Association between Firearm Laws and Homicide in Urban Counties

Laws related to the sale, use, and carrying of firearms have been associated with differences in firearm homicide rates at the state level. Right-to-carry (RTC) and stand your ground (SYG) laws are associated with increases in firearm homicide; permit-to-purchase (PTP) laws and those prohibiting individuals convicted of violent misdemeanors (VM) have been associated with decreases in firearm homicide. Evidence for the effect of comprehensive background checks (CBC) not tied to PTP is inconclusive. Because firearm homicide tends to concentrate in urban areas, this study was designed to test the effects of firearm laws on homicide in large, urban U.S. counties. We conducted a longitudinal study using an interrupted time series design to evaluate the effect of firearm laws on homicide in large, urban U.S. counties from 1984 to 2015 (N?=?136). We used mixed effects Poisson regression models with random intercepts for counties and year fixed effects to account for national trends. Models also included county and state characteristics associated with violence. Homicide was stratified by firearm versus all other methods to test for specificity of the laws’ effects. PTP laws were associated with a 14% reduction in firearm homicide in large, urban counties (IRR?=?0.86, 95% CI 0.82–0.90). CBC-only, SYG, RTC, and VM laws were all associated with increases in firearm homicide. None of the laws were associated with differences in non-firearm homicide rates. These findings are consistent with prior research at the state level showing PTP laws are associated with decreased firearm homicide. Testing the effects of PTP laws specifically in large, urban counties strengthens available evidence by isolating the effects in the geographic locations in which firearm homicides concentrate.     

Bayesian rates for homicide mapping in Brazilian municipalities

This study analyzes homicide incidence per municipality (county) in Brazil in the year 2008. The authors estimate and compare homicide rates according to different methods, finding evidence that depending on the method employed, the results can differ significantly, especially for small municipalities. Bayesian spatial procedures were employed, allowing minimization of variation in the rate estimates. The methods consider a priori distributions and information on contiguity of municipalities. According to the findings, the impact of corrective procedures is not relevant for large municipalities, but such estimates present significant differences for small municipalities. Comparing the different estimates, the authors conclude that there may be distortions in the rates published in the literature. To overcome such potential distortions, it is necessary to take the main goal in each analysis into account. When the emphasis is on overall visualization of the homicide phenomenon, the best option is spatial corrections. However, to obtain more accurate local estimates, Bayesian methods are more appropriate.     

The Effect of State Gun Laws on the Supply of Guns and Gun Crimes: A Multilevel Modeling Analysis

Gun homicides alone cause 11 000 deaths each year. Numerous shootings in recent years, terrorist and otherwise, have sparked heated debates about gun control policy, particularly with regard to Second Amendment issues. Such discussions are more often than not driven by political and philosophical ideology, rather than empirical data. The self-imposed ban on the dissemination of gun-related statistics on the part of the Centers for Disease Control has created difficulty in obtaining data, particularly at the sub-state level. Furthermore, the majority of previous research has not examined the effect on crimes that are gun specific, that is, homicides or robberies committed using guns versus all homicides or robberies. The objective of this research is to examine the multiple relationships between gun ownership and gun crimes at the sub-state level and to then examine the effect of state-level gun laws on the sub-state relationships via multilevel modeling. This research involves two separate units of analysis, counties and states. Initial multivariate regression was conducted on a subset of counties in the United States examining the relationship between the percentage of all households with at least one gun, and both the gun homicide rate as well as the gun robbery rate. Data on gun ownership were obtained from the CDC's BRFSS SMART County-level data files from the mid-2000s. These data are sampled and weighted to be representative of the entire county. A total of 218 weighted, aggregated counties were used from two separate non-overlapping years combined. Gun crime data were obtained from the FBI Uniform Crime Reports, which provides total universe data on crimes, including the number committed with a gun. State gun law rankings were obtained from the Law Center to Prevent Gun Violence, which quantifies state gun laws into a numeric index. Standard county-level control variables such as race, income, poverty, region of the country, level of education, and population density were employed. Multilevel models were then run overlaying state gun laws with county-level data. The population studied included adults age 18 and older who responded to the Centers for Disease Control's BRFSS survey questions, as well as all gun crimes known to the police. In traditional multivariate analysis, a higher percentage of households with guns was associated with lower gun robbery rates as well as lower gun homicide rates. Multilevel modeling using state gun laws indicated that more restrictive gun laws were also associated with lower gun robbery and homicide rates. The percentage of households with guns remained negative and significant in the multilevel models. Moreover, restrictive state gun laws were associated with lower gun ownership rates at the county level. Both legal gun ownership and more restrictive state gun laws are associated with lower rates of violent gun crimes (robbery and homicide). The results unexpectedly provide support for both pro- and anti-gun positions. This suggests that causal patterns are more complex than either position might indicate, and public health policies on gun supply and restriction would be better informed by empirical analysis.     

Adjusting for confounding by neighborhood using generalized linear mixed models and complex survey data

When investigating health disparities, it can be of interest to explore whether adjustment for socioeconomic factors at the neighborhood level can account for, or even reverse, an unadjusted difference. Recently, we proposed new methods to adjust the effect of an individual‐level covariate for confounding by unmeasured neighborhood‐level covariates using complex survey data and a generalization of conditional likelihood methods. Generalized linear mixed models (GLMMs) are a popular alternative to conditional likelihood methods in many circumstances. Therefore, in the present article, we propose and investigate a new adaptation of GLMMs for complex survey data that achieves the same goal of adjusting for confounding by unmeasured neighborhood‐level covariates. With the new GLMM approach, one must correctly model the expectation of the unmeasured neighborhood‐level effect as a function of the individual‐level covariates. We demonstrate using simulations that even if that model is correct, census data on the individual‐level covariates are sometimes required for consistent estimation of the effect of the individual‐level covariate. We apply the new methods to investigate disparities in recency of dental cleaning, treated as an ordinal outcome, using data from the 2008 Florida Behavioral Risk Factor Surveillance System (BRFSS) survey. We operationalize neighborhood as zip code and merge the BRFSS data with census data on ZIP Code Tabulated Areas to incorporate census data on the individual‐level covariates. We compare the new results to our previous analysis, which used conditional likelihood methods. We find that the results are qualitatively similar. Copyright © 2012 John Wiley & Sons, Ltd.     

The relationship between state abortion-restrictions and homicide deaths among children under 5 years of age: a longitudinal study

The purpose of this study is to explore whether, in the U.S., there are associations between state-level variations in mortality among young children and state abortion restriction policies - such as parental-consent requirements, parental-notification requirements, mandatory delay laws, and restrictions on Medicaid funding for abortion. To investigate this, we used NCHS Multiple Cause of Deaths public-use data files for the period 1983-2002, and compiled data on children ages 0-4 identified as having died as a result of assault/homicide in each state and year. Medicaid funding of abortion, mandatory delay laws, and parental involvement laws for minors seeking abortions were included as the main predictor variables of interest. Multivariate count data models using pooled state-year-age cohort data, with state and time fixed effects and other state-level controls, were estimated. Results indicated that, between 1983 and 2002, the average increase in the number of homicide deaths for children under 5 years of age was 5.70 per state among states that implemented stricter abortion policies over that time, and 2.00 per state for states that did not. In the count data models, parental-consent laws were associated with a 13 percent increase in child homicide deaths; parental-notification laws were associated with an 8 percent increase in child homicide deaths though the results were less robust to alternate model specifications; mandatory delay requirements were associated with a 13 percent increase in child homicide deaths. While these data do not allow us to discern precise pathways via which state abortion-restrictions can lead to more child homicide deaths, we speculate that state restrictions on abortion may result in a disproportionate increase in children born into relatively high-risk environments. Additional research is called for to explore the association of state abortion-restrictions with other measures of infant/child health and well-being.     

A general framework for estimating volume-outcome associations from longitudinal data

Recently, there has been much interest in using volume-outcome data to establish causal associations between measures of surgical experience or quality and patient outcomes following a surgical procedure, such as coronary artery bypass graft, total hip replacement, and radical prostatectomy. However, there does not appear to be a standard approach to a volume-outcome analysis with respect to specifying a volume measure and selecting an estimation method. We establish the recurrent marked point process as a general framework from which to approach a longitudinal volume-outcome analysis and examine the statistical issues associated with using longitudinal data analysis methods to model aggregate volume-outcome data. We review assumptions to ensure that linear or generalized linear mixed models and generalized estimating equations provide valid estimates of the volume-outcome association. In addition, we provide theoretical and empirical evidence that bias may be introduced when an aggregate volume measure is used to address a scientific question regarding the effect of cumulative experience. We conclude with the recommendation that analysts carefully specify a volume measure that most accurately reflects their scientific question of interest and select an estimation method that is appropriate for their scientific context.     

Social capital and violence in the United States, 1974-1993

Social capital is a characteristic of communities. Cross-sectional studies have shown that social capital is inversely associated with homicide and violent crime. We hypothesized that variations in social capital in US states over time can predict variations in regional homicide mortality both across and within time periods. We analyzed serial cross-sectional data for measures of social capital and age-adjusted homicide rates between 1974 and 1993. We used perception of social trust and per capita membership in voluntary associations, obtained from responses to the General Social Surveys, as the principal measures of social capital. We controlled for potential confounding by mean levels of income, urbanization, and region. Measures of perceived trust were strongly inversely correlated with homicide rates in an aggregate cross-sectional analysis (r=-0.51, p<0.001) and also within each time period. Social capital was an independent predictor of rates of violence when controlling for income, region, and urbanization (p<0.001). Homicide rates also predicted levels of social capital in adjusted models (p<0.001). To investigate directionality of this relationship we developed Markov transition matrices that described the change in the states' levels of social capital and homicide across time intervals. Analysis of the transitional probabilities confirmed that a simple unidirectional association between social capital and violence was not sufficient to describe this association. There is likely an impact of violence on levels of perceived trust in communities that complements the hypothesized effect of social capital on homicide. We conclude that the relationship between social capital and violence over time is non-linear and dynamic. More complex analytic models describing the relationship between violence and ecological social determinants need to be considered.     

Analysis of cluster randomized trials with repeated cross-sectional binary measurements

Analytical techniques appropriate for cluster randomized trials that utilize a repeated cross-sectional design have not been extensively evaluated. This paper compares methods that can be used to evaluate the impact of an intervention on dichotomous outcomes. The methods are applied to data from a study on the implementation of Cochrane review evidence, in which 25 hospital obstetric units were randomized. Assessments were made for 30 pregnancies in each obstetric unit at baseline, and for 30 separate pregnancies at follow-up. The principal issues addressed are how best to take clustering into account and to allow for baseline imbalance. We compare cluster level analyses, the clustered Woolf method, marginal models based on generalized estimating equations, multilevel models, and methods based on random effects meta-analysis. Analyses which ignored the baseline assessments showed no effect of the intervention. There was substantial baseline imbalance, however, so that analyses taking into account the baseline were necessary. Yet, while analyses of change from baseline showed evidence of an effect of intervention, adjusting for baseline using analysis of covariance did not. Analysis of covariance required the use of cluster level rather than individual level responses, since different pregnancies were evaluated at baseline and follow-up. Also, when analysing change from baseline, we show it is important to allow for variation in the effect of secular trend between clusters in a multilevel model, or use robust variance estimates in a marginal model, for otherwise confidence intervals for the effect of intervention will be too narrow. We conclude however that analyses of change from baseline can be misleading since they are affected too much by baseline results, and that analysis of covariance approaches are preferable. To prevent difficulties in interpreting the results from repeated cross-sectional cluster trial designs, one should either attempt to achieve baseline balance by careful stratification of the clusters prior to randomization, or have sufficiently large samples for precise estimation of the effects of imbalance.     

Assessing health care interventions via an interrupted time series model: Study power and design considerations

The delivery and assessment of quality health care is complex with many interacting and interdependent components. In terms of research design and statistical analysis, this complexity and interdependency makes it difficult to assess the true impact of interventions designed to improve patient health care outcomes. Interrupted time series (ITS) is a quasi-experimental design developed for inferring the effectiveness of a health policy intervention while accounting for temporal dependence within a single system or unit. Current standardized ITS methods do not simultaneously analyze data for several units nor are there methods to test for the existence of a change point and to assess statistical power for study planning purposes in this context. To address this limitation, we propose the “Robust Multiple ITS” (R-MITS) model, appropriate for multiunit ITS data, that allows for inference regarding the estimation of a global change point across units in the presence of a potentially lagged (or anticipatory) treatment effect. Under the R-MITS model, one can formally test for the existence of a change point and estimate the time delay between the formal intervention implementation and the over-all-unit intervention effect. We conducted empirical simulation studies to assess the type one error rate of the testing procedure, power for detecting specified change-point alternatives, and accuracy of the proposed estimating methodology. R-MITS is illustrated by analyzing patient satisfaction data from a hospital that implemented and evaluated a new care delivery model in multiple units.     

Joint modeling of binary response and survival for clustered data in clinical trials

In clinical trials, it is often desirable to evaluate the effect of a prognostic factor such as a marker response on a survival outcome. However, the marker response and survival outcome are usually associated with some potentially unobservable factors. In this case, the conventional statistical methods that model these two outcomes separately may not be appropriate. In this paper, we propose a joint model for marker response and survival outcomes for clustered data, providing efficient statistical inference by considering these two outcomes simultaneously. We focus on a special type of marker response: a binary outcome, which is investigated together with survival data using a cluster-specific multivariate random effect variable. A multivariate penalized likelihood method is developed to make statistical inference for the joint model. However, the standard errors obtained from the penalized likelihood method are usually underestimated. This issue is addressed using a jackknife resampling method to obtain a consistent estimate of standard errors. We conduct extensive simulation studies to assess the finite sample performance of the proposed joint model and inference methods in different scenarios. The simulation studies show that the proposed joint model has excellent finite sample properties compared to the separate models when there exists an underlying association between the marker response and survival data. Finally, we apply the proposed method to a symptom control study conducted by Canadian Cancer Trials Group to explore the prognostic effect of covariates on pain control and overall survival.     

Commentary: the role of toxicology in prevention and precaution

Advocates of the Precautionary Principle have recently called for a "new science" to support the goals of precaution-based environmental and occupational health policy. While much attention has been given to epidemiology, the evidentiary science most relevant to precaution, or prevention, is toxicology. Opportunities for enhancing the role of toxicology in public policy must consider current biases in the field. Thus, rather than a "new science", advocates for change should focus upon ensuring that current scientific methods are appropriate and that interpretations of scientific data are accurate.     

Disclosure control using partially synthetic data for large‐scale health surveys,with applications to CanCORS

Statistical agencies have begun to partially synthesize public‐use data for major surveys to protect the confidentiality of respondents’ identities and sensitive attributes by replacing high disclosure risk and sensitive variables with multiple imputations. To date, there are few applications of synthetic data techniques to large‐scale healthcare survey data. Here, we describe partial synthesis of survey data collected by the Cancer Care Outcomes Research and Surveillance (CanCORS) project, a comprehensive observational study of the experiences, treatments, and outcomes of patients with lung or colorectal cancer in the USA. We review inferential methods for partially synthetic data and discuss selection of high disclosure risk variables for synthesis, specification of imputation models, and identification disclosure risk assessment. We evaluate data utility by replicating published analyses and comparing results using original and synthetic data and discuss practical issues in preserving inferential conclusions. We found that important subgroup relationships must be included in the synthetic data imputation model, to preserve the data utility of the observed data for a given analysis procedure. We conclude that synthetic CanCORS data are suited best for preliminary data analyses purposes. These methods address the requirement to share data in clinical research without compromising confidentiality. Copyright © 2013 John Wiley & Sons, Ltd.     

Taxing sin and saving lives: Can alcohol taxation reduce female homicides?

With costs exceeding $5.8 billion per year, violence against women has significant ramifications for victims, their families, the health care systems that treat them, and the employers who depend on their labor. Prior research has found that alcohol abuse contributes to violence against both men and women, and that stringent alcohol control policies can reduce alcohol consumption and in turn some forms of violence. In this paper, we estimate the direct relationship between an important alcohol control measure, excise taxes, and the most extreme form of violence, homicide. We use female homicide rates as our measure of severe violence, as this measure is consistently and accurately reported across multiple years. Our results provide evidence that increased alcohol taxes reduce alcohol consumption and that reductions in alcohol consumption can reduce femicide. Unfortunately, a direct test of the relationship does not have the power to determine whether alcohol taxes effectively reduce female homicide rates. We conclude that while alcohol taxes have been shown to effectively reduce other forms of violence against women, policy makers may need alternative policy levers to reduce the most severe form of violence against women.     

Evaluating the Effect of State Regulation of Federally Licensed Firearm Dealers on Firearm Homicide

Effective federal regulation of firearm dealers has proven difficult. Consequently, many states choose to implement their own regulations. We examined the impact of state-required licensing, record keeping of sales, allowable inspections, and mandatory theft reporting on firearm homicide from 1995 to 2010. We found that lower homicide rates were associated with states that required licensing and inspections. We concluded that firearm dealer regulations might be an effective harm reduction strategy for firearm homicide.Current federal regulations and enforcement practices limit the federal government’s ability to effectively deter illegal firearm sales by federally licensed firearm dealers.1–4 Several states have enacted their own firearm laws or require an additional layer of oversight, but evidence in support of these laws as injury reduction strategies vary.5–7 Firearm dealer regulations aimed at decreasing trafficking have been successful, yet little data exist regarding the effect of these regulations on firearm homicides.8 In this study, we examined state licensing and other lawful sales promoting dealer regulations, and hypothesized that they decrease firearm homicide.     

Effects of Managed Care Contracting on Physician Labor Supply

We examine the effect of managed care contracting on physician labor supply for office-based medical practices. We extend the standard labor supply model to incorporate choices regarding the patient base. Empirical tests use data from the 1985 and 1988 national HCFA Physician Practice Costs and Income Surveys and InterStudy Managed Care Surveys. We use physician-level information on participation in managed care contracting to estimate changes in work hours. Managed care contracting is generally associated with lower physician work hours. However, accounting for motivations to participate in contracts and the extent of contracting, the effect on hours is reduced in magnitude and significance. We conclude that relying on broad aggregate measures for policy analysis will likely be misleading as underlying motivations and contracting incentives change over time.     

High School Physical Education Requirements and Youth Body Weight: New Evidence from the YRBS

Previous research has found that high school physical education (PE) requirements are largely ineffective at reducing youth body weight. However, these studies were forced to rely on cross‐state variation in PE requirements to identify their impacts, raising concerns that estimated policy effects may be confounded by state‐level unobservables. Using data from the State and National Youth Risk Behavior Surveys and exploiting recent changes in state high school PE laws, we re‐examine the effect of PE requirements on body weight. Our estimates show that a one‐semester increase in PE requirements is associated with a 10 to 13% increase in minutes per week spent physically active in PE classes, but with no change in net vigorous exercise and little change in youth body weight. We conclude that substitution of in‐school for outside‐of‐school physical activity and small resultant net energy expenditures can explain the absence of body weight effects. Copyright © 2016 John Wiley & Sons, Ltd.     

Forcing the task.

Although not generally included in public health curricula, public health policy formulation is essential to local agency public health operations. When public health measures are mandated by referendum of the electorate, such as the HIV exposure notification laws passed in California in 1988, health agencies must act prudently and quickly to formulate effective policies. We found it prudent to convene a task force that brought together the critical, albeit diverse, legal/judicial/health players. This action not only served to bring about consensus, but provided us with an opportunity to educate participants about the policy, and the implementation of appropriate communicable disease measures. Through consensus and education, we were able to implement the disclosure laws in a timely manner, create an atmosphere of trust, and enable a skilled and confident workforce.     

我国健康投资经济效益在国际间的位置及其分析

目的 通过分析世界主要国家（地区）健康投资效益的大小、来评价我国卫生工作水平和国家卫生政策的质量,为决策机构提供参考依据。方法 依据世界银行公布的数据和国内资料,结合其经济规模,从考察投入、产出、效益入手,比较它们在国际中的位置,以评判各国国民健康水平的差异。结果 经实证调研,认识到我国几十年来的健康投资在国际间具有较高效率,位于世界前列。结论 从建国至90年代,我国的卫生政策是适宜的,有效的,符合客观规律。因此,未来卫生机制改革中,应把握优势,使有限资源发挥出较高健康效益。     

SNP selection in genome-wide and candidate gene studies via penalized logistic regression

Penalized regression methods offer an attractive alternative to single marker testing in genetic association analysis. Penalized regression methods shrink down to zero the coefficient of markers that have little apparent effect on the trait of interest, resulting in a parsimonious subset of what we hope are true pertinent predictors. Here we explore the performance of penalization in selecting SNPs as predictors in genetic association studies. The strength of the penalty can be chosen either to select a good predictive model (via methods such as computationally expensive cross validation), through maximum likelihood-based model selection criterion (such as the BIC), or to select a model that controls for type I error, as done here. We have investigated the performance of several penalized logistic regression approaches, simulating data under a variety of disease locus effect size and linkage disequilibrium patterns. We compared several penalties, including the elastic net, ridge, Lasso, MCP and the normal-exponential-γ shrinkage prior implemented in the hyperlasso software, to standard single locus analysis and simple forward stepwise regression. We examined how markers enter the model as penalties and P-value thresholds are varied, and report the sensitivity and specificity of each of the methods. Results show that penalized methods outperform single marker analysis, with the main difference being that penalized methods allow the simultaneous inclusion of a number of markers, and generally do not allow correlated variables to enter the model, producing a sparse model in which most of the identified explanatory markers are accounted for.