Efforts to simplify the construction of the DHS wealth index are encouraged (while recognizing it is constructed differently in each country), but attempts to assess equity in health programs should bear in mind that it is not sufficient to calculate the wealth index just for the participants in the program. The quintile distributions can vary dramatically within sub-populations. Assessments of equity require knowledge of the distribution of potential participants as well as actual participants.

While the global objective is exclusive breastfeeding (EBF) for a full 6 months duration, the standard indicator is a "prevalence" indicator, that is, the percentage of all children under age 6 months who are exclusively breastfed at a point in time. That yields a higher percentage than a more direct indicator of duration and can be easily misunderstood, exaggerating the amount of EBF. A measurement of actual percentage of children exclusively breastfeeding for a full 6 months can be easily calculated from standard DHS and MICS data.

Between 1990 and 2010, maternal mortality decreased globally by nearly 50%, from 543 000 maternal deaths per year to 287 000,1WHOUNICEFUNFPAthe World BankTrends in maternal mortality 1990 to 2010. World Health Organization, Geneva2012Google Scholar with the greatest reductions in the second half of this period. A major catalyst for this progress was the target set by Millennium Development Goal 5: reduction of the maternal mortality ratio by 75% between 1990 and 2015. Later, a second target on reproductive health was added, which has undoubtedly contributed to accelerated progress. In tandem, countries and the international community enhanced their commitment to support, develop, and implement effective interventions to improve the health of women, especially during pregnancy and childbirth, and post-partum. As 2015 approaches, attention turns to what has been achieved—and what lies ahead for the global development goals, including those for maternal and child health.2Langer A Horton R Chalamilla G A manifesto for maternal health post-2015.Lancet. 2013; 381: 601-602Summary Full Text Full Text PDF PubMed Scopus (32) Google Scholar In 2012, encouraged by the substantial reduction in mortality for children younger than 5 years, the global community (spearheaded by the governments of Ethiopia, India, and the USA, in collaboration with UNICEF, WHO, and other partners) put forward a vision of ending preventable child deaths.3UNICEFState of the world's children. United Nations International Children's Emergency Fund, New York, NY2012Google Scholar, 4Glass RI Guttmacher AE Black RE Ending preventable child death in a generation.JAMA. 2012; 308: 141-142Crossref PubMed Scopus (26) Google Scholar As a benchmark of this vision, a target for post-2015 was proposed. The target that "all countries should achieve a national under-5 mortality rate (U5MR) of below 20 deaths per 1000 live births by 2035, and the global average U5MR should decline to 15 deaths per 1000 in 2035" was ambitious but achievable, according to analyses and modelling of the potential effect of interventions to reduce child mortality.4Glass RI Guttmacher AE Black RE Ending preventable child death in a generation.JAMA. 2012; 308: 141-142Crossref PubMed Scopus (26) Google Scholar, 5Winfrey W McKinnon R Stover J Methods used in the Lives Saved Tool (LIST).BMC Public Health. 2011; 11: S32Crossref PubMed Scopus (59) Google Scholar As countries and international platforms engage in post-2015 planning, now is the time to envision the ending of preventable maternal deaths, as stressed in a recent manifesto for maternal health.2Langer A Horton R Chalamilla G A manifesto for maternal health post-2015.Lancet. 2013; 381: 601-602Summary Full Text Full Text PDF PubMed Scopus (32) Google Scholar The global health community should build on past successes, and accelerate progress towards elimination of all preventable maternal mortality within a specified period. WHO, the US Government, and many others have started processes to define a new vision for maternal health, including discussion of potential targets. The initial technical discussions convened by WHO as part of this process, with country-level programme managers, scientists, professional organisations, donors, and other partner agencies, resulted in proposals to take forward a bold vision of ending preventable maternal deaths, around which the international community can rally. The analyses done as a background to the discussions show the feasibility of this vision. An ambitious but realistic global target is to reduce maternal mortality ratios to less than 50 per 100 000 livebirths by 2035. Dependent on which target year is selected, this would translate to a maternal mortality ratio of less than 90 by 2025, or 70 by 2030 (figure). In view of the accumulated experience of the global health community, with many countries achieving rapid reductions in maternal mortality in recent years, we believe that this target can be reached worldwide. Additionally, we propose a new method to measure and track progress in every country, based on achievement of defined milestone values according to initial maternal mortality ratios. This method would help to focus planning for maternal survival. For all countries with estimated maternal mortality ratios of less than 400 in 2010, the goal would be a steady progression past a series of 5 year milestones to reach the global target. The expectation that every country would cross one milestone within every 5 year interval will provide a method to measure each country's progress, and will also contribute to global progress. The 5 year milestones for countries with high initial maternal mortality ratios (>400) would be individually designed and tracked. Countries with an estimated maternal mortality ratio of less than 100 would be expected to move to lower values according to defined milestones, but with a focus on internal subpopulations whose maternal mortality is higher than the national rate. New strategies to help to achieve rapid reductions in maternal mortality are needed. They need to be innovative and dynamic, with the aim to reach equity between and within countries. Targeted interventions should be implemented at scale, to address causes and conditions contributing to maternal deaths that vary across countries. Collection of data about the causes and conditions of every maternal death—eg, through maternal death surveillance and response—is a necessary first step. Improved information enables development of strategies to respond to needs, and to address both clinical and distal causes of deaths, and will allow for targeted monitoring, accountability, and action. Strategies can be adapted to specific local causes, whether they are directly related to maternal mortality (eg, haemorrhage, pre-eclampsia or eclampsia, sepsis, or unsafe abortion) or indirectly related (eg, HIV, malaria, tuberculosis, anaemia, or non-communicable diseases). Strategies to implement targeted interventions to reduce maternal mortality need to address more than the clinical causes of death—they should respond to changing demographics, meet the specific needs of women for reproductive health, and address contextual features such as challenges caused by changes in health-care systems. These challenges include financial incentives, the effects of decentralisation, the role of the private sector, and urbanisation. Universal access to high-quality health services, including family planning and information and services for reproductive health (especially for vulnerable and at-risk populations), should be put at the centre of efforts to achieve the vision of ending maternal deaths. To accelerate and sustain the gains already achieved, international and national commitments are needed, such as those made in response to the UN Secretary General's Every Woman Every Child report. Additionally, national-level processes to sharpen strategies for improvement of health and development beyond 2015 have already started in many countries. Ending of preventable maternal death is not a dream. Mahmoud Fathalla, past president of the International Federation of Obstetricians and Gynaecologists and former Chair of the WHO Advisory Committee on Health Research, said: "Women are not dying because of untreatable diseases. They are dying because societies have yet to make the decision that their lives are worth saving."6Fathalla MF Human rights aspects of safe motherhood.Clin Obstet Gynecol. 2006; 20: 409-419Scopus (38) Google Scholar Momentum to value women's lives is growing—let us strengthen it. We declare that we have no conflicts of interest. The views expressed here are those of the authors themselves and they do not necessarily represent the views of WHO or other organisations. © 2013 World Health Organization; licensee Elsevier. This is an Open Access article published without any waiver of WHO's privileges and immunities under international law, convention, or agreement. This article should not be reproduced for use in association with the promotion of commercial products, services or any legal entity. There should be no suggestion that WHO endorses any specific organisation or products. The use of the WHO logo is not permitted. This notice should be preserved along with the article's original URL.

Chapter 31 Identifying and Adjusting for Recall Error with Application to Fertility Surveys Thomas W. Pullum, Thomas W. PullumSearch for more papers by this authorS. Lynne Stokes, S. Lynne StokesSearch for more papers by this author Thomas W. Pullum, Thomas W. PullumSearch for more papers by this authorS. Lynne Stokes, S. Lynne StokesSearch for more papers by this author Book Editor(s):Lars Lyberg, Lars LybergSearch for more papers by this authorPaul Biemer, Paul BiemerSearch for more papers by this authorMartin Collins, Martin CollinsSearch for more papers by this authorEdith De Leeuw, Edith De LeeuwSearch for more papers by this authorCathryn Dippo, Cathryn DippoSearch for more papers by this authorNorbert Schwarz, Norbert SchwarzSearch for more papers by this authorDennis Trewin, Dennis TrewinSearch for more papers by this author First published: 10 February 1997 https://doi.org/10.1002/9781118490013.ch31Citations: 1Book Series:Wiley Series in Probability and Statistics AboutPDFPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShareShare a linkShare onFacebookTwitterLinked InRedditWechat Summary This chapter contains sections titled: Introduction Background Individual-Level Estimation of True Rates and Misreporting Parameters Aggregate-Level Estimation of True Rates and Misreporting Parameters Discussion Acknowledgments Citing Literature Survey Measurement and Process Quality RelatedInformation

Recent years have seen the development of formal and microsimulation models of the structure and dynamics of kin networks. These models generally assume uncorrelated fertility within and across generations. Several sets of real data, however, show positive correlations between the frequencies of various categories of kin. This paper uses formal models to calculate the correlations that will exist between certain categories of kin even if mothers and daughters have independent fertility. Mechanisms by which fertility might be transmitted from mothers to their daughters are considered and the implications for kin correlations are evaluated.

David, Mroz, Sanderson, Wachter, and Weir (1988) proposed procedure for inferring the existence of fertility control by using distributions.' There is need for such procedures to complement those that use age-specific marital fertility rates (Coale & Trussell 1974). The reasoning and techniques developed by David et al. constitute major and promising step in this direction. Their method consists of two components. The first is the calculation of distribution, specific for given population, which serves as reference against which the observed distribution is compared. This discussion will not deal with the calculation of the natural distribution, despite its critical importance. My focus is on the second component, which compares the observed and natural distributions to generate an interval estimate of the proportion at each observed who have practiced control. David et al. produce a pair of easily implemented formulas for upper- and lowerbound estimates of the expected proportion of the population ever controlling and the distribution of controllers by parity (p. 163). They imply that their upper- and lower-bound estimates, derived from the polar cases of pure stopping and pure spacing (p. 179), embrace the full range of the proportion of controllers under any possible pattern of control. The authors stated that they yield upper and lower bounds . . . even when the actual behavior of effective controllers in the target [observed] cohort does not conform to either of the polarcase specifications (p. 172). The main example in David et al.'s article will be reanalyzed in this comment. It consists of urban women from the 1911 Census of Ireland who married at the ages of 2024 and at the time of the census had been married for 5-9 years. David et al. calculated that maximum of 30.68% and minimum of 20.57% of these women had effectively practiced some form of control. It is noteworthy that the minimum is so high in cohort that was only married for 5-9 years and preceded modern contraception, and that the range between the lower and upper estimates is relatively small. Findings such as these have great significance for inferring the adoption of contraception in historical populations. I will show that in this example, and more generally, the range of controllers may be much wider than implied by David et al. Of particular importance, the minimum number may be much smaller. I shall suggest that their method is more sensitive to its assumptions than potential users might at first recognize. For notation, t( j) and n( j) are the observed and natural distributions, and T( j) and N( j) are the corresponding cumulative distributions, cumulated from the high to the low parities. Upper and lower estimates of the proportion of women at j who may be regarded as effective controllers are labelled cu( j) and CL( I); their sums across j are the overall limits Cu and CL. If one wished to estimate the maximum proportion of controllers without David et al.'s methodology, the simplest approach would be to assume that each controller had prevented the minimum number of births, that is, exactly one birth. Let C be the proportion of

Procedures are developed to allocate the change in mean fertility to the change in specific parities or groups of parities. One procedure uses the proportion at each parity and another uses parity progression ratios. Both are based on the delta method for approximating change in a function of several variables. Drawing on an analogy to survival in a life table, the relational logit model is applied to parity progression. This method allows several parity distributions to be synthesized and to have differences summarized with two parameters. The three procedures are applied to successive cohorts of white U.S. women who completed their childbearing between 1920 and 1980.

This study was aimed at deepening understanding of the Mexican fertility transition through a comparison of data from the 1977 Mexican Fertility Survey and the 1982 National Demographic Survey. In the 5 years between surveys the total fertility rate for women ages 20-44 years dropped by 24% from 5.4 to 4.1. These reductions were relatively uniform across age groups. The most striking declines were recorded among less educated and rural women--groups that had the highest fertility in 1977 and are considered difficult to reach with family planning programs. Women with 0-3 years of education experienced a 31% decline in fertility between 1977 and 1982 compared to a 21% drop among those with 4-6 years and no change among women with 7 or more years of schooling. Moreover fertility declined by 27% among rural women compared to 20% among urban women. Overall contraceptive prevalence increased by 52% between the 2 surveys from 27% to 41%; the greatest increase (59%) was observed among women in both the 30-34 and 40-44 year age groups. Female sterilization showed the greatest growth in use (352%) in the intersurvey interval and is now the most important contraceptive method among women 30 years of age and over. Oral contraceptive use which increased by 29% remains the method of choice for women under 30 years of age. The greatest increases in contraceptive use ocurred among women at the lowest educational levels and rural women. Striking among these results are the rapidity of the fertility decline and the convergence of the reproductive behavior of different socioeconomic strata of the population. Because the decline in fertility began so soon after the official family planning program was started in 1974 it can be assumed that the demand for children in Mexico had already fallen. The impact of the program was most likely achieved through changing norms about women and the family and through reducing the costs of fertility regulation rather than any perceived increase in the costs of children.

Existing methods for estimating population parameters in settings of data deficiency do not provide techniques for analysis of commonly available longitudinal data. In setting where complete population data is unavailable, longitudinal data recorded for only a subset of the total population are often available (e.g., event registers, genealogies). In this article we present and evaluate models which derive population parameters for the population subgroup underlying such longitudinal data. Using the distribution of individual times until first recorded event within a measurement interval, population parameters are estimated which provide basic denominator data for analyzing event occurrence. The models which we derive are especially suited to records which may include migration and population growth trends. The use of the models is demonstrated and evaluated through an application to genealogical records for a nineteenth-century population. Possible extensions of these models and their major limitations are also discussed.

Associated with every real birth cohort of women is a set of probabilities {f k} of eventually having k daughters. With a variant of stable population theory, these probabilities are used to generate the entire probability distributions, as well as all moments, for all categories of kin who are female and female-related. With additional assumptions, a full two-sex model for all kin also is given. The two-sex model is applied to a cohort of U.S. women born in the mid-twentieth century, suggesting plausible frequencies of kin in a stationary population.

This paper describes policy-related research conducted in 1979 in response to a specific request from the Seattle (Washington) School District (SSD). The background of this request is as follows: after unsuccessful attempts to eliminate racial imbalance in Seattle's schools by purely voluntary methods, the SSD began a mandatory busing program based on pairs and triads of schools, and modified feeder patterns to 20 schools, in September 1978. In anticipation of this action an organization (Citizens for Voluntary Integration Committee; CIVIC) that advocated exclusively voluntary desegregation was able to place on the ballot for November 1978, an initiative (Initiative 350) to limit the legal capacity of any school district in the state of Washington to reassign students for desegregation. This initiative passed by a twoto-one margin (a simple majority was required). The SSD and other affected school districts brought suit challenging the constitutionality of the initiative, obtained an injunction, and continued with their desegregation programs. The Attorney General of Washington defended the legality of the initiative. The SSD alleged that the initiative violated the Fourteenth Amendment's equal protection clause because (1) the initiative treated racial student assignments differently from all other student assignments; (2) the discriminatory intent to stop racial desegregation was among the factors leading to the initiative's adoption; and (3) the initiative prohibited necessary desegregation tools in all school districts, even districts with a constitutional obligation to desegregate. The U.S. District Court declared the initiative unconstitutional on all these grounds and issued a permanent injunction against its enforcement. An appeal currently is being heard by the U.S. Supreme Court, with a decision expected in early 1982. We shall not comment further on the legal issues and implications of this suit.

A method for estimating conception rates, using vital statistics data, is developed and applied to data on five-year age groups of California women for 1971. The approach is deterministic and allocates total exposure time to the known pregnancy outcomes of live birth, spontaneous abortion, and induced abortion. The population at risk is defined to exclude women who are known to be sterile or sexually inactive. Early fetal loss, premarital conception, and contraceptive use are taken into account. Estimates are made of the fecundability which would obtain ifno contraception were used.