-
Journal of Clinical Epidemiology Feb 2022Directed acyclic graphs (DAGs) are an intuitive yet rigorous tool to communicate about causal questions in clinical and epidemiologic research and inform study design...
Directed acyclic graphs (DAGs) are an intuitive yet rigorous tool to communicate about causal questions in clinical and epidemiologic research and inform study design and statistical analysis. DAGs are constructed to depict prior knowledge about biological and behavioral systems related to specific causal research questions. DAG components portray who receives treatment or experiences exposures; mechanisms by which treatments and exposures operate; and other factors that influence the outcome of interest or which persons are included in an analysis. Once assembled, DAGs - via a few simple rules - guide the researcher in identifying whether the causal effect of interest can be identified without bias and, if so, what must be done either in study design or data analysis to achieve this. Specifically, DAGs can identify variables that, if controlled for in the design or analysis phase, are sufficient to eliminate confounding and some forms of selection bias. DAGs also help recognize variables that, if controlled for, bias the analysis (e.g., mediators or factors influenced by both exposure and outcome). Finally, DAGs help researchers recognize insidious sources of bias introduced by selection of individuals into studies or failure to completely observe all individuals until study outcomes are reached. DAGs, however, are not infallible, largely owing to limitations in prior knowledge about the system in question. In such instances, several alternative DAGs are plausible, and researchers should assess whether results differ meaningfully across analyses guided by different DAGs and be forthright about uncertainty. DAGs are powerful tools to guide the conduct of clinical research.
Topics: Bias; Causality; Confounding Factors, Epidemiologic; Data Interpretation, Statistical; Humans; Selection Bias
PubMed: 34371103
DOI: 10.1016/j.jclinepi.2021.08.001 -
Epidemiology (Cambridge, Mass.) Sep 2022Selection bias remains a subject of controversy. Existing definitions of selection bias are ambiguous. To improve communication and the conduct of epidemiologic research...
Selection bias remains a subject of controversy. Existing definitions of selection bias are ambiguous. To improve communication and the conduct of epidemiologic research focused on estimating causal effects, we propose to unify the various existing definitions of selection bias in the literature by considering any bias away from the true causal effect in the referent population (the population before the selection process), due to selecting the sample from the referent population, as selection bias. Given this unified definition, selection bias can be further categorized into two broad types: type 1 selection bias owing to restricting to one or more level(s) of a collider (or a descendant of a collider) and type 2 selection bias owing to restricting to one or more level(s) of an effect measure modifier. To aid in explaining these two types-which can co-occur-we start by reviewing the concepts of the target population, the study sample, and the analytic sample. Then, we illustrate both types of selection bias using causal diagrams. In addition, we explore the differences between these two types of selection bias, and describe methods to minimize selection bias. Finally, we use an example of "M-bias" to demonstrate the advantage of classifying selection bias into these two types.
Topics: Bias; Causality; Humans; Selection Bias
PubMed: 35700187
DOI: 10.1097/EDE.0000000000001516 -
Journal of Assisted Reproduction and... Jul 2021Embryo selection within in vitro fertilization (IVF) is the process of evaluating qualities of fertilized oocytes (embryos) and selecting the best embryo(s) available...
Embryo selection within in vitro fertilization (IVF) is the process of evaluating qualities of fertilized oocytes (embryos) and selecting the best embryo(s) available within a patient cohort for subsequent transfer or cryopreservation. In recent years, artificial intelligence (AI) has been used extensively to improve and automate the embryo ranking and selection procedure by extracting relevant information from embryo microscopy images. The AI models are evaluated based on their ability to identify the embryo(s) with the highest chance(s) of achieving a successful pregnancy. Whether such evaluations should be based on ranking performance or pregnancy prediction, however, seems to divide studies. As such, a variety of performance metrics are reported, and comparisons between studies are often made on different outcomes and data foundations. Moreover, superiority of AI methods over manual human evaluation is often claimed based on retrospective data, without any mentions of potential bias. In this paper, we provide a technical view on some of the major topics that divide how current AI models are trained, evaluated and compared. We explain and discuss the most common evaluation metrics and relate them to the two separate evaluation objectives, ranking and prediction. We also discuss when and how to compare AI models across studies and explain in detail how a selection bias is inevitable when comparing AI models against current embryo selection practice in retrospective cohort studies.
Topics: Area Under Curve; Artificial Intelligence; Blastocyst; Calibration; Cryopreservation; Databases, Factual; Decision Making, Computer-Assisted; Embryo Transfer; Female; Fertilization in Vitro; Humans; Image Processing, Computer-Assisted; Pregnancy; Sample Size; Sensitivity and Specificity
PubMed: 34173914
DOI: 10.1007/s10815-021-02254-6 -
The Cochrane Database of Systematic... Oct 2020Respiratory distress, particularly respiratory distress syndrome (RDS), is the single most important cause of morbidity and mortality in preterm infants. In infants with... (Meta-Analysis)
Meta-Analysis
BACKGROUND
Respiratory distress, particularly respiratory distress syndrome (RDS), is the single most important cause of morbidity and mortality in preterm infants. In infants with progressive respiratory insufficiency, intermittent positive pressure ventilation (IPPV) with surfactant has been the usual treatment, but it is invasive, potentially resulting in airway and lung injury. Continuous positive airway pressure (CPAP) has been used for the prevention and treatment of respiratory distress, as well as for the prevention of apnoea, and in weaning from IPPV. Its use in the treatment of RDS might reduce the need for IPPV and its sequelae.
OBJECTIVES
To determine the effect of continuous distending pressure in the form of CPAP on the need for IPPV and associated morbidity in spontaneously breathing preterm infants with respiratory distress.
SEARCH METHODS
We used the standard strategy of Cochrane Neonatal to search CENTRAL (2020, Issue 6); Ovid MEDLINE and Epub Ahead of Print, In-Process & Other Non-Indexed Citations, Daily and Versions; and CINAHL on 30 June 2020. We also searched clinical trials databases and the reference lists of retrieved articles for randomised controlled trials and quasi-randomised trials.
SELECTION CRITERIA
All randomised or quasi-randomised trials of preterm infants with respiratory distress were eligible. Interventions were CPAP by mask, nasal prong, nasopharyngeal tube or endotracheal tube, compared with spontaneous breathing with supplemental oxygen as necessary.
DATA COLLECTION AND ANALYSIS
We used standard methods of Cochrane and its Neonatal Review Group, including independent assessment of risk of bias and extraction of data by two review authors. We used the GRADE approach to assess the certainty of evidence. Subgroup analyses were planned on the basis of birth weight (greater than or less than 1000 g or 1500 g), gestational age (groups divided at about 28 weeks and 32 weeks), timing of application (early versus late in the course of respiratory distress), pressure applied (high versus low) and trial setting (tertiary compared with non-tertiary hospitals; high income compared with low income) MAIN RESULTS: We included five studies involving 322 infants; two studies used face mask CPAP, two studies used nasal CPAP and one study used endotracheal CPAP and continuing negative pressure for a small number of less ill babies. For this update, we included one new trial. CPAP was associated with lower risk of treatment failure (death or use of assisted ventilation) (typical risk ratio (RR) 0.64, 95% confidence interval (CI) 0.50 to 0.82; typical risk difference (RD) -0.19, 95% CI -0.28 to -0.09; number needed to treat for an additional beneficial outcome (NNTB) 6, 95% CI 4 to 11; I = 50%; 5 studies, 322 infants; very low-certainty evidence), lower use of ventilatory assistance (typical RR 0.72, 95% CI 0.54 to 0.96; typical RD -0.13, 95% CI -0.25 to -0.02; NNTB 8, 95% CI 4 to 50; I = 55%; very low-certainty evidence) and lower overall mortality (typical RR 0.53, 95% CI 0.34 to 0.83; typical RD -0.11, 95% CI -0.18 to -0.04; NNTB 9, 95% CI 2 to 13; I = 0%; 5 studies, 322 infants; moderate-certainty evidence). CPAP was associated with increased risk of pneumothorax (typical RR 2.48, 95% CI 1.16 to 5.30; typical RD 0.09, 95% CI 0.02 to 0.16; number needed to treat for an additional harmful outcome (NNTH) 11, 95% CI 7 to 50; I = 0%; 4 studies, 274 infants; low-certainty evidence). There was no evidence of a difference in bronchopulmonary dysplasia, defined as oxygen dependency at 28 days (RR 1.04, 95% CI 0.35 to 3.13; I = 0%; 2 studies, 209 infants; very low-certainty evidence). The trials did not report use of surfactant, intraventricular haemorrhage, retinopathy of prematurity, necrotising enterocolitis and neurodevelopment outcomes in childhood.
AUTHORS' CONCLUSIONS
In preterm infants with respiratory distress, the application of CPAP is associated with reduced respiratory failure, use of mechanical ventilation and mortality and an increased rate of pneumothorax compared to spontaneous breathing with supplemental oxygen as necessary. Three out of five of these trials were conducted in the 1970s. Therefore, the applicability of these results to current practice is unclear. Further studies in resource-poor settings should be considered and research to determine the most appropriate pressure level needs to be considered.
Topics: Bronchopulmonary Dysplasia; Continuous Positive Airway Pressure; Humans; Infant, Low Birth Weight; Infant, Newborn; Infant, Premature; Intermittent Positive-Pressure Ventilation; Outcome Assessment, Health Care; Pneumothorax; Pulmonary Surfactants; Randomized Controlled Trials as Topic; Respiratory Distress Syndrome, Newborn; Respiratory Insufficiency; Selection Bias; Treatment Failure
PubMed: 33058208
DOI: 10.1002/14651858.CD002271.pub3 -
Fertility and Sterility Dec 2020
Topics: COVID-19; Humans; Pandemics; Parents; SARS-CoV-2; Selection Bias
PubMed: 33280724
DOI: 10.1016/j.fertnstert.2020.10.057 -
Nephrology (Carlton, Vic.) Jun 2020Study quality depends on a number of factors, one of them being internal validity. Such validity can be affected by random and systematic error, the latter also known as... (Review)
Review
Study quality depends on a number of factors, one of them being internal validity. Such validity can be affected by random and systematic error, the latter also known as bias. Both make it more difficult to assess a correct frequency or the true relationship between exposure and outcome. Where random error can be addressed by increasing the sample size, a systematic error in the design, the conduct or the reporting of a study is more problematic. In this article, we will focus on bias, discuss different types of selection bias (sampling bias, confounding by indication, incidence-prevalence bias, attrition bias, collider stratification bias and publication bias) and information bias (recall bias, interviewer bias, observer bias and lead-time bias), indicate the type of studies where they most frequently occur and provide suggestions for their prevention.
Topics: Biomedical Research; Humans; Interviews as Topic; Observer Variation; Research Design; Selection Bias; Self Report
PubMed: 32133725
DOI: 10.1111/nep.13706 -
Epidemiology (Cambridge, Mass.) Sep 2021Confounding, selection bias, and measurement error are well-known sources of bias in epidemiologic research. Methods for assessing these biases have their own...
Confounding, selection bias, and measurement error are well-known sources of bias in epidemiologic research. Methods for assessing these biases have their own limitations. Many quantitative sensitivity analysis approaches consider each type of bias individually, although more complex approaches are harder to implement or require numerous assumptions. By failing to consider multiple biases at once, researchers can underestimate-or overestimate-their joint impact. We show that it is possible to bound the total composite bias owing to these three sources and to use that bound to assess the sensitivity of a risk ratio to any combination of these biases. We derive bounds for the total composite bias under a variety of scenarios, providing researchers with tools to assess their total potential impact. We apply this technique to a study where unmeasured confounding and selection bias are both concerns and to another study in which possible differential exposure misclassification and confounding are concerns. The approach we describe, though conservative, is easier to implement and makes simpler assumptions than quantitative bias analysis. We provide R functions to aid implementation.
Topics: Bias; Confounding Factors, Epidemiologic; Epidemiologic Studies; Humans; Research Design; Selection Bias
PubMed: 34224471
DOI: 10.1097/EDE.0000000000001380 -
JAMA Jun 2021Continuous glucose monitoring (CGM) is recommended for patients with type 1 diabetes; observational evidence for CGM in patients with insulin-treated type 2 diabetes is... (Observational Study)
Observational Study
IMPORTANCE
Continuous glucose monitoring (CGM) is recommended for patients with type 1 diabetes; observational evidence for CGM in patients with insulin-treated type 2 diabetes is lacking.
OBJECTIVE
To estimate clinical outcomes of real-time CGM initiation.
DESIGN, SETTING, AND PARTICIPANTS
Exploratory retrospective cohort study of changes in outcomes associated with real-time CGM initiation, estimated using a difference-in-differences analysis. A total of 41 753 participants with insulin-treated diabetes (5673 type 1; 36 080 type 2) receiving care from a Northern California integrated health care delivery system (2014-2019), being treated with insulin, self-monitoring their blood glucose levels, and having no prior CGM use were included.
EXPOSURES
Initiation vs noninitiation of real-time CGM (reference group).
MAIN OUTCOMES AND MEASURES
Ten end points measured during the 12 months before and 12 months after baseline: hemoglobin A1c (HbA1c); hypoglycemia (emergency department or hospital utilization); hyperglycemia (emergency department or hospital utilization); HbA1c levels lower than 7%, lower than 8%, and higher than 9%; 1 emergency department encounter or more for any reason; 1 hospitalization or more for any reason; and number of outpatient visits and telephone visits.
RESULTS
The real-time CGM initiators included 3806 patients (mean age, 42.4 years [SD, 19.9 years]; 51% female; 91% type 1, 9% type 2); the noninitiators included 37 947 patients (mean age, 63.4 years [SD, 13.4 years]; 49% female; 6% type 1, 94% type 2). The prebaseline mean HbA1c was lower among real-time CGM initiators than among noninitiators, but real-time CGM initiators had higher prebaseline rates of hypoglycemia and hyperglycemia. Mean HbA1c declined among real-time CGM initiators from 8.17% to 7.76% and from 8.28% to 8.19% among noninitiators (adjusted difference-in-differences estimate, -0.40%; 95% CI, -0.48% to -0.32%; P < .001). Hypoglycemia rates declined among real-time CGM initiators from 5.1% to 3.0% and increased among noninitiators from 1.9% to 2.3% (difference-in-differences estimate, -2.7%; 95% CI, -4.4% to -1.1%; P = .001). There were also statistically significant differences in the adjusted net changes in the proportion of patients with HbA1c lower than 7% (adjusted difference-in-differences estimate, 9.6%; 95% CI, 7.1% to 12.2%; P < .001), lower than 8% (adjusted difference-in-differences estimate, 13.1%; 95% CI, 10.2% to 16.1%; P < .001), and higher than 9% (adjusted difference-in-differences estimate, -7.1%; 95% CI, -9.5% to -4.6%; P < .001) and in the number of outpatient visits (adjusted difference-in-differences estimate, -0.4; 95% CI, -0.6 to -0.2; P < .001) and telephone visits (adjusted difference-in-differences estimate, 1.1; 95% CI, 0.8 to 1.4; P < .001). Initiation of real-time CGM was not associated with statistically significant changes in rates of hyperglycemia, emergency department visits for any reason, or hospitalizations for any reason.
CONCLUSIONS AND RELEVANCE
In this retrospective cohort study, insulin-treated patients with diabetes selected by physicians for real-time continuous glucose monitoring compared with noninitiators had significant improvements in hemoglobin A1c and reductions in emergency department visits and hospitalizations for hypoglycemia, but no significant change in emergency department visits or hospitalizations for hyperglycemia or for any reason. Because of the observational study design, findings may have been susceptible to selection bias.
Topics: Adult; Biosensing Techniques; Blood Glucose Self-Monitoring; Confidence Intervals; Diabetes Mellitus, Type 1; Diabetes Mellitus, Type 2; Female; Glycated Hemoglobin; Health Services Needs and Demand; Hospitalization; Humans; Hyperglycemia; Hypoglycemia; Hypoglycemic Agents; Insulin; Male; Middle Aged; Numbers Needed To Treat; Propensity Score; Retrospective Studies; Selection Bias; Time Factors; Treatment Outcome
PubMed: 34077502
DOI: 10.1001/jama.2021.6530 -
JCPP Advances Oct 2021Sophisticated statistical analyses of data from large-scale prospective birth cohort studies combined with thoughtful study designs have advanced understanding about the...
Sophisticated statistical analyses of data from large-scale prospective birth cohort studies combined with thoughtful study designs have advanced understanding about the causes, consequences and developmental course of child and adolescent mental health problems. Available large-scale prospective cohort studies, such as ALSPAC, MoBA, and TEDS have many noteworthy strengths, but they all suffer from non-random non-participation and attrition over time. Recent findings have highlighted that prospective birth cohort studies need to carefully consider the importance of selection bias.
PubMed: 37431437
DOI: 10.1002/jcv2.12043 -
The Cochrane Database of Systematic... Jun 2019Frequent consumption of excess amounts of sugar-sweetened beverages (SSB) is a risk factor for obesity, type 2 diabetes, cardiovascular disease and dental caries....
BACKGROUND
Frequent consumption of excess amounts of sugar-sweetened beverages (SSB) is a risk factor for obesity, type 2 diabetes, cardiovascular disease and dental caries. Environmental interventions, i.e. interventions that alter the physical or social environment in which individuals make beverage choices, have been advocated as a means to reduce the consumption of SSB.
OBJECTIVES
To assess the effects of environmental interventions (excluding taxation) on the consumption of sugar-sweetened beverages and sugar-sweetened milk, diet-related anthropometric measures and health outcomes, and on any reported unintended consequences or adverse outcomes.
SEARCH METHODS
We searched 11 general, specialist and regional databases from inception to 24 January 2018. We also searched trial registers, reference lists and citations, scanned websites of relevant organisations, and contacted study authors.
SELECTION CRITERIA
We included studies on interventions implemented at an environmental level, reporting effects on direct or indirect measures of SSB intake, diet-related anthropometric measures and health outcomes, or any reported adverse outcome. We included randomised controlled trials (RCTs), non-randomised controlled trials (NRCTs), controlled before-after (CBA) and interrupted-time-series (ITS) studies, implemented in real-world settings with a combined length of intervention and follow-up of at least 12 weeks and at least 20 individuals in each of the intervention and control groups. We excluded studies in which participants were administered SSB as part of clinical trials, and multicomponent interventions which did not report SSB-specific outcome data. We excluded studies on the taxation of SSB, as these are the subject of a separate Cochrane Review.
DATA COLLECTION AND ANALYSIS
Two review authors independently screened studies for inclusion, extracted data and assessed the risks of bias of included studies. We classified interventions according to the NOURISHING framework, and synthesised results narratively and conducted meta-analyses for two outcomes relating to two intervention types. We assessed our confidence in the certainty of effect estimates with the GRADE framework as very low, low, moderate or high, and presented 'Summary of findings' tables.
MAIN RESULTS
We identified 14,488 unique records, and assessed 1030 in full text for eligibility. We found 58 studies meeting our inclusion criteria, including 22 RCTs, 3 NRCTs, 14 CBA studies, and 19 ITS studies, with a total of 1,180,096 participants. The median length of follow-up was 10 months. The studies included children, teenagers and adults, and were implemented in a variety of settings, including schools, retailing and food service establishments. We judged most studies to be at high or unclear risk of bias in at least one domain, and most studies used non-randomised designs. The studies examine a broad range of interventions, and we present results for these separately.Labelling interventions (8 studies): We found moderate-certainty evidence that traffic-light labelling is associated with decreasing sales of SSBs, and low-certainty evidence that nutritional rating score labelling is associated with decreasing sales of SSBs. For menu-board calorie labelling reported effects on SSB sales varied.Nutrition standards in public institutions (16 studies): We found low-certainty evidence that reduced availability of SSBs in schools is associated with decreased SSB consumption. We found very low-certainty evidence that improved availability of drinking water in schools and school fruit programmes are associated with decreased SSB consumption. Reported associations between improved availability of drinking water in schools and student body weight varied.Economic tools (7 studies): We found moderate-certainty evidence that price increases on SSBs are associated with decreasing SSB sales. For price discounts on low-calorie beverages reported effects on SSB sales varied.Whole food supply interventions (3 studies): Reported associations between voluntary industry initiatives to improve the whole food supply and SSB sales varied.Retail and food service interventions (7 studies): We found low-certainty evidence that healthier default beverages in children's menus in chain restaurants are associated with decreasing SSB sales, and moderate-certainty evidence that in-store promotion of healthier beverages in supermarkets is associated with decreasing SSB sales. We found very low-certainty evidence that urban planning restrictions on new fast-food restaurants and restrictions on the number of stores selling SSBs in remote communities are associated with decreasing SSB sales. Reported associations between promotion of healthier beverages in vending machines and SSB intake or sales varied.Intersectoral approaches (8 studies): We found moderate-certainty evidence that government food benefit programmes with restrictions on purchasing SSBs are associated with decreased SSB intake. For unrestricted food benefit programmes reported effects varied. We found moderate-certainty evidence that multicomponent community campaigns focused on SSBs are associated with decreasing SSB sales. Reported associations between trade and investment liberalisation and SSB sales varied.Home-based interventions (7 studies): We found moderate-certainty evidence that improved availability of low-calorie beverages in the home environment is associated with decreased SSB intake, and high-certainty evidence that it is associated with decreased body weight among adolescents with overweight or obesity and a high baseline consumption of SSBs.Adverse outcomes reported by studies, which may occur in some circumstances, included negative effects on revenue, compensatory SSB consumption outside school when the availability of SSBs in schools is reduced, reduced milk intake, stakeholder discontent, and increased total energy content of grocery purchases with price discounts on low-calorie beverages, among others. The certainty of evidence on adverse outcomes was low to very low for most outcomes.We analysed interventions targeting sugar-sweetened milk separately, and found low- to moderate-certainty evidence that emoticon labelling and small prizes for the selection of healthier beverages in elementary school cafeterias are associated with decreased consumption of sugar-sweetened milk. We found low-certainty evidence that improved placement of plain milk in school cafeterias is not associated with decreasing sugar-sweetened milk consumption.
AUTHORS' CONCLUSIONS
The evidence included in this review indicates that effective, scalable interventions addressing SSB consumption at a population level exist. Implementation should be accompanied by high-quality evaluations using appropriate study designs, with a particular focus on the long-term effects of approaches suitable for large-scale implementation.
Topics: Adolescent; Adult; Animals; Artificially Sweetened Beverages; Child; Commerce; Controlled Before-After Studies; Drinking Behavior; Drinking Water; Environment; Fast Foods; Food Supply; Fruit; Humans; Interrupted Time Series Analysis; Milk; Nutritive Value; Product Labeling; Randomized Controlled Trials as Topic; Schools; Selection Bias; Social Environment; Sugar-Sweetened Beverages; Young Adult
PubMed: 31194900
DOI: 10.1002/14651858.CD012292.pub2