Price elasticities of demand for curative health care with control for sample selectivity on endogenous illness: an analysis for Sri Lanka.

Estimation of demand for health care with samples of only the ill may bias estimates. Additionally, the lack of exogenous information, especially distance, about the alternative care providers causes omitted variable problems. This paper alleviates both problems through geographic mapping of facility information to individuals, combined with joint estimation of illness (health production) and health care demand. The joint estimation full sample demand results are compared to those from one equation estimation for only the ill sample. The results indicate that the selectivity problem is significant, but that for this sample the magnitude of the bias on the price coefficient is small.

The determinants of child mortality in the Philippines: estimation of a structural model.

"This paper employs a longitudinal data set from the Philippines to estimate a structural model of the determinants of infant mortality. To control for the endogeneity of some explanatory variables, a set of reduced form equations for these outcomes is estimated simultaneously with the mortality hazard. A non-parametric discrete factor method is used to control for common unobservable factors in the error terms. Birth interval and parity lose explanatory power once biological measures are considered in the model. The results indicate significant correlations across the equations' error terms. The significant effects of breastfeeding and growth on mortality are quantified in simulation experiments." The data are from a longitudinal survey conducted in metropolitan Cebu between 1983 and 1986.

Community-level determinants of contraceptive use in the Philippines: a structural analysis.

We use household and community data from the Philippines to estimate a multilevel model of contraceptive use. We go beyond previous efforts in this field by developing a structural model that recognizes joint endogeneity and the temporal ordering of variables. by considering a wider range of community influences on fertility behavior, and by employing an econometric procedure allowing for a multilevel error structure. The results suggest that there are significant effects on fertility behavior of community-level family planning services, labor-conditions, and infrastructure development. These results provide insights regarding the structural determinants of contraceptive use and fertility that are useful for drawing policy implications.

Age-specific determinants of stunting in Filipino children.

This study identifies age-specific factors related to new cases of stunting that develop in Filipino children from birth to 24 mo of age. Data come from nearly 3000 participants in the Cebu Longitudinal Health and Nutrition Survey, a community-based study conducted from 1983 to 1995. Length, morbidity, feeding and health-related data were collected bimonthly during home visits. Stunting (length >2 SD below the WHO age- and sex-specific medians) occurred in 69% of rural and 60% of urban children by 24 mo of age. We used a multivariate discrete time hazard model to estimate the likelihood of becoming stunted in each 2-mo interval. The likelihood of stunting was significantly increased by diarrhea, febrile respiratory infections, early supplemental feeding and low birth weight. The effect of birth weight was strongest in the first year. Breast-feeding, preventive health care and taller maternal stature significantly decreased the likelihood of stunting. Males were more likely to become stunted in the first year, whereas females were more likely to become stunted in the second year of life. Because stunting is strongly related to poor functional outcomes such as impaired intellectual development during childhood, and to short stature in adulthood, these results emphasize the need for prevention of growth retardation through promotion of prenatal care and breast-feeding, as well as control of infectious diseases.

PIP: The age-specific factors related to new cases of stunting that developed in Filipino children from birth to 24 months of age were identified. The initial cohort included all pregnant women who gave birth during 1983-1984. Data were derived from the Cebu Longitudinal Health and Nutrition Survey conducted from 1983 to 1995. 3080 single live births were followed bimonthly for 24 months and then revisited in 1991 and 1995. Length, morbidity, feeding and health-related data were collected bimonthly during home visits. Stunting (length 2 months below the World Health Organization's age- and sex-specific medians) occurred in 69% of rural and 60% of urban children by 24 months of age. A multivariate discrete time hazard model was used to estimate the likelihood of becoming stunted in each 2-month interval, analyzing the data of 2859 infants with length data at 2 months of age. By 12 months of age, 37.7% of rural and 35.1% of urban children were stunted. By 24 months, 68.8% of rural and 61.9% of urban children were stunted. The likelihood of stunting was significantly increased by diarrhea, febrile respiratory infections, early supplemental feeding, and low birth weight ( 2500 g). The effect of birth weight was strongest in the first year. The likelihood of stunting was dramatically higher for low-birth-weight infants throughout the first 2 years. Breast-feeding, preventive health care (maternal education and immunization) and taller maternal stature decreased the likelihood of stunting. Males were more likely to become stunted in the first year, whereas females were more likely to become stunted in the second year of life. The combined effects of morbidity and feeding showed at 12 months the likelihood of stunting of 0.25 for the low morbidity, optimal feeding group, whereas it was 0.51 in the high morbidity, inappropriately fed group. The prevention of growth retardation through the promotion of prenatal care and breast-feeding as well as the control of infectious diseases is stressed.

Trends in breakfast consumption of US adults between 1965 and 1991.

OBJECTIVE: To examine breakfast consumption patterns and trends between 1965 and 1991 for adults in the United States. DESIGN: Trends analysis pooling three cross-sectional surveys. SETTING: Nationally representative samples obtained from the Nationwide Food Consumption Survey (NFCS) of 1965, the NFCS of 1977-1978, and the 1989-1991 Continuing Survey of Food Intakes by Individuals. SUBJECTS: Adults aged 18 years and older: 6,274 in 1965, 18,033 in 1977-1978, and 10,812 in 1989-1991. All results reflect use of sampling weights, so results reflect nationally representative samples in each time period. MAIN OUTCOME MEASURES: Breakfast consumption, defined as the consumption of food and/or beverage between the hours of 5 AM and 9 AM, was the focus of the trends analysis. Population prevalence rates are reported for the entire population and population subgroups. STATISTICAL ANALYSIS PERFORMED: Probit analysis was used to identify factors associated with changes in breakfast consumption over time. RESULTS: Breakfast consumption declined in the 26-year period between 1965 and 1991 from 86% to 75% for US adults. Breakfast consumption increased with age, and the age differential increased over time. Urban-rural and South-non-South differences in breakfast consumption narrowed over time, whereas black-nonblack and college-noncollege differences increased slightly or remained constant. The nutritional quality of food consumed at breakfast has improved since 1965. CONCLUSIONS: Although part of the decline in breakfast consumption can be explained by personal and demographic determinants, other unknown factors contributed to the trends. Elucidation of such factors is necessary to predict differences in breakfast as a health-related behavior.

The initiation of adolescent sexual and contraceptive behavior during changing times.

This paper examines changes in the initiation of adolescent sexual and contraceptive behavior in the United States between 1978 and 1988. Since a number of contextual changes occurred during this time period including a focusing of public attention on adolescent pregnancy along with the widespread publicity surrounding AIDS, we expected that the response to these events not only would change over time, but would also vary across social groups. Using data from Cycles III and IV of the National Survey of Family Growth, we find that the overall population patterns of earlier initiation of sexual intercourse and increased use of condoms at first intercourse are not found in all segments of the population. In general, the effects of race, religion, mother's education, and age changed during this time period. The long-term trend of younger age at first intercourse was halted for Blacks, and reversed for White, fundamentalist Protestants, but continued for all other Whites. Overall, patterns throughout the decade suggest that pressures from parents, religious groups, and others either lead to a later age at first intercourse, or use of contraception, but not both. A notable exception is that increased maternal education leads to both a later age at first intercourse and a higher likelihood of using contraception at first intercourse.

PIP: Over the period 1978-88, the US public gave increasing attention to adolescent pregnancy and AIDS. In that context, one might expect changes in the initiation of adolescent sexual and contraceptive behaviors. The authors examined such changes. The analysis of data from the third and fourth cycles of the National Survey of Family Growth failed to demonstrate the existence of trends toward the earlier initiation of sexual intercourse and increased use of condoms at first intercourse in all population segments. The effects of race, religion, mother's education, and age changed during the period. The long-term trend of younger age at first intercourse continued for all Whites except those who are fundamentalist Protestants, ended for Blacks, and reversed for White, fundamentalist Protestants. Overall patterns throughout the decade suggest that pressures from parents, religious groups, and others lead to either a later age at first intercourse or use of contraception, but not both. Increased maternal education, however, leads to both a later age at first intercourse and a higher likelihood of using contraception at first intercourse.

Community and contraceptive choice in rural Thailand: a case study of Nang Rong.

This paper blends quantitative with qualitative data in an investigation of community and contraceptive choice in Nang Rong, Thailand. Specifically, it develops an explanation of 1) method dominance within villages, coupled with 2) marked differences between villages in the popularity of particular methods. The quantitative analysis demonstrates the importance of village location and placement of family planning services for patterns of contraceptive choice. The qualitative data provide a complementary perspective, emphasizing the importance of social as well as physical space and giving particular attention to the structure of conversational networks.

Quality of services and demand for health care in Nigeria: a multinomial probit estimation.

This study attempts to empirically answer three important policy questions for a population sample from Ogun State, Nigeria: 1. Would price (fee) increases for health care lead to large reductions of care usage or to shifts across types of care used? 2. Would price increases lead to net increases in revenues for the health system? 3. Would the price increases have larger impacts (in the form of reductions in health care usage) on lower income members of the population? Household data are combined with data on prices and quality of care, collected directly from facilities, to estimate the demand for outpatient health care. Many of the statistical problems of demand estimation with micro level data are avoided by an innovation--the first use of the multinomial probit estimation method for health demand. A separate but related problem, that the price data used in such studies are usually endogenous (in fact usually are expenditures, which are to a great degree determined by the actual care choice) is avoided by the collection of a specific exogenous price variable directly from the health providers. Because the health care 'good'--outpatient health care--can vary to such a degree across providers, quality of care must be controlled in order that the coefficients on prices and other variables will not be biased. A strong circumstantial case can be made that past estimation efforts probably underestimated the impact of prices of care on provider choices, because those providers charging higher prices also tend to provide higher quality care and those charging lower prices to provide care of lower quality. Because of this fear of bias on the extremely important price coefficient, effective control of the quality of the care available at the alternative accessible care providers is almost certainly at this time the most important marginal innovation to demand estimation. Most past researchers simply have not had available to them exogenous quality of care information collected via a facility (provider) survey. This study tried several health care provider quality variables and finally used three distinct variables which were statistically significant: (a) expenditure per person in population served; (b) percentage of times drugs are available; and (c) interviewers evaluation of the physical condition of the facility. Price of a visit to the facility is also included, and also is an exogenous variable collected directly from the alternative available providers. For the variables of most interest for this study, price and quality of care, the results are quite reasonable and much as expected.(ABSTRACT TRUNCATED AT 250 WORDS)

Binary outcomes and endogenous explanatory variables: tests and solutions with an application to the demand for contraceptive use in Tunisia.

Many demographic studies examine discrete outcomes, and researchers often suspect that some of the explanatory variables may be influenced by the same unobserved factors that determine the discrete outcome under examination. In linear models, the standard solution to this potential endogeneity bias is an estimator such as two-stage least squares. These methods have been extended to models with limited dependent variables, but there is little information on the performance of the methods in the types of data sets typically used in demographic research. This paper helps to fill this gap. It describes a simple analytic framework for estimating the effects of explanatory variables on discrete outcomes, which controls for the potential endogeneity of explanatory variables. It also discusses tests for exogeneity and joint determination of the outcomes and the explanatory variables. It summarizes the results of a Monte Carlo study of the performance of these techniques and uses these results to suggest how researchers should approach these problems in practice. We apply these methods to the examination of the impact of fertility intentions on contraceptive use, based on data from the 1988 Tunisia Demographic and Health Survey.

Nutrition, lactation, and birth spacing in Filipino women.

The Cebu Longitudinal Health and Nutrition Survey is used to examine the roles of women's nutrition and infant feeding in determining time from birth to menses and time from menses to conception. The analysis sample includes 2,648 Filipino women followed for 24 months postpartum. Recently devised statistical estimation techniques to control for unobserved heterogeneity and endogeneity are employed in estimating a two-state hazard model. Low body mass index and lower dietary fat intake are associated with increased duration of postpartum amenorrhea. Contraceptive use, high dietary fat consumption, higher parity, and absence of spouse predict a longer waiting time to conception once menses have returned. Simulation of the hazard model is used to examine the effects of the key nutrition and lactation factors.

Survival in the perinatal period: a prospective analysis.

A prospective study of 3080 Filipino mothers and non-twin births in 33 communities is used to study the determinants of mortality in week 1 postpartum. The results show significant nonlinear birth weight effects and the importance of environmental contamination, particularly for infants born by traditional methods at home, and several other intermediate and underlying determinants of perinatal mortality. The pathways through which important sociodemographic factors affect perinatal mortality are also presented.

PIP: The Cebu Study Team collected and analyzed prospective data on 3080 women living in Metropolitan Cebu, the Philippines, who delivered singletons between May 1983 and April 1984 to learn what the determinants of mortality during the first 7 days of life are. The team used a generalized method prohibit estimation technique to control for confounding relationships from unobserved heterogeneity and the endogeneity of intermediate variables, thus significantly reducing bias. The strongest determinant of early neonatal mortality was birth weight. The probability of dying among newborns weighing less than 2500 gm was 7% compared with .25% for newborns weighing about 3000 gm (p .01). A relatively high birth weight was also a determinant of early neonatal mortality. Fecal contamination around the home coupled with a traditional delivery also had a significant effect on mortality (probability of death, 1.39% vs. .22% for no fecal contamination p .01). Modern delivery in a home with fecal contamination was not a determinant of neonatal mortality (probability of death, .29% with contamination, .21% with no contamination, and .22% for traditional delivery with no contamination). Delivery during the first days of life (.27% vs. .19%; p .05). Yet seasonality was also a significant determinant of many endogenous variables (e.g., birth weight and environmental quality), so this finding could not be easily explained. These results should be replicated in other settings with larger sample sizes. Nevertheless, they did show the benefit of the underlying=intermediate-outcome behavioral framework (the estimation and simulation procedure) which elucidates effects of key socioeconomic factors on biological outcomes.

The duration of breast-feeding: how is it affected by biological, sociodemographic, health sector, and food industry factors?

Breast-feeding duration has important effects on a wide array of demographic and maternal and infant health outcomes. This study uses a discrete-time logit hazards model to explore the relationship of biological, sociodemographic, health sector, and food industry practices on the duration of full and partial breast-feeding in a sample of more than 2,600 infants from the Cebu Longitudinal Health and Nutrition Survey. Food industry and health sector practices (such as distribution of free samples of infant formula), the mother's perception of being pregnant, infant fatness, and a set of standard sociodemographic factors all affect breast-feeding duration significantly.

Eating patterns and energy and nutrient intakes of US women.

A longitudinal multivariate analysis was used to determine whether differences in energy and nutrient intakes were present for women classified into different eating patterns. Ten multidimensional eating patterns were created based on the proportion of energy consumed at home and at seven away-from-home locations. Data were from 1,120 women aged 19 through 50 years who were surveyed up to six times over a 1-year period as part of the 1985 Continuing Survey of Food Intake by Individuals, US Department of Agriculture. Data from 5,993 days were analyzed. To examine differences in energy and nutrient intakes, longitudinal multivariate analyses were used to control for eating pattern and factors such as demographics, season, and day of week. Younger women in the Fast Food eating pattern consumed the greatest intakes of energy, total fat, saturated fat, cholesterol, and sodium. Well-educated, higher-income women in the Restaurant pattern consumed diets with the highest overall fat density. Nutrient densities for dietary fiber, calcium, vitamin C, and folacin were particularly low in away-from-home eating patterns. In contrast, moderately educated, middle-aged and middle-income women in the Home Mixed eating pattern (70% at home, 30% away from home) consumed the most healthful diets. We conclude that knowledge of demographics such as income and education is not enough to target dietary interventions. Rather, educational efforts must consider both demographics and the location of away-from-home eating. This will allow development of behavioral change strategies that consider food choices dictated by the eating environment as well as personal knowledge and attitude factors related to adoption of healthful food choices.

Influences on the extent of breast-feeding: a prospective study in the Philippines.

There has been much controversy about the impact of the health sector and the infant food industry on breast-feeding behavior. This study links causal factors to breast-feeding decisions, using a longitudinal survey of more than 3,000 Filipino mother-infant pairs. Most factors decreasing the likelihood that mothers will breast-feed seem to be related to family economics. Delivery in a private hospital, urban residence, high income, absence of spouse, and having worked for wages affect adversely the initiation of breast-feeding. Formula advertising and distribution of samples appear to have relatively little impact on feeding decisions.

The production of infant health: input demand and health status differences related to gender of the infant.

The authors collected data on all births between May 1, 1983, and April 30, 1984, in the Cebu City area of the Philippines island of Cebu. Some 3,080 mothers were studied in order to "test the effects of gender preferences on infant health within the context of a well-defined model. This model considers the mechanisms by which sociocultural and biological factors affect infant health." The authors conclude that although there are some minor gender-related differences in health input demands, they should have little impact on the cohort's actual health outcomes.

Adolescent contraceptive method choices.

This article analyzes determinants of contraceptive method choices among adolescent women in the United States. By using data from the 1982 National Survey of Family Growth, we examined factors that differentiate users of various methods early in the sexual careers of teenaged women. We find that patterns of method choice not only vary by race and region within the United States but also change over the teenager's life course. In addition, among teenagers who did not use a method at first sex, the likelihood of adopting a method soon thereafter was low for both whites and blacks and was unaffected by social structural characteristics.

Changes in breast-feeding in the Philippines 1973-1983.

Only limited research has been conducted on the patterns and determinants of changes over time in the extent and duration of breast-feeding. Methods are used that allow the examination of the effects of changes in population characteristics and behavioral factors on breast-feeding outcomes. Comparable national surveys are employed to examine the factors associated with the combination of a decline in the extent of and an increase in the duration of breast-feeding between 1973 and 1983 in the Philippines. Important differences are found in the factors explaining the extent and duration of breast-feeding between 1973 and 1983.

Accessibility, quality of care and prenatal care use in the Philippines.

The patterns and determinants of prenatal care are examined through the use of a randomly selected sample of 3000 rural and urban women who were studied prospectively during pregnancy and at three or four days postpartum. A large number of policy factors were found to influence the choice of most frequently used type of traditional, modern public or modern private prenatal care and the number of visits to each type of care, but few affected the first month of visit. The quality of care provided, accessibility to this care, and insurance available to the mother all had important effects on prenatal patterns. Large differences exist in the set of feasible policy options for improving prenatal care in urban and rural areas.

Breast-feeding patterns and determinants in Jordan.

PIP: The purpose of this paper is to analyze the patterns and determinants of the extent and duration of breastfeeding in Jordan. Data is from a national sample of women surveyed by the Government of Jordan in 1976 as part of the World Fertility Survey program. Women's level of education appears to have the expected negative impact on the decision to ever breastfeed the child or not. Having other children aged 4-6 in the home seems to reduce the probability of ever breastfeeding. In terms of short and long-diration breastfeeders, mother's education has the expected negative effect within both groups of women, but father's education has a significant impact only for the longer duration breastfeeders. Working away from home seems to interfere with breastfeeding only at the longer durations. The effects on breastfeeding of being an urban resident or in-migrant are generally negative. The child's being male increases the likelihood of its being breast-fed only at longer durations. The use of the pill or other contracetpives has a larger negative impact on the short duration decision, but no significant effect on the long duration choice. Having female children 7 to 12 in the home is found to increase the probability of breastfeeding beyond 4.5 months among the short duration group. Policy inplications focus on factors influencing the decisions of women to continue breastfeeding for short to moderate durations. The statistical techniques used in this paper are contained in a technical appendix.^ieng

Cluster analysis of food consumption patterns of older Americans.

This study uses data from the 1977-78 Nationwide Food Consumption Survey (NFCS) to develop a classification scheme for differentiating individuals into groups that have similar patterns of food consumption. The article examines the nutritional adequacy of those food patterns and identifies the socioeconomic factors associated with each pattern. Cluster analysis is used to identify the food consumption patterns of a nationally representative sample of persons aged 65 through 74 years. The results indicate that the food patterns of older persons can be well categorized as light eaters, heavy eaters, or consumers of large amounts of alcoholic beverages, salty snack products, animal fat products, legumes, or sweets and desserts. Following those different food patterns leads to noteworthy differences in nutrient intakes. Ethnic group membership and residence status are found to be the most important socioeconomic factors associated with differences in the food patterns followed.