Effects of disease misclassification on exposure-disease association

Observational Dose-Response Meta-Analysis Methods May Bias Risk Estimates at Low Consumption Levels: The Case of Meat and Colorectal Cancer

Observational studies of foods and health are susceptible to bias, particularly from confounding between diet and other lifestyle factors. Common methods for deriving dose-response meta-analysis (DRMA) may contribute to biased or overly certain risk estimates. We used DRMA models to evaluate the empirical evidence for colorectal cancer (CRC) association with unprocessed red meat (RM) and processed meats (PM), and the consistency of this association for low and high consumers under different modeling assumptions. Using the Global Burden of Disease project's systematic reviews as a start, we compiled a data set of studies of PM with 29 cohorts contributing 23,522,676 person-years and of 23 cohorts for RM totaling 17,259,839 person-years. We fitted DRMA models to lower consumers only [consumption < United States median of PM (21 g/d) or RM (56 g/d)] and compared them with DRMA models using all consumers. To investigate impacts of model selection, we compared classical DRMA models against an empirical model for both lower consumers only and for all consumers. Finally, we assessed if the type of reference consumer (nonconsumer or mixed consumer/nonconsumer) influenced a meta-analysis of the lowest consumption arm. We found no significant association with consumption of 50 g/d RM using an empirical fit with lower consumption (relative risk [RR] 0.93 (0.8-1.02) or all consumption levels (1.04 (0.99-1.10)), while classical models showed RRs as high as 1.09 (1.00-1.18) at 50g/day. PM consumption of 20 g/d was not associated with CRC (1.01 (0.87-1.18)) when using lower consumer data, regardless of model choice. Using all consumption data resulted in association with CRC at 20g/day of PM for the empirical models (1.07 (1.02-1.12)) and with as little as 1g/day for classical models. The empirical DRMA showed nonlinear, nonmonotonic relationships for PM and RM. Nonconsumer reference groups did not affect RM (P = 0.056) or PM (P = 0.937) association with CRC in lowest consumption arms. In conclusion, classical DRMA model assumptions and inclusion of higher consumption levels influence the association between CRC and low RM and PM consumption. Furthermore, a no-risk limit of 0 g/d consumption of RM and PM is inconsistent with the evidence.

Building Learning Healthcare Systems for Critical Care Medicine

Learning healthcare systems are an evolving way of integrating informatics, analytics, and continuous improvement into daily practice in healthcare. This article discusses strategies to build learning healthcare systems for critical care medicine.

Double-Negative Results Matter: A Reevaluation of Sensitivities for Detecting SARS-CoV-2 Infection Using Saliva Versus Nasopharyngeal Swabs

In a recent systematic review, Bastos et al. (Ann Intern Med. 2021;174(4):501-510) compared the sensitivities of saliva sampling and nasopharyngeal swabs in the detection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection by assuming a composite reference standard defined as positive if either test is positive and negative if both tests are negative (double negative). Even under a perfect specificity assumption, this approach ignores the double-negative results and risks overestimating the sensitivities due to residual misclassification. In this article, we first illustrate the impact of double-negative results in the estimation of the sensitivities in a single study, and then propose a 2-step latent class meta-analysis method for reevaluating both sensitivities using the same published data set as that used in Bastos et al. by properly including the observed double-negative results. We also conduct extensive simulation studies to compare the performance of the proposed method with Bastos et al.'s method for varied levels of prevalence and between-study heterogeneity. The results demonstrate that the sensitivities are overestimated noticeably using Bastos et al.'s method, and the proposed method provides a more accurate evaluation with nearly no bias and close-to-nominal coverage probability. In conclusion, double-negative results can significantly impact the estimated sensitivities when a gold standard is absent, and thus they should be properly incorporated.

Understanding the Extent of Polypharmacy and its Association With Health Service Utilization Among Persons With Cancer and Multimorbidity: A Population-Based Retrospective Cohort Study in Ontario, Canada

Background: Cancer often co-occurs with other chronic conditions, which may result in polypharmacy. Polypharmacy is associated with adverse outcomes, including increased health service utilization. Objectives: This study examines the overall prevalence of polypharmacy (5 or more medications) among adults with cancer and multimorbidity, as well as the association of both minor polypharmacy (5-9 medications) and hyper-polypharmacy (10 or more medications) on high use of emergency room visits and hospitalizations, while controlling for age, sex, and type and stage of cancer. Methods: This retrospective longitudinal study used linked health administrative databases and included persons 18 years and older diagnosed with cancer between April 2010 and March 2013 in Ontario, Canada. Data on the number of health service utilizations at or above the 90th percentile (high users), was collected up to March 2014 and multivariate logistic regression was used to determine the impact of polypharmacy. Results: The prevalence of polypharmacy was 46% prior to cancer diagnosis, and 57% one year after diagnosis. Polypharmacy prior to and after cancer diagnosis increased with the level of multimorbidity, increasing age, but did not differ by sex. It was also highest in persons with lung cancer (52.4%) and those diagnosed with stage 4 cancer (51.3%). Minor polypharmacy increased the odds of being a high user of emergency rooms (1.16; 99% CI: 1.09-1.24) and hospitalizations (1.03; 0.98-1.09) and the odds of high use was greater with hyper-polypharmacy (1.41; 1.33-1.51) and (1.23; 1.17-1.29) respectively. Conclusion: Polypharmacy is highly prevalent and is associated with high health service utilization among adults with cancer and multimorbidity.

Nightmares: an independent risk factor for cardiovascular disease?

STUDY OBJECTIVES

Prior work has established associations between post-traumatic stress disorder (PTSD), disrupted sleep, and cardiovascular disease (CVD), but few studies have examined health correlates of nightmares beyond risks conferred by PTSD. This study examined associations between nightmares and CVD in military veterans.

METHODS

Participants were veterans (N = 3468; 77% male) serving since September 11, 2001, aged 38 years (SD = 10.4); approximately 30% were diagnosed with PTSD. Nightmare frequency and severity were assessed using the Davidson Trauma Scale (DTS). Self-reported medical issues were assessed using the National Vietnam Veterans Readjustment Study Self-report Medical Questionnaire. Mental health disorders were established using the Structured Clinical Interview for DSM-IV. The sample was stratified by the presence or absence of PTSD. Within-group associations between nightmare frequency and severity and self-reported CVD conditions, adjusting for age, sex, race, current smoking, depression, and sleep duration.

RESULTS

Frequent and severe nightmares during the past week were endorsed by 32% and 35% of participants, respectively. Those endorsing nightmares that were frequent, severe, and the combination thereof were more likely to also evidence high blood pressure (ORs 1.42, OR 1.56, and OR 1.47, respectively) and heart problems (OR 1.43, OR 1.48, and OR 1.59, respectively) after adjusting for PTSD diagnosis and other covariates.

CONCLUSIONS

Nightmare frequency and severity among veterans are associated with cardiovascular conditions, even after controlling for PTSD diagnosis. Study findings suggest that nightmares may be an independent risk factor for CVD. Additional research is needed to validate these findings using confirmed diagnoses and explore potential mechanisms.

Long-Term Exposure to Traffic-Related Air Pollution and Diabetes: A Systematic Review and Meta-Analysis

Objectives: We report results of a systematic review on the health effects of long-term traffic-related air pollution (TRAP) and diabetes in the adult population. Methods: An expert Panel appointed by the Health Effects Institute conducted this systematic review. We searched the PubMed and LUDOK databases for epidemiological studies from 1980 to July 2019. TRAP was defined based on a comprehensive protocol. Random-effects meta-analyses were performed. Confidence assessments were based on a modified Office for Health Assessment and Translation (OHAT) approach, complemented with a broader narrative synthesis. We extended our interpretation to include evidence published up to May 2022. Results: We considered 21 studies on diabetes. All meta-analytic estimates indicated higher diabetes risks with higher exposure. Exposure to NO₂ was associated with higher diabetes prevalence (RR 1.09; 95% CI: 1.02; 1.17 per 10 μg/m³), but less pronounced for diabetes incidence (RR 1.04; 95% CI: 0.96; 1.13 per 10 μg/m³). The overall confidence in the evidence was rated moderate, strengthened by the addition of 5 recently published studies. Conclusion: There was moderate evidence for an association of long-term TRAP exposure with diabetes.

Vaginal Microbiome Metagenome Inference Accuracy: Differential Measurement Error according to Community Composition

Several studies have compared metagenome inference performance in different human body sites; however, none specifically reported on the vaginal microbiome. Findings from other body sites cannot easily be generalized to the vaginal microbiome due to unique features of vaginal microbial ecology, and investigators seeking to use metagenome inference in vaginal microbiome research are "flying blind" with respect to potential bias these methods may introduce into analyses. We compared the performance of PICRUSt2 and Tax4Fun2 using paired 16S rRNA gene amplicon sequencing and whole-metagenome sequencing data from vaginal samples from 72 pregnant individuals enrolled in the Pregnancy, Infection, and Nutrition (PIN) cohort. Participants were selected from those with known birth outcomes and adequate 16S rRNA gene amplicon sequencing data in a case-control design. Cases experienced early preterm birth (<32 weeks of gestation), and controls experienced term birth (37 to 41 weeks of gestation). PICRUSt2 and Tax4Fun2 performed modestly overall (median Spearman correlation coefficients between observed and predicted KEGG ortholog [KO] relative abundances of 0.20 and 0.22, respectively). Both methods performed best among Lactobacillus crispatus-dominated vaginal microbiotas (median Spearman correlation coefficients of 0.24 and 0.25, respectively) and worst among Lactobacillus iners-dominated microbiotas (median Spearman correlation coefficients of 0.06 and 0.11, respectively). The same pattern was observed when evaluating correlations between univariable hypothesis test P values generated with observed and predicted metagenome data. Differential metagenome inference performance across vaginal microbiota community types can be considered differential measurement error, which often causes differential misclassification. As such, metagenome inference will introduce hard-to-predict bias (toward or away from the null) in vaginal microbiome research. IMPORTANCE Compared to taxonomic composition, the functional potential within a bacterial community is more relevant to establishing mechanistic understandings and causal relationships between the microbiome and health outcomes. Metagenome inference attempts to bridge the gap between 16S rRNA gene amplicon sequencing and whole-metagenome sequencing by predicting a microbiome's gene content based on its taxonomic composition and annotated genome sequences of its members. Metagenome inference methods have been evaluated primarily among gut samples, where they appear to perform fairly well. Here, we show that metagenome inference performance is markedly worse for the vaginal microbiome and that performance varies across common vaginal microbiome community types. Because these community types are associated with sexual and reproductive outcomes, differential metagenome inference performance will bias vaginal microbiome studies, obscuring relationships of interest. Results from such studies should be interpreted with substantial caution and the understanding that they may over- or underestimate associations with metagenome content.

Association of Uncontrolled Hypertension or Diabetes Mellitus With Major Adverse Cardiovascular Events and Mortality in South Korea: Population-Based Cohort Study

BACKGROUND

Managing hypertension (HT) and diabetes mellitus (DM) is crucial to preventing cardiovascular diseases. Few studies have investigated the incidence and risk of cardiovascular diseases or mortality in uncontrolled HT or DM in the Asian population. Epidemiological studies of cardiovascular disease should be conducted with continuous consideration of the changing disease risk profiles, lifestyles, and socioeconomic status over time.

OBJECTIVE

We aimed to examine the association of uncontrolled HT or DM with the incidence of cardiovascular events or deaths from any cause.

METHODS

This population-based retrospective study was conducted using data from the Korean National Health Insurance Service-National Health Screening Cohort, including patients aged 40-79 years who participated in national screening from 2002 to 2003 and were followed up until 2015. The health screening period from 2002 to 2013 was stratified into 6 index periods in 2-year cycles, and the follow-up period from 2004 to 2015 was stratified accordingly into 6 subsequent 2-year periods. The incidence rates and hazard ratio (HR) for major adverse cardiovascular events (MACE) and death from any cause were estimated according to HT or DM control status. Extended Cox models with time-dependent variables updated every 2 years, including sociodemographic characteristics, blood pressure (BP), fasting blood glucose (FBG), medication prescription, and adherence, were used.

RESULTS

Among the total cohort of 440,249 patients, 155,765 (35.38%) were in the uncontrolled HT or DM group. More than 60% of the patients with HT or DM who were prescribed medications did not achieve the target BP or FBG. The incidence of MACE was 10.8-15.5 and 9.6-13.3 per 1000 person-years in the uncontrolled DM and uncontrolled HT groups, respectively, and increased with age. In the uncontrolled HT and DM group, the incidence of MACE was high (15.2-17.5 per 1000 person-years) at a relatively young age and showed no age-related trend. Adjusted HR for MACE were 1.28 (95% CI 1.23-1.32) for the uncontrolled DM group, 1.32 (95% CI 1.29-1.35) for the uncontrolled HT group, and 1.54 (95% CI 1.47-1.60) for the uncontrolled HT and DM group. Adjusted HR for death from any cause were 1.05 (95% CI 1.01-1.10) for the uncontrolled DM group, 1.13 (95% CI 1.10-1.16) for the uncontrolled HT group, and 1.17 (95% CI 1.12-1.23) for the uncontrolled HT and DM group.

CONCLUSIONS

This up-to-date evidence of cardiovascular epidemiology in South Korea serves as the basis for planning public health policies to prevent cardiovascular diseases. The high uncontrolled rates of HT or DM, regardless of medication prescription, have led us to suggest the need for a novel system for effective BP or glycemic control, such as a community-wide management program using mobile health technology.

Food insecurity and its associations with bulimic-spectrum eating disorders, mood disorders, and anxiety disorders in a nationally representative sample of U.S. adults

PURPOSE

To examine cross-sectional associations between food insecurity and 12-month eating disorders, mood disorders, and anxiety disorders among U.S. adults.

METHODS

This study used data collected between 2001 and 2003 from 2914 participants in the National Comorbidity Survey-Replication, a nationally representative sample of U.S. adults (mean age = 44.9 years; 53.4% female). Twelve-month food insecurity was assessed with a modified version of the Short Form U.S. Household Food Security Scale. Twelve-month DSM-IV diagnoses of mental disorders were based on the World Health Organization Composite International Diagnostic Interview. Modified Poisson regression models were conducted, adjusting for age, sex, race/ethnicity, education, and income-to-poverty ratio.

RESULTS

Food insecurity was experienced by 11.1% of participants. Food insecurity was associated with greater prevalence of bulimic-spectrum eating disorders (prevalence ratio [PR] = 3.81; 95% confidence interval [CI] 2.26-6.42), mood disorders (PR = 2.53; 95% CI 1.96-3.29), and anxiety disorders (PR = 1.69; 95% CI 1.39-2.07).

CONCLUSION

Results indicate that food insecurity is associated with a range of internalizing mental disorders, though these findings should be confirmed with contemporary data to reflect DSM-5 diagnostic updates and the economic effects of the COVID-19 pandemic. Findings from this study emphasize the need to expand food insecurity interventions and improve access to mental health services for food-insecure populations.

Recall and Interviewer Bias in Parental Report of Pediatric Exposure to Aromatic Plant Ingredients in Personal Care Products: Development and Validation of a More Accurate Approach

Lavender and tea tree essential oils are traditionally considered to be mild, gentle, and safe for pediatric populations and are ubiquitous in personal care products. Recent case reports have proposed a potential association between exposure to these ingredients and endocrine disruption, but these reports contain misclassification bias. The purpose of this study is to develop a reliable and valid measurement instrument for the accurate classification of exposure to aromatic plant ingredients in personal care products to be used in epidemiological studies. This study tested the Aromatic Plant Ingredients and Child Health Survey (APICHS) for validity and reliability, contrasting it with the current approach used in clinician's offices. The APICHS was found to have exceptional sensitivity and specificity (100% and 92.86%, respectively) with a positive predictive value of 97.22%, far exceeding the sensitivity and specificity of the method currently in use. The APICHS is a valid, reliable tool for accurate classification of exposure to aromatic plant ingredients in personal care products and should be used for the avoidance of misclassification.

Addressing Bias in Responder Analysis of Patient-Reported Outcomes

INTRODUCTION

Quantitative patient-reported outcome (PRO) measures ideally are analyzed on their original scales and responder analyses are used to aid the interpretation of those primary analyses. As stated in the FDA PRO Guidance for Medical Product Development (2009), one way to lend meaning and interpretation to such a PRO measure is to dichotomize between values where within-patient changes are considered clinically important and those that are not. But even a PRO scale with a cutoff score that discriminates well between responder and non-responders is fraught with some misclassification.

METHODS

Using estimates of sensitivity and specificity on classification of responder status from a PRO instrument, formulas are provided to correct for such responder misclassification under the assumption of no treatment misclassification. Two case studies from sexual medicine illustrate the methodology.

RESULTS

Adjustment formulas on cell counts for responder misclassification are a direct extension of correction formulas for misclassification on disease from a two-way cross-classification table of disease (yes, no) and exposure (yes, no). Unadjusted and adjusted estimates of treatment effect are compared in terms of odds ratio, response ratio, and response difference. In the two case studies, there was considerable underestimation of treatment effect.

DISCUSSION AND CONCLUSIONS

The methodology can be applied to different therapeutic areas. Limitations of the methodology, such as when adjusted cell estimates become negative, are highlighted. The role of anchor-based methodology is discussed for obtaining estimates of sensitivity and specificity on responder classification. Correction for treatment effect bias from misclassification of responder status on PRO measures can lead to more trustworthy interpretation and effective decision-making. Clinicaltrials.gov: NCT00343200.

To screen or not to screen? The development of a prediction model for aorto-iliac stenosis in kidney transplant candidates

Screening for aorto‐iliac stenosis is important in kidney transplant candidates as its presence affects pre‐transplantation decisions regarding side of implantation and the need for an additional vascular procedure. Reliable imaging techniques to identify this condition require contrast fluid, which can be harmful in these patients. To guide patient selection for these imaging techniques, we aimed to develop a prediction model for the presence of aorto‐iliac stenosis. Patients with contrast‐enhanced imaging available in the pre‐transplant screening between January 1st, 2000 and December 31st, 2018 were included. A prediction model was developed using multivariable logistic regression analysis and internally validated using bootstrap resampling. Model performance was assessed with the concordance index and calibration slope. Three hundred and seventy‐three patients were included, 90 patients (24.1%) had imaging‐proven aorto‐iliac stenosis. Our final model included age, smoking, peripheral arterial disease, coronary artery disease, a previous transplant, intermittent claudication and the presence of a femoral artery murmur. The model yielded excellent discrimination (optimism‐corrected concordance index: 0.83) and calibration (optimism‐corrected calibration slope: 0.91). In conclusion, this prediction model can guide the development of standardized protocols to decide which patients should receive vascular screening to identify aorto‐iliac stenosis. External validation is needed before this model can be implemented in patient care.

Intergenerational nutrition benefits of India's national school feeding program

India has the world's highest number of undernourished children and the largest school feeding program, the Mid-Day Meal (MDM) scheme. As school feeding programs target children outside the highest-return "first 1000-days" window, they have not been included in the global agenda to address stunting. School meals benefit education and nutrition in participants, but no studies have examined whether benefits carry over to their children. Using nationally representative data on mothers and their children spanning 1993 to 2016, we assess whether MDM supports intergenerational improvements in child linear growth. Here we report that height-for-age z-score (HAZ) among children born to mothers with full MDM exposure was greater (+0.40 SD) than that in children born to non-exposed mothers. Associations were stronger in low socioeconomic strata and likely work through women's education, fertility, and health service utilization. MDM was associated with 13-32% of the HAZ improvement in India from 2006 to 2016.

Risk of Stevens-Johnson syndrome and toxic epidermal necrolysis associated with anticonvulsants in a Japanese population: Matched case-control and cohort studies

BACKGROUND

Evidence for the risk and incidence of anticonvulsant-induced Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) in Japan is scarce.

METHODS

We conducted a matched case-control study using a large-scale Japanese claims database. SJS/TEN cases were identified using a claims-based algorithm developed in a previous study (sensitivity 76.9%, specificity 99.0%). Conditional logistic regression with Firth's bias correction to address an issue of rare events was used to estimate odds ratios (ORs) for SJS/TEN for each anticonvulsant use (90 days before the index date) versus non-use. 90-day cumulative incidence of SJS/TEN per 100,000 new users was calculated for 33 anticonvulsants. Causality between anticonvulsant use and SJS/TEN in each exposed case was assessed using the algorithm of drug causality for epidermal necrolysis (ALDEN) score.

RESULTS

From 5,114,492 subjects, we selected 71 SJS/TEN cases and 284 controls. We observed significantly increased ORs for SJS/TEN among new users of carbamazepine (OR 68.00) and lamotrigine (OR 36.00) with ALDEN scores of "probable" or higher. Cumulative incidence of SJS/TEN was 93.83 for carbamazepine and 84.33 for lamotrigine. One case newly exposed to phenytoin which developed SJS/TEN was rated "unlikely" in ALDEN causality, resulting in cumulative incidence of 66.27. Cumulative incidence of SJS/TEN was 25.23 for levetiracetam, 7.52 for clonazepam, and 1.23 for diazepam, but their ALDEN scores were "very unlikely".

CONCLUSIONS

This study is the first to document the differential risk of SJS/TEN for anticonvulsants in a real-world setting in Japan. Exposure to carbamazepine and lamotrigine was associated with an increased risk of SJS/TEN.

We need to do better: A systematic review and meta-analysis of diagnostic test accuracy of restless legs syndrome screening instruments

This systematic review and meta-analysis evaluated the diagnostic accuracy of screening instruments for restless legs syndrome (RLS) and reports sensitivity, specificity, positive (PPV) and negative predictive values (NPV). Searches for primary studies were conducted in electronic databases. Of the 1541 citations identified, 52 were included in the meta-analysis. The methodological quality of each study was evaluated using QUADAS-2. Only 14 studies assessed the reference standard in all participants or in all screen-positives and a selection of screen-negatives. Bivariate meta-analysis of these 14 studies estimated median sensitivity to be 0.88 (0.72-0.96) and specificity 0.90 (0.84-0.93); based on a population prevalence of 5%, the calculated PPV was 0.31 (0.27-0.34). For all 52 studies, with either full or partial verification of RLS status, we constructed best-case scenario sensitivities and specificities at pre-defined levels of prevalence: across all samples, when prevalence is 5%, the median best-case scenario PPV is 0.48 with significant between-study heterogeneity. No RLS screening instruments can currently be recommended for use without an expert clinical interview in epidemiological studies. For conditions with statistically low prevalence such as RLS, the specificity, not the sensitivity, of a screening instrument determines true prevalence. Therefore, future instruments should maximize specificity. We provide guidelines on RLS ascertainment in epidemiological studies that requires a two-step process with clinical interview following a screening test, and given the poor reporting quality of many RLS epidemiological studies, we include an RLS reporting checklist.

Design and conduct of facility-based surveillance for severe childhood pneumonia in the Household Air Pollution Intervention Network (HAPIN) trial

Pneumonia is both a treatable and preventable disease but remains a leading cause of death in children worldwide. Household air pollution caused by burning biomass fuels for cooking has been identified as a potentially preventable risk factor for pneumonia in low- and middle-income countries. We are conducting a randomised controlled trial of a clean energy intervention in 3200 households with pregnant women living in Guatemala, India, Peru and Rwanda. Here, we describe the protocol to ascertain the incidence of severe pneumonia in infants born to participants during the first year of the study period using three independent algorithms: the presence of cough or difficulty breathing and hypoxaemia (≤92% in Guatemala, India and Rwanda and ≤86% in Peru); presence of cough or difficulty breathing along with at least one World Health Organization-defined general danger sign and consolidation on chest radiography or lung ultrasound; and pneumonia confirmed to be the cause of death by verbal autopsy. Prior to the study launch, we identified health facilities in the study areas where cases of severe pneumonia would be referred. After participant enrolment, we posted staff at each of these facilities to identify children enrolled in the trial seeking care for severe pneumonia. To ensure severe pneumonia cases are not missed, we are also conducting home visits to all households and providing education on pneumonia to the mother. Severe pneumonia reduction due to mitigation of household air pollution could be a key piece of evidence that sways policymakers to invest in liquefied petroleum gas distribution programmes.

Challenges in the diagnosis of paediatric pneumonia in intervention field trials: recommendations from a pneumonia field trial working group

Pneumonia is a leading killer of children younger than 5 years despite high vaccination coverage, improved nutrition, and widespread implementation of the Integrated Management of Childhood Illnesses algorithm. Assessing the effect of interventions on childhood pneumonia is challenging because the choice of case definition and surveillance approach can affect the identification of pneumonia substantially. In anticipation of an intervention trial aimed to reduce childhood pneumonia by lowering household air pollution, we created a working group to provide recommendations regarding study design and implementation. We suggest to, first, select a standard case definition that combines acute (≤14 days) respiratory symptoms and signs and general danger signs with ancillary tests (such as chest imaging and pulse oximetry) to improve pneumonia identification; second, to prioritise active hospital-based pneumonia surveillance over passive case finding or home-based surveillance to reduce the risk of non-differential misclassification of pneumonia and, as a result, a reduced effect size in a randomised trial; and, lastly, to consider longitudinal follow-up of children younger than 1 year, as this age group has the highest incidence of severe pneumonia.

Association of Previous Measles Infection With Markers of Acute Infectious Disease Among 9- to 59-Month-Old Children in the Democratic Republic of the Congo

BACKGROUND

Transient immunosuppression and increased susceptibility to other infections after measles infection is well known, but recent studies have suggested the occurrence of an "immune amnesia" that could have long-term immunosuppressive effects.

METHODS

We examined the association between past measles infection and acute episodes of fever, cough, and diarrhea among 2350 children aged 9 to 59 months whose mothers were selected for interview in the 2013-2014 Democratic Republic of the Congo (DRC) Demographic and Health Survey (DHS). Classification of children who had had measles was completed using maternal recall and measles immunoglobulin G serostatus obtained via dried-blood-spot analysis with a multiplex immunoassay. The association with time since measles infection and fever, cough, and diarrhea outcomes was also examined.

RESULTS

The odds of fever in the previous 2 weeks were 1.80 (95% confidence interval [CI], 1.25-2.60) among children for whom measles was reported compared to children with no history of measles. Measles vaccination demonstrated a protective association against selected clinical markers of acute infectious diseases.

CONCLUSION

Our results suggest that measles might have a long-term effect on selected clinical markers of acute infectious diseases among children aged 9 to 59 months in the DRC. These findings support the immune-amnesia hypothesis suggested by others and underscore the need for continued evaluation and improvement of the DRC's measles vaccination program.

Selection and Misclassification Biases in Longitudinal Studies

Using imperfect tests may lead to biased estimates of disease frequency and measures of association. Many studies have looked into the effect of misclassification on statistical inferences. These evaluations were either within a cross-sectional study framework, assessing biased prevalence, or for cohort study designs, evaluating biased incidence rate or risk ratio estimates based on misclassification at one of the two time-points (initial assessment or follow-up). However, both observations at risk and incident cases can be wrongly identified in longitudinal studies, leading to selection and misclassification biases, respectively. The objective of this paper was to evaluate the relative impact of selection and misclassification biases resulting from misclassification, together, on measures of incidence and risk ratio. To investigate impact on measure of disease frequency, data sets from a hypothetical cohort study with two samples collected one month apart were simulated and analyzed based on specific test and disease characteristics, with no elimination of disease during the sampling interval or clustering of observations. Direction and magnitude of bias due to selection, misclassification, and total bias was assessed for diagnostic test sensitivity and specificity ranging from 0.7 to 1.0 and 0.8 to 1.0, respectively, and for specific disease contexts, i.e., disease prevalences of 5 and 20%, and disease incidences of 0.01, 0.05, and 0.1 cases/animal-month. A hypothetical exposure with known strength of association was also generated. A total of 1,000 cohort studies of 1,000 observations each were simulated for these six disease contexts where the same diagnostic test was used to identify observations at risk at beginning of the cohort and incident cases at its end. Our results indicated that the departure of the estimates of disease incidence and risk ratio from their true value were mainly a function of test specificity, and disease prevalence and incidence. The combination of the two biases, at baseline and follow-up, revealed the importance of a good to excellent specificity relative to sensitivity for the diagnostic test. Small divergence from perfect specificity extended quickly to disease incidence over-estimation as true prevalence increased and true incidence decreased. A highly sensitive test to exclude diseased subjects at baseline was of less importance to minimize bias than using a highly specific one at baseline. Near perfect diagnostic test attributes were even more important to obtain a measure of association close to the true risk ratio, according to specific disease characteristics, especially its prevalence. Low prevalent and high incident disease lead to minimal bias if disease is diagnosed with high sensitivity and close to perfect specificity at baseline and follow-up. For more prevalent diseases we observed large risk ratio biases towards the null value, even with near perfect diagnosis.

From menarche to menopause: A population-based assessment of water, sanitation, and hygiene risk factors for reproductive tract infection symptoms over life stages in rural girls and women in India

Women face greater challenges than men in accessing water, sanitation, and hygiene (WASH) resources to address their daily needs, and may respond to these challenges by adopting unsafe practices that increase the risk of reproductive tract infections (RTIs). WASH practices may change as women transition through socially-defined life stage experiences, like marriage and pregnancy. Thus, the relationship between WASH practices and RTIs might vary across female reproductive life stages. This cross-sectional study assessed the relationship between WASH exposures and self-reported RTI symptoms in 3,952 girls and women from two rural districts in India, and tested whether social exposures represented by reproductive life stage was an effect modifier of associations. In fully adjusted models, RTI symptoms were less common in women using a latrine without water for defecation versus open defecation (Odds Ratio (OR) = 0.69; Confidence Interval (CI) = 0.48, 0.98) and those walking shorter distances to a bathing location (OR = 0.79, CI = 0.63, 0.99), but there was no association between using a latrine with a water source and RTIs versus open defecation (OR = 1.09; CI = 0.69, 1.72). Unexpectedly, RTI symptoms were more common for women bathing daily with soap (OR = 6.55, CI = 3.60, 11.94) and for women washing their hands after defecation with soap (OR = 10.27; CI = 5.53, 19.08) or ash/soil/mud (OR = 6.02; CI = 3.07, 11.77) versus water only or no hand washing. WASH practices of girls and women varied across reproductive life stages, but the associations between WASH practices and RTI symptoms were not moderated by or confounded by life stage status. This study provides new evidence that WASH access and practices are associated with self-reported reproductive tract infection symptoms in rural Indian girls and women from different reproductive life stages. However, the counterintuitive directions of effect for soap use highlights that causality and mechanisms of effect cannot be inferred from this study design. Future research is needed to understand whether improvements in water and sanitation access could improve the practice of safe hygiene behaviors and reduce the global burden of RTIs in women.

Corrected ROC analysis for misclassified binary outcomes

Creating accurate risk prediction models from Big Data resources such as Electronic Health Records (EHRs) is a critical step toward achieving precision medicine. A major challenge in developing these tools is accounting for imperfect aspects of EHR data, particularly the potential for misclassified outcomes. Misclassification, the swapping of case and control outcome labels, is well known to bias effect size estimates for regression prediction models. In this paper, we study the effect of misclassification on accuracy assessment for risk prediction models and find that it leads to bias in the area under the curve (AUC) metric from standard ROC analysis. The extent of the bias is determined by the false positive and false negative misclassification rates as well as disease prevalence. Notably, we show that simply correcting for misclassification while building the prediction model is not sufficient to remove the bias in AUC. We therefore introduce an intuitive misclassification-adjusted ROC procedure that accounts for uncertainty in observed outcomes and produces bias-corrected estimates of the true AUC. The method requires that misclassification rates are either known or can be estimated, quantities typically required for the modeling step. The computational simplicity of our method is a key advantage, making it ideal for efficiently comparing multiple prediction models on very large datasets. Finally, we apply the correction method to a hospitalization prediction model from a cohort of over 1 million patients from the Veterans Health Administrations EHR. Implementations of the ROC correction are provided for Stata and R. Published 2017. This article is a U.S. Government work and is in the public domain in the USA.

Bias in logistic regression due to imperfect diagnostic test results and practical correction approaches

Background

Logistic regression is a statistical model widely used in cross-sectional and cohort studies to identify and quantify the effects of potential disease risk factors. However, the impact of imperfect tests on adjusted odds ratios (and thus on the identification of risk factors) is under-appreciated. The purpose of this article is to draw attention to the problem associated with modelling imperfect diagnostic tests, and propose simple Bayesian models to adequately address this issue.

Methods

A systematic literature review was conducted to determine the proportion of malaria studies that appropriately accounted for false-negatives/false-positives in a logistic regression setting. Inference from the standard logistic regression was also compared with that from three proposed Bayesian models using simulations and malaria data from the western Brazilian Amazon.

Results

A systematic literature review suggests that malaria epidemiologists are largely unaware of the problem of using logistic regression to model imperfect diagnostic test results. Simulation results reveal that statistical inference can be substantially improved when using the proposed Bayesian models versus the standard logistic regression. Finally, analysis of original malaria data with one of the proposed Bayesian models reveals that microscopy sensitivity is strongly influenced by how long people have lived in the study region, and an important risk factor (i.e., participation in forest extractivism) is identified that would have been missed by standard logistic regression.

Conclusion

Given the numerous diagnostic methods employed by malaria researchers and the ubiquitous use of logistic regression to model the results of these diagnostic tests, this paper provides critical guidelines to improve data analysis practice in the presence of misclassification error. Easy-to-use code that can be readily adapted to WinBUGS is provided, enabling straightforward implementation of the proposed Bayesian models.

Electronic supplementary material

The online version of this article (doi:10.1186/s12936-015-0966-y) contains supplementary material, which is available to authorized users.

Effects of changing exposure to neighbourhood greenness on general and mental health: A longitudinal study

Green neighbourhood environments have been associated with physical and psychological wellbeing in adults. Access to greenness is potentially more important in vulnerable subgroups. In this study based on longitudinal survey data from southern Sweden the cohort was divided into prognostic groups for good self-reported general (n=8891) and mental (n=9444) health. We used independent survey data to assess perceived neighbourhood greenness in 1km(2) areas, and estimated effects of changing exposure longitudinally stratified by prognostic group. The overall effect on health was small and statistically uncertain (for general health OR 1.04, 95% CI 0.98-1.10, for mental health OR 1.07, 95% CI 1.00-1.14). A more beneficial effect of increased greenness was indicated among subjects with lowest prognostic of good general health (OR 1.24, 95% CI 1.01-1.52). The study provided only weak evidence for beneficial effects of increased neighbourhood greenness triggered by changing residence. It seems that altered life circumstances, e.g. changed civil or socioeconomic status that often trigger a decision to move, are also the key determinants of the health consequences of changing residence.

Problem drinking as a risk factor for tuberculosis: a propensity score matched analysis of a national survey

Background

Epidemiological and other evidence strongly supports the hypothesis that problem drinking is causally related to the incidence of active tuberculosis and the worsening of the disease course. The presence of a large number of potential confounders, however, complicates the assessment of the actual size of this causal effect, leaving room for a substantial amount of bias. This study aims to contribute to the understanding of the role of confounding in the observed association between problem drinking and tuberculosis, assessing the effect of the adjustment for a relatively large number of potential confounders on the estimated prevalence odds ratio of tuberculosis among problem drinkers vs. moderate drinkers/abstainers in a cross-sectional, nationally representative sample of the South African adult population.

Methods

A propensity score approach was used to match each problem drinker in the sample with a subset of moderate drinkers/abstainers with similar characteristics in respect to a set of potential confounders. The prevalence odds ratio of tuberculosis between the matched groups was then calculated using conditional logistic regression. Sensitivity analyses were conducted to assess the robustness of the results in respect to misspecification of the model.

Results

The prevalence odds ratio of tuberculosis between problem drinkers and moderate drinkers/abstainers was 1.97 (95% CI: 1.40 to 2.77), and the result was robust with respect to the matching procedure as well as to incorrect adjustment for potential mediators and to the possible presence of unmeasured confounders. Sub-population analysis did not provide noteworthy evidence for the presence of interaction between problem drinking and the observed confounders.

Conclusion

In a cross-sectional national survey of the adult population of a middle income country with high tuberculosis burden, problem drinking was associated with a two fold increase in the odds of past TB diagnosis after controlling for a large number of socio-economic and biological confounders. Within the limitations of a cross-sectional study design with self-reported tuberculosis status, these results adds to previous evidence of a causal link between problem drinking and tuberculosis, and suggest that the observed higher prevalence of tuberculosis among problem drinkers commonly found in population studies cannot be attributed to the confounding effect of the uneven distribution of other risk factors.

Accuracy of International Classification of Diseases, 9th Revision, Clinical Modification codes for upper gastrointestinal complications varied by position and age: a validation study in a cohort of nonsteroidal anti-inflammatory drugs users in Friuli Venezia Giulia, Italy

PURPOSE

To validate the International Classification of Diseases, 9th Revision, Clinical Modification discharge codes used to identify cases of upper gastrointestinal complications (UGICs) in hospitals of Friuli Venezia Giulia, Italy.

METHODS

Cohort study on the risk of UGIC in users of nonsteroidal anti-inflammatory drugs conducted in Friuli Venezia Giulia between 2001 and 2008. Cases were identified through primary and secondary International Classification of Diseases, 9th Revision Clinical specific codes 531 (gastric ulcer), 532 (duodenal ulcer), 533 (peptic ulcer), 534 (gastrojejunal ulcer), and nonspecific code 578 (gastrointestinal hemorrhage). Potential cases were confirmed through hospital chart review.

RESULTS

The chart retrieval percentage was 98.4%.The positive predictive value (PPV) was 94.3% for primary codes 531 and 532, 79.5% for code 533, 83.1% for code 534, 40.2% for code 578. The PPV for secondary codes was 34.7% but increased to 88.9% and 79.2% when the primary code was for peritonitis or acute post-hemorrhagic anemia, respectively. Validation of secondary codes increased case ascertainment by 4.9%. Endoscopy confirmed 79.4% of cases but only 67.2% of those above age 84 years.

CONCLUSIONS

The PPV was high for specific primary codes and moderate to low for nonspecific primary and secondary codes. The inclusion of confirmed cases identified by nonspecific and secondary codes can be of value in studies that need a complete ascertainment of cases occurring in the study population. In this cohort, not including these cases would underestimate the incidence of UGICs. A potential for case misclassification exists in particular in eldest ages.