We performed a large, multicenter, randomized controlled trial to determine the efficacy and safety of early colonoscopy on outcomes of patients with acute lower gastrointestinal bleeding (ALGIB). We performed an open-label study at 15 hospitals in Japan of 170 patients with ALGIB randomly assigned (1:1) to groups that underwent early colonoscopy (within 24 hours of initial visit to the hospital) or elective colonoscopy (24-96 hours after hospital admission). The primary outcome was identification of stigmata of recent hemorrhage (SRH). Secondary outcomes were rebleeding within 30 days, endoscopic treatment success, need for transfusion, length of stay, thrombotic events within 30 days, death within 30 days, and adverse events. SRH were identified in 17 of 79 patients (21.5%) in the early colonoscopy group vs 17 of 80 patients (21.3%) in the elective colonoscopy group (difference, 0.3; 95% confidence interval, -12.5 to 13.0; P = .967). Rebleeding within 30 days of hospital admission occurred in 15.3% of patients in the early colonoscopy group and 6.7% of patients in the elective colonoscopy group (difference, 8.6; 95% confidence interval, -1.4 to 18.7); there were no significant differences between groups in successful endoscopic treatment rate, transfusion rate, length of stay, thrombotic events, or death within 30 days. The adverse event of hemorrhagic shock occurred during bowel preparation in no patient in the early group vs 2 patients (2.5%) in the elective colonoscopy group. In a randomized controlled study, we found that colonoscopy within 24 hours after hospital admission did not increase SRH or reduce rebleeding compared with colonoscopy at 24-96 hours in patients with ALGIB. ClinicalTrials.gov, Numbers: UMIN000021129 and NCT03098173.

Graphical models are useful tools in causal inference, and causal directed acyclic graphs (DAGs) are used extensively to determine the variables for which it is sufficient to control for confounding to estimate causal effects. We discuss the following ten pitfalls and tips that are easily overlooked when using DAGs; 1) Each node on DAGs corresponds to a random variable, and not its realized values; 2) The presence or absence of arrow in DAGs corresponds to the presence or absence of individual causal effect in the population; 3) “Non-manipulable” variables and their arrows should be drawn with care; 4) It is preferable to draw DAGs for the total population, rather than for the exposed or unexposed groups; 5) DAGs are primarily useful to examine the presence of confounding in distribution in the notion of confounding in expectation; 6) Although DAGs provide qualitative differences of causal structures, they cannot describe details of how to adjust for confounding; 7) DAGs can be used to illustrate the consequences of matching and the appropriate handling of matched variables in cohort and case-control studies; 8) When explicitly accounting for temporal order in DAGs, it is necessary to use separate nodes for each timing; 9) In certain cases, DAGs with signed edges can be used in drawing conclusions about the direction of bias; and 10) DAGs can be (and should be) used to describe not only confounding bias but also other forms of bias. We also discuss recent developments of graphical models and their future directions.

Epidemiologists are increasingly encountering complex longitudinal data, in which exposures and their confounders vary during follow-up. When a prior exposure affects the confounders of the subsequent exposures, estimating the effects of the time-varying exposures requires special statistical techniques, possibly with structural (i.e., counterfactual) models for targeted effects, even if all confounders are accurately measured. Among the methods used to estimate such effects, which can be cast as a marginal structural model in a straightforward way, one popular approach is inverse probability weighting. Despite the seemingly intuitive theory and easy-to-implement software, misunderstandings (or “pitfalls”) remain. For example, one may mistakenly equate marginal structural models with inverse probability weighting, failing to distinguish a marginal structural model encoding the causal parameters of interest from a nuisance model for exposure probability, and thereby failing to separate the problems of variable selection and model specification for these distinct models. Assuming the causal parameters of interest are identified given the study design and measurements, we provide a step-by-step illustration of generalized computation of standardization (called the g-formula) and inverse probability weighting, as well as the specification of marginal structural models, particularly for time-varying exposures. We use a novel hypothetical example, which allows us access to typically hidden potential outcomes. This illustration provides steppingstones (or “tips”) to understand more concretely the estimation of the effects of complex time-varying exposures.

Objective To investigate the factors associated with serious psychological distress (SPD) during the COVID-19 pandemic in Japan. Design Nationwide cross-sectional study using survey data. Setting Internet survey using sampling weights for national estimates conducted between 25 August and 30 September 2020 in Japan. Exposures Demographics (age, gender), socioeconomic status (income level, employment type, educational attainment, marital status, family composition and caregiving burden); the experience of domestic violence (DV), the state of emergency and fear of and stigma related to COVID-19. Main outcome measures Prevalence of SPD, defined as Kessler 6 Scale score ≥13. Results Among 25 482 respondents, 10.0% met the criteria of SPD. Overall, women (adjusted OR (aOR) 1.59; 95% CI 1.17 to 2.16; p=0.003), ages 15–29 (aOR 2.35 compared with ages 45–59 years; 95% CI 1.64 to 3.38; p<0.001), low-income level (aOR 1.70 compared with intermediate income; 95% CI 1.16 to 2.49; p=0.007), providing caregiving to family members (aOR 5.48; 95% CI 3.51 to 8.56; p<0.001), experiencing DV (aOR 5.72; 95% CI 3.81 to 8.59; p<0.001) and fear of COVID-19 (aOR 1.96; 95% CI 1.55 to 2.48; p<0.001) were associated with SPD. Among women aged 15–29 years, who have a higher risk of suicide during the COVID-19 pandemic in Japan, caregiving, DV, fear of COVID-19 and COVID-19-related stigma were associated with SPD. Conclusions Economic situation, caregiving burden, DV and fear of COVID-19 were independently associated with SPD during the COVID-19 pandemic. Among young women, similar factors, except economic situation, were associated with SPD. Targeted interventions based on age and gender may be more effective in mitigating the negative impact of the COVID-19 pandemic on the population’s mental health.

Aims: The criteria for ‘good’ Fontan haemodynamics have been poorly defined in relation to long-term outcomes. The aim of this study was to identify the risk factors for mortality among haemodynamic parameters obtained early after the Fontan operation. Methods and results: Clinical data of all perioperative survivors of the Fontan operation performed before 2011, from nine institutions, were collected through a retrospective chart review. In total, 1260 patients were included. The median age at the time of Fontan operation was 3.6 years. Post-operative cardiac catheterization was conducted in 1117 patients at a median period of 1.0 years after the operation. During the median follow-up period of 10.2 years, 107 patients died. The mortality rates at 10, 20, and 25 years after the operation were 5%, 12%, and 22%, respectively. On multivariable analysis, older age at the time of the operation {≥15 years, hazard ratio (HR) [95% confidence interval (CI)]: 3.2 (1.7–5.9)} and haemodynamic parameters obtained at post-operative catheterization, such as low ejection fraction [<30%, HR (95% CI): 7.5 (3.2–18)], low systemic oxygen saturation [<80%, HR (95% CI): 3.8 (1.6–9.1)], high central venous pressure [≥16 mmHg, HR (95% CI): 2.3 (1.3–3.9)], and low mean systemic arterial pressure [<60 mmHg, HR (95% CI): 3.0 (1.4–6.2)] were identified as independent predictors of mortality. The predictive model based on these parameters had a c-index of 0.75 at 10 years. Conclusions: Haemodynamic parameters obtained at a median period of 1.0 years, post-operatively, can accurately identify patients with a high mortality risk, who may need intensive management to improve long-term outcomes.