Misunderstandings about the effects of race and sex on physicians' referrals for cardiac catheterization

A nonparametric proportional risk model to assess a treatment effect in time-to-event data

Time-to-event analysis often relies on prior parametric assumptions, or, if a semiparametric approach is chosen, Cox's model. This is inherently tied to the assumption of proportional hazards, with the analysis potentially invalidated if this assumption is not fulfilled. In addition, most interpretations focus on the hazard ratio, that is often misinterpreted as the relative risk (RR), the ratio of the cumulative distribution functions. In this paper, we introduce an alternative to current methodology for assessing a treatment effect in a two-group situation, not relying on the proportional hazards assumption but assuming proportional risks. Precisely, we propose a new nonparametric model to directly estimate the RR of two groups to experience an event under the assumption that the risk ratio is constant over time. In addition to this relative measure, our model allows for calculating the number needed to treat as an absolute measure, providing the possibility of an easy and holistic interpretation of the data. We demonstrate the validity of the approach by means of a simulation study and present an application to data from a large randomized controlled trial investigating the effect of dapagliflozin on all-cause mortality.

Telehealth use among pediatric Alabama Medicaid enrollees, March-December 2020: Variations by race/ethnicity & place of residence

During the early days and months of the COVID-19 pandemic, healthcare facilities experienced a slump in non-COVID-related visits, and there was an increasing interest in telehealth to deliver healthcare services for adult and pediatric patients. The study investigated telehealth use variation by race/ethnicity and place of residence for the pediatric enrollees of the Alabama Medicaid program. This retrospective observational study examined Alabama Medicaid claims data from March to December 2020 for enrollees less than 19 years. There were 637,792 pediatric enrollees in the Alabama Medicaid program during the study period, and 16.9% of them had used telehealth to meet healthcare needs. This study employed a multivariate Poisson mixed-effects model with robust error variance to obtain differences in telehealth utilization and found that Non-Hispanic Black children were 80% as likely, Hispanic children were 55% as likely, and Asian Children were 46% as likely to have used telehealth compared to Non-Hispanic White children. Pediatric enrollees in large rural areas and isolated areas were significantly less likely (IRR: 0.90 for both, p<0.05) to use telehealth than those in urban areas. This study's findings suggest that attention needs to be paid to addressing race/ethnicity disparities in accessing telehealth services.

A rank-based approach to design and analysis of pretest-posttest randomized trials, with application to COVID-19 ordinal scale data

Randomized controlled trials with a pretest-posttest design frequently yield ordered categorical outcome data. Focusing on the estimation of the win probability that a treated participant would have a better score than (or win over) a control participant, we developed methods for analysis and sample size planning for such trials. We exploited the analysis of covariance framework with the dependent variable being individual participants' win fractions at posttest and the covariate being the win fractions at pretest. The win fractions were obtained using the mid-ranks of the ordinal data. Simulation evaluation based on a recent randomized trial on COVID-19 suggests that the methods perform very well. A sample SAS code for data analysis is presented.

Exploring the Application of Intersectionality as a Path toward Equity in Perinatal Health: A Scoping Review

OBJECTIVE

To conduct a scoping review to determine how past studies have applied the theory of intersectionality, a critical feminist research paradigm, to understand the physical health and mental health outcomes of perinatal people as a step toward addressing maternal health disparities and injustice. The study includes a review of existing research on maternal physical and mental health outcomes, presents the strengths and limitations of existing studies, and provides recommendations on best practices in applying intersectionality in research to address systemic issues and improve outcomes for the perinatal population.

METHODS

We conducted an extensive literature search across four search engines, yielding 28 publications using the intersectionality framework that focused on the outcomes of perinatal people, with a total sample of 9,856,042 participants. We examined how these studies applied intersectionality and evaluated them based on three areas: conceptualization, research method, and interpretation/findings.

RESULTS

Our findings indicate that maternal health researchers have provided good descriptions of the interaction of systemic inequalities and have used analysis that allows for the examination of interlocking and mutually reinforcing social positions or systems. We find that improvement is needed in the areas of conceptualization, reflexivity, and understanding of power structure. Recommendations are provided in the form of a checklist to guide future research toward an impactful approach to addressing perinatal health disparities.

RELEVANCE

Our scoping review has implications for improving applied health research to address perinatal health disparities, mortality, and morbidity. Recommendations are given along with references to other tools, and a guidance checklist is provided to support scholars in creating an impactful approach to applying intersectionality in the goal of addressing maternal health disparities.

Distribution-Free Approach to the Design and Analysis of Randomized Stroke Trials With the Modified Rankin Scale

BACKGROUND

Many methods have been suggested for analyzing the modified Rankin Scale (mRS). However, there lacks a unified approach to analysis and sample size determination that properly uses the ordinal nature of the data. We propose a simple method for CI estimation and corresponding sample size determination.

METHODS

We quantify treatment effect by the win probability (WinP) that a randomly selected patient in the treatment group has an equal or a better mRS score than a patient in the control group. Thus, a win probability of 0.5 means no effect, likened to a draw in competitive sports. We estimate the win probability and its SE based on the ranks of mRS scores, where tied scores are handled by average ranks. Corresponding methods for hypothesis testing, CI estimation, and sample size determination are derived. The methods are evaluated with a simulation study based on real data from 10 randomized stroke trials that used mRS as the outcome measure.

RESULTS

Simulation results demonstrated that the methods performed very well in terms of CI coverage, tail errors, and assurance to achieving the prespecified precision. Because the methods are very simple, we implemented them in an Excel spreadsheet, requiring only user inputs on frequencies of mRS scores in 2 comparison groups.

CONCLUSIONS

Sound statistical methods are important for the success of randomized stroke trials. The proposed methods and associated spreadsheet should prove useful for stroke researchers in the planning and analysis of randomized trials. Meta-analysis has also been made easy for trials with ordinal scores.

Use of Electronic Cigarettes Among Cannabis-Naive Adolescents and Its Association With Future Cannabis Use

IMPORTANCE

Electronic cigarette (e-cigarette) use has been reported to increase the likelihood of future cigarette smoking among adolescents. The prospective association between e-cigarette use and cannabis use has been less clear, especially in recent years.

OBJECTIVE

To examine the association between e-cigarette use among cannabis-naive adolescents and cannabis use 1 year later.

DESIGN, SETTING, AND PARTICIPANTS

The Population Assessment of Tobacco and Health (PATH) Study, a nationally representative cohort study, uses a 4-stage, stratified probability sample design to select participants aged 12 years or older from the US civilian, noninstitutionalized population. This study sample included 9828 cannabis-naive adolescents at the baseline survey who participated in both wave 4.5 (2017-2018) and wave 5 (2018-2019) of PATH.

EXPOSURES

e-Cigarette use, assessed by ever use, past 12-month use, and past 30-day use.

MAIN OUTCOMES AND MEASURES

Cannabis use in wave 5, assessed by past 12-month and past 30-day use. Multivariable logistic regressions assessed the association between e-cigarette use and cannabis use 1 year later. Results were weighted to produce nationally representative findings.

RESULTS

Of the 9828 adolescents included in the analysis, 5361 (57.3%) were aged 12 to 14 years, 5056 (50.7%) were male, and 4481 (53.0%) were non-Hispanic White. After adjustment for sociodemographic characteristics, environmental factors, other substance use, and sensation seeking, e-cigarette use among cannabis-naive adolescents was associated with increased likelihoods of both self-reported past 12-month and past 30-day cannabis use 1 year later. The adjusted relative risks (aRRs) of subsequent past 12-month cannabis use with ever use of e-cigarettes was 2.57 (95% CI, 2.04-3.09), with past 12-month use of e-cigarettes was 2.62 (95% CI, 2.10-3.15), and with past 30-day use of e-cigarettes was 2.18 (95% CI, 1.50-2.85). The aRRs of subsequent past 30-day cannabis use with ever use of e-cigarettes was 3.20 (95% CI, 2.10-4.31), with past 12-month use of e-cigarettes was 3.40 (95% CI, 2.17-4.63), and with past 30-day use of e-cigarettes was 2.96 (95% CI, 1.52-4.40).

CONCLUSIONS AND RELEVANCE

This cohort study's findings suggest a strong association between adolescent e-cigarette use and subsequent cannabis use. However, despite the strong association at the individual level, e-cigarette use seems to have had a minimal association with the prevalence of youth cannabis use at the population level.

Patient traits shape health-care stakeholders' choices on how to best allocate life-saving care

During global pandemics, health-care decision makers often face critical shortages of life-saving medical equipment. How do medical stakeholders prioritize which patients are most deserving of scarce treatment? We report the results of three conjoint experiments conducted in the United States in 2020, testing for biases in US physicians', citizens' and elected politicians' preferences for scarce ventilator distribution. We found that all stakeholders prioritized younger patients and patients who had a higher probability of surviving with ventilator access. When patients' survivability was tied, physicians prioritized patients from racial/ethnic minorities (that is, Asian, Black and Hispanic patients) over all-else-equal white patients, religious minorities (that is, Muslim patients) over religious majority group members (that is, Catholic patients) and patients of lower socio-economic status over wealthier patients. The public also prioritized Black and Hispanic patients over white patients but were biased against religious minorities (that is, Atheist and Muslim patients) relative to religious majority group members. Elected politicians were also biased against Atheist patients. Our effects varied by political party-with Republican physicians, politicians and members of the public showing bias against religious minority patients and Democratic physicians showing preferential treatment of racial and religious minorities. Our results suggest that health-care stakeholders' personal biases impact decisions on who deserves life-saving medical equipment.

An investigation of quantitative methods for assessing intersectionality in health research: A systematic review

Intersectionality is a theoretical framework that investigates how interlocking systems of power and oppression at the societal level influence the lived experiences of historically and socially marginalized groups. Currently, there are no consistent or widely adopted quantitative methods to investigate research questions informed by intersectionality theory. The objective of this systematic review is to describe the current landscape of quantitative methods used to assess intersectionality and to provide recommendations on analytic best practices for future research. We searched PubMed, EMBASE, and the Web of Science in December 2019 to identify studies using analytic quantitative intersectionality approaches published up to December 2019 (PROSPERO CRD42020162686). To be included in the study, articles had to: (1) be empirical research, (2) use a quantitative statistical method, (3) be published in English, and (4) incorporate intersectionality. Our initial search yielded 1889 articles. After screening by title/abstract, methods, and full text review, our final analytic sample included 153 papers. Eight unique classes of quantitative methods were identified, with the majority of studies employing regression with an interaction term. We additionally identified several methods which appear to be at odds with the key tenets of intersectionality. As quantitative intersectionality continues to expand, careful attention is needed to avoid the dilution of the core tenets. Specifically, emphasis on social power is needed as methods continue to be adopted and developed. Additionally, clear explanation of the selection of statistical approaches is needed and, when using regression with interaction terms, researchers should opt for use of the additive scale. Finally, use of methods that are potentially at odds with the tenets of intersectionality should be avoided.

Occupational factors and miscarriages in the US fire service: a cross-sectional analysis of women firefighters

BACKGROUND

Evidence from previous studies suggests that women firefighters have greater risk of some adverse reproductive outcomes. The purpose of this study was to investigate whether women firefighters had greater risk of miscarriage compared to non-firefighters and whether there were occupational factors associated with risk of miscarriage among firefighters.

METHODS

We studied pregnancies in the United States fire service using data from the Health and Wellness of Women Firefighters Study (n = 3181). We compared the prevalence of miscarriage among firefighters to published rates among non-firefighters using age-standardized prevalence ratios. We used generalized estimating equations to estimate relative risks (RRs) and 95% confidence intervals (CIs) between occupational factors (employment (career/volunteer), wildland firefighter status (wildland or wildland-urban-interface/structural), shift schedule, fire/rescue calls at pregnancy start) and risk of miscarriage, adjusted for age at pregnancy, education, gravidity, BMI, and smoking. We evaluated if associations varied by age at pregnancy or employment.

RESULTS

Among 1074 firefighters and 1864 total pregnancies, 404 pregnancies resulted in miscarriages (22%). Among most recent pregnancies, 138 resulted in miscarriage (13%). Compared to a study of US nurses, firefighters had 2.33 times greater age-standardized prevalence of miscarriage (95% CI 1.96-2.75). Overall, we observed that volunteer firefighters had an increased risk of miscarriage which varied by wildland status (interaction p-value< 0.01). Among structural firefighters, volunteer firefighters had 1.42 times the risk of miscarriage (95% CI 1.11-1.80) compared to career firefighters. Among wildland/wildland-urban-interface firefighters, volunteer firefighters had 2.53 times the risk of miscarriage (95% CI 1.35-4.78) compared to career firefighters.

CONCLUSIONS

Age-standardized miscarriage prevalence among firefighters may be greater than non-firefighters and there may be variation in risk of miscarriage by fire service role. Further research is needed to clarify these associations to inform policy and decision-making.

Title: Is OR "portable" in meta-analysis? Time to consider bivariate generalized linear mixed model

OBJECTIVES

 A recent paper by Doi et al. advocated completely replacing the relative risk (RR) with the odds ratio (OR) as the effect measure in clinical trials and meta-analyses with binary outcomes. Besides some practical advantages of RR over OR, Doi et al.'s key assumption that the OR is "portable" in the meta-analysis, i.e., study-specific ORs are likely not correlated with baseline risks, was not well justified. Study designs and settings: We summarized Spearman's rank correlation coefficient between study-specific ORs and baseline risks in 40,243 meta-analyses from the Cochrane Database of Systematic Reviews (CDSR).

RESULTS

Study-specific ORs tend to be higher in studies with lower baseline risks of disease for most meta-analyses in CDSR. Using an actual meta-analysis example, we demonstrate that there is a strong negative correlation between OR (RR or RD) with the baseline risk and the conditional effects notably vary with baseline risks.

CONCLUSIONS

 Replacing RR or RD with OR is currently unadvisable in clinical trials and meta-analyses. It is possible that no effect measure is "portable" in a meta-analysis. In addition to the overall (or marginal) effect, we suggest presenting the conditional effect based on the baseline risk using a bivariate generalized linear mixed model.

Completeness of reporting and risks of overstating impact in cluster randomised trials: a systematic review

Overstating the impact of interventions through incomplete or inaccurate reporting can lead to inappropriate scale-up of interventions with low impact. Accurate reporting of the impact of interventions is of great importance in global health research to protect scarce resources. In global health, the cluster randomised trial design is commonly used to evaluate complex, multicomponent interventions, and outcomes are often binary. Complete reporting of impact for binary outcomes means reporting both relative and absolute measures. We did a systematic review to assess reporting practices and potential to overstate impact in contemporary cluster randomised trials with binary primary outcome. We included all reports registered in the Cochrane Central Register of Controlled Trials of two-arm parallel cluster randomised trials with at least one binary primary outcome that were published in 2017. Of 73 cluster randomised trials, most (60 [82%]) showed incomplete reporting. Of 64 cluster randomised trials for which it was possible to evaluate, most (40 [63%]) reported results in such a way that impact could be overstated. Care is needed to report complete evidence of impact for the many interventions evaluated using the cluster randomised trial design worldwide.

Is OR "portable" in meta-analysis? Time to consider bivariate generalized linear mixed model

OBJECTIVES

A recent paper by Doi et al. advocated completely replacing the relative risk (RR) with the odds ratio (OR) as the effect measure used to report the association between a treatment and a binary outcome in clinical trials and meta-analyses. Besides some practical advantages of RR over OR and the well-known issue of the OR being non-collapsible, Doi et al.'s key assumption that the OR is "portable" in the meta-analysis, i.e., study-specific ORs are likely not correlated with baseline risks, was not well justified. Study designs and settings: We summarized the Spearman's rank correlation coefficient between study-specific OR and the baseline risk in 40,243 meta-analyses from the Cochrane Database of Systematic Reviews (CDSR).

RESULTS

Study-specific ORs are negatively correlated with baseline risk of disease (i.e., higher ORs tend to be observed in studies with lower baseline risks of disease) for most meta-analyses in CDSR. Using a meta-analysis comparing the effect of oral sumatriptan (100 mg) versus placebo on mitigating the acute headache at 2 hours after drug administration, we demonstrate that there is a strong negative correlation between OR (RR or RD) with the baseline risk and the conditional effects notably vary with baseline risks.

CONCLUSIONS

Replacing RR or RD with OR is currently unadvisable in clinical trials and meta-analyses. It is possible that no effect measure is "portable" in a meta-analysis. In cases where portability of the effect measure is challenging to satisfy, we suggest presenting the conditional effect based on the baseline risk using a bivariate generalized linear mixed model. The bivariate generalized linear mixed model can be used to account for correlation between the effect measure and baseline disease risk. Furthermore, in addition to the overall (or marginal) effect, we recommend that investigators also report the effects conditioning on the baseline risk.

Livestock ownership and anaemia in Sub-Saharan African households

OBJECTIVE

To determine the association between livestock ownership and Hb concentration of women of child-bearing age (WCBA) and preschool-aged children in Sub-Saharan Africa (SSA).

DESIGN

A prospective analysis of publicly available cross-sectional data, using linear and logistic regressions controlling for potential confounders.

SETTING

Twenty-eight countries in SSA.

PARTICIPANTS

162 305 WCBA and 118 607 children aged 6-59 months.

RESULTS

More than half of WCBA (62·5 %) and children (58 %) belonged to households that owned livestock. The average altitude-adjusted blood Hb concentration for WCBA and children was 12.23 and 10·24 g/dL, respectively. In adjusted models, higher number of livestock owned was associated with lower Hb concentration for children but not for WCBA. The magnitude of the association for children was small, with one additional unit of livestock owned reducing Hb concentration by 0·001 g/dL. Higher numbers of cattle, cows and bulls, sheep, and goats were associated with lower Hb concentration for both groups. The number of certain categories of livestock owned was associated with the consumption of relevant foods by children. There was no association between the consumption of animal-source foods and Hb concentration or between livestock ownership and diarrhoeal diseases or fever among children.

CONCLUSIONS

Livestock ownership in SSA had a net negative association with the Hb concentration of children and no association with that of WCBA. The results highlight the need for research aimed at clarifying the mechanisms linking livestock ownership and nutritional status, and identifying entry points for leveraging livestock ownership to improve the health of women and children in SSA.

Alive and Ventilator Free: A Hierarchical, Composite Outcome for Clinical Trials in the Acute Respiratory Distress Syndrome

OBJECTIVES

Survival from acute respiratory distress syndrome is improving, and outcomes beyond mortality may be important for testing new treatments. The "ventilator-free days" score, is an established composite that equates ventilation on day 28 to death. A hierarchical outcome treating death as a worse than prolonged ventilation would enhance face validity, but performance characteristics and reporting of such an outcome are unknown. We therefore evaluated the performance of a novel hierarchical composite endpoint, the Alive and Ventilator Free score.

DESIGN

Using data from four Acute Respiratory Distress Syndrome Network clinical trials, we compared Alive and Ventilator Free to the ventilator-free days score. Alive and Ventilator Free compares each patient with every other patient in a win-lose-tie for each comparison. Duration of mechanical ventilation is only compared if both patients survived. We evaluated power of Alive and Ventilator Free versus ventilator-free days score under various circumstances.

SETTING

ICUs within the Acute Respiratory Distress Syndrome Network.

PATIENTS

Individuals enrolled in four Acute Respiratory Distress Syndrome Network trials.

INTERVENTIONS

None for this analysis.

MEASUREMENTS AND MAIN RESULTS

Within the four trials (n = 2,410 patients), Alive and Ventilator Free and ventilator-free days score had similar power, with Alive and Ventilator Free slightly more powerful when a mortality difference was present, and ventilator-free days score slightly more powerful with a difference in duration of mechanical ventilation. Alive and Ventilator Free less often found in favor of treatments that increased mortality and increased days free of ventilation among survivors.

CONCLUSIONS

A hierarchical composite endpoint, Alive and Ventilator Free, preserves statistical power while improving face validity. Alive and Ventilator Free is less prone to favor a treatment with discordant effects on survival and days free of ventilation. This general approach can support complex outcome hierarchies with multiple constituent outcomes. Approaches to interpretation of differences in Alive and Ventilator Free are also presented.

Relative Measures of Association for Binary Outcomes: Challenges and Recommendations for the Global Health Researcher

Background:

Binary outcomes—which have two distinct levels (e.g., disease yes/no)—are commonly collected in global health research. The relative association of an exposure (e.g., a treatment) and such an outcome can be quantified using a ratio measure such as a risk ratio or an odds ratio. Although the odds ratio is more frequently reported than the risk ratio, many researchers, policymakers, and the general public frequently interpret it as a risk ratio. This is particularly problematic when the outcome is common because the magnitude of association is larger on the odds ratio scale than the risk ratio scale. Some recently published global health studies included misinterpretation of the odds ratio, which we hypothesize is because statistical methods for risk ratio estimation are not well known in the global health research community.

Objectives:

To compare and contrast available statistical methods to estimate relative measures of association for binary outcomes and to provide recommendations regarding their use.

Methods:

Logistic regression for odds ratios and four approaches for risk ratios: two direct regression approaches (modified log-Poisson and log-binomial) and two indirect methods (standardization and substitution) based on logistic regression.

Findings:

Illustrative examples demonstrate that misinterpretation of the odds ratio remains a common issue in global health research. Among the four methods presented for estimation of risk ratios, the modified log-Poisson approach is generally preferred because it has the best numerical performance and it is as easy to implement as is logistic regression for odds ratio estimation.

Conclusions:

We conclude that, when study design allows, studies with binary outcomes should preferably report risk ratios to measure relative association.

Methods for analytic intercategorical intersectionality in quantitative research: Discrimination as a mediator of health inequalities

RATIONALE

Intersectionality as a theoretical framework has gained prominence in qualitative research on social inequity. Intercategorical quantitative applications have focused primarily on describing health or social inequalities across intersectional groups, coded using cross-classified categories or interaction terms. This descriptive intersectionality omits consideration of the mediating processes (e.g., discrimination) through which intersectional positions impact outcome inequalities, which offer opportunities for intervention.

OBJECTIVE

We argue for the importance of a quantitative analytic intersectionality. We identify methodological challenges and potential solutions in structuring studies to allow for both intersectional heterogeneity in outcomes and in the ways that processes such as discrimination may cause these outcomes for those at different intersections.

METHOD

To incorporate both mediation and exposure-mediator interaction, we use VanderWeele's three-way decomposition methodology, adapt the interpretation for application to analytic intersectionality studies, and present a step-by-step analytic approach. Using online panel data collected from Canada and the United States in 2016 (N = 2542), we illustrate this approach with a statistical analysis of whether and to what extent observed inequalities in psychological distress across intersections of ethnoracial group and sexual or gender minority (SGM) status may be explained by past-year experiences of day-to-day discrimination, assessed using the Intersectional Discrimination Index (InDI).

RESULTS AND CONCLUSIONS

We describe actual and adjusted intersectional inequalities in psychological distress and decompose them to identify three component effects for each of 11 intersectional comparison groups (e.g., Indigenous SGM), versus the reference intersectional group that experienced the lowest levels of discrimination (white non-SGM). These reflect the expected inequality in outcome: 1) due to membership in the more discriminated-against group, if its members had experienced the same lower levels of discrimination as the reference intersection; 2) due to unequal levels of discrimination; and 3), due to unequal effects of discrimination. We present considerations for use and interpretation of these methods.

Relative rates not relative risks: addressing a widespread misinterpretation of hazard ratios

The use of the Cox proportional hazards model is ubiquitous in modern medical research. Despite the widespread implementation of this model, the terminology and interpretation that is used to describe the estimate hazard ratio (HR) has become loose and, unfortunately, often incorrect. Although some journals offer guidelines that advise against reporting HRs as relative risks, these guidelines are frequently overlooked. Perhaps due to a lack of understanding, authors continue to interpret the resultant HR as a relative risk-such an interpretation is inappropriate and can be misleading. The HR should be described as a relative rate, not as a relative risk. This article demonstrates that although the direction of the HR can be used to explain the direction of the relative risk, the magnitude of the HR alone cannot be used to explain the magnitude of the relative risk. This article clarifies the relationship between HRs and relative risks in a way that may be better suited for the applied clinical researcher. We also provide a convenient table illustrating the magnitude of relative risk under various values of the HR; the table demonstrates that for a given constant HR, the magnitude of the relative risk can vary substantially. As a take-home message, authors should refrain from using the magnitude of the HR to describe the magnitude of the relative risk. Authors should be strongly encouraged to ascribe accurate interpretations to the statistics derived from fitted Cox proportional hazards regression models.

Did I Tell You That? Ethical Issues Related to Using Computational Methods to Discover Non-Disclosed Patient Characteristics

Widespread availability of electronic health records coupled with sophisticated statistical methods offer great potential for a variety of applications for health and disease surveillance, developing predictive models and advancing decision support for clinicians. However, use of "big data" mining and discovery techniques has also raised ethical issues such as how to balance privacy and autonomy with the wider public benefits of data sharing. Furthermore, electronic data are being increasingly used to identify individual characteristics, which can be useful for clinical prediction and management, but were not previously disclosed to a clinician. This process in computer parlance is called electronic phenotyping, and has a number of ethical implications. Using the Belmont Report's principles of respect for persons, beneficence, and justice as a framework, we examined the ethical issues posed by electronic phenotyping. Ethical issues identified include the ability of the patient to consent for the use of their information, the ability to suppress pediatric information, ensuring that the potential benefits justify the risks of harm to patients, and acknowledging that the clinician's biases or stereotypes, conscious or unintended, may become a factor in the therapeutic interaction. We illustrate these issues with two vignettes, using the person characteristic of gender minority status (i.e., transgender identity) and health history characteristic of substance abuse. Data mining has the potential to uncover patient characteristics previously obscured, which can provide clinicians with beneficial clinical information. Hence, ethical guidelines must be updated to ensure that electronic phenotyping supports the principles of respect for persons, beneficence, and justice.

Evaluating Public Health Interventions: 6. Modeling Ratios or Differences? Let the Data Tell Us

We provide an overview of the relative merits of ratio measures (relative risks, risk ratios, and rate ratios) compared with difference measures (risk and rate differences). We discuss evidence that the multiplicative model often fits the data well, so that rarely are interactions with other risk factors for the outcome observed when one uses a logistic, relative risk, or Cox regression model to estimate the intervention effect. As a consequence, additive models, which estimate the risk or rate difference, will often exhibit interactions. Under these circumstances, absolute measures of effect, such as years of life lost, disability- or quality-adjusted years of life lost, and number needed to treat, will not be externally generalizable to populations other than those with similar risk factor distributions as the population in which the intervention effect was estimated. Nevertheless, these absolute measures are often of the greatest importance in public health decision-making. When studies of high-risk study populations are used to more efficiently estimate effects, these populations will not be representative of the general population's risk factor distribution. The relative homogeneity of ratio versus absolute measures will thus have important implications for the generalizability of results across populations.

Log Odds and the Interpretation of Logit Models

OBJECTIVE

We discuss how to interpret coefficients from logit models, focusing on the importance of the standard deviation (σ) of the error term to that interpretation.

STUDY DESIGN

We show how odds ratios are computed, how they depend on the standard deviation (σ) of the error term, and their sensitivity to different model specifications. We also discuss alternatives to odds ratios.

PRINCIPAL FINDINGS

There is no single odds ratio; instead, any estimated odds ratio is conditional on the data and the model specification. Odds ratios should not be compared across different studies using different samples from different populations. Nor should they be compared across models with different sets of explanatory variables.

CONCLUSIONS

To communicate information regarding the effect of explanatory variables on binary {0,1} dependent variables, average marginal effects are generally preferable to odds ratios, unless the data are from a case-control study.

Constraint approaches to the estimation of relative risk

In medical and epidemiologic studies, relative risk is usually the parameter of interest. However, calculating relative risk using standard log-Binomial regression approach often encounters non-convergence. A modified Poisson regression, which uses robust variance, was proposed by Zou in 2004. Although the modified Poisson regression with sandwich variance estimator is valid for the estimation of relative risk, the predicted probability of the outcome may be greater than the natural boundary 1 for the unobserved but plausible covariate combinations. Moreover, the lower and upper bounds of confidence intervals for predicted probabilities could fall out of (0, 1). Chu and Cole, in 2010, proposed a Bayesian approach to overcome this issue. Posterior median was used to get the parameter estimation. However, the Bayesian approach may provide biased estimation, especially when the probability of outcome is high. In this article, we propose an alternative constraint optimization approach for estimating relative risk. Our approach can reach similar or better performance than Bayesian approach in terms of bias, root mean square error, coverage rate, and predictive probabilities. Simulation studies are conducted to demonstrate the usefulness of this approach. Our method is also illustrated by Prospective Registry Evaluating Myocardial Infarction: Event and Recovery data.

Medial femur T2 Z-scores predict the probability of knee structural worsening over 4-8 years: Data from the osteoarthritis initiative

OBJECTIVE

The purpose of this study was to determine the probability of structural worsening of knee cartilage and whole joint degeneration over 4-8 years based on cartilage T₂ Z-scores at baseline.

DESIGN

Right knees with Kellgren-Lawrence (KL) grades of 0-2 in 587 participants from the Osteoarthritis Initiative were studied. 3T MR images were used to perform baseline cartilage T₂ quantification and assess 4-year changes in cartilage morphology (WORMS scoring) in 5 regions. Changes in joint space narrowing (JSN) and KL were assessed over 8 years. T₂ Z-scores were based on a reference database of knees without morphologic cartilage degeneration at baseline. Odds ratios for, and predicted probabilities of any worsening in WORMS cartilage, JSN and KL grade were obtained from logistic regression models.

RESULTS

A one-unit increase in the baseline medial femur T₂ Z-score was associated with cartilage worsening in the same region (OR = 1.59; P < 0.0001) and in any region (OR = 1.37; P < 0.0001), and with worsening JSN (OR = 1.82; P < 0.0001) and KL grades (OR = 1.69; P < 0.0001). Predicted probabilities of worsening in knees with a medial femur T₂ Z-score from 2-4 were 38% for medial femur cartilage WORMS, 70% for any cartilage region, 28% for increasing JSN and 31% for increasing KL grade.

CONCLUSION

Knees with elevated cartilage T₂ (especially in the medial femur and those that are 2 to 4 SDs above the mean reference values) are significantly more likely to have structural worsening over 4 to 8 years. Knowing cartilage T₂ Z-scores may aid in targeting prevention efforts at early stages of osteoarthritis.

LEVEL OF EVIDENCE

2 Technical Efficacy: Stage 3 J. Magn. Reson. Imaging 2017;46:1128-1136.

Dialogues in Diversity: An Invited Series of Papers, Advance Directives, DNRs, and End-of-Life Care for African Americans

The article utilizes a meta-analysis of the existing empirical research and theory on health care directives to provide some insights into the documented pattern of African Americans to use advance directives less than Whites. A number of relevant factors are highlighted and examined. In addition the article attempts to provide some insights into African American family life and traditional values regarding the care of the elderly and end-of-life care. The African American tradition of employing a family-centered decision making process during family crisis, as well as a significant cultural mistrust of institutionalized care is also explored. The article also attempts to offer some practical suggestions for clinical care givers working with African Americans to enhance culturally sensitive care giving and the utilization of advanced directives among African Americans at the end-of-life.

Helping Doctors and Patients Make Sense of Health Statistics

Many doctors, patients, journalists, and politicians alike do not understand what health statistics mean or draw wrong conclusions without noticing. Collective statistical illiteracy refers to the widespread inability to understand the meaning of numbers. For instance, many citizens are unaware that higher survival rates with cancer screening do not imply longer life, or that the statement that mammography screening reduces the risk of dying from breast cancer by 25% in fact means that 1 less woman out of 1,000 will die of the disease. We provide evidence that statistical illiteracy (a) is common to patients, journalists, and physicians; (b) is created by nontransparent framing of information that is sometimes an unintentional result of lack of understanding but can also be a result of intentional efforts to manipulate or persuade people; and (c) can have serious consequences for health.

The causes of statistical illiteracy should not be attributed to cognitive biases alone, but to the emotional nature of the doctor–patient relationship and conflicts of interest in the healthcare system. The classic doctor–patient relation is based on (the physician's) paternalism and (the patient's) trust in authority, which make statistical literacy seem unnecessary; so does the traditional combination of determinism (physicians who seek causes, not chances) and the illusion of certainty (patients who seek certainty when there is none). We show that information pamphlets, Web sites, leaflets distributed to doctors by the pharmaceutical industry, and even medical journals often report evidence in nontransparent forms that suggest big benefits of featured interventions and small harms. Without understanding the numbers involved, the public is susceptible to political and commercial manipulation of their anxieties and hopes, which undermines the goals of informed consent and shared decision making.

What can be done? We discuss the importance of teaching statistical thinking and transparent representations in primary and secondary education as well as in medical school. Yet this requires familiarizing children early on with the concept of probability and teaching statistical literacy as the art of solving real-world problems rather than applying formulas to toy problems about coins and dice. A major precondition for statistical literacy is transparent risk communication. We recommend using frequency statements instead of single-event probabilities, absolute risks instead of relative risks, mortality rates instead of survival rates, and natural frequencies instead of conditional probabilities. Psychological research on transparent visual and numerical forms of risk communication, as well as training of physicians in their use, is called for.

Statistical literacy is a necessary precondition for an educated citizenship in a technological democracy. Understanding risks and asking critical questions can also shape the emotional climate in a society so that hopes and anxieties are no longer as easily manipulated from outside and citizens can develop a better-informed and more relaxed attitude toward their health.

Methodological approaches to evaluate the impact of FDA drug safety communications

BACKGROUND

When the US FDA approves a new prescription drug there is still a great deal remaining to be learned about the safe and proper use of that product. When new information addressing these topics emerges post-approval, the FDA may issue a Drug Safety Communication (DSC) to alert patients and physicians. The effectiveness of the communication-how drug safety messaging conveyed in FDA DSCs changes patient or prescriber behavior-may depend on multiple factors, including the way physicians and patients learn about the information, their understanding of the issues conveyed, and their perception of the importance of the information. In 2013, the FDA issued two DSCs addressing critical new warnings related to products containing the sedative/hypnotic zolpidem.

OBJECTIVE

In this article, we describe a core set of research initiatives that can be used to study how zolpidem-related DSCs affected subsequent physician and patient decision making.

METHODS

These research initiatives include analyzing drug utilization patterns and related health outcomes; comparing zolpidem-containing products against a comparator with similar indications [eszopiclone (Lunesta)] not covered by the 2013 DSCs; and surveying patients and qualitatively evaluating the dissemination of information regarding these drugs in traditional and social-media channels.

CONCLUSIONS

Using an integrated, multidisciplinary approach, we can obtain information that can be used to optimize regulatory communications by seeking to understand the impact of the information contained in FDA risk communications.

Interpretation of Results of Studies Evaluating an Intervention Highlighted in Google Health News: A Cross-Sectional Study of News

BACKGROUND

Mass media through the Internet is a powerful means of disseminating medical research. We aimed to determine whether and how the interpretation of research results is misrepresented by the use of "spin" in the health section of Google News. Spin was defined as specific way of reporting, from whatever motive (intentional or unintentional), to emphasize that the beneficial effect of the intervention is greater than that shown by the results.

METHODS

We conducted a cross-sectional study of news highlighted in the health section of US, UK and Canada editions of Google News between July 2013 and January 2014. We searched for news items for 3 days a week (i.e., Monday, Wednesday, and Friday) during 6 months and selected a sample of 130 news items reporting a scientific article evaluating the effect of an intervention on human health.

RESULTS

In total, 78% of the news did not provide a full reference or electronic link to the scientific article. We found at least one spin in 114 (88%) news items and 18 different types of spin in news. These spin were mainly related to misleading reporting (59%) such as not reporting adverse events that were reported in the scientific article (25%), misleading interpretation (69%) such as claiming a causal effect despite non-randomized study design (49%) and overgeneralization/misleading extrapolation (41%) of the results such as extrapolating a beneficial effect from an animal study to humans (21%). We also identified some new types of spin such as highlighting a single patient experience for the success of a new treatment instead of focusing on the group results.

CONCLUSIONS

Interpretation of research results was frequently misrepresented in the health section of Google News. However, we do not know whether these spin were from the scientific articles themselves or added in the news.

The Impact of Vaccine Concerns on Racial/Ethnic Disparities in Influenza Vaccine Uptake Among Health Care Workers

OBJECTIVES

We explored whether collective concerns about the safety, effectiveness, and necessity of influenza vaccines mediate racial/ethnic disparities in vaccine uptake among health care workers (HCWs).

METHODS

We used a self-administered Web-based survey to assess race/ethnicity (exposure), concerns about influenza vaccination (mediator; categorized through latent class analysis), and influenza vaccine uptake (outcome) for the 2012 to 2013 influenza season among HCWs at St. Jude Children's Research Hospital in Memphis, Tennessee. We used mediation analysis to estimate prevalence ratios (PRs) and 95% confidence intervals (CIs) for the total, direct, and indirect effects of race/ethnicity on influenza vaccine uptake.

RESULTS

Non-Hispanic Blacks had lower influenza vaccine uptake than non-Hispanic Whites (total effect: PR = 0.87; 95% CI = 0.75, 0.99), largely mediated by high concern about influenza vaccines (natural indirect effect: PR = 0.89; 95% CI = 0.84, 0.94; controlled direct effect: PR = 0.98; 95% CI = 0.85, 1.1). Hispanic and Asian HCWs had modestly lower uptake than non-Hispanic Whites, also mediated by high concern about influenza vaccines.

CONCLUSIONS

Racial/ethnic disparities among HCWs could be attenuated if concerns about the safety, effectiveness, and necessity of influenza vaccines were reduced.

Relative risk reduction is useful metric to standardize effect size for public heath interventions for translational research

OBJECTIVES

Heterogeneity of effect measures in intervention studies undermines the use of evidence to inform policy. Our objective was to develop a comprehensive algorithm to convert all types of effect measures to one standard metric, relative risk reduction (RRR).

STUDY DESIGN AND SETTING

This work was conducted to facilitate synthesis of published intervention effects for our epidemic modeling of the health impact of human immunodeficiency virus [HIV testing and counseling (HTC)]. We designed and implemented an algorithm to transform varied effect measures to RRR, representing the proportionate reduction in undesirable outcomes.

RESULTS

Our extraction of 55 HTC studies identified 473 effect measures representing unique combinations of intervention-outcome-population characteristics, using five outcome metrics: pre-post proportion (70.6%), odds ratio (14.0%), mean difference (10.2%), risk ratio (4.4%), and RRR (0.9%). Outcomes were expressed as both desirable (29.5%, eg, consistent condom use) and undesirable (70.5%, eg, inconsistent condom use). Using four examples, we demonstrate our algorithm for converting varied effect measures to RRR and provide the conceptual basis for advantages of RRR over other metrics.

CONCLUSION

Our review of the literature suggests that RRR, an easily understood and useful metric to convey risk reduction associated with an intervention, is underused by original and review studies.

Coronary artery disease and the contours of pharmaceuticalization

Coronary artery disease (CAD) has dominated mortality for most of the past century, not just in Europe and North America but worldwide. Treatments for CAD, both pharmaceutical and surgical, have become leading sectors of the healthcare economy. This paper focuses on the therapeutic landscape for CAD in the United States. We hope to add texture to the broader conversation of pharmaceuticalization explored in this issue by situating pharmaceutical therapies as just one element in the broader therapeutic terrain, alongside cardiac surgery and interventional cardiology. Patients with CAD must navigate a therapeutic landscape with three intersecting paths: lifestyle change, pharmaceuticals, and surgery. While pharmaceuticals are often seen as a quick fix, a way of avoiding more difficult lifestyle changes, it is surgery and angioplasty that promise patients the quickest fix of all. There also is another option, often overlooked by analysts but popular among physicians and patients: inaction. The U.S. context is often critiqued as a site of excessive treatment with respect to both drugs and procedures, and yet there is deep stratification within it--over-treatment in many populations and under-treatment in others. People who experience the serious risks of CAD do so in a racialized terrain of durable preoccupations with difference and unequal access to care. While the pharmaceuticalization literature disproportionately attends to lifestyle drugs, which some observers consider to be medically inappropriate or unnecessary, CAD does remain the leading cause of death. Thus, the stakes are high. Examination of the pharmaceuticalization of CAD in light of surgical treatments and racial disparities offers a window into the pervasiveness and persuasiveness of pharmaceuticals in an increasingly consumer-driven medicine, as well as the limits of their appeal and their reach.

Which pathways to psychiatric care lead to earlier treatment and a shorter duration of first-episode psychosis?

BACKGROUND

The pathways to care in a first onset psychosis are diverse and may influence the chances of early treatment and therefore the duration of untreated psychosis. We test which pathways to care are associated with a delay in receiving treatment and a longer duration of untreated psychosis (DUP).

METHODS

In a population based survey, we interviewed 480 people with first episode psychosis aged 18 to 64 years over a 2-year period. Information from structured interview and case files provided DSM-IV diagnostic, clinical, and demographic information. Consecutive contacts in the care pathway were mapped using the World Health Organisation's Encounter Form. Using information from all sources, DUP was defined as time from symptom onset to first treatment with antipsychotic medication.

RESULTS

The most common first contacts were primary care physicians (35.2%), emergency rooms in general hospital settings (21.3%), and criminal justice agencies (25.4%). In multivariate regression models, compared to DUP for those first in contact with primary care, DUP was shortest for first encounters with psychiatric emergency clinics (RR = 0.4, 95% CI: 0.23-0.71) and longest for first encounters with criminal justice agencies (RR = 1.61, 95% CI: 1-2.58). Older age was associated with a longer DUP (RR = 1.01 per year, 95% CI: 1-1.04). A shorter DUP was associated with a diagnosis of mania and affective psychoses-NOS compared with schizophrenia (RR = 0.22, 95% CI: 0.14-0.35; RR = 0.18, 95% CI: 0.06-0.54, respectively), for Black compared with White ethnicity (RR = 0.52, 95% CI: 0.34-0.82), and for each close person in the social network (RR = 0.9, 95% CI: 0.84-0.96).

CONCLUSIONS

To further reduce DUP, better links are needed between primary care, emergency rooms, criminal justice and psychiatric services.

Detecting moderator effects using subgroup analyses

In the analysis of prevention and intervention studies, it is often important to investigate whether treatment effects vary among subgroups of patients defined by individual characteristics. These "subgroup analyses" can provide information about how best to use a new prevention or intervention program. However, subgroup analyses can be misleading if they test data-driven hypotheses, employ inappropriate statistical methods, or fail to account for multiple testing. These problems have led to a general suspicion of findings from subgroup analyses. This article discusses sound methods for conducting subgroup analyses to detect moderators. Multiple authors have argued that, to assess whether a treatment effect varies across subgroups defined by patient characteristics, analyses should be based on tests for interaction rather than treatment comparisons within the subgroups. We discuss the concept of heterogeneity and its dependence on the metric used to describe treatment effects. We discuss issues of multiple comparisons related to subgroup analyses and the importance of considering multiplicity in the interpretation of results. We also discuss the types of questions that would lead to subgroup analyses and how different scientific goals may affect the study at the design stage. Finally, we discuss subgroup analyses based on post-baseline factors and the complexity associated with this type of subgroup analysis.

Education and obesity at age 40 among American adults

Although many have studied the association between educational attainment and obesity, studies to date have not fully examined prior common causes and possible interactions by race/ethnicity or gender. It is also not clear if the relationship between actual educational attainment and obesity is independent of the role of aspired educational attainment or expected educational attainment. The authors use generalized linear log link models to examine the association between educational attainment at age 25 and obesity (BMI≥30) at age 40 in the USA's National Longitudinal Survey of Youth 1979 cohort, adjusting for demographics, confounders, and mediators. Race/ethnicity but not gender interacted with educational attainment. In a complete case analysis, after adjusting for socioeconomic covariates from childhood, adolescence, and adulthood, among whites only, college graduates were less likely than high school graduates to be obese (RR = 0.69, 95%CI: 0.57, 0.83). The risk ratio remained similar in two sensitivity analyses when the authors adjusted for educational aspirations and educational expectations and analyzed a multiply imputed dataset to address missingness. This more nuanced understanding of the role of education after controlling for a thorough set of confounders and mediators helps advance the study of social determinants of health and risk factors for obesity.

Maternal and Early Childhood Risk Factors for Overweight and Obesity among Low-Income Predominantly Black Children at Age Five Years: A Prospective Cohort Study

Objective. To identify maternal and early childhood risk factors for obesity and overweight among children at age 5 in the state of Alabama. Methods. We recruited 740 mothers during early pregnancy from University of Alabama Prenatal Clinics in a prospective cohort study and followed them throughout pregnancy. We followed their children from birth until 5 years of age. The main outcome measure was obesity (BMI for age and sex ≥ 95th percentile) at 5 years of age. We used poisson regression with robust variance estimation to compute risk ratio (RR). Results. At the 5th year of followup, 71 (9.6%) of the children were obese and 85 (11.5%) were overweight (BMI ≥ 85th-<95th percentile). In multivariable analysis, maternal prepregnancy overweight (RR: 2.30, 95% CI: 1.29-4.11) and obesity (RR: 2.53, 95% CI: 1.49-4.31), and child's birth weight >85th percentile (RR: 2.04, 95% CI: 1.13-3.68) were associated with childhood obesity. Maternal prepregnancy BMI, birth weight, and maternal smoking were associated with the child being overweight 1-12 cigarettes/day versus 0 cigarettes/day (RR: 1.40, 95% CI: 1.02-1.91). Conclusion. Children of overweight and obese mothers, and children with higher birth weight, are more likely to be obese and overweight at age 5. Maternal smoking 1-12 cigarettes per day is associated with the child being overweight.

Misuse of odds ratios in obesity literature: an empirical analysis of published studies

Odds ratios (ORs) are widely used in scientific research to demonstrate the associations between outcome variables and covariates (risk factors) of interest, and are often described in language suitable for risks or probabilities, but odds and probabilities are related, not equivalent. In situations where the outcome is not rare (e.g., obesity), ORs no longer approximate the relative risk ratio (RR) and may be misinterpreted. Our study examines the extent of misinterpretation of ORs in Obesity and International Journal of Obesity. We reviewed all 2010 issues of these journals to identify all articles that presented ORs. Included articles were then primarily reviewed for correct presentation and interpretation of ORs; and secondarily reviewed for article characteristics that may have been associated with how ORs are presented and interpreted. Of the 855 articles examined, 62 (7.3%) presented ORs. ORs were presented incorrectly in 23.2% of these articles. Clinical articles were more likely to present ORs correctly than social science or basic science articles. Studies with outcome variables that had higher relative prevalence were less likely to present ORs correctly. Overall, almost one-quarter of the studies presenting ORs in two leading journals on obesity misinterpreted them. Furthermore, even when researchers present ORs correctly, the lay media may misinterpret them as relative RRs. Therefore, we suggest that when the magnitude of associations is of interest, researchers should carefully and accurately present interpretable measures of association--including RRs and risk differences--to minimize confusion and misrepresentation of research results.

[Do media reports and public brochures facilitate informed decision making about cervical cancer prevention?]

With the introduction and recommendation of the new HPV (human papillomavirus) vaccination in 2007, cervical cancer prevention has evoked large public interest. Is the public able to make informed decisions on the basis of media reports and brochures? To answer this question, an analysis of media coverage of HPV vaccination (Gardasil®) and Pap (Papanicolaou) screening was conducted from 2007-2009, which investigated the minimum requirement of completeness (pros and cons), transparency (absolute numbers), and correctness (references concerning outcome, uncertainty, magnitude) of the information. As a bench mark, facts boxes with concise data on epidemiology, etiology, benefits, harms, and costs were compiled in advance. Although all vaccination reports and brochures covered the impact of prevention, only 41% provided concrete numbers on effectiveness (90/220) and 2% on absolute risk reductions for the cancer surrogate dysplasia (5/220), whereby none of the latter numbers was correct. The prevention potential was correctly presented once. Only 48% (105/220) mentioned pros and cons. With regard to screening, 20% (4/20) provided explicit data on test quality and one expressed these in absolute numbers, while 25% (5/20) reported the prevention potential; all given numbers were correct. Finally, 25% (5/20) mentioned the possibility of false positive results. Minimum requirements were fulfilled by 1/220 vaccination and 1/20 screening reports. At present, informed decision making based on media coverage is hardly possible.

Ratio of geometric means to analyze continuous outcomes in meta-analysis: comparison to mean differences and ratio of arithmetic means using empiric data and simulation

Meta-analyses pooling continuous outcomes can use mean differences (MD), standardized MD (MD in pooled standard deviation units, SMD), or ratio of arithmetic means (RoM). Recently, ratio of geometric means using ad hoc (RoGM (ad hoc) ) or Taylor series (RoGM (Taylor) ) methods for estimating variances have been proposed as alternative effect measures for skewed continuous data. Skewed data are suggested for summary measures of clinical parameters restricted to positive values which have large coefficients of variation (CV). Our objective was to compare performance characteristics of RoGM (ad hoc) and RoGM (Taylor) to MD, SMD, and RoM. We used empiric data from systematic reviews reporting continuous outcomes and selected from each the meta-analysis with the most and at least 5 trials (Cochrane Database [2008, Issue 1]). We supplemented this with simulations conducted with representative parameters. Pooled results were calculated using each effect measure. Of the reviews, 232/5053 met the inclusion criteria. Empiric data and simulation showed that RoGM (ad hoc) exhibits more extreme treatment effects and greater heterogeneity than all other effect measures. Compared with MD, SMD, and RoM, RoGM (Taylor) exhibits similar treatment effects, more heterogeneity when CV ≤0.7, and less heterogeneity when CV > 0.7. In conclusion, RoGM (Taylor) may be considered for pooling continuous outcomes in meta-analysis when data are skewed, but RoGM (ad hoc) should not be used. However, clinicians' lack of familiarity with geometric means combined with acceptable performance characteristics of RoM in most situations suggests that RoM may be the preferable ratio method for pooling continuous outcomes in meta-analysis.

A simple method for estimating relative risk using logistic regression

BACKGROUND

Odds ratios (OR) significantly overestimate associations between risk factors and common outcomes. The estimation of relative risks (RR) or prevalence ratios (PR) has represented a statistical challenge in multivariate analysis and, furthermore, some researchers do not have access to the available methods.

OBJECTIVE

To propose and evaluate a new method for estimating RR and PR by logistic regression.

METHODS

A provisional database was designed in which events were duplicated but identified as non-events. After, a logistic regression was performed and effect measures were calculated, which were considered RR estimations. This method was compared with binomial regression, Cox regression with robust variance and ordinary logistic regression in analyses with three outcomes of different frequencies.

RESULTS

ORs estimated by ordinary logistic regression progressively overestimated RRs as the outcome frequency increased. RRs estimated by Cox regression and the method proposed in this article were similar to those estimated by binomial regression for every outcome. However, confidence intervals were wider with the proposed method.

CONCLUSION

This simple tool could be useful for calculating the effect of risk factors and the impact of health interventions in developing countries when other statistical strategies are not available.

Misrepresentation of randomized controlled trials in press releases and news coverage: a cohort study

BACKGROUND

Previous studies indicate that in published reports, trial results can be distorted by the use of "spin" (specific reporting strategies, intentional or unintentional, emphasizing the beneficial effect of the experimental treatment). We aimed to (1) evaluate the presence of "spin" in press releases and associated media coverage; and (2) evaluate whether findings of randomized controlled trials (RCTs) based on press releases and media coverage are misinterpreted.

METHODS AND FINDINGS

We systematically searched for all press releases indexed in the EurekAlert! database between December 2009 and March 2010. Of the 498 press releases retrieved and screened, we included press releases for all two-arm, parallel-group RCTs (n = 70). We obtained a copy of the scientific article to which the press release related and we systematically searched for related news items using Lexis Nexis. "Spin," defined as specific reporting strategies (intentional or unintentional) emphasizing the beneficial effect of the experimental treatment, was identified in 28 (40%) scientific article abstract conclusions and in 33 (47%) press releases. From bivariate and multivariable analysis assessing the journal type, funding source, sample size, type of treatment (drug or other), results of the primary outcomes (all nonstatistically significant versus other), author of the press release, and the presence of "spin" in the abstract conclusion, the only factor associated, with "spin" in the press release was "spin" in the article abstract conclusions (relative risk [RR] 5.6, [95% CI 2.8-11.1], p < 0.001). Findings of RCTs based on press releases were overestimated for 19 (27%) reports. News items were identified for 41 RCTs; 21 (51%) were reported with "spin," mainly the same type of "spin" as those identified in the press release and article abstract conclusion. Findings of RCTs based on the news item was overestimated for ten (24%) reports.

CONCLUSION

"Spin" was identified in about half of press releases and media coverage. In multivariable analysis, the main factor associated with "spin" in press releases was the presence of "spin" in the article abstract conclusion.