Regression with autocorrelated data: A study of time trend of global infant mortality rate

The infant mortality rate (IMR) is considered to be one of the most important indices of a country's well-being. Countries around the world and other health organizations like the World Health Organization are dedicating their resources, knowledge and energy to reduce the infant mortality rates. The well-known Millennium Development Goal 4 (MDG 4), whose aim is to archive a two thirds reduction of the under-five mortality rate between 1990 and 2015, is an example of the commitment. ^ In this study our goal is to model the trends of IMR between the 1950s to 2010s for selected countries. We would like to know how the IMR is changing overtime and how it differs across countries. ^ IMR data collected over time forms a time series. The repeated observations of IMR time series are not statistically independent. So in modeling the trend of IMR, it is necessary to account for these correlations. We proposed to use the generalized least squares method in general linear models setting to deal with the variance-covariance structure in our model. In order to estimate the variance-covariance matrix, we referred to the time-series models, especially the autoregressive and moving average models. Furthermore, we will compared results from general linear model with correlation structure to that from ordinary least squares method without taking into account the correlation structure to check how significantly the estimates change.^

Overcrowding and Infant Mortality: A Tale of Two Cities

Using detailed historical data for the cities of Glasgow and Edinhurgh, evidence is found in support of the hypothesis that overcrowding is a significant cause of infant mortality. We distinguish between voluntary overcrowding (due to the budgetary choices of poor families) and involuntary overcrowding (due to market failure in the provision of an adequate supply of appropriate housing). We found that, over the fifty year period, 1911-1961, Glasgow's infant mortality rate was significantly higher than that of Edinburgh, despite their close geographical proximity, and that a large part of the difference can he attributed to involuntary overcrowding in the first half of the twentieth century. We argue that this was due to the distinctly different housing policies adopted by the two cities, with important lessons for present day public authorities.

Infant mortality rates according to socioeconomic status in a Brazilian city

OBJECTIVE: Data from municipal databases can be used to plan interventions aimed at reducing inequities in health care. The objective of the study was to determine the distribution of infant mortality according to an urban geoeconomic classification using routinely collected municipal data. METHODS: All live births (total of 42,381) and infant deaths (total of 731) that occurred between 1994 and 1998 in Ribeirão Preto, Brazil, were considered. Four different geoeconomic areas were defined according to the family head's income in each administrative urban zone. RESULTS: The trends for infant mortality rate and its different components, neonatal mortality rate and post-neonatal mortality rate, decreased in Ribeirão Preto from 1994 to 1998 (chi-square for trend, p<0.05). These rates were inversely correlated with the distribution of lower salaries in the geoeconomic areas (less than 5 minimum wages per family head), in particular the post-neonatal mortality rate (chi-square for trend, p<0.05). Finally, the poor area showed a steady increase in excess infant mortality. CONCLUSIONS: The results indicate that infant mortality rates are associated with social inequality and can be monitored using municipal databases. The findings also suggest an increase in the impact of social inequality on infant health in Ribeirão Preto, especially in the poor area. The monitoring of health inequalities using municipal databases may be an increasingly more useful tool given the continuous decentralization of health management at the municipal level in Brazil.

Narrowing inequalities in infant mortality in Southern Brazil

OBJECTIVE: To determine the trends of infant mortality from 1995 to 1999 according to a geographic area-based measure of maternal education in Porto Alegre, Brazil. METHODS: A registry-based study was carried out and a municipal database created in 1994 was used. All live births (n=119,170) and infant deaths (n=1,934) were considered. Five different geographic areas were defined according to quintiles of the percentage of low maternal educational level (<6 years of schooling): high, medium high, medium, medium low, and low. The chi-square test for trend was used to compare rates between years. Incidence rate ratio was calculated using Poisson regression to identify excess infant mortality in poorer areas compared to higher schooling areas. RESULTS: The infant mortality rate (IMR) decreased steadily from 18.38 deaths per 1,000 live births in 1995 to 12.21 in 1999 (chi-square for trend p<0.001). Both neonatal and post-neonatal mortality rates decreased although the drop seemed to be steeper for the post-neonatal component. The higher decline was seen in poorer areas. CONCLUSION: Inequalities in IMR seem to have decreased due to a steeper reduction in both neonatal and post-neonatal components of infant mortality in lower maternal schooling area.

Infant mortality: comparison between two birth cohorts from Southeast and Northeast, Brazil

OBJECTIVE: To obtain population estimates and profile risk factors for infant mortality in two birth cohorts and compare them among cities of different regions in Brazil. METHODS: In Ribeirão Preto, southeast Brazil, infant mortality was determined in a third of hospital live births (2,846 singleton deliveries) in 1994. In São Luís, northeast Brazil, data were obtained using systematic sampling of births stratified by maternity unit (2,443 singleton deliveries) in 1997-1998. Mothers answered standardized questionnaires shortly after delivery and information on infant deaths was retrieved from hospitals, registries and the States Health Secretarys' Office. The relative risk (RR) was estimated by Poisson regression. RESULTS: In São Luís, the infant mortality rate was 26.6/1,000 live births, the neonatal mortality rate was 18.4/1,000 and the post-neonatal mortality rate was 8.2/1,000, all higher than those observed in Ribeirão Preto (16.9, 10.9 and 6.0 per 1,000, respectively). Adjusted analysis revealed that previous stillbirths (RR=3.67 vs 4.13) and maternal age <18 years (RR=2.62 vs 2.59) were risk factors for infant mortality in the two cities. Inadequate prenatal care (RR=2.00) and male sex (RR=1.79) were risk factors in São Luís only, and a dwelling with 5 or more residents was a protective factor (RR=0.53). In Ribeirão Preto, maternal smoking was associated with infant mortality (RR=2.64). CONCLUSIONS: In addition to socioeconomic inequalities, differences in access to and quality of medical care between cities had an impact on infant mortality rates.

Case-control study on infant mortality in Southern Brazil

OBJECTIVE: To identify risk factors associated with infant mortality and, more specifically, with neonatal mortality. METHODS: A case-control study was carried out in the municipality of Caxias do Sul, Southern Brazil. Characteristics of prenatal care and causes of mortality were assessed for all live births in the 2001-2002 period with a completed live-birth certificate and whose mothers lived in the municipality. Cases were defined as all deaths within the first year of life. As controls, there were selected the two children born immediately after each case in the same hospital, who were of the same sex, and did not die within their first year of life. Multivariate analysis was performed using conditional logistic regression. RESULTS: There was a reduction in infant mortality, the greatest reduction was observed in the post-neonatal period. The variables gestational age (<36 weeks), birth weight (<2,500 g), and 5-minute Apgar (<6) remained in the final model of the multivariate analysis, after adjustment. CONCLUSIONS: Perinatal conditions comprise almost the totality of neonatal deaths, and the majority of deaths occur at delivery. The challenge for reducing infant mortality rate in the city is to reduce the mortality by perinatal conditions in the neonatal period.

Infant Mortality in Novo Hamburgo: Associated Factors and Cardiovascular Causes

Background: Infant mortality has decreased in Brazil, but remains high as compared to that of other developing countries. In 2010, the Rio Grande do Sul state had the lowest infant mortality rate in Brazil. However, the municipality of Novo Hamburgo had the highest infant mortality rate in the Porto Alegre metropolitan region. Objective: To describe the causes of infant mortality in the municipality of Novo Hamburgo from 2007 to 2010, identifying which causes were related to heart diseases and if they were diagnosed in the prenatal period, and to assess the access to healthcare services. Methods: This study assessed infants of the municipality of Novo Hamburgo, who died, and whose data were collected from the infant death investigation records. Results: Of the 157 deaths in that period, 35.3% were reducible through diagnosis and early treatment, 25% were reducible through partnership with other sectors, 19.2% were non-preventable, 11.5% were reducible by means of appropriate pregnancy monitoring, 5.1% were reducible through appropriate delivery care, and 3.8% were ill defined. The major cause of death related to heart disease (13.4%), which was significantly associated with the variables ‘age at death’, ‘gestational age’ and ‘birth weight’. Regarding access to healthcare services, 60.9% of the pregnant women had a maximum of six prenatal visits. Conclusion: It is mandatory to enhance prenatal care and newborn care at hospitals and basic healthcare units to prevent infant mortality.

Infant Mortality in Iowa Interim Study Committee, January 1992

The Infant Mortality in Iowa Interim Study Committee was established by the Legislative Council to review historical trends in Iowa's infant mortality rates to identify the extent of the problem on a statewide basis. Identify areas in the state with the greatest incidence of infant mortality, and research health complications. Identify factors which lead to impoverished families, and research access to health care services. Survey and review the current structure of service provided to pregnant women in Iowa health care facilities, and solicit information on the level of existing prenatal services. Recommend changes in Iowa's health care system which would lower Iowa's infant mortality rate.

Black/white disparities in infant mortality rates in the U.S.: A quality of care perspective

The infant mortality rate for non-Hispanic Black infants in the U.S. is 13.63 deaths per 1,000 live births while the IMR for non-Hispanic White persons in the U.S. is 5.76 deaths per 1,000 live births. Black women are 2 times as likely as White women to deliver preterm infants and Black women are 2 times as likely as White women to deliver low birth weight infants (weighing less than 2,500 grams at birth). Differential underlying risk factors among mothers of different racial/ethnic groups for delivering pre-term and low birth weight infants have been historically accepted as the cause of racial disparities in IMRs. However, differential underlying risk status may not be the only major causative factor. Differential or unequal access to and provision of care is widely speculated to be a leading contributing factor to the wide racial disparity in infant mortality.2 This paper conducts a systematic review of existing literature investigating racial disparities in obstetrical care provided by healthcare practitioners to evaluate whether inequities in healthcare services provided to pregnant mothers and their neonates exist. The search terms "racial disparities obstetrical care," "racial differences quality of prenatal care," and "infant mortality racial disparities" were entered into the EBSCO Medline, Ovid Medline, PubMed, and Academic Search Complete databases, and articles between years 1990–2011 were selected for abstract review. The only articles included were those that used statistical methods to assess whether racial inequalities were present in the obstetrical services provided to pregnant women. My literature search returned 5 articles. Four of the five studies yielded significant racial differences in obstetrical care. However, the one study that used a large, nationally representative valid sample did not represent significant differences. Thus, this review provides initial evidence for racial disparities in obstetrical care, but concludes that more studies are needed in this area. Not all of the studies reviewed were consistent in the use and measurement of services, and not all studies were significant. The policy and public health implications of possible racial disparities in obstetrical care include the need to develop standard of care protocols for ALL obstetrical patients across the United States to minimize and/or eliminate the inequities and differences in obstetrical services provided.^

Time-trends in infant mortality rates in the southern Europe

Infant Mortality Rate (IMR) has been considered an important health indicator in monitoring quality of health care. Objectives: To examine trends in IMR in Southern Europe (SE) over the last two decades. Methods: Number of live births and infant deaths reported for SE (Portugal, Italy, Greece and Spain) between 1990 and 2013 were abstracted from World Health Organization Database. Annual IMR per 1,000 live births was computed for each country and for SE as a whole. Joinpoint regression models were used to estimate average annual percent change (AAPC) in IMR and respective 95% Conidence Interval (95% CI) and to identify points in time when signiicant changes in trend occurred (joinpoints). Results: Between 1990 and 2013, IMR signiicantly decreased from 9.2 to 3.4 in SE and the highest decrease was observed from 1992 to 1998 (AAPC = -6.1%; 95%CI: -6.5%; -5.8%). Signiicant decline in IMR was observed in all countries but the pattern was different across countries. IMR varied across countries between 11.5 and 8.3 and between 3.1 and 3.8 in 2013; highest IMR were observed in Portugal for 1990 and in Greece for 2013. Most notable decreases in IMR were observed from 1990 to 1995 in Portugal (AAPC = -8.4% 95%CI: -8.8; -8.1), from 1993 to 1997 in Italy (AAPC = -6.6% 95%CI: -7.8; -5.5) from 1998 to 2006 in Greece (AAPC = -6.8% 95%CI: -7.0; -6.5), and from 1993 to 1996 in Spain (AAPC = -7.3% 95%CI: -9.0; -5.6). Decreases in IMR were signiicant during all time period in Italy and Greece but in Portugal and Spain IMR became unchangeable after 2010-2011. Conclusions: Decrease in IMR in countries of SE suggests steadily improvement in the quality of health care. However differences in the pattern of decrease across countries during the last years deserve particular attention.

Infant mortality and low birth weight in cities of Northeastern and Southeastern Brazil

OBJECTIVE: To compare estimates of low birth weight (LBW), preterm birth, small for gestational age (SGA), and infant mortality in two birth cohorts in Brazil. METHODS: The two cohorts were performed during the 1990s, in São Luís, located in a less developed area in Northeastern Brazil, and Ribeirão Preto, situated in a more developed region in Southeastern Brazil. Data from one-third of all live births in Ribeirão Preto in 1994 were collected (2,839 single deliveries). In São Luís, systematic sampling of deliveries stratified by maternity hospital was performed from 1997 to 1998 (2,439 single deliveries). The chi-squared (for categories and trends) and Student t tests were used in the statistical analyses. RESULTS: The LBW rate was lower in São Luís, thus presenting an epidemiological paradox. The preterm birth rates were similar, although expected to be higher in Ribeirão Preto because of the direct relationship between preterm birth and LBW. Dissociation between LBW and infant mortality was observed, since São Luís showed a lower LBW rate and higher infant mortality, while the opposite occurred in Ribeirão Preto. CONCLUSIONS: Higher prevalence of maternal smoking and better access to and quality of perinatal care, thereby leading to earlier medical interventions (cesarean section and induced preterm births) that resulted in more low weight live births than stillbirths in Ribeirão Preto, may explain these paradoxes. The ecological dissociation observed between LBW and infant mortality indicates that the LBW rate should no longer be systematically considered as an indicator of social development.

The circadian rhythm of the perinatal mortality rate in Switzerland.

The authors examine the relation between the perinatal mortality rate (PMR), birth weight in four categories, and hour of birth throughout the week in Switzerland, using data on 672,013 births and 5,764 perinatal deaths recorded between 1979 and 1987. From Monday to Friday, the PMR follows a circadian rhythm with a regular increase from early morning to evening, with a peak for babies born between 7 and 8 p.m. This pattern of variation has two main components: The circadian rhythms for the proportion of births in the four weight categories and the PMR circadian rhythm for babies weighing more than 2.5 kg. According to a cosinor model, which describes about 40% of the total variation in the PMR, the most important determinants are changes in the proportions of births: Low birth weight increases toward the afternoon and night. Mechanisms underlying the weight-specific timing of birth are discussed, including time selection of birth according to obstetric risks, the direct effect of neonatal and obstetric care, and chronobiologic behavior.

The circadian rhythm of the perinatal mortality rate in Switzerland.

The authors examine the relation between the perinatal mortality rate (PMR), birth weight in four categories, and hour of birth throughout the week in Switzerland, using data on 672,013 births and 5,764 perinatal deaths recorded between 1979 and 1987. From Monday to Friday, the PMR follows a circadian rhythm with a regular increase from early morning to evening, with a peak for babies born between 7 and 8 p.m. This pattern of variation has two main components: The circadian rhythms for the proportion of births in the four weight categories and the PMR circadian rhythm for babies weighing more than 2.5 kg. According to a cosinor model, which describes about 40% of the total variation in the PMR, the most important determinants are changes in the proportions of births: Low birth weight increases toward the afternoon and night. Mechanisms underlying the weight-specific timing of birth are discussed, including time selection of birth according to obstetric risks, the direct effect of neonatal and obstetric care, and chronobiologic behavior.

Indigent care and infant mortality in Texas

Infant Mortality has been made a point of emphasis by the Department of Health and Human Services in the Healthy People 2000, 2010, and 2020 priorities. This study used the Behavioral Model for Vulnerable Populations to consider a number of factors which impact infant mortality in the indigent populations in the State of Texas. The primary focus of this study was the enabling factor of community resources, specifically the program used by each county to provide care to their indigent population. The Legislature of the State of Texas requires that each state have a program set up within a Hospital District, Public Hospital or develop a County Indigent Health Care Program (CIHCP) in order to provide the basic health care needs of their most vulnerable residents. We sought to determine whether the development of a CICHP without an appointed hospital to provide the care would have an adverse effect on residents seeking care and increase infant mortality. A Poisson Regression Analysis was used to analyze incidence rate ratios adjusting for race/ethnicity and wealth/poverty variables. Our study showed that counties using a CIHCP had significantly lower infant mortality rates when compared to counties using a hospital district and were statistically equivalent to counties using a public hospital program or a combination of service programs. This relationship was maintained when adjusted incidence rate ratios were calculated. This may give evidence that counties struggling to fund a public hospital or hospital district may be able to find a more cost-effective alternative in the CIHCP without adversely affecting the health status of their residents. More cost-benefit analysis and controlling analysis must be done to further characterize this relationship. ^