Reducing Infant Mortality Rates Using the Perinatal Periods of Risk Model

Transcription

1 Public Health Nursing Vol. 22 No. 1, pp /05 # Blackwell Publishing, Inc. POPULATIONS AT RISK ACROSS THE LIFESPAN Reducing Infant Mortality Rates Using the Perinatal Periods of Risk Model Paulette G. Burns ABSTRACT Despite decreases in the last 50 years, infant mortality rates in the United States remain higher than in other industrialized countries. Using overall infant mortality rates to determine the effectiveness of interventions does not help communities focus on particular underlying factors contributing to static, and sometimes increasing, community rates. This study was designed to determine and rank contributing factors to fetal-infant mortality in a specific community using the Perinatal Periods of Risk (PPOR) model. The PPOR model was used to map fetal-infant mortality for 1995 to 1998 in the Tulsa, Oklahoma, Healthy Start Program as compared to traditional calculation methods. The overall fetal-infant mortality rate using the PPOR model was 12.7 compared to 7.11 calculated using the traditional method. The maternal health cell rate was 5.4, maternal care cell rate was 2.9, newborn care cell was 1.9 compared to a 4.1 neonatal death rate calculated using the traditional method, and the infant health cell was 2.4 compared to a 2.9 postneonatal rate calculated using the traditional method. Because the highest infant mortality was in the maternal health cell, intervention strategies were designed to promote the health of women prior to and between pregnancies. The PPOR model was helpful in targeting interventions to reduce fetal-infant mortality based on the prioritization of contributing factors. Key words: fetal death, infant mortality, pregnancy outcomes, stillbirth. Infant mortality rates are an accepted indicator for measuring a nationõs health. The rate is representative of the health status and social well-being of any nation. Since 1950, the U.S. infant mortality rate has dropped from a high of 29.2 deaths per 1,000 live births (1950) to a low of 6.9 deaths per 1,000 live births (2000) [Centers for Disease Control (CDC), 2002a; Hoyert, Smith, Arias, & Murphy, 2001], with a target rate of 4.1 deaths per 1,000 live births in 2010 (CDC, 2002b). Even though the U.S. infant mortality rate has dropped significantly from 9.2 deaths per 1,000 live births (1990) to 6.9 deaths per 1,000 live births (2000), 24 other industrialized countries have infant mortality rates lower than those of the United States (Matthews, Menacker, & MacDorman, 2002). 2 Paulette G. Burns, R.N., Ph.D., Texas Christian University, Fort Worth, Texas. Correspondence to: Paulette G. Burns, Texas Christian University, TCU Box , 2800 University Drive, Fort Worth, TX p.burns@tcu.edu U.S. infant mortality rates vary widely by race of the mother, with the lowest rate of 3.5 deaths per 1,000 babies of Chinese mothers to a high of 13.5 deaths per 1,000 infants of Black mothers (Matthews et al., 2002). Infant mortality rates are greater among women who received no prenatal care, have less education, are unmarried, teenaged, and smokers, and for male infants, multiple births, and preterm low-birthweight babies. States in the South generally have higher infant mortality rates with the highest being 10.3 deaths per 1,000 births in Mississippi. In comparison, states in the North generally have lower infant mortality rates with the lowest being in Massachusetts at 5.0 deaths per 1,000 live births (Matthews et al., 2002). The three leading causes of infant death in 1999 were congenital malformations (21%), disorders related to short gestation and low birthweight (16%), and sudden infant death syndrome (SIDS) (9%) (Matthews et al., 2002). Using the yardstick of overall infant mortality rates to determine the effectiveness of interventions does not help communities focus on specific underlying factors contributing to the static, or sometimes

2 Burns et al.: Reducing Infant Mortality Rates Using the PPOR Model 3 increasing, infant mortality rates in communities. The urgent need to decrease infant mortality rates demands more focused and innovative strategies. Infant mortality rates are generally defined and calculated as the total number of infant deaths per 1,000 births divided by the total number of births per year. Infant mortality can further be differentiated by time of death after a live birth. The neonatal mortality rate is the number of infant deaths between less than 1 hr and 27 days after birth per 1,000 births divided by the total number of births for the year. The postneonatal mortality rate is the number of infant deaths between 28 and 364 days per 1,000 births divided by the total number of births. These methods of calculating mortality do not include fetal mortality or stillbirths. Calculating disparities in rates among different ethnic and racial groups, correlating variables associated with poor infant outcomes, and understanding the underlying causes of fetal-infant deaths are prerequisites in trying to monitor and improve infant outcomes. The complexity and multifactor etiologies of infant mortality can promote inertia because the problem seems too big to solve. A more parsimonious model that provides direction, focus, and suggests effective interventions has been developed. The Perinatal Periods of Risk (PPOR) model provides a different paradigm for approaching infant mortality rates. The model helps users to identify and rank four factors as they contribute to the overall infant mortality rate: (a) the motherõs health prior to and between pregnancies, (b) maternal health care systems, (c) neonatal health care systems, and (d) infant health during the first year of life. Targeting interventions to the factors most responsible for the infant mortality rate should help to reduce the rate more rapidly and effectively. The article reports on one project designed to test the use of the PPOR model in a specific population. The perinatal periods of risk model The Institute of MedicineÕs landmark (1988) document, The Future of Public Health, states that to provide effective public health programs and services focused on population groups, public health practice must be grounded in the functions of assessment, assurance, and policy development. One approach to reducing infant mortality that includes all three of the core public health functions is the use of the PPOR model developed in the late 1980s by Dr Brian McCarthy in collaboration with the World Health Organization and the Perinatal Collaborative Center at the CDC. The model has been used for over a decade to monitor and investigate fetal-infant mortality in developing countries. Since 1997, the model has been used and tested in the United States in urban settings through the Perinatal Periods of Risk work-group efforts. The work group consists of several urban centers under the name of CityMatch (2000), in conjunction with the CDC, the March of Dimes, and the University of Nebraska Medical Center ( The PPOR model involves a five-step process including (a) engaging community partners early to gain consensus and support in reducing infant mortality, (b) mapping fetal-infant mortality by birthweight and age of death, (c) focusing on reducing the overall fetal-infant mortality rate, (d) examining potential opportunity gaps between population groups, and (e) targeting further investigations and prevention efforts on the gaps ( The author and the work group focus intensely on the first step and provide strategies and tools on the website to engage community partners. Convincing community partners of the importance of targeting the problem of infant mortality through endeavors that make a difference in infant outcomes cannot be overemphasized. No step should be overlooked when using the model; however, the focus of this article will be on the second step of the model as used in the Tulsa Healthy Start Program. Theoretical Constructs of the PPOR Model The PPOR model is based on two major theoretical constructs: age of fetus-infant at death and birthweight. Traditionally, the age of the infant at death, whether neonatal or postneonatal, has been calculated and reported, but fetal deaths and stillbirths have not usually been a focus of attention. Each developmental stage is associated with different risk factors for poor birth outcome. For example, during the neonatal period, a baby is more at risk of death due to such factors as birth traumas, congenital anomalies, or sepsis. A major health system-level intervention focused on improving neonatal outcomes has included developing and providing access

3 4 Public Health Nursing Volume 22 Number 1 January/February 2005 to neonatal intensive care units (Goodman et al., 2002; Richardson et al., 1998). Postneonatal infants are at increased risk of death from SIDS; hence, health system-level interventions have focused on prevention of SIDS. One very successful prevention effort is the national Back to Sleep campaign (American Academy of Pediatrics, Task Force on Infant Sleep Position and Sudden Death Syndrome, 2002; Pollack & Frohna, 2002). Preterm babies of viable gestational age, generally considered 24 weeks, are at higher risk of death due to the nutritional state of the mother, medical conditions of the mother, smoking status of mother, genetic factors, and many others (Maloni, 2000; McCloskey et al., 1999). Interventions for this stage of development should focus on efforts to improve the health of the mother prenatally, and the general health and well-being of women of childbearing age. The birthweight of an infant is the strongest predictor at birth of the childõs survival. Risk factors related to a lower birthweight and risk factors related to mortality at a particular birthweight are often different. For example, a low birthweight at any gestational age may be related to socioeconomic status, race, or medical conditions of the mother (Maloni, 2000; MacDorman & Singh, 1998; Sable & Herman, 1997). Mortality differences in infants of low birthweight are often related to access to perinatal care, referral systems, and transport systems (Goodman et al., 2002). The PPOR model maps each death in a geographic region based on birthweight and age at death, including fetal, neonatal, and postneonatal periods as shown in Fig. 1. The model only includes infants weighing 500 g, including live births or fetal deaths. This cut-off point is due to large reporting differences in vital records across the United States. Also, infants are usually considered physically viable at 500 g. Only fetal-infant deaths occurring at 24+ weeks gestation are counted using the PPOR model. The PPOR model combines the top three cells (fetal, neonatal, and postneonatal) with birthweights of 500 1,499 g into one cell, designated the maternal health cell. Essentially, the very low birthweight infant deaths are combined into this one cell. The three remaining groups of fetal, neonatal, and postneonatal 1,500 g and over are renamed maternal care cell, newborn care cell, and infant health cell, respectively (Fig. 2). The name of each cell suggests the primary prevention focus for that group: maternal health/prematurity, maternal health care system changes, neonatal health care system changes, or infant health Birthweight 500 1,499 g 1,500 + g Fetal Fetal Age at death Neonatal Neonatal Postneonatal Postneonatal Figure 1. Fetal-infant mortality according to the PPOR model. F, fetal mortality; N, neonatal mortality; and PN, postneonatal mortality. The top three cells are for fetal, neonatal, and postneonatal infant deaths with birthweights of 500 1,499 g and 24+ weeks gestation. The bottom three cells are fetal, neonatal, and postneonatal deaths of infants with birthweights of 1,500+ g and 24+ weeks gestation at birth. during the first year of life. Because fetal deaths earlier than 24 weeks, live births less than 500 g, spontaneous abortions, and induced abortions are not included in the model, fetal-infant mortality may still be underestimated which is a limitation of this method. Resources for Fetal-Infant Mortality Data Infant mortality data for each state for 1995 to 1998 can be found on the CDC Wonder website ( A database Birthweight 500 1,499 g 1,500 + g Maternal care Age at death Maternal health/prematurity Newborn care Infant health Figure 2. PPOR fetal infant mortality with transformed cells. The upper fetal, neonatal, and postneonatal cells combined to form the maternal health cell, generally including very low-birthweight fetal-infant deaths. The lower three fetal, neonatal, and postneonatal cells including 1,500+ g and at least 24 weeks gestation transformed to the maternal care cell, newborn care cell, and infant health cell, respectively.

4 Burns et al.: Reducing Infant Mortality Rates Using the PPOR Model 5 of linked birth and death certificates for each state and each county per state is available. Other variables in the database include the motherõs age, motherõs ethnicity, marital status, month prenatal care began, gestational period, live birth order, birthweight, International Classification of Diseases (ICD) codes for death, and age of child at death. The database does not include fetal deaths; however, it contains information for infant neonatal deaths starting at 20 weeks gestation as opposed to the PPOR model which should be used for deaths occurring after 24 weeks, as noted above. The National Center for Health Statistics (2002) has a database of fetal death information on data tapes that can be ordered from the Center for a fee per tape. Data used to determine Tulsa County fetal deaths were obtained from the Oklahoma State Department of Health (1998a) and the Tulsa City County Health Department for 1995 to Targeting Nursing Interventions and Services Multiple interventions are important in reducing infant mortality, and the PPOR model guides prioritizing interventions based on the cell contributing the most to infant mortality rates. Interventions related to the maternal health cell are aimed at helping mothers be healthy prior to and during any pregnancy and ways to prevent prematurity. Some interventions related to this area include family planning and spacing, adequate nutrition, substanceabuse prevention and treatment, stress reduction efforts, smoking prevention and cessation, recognition of signs and symptoms of pregnancy, adequate exercise and sleep, education, and empowerment. Interventions focused on the maternal care cell include community services for high-risk follow-up, access to and utilization of prenatal care, trimester of entry into care, genetic counseling, pregnancy testing, surgical services, intrapartum and postpartum monitoring and education, anticipatory guidance, and surgical services for cesarean sections and permanent sterilization procedures. Interventions appropriate to decrease mortality in the newborn care cell focus on the health care available to new parents during this developmental stage such as access to obstetrical/pediatric providers and neonatal intensive care and newborn nursery services, resuscitation, thermal control, high-risk infant follow-up, and breastfeeding assistance and education. Interventions for the infant care cell focus on infant needs during the first year of life including adequate nutrition, injury control/safety, immunizations, no secondhand smoke, access to a medical home, developmental screening/well baby care, high-risk follow-up, and anticipatory guidance with the parents. The community project According to Matthews et al. (2002), Oklahoma had the highest infant mortality rate for non-hispanic white infants at 8.2 deaths per 1,000 live births in Between 1995 and 1998, OklahomaÕs infant mortality rate ranged from 7.5 to 8.5 deaths per 1,000 live births, consistently exceeding the national rate (Hoyert et al., 2001; Oklahoma State Department of Health, 1998b). Infant mortality rates in OklahomaÕs second largest population center, Tulsa County, are consistently higher than the U.S. rate and the Oklahoma rate except in 1997, ranging from 6.9to 9.4 deaths per 1,000 live births (Oklahoma State Department of Health, 1998a). The Tulsa Perinatal Coalition, a community-based organization representing over 30 community agencies, received federal funding through the Maternal Child Health Bureau ÒHealthy StartÓ program in The goal of the program was to reduce infant mortality by 50% in selected census tracts with at least a 9.0 rate of infant mortality. The PPOR model was used during phase II of the Tulsa Program beginning in Calculating Fetal-Infant Mortality Rates Using the PPOR Model The PPOR model definitions and calculations for fetal-infant mortality rates may be found at Originators of the model caution that there should be at least 60 deaths per cell before calculating the rate due to irregular proportional differences if less than 60. Using the PPOR model, rates for Tulsa County were calculated for cumulative years 1995 to The overall fetal-infant mortality rate was 12.7 compared to 7.11 calculated using the traditional method. The maternal health cell rate was 5.4, maternal care cell rate was 2.9, newborn care cell was 1.9 compared to a 4.1 neonatal death rate calculated using the traditional method, and the infant health cell was 2.4 compared to a 2.9postneonatal rate calculated

5 6 Public Health Nursing Volume 22 Number 1 January/February 2005 using the traditional method. Because the CDC database, Wonder, does not offer the category of 24 weeks gestation, 20 weeks gestation was used in calculating the Tulsa County rates. The PPOR calculation helps clearly distinguish which factor is contributing the most to the overall fetal-infant mortality rate. The maternal health cell was contributing 42% to the overall rate in Tulsa County. Focused Interventions of the Tulsa Healthy Start Program The focus of the Tulsa Healthy Start Program in the first 4 years was initiating services and general supports primarily for the prenatal period. New services included transportation to clinic visits, translation services during clinic visits for women with English as a second language, case management services, outreach efforts to recruit pregnant women into prenatal care, and health care education for previously underserved community groups through schools, juvenile facilities, and the city jail. After using the PPOR model to identify the highest risk groups, the staff of the Healthy Start Project, in concert with the Tulsa Perinatal Coalition, decided to target health promotion efforts to impact the health of the mother, as the maternal health cell was the major contributor to mortality. A focus on interconceptional care was added to the program through outreach, case management, and education staff. In addition, the Tulsa Fetal Infant Mortality Review Board findings were compared with the PPOR mortality rates. Many of the recommendations of the review board were related to improving the motherõs health prior to pregnancy. Recommendations for use of the PPOR model The Healthy Start staff and the Perinatal Coalition evaluation committee members found the model useful in planning targeted interventions for childbearing-age women. Many state and national database links were identified and used in the assessment, which strengthened the population focus. However, the quality and sufficiency of the data determines the accuracy of the assessment. Therefore, it is imperative to understand each process of data collection, storage, and retrieval in the public health-record-keeping system. Some of the important lessons learned using the PPOR model include: 1 Identify definitions and processes for completing and submitting birth and death certificates in your county and state. Ascertain if fetal deaths are part of the database or in a separate database and how they are completed and submitted. 2 Obtain direct access to the birth and death certificate data if possible. Ascertain the quality of the data by the percentage of linked birth and death certificates. The higher the percentage of unmatched birth and death certificates, the more chance for error in calculations. 3 Make sure there are at least 60 fetal-infant deaths to make sense of the four mortality groups. In order to achieve the 60 deaths, a larger geographic area may need to be used or a larger time period. However, the PPOR work group suggests that not more than a 5-year period should be combined due to health system and community changes over time. 4 Be ready to help community groups understand the differences in the mortality cells and then prioritize the intervention efforts. 5 Use the PPOR model to evaluate community interventions over time, so that comparisons are meaningful and engender further action. Conclusion Infant mortality remains a significant health problem of the United States. The PPOR model offers a new paradigm for approaching the problem and helping communities develop strategies and focus resources on population groups that contribute most to the mortality statistics. Using the PPOR model in Tulsa showed the need to focus interventions on the maternal health cell of the model. The PPOR model is accessible and easy to use with existing datasets to develop area-specific interventions. Population-focused public health nursing requires community partnerships and collaborations to improve the health outcomes for our citizens. Public

6 Burns et al.: Reducing Infant Mortality Rates Using the PPOR Model 7 health nurses play a vital role in initiating and developing such community coalitions. The PPOR model is a useful tool for community partnerships to use in targeting actions for the population group contributing most to infant mortality rates. Nurses working in other communities can use the model to mobilize their communities into action. Acknowledgments The Tulsa Perinatal Coalition and the Tulsa Healthy Start Staff were instrumental in directing and implementing all phases of the reported project. Their persistent commitment to helping underserved women have healthier pregnancies and infants is deeply appreciated and recognized. References American Academy of Pediatrics, Task Force on Infant Sleep Position and Sudden Death Syndrome. (2002). Changing concepts of sudden infant death syndrome: Implications for infant sleeping environment and sleep position. Pediatrics, 105, Centers for Disease Control & Prevention. (2002a). Infant mortality and low birth weight among black and white infants, United States, MMWR: Morbidity and Mortality Weekly Report, 51(27), Retrieved August 2, 2004, from mmwr/preview/mmwrhtml/mm5127a1.htm Centers for Disease Control & Prevention. (2002b). CDC Wonder Database Atlanta, GA: Author. Retrieved October 14, 2002, from Citymatch. (2000). Perinatal periods of risk approach in US cities. General description. Retrieved July 9, 2003, from Goodman,D.,Fisher,E.,Little,G.,Stukel,T.,Chang,C., & Schoendorf, K. (2002). The relation between the availability of neonatal intensive care and neonatal mortality. The New England Journal of Medicine, 346(20), Hoyert, D. L., Smith, B. L., Arias, E. & Murphy, S. L. (2001). Deaths: Final data for National Vital Statistics Reports: From the Centers for Disease Control and Prevention, National Center for Health Statistics, National Vital Statistics System, 49(8), Institute of Medicine. (1988). The future of public health. Washington, DC: Author. MacDorman, M. F., & Singh, G. K. (1998). Midwifery care, social and medical risk factors, and birth outcomes in the USA. Journal of Epidemiology and Community Health, 52(5), Maloni, J. (2000). Preventing preterm birth: Evidence-based interventions shift toward prevention. AWHONN Lifelines, 4, Matthews, T. J., Menacker, F., & MacDorman, M. (2002). Infant mortality statistics from the 2000 period linked birth/infant death data set. National Vital Statistics Reports: From the Centers for Disease Control and Prevention, National Center for Health Statistics, National Vital Statistics System, 50(12), McCloskey, L., Plough, A., Power, K., Higgins, C., Cruz, A., & Brown, E. (1999). A communitywide infant mortality review: Findings and implications. Public Health Reports, 114(2), National Center for Health Statistics. (2002). Data definitions. Washington, DC: Author. Retrieved October 14, 2002, from nchs/datawh/nchsdefs/rates.htm Oklahoma State Department of Health. (1998a). Profile of health status indicators, Oklahoma report of health, Tulsa County. Oklahoma City, OK: Author. Retrieved October, 2002, from planning/hsip/hsip98/tulsa Oklahoma State Department of Health. (1998b). State and County Health Status Indicator Profiles, Oklahoma Health Statistics Table I, Resident births, fetal deaths (stillbirths), deaths and infant deaths by 34 selected causes, number and rate per 100,000 population. Oklahoma City, OK: Author. Retrieved October 14, 2002, from program/phs/ohs/ohs98/index Pollack, H. A., & Frohna, J. G. (2002). Infant sleep placement after the back to sleep campaign. Pediatrics, 109, Richardson, D. K., Gray, J. E., Gortmaker, S. L., Goldmann, D. A., Pursley, D. M., & McCormick, M. C. (1998). Declining severity adjusted mortality: Evidence of improving neonatal intensive care. Pediatrics, 102, Sable, M. R., & Herman, A. A. (1997). The relationship between prenatal health behavior advice and low birth weight. Public Health Reports, 112,