Surrogacy assessment using principal stratification and a Gaussian copula model.

In clinical trials, a surrogate outcome ( S) can be measured before the outcome of interest ( T) and may provide early information regarding the treatment ( Z) effect on T. Many methods of surrogacy validation rely on models for the conditional distribution of T given Z and S. However, S is a post-randomization variable, and unobserved, simultaneous predictors of S and T may exist, resulting in a non-causal interpretation. Frangakis and Rubin developed the concept of principal surrogacy, stratifying on the joint distribution of the surrogate marker under treatment and control to assess the association between the causal effects of treatment on the marker and the causal effects of treatment on the clinical outcome. Working within the principal surrogacy framework, we address the scenario of an ordinal categorical variable as a surrogate for a censored failure time true endpoint. A Gaussian copula model is used to model the joint distribution of the potential outcomes of T, given the potential outcomes of S. Because the proposed model cannot be fully identified from the data, we use a Bayesian estimation approach with prior distributions consistent with reasonable assumptions in the surrogacy assessment setting. The method is applied to data from a colorectal cancer clinical trial, previously analyzed by Burzykowski et al.

Bleeding patterns during the menopausal transition in the multi-ethnic Study of Women's Health Across the Nation (SWAN): a prospective cohort study.

OBJECTIVE: Previous studies describing menses duration and heaviness of flow during the menopausal transition (MT) have been short in duration and limited to white women. We estimated the frequency of and risk factors for prolonged bleeding, spotting and heavy bleeding during the MT in an ethnically diverse population. DESIGN: Prospective community-based cohort study. SETTING USA: southeastern Michigan, northern California and Los Angeles, California. POPULATION: A total of 1320 midlife women who participated in the Study of Women's Health Across the Nation (SWAN) Menstrual Calendar Substudy. Participants included African-American, white, Chinese, and Japanese women. METHODS: Women completed daily menstrual calendars from 1996 to 2006, and provided information on hormone therapy, smoking and physical activity. Annual measures included height and weight. Kaplan-Meier survival analysis and multivariable regression were used to analyse the data. MAIN OUTCOME MEASURES: Menses of 10+ days, spotting of 6+ days, heavy bleeding of 3+ days. RESULTS: At least three occurrences of menses 10+ days was reported by 77.7% (95% confidence interval [95% CI] 56.7-93.2), of 6+ days of spotting by 66.8% (95% CI 55.2-78.0) and of 3+ days of heavy bleeding by 34.5% (95% CI 30.2-39.2) of women. Menses of 10+ days, 6+ days of spotting, and 3+ days of heavy bleeding were associated with MT stage, uterine fibroids, hormone use and ethnicity. Body mass index was associated with 3+ days of heavy bleeding. CONCLUSIONS: These data provide clinicians and women with important information about the expected frequency of prolonged and heavy bleeding and spotting during the menopausal transition that may facilitate clinical decision making.

Non-fatal and fatal crash injury risk for children in minivans compared with children in sport utility vehicles.

OBJECTIVE: To compare the fatal and non-fatal crash injury risk for children in minivans compared with midsize and large sport utility vehicles (SUVs). DESIGN: Three large population-based sources of US crash data were used--a nationwide cohort of sampled police-reported crashes (NASS-CDS) along with a census of fatal crashes (FARS), plus a large child crash surveillance system, Partners for Child Passenger Safety (PCPS)--collected in 16 states via insurance claim records and validated telephone survey. Each included: 2000-2006 data, occupants aged 0-15 years, traveling in minivan or (midsize/large) SUV, model year 1998-2007. Outcome of interest was parent/driver report of non-fatal injury (Abbreviated Injury Scale scores of 2 or higher) in PCPS and fatal injury in NASS-CDS/FARS. RESULTS: Compared with children riding in SUVs, those in minivans experienced a similar crude reduction in the relative risk of non-fatal injury (PCPS: unadjusted odds ratio (OR) = 0.55) and fatality (NASS-CDS/FARS cohort: unadjusted OR = 0.58). In PCPS, this reduction in injury risk changed little after adjustment for child, driver, and vehicle factors (adjusted OR = 0.56, 95% CI 0.38 to 0.82). Lower fatality risk in the NASS-CDS/FARS cohort was partially explained by the same factors (adjusted OR = 0.76, 95% CI 0.51 to 1.13). CONCLUSIONS: There may be important safety differences in vehicles during a crash that lead to fewer non-fatal injuries to child occupants of minivans compared with SUVs.

Trip characteristics of vehicle crashes involving child passengers.

OBJECTIVES: To describe the trip characteristics of vehicle crashes involving children, and to examine the effect of situational factors on front row seating or inappropriate restraint for young children. METHODS: A cross sectional study was conducted on children <16 years in crashes of insured vehicles in 15 US states, with data collected using insurance claims records and a telephone interview. A descriptive analysis of the characteristics of vehicle crashes involving children was performed. Multivariate Poisson regression was used to identify situational factors associated with inappropriate restraint or front row seating. RESULTS: These data suggest that children were traveling in vehicles involved in crashes that occurred under usual driving circumstances-that is, closer to home (60%), on a local road (56%), during normal daytime hours (71%), within areas with relatively lower posted speed limits (76%). Compared with children involved in morning crashes, those in daytime crashes (RR = 1.65, 95% CI 1.13 to 2.49) or in night-time crashes (RR = 1.63, 95% CI 1.09 to 2.67) were more likely to be sitting in the front seat. Children involved in night-time crashes were more likely to be inappropriately restrained (RR = 1.12, 95% CI 1.01 to 1.22) than those in daytime crashes. Children riding with two or more additional passengers were more likely to be inappropriately restrained (RR = 1.12, 95% CI 1.02 to 1.27) than those with no other passengers. CONCLUSIONS: Educational initiatives should aim to increase the perception that parents have about the potential crash risk of everyday trips. Some situational characteristics of trips were associated with inappropriate restraint and front row seating behaviors for young children.

Teen drivers and the risk of injury to child passengers in motor vehicle crashes.

OBJECTIVES: The first aim was to examine the relationship between driver's age (novice teens, older teens, and adults) and child passenger's restraint status, front row seating, and injury risk. The second aim was to explore whether there was an excess injury risk to child passengers in teen crashes compared to those in adult crashes by examining the contributing factors. METHODS: A cross sectional study involving telephone interviews with insured drivers in a probability sample of 12 163 crashes involving 19 111 children was conducted. Sequential logistic regressions were employed. RESULTS: Among child passengers aged 4-8, appropriate restraint was <1% for novice teens, 4.5% for older teens, and 23.6% for adults. Front row seating for children <13 years was more common in the novice teen group (26.8%) than in the other two groups. Compared with children riding with adults, those with both teen groups experienced excess injury risk. After adjusting for crash severity, there was a 43% reduction in the odds ratio (OR) for novice teens (OR 1.58, 95% confidence interval (CI) 1.14 to 2.19) and a 24% reduction for older teens (OR 2.15, 95% CI 1.42 to 3.26). After adjusting for vehicle type, child's restraint status and front row seating, there was a further 19% reduction in the OR for novice teens (OR 1.37, 95% CI 1.00 to 1.88) and a further 13% reduction for older teens (OR 1.74, 95% CI 1.14 to 2.66). CONCLUSION: These findings suggest ways in which graduated driver licensing laws may be further enhanced to better protect child passengers from the excess injury risk associated with teen crashes.

Environmental justice: frequency and severity of US chemical industry accidents and the socioeconomic status of surrounding communities.

STUDY OBJECTIVES: The Clean Air Act Amendments of 1990 requires that chemical facilities in the US with specified quantities of certain toxic or flammable chemicals file a five year history of accidents. This study considers the relation between the reported accidents and surrounding community characteristics. DESIGN: This study is a retrospective analysis of the association between the demographics of counties in which facilities are located and the risk of accidental chemical release and resulting injuries at those facilities. The "location risk" (the risk that a facility having large volumes of hazardous chemicals is located in a community) and "operations risk" (the risk of an accident itself) are investigated. SETTING: 1994-2000 accident history data from 15 083 US industrial facilities using one or more of 140 flammable or toxic substances above a threshold level. Demographic makeup of 2333 counties surrounding these facilities was determined from the 1990 US census. MAIN RESULTS: Larger and more chemical intensive facilities tend to be located in counties with larger African-American populations and in counties with both higher median incomes and high levels of income inequality. Even after adjusting for location risk there is greater risk of accidents for facilities in heavily African-American counties (OR of accident = 1.9, 95% CI = 1.5 to 2.4). CONCLUSIONS: Further research and policy interventions are required to reduce the probability of locating facilities in an inequitable fashion, as well as health surveillance, and regulatory monitoring and enforcement activities to ensure that hazardous facilities in minority communities prepare and prevent accidental chemical releases to the same standards as elsewhere.

Graduated driver licensing in Michigan: early impact on motor vehicle crashes among 16-year-old drivers.

CONTEXT: Graduated driver licensing (GDL) programs are being adopted in many states to address the high rate of motor vehicle fatalities among teens by requiring teenaged drivers to gain experience and maturity under conditions of relatively low crash risk before gaining full driving privileges. OBJECTIVE: To evaluate the early impact of Michigan's GDL program on traffic crashes among 16-year-old drivers. DESIGN, SETTING, AND SUBJECTS: Analysis of Michigan motor vehicle crash data from 1996 (before GDL program implementation) vs 1998 and 1999 (after GDL program implementation) for 16-year-olds, adjusting for trends among persons 25 years or older. INTERVENTION: Michigan's GDL program, instituted April 1, 1997, for teens younger than 18 years entering the driver license system, includes 3 licensure levels, each with driving restrictions and requirements to progress to the next level. Requirements include extended, supervised practice in the learning level, 2-phase driver education, and night driving restrictions in the intermediate level. MAIN OUTCOME MEASURES: Rates in 1996 vs 1998 and 1999 for all police-reported crashes; for fatal injury, nonfatal injury, and fatal/nonfatal injury combined crashes; for day, evening, and night crashes; for single-vehicle and multivehicle crashes; and for alcohol-related crashes. RESULTS: Overall, the rate of 16-year-old drivers (per 1000 population) involved in crashes declined from 154 in 1996 to 111 in 1999 (relative risk [RR], 0.72; 95% confidence interval [CI], 0.71-0.73). After adjusting for populationwide trends, the overall crash risk for 16-year-olds was significantly reduced in 1999 from 1996 by 25% (adjusted RR, 0.75; 95% CI, 0.74-0.77). There were also significant reductions for nonfatal injury and combined fatal and nonfatal crashes; for day, evening, and night crashes; and for single-vehicle and multivehicle crashes. Fatal crashes declined from 1996 to 1999, but not significantly (RR, 0.74; 95% CI, 0.49-1.14), and alcohol-related crashes continued at a low rate (RR, 1.01; 95% CI, 0.80-1.29). CONCLUSIONS: Analysis of the first 2 full calendar years following Michigan's GDL program implementation indicates substantial crash reductions among 16-year-olds. Future research is necessary to determine if these reductions are maintained and if other jurisdictions achieve similar results.

Long-term follow-up of a high school alcohol misuse prevention program's effect on students' subsequent driving.

BACKGROUND: Alcohol-related injuries, particularly motor vehicle, are an important cause of adolescent mortality. School-based alcohol prevention programs have not been evaluated in terms of driving outcomes. This study examined the effects on subsequent driving of a high school-based alcohol prevention program. METHODS: The Alcohol Misuse Prevention Study included a randomized test of the effectiveness of an alcohol misuse prevention curriculum conducted among 4,635 10th-grade students. Students were assigned to intervention (n = 1,820) or control (n = 2,815) groups and were followed for an average of 7.6 years after licensure, which typically occurred during or shortly after 10th grade. Outcomes examined included alcohol-related and other serious offenses, and at-fault, single-vehicle, and alcohol-related crashes. RESULTS: Only serious offenses (which included alcohol-related) had a significant treatment effect (statistically marginal) after we adjusted for sex, age, race, alcohol use/misuse, family structure, presence of prelicense offenses, age of driver licensure, and parental attitudes toward teen drinking. The effect was found only during the first year of licensure (estimated adjusted relative risk = 0.80, confidence interval = 0.63-1.01). Two first-year serious offense interactions were found. The positive effect was strongest among the largest subgroup of students, those who were drinking less than one drink per week on average before the curriculum, compared with those who drank more than one drink per week (p = 0.009). The effect was also stronger for the small subgroup of students whose parents had not expressed disapproval of teens' drinking, compared with those whose parents had disapproved (p = 0.004). CONCLUSIONS: These findings suggest that a high school-based alcohol prevention program can positively affect subsequent driving, particularly that of students who do not use alcohol regularly. The results highlight the need to start prevention efforts early and extend them beyond the initial exposure to driving. Programs should incorporate the differing backgrounds of the students.

Changes in young adult offense and crash patterns over time.

A study of 13,809 young adult drivers in Michigan examined offenses and crashes ('incidents') for an average of 7 years after their original license date. During this period, 73% of subjects committed an offense that resulted in a conviction and 58% had a crash that was reported to the police. Forty-two percent had committed an offense classified as 'serious,' and 21% had an 'at-fault' crash. The odds of an offense being serious decreased approximately 8% per year of licensure, independent of gender or age at licensure. Similarly, the odds of a crash being at-fault decreased overall about 6% per year of licensure, but the decline was more than twice as fast for women as for men. Examining the empirical rates directly, it was found that the rate for minor offenses increased somewhat with time and then stabilized, while the rate for serious offenses declined. Also, offenses were less likely to be serious the later they occurred in the sequence of offenses for an individual. For crashes, the risk of having an at-fault crash declined more rapidly than the risk of a not-at-fault crash, although the rate of decrease began to equalize after approximately 5 years of licensure. The proportion of crashes that were at-fault did not decline over the sequence of crashes for an individual. Although crashes and offenses are positively correlated, they follow different trajectories over the early years of licensure.

Rear seating and risk of injury to child occupants by vehicle type.

The safety of rear-seated child passengers was evaluated across vehicle types. 113,887 children under age 16 in crashes were enrolled as part of an on-going crash surveillance system which links insurance claims data to telephone survey and crash investigation data. Children in the second row suffered less significant injuries than those in the front in all vehicle types except compact extended cab pickup trucks in which the risk for children in the rear was 13% as compared to 2.8% for front-seated occupants. Further research is needed to identify the child and vehicle characteristics which might explain this increased injury risk.