scholarly journals Measuring the sensitivity of difference-in-difference estimates to the parallel trends assumption

2021 ◽  
pp. 263208432110613
Author(s):  
Landon Gibson ◽  
Frederick Zimmerman
Keyword(s):  
Sensitivity Analysis ◽  
Treatment Period ◽  
Treatment Effect ◽  
Statistical Tests ◽  
Sensitivity Analyses ◽  
Difference In Difference ◽  
Control Groups ◽  
Pre Treatment ◽  
Background Difference ◽  
And Control

Background. Difference-in-Difference makes a critical assumption that the changes in the outcomes, over the post-treatment period, are similar between the treated and control groups—the parallel trends assumption. Evaluation of this assumption is often done either by graphical examination or by statistical tests in the pre-treatment period. They result in a binary conclusion about the validity of the assumption. Purpose. This paper proposes a sensitivity analysis that quantifies the departure from parallel trends necessary to meaningfully change the estimated treatment effect. Results. Sensitivity analyses have an advantage over traditional parallel trends tests: they use all available data and thereby work even if only one pre-period is available, and they quantify the strength of unobserved confounder(s) required to change the conclusions of a study. Conclusions. We apply the sensitivity analysis metrics developed by Cinelli and Hazlett (2020) and illustrate them on two studies.

Indicated preventive interventions for anxiety in children and adolescents: a review and meta-analysis of school based programs

2019 ◽  
Author(s):  
Siobhan Hugh-Jones ◽  
Sophie Beckett ◽  
Pavan Mallikarjun
Keyword(s):  
Meta Analysis ◽  
Child And Adolescent ◽  
Sensitivity Analyses ◽  
Cochrane Library ◽  
Control Groups ◽  
Adolescent Anxiety ◽  
School Based ◽  
Intervention Intensity ◽  
Randomised Controlled ◽  
And Control

Schools are promising sites for the delivery of prevention and early intervention programs to reduce child and adolescent anxiety. It is unclear whether universal or targeted approaches are most effective. This review and meta-analysis examines the effectiveness of school-based indicated interventions and was registered with PROSPERO [CRD42018087628].MEDLINE, EMBASE, PsycINFO and the Cochrane Library were searched for randomised controlled trials comparing indicated school programs for child and adolescent anxiety to active or inactive control groups. Twenty original studies, with 2076 participants, met the inclusion criteria and 18 were suitable for meta-analysis. Sub-group and sensitivity analyses explored intervention intensity, delivery agent and control type. A small beneficial effect was found for indicated programs compared to controls on self-reported anxiety symptoms at post-test (g = -0.28, CI = -0.50, -0.05, k= 18). The small effect was maintained at 6 (g = -0.35, CI= -0.58, -0.13, k = 9) and 12 months (g = -0.24, CI = -0.48, 0.00, k = 4). Based on two studies, >12 month effects were very small (g = -0.01, CI= -0.38, 0.36). No differences were found based on intervention intensity, delivery agent and control type. There was evidence of publication bias and a relatively high risk of contamination in studies. Findings support the value of school based indicated programs for child and adolescent anxiety. Effects at 12 months outperform many universal programs. High quality, randomised controlled and pragmatic trials are needed, with attention control groups and beyond 12 month diagnostic assessments are needed.

scholarly journals In Vivo Study of the Action of a Topical Anti-Inflammatory Drug In Rat Teeth Submitted To Dental Bleaching

2018 ◽  
Vol 29 (6) ◽  
pp. 555-561 ◽  
Author(s):  
Francine Benetti ◽  
André Luiz Fraga Briso ◽  
Luciana Louzada Ferreira ◽  
Marina Carminatti ◽  
Larissa Álamo ◽  
...  
Keyword(s):  
Statistical Tests ◽  
Control Groups ◽  
Pulp Tissue ◽  
Necrosis Factor Alpha ◽  
Dental Bleaching ◽  
Tnf Α ◽  
Pulp Inflammation ◽  
Significant Difference ◽  
Anti Inflammatory ◽  
And Control

Abstract Bleaching gel containing hydrogen peroxide (H2O2) cause damages in pulp tissue. This study investigated the action of a topical anti-inflammatory, the Otosporin®, in rats’ bleached teeth with the null hypothesis of which the Otosporin® is no able to minimize the pulp inflammation that bleaching gel generates. The rat’s molars were divided into groups: BLE: bleached (35% H2O2 concentration /single application of 30 min); BLE-O: bleached followed by Otosporin® (10 min); and control: placebo gel. In the second day after dental bleaching, the rats were killed, and the jaws were processed for hematoxylin-eosin and immunohistochemistry analysis for tumor necrosis factor alpha (TNF-α), interleukin (IL)-6 and IL-17. The data collected were subjected to Kruskal-Wallis and Dunn statistical tests with at a 5% level of significance (p<0.05). The BLE group had moderate to strong inflammation in the occlusal third of the coronary pulp, with necrotic areas; and BLE-O, mild inflammation (p<0.05). There was a significant difference in the occlusal and middle thirds of the coronary pulp between the BLE with BLE-O and control groups (p<0.05). There was no difference in the cervical third (p>0.05). The BLE group had a high immunoexpression of TNF-α than BLE-O and control groups (p<0.05), with moderate and mild immunoexpression, respectively. Regarding IL-6 and IL-17, the BLE group had higher immunoexpression than control (p<0.05); the BLE-O was similar to the control (p>0.05). The topical anti-inflammatory Otosporin® can reduce pulp inflammation after dental bleaching in the rat teeth.

scholarly journals Bootstrap bias correction for average treatment effects with inverse propensity weights

2019 ◽  
Vol 52 (2) ◽  
pp. 187-200
Author(s):  
GUBHINDER KUNDHI ◽  
MARCEL VOIA
Keyword(s):  
Propensity Score ◽  
Treatment Effect ◽  
Bias Correction ◽  
Treatment Effects ◽  
Potential Problem ◽  
Average Treatment Effect ◽  
Omitted Variables ◽  
Control Groups ◽  
Average Treatment ◽  
And Control

The estimated average treatment effect in observational studies is biased if the assumptions of ignorability and overlap are not satisfied. To deal with this potential problem when propensity score weights are used in the estimation of the treatment effects, in this paper we propose a bootstrap bias correction estimator for the average treatment effect (ATE) obtained with the inverse propensity score (BBC-IPS) estimator. We show in simulations that the BBC-IPC performs well when we have misspecifications of the propensity score (PS) due to: omitted variables (ignorability property may not be satisfied), overlap (imbalances in distribution between treatment and control groups) and confounding effects between observables and unobservables (endogeneity). Further refinements in bias reductions of the ATE estimates in smaller samples are attained by iterating the BBC-IPS estimator.

scholarly journals A four-step strategy for handling missing outcome data in randomised trials affected by a pandemic

2020 ◽  
Author(s):  
Suzie Cro ◽  
Tim P Morris ◽  
Brennan C Kahan ◽  
Victoria R Cornelius ◽  
James R Carpenter
Keyword(s):  
Sensitivity Analysis ◽  
Missing Data ◽  
Treatment Effect ◽  
Missing At Random ◽  
Outcome Data ◽  
Sensitivity Analyses ◽  
Free World ◽  
Randomised Trials ◽  
Primary Analysis ◽  
Missing Not At Random

Abstract Background: The coronavirus pandemic (Covid-19) presents a variety of challenges for ongoing clinical trials, including an inevitably higher rate of missing outcome data, with new and non-standard reasons for missingness. International drug trial guidelines recommend trialists review plans for handling missing data in the conduct and statistical analysis, but clear recommendations are lacking.Methods: We present a four-step strategy for handling missing outcome data in the analysis of randomised trials that are ongoing during a pandemic. We consider handling missing data arising due to (i) participant infection, (ii) treatment disruptions and (iii) loss to follow-up. We consider both settings where treatment effects for a ‘pandemic-free world’ and ‘world including a pandemic’ are of interest. Results: In any trial, investigators should; (1) Clarify the treatment estimand of interest with respect to the occurrence of the pandemic; (2) Establish what data are missing for the chosen estimand; (3) Perform primary analysis under the most plausible missing data assumptions followed by; (4) Sensitivity analysis under alternative plausible assumptions. To obtain an estimate of the treatment effect in a ‘pandemic-free world’, participant data that are clinically affected by the pandemic (directly due to infection or indirectly via treatment disruptions) are not relevant and can be set to missing. For primary analysis, a missing-at-random assumption that conditions on all observed data that are expected to be associated with both the outcome and missingness may be most plausible. For the treatment effect in the ‘world including a pandemic’, all participant data is relevant and should be included in the analysis. For primary analysis, a missing-at-random assumption – potentially incorporating a pandemic time-period indicator and participant infection status – or a missing-not-at-random assumption with a poorer response may be most relevant, depending on the setting. In all scenarios, sensitivity analysis under credible missing-not-at-random assumptions should be used to evaluate the robustness of results. We highlight controlled multiple imputation as an accessible tool for conducting sensitivity analyses.Conclusions: Missing data problems will be exacerbated for trials active during the Covid-19 pandemic. This four-step strategy will facilitate clear thinking about the appropriate analysis for relevant questions of interest.

Association between serum amyloid A levels and cancers: a systematic review and meta-analysis

2018 ◽  
Vol 94 (1115) ◽  
pp. 499-507 ◽  
Author(s):  
Jielin Zhou ◽  
Jie Sheng ◽  
Yong Fan ◽  
Xingmeng Zhu ◽  
Qi Tao ◽  
...  
Keyword(s):  
Systematic Review ◽  
Serum Amyloid A ◽  
Meta Analysis ◽  
Serum Amyloid ◽  
Sensitivity Analyses ◽  
Cochrane Library ◽  
Control Groups ◽  
Amyloid A ◽  
Mean Differences ◽  
And Control

ObjectiveIncreased serum amyloid A (SAA) levels have been investigated in various human malignancies, but a consistent perspective has not been established to date. This study systematically reviewed the association between SAA levels and cancers.MethodsCochrane Library, PubMed and Embase were carefully searched for available studies. The following keywords were used in database searches: ‘serum amyloid A’, ‘SAA’, ‘cancer’, ‘tumour’, ‘carcinoma’, ‘nubble’, ‘knurl’ and ‘lump’. Pooled standard mean differences (SMDs) with corresponding 95% CIs were calculated using random-effects model analysis.ResultsTwenty studies, which contained 3682 cancer cases and 2424 healthy controls, were identified in this systematic review and meta-analysis. Our study suggested that the average SAA concentrations in the case groups were significantly higher than those in control groups (SMD 0.77, 95% CI 0.55 to 1.00, p<0.001). Subgroup analysis revealed that continent, age and cancer location were associated with SAA level differences between case groups and control groups. Sensitivity analyses showed the robustness and credibility of our results. In addition, we further stratified analyses for cancer stages and found that the concentrations of SAA increased gradually with the aggravation of cancer stages.ConclusionHigh circulating SAA levels were markedly associated with the developing risks of cancer, especially for participants from Asia, Oceania and Europe, or subject age more than 50, or locations in oesophageal squamous cell, ovarian, breast, lung, renal and gastric cancers. In addition, our study found that the concentrations of SAA increased with the severity of cancer stages.

Looking for twins: how to build better counterfactuals with matching

2021 ◽  
pp. 1-16
Author(s):  
Stefano Costalli ◽  
Fedra Negri
Keyword(s):  
Observational Data ◽  
Selection Bias ◽  
Empirical Distribution ◽  
Exact Matching ◽  
Control Groups ◽  
Units Of Analysis ◽  
Pre Treatment ◽  
Peacekeeping Missions ◽  
And Control ◽  
Matching Techniques

Abstract A primary challenge for researchers that make use of observational data is selection bias (i.e. the units of analysis exhibit systematic differences and dis-homogeneities due to non-random selection into treatment). This article encourages researchers in acknowledging this problem and discusses how and – more importantly – under which assumptions they may resort to statistical matching techniques to reduce the imbalance in the empirical distribution of pre-treatment observable variables between the treatment and control groups. With the aim of providing a practical guidance, the article engages with the evaluation of the effectiveness of peacekeeping missions in the case of the Bosnian civil war, a research topic in which selection bias is a structural feature of the observational data researchers have to use, and shows how to apply the Coarsened Exact Matching (CEM), the most widely used matching algorithm in the fields of Political Science and International Relations.

scholarly journals Lasso adjustments of treatment effect estimates in randomized experiments

2016 ◽  
Vol 113 (27) ◽  
pp. 7383-7390 ◽  
Author(s):  
Adam Bloniarz ◽  
Hanzhong Liu ◽  
Cun-Hui Zhang ◽  
Jasjeet S. Sekhon ◽  
Bin Yu
Keyword(s):  
Treatment Effect ◽  
Multivariate Regression ◽  
Asymptotic Variance ◽  
Ordinary Least Squares ◽  
Randomized Experiments ◽  
Control Groups ◽  
The Difference ◽  
Selection Operator ◽  
Simple Difference ◽  
And Control

We provide a principled way for investigators to analyze randomized experiments when the number of covariates is large. Investigators often use linear multivariate regression to analyze randomized experiments instead of simply reporting the difference of means between treatment and control groups. Their aim is to reduce the variance of the estimated treatment effect by adjusting for covariates. If there are a large number of covariates relative to the number of observations, regression may perform poorly because of overfitting. In such cases, the least absolute shrinkage and selection operator (Lasso) may be helpful. We study the resulting Lasso-based treatment effect estimator under the Neyman–Rubin model of randomized experiments. We present theoretical conditions that guarantee that the estimator is more efficient than the simple difference-of-means estimator, and we provide a conservative estimator of the asymptotic variance, which can yield tighter confidence intervals than the difference-of-means estimator. Simulation and data examples show that Lasso-based adjustment can be advantageous even when the number of covariates is less than the number of observations. Specifically, a variant using Lasso for selection and ordinary least squares (OLS) for estimation performs particularly well, and it chooses a smoothing parameter based on combined performance of Lasso and OLS.

scholarly journals D3M: Detection of differential distributions of methylation levels

2015 ◽  
Author(s):  
Yusuke Matsui ◽  
Masahiro Mizuta ◽  
Satoru Miyano ◽  
Teppei Shimamura
Keyword(s):  
Epigenetic Modification ◽  
Statistical Tests ◽  
Kernel Functions ◽  
The Cancer Genome Atlas ◽  
Underlying Structure ◽  
Lower Grade ◽  
Summary Statistics ◽  
Control Groups ◽  
The Difference ◽  
And Control

DNA methylation is an important epigenetic modification related to a variety of diseases including cancers. One of the key issues of methylation analysis is to detect the differential methylation sites between case and control groups. Previous approaches describe data with simple summary statistics and kernel functions, and then use statistical tests to determine the difference. However, a summary statistics-based approach cannot capture complicated underlying structure, and a kernel functions-based approach lacks interpretability of results. We propose a novel method D3M, for detection of differential distribution of methylation, based on distribution-valued data. Our method can detect high-order moments, such as shapes of underlying distributions in methylation profiles, based on the Wasserstein metric. We test the significance of the difference between case and control groups and provide an interpretable summary of the results. The simulation results show that the proposed method achieves promising accuracy and outperforms previous methods. Glioblastoma multiforme and lower grade glioma data from The Cancer Genome Atlas and show that our method supports recent biological advances and suggests new insights.

scholarly journals A pilot study to determine the feasibility of a cluster randomised controlled trial of an intervention to change peer attitudes towards children who stutter

2018 ◽  
Vol 65 (1) ◽  
Author(s):  
Rizwana B. Mallick ◽  
Lehana Thabane ◽  
A.S.M. Borhan ◽  
Harsha Kathard
Keyword(s):  
Pilot Study ◽  
Treatment Effect ◽  
Dropout Rate ◽  
Control Group ◽  
Control Groups ◽  
Post Intervention ◽  
Peer Attitudes ◽  
Significant Difference ◽  
Randomised Controlled ◽  
And Control

Background: While randomised controlled trials (RCTs) are considered the gold standard of research, prior study is needed to determine the feasibility of a future large-scale RCT study. Objectives: This pilot study, therefore, aimed to determine feasibility of an RCT by exploring: (1) procedural issues and (2) treatment effect of the Classroom Communication Resource (CCR), an intervention for changing peer attitudes towards children who stutter. Method: A pilot cluster stratified RCT design was employed whereby the recruitment took place first at school-level and then at individual level. The dropout rate was reported at baseline, 1 and 6 months post-intervention. For treatment effect, schools were the unit of randomisation and were randomised to receive either the CCR intervention administered by teachers or usual practice, using a 1:1 allocation ratio. The stuttering resource outcomes measure (SROM) measured treatment effect at baseline, 1 and 6 months post-intervention overall and within the constructs (positive social distance, social pressure and verbal interaction). Results: For school recruitment, 11 schools were invited to participate and 82% (n = 9) were recruited. Based on the school recruitment, N = 610 participants were eligible for this study while only n = 449 were recruited, where there was n = 183 in the intervention group and n = 266 in the control group. The dropout rate from recruitment to baseline was as follows: intervention, 23% (n = 34), and control, 6% (n = 15). At 1 month a dropout rate of 7% (n = 10) was noted in the intervention and 6% (n = 15) in the control group, whereas at 6 months, dropout rates of 7% (n = 10) and 17% (n = 44) were found in the intervention and control groups, respectively. For treatment effect on the SROM, the estimated mean differences between intervention and control groups were (95% Confidence Interval (CI): -1.07, 5.11) at 1 month and 3.01 (95% CI: -0.69, 6.69) at 6 months. A statistically significant difference was observed at 6 months on the VI subscale of the SROM, with 1.35 (95% CI: 0.58, 2.13). Conclusion: A high recruitment rate of schools and participants was observed with a high dropout rate of participants. Significant differences were only noted at 6 months post-intervention within one of the constructs of the SROM. These findings suggest that a future RCT study is warranted and feasible.

scholarly journals Comparative studies of early season moxidectin treatment and conventional ivermectin/ benzimidazole treatments in the control of cyathostomes in horses

2012 ◽  
Vol 48 (No. 1 - 2) ◽  
pp. 9-17
Author(s):  
I. Langrová ◽  
I. Jankovská ◽  
M. Borovský
Keyword(s):  
Live Weight ◽  
Treated Group ◽  
Control Group ◽  
Control Groups ◽  
Grazing Period ◽  
The Mean ◽  
Faecal Egg Counts ◽  
Pre Treatment ◽  
And Control ◽  
First Time

Moxidectin administered in January or February at a single dose was tested for efficacy in horses on two farms for 12 and 11 months, respectively. Horses were infected with cyathostomes naturally in the previous grazing period. Forty horses of farm 1 and 20 horses of farm 2 were used in controlled tests to evaluate the efficacy of moxidectin 2% gel formulation at the dosage 0.4 mg moxidectin per kg of live weight, ivermectin commercial paste formulation at the dosage 0.2 mg ivermectin per kg of live weight, mebendazole and fenbendazole commercial paste formulation at the dosage both 7.5 mg mebendazole and fenbendazole per kg of live weight, all applied orally. Three control groups of 10 horses each (farm 1) were treated twice a year with ivermectin and benzimidazoles, respectively. Individual faecal egg counts, faecal cultures and larval differentiation were performed. Moxidectin had more prolonged and greater suppressive effects on the post-treatment reappearance and magnitude of strongyle egg counts than did ivermectin or benzimidazoles. In the moxidectin treated group (M1) strongyle eggs were seen for the first time in April and a slight increase in the mean count of eggs per gram of faeces (EPG) was observed during the rest of the season. Litter larval counts significantly reflected levels of exposure during the tested season. Twenty animals of farm 2 were allocated into two groups of ten horses each based on pre-treatment eggs per gram (EPG) counts (moxidectin treated group and control group). In the moxidectin treated group mean egg counts remained very low throughout the study. A plateau was reached by autumn, with egg counts ranging from 74 to 145 EPG. The faecal egg counts of moxidectin treated group (M2) were significantly higher in March, April, May and June.

Sign in / Sign up

Export Citation Format

Share Document