Exploring the possible reasons for heterogeneity between studies is an important aspect of conducting a meta-analysis. This paper compares a number of methods which can be used to investigate whether a particular covariate, with a value defined for each study in the meta-analysis, explains any heterogeneity. The main example is from a meta-analysis of randomized trials of serum cholesterol reduction, in which the log-odds ratio for coronary events is related to the average extent of cholesterol reduction achieved in each trial. Different forms of weighted normal errors regression and random effects logistic regression are compared. These analyses quantify the extent to which heterogeneity is explained, as well as the effect of cholesterol reduction on the risk of coronary events. In a second example, the relationship between treatment effect estimates and their precision is examined, in order to assess the evidence for publication bias. We conclude that methods which allow for an additive component of residual heterogeneity should be used. In weighted regression, a restricted maximum likelihood estimator is appropriate, although a number of other estimators are also available. Methods which use the original form of the data explicitly, for example the binomial model for observed proportions rather than assuming normality of the log-odds ratios, are now computationally feasible. Although such methods are preferable in principle, they often give similar results in practice.

/pdf/explaining-heterogeneity-in-meta-analysis-a-comparison-of-52mdg0qtyq.pdf

Explaining heterogeneity in meta-analysis: a comparison of methods.

Poverty in the United Kingdom

This volume argues that international human rights law has made a positive contribution to the realization of human rights in much of the world. Although governments sometimes ratify human rights treaties, gambling that they will experience little pressure to comply with them, this is not typically the case. Focusing on rights stakeholders rather than the United Nations or state pressure, Beth Simmons demonstrates through a combination of statistical analyses and case studies that the ratification of treaties leads to better rights practices on average. Simmons argues that international human rights law should get more practical and rhetorical support from the international community as a supplement to broader efforts to address conflict, development, and democratization.

/pdf/mobilizing-for-human-rights-international-law-in-domestic-1kn9xhxi9f.pdf

Mobilizing for Human Rights: International Law in Domestic Politics

Skilled attendance at childbirth is crucial for decreasing maternal and neonatal mortality, yet many women in low- and middle-income countries deliver outside of health facilities, without skilled help. The main conceptual framework in this field implicitly looks at home births with complications. We expand this to include "preventive" facility delivery for uncomplicated childbirth, and review the kinds of determinants studied in the literature, their hypothesized mechanisms of action and the typical findings, as well as methodological difficulties encountered. We searched PubMed and Ovid databases for reviews and ascertained relevant articles from these and other sources. Twenty determinants identified were grouped under four themes: (1) sociocultural factors, (2) perceived benefit/need of skilled attendance, (3) economic accessibility and (4) physical accessibility. There is ample evidence that higher maternal age, education and household wealth and lower parity increase use, as does urban residence. Facility use in the previous delivery and antenatal care use are also highly predictive of health facility use for the index delivery, though this may be due to confounding by service availability and other factors. Obstetric complications also increase use but are rarely studied. Quality of care is judged to be essential in qualitative studies but is not easily measured in surveys, or without linking facility records with women. Distance to health facilities decreases use, but is also difficult to determine. Challenges in comparing results between studies include differences in methods, context-specificity and the substantial overlap between complex variables. Studies of the determinants of skilled attendance concentrate on sociocultural and economic accessibility variables and neglect variables of perceived benefit/need and physical accessibility. To draw valid conclusions, it is important to consider as many influential factors as possible in any analysis of delivery service use. The increasing availability of georeferenced data provides the opportunity to link health facility data with large-scale household data, enabling researchers to explore the influences of distance and service quality.

/pdf/still-too-far-to-walk-literature-review-of-the-determinants-2vt2nanmuk.pdf

Still too far to walk: literature review of the determinants of delivery service use.

HEALTH AND DEPRIVATION: Inequality and the north

In this paper random-effects logistic models are used to analyze the effects of the preceding birth interval on postneonatal mortality in Brazil, controlling for the correlation of survival outcomes between siblings. The results are compared to those obtained by using ordinary logistic regression. Family effects are found to be highly significant in the random-effects model, but the substantive conclusions of the ordinary logistic model are preserved. In particular, birth interval effects remain highly significant.

Birth interval and family effects on postneonatal mortality in Brazil

This article examines data from a study on garment-factory workers in Bangladesh to explore the implications of work for the early socialization of young women. For the first time, large numbers of young Bangladeshi women are being given an alternative to lives in which they move directly from childhood to adulthood through early marriage and childbearing. Employment creates a period of transition in contrast to the abrupt assumption of adult roles at very young ages that marriage and childbearing mandate. This longer transition creates a period of adolescence for young women working in the garment sector that is shown to have strong implications for the women's long-term reproductive health.

/pdf/transition-to-adulthood-of-female-garment-factory-workers-in-1cilwvq4e5.pdf

Transition to Adulthood of Female Garment-factory Workers in Bangladesh

The Australian People. An Encyclopedia of the Nation, its People and their Origins

This paper investigates the effect of previous cohabitation on marital stability among the 1958 British birth cohort. Prospective data from the National Child Development Study are used to investigate the way in which family background factors and early lifecourse experiences, including cohabitation, affect the risk of first marriage dissolution by age 33. Discrete time logistic regression hazards models are used to analyse the risk of separation in the first eight years of marriage. Many socio-economic and family background factors are found to act through more intermediate determinants, such as age at marriage and the timing of childbearing, to affect the risk of separation. Previous cohabitation with another partner and premarital cohabitation are both associated with higher rates of marital breakdown. The effect of premarital cohabitation is attenuated but remains significant once the characteristics of cohabitors are controlled, and cannot be explained by the longer time spent in a partnership.

Marital dissolution among the 1958 British birth cohort: The role of cohabitation

Analysis of Binary Data, 2nd edn. By D. R. Cox and E. J. Snell. ISBN 0 412 30620 4. Chapman and Hall, London, 1989. 236 pp. £20.00.

Ian Diamond

Papers

Birth interval and family effects on postneonatal mortality in Brazil

Transition to Adulthood of Female Garment-factory Workers in Bangladesh

The Australian People. An Encyclopedia of the Nation, its People and their Origins

Marital dissolution among the 1958 British birth cohort: The role of cohabitation

Analysis of Binary Data. 2nd Edn.