstetric embolism3.1 (#11)2.7 (#13)3.2 (#10)1.6 (#18)2.7 (#16)2.4 (#16)3.3 (#17)5.7 (#10)3.7 (#10)
      HIV infection1.9 (#19)8.0 (#5)1.5 (#16)1.5 (#19)(#25)
      Data were suppressed because of disclosure rules (case count ≤6).
      4.2 (#10)77.7 (#1)10.2 (#5)4.7 (#8)
      Sickle cell anemia with crisis0.3 (#30)0.7 (#25)(#32)
      Data were suppressed because of disclosure rules (case count ≤6).
      0.0 (#43)0.0 (#38)0.0 (#38)2.8 (#18)6.1 (#9)0.0 (#29)
      PPH: postpartum hemorrhage.
      a Data were suppressed because of disclosure rules (case count ≤6).
      Table 3 shows disparities in each of the top SMM components according to maternal birthplace. Disparities between all immigrants, considered as a single group, and the Canadian-born women do not accurately represent the actual disparities observed across different immigrant groups. Compared with Canadian-born women, immigrants as a single group are not statistically different. However, women from Europe or Western countries and South Asia had lower rates of blood transfusion with or without postpartum hemorrhage, but women from East Asia, Southeast Asia, and the Pacific and Sub-Saharan Africa had higher rates than Canadian-born women. Eclampsia was also less common among most immigrant groups, with the only exception being women from Sub-Saharan Africa. Women from East Asia, Southeast Asia, and the Pacific and Sub-Saharan Africa had higher rates of several components of SMM compared with Canadian-born women. The strongest associations were found in HIV infection for Sub-Saharan Africans (RR 43.58; 95% CI 34.39 to 55.23), followed by Caribbean, Latin American, and Middle Eastern or North African women. Women from the Caribbean had much higher rates of sickle cell anemia with crisis relative to Canadian-born women (RR 17.34; 95% CI 9.52 to 31.60).
      Table 3Unadjusted RRs (and 95% CIs) in the top 10 component conditions of SMM for different immigrant groups compared with non-immigrants
      SMM componentCanadian-born (N = 916 999)All immigrants (N = 335 544)Europe and Western countries (N = 52 425)South Asia (N = 104 458)East Asia, Southeast Asia, and Pacific (N = 74 970)Middle East and North Africa (N = 33 686)Sub Saharan Africa (N = 24 575)Caribbean (N = 26 437)Latin America (N = 18 993)
      Blood transfusion (whole blood: platelet plasma or red cell)1.000.98 (0.93–1.03)0.69
      P < 0.05.
      (0.60–0.80)
      0.85
      P < 0.05.
      (0.77–0.93)
      1.20
      P < 0.05.
      (1.09–1.31)
      0.85 (0.73–1.00)1.36
      P < 0.05.
      (1.18–1.58)
      1.16 (1.00–1.35)1.12 (0.93–1.34)
      PPH and blood transfusion1.000.94 (0.87–1.00)0.72
      P < 0.05.
      (0.60–0.86)
      0.80
      P < 0.05.
      (0.70–0.90)
      1.22
      P < 0.05.
      (1.08–1.37)
      0.88 (0.71–1.07)1.18 (0.96–1.45)0.85 (0.67–1.07)1.10 (0.87–1.40)
      Eclampsia1.000.76
      P < 0.05.
      (0.66–0.87)
      0.47
      P < 0.05.
      (0.31–0.69)
      0.76
      P < 0.05.
      (0.60–0.95)
      0.65
      P < 0.05.
      (0.49–0.87)
      0.46
      P < 0.05.
      (0.28–0.76)
      1.11 (0.76–1.62)1.59
      P < 0.05.
      (1.17–2.16)
      0.87 (0.54–1.41)
      Cardiac arrest or failure, myocardial infarction, or pulmonary edema1.001.10 (0.97–1.25)0.93 (0.69–1.25)0.99 (0.80–1.23)1.28
      P < 0.05.
      (1.03–1.60)
      0.98 (0.68–1.40)1.73
      P < 0.05.
      (1.26–2.37)
      1.24 (0.87–1.78)0.67 (0.38–1.19)
      Rupture of uterus before onset of labor1.000.75
      P < 0.05.
      (0.65–0.87)
      0.61
      P < 0.05.
      (0.42–0.87)
      0.80 (0.63–1.01)0.63
      P < 0.05.
      (0.46–0.85)
      0.85 (0.57–1.27)1.60
      P < 0.05.
      (1.16–2.24)
      0.56
      P < 0.05.
      (0.33–0.96)
      0.40
      P < 0.05.
      (0.19–0.84)
      Puerperal sepsis1.001.18
      P < 0.05.
      (1.04–1.34)
      0.75 (0.53–1.05)1.20 (0.99–1.47)1.32
      P < 0.05.
      (1.06–1.65)
      1.27 (0.91–1.76)1.69
      P < 0.05.
      (1.21–2.36)
      1.09 (0.73–1.63)0.91 (0.55–1.52)
      Repair of bladder, urethra or intestine1.000.99 (0.85–1.15)1.57
      P < 0.05.
      (1.20–2.05)
      0.80 (0.61–1.04)0.96 (0.72–1.28)1.22 (0.84–1.77)0.81 (0.48–1.37)0.70 (0.40–1.21)0.82 (0.45–1.49)
      Embolization or ligation of pelvic vessels and PPH1.001.46
      P < 0.05.
      (1.25–1.70)
      1.38 (1.00–1.93)0.90 (0.67–1.20)2.19
      P < 0.05.
      (1.74–2.76)
      1.70
      P < 0.05.
      (1.18–2.46)
      1.40 (0.87–2.24)1.44 (0.92–2.26)1.51 (0.90–2.52)
      Assisted vent tracheostomy or endotracheal tube1.001.37
      P < 0.05.
      (1.15–1.64)
      0.86 (0.54–1.38)0.94 (0.67–1.30)2.04
      P < 0.05.
      (1.56–2.68)
      1.27 (0.78–2.06)2.45
      P < 0.05.
      (1.62–3.71)
      1.90
      P < 0.05.
      (1.21–2.98)
      Data were suppressed because of disclosure rules (case count ≤6).
      Placenta previa with hemorrhage and blood transfusion1.001.12 (0.92–1.36)0.72 (0.42–1.22)0.80 (0.55–1.15)1.79
      P < 0.05.
      (1.33–2.41)
      0.72 (0.37–1.39)1.96
      P < 0.05.
      (1.22–3.15)
      0.91 (0.47–1.77)1.27 (0.66–2.46)
      Obstetric embolism1.000.87 (0.69–1.11)1.03 (0.62–1.73)0.52
      P < 0.05.
      (0.32–0.84)
      0.85 (0.54–1.34)0.76 (0.37–1.53)1.04 (0.52–2.09)1.80
      P < 0.05.
      (1.07–3.03)
      1.17 (0.55–2.48)
      HIV infection1.004.21
      P < 0.05.
      (3.38–5.25)
      0.81 (0.37–1.77)0.84 (0.48–1.48)
      Data were suppressed because of disclosure rules (case count ≤6).
      2.24
      P < 0.05.
      (1.30–3.87)
      43.58
      P < 0.05.
      (34.39–55.23)
      5.72
      P < 0.05.
      (3.67–8.92)
      2.65
      P < 0.05.
      (1.28–5.50)
      Sickle cell anemia with crisis1.002.15 (1.21–3.81)0.55 (0.07–4.00)
      Data were suppressed because of disclosure rules (case count ≤6).
      Data were suppressed because of disclosure rules (case count ≤6).
      Data were suppressed because of disclosure rules (case count ≤6).
      8.16 (3.60–18.49)17.34 (9.52–31.60)
      Data were suppressed because of disclosure rules (case count ≤6).
      PPH: postpartum hemorrhage.
      a P < 0.05.
      b Data were suppressed because of disclosure rules (case count ≤6).
      Table 4 shows the RRs of SMM between immigrant subgroups versus Canadian-born women, before and after removing women with HIV infection. Immigrant subgroups were divided according to regions, which were further subdivided into specific countries whenever possible. Women from Europe or Western and South Asian countries had lower or similar rates of SMM than Canadian-born women, whereas women from the Middle East and North Africa and Latin America did not differ from the reference group. Higher rates of SMM were seen among women from some countries in Southeast Asia (i.e., the Philippines and Vietnam), Sub-Saharan Africa, and the Caribbean (i.e., Jamaica and Haiti), before and after adjustment for maternal characteristics. Exclusion of HIV infection cases did not affect the observed associations for most women, with the only exception being women from Sub-Saharan African countries. The RRs for women from Nigeria, Ethiopia, Ghana, and the Congo were substantially attenuated after removing HIV infection cases, but they remained positively associated with SMM. For women from the Sudan and Zimbabwe, the association disappeared altogether, thus indicating that HIV infections were responsible for these associations.
      Table 4RRs in SMM for immigrants from different world regions and countries of birth compared with Canadian-born women, before and after excluding women with HIV infection, in Ontario
      Women's birthplaceRate (per 1000 deliveries)RR (95% CI)Adjusted RR (95% CI)
      Adjusted for year of delivery, maternal age at delivery (15 to 19, 20 to 24, 25 to 29, 30 to 34, 35 to 39, 40 to 49), parity (0, 1, 2–3, 4+), multifetal pregnancy (yes vs. no), rural residence (yes vs. no), neighbourhood income quintiles (Q1 to Q5).
      Adjusted RR (95% CI)
      Adjusted for year of delivery, maternal age at delivery (15 to 19, 20 to 24, 25 to 29, 30 to 34, 35 to 39, 40 to 49), parity (0, 1, 2–3, 4+), multifetal pregnancy (yes vs. no), rural residence (yes vs. no), neighbourhood income quintiles (Q1 to Q5).
      (excluding HIV infection)
      Canadian-born (N = 916 999)11.81.001.001.00
      All immigrants (N = 335 544)12.11.01 (0.97–1.05)1.01 (0.97–1.05)0.97 (0.93–1.01)
      Europe and Western countries (N = 52 425)9.20.77 (0.71–0.85)0.78 (0.71–0.85)0.78 (0.71–0.85)
       United States (N = 5466)13.01.09 (0.86–1.37)1.08 (0.86–1.36)1.08 (0.86–1.36)
       Former Yugoslavia (N = 6601)11.10.92 (0.74–1.16)0.92 (0.73–1.15)0.92 (0.73–1.16)
       United Kingdom and colonies (N = 4932)10.10.85 (0.64–1.12)0.85 (0.64–1.11)0.86 (0.65–1.13)
       Portugal and Azores (N = 3892)9.50.80 (0.58–1.10)0.84 (0.61–1.16)0.86 (0.62–1.19)
       Rest of Western countries (N = 10 711)8.70.73 (0.59–0.89)0.72 (0.59–0.89)0.71 (0.58–0.88)
       Former Soviet Union (N = 7883)8.50.71 (0.56–0.90)0.73 (0.58–0.93)0.67 (0.53–0.86)
       Romania (N = 4741)7.60.64 (0.46–0.88)0.62 (0.45–0.86)0.63 (0.45–0.87)
       Poland (N = 8073)6.90.58 (0.44–0.75)0.60 (0.46–0.78)0.61 (0.47–0.79)
      South Asia (N = 104 458)10.20.85 (0.80–0.91)0.89 (0.84–0.95)0.90 (0.85–0.97)
       Afghanistan (N = 6411)11.10.93 (0.73–1.17)0.93 (0.74–1.18)0.92 (0.73–1.17)
       Rest of South Asia (N = 5607)11.10.92 (0.72–1.19)0.94 (0.73–1.20)0.96 (0.75–1.24)
       Pakistan (N = 26 956)10.60.88 (0.79–0.99)0.94 (0.83–1.05)0.96 (0.85–1.08)
       India (N = 46 216)9.90.83 (0.76–0.91)0.88 (0.80–0.96)0.88 (0.80–0.96)
       Sri Lanka (N = 19 268)9.90.82 (0.72–0.95)0.84 (0.73–0.97)0.85 (0.74–0.98)
      East Asia, Southeast Asia, and Pacific (N = 74 970)12.91.08 (1.01–1.15)1.05 (0.98–1.12)1.07 (1.00–1.14)
       Philippines (N = 22 066)16.91.42 (1.28–1.57)1.36 (1.23–1.51)1.39 (1.25–1.54)
       Vietnam (N = 8411)14.61.22 (1.03–1.46)1.23 (1.03–1.47)1.24 (1.03–1.48)
       China (N = 31 753)11.00.92 (0.82–1.02)0.90 (0.81–1.00)0.92 (0.82–1.02)
       Rest of East Asia and Pacific (N = 12 740)9.30.78 (0.65–0.94)0.78 (0.65–0.94)0.79 (0.66–0.94)
      Middle East and North Africa (N = 33 686)10.70.90 (0.81–1.00)0.91 (0.82–1.01)0.90 (0.81–1.00)
       Iran (N = 5442)10.80.91 (0.70–1.17)0.87 (0.68–1.13)0.89 (0.69–1.15)
       Rest of Middle East and North Africa (N = 12 740)10.60.89 (0.76–1.03)0.92 (0.79–1.07)0.93 (0.80–1.08)
       Iraq (N = 5415)9.80.82 (0.63–1.07)0.82 (0.63–1.07)0.84 (0.64–1.10)
       Lebanon (N = 4683)9.00.75 (0.55–1.01)0.78 (0.58–1.06)0.79 (0.58–1.06)
      Sub–Saharan Africa (N = 24 575)23.01.93 (1.77–2.09)1.84 (1.69–2.00)1.26 (1.14–1.40)
       Zimbabwe (N = 607)49.44.13 (2.91–5.86)4.20 (2.95–5.96)0.30 (0.08–1.19)
       Congo, DRC (N = 1120)37.53.14 (2.33–4.22)3.03 (2.25–4.07)1.83 (1.23–2.72)
       Nigeria (N = 3069)28.72.40 (1.95–2.95)2.38 (1.93–2.93)1.94 (1.53–2.45)
       Ethiopia (N = 3587)26.82.24 (1.84–2.73)2.05 (1.67–2.50)1.29 (1.00–1.67)
       Rest of Sub-Saharan Africa (N = 5859)21.51.80 (1.51–2.14)1.78 (1.49–2.11)1.07 (0.85–1.34)
       Ghana (N = 3308)20.91.74 (1.38–2.20)1.71 (1.35–2.16)1.34 (1.02–1.75)
       Sudan (N = 1841)19.01.59 (1.14–2.21)1.47 (1.06–2.04)1.01 (0.67–1.51)
       Somalia (N = 5868)16.41.37 (1.12–1.67)1.16 (0.94–1.42)1.09 (0.88–1.35)
       Kenya (N = 1157)16.41.37 (0.88–2.15)1.36 (0.87–2.12)0.81 (0.45–1.46)
      Caribbean (N = 26 437)14.61.22 (1.11–1.35)1.22 (1.10–1.35)1.17 (1.05–1.29)
       Haiti (N = 598)31.82.66 (1.71–4.14)2.53 (1.62–3.95)2.47 (1.56–3.90)
       Rest of Caribbean (N = 3535)16.41.37 (1.06–1.77)1.34 (1.04–1.73)1.24 (0.95–1.63)
       Jamaica (N = 11 736)15.11.26 (1.09–1.46)1.24 (1.07–1.44)1.22 (1.04–1.42)
       Trinidad and Tobago (N = 3839)13.01.09 (0.83–1.44)1.08 (0.82–1.42)1.03 (0.78–1.37)
       Guyana (N = 6729)12.31.03 (0.83–1.28)1.04 (0.84–1.29)0.96 (0.77–1.21)
      Hispanic America (N = 18 993)11.80.99 (0.87–1.13)1.00 (0.88–1.14)0.98 (0.86–1.13)
       Mexico (N = 3179)15.11.25 (0.94–1.66)1.23 (0.93–1.63)1.23 (0.93–1.64)
       El Salvador (N = 2572)12.81.07 (0.76–1.51)1.11 (0.79–1.56)1.08 (0.76–1.53)
       Colombia (N = 2926)11.61.05 (0.75–1.47)1.06 (0.76–1.49)1.02 (0.72–1.44)
       Rest of Hispanic America (N = 10 546)10.40.87 (0.72–1.05)0.89 (0.74–1.07)0.87 (0.72–1.06)
      a Adjusted for year of delivery, maternal age at delivery (15 to 19, 20 to 24, 25 to 29, 30 to 34, 35 to 39, 40 to 49), parity (0, 1, 2–3, 4+), multifetal pregnancy (yes vs. no), rural residence (yes vs. no), neighbourhood income quintiles (Q1 to Q5).

      Discussion

      In a setting with both high immigration from diverse parts of the world and universal access to health care services, we found substantial variability in the risk of SMM and its components according to maternal birthplace. However, overall rates of SMM among immigrants and Canadian-born women were approximately similar.
      Our study is consistent with previous studies documenting a higher risk of SMM among women originating from Sub-Saharan Africa and the Caribbean.
      • Urquia M.L.
      • Glazier R.H.
      • Mortensen L.
      • et al.
      Severe maternal morbidity associated with maternal birthplace in three high-immigration settings.
      • Knight M.
      • Kurinczuk J.J.
      • Spark P.
      • et al.
      Inequalities in maternal health: national cohort study of ethnic variation in severe maternal morbidities.
      • Zwart J.J.
      • Jonkers M.D.
      • Richters A.
      • et al.
      Ethnic disparity in severe acute maternal morbidity: a nationwide cohort study in the Netherlands.
      However, our study provides additional details by identifying heterogeneity within maternal world regions of origin and the specific countries with higher rates of SMM (e.g., Jamaica and Haiti among Caribbean women). Although the overall patterns of risk are consistent across studies, small differences in results between our studies and previous studies from Britain
      • Knight M.
      • Kurinczuk J.J.
      • Spark P.
      • et al.
      Inequalities in maternal health: national cohort study of ethnic variation in severe maternal morbidities.
      and the Netherlands
      • Zwart J.J.
      • Jonkers M.D.
      • Richters A.
      • et al.
      Ethnic disparity in severe acute maternal morbidity: a nationwide cohort study in the Netherlands.
      reflect differences in the definition of SMM and ethnicity or immigration, reporting of specific conditions, a different baseline risk in the native-born population, different migration patterns and selection processes, or differential health care. Our study also contributes to our knowledge of immigrants from East Asia, who are not very common in European studies. Women from the Philippines and Vietnam were at higher risk of SMM in our study, unlike other groups, such as Chinese women.
      Our study has implications for future research. First, it illustrates that pooling all immigrants into a single group yields an SMM risk profile for immigrants that is similar to that of the Canadian-born population. Such aggregation bias resulting from grouping distinct but smaller subgroups of women with higher and lower risks of SMM leads to masking of underlying differences among immigrant subgroups. Comparisons of all immigrants as a single group should be avoided whenever possible or should be made only after having confirmed homogeneity of risk across the immigrant subgroups. Second, the large heterogeneity according to country of birth observed in our study suggests that differential mechanisms underlie risks among women from specific countries of origin (e.g., HIV infection and sickle cell anemia among women from Sub-Saharan Africa and the Caribbean). Such heterogeneity warrants further research at the country of origin level to disentangle further the pathways leading to higher risk among specific subgroups.
      Limitations present in our study include the lack of information on pre-migration, early life exposures, obstetric history, and other potential confounders around the time of delivery, such as smoking and other behavioural and constitutional factors such as obesity. However, given that many of these lifestyle factors may be related to the country of origin, adjustment may result in overcontrol and suppress or attenuate the associations with maternal birthplace. Our findings may not be generalizable to settings with migrants from countries different from those settling in Canada. Finally, the SMM composite indicator was developed for surveillance. The combination of prevalent (i.e., pre-existing or chronic) and incident or acute conditions in a single indicator makes it difficult to interpret the results as purely reflecting a general state of maternal health or serious complications arising during pregnancy, labour, and delivery. Conversely, the use of a single comprehensive indicator captures differences in overall reproductive health status and care among immigrant women that may not manifest in unpowered analysis of a single, less frequent condition. Our study identified patterns of risk among immigrant women, but further research is needed to clarify the mechanisms and causes underlying the associations, including the continuum of care, which may provide opportunities for timely intervention. Maternal birthplace should be regarded as a marker of risk that provides clues for the identification of individual risk factors.

      Conclusion

      Although immigrants in Canada have rates of SMM that are similar to those of Canadian-born women, SMM rates vary substantially according to maternal birthplace. The variability increases when results are disaggregated by country of birth and specific morbidity.

      Acknowledgements

      This study was supported by the Institute for Clinical Evaluative Sciences (ICES), which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care (MOHLTC). The opinions, results, and conclusions reported in this paper are those of the authors and are independent of the funding sources. No endorsement by ICES or the Ontario MOHLTC is intended or should be inferred. The basis of parts of this material consists of data and information compiled and provided by CIHI. However, the analyses, conclusions, opinions, and statements expressed herein are those of the authors and not necessarily those of CIHI. This study was funded by a grant of the Canadian Institutes of Health Research (CIHR) (MOP- 123267). M.U. is a Canada Research Chair in Applied Population Health. J.R. and K.S.J. hold CIHR chairs in Reproductive and Child Health Services and Policy Research.

      Supplementary Data

      References

        • Boyd M.
        • Vickers M.
        100 years of immigration in Canada.
        Can Social Trends. 2000; (Autumn:2–12. Statistics Canada Catalogue No. 11-008)
        • WHO, UNICEF, UNFPA, World Bank Group and the United Nations Population Division
        Trends in maternal mortality: 1990 to 2015. Estimates by WHO, UNICEF, UNFPA, World Bank Group and the United Nations Population Division.
        World Health Organization, Geneva2015
        • Pedersen G.S.
        • Grontved A.
        • Mortensen L.H.
        • et al.
        Maternal mortality among migrants in Western Europe: a meta-analysis.
        Matern Child Health J. 2014; 18: 1628-1638
        • Urquia M.L.
        • Vang Z.M.
        • Bolumar F.
        Birth outcomes of Latin Americans in two countries with contrasting immigration admission policies: Canada and Spain.
        PLoS One. 2015; 10: e0136308
        • Wanigaratne S.
        • Cole D.C.
        • Bassil K.
        • et al.
        Contribution of HIV to maternal morbidity among refugee women in Canada.
        Am J Public Health. 2015; 105: 2449-2456
        • Urquia M.L.
        • Glazier R.H.
        • Mortensen L.
        • et al.
        Severe maternal morbidity associated with maternal birthplace in three high-immigration settings.
        Eur J Public Health. 2015; 25: 620-625
        • van den Akker T.
        • van Roosmalen J.
        Maternal mortality and severe morbidity in a migration perspective.
        Best Pract Res Clin Obstet Gynaecol. 2016; 32: 26-38
        • Joseph K.S.
        • Liu S.
        • Rouleau J.
        • et al.
        Severe maternal morbidity in Canada, 2003 to 2007: surveillance using routine hospitalization data and ICD-10CA codes.
        J Obstet Gynaecol Can. 2010; 32: 837-846
        • Roberts C.L.
        • Cameron C.A.
        • Bell J.C.
        • et al.
        Measuring maternal morbidity in routinely collected health data: development and validation of a maternal morbidity outcome indicator.
        Med Care. 2008; 46: 786-794
        • Sousa M.H.
        • Cecatti J.G.
        • Hardy E.E.
        • et al.
        Severe maternal morbidity (near miss) as a sentinel event of maternal death: an attempt to use routine data for surveillance.
        Reprod Health. 2008; 5: 6
        • Kayem G.
        • Kurinczuk J.
        • Lewis G.
        • et al.
        Risk factors for progression from severe maternal morbidity to death: a national cohort study.
        PLoS One. 2011; 6: e29077
        • Wahlberg A.
        • Roost M.
        • Haglund B.
        • et al.
        Increased risk of severe maternal morbidity (near-miss) among immigrant women in Sweden: a population register-based study.
        BJOG. 2013; 120: 1605-1611
        • Knight M.
        • Kurinczuk J.J.
        • Spark P.
        • et al.
        Inequalities in maternal health: national cohort study of ethnic variation in severe maternal morbidities.
        BMJ. 2009; 338: b542
        • Zwart J.J.
        • Jonkers M.D.
        • Richters A.
        • et al.
        Ethnic disparity in severe acute maternal morbidity: a nationwide cohort study in the Netherlands.
        Eur J Public Health. 2011; 21: 229-234
        • Canadian Institute for Health Information
        Final report. The Canadian enhancement of ICD-10 (International Statistical Classification of Diseases and Related Health Problems, tenth revision).
        Canadian Institute for Health Information, Ottawa2001
        • Canadian Institute for Health Information
        International Statistical Classification of Diseases and Related Health Problems, tenth revision. CCICD-10. 2009.
        Canadian Institute for Health Information, Ottawa2010
        • Liu S.
        • Joseph K.S.
        • Bartholomew S.
        • et al.
        Temporal trends and regional variations in severe maternal morbidity in Canada, 2003 to 2007.
        J Obstet Gynaecol Can. 2010; 32: 847-855
        • Joseph K.S.
        • Fahey J.
        Validation of perinatal data in the Discharge Abstract Database of the Canadian Institute for Health Information.
        Chronic Dis Can. 2009; 29: 96-100
        • Hutton E.K.
        • Cappelletti A.
        • Reitsma A.H.
        • et al.
        Outcomes associated with planned place of birth among women with low-risk pregnancies.
        CMAJ. 2016; 188: E80-E90
        • Zou G.
        A modified Poisson regression approach to prospective studies with binary data.
        Am J Epidemiol. 2004; 159: 702-706