Advertisement

Beyond the numbers: classifying contributory factors and potentially avoidable maternal deaths in New Zealand, 2006–2009

Published:August 01, 2011DOI:https://doi.org/10.1016/j.ajog.2011.07.044

      Objective

      We sought to describe a new classification system for contributory factors in, and potential avoidability of, maternal deaths and to determine the contributory factors and potential avoidability among 4 years of maternal deaths in New Zealand.

      Study Design

      A new classification system for reporting contributory factors in all maternal deaths was developed from previous tools and applied to all maternal deaths in New Zealand from 2006 through 2009.

      Results

      There were 49 deaths and the maternal mortality ratio was 19.2/100,000 maternities. Contributory factors were identified in 55% of cases. An expert panel identified 35% of maternal deaths as potentially avoidable. In cases where potential avoidability was determined, there were nearly always 2 or 3 domains where contributory factors were identified.

      Conclusion

      Almost one third of maternal deaths in New Zealand can be considered to be potentially avoidable. This methodology has the potential to identify areas for improvement in the quality of maternity care.

      Key words

      Maternal deaths are devastating events for families. The number of maternal deaths each year in New Zealand varies from 9-15 (Perinatal and Maternal Mortality Review Committee [PMMRC] 2008).
      PMMRC 2010
      Perinatal and Maternal Mortality in New Zealand 2008. Fourth report to the Minister of Health. July 2009 to June 2010. Wellington: Ministry of Health.
      In 2005, the New Zealand Minister of Health established the PMMRC for the purpose of reviewing both perinatal and maternal deaths. The strength of mortality review is the ability to review deaths both individually and as aggregated data. From the single examination of cases by experts comes information that might otherwise have been overlooked. From the aggregated data come broader themes and trends that can be identified and monitored and with appropriate policy changes and interventions might lead to improvements in outcomes.
      NHS Institute for Innovation and Improvement
      Reducing avoidable mortality.
      See related editorial, page 293
      Mortality review should not only focus on definitions and causation of disease but also needs to focus on modifiable features in health systems and the quality of clinical care.
      NHS Institute for Innovation and Improvement
      Reducing avoidable mortality.
      With this in mind, the PMMRC sought to report not only clinical data but also contributory factors and potential avoidability. For example, a woman with preeclampsia dies at 38 weeks from a cerebral hemorrhage and on review of the case the management of hypertension was found to be outside standard practice. This death would be classified as a direct maternal death from hypertension with contributory factors that might include inadequate protocols or guidelines, failure to follow standard practice, and failure to appreciate the seriousness of the condition. The death can therefore be considered potentially avoidable.
      It has been suggested that even in the developed world 50% of all maternal deaths are potentially avoidable.
      • Cantwell R.
      • Clutton-Brock T.
      • Cooper G.
      • et al.
      Saving mother' lives: reviewing maternal deaths to make motherhood safer: 2006-2008. The eighth report of the confidential enquiries into maternal deaths in the United Kingdom.
      A number of models of reporting contributory factors have already been developed.
      • De Reu P.
      • Van Diem M.
      • Eskes M.
      • et al.
      The Dutch perinatal audit project: a feasibility study for nationwide perinatal audit in the Netherlands.
      The 13th report of the perinatal and infant mortality committee of Western Australia for deaths in the triennium 2005-2007.
      • Geller S.E.
      • Rosenberg D.
      • Cox S.
      • Brown M.
      • Simonson L.
      • Kilpatrick S.
      A scoring system identified near-miss maternal morbidity during pregnancy.
      • Cronin C.
      Five years of learning from analysis of clinical occurrences in pediatric care using the London protocol.
      • Szekendi M.
      Using patient safety morbidity and mortality conferences to promote transparency and a culture of safety.
      None of these systems adequately met our requirements for identifying potential avoidability. Some failed to provide adequate documentation of the process. For example, we were unable to find a definition of substandard care or the methodology used by Centre for Maternal and Child Enquiries (CMACE).
      • Cantwell R.
      • Clutton-Brock T.
      • Cooper G.
      • et al.
      Saving mother' lives: reviewing maternal deaths to make motherhood safer: 2006-2008. The eighth report of the confidential enquiries into maternal deaths in the United Kingdom.
      • Cronin C.
      Five years of learning from analysis of clinical occurrences in pediatric care using the London protocol.
      Other models covered some dimensions well but were not comprehensive.
      • Geller S.E.
      • Rosenberg D.
      • Cox S.
      • Brown M.
      • Simonson L.
      • Kilpatrick S.
      A scoring system identified near-miss maternal morbidity during pregnancy.
      For example, we sought to report on barriers to accessing or engaging with care and yet none of the classification systems encompassed this dimension. In the case of root cause analysis, it was thought that this only addressed system issues and did not consider the contribution of clinical competence.
      • Szekendi M.
      Using patient safety morbidity and mortality conferences to promote transparency and a culture of safety.
      The aim of this report is to, first, describe a new classification system for contributory factors incorporating the best of these approaches and, second, to determine the contributory factors and potential avoidability among 4 years of maternal deaths in New Zealand.

      Materials and Methods

      The Maternal Mortality Review Working Group (MMRWG) of the PMMRC is responsible for reviewing all maternal deaths in New Zealand. The members of the working group include obstetricians, midwives, an anesthetist, a psychiatrist, a health care manager, and a pathologist.

      Definitions

      In New Zealand, a maternal death is defined as the death of a woman while pregnant or within 42 days of termination of 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.
      The maternal mortality ratio is calculated per 100,000 maternities and follows this CMACE approach.
      • Cantwell R.
      • Clutton-Brock T.
      • Cooper G.
      • et al.
      Saving mother' lives: reviewing maternal deaths to make motherhood safer: 2006-2008. The eighth report of the confidential enquiries into maternal deaths in the United Kingdom.
      Maternities are defined as all live births and fetal deaths at ≥20 weeks or weighing ≥400 g if gestation unknown. Pregnancies ending <20 weeks are not included in this working definition because the absolute number of pregnancies ending before this time is unknown.
      The definitions adopted by the MMRWG are based on the World Health Organization definitions from the International Statistical Classification of Diseases, 10th Revision and the cause of each death is subclassified, using the CMACE system
      • Cantwell R.
      • Clutton-Brock T.
      • Cooper G.
      • et al.
      Saving mother' lives: reviewing maternal deaths to make motherhood safer: 2006-2008. The eighth report of the confidential enquiries into maternal deaths in the United Kingdom.
      :
      • Direct maternal deaths: those resulting from obstetric complications of the pregnant state (pregnancy, labor, or puerperium); from interventions, omissions, incorrect treatment; or from a chain of events resulting from the above.
      • Indirect maternal deaths: those resulting from previous existing disease or disease that developed during pregnancy and not due to direct obstetric causes, but aggravated by the physiologic effects of pregnancy.
      • Coincidental maternal deaths: those resulting from unrelated causes that happen to occur in pregnancy or the puerperium.

      Development of a new classification system for contributory factors

      The classification system was developed iteratively following a literature search for similar models, in discussion with the PMMRC, the MMRWG, local coordinators, and Ministry of Health officials of the Quality Improvement Committee. The published classification systems considered in the development of this system for New Zealand include the following:
      • 1
        CMACE identifies cases of substandard care and considers the following contributory factors: failures in diagnosis and treatment, errors in management of complications, failures in the organization of health care (structure), and patient factors such as late presentation for treatment, barriers to access to care, social situation, family violence, and drug use.
        • Cantwell R.
        • Clutton-Brock T.
        • Cooper G.
        • et al.
        Saving mother' lives: reviewing maternal deaths to make motherhood safer: 2006-2008. The eighth report of the confidential enquiries into maternal deaths in the United Kingdom.
        The definition for substandard care (major) is “Contributed significantly to the death of the mother. In many, but not all cases, different treatment may have altered the outcome.” No description of the methods used to identify the contributory factors or substandard care is provided within the report or referenced.
      • 2
        The London protocol
        • Cronin C.
        Five years of learning from analysis of clinical occurrences in pediatric care using the London protocol.
        • Taylor-Adams S.
        for reviewing critical incidents contains a framework of contributory factors that includes the following: institutional (medicolegal and regulatory), organization and management (financial and policies), work environment (staffing levels and skills mix, the equipment and administrative and management support), team (communication, supervision, leadership), individual staff member (knowledge and skills), tasks (protocols, accuracy and availability of diagnostic tests), and the patient (the complexity of their condition, language, and social factors). This tool was a useful starting point for the methodology developed.
      • 3
        The Geller model for scoring preventability includes provider- and system-related preventability.
        • Geller S.E.
        • Rosenberg D.
        • Cox S.
        • Brown M.
        • Simonson L.
        • Kilpatrick S.
        A scoring system identified near-miss maternal morbidity during pregnancy.
        • Geller S.E.
        • Rosenberg D.
        • Cox S.M.
        • et al.
        The continuum of maternal morbidity and mortality: factors associated with severity.
        • Geller S.E.
        • Cox S.M.
        • Callaghan W.M.
        • Berg C.J.
        Morbidity and mortality in pregnancy: laying the groundwork for safe motherhood.
        Morbidity and mortality cases were assessed by a multidisciplinary team of experts to reach a consensus on classification of death as preventable or not. Preventability was described as “action or inaction on the part of the health care provider, system or patient that may have caused or contributed” to the adverse outcome.
        • Geller S.E.
        • Rosenberg D.
        • Cox S.M.
        • et al.
        The continuum of maternal morbidity and mortality: factors associated with severity.
        A descriptive model for identifying provider- and system-related preventability notes 10 categories to assess but they have not included patient factors in this model. Geller et al,
        • Geller S.E.
        • Cox S.M.
        • Callaghan W.M.
        • Berg C.J.
        Morbidity and mortality in pregnancy: laying the groundwork for safe motherhood.
        however, does comment on noncompliance as a contributory factor, stating patient factors as being social factors that limit their ability to access health care as well as actions or delays by the patient that contributed to the death.
      • 4
        Root cause analysis approach for identifying factors includes policies, communication, training and competency, fatigue, scheduling, environment, and equipment factors.
        • Szekendi M.
        Using patient safety morbidity and mortality conferences to promote transparency and a culture of safety.
        Patient factors are not included as causal factors nor is the family included in the mortality review process, citing confidentiality.
        • Szekendi M.
        Using patient safety morbidity and mortality conferences to promote transparency and a culture of safety.
      • 5
        In the preventability scale of the Perinatal and Infant Mortality Committee of Western Australia,
        The 13th report of the perinatal and infant mortality committee of Western Australia for deaths in the triennium 2005-2007.
        the preventability of an adverse event is defined as “an error in management due to failure to follow accepted practice at the individual or system level” and accepted practice is taken to be “the current level of expected performance for the average practitioner or system that manages the patient”
        The 13th report of the perinatal and infant mortality committee of Western Australia for deaths in the triennium 2005-2007.
        and are based on the preventability score used in the Quality in Australian Health Care Study.
        • Wilson R.W.
        • Runciman W.B.
        • Gibberd R.W.
        • Harrison B.T.
        • Newby L.
        • Hamilton J.D.
        The quality in Australian health care study.
        This system does not appear to address barriers to accessing and engaging with care.

      The new system for classifying potentially avoidable deaths

      Following the presentation of each maternal death the role of contributory factors is considered. There are 2 steps. In the first step, the following questions that identify contributory factors are considered:
      • Have any organizational and/or management factors been identified?
      • Have factors relating to personnel been identified?
      • Have factors relating to technology and equipment been identified?
      • Have factors relating to environment been identified?
      • Have barriers to accessing/engaging with care been identified?
      Examples of these and the subcategories within each are given in the PMMRC Contributory Factors Form (Table 1).
      TABLE 1Perinatal and Maternal Mortality Review Committee Contributory Factors form
      Have any organizational and/or management factors been identified?
       Poor organizational arrangements of staff
       Inadequate education and training
       Lack of policies, protocols, or guidelines
       Inadequate numbers of staff
       Poor access to senior clinical staff
       Failure or delay in emergency response
       Delay in procedure, eg, cesarean section
       Inadequate systems/process for sharing of clinical information between services
       Delayed access to test results or inaccurate results
       Other reason
      Have factors relating to personnel been identified?
       Knowledge and skills of staff were lacking (includes failure to maintain competence)
       Delayed emergency response by staff
       Failure of communication between staff
       Failure to seek help/supervision
       Failure to offer or follow recommended best practice
       Lack of recognition of complexity or seriousness of condition
       Other reason
      Have factors relating to technology and equipment been identified?
       Essential equipment not available
       Lack of maintenance of equipment
       Malfunction/failure of equipment
       Failure/lack of information technology
       Other reason
      Have factors relating to environment been identified?
       Geography, eg, long-distance transfer
       Building and design functionality limited clinical response
       Other reason
      Have barriers to accessing/engaging with care (eg, no, infrequent, or late booking for antenatal care; woman declined treatment/advice) been identified?
       Substance use
       Lack of recognition of complexity or seriousness of condition (by either woman or her family)
       Maternal mental illness
       Cultural barriers
       Language barriers
       Not eligible to access free care
       Family violence
       Other reason
      Farquhar. Potential avoidability in maternal deaths in New Zealand 2006 through 2009. Am J Obstet Gynecol 2011.
      If contributory factors were identified, the second step was to consider if the maternal death was potentially avoidable. Potential avoidabililty is defined where there are aspects of care that may have changed the clinical outcome had they been identified. A maternal death was considered potentially avoidable if the absence of the contributory factors would have prevented the death.

      Process for data collection for reporting potentially avoidable factors

      The process for maternal death review in New Zealand is summarized in the Figure. Following notification of a maternal death, the national coordinator issues maternal death reporting forms to the appropriate local coordinator, who is then responsible for gathering the relevant clinical information from staff involved with the woman's care. Each completed reporting form, along with relevant clinical records, is reviewed by a designated member or members of the working group, who presents a summary of the case and findings to the working group. Each case is then discussed by the MMRWG in detail, including review of contributory factors and potential avoidability.
      Figure thumbnail gr1
      FIGUREProcess for maternal death review in New Zealand
      aAll health care workers are required to provide information about the mother and baby where a maternal death has occurred under the New Zealand Public Health and Disability Act 2000; bLocal review is protected under the Protected Quality Assurance Activities of the Health Practitioners Competency Act and data collection for the MMRWG is protected under the New Zealand Public Health and Disability Act 2000.
      MMRWG, Maternal Mortality Review Working Group; PMMRC, Perinatal and Maternal Mortality Review Committee.
      Farquhar. Potential avoidability in maternal deaths in New Zealand 2006 through 2009. Am J Obstet Gynecol 2011.
      The new system for classifying contributory factors and potentially avoidable deaths was applied to all maternal deaths from 2006 through 2009 inclusive. From 2006 through 2008, the MMRWG of the PMMRC prospectively assessed potential avoidability of all maternal deaths but did not use a tool for identifying contributory factors. In early 2010 an expert panel that included a midwife researcher, an obstetrician, and an epidemiologist, one of whom was also a member of the working group, considered each death from 2006 through 2008 individually and completed the classification form retrospectively using the tool (Table 1). During this process small refinements were made to the system and these were applied retrospectively to all cases reviewed. The expert panel's retrospective assessment of potential avoidability was consistent, in 32 of the 35 deaths, with the original assessment of the working group. For 2009 deaths, the working group applied the new tool described here prospectively in reviewing the 2009 maternal deaths.

      Details of institutional review board approval

      This was not required as this report and the data it contains are covered under the New Zealand Public Health and Disability Act of 2000.

      Results

      Forty-nine deaths were identified in the 4 years of this study. This is believed to represent complete ascertainment of direct and indirect maternal deaths in New Zealand over this time period. The causes of these 49 deaths are presented in Table 2.
      PMMRC 2010
      Perinatal and Maternal Mortality in New Zealand 2008. Fourth report to the Minister of Health. July 2009 to June 2010. Wellington: Ministry of Health.
      PMMRC 2008
      Perinatal and Maternal Mortality in New Zealand 2006. Second report to the Minister of Health. July 2007 to June 2008. Wellington: Ministry of Health.
      PMMRC 2009
      Perinatal and Maternal Mortality in New Zealand 2007. Third report to the Minister of Health. July 2008 to June 2009. Wellington: Ministry of Health.
      PMMRC 2011
      Perinatal and Maternal Mortality in New Zealand 2009. Fifth report to the Minister of Health. July 2010 to June 2011. Wellington: Ministry of Health.
      The maternal mortality ratio for the years 2006 through 2009 was 19.2/100,000 maternities.
      TABLE 2Maternal mortalities and cause of maternal deaths 2006 through 2009
      Classification and cause of maternal deathTotalContributory factors presentPotentially avoidable deaths
      Direct maternal death20116
       Amniotic fluid embolism84
       Postpartum hemorrhage322
       Pulmonary embolism211
       Peripartum cardiomyopathy1
       Preeclampsia433
       Sepsis21
      Indirect maternal death261611
       Preexisting medical condition974
       Nonobstetric sepsis622
       Intracranial hemorrhage1
       Suicide1075
      Unclassifiable3
      Total492717
      Maternal mortality ratio19.2/100,000 maternities
      Farquhar. Potential avoidability in maternal deaths in New Zealand 2006 through 2009. Am J Obstet Gynecol 2011.
      The findings of the retrospective review of contributory factors, and potentially avoidable maternal deaths in 2006 through 2009, are shown in Table 3. Contributory factors were identified in 55% of cases overall, and in many cases >1 factor applied. Factors relating to organizational and/or management, personnel, and barriers to access and engagement were more frequent and factors relating to the environment and technology and equipment were less common and not found in cases where the death was potentially avoidable.
      TABLE 3Contributory factors identified by category in maternal deaths 2006 through 2009
      Contributory features identified: (multiple categories may apply)n = 49
      DirectIndirectn%
      Organizational and/or management1633
       Poor organizational arrangements of staff3148
       Inadequate education and training42612
       Lack of policies, protocols, or guidelines671327
       Inadequate numbers of staff
       Poor access to senior clinical staff1124
       Failure or delay in emergency response3148
       Delay in procedure, eg, cesarean1124
       Inadequate systems/process for sharing clinical information between services1348
       Delayed access to test results or inaccurate results112
       Other reason1124
      Personnel1735
       Knowledge and skills of staff lacking (includes failure to maintain competence)44810
       Delayed emergency response by staff32510
       Failure of communication between staff35816
       Failure to seek help/supervision2136
       Failure to offer or follow recommended best practice1124
       Lack of recognition of complexity/seriousness of condition33612
      Technology and equipment12
       Essential equipment not available
       Lack of maintenance of equipment112
       Malfunction/failure of equipment
       Failure/lack of information technology
      Environment36
       Geography, eg, long-distance transfer112
       Building and design functionality limited clinical response1124
      Barriers to accessing or engaging with care1429
       Substance use5510
       Family violence336
       Lack of recognition of complexity/seriousness of condition (by either woman or her family)6612
       Maternal mental illness112
       Cultural barriers
       Language barriers112
       Not eligible to access free care112
       Other, specify
      Unbooked (1), late booking (2), infrequent attendance (2), social circumstances–inability to engage in care (2).
      16714
      Farquhar. Potential avoidability in maternal deaths in New Zealand 2006 through 2009. Am J Obstet Gynecol 2011.
      a Unbooked (1), late booking (2), infrequent attendance (2), social circumstances–inability to engage in care (2).
      Of factors relating to organization/management, the most commonly identified subcategory was lack of policies, protocols, or guidelines. Among factors relating to personnel, the most common were that knowledge and skills of staff were lacking (including failure to maintain competence) and that there had been failures of communication between staff. Reasons for barriers to access or engagement with care most often identified were factors such as substance use and family violence, and the mother or her family's perceived failure to recognize the complexity or seriousness of her condition.
      The death was not thought to have been potentially avoidable in all cases where there were contributory factors identified. The expert panel identified potentially avoidability in 35% of maternal deaths. In cases where death was potentially avoidable, there were nearly always 2 or 3 domains where contributory factors were identified. Table 4 reports on the deaths where contributory factors were present from the 4 years.
      TABLE 4Potential avoidability among maternal deaths with contributory factors 2006 through 2009
      CaseWas death potentially avoidable?DirectIndirectContributory factors
      No. of categories presentOrganizational and/or management factorsFactors relating to personnelTechnology and equipmentEnvironmentBarriers to accessing or engaging with care
      Case 1YesPostpartum hemorrhage3YesYesYes
      Case 2YesPreeclampsia3YesYesYes
      Case 3YesPreeclampsia2YesYes
      Case 4YesPulmonary embolism2YesYes
      Case 5YesPostpartum hemorrhage2YesYes
      Case 6YesSuicide3YesYesYes
      Case 7YesPreexisting medical condition3YesYesYes
      Case 8YesSuicide3YesYesYes
      Case 9YesNonobstetric sepsis3YesYesYes
      Case 10YesSuicide3YesYesYes
      Case 11YesNonobstetric sepsis3YesYesYes
      Case 12YesPreexisting medical condition2YesYes
      Case 13YesSuicide2YesYes
      Case 14YesPreexisting medical condition2YesYes
      Case 15YesSuicide2YesYes
      Case 16YesPreexisting medical condition1Yes
      Case 17NoAmniotic fluid embolism3YesYesYes
      Case 18NoAmniotic fluid embolism3YesYesYes
      Case 19NoAmniotic fluid embolism1Yes
      Case 20NoSepsis1Yes
      Case 21NoPreeclampsia1Yes
      Case 22NoAmniotic fluid embolism1Yes
      Case 23NoSuicide2YesYes
      Case 24NoSuicide1Yes
      Case 25NoPreexisting medical condition1Yes
      Case 26NoPreexisting medical condition1Yes
      Case 27NoPreexisting medical condition1Yes
      Total17171317
      Farquhar. Potential avoidability in maternal deaths in New Zealand 2006 through 2009. Am J Obstet Gynecol 2011.

      Comment

      This article describes a tool for measuring contributory factors and then determining potentially avoidable maternal deaths in New Zealand. In over one half of cases there were contributory factors present and in over one third of all cases the deaths were considered to be potentially avoidable. The tool was developed after reviewing the literature for root cause analysis and quality improvement as well as previously described classification systems for factors contributing to maternal mortality. Our review of current models for assessing contributory factors revealed many common themes but did not identify a system that adequately met the requirement of our national maternal mortality review. The approach that we have developed draws heavily on the London protocol but goes one step further to determine potential avoidability.
      • Taylor-Adams S.
      • Donabedian A.
      Advantages and limitations of explicit criteria for assessing the quality of health care.
      Establishing a tool for identifying contributory factors and potentially avoidable deaths provides a standardized framework that we consider will allow maternal deaths and severe maternal morbidities to be “measured” in the same way.
      There are limitations to our approach. One limitation is that the review is performed at arms length from the health care setting and that there may be information that the reviewers are unaware of that might have changed the outcome of the review. This concern could be addressed further by a study comparing local review with an external expert panel. A further limitation is that although we have shown the assessment of potential avoidability to be reliable over time (comparing retrospective review by the expert panel with prospective review by the working group), we have not provided definitive evidence that the tool, including its assessment of contributory factors, can be applied reliably across different multidisciplinary groups.
      International approaches were considered during the development phase although only CMACE
      • Cantwell R.
      • Clutton-Brock T.
      • Cooper G.
      • et al.
      Saving mother' lives: reviewing maternal deaths to make motherhood safer: 2006-2008. The eighth report of the confidential enquiries into maternal deaths in the United Kingdom.
      in the United Kingdom and Geller et al
      • Geller S.E.
      • Rosenberg D.
      • Cox S.
      • Brown M.
      • Simonson L.
      • Kilpatrick S.
      A scoring system identified near-miss maternal morbidity during pregnancy.
      in the United States reported on avoidability in maternal mortality. CMACE classified 70% of direct deaths and 55% of indirect deaths as receiving substandard care in 2006 through 2008 although no description of the method of identifying or classifying substandard care is provided in the report.
      • Cantwell R.
      • Clutton-Brock T.
      • Cooper G.
      • et al.
      Saving mother' lives: reviewing maternal deaths to make motherhood safer: 2006-2008. The eighth report of the confidential enquiries into maternal deaths in the United Kingdom.
      However, common themes between our report and the CMACE report include clinical knowledge and skills, failure to identify very sick women, best management of high-risk women, and improving communication. The Dutch Perinatal Audit Project identified 3 levels of substandard care; professional, organizational, and patient-related and substandard care factors were present in 32% of cases.
      • De Reu P.
      • Van Diem M.
      • Eskes M.
      • et al.
      The Dutch perinatal audit project: a feasibility study for nationwide perinatal audit in the Netherlands.
      Deviations from appropriate guidelines or best practice were used to identify substandard care, and smoking and obesity were included in the list of patient factors. Including these factors in the assessment of potential avoidability is contentious although the link with adverse outcomes is clearly established and addressing these risk factors with weight reduction and smoking cessation programs prior to and during pregnancy has been shown to improve outcomes.
      • McCowan L.M.
      • Dekker G.A.
      • Chan E.
      • et al.
      SCOPE Consortium
      Spontaneous preterm birth and small for gestational age infants in women who stop smoking early in pregnancy: prospective cohort study.
      We have chosen to exclude these risk factors from our list of contributory factors and to report these comorbidities separately. We also considered the system developed by Geller et al
      • Geller S.E.
      • Rosenberg D.
      • Cox S.
      • Brown M.
      • Simonson L.
      • Kilpatrick S.
      A scoring system identified near-miss maternal morbidity during pregnancy.
      • Geller S.E.
      • Rosenberg D.
      • Cox S.M.
      • et al.
      The continuum of maternal morbidity and mortality: factors associated with severity.
      • Geller S.E.
      • Cox S.M.
      • Callaghan W.M.
      • Berg C.J.
      Morbidity and mortality in pregnancy: laying the groundwork for safe motherhood.
      for reporting maternal morbidity. This approach identifies contributory provider and system factors and determines preventability and has been used in 1 region in the United States. One reason for not adopting this approach was that the delivery of maternity services in the United States is very different from New Zealand where 85% of maternity care is provided by midwives. A second reason was the lack of recognition of the importance of barriers to accessing and engaging with care.
      The implication of this report is that national maternal mortality review with appropriate recommendations and suggestions for implementation could improve maternity services in New Zealand as a proportion of maternal deaths could be prevented. Recommendations could be directed at any level, eg, government, health care providers, clinicians, or to pregnant women and their families, and may include education, a change in practice or system, or development of clinical guidelines or new policies.
      World Health Organization
      Beyond the numbers: reviewing maternal deaths and complications to make pregnancy safer WHO 20074.
      In the United Kingdom where evaluation of substandard care has been undertaken for 4 triennia there is a reduction in maternal deaths.
      • Cantwell R.
      • Clutton-Brock T.
      • Cooper G.
      • et al.
      Saving mother' lives: reviewing maternal deaths to make motherhood safer: 2006-2008. The eighth report of the confidential enquiries into maternal deaths in the United Kingdom.
      The Dutch perinatal project was a feasibility study and the impact of this methodology on perinatal mortality rates has not yet been reported.
      • De Reu P.
      • Van Diem M.
      • Eskes M.
      • et al.
      The Dutch perinatal audit project: a feasibility study for nationwide perinatal audit in the Netherlands.
      Although there are many studies of hospital mortality review few have considered avoidability or standard of care. One example is the National Health Service Institute for Innovation and Improvement in the United Kingdom. They reported that hospitalized standardized mortality rates were reduced 2% in 1 year after causes of avoidable mortality were identified and a range of interventions were introduced.
      NHS Institute for Innovation and Improvement
      Scoping of evidence in relation to interventions to reduce avoidable mortality.
      In New Zealand, suicide was the most common cause of maternal mortality and 7 of 10 suicides had contributory features and 5 of 10 were found to be potentially avoidable. Barriers to accessing and engaging care were the most common contributory factor, specifically due to the woman's mental health. Other barriers were substance abuse and family violence. Common organizational factors were lack of policies, protocols, or guidelines and inadequate systems/process for sharing of clinical information between services. Factors relating to personnel identified included knowledge and skills of staff were lacking, failure of communication between staff, and lack of recognition of complexity or seriousness of condition. In the CMACE report, family violence and substance abuse was also found to be a contributory factor in deaths from suicide. Of the women known to be involved with psychiatric services (69%), psychiatric care was found to be less than optimal. As a result of maternal mortality review, there have been recommendations for better coordination of mental health services within maternity care, greater recognition of mental health problems in pregnant women, and improved access to mother and baby units. If our approach is successful then the number of maternal deaths from suicide will hopefully decline.
      Reporting maternal mortality is an important first step in measuring the quality and safety of a maternity system. Identifying and reporting on contributory factors associated with mortality and morbidity is the next obvious step and may eventually become a more useful and meaningful measure of the safety and quality of care provided. We consider that this tool could be applied in mortality and morbidity reviews to identify where changes could be made to improve the quality of care. In the future we hope to report on the use of this tool for perinatal mortality and morbidity as well as maternal morbidity. At a national level confidential enquiry and identification of contributory factors and potentially avoidable deaths should lead to the introduction of policy changes and interventions that will contribute to clinical excellence, improve outcomes, and reduce maternal deaths.

      Acknowledgments

      We are grateful to the members of the MMRWG of the PMMRC for reviewing all the cases (http://www.pmmrc.health.govt.nz/moh.nsf/indexcm/pmmrc-home?Open&m_id=1). We are also grateful to the clinicians for providing the information to the committee and the secretariat of the PMMRC at the Ministry of Health, New Zealand.

      References

        • PMMRC 2010
        Perinatal and Maternal Mortality in New Zealand 2008. Fourth report to the Minister of Health. July 2009 to June 2010. Wellington: Ministry of Health.
        (Accessed Aug. 31, 2011)
        • NHS Institute for Innovation and Improvement
        Reducing avoidable mortality.
        (Accessed Aug. 31, 2011)
        • Cantwell R.
        • Clutton-Brock T.
        • Cooper G.
        • et al.
        Saving mother' lives: reviewing maternal deaths to make motherhood safer: 2006-2008. The eighth report of the confidential enquiries into maternal deaths in the United Kingdom.
        (Accessed Aug. 31, 2011)
        • De Reu P.
        • Van Diem M.
        • Eskes M.
        • et al.
        The Dutch perinatal audit project: a feasibility study for nationwide perinatal audit in the Netherlands.
        Acta Obstet Gynecol. 2009; 88: 1201-1208
      1. The 13th report of the perinatal and infant mortality committee of Western Australia for deaths in the triennium 2005-2007.
        (Accessed Aug. 31, 2011)
        • Geller S.E.
        • Rosenberg D.
        • Cox S.
        • Brown M.
        • Simonson L.
        • Kilpatrick S.
        A scoring system identified near-miss maternal morbidity during pregnancy.
        J Clin Epidemiol. 2004; 57: 716-720
      2. Centre for Maternal and Child Enquiries.
        (Accessed Aug. 31, 2011)
        • Cronin C.
        Five years of learning from analysis of clinical occurrences in pediatric care using the London protocol.
        Healthc Q. 2006; 9: 16-21
        • Szekendi M.
        Using patient safety morbidity and mortality conferences to promote transparency and a culture of safety.
        Jt Comm J Qual Patient Saf. 2010; 36: 3-9
        • Institute for Healthcare Improvement
        The London protocol.
        (Accessed Feb. 18, 2011)
        • Taylor-Adams S.
        Systems analysis of clinical incidents.
        (Accessed Aug. 31, 2011)
        • Geller S.E.
        • Rosenberg D.
        • Cox S.M.
        • et al.
        The continuum of maternal morbidity and mortality: factors associated with severity.
        Am J Obstet Gynaecol. 2004; 191: 939-944
        • Geller S.E.
        • Cox S.M.
        • Callaghan W.M.
        • Berg C.J.
        Morbidity and mortality in pregnancy: laying the groundwork for safe motherhood.
        Womens Health Issues. 2006; 16: 176-188
        • Wilson R.W.
        • Runciman W.B.
        • Gibberd R.W.
        • Harrison B.T.
        • Newby L.
        • Hamilton J.D.
        The quality in Australian health care study.
        Med J Aust. 1995; 163: 458-471
        • PMMRC 2008
        Perinatal and Maternal Mortality in New Zealand 2006. Second report to the Minister of Health. July 2007 to June 2008. Wellington: Ministry of Health.
        (Accessed Aug. 31, 2011)
        • PMMRC 2009
        Perinatal and Maternal Mortality in New Zealand 2007. Third report to the Minister of Health. July 2008 to June 2009. Wellington: Ministry of Health.
        (Accessed Aug. 31, 2011)
        • PMMRC 2011
        Perinatal and Maternal Mortality in New Zealand 2009. Fifth report to the Minister of Health. July 2010 to June 2011. Wellington: Ministry of Health.
        (Accessed Aug. 31, 2011)
        • Donabedian A.
        Advantages and limitations of explicit criteria for assessing the quality of health care.
        Milbank Mem Fund Q Health Soc. 1981; 59: 99-106
        • McCowan L.M.
        • Dekker G.A.
        • Chan E.
        • et al.
        • SCOPE Consortium
        Spontaneous preterm birth and small for gestational age infants in women who stop smoking early in pregnancy: prospective cohort study.
        BMJ. 2009; 338: b1081
        • World Health Organization
        Beyond the numbers: reviewing maternal deaths and complications to make pregnancy safer.
        (Accessed Aug. 31, 2011)
        • NHS Institute for Innovation and Improvement
        Scoping of evidence in relation to interventions to reduce avoidable mortality.
        (Accessed Aug. 31, 2011)