OBJECTIVE: To identify potential bias in non-inferiority design of published cancer trials, and to provide suggestions for future practice.
METHODS: We systematically searched MEDLINE, Embase and CENTRAL databases (until April 17, 2020) to obtain non-inferiority phase III cancer trials and protocols. Distribution of essential characteristics and study design parameters was compared between trials with and without concluding non-inferiority using multivariable logistic regression.
RESULTS: A total of 291 eligible trials were included. We observed that increased odds of concluding non-inferiority was significantly associated with more lenient non-inferiority margins (OR = 1•94, 95% CI 1•02-3•69) and higher hypothesized event rate (OR = 1•24, 95% CI 1•06-1•47). Trials that established non-inferiority adopted margins that were more dispersedly distributed (dispersion OR = 2•90, 95% CI 1•88-4.48).
CONCLUSIONS: Although limited by the exploratory nature, our study demonstrated existence of possible distorted non-inferiority design which could incur excess non-inferiority in cancer clinical trials. Pre-registration and transparent reporting of detailed non-inferiority design is imperative for future research.

Coronavirus disease (COVID-19) is a respiratory disease affecting many people and able to be transmitted through direct and perhaps indirect contact. Direct contact transmission, mediated by aerosols or droplets, is widely demonstrated, whereas indirect transmission is only supported by collateral evidence such as virus persistence on inanimate surfaces and data from other similar viruses. The present systematic review aims to estimate SARS-CoV-2 prevalence on inanimate surfaces, identifying risk levels according to surface characteristics. Data were obtained from studies in published papers collected from two databases (PubMed and Embase) with the last search on 1 September 2020. Included studies had to be papers in English, had to deal with coronavirus and had to consider inanimate surfaces in real settings. Studies were coded according to our assessment of the risk that the investigated surfaces could be contaminated by SARS-CoV-2. A meta-analysis and a metaregression were carried out to quantify virus RNA prevalence and to identify important factors driving differences among studies. Thirty-nine out of forty retrieved paper reported studies carried out in healthcare settings on the prevalence of virus RNA, five studies carry out also analyses through cell culture and six tested the viability of isolated viruses. Overall prevalences of SARS-CoV-2 RNA on high-, medium- and low-risk surfaces were 0.22 (CI95 [0.152-0.296]), 0.04 (CI95 [0.007-0.090]), and 0.00 (CI95 [0.00-0.019]), respectively. The duration surfaces were exposed to virus sources (patients) was the main factor explaining differences in prevalence.

The study evaluated the quality of cardiovascular prevention in real-world clinical practice. The recurrence of up to five cardiovascular events was assessed, as data on recurrence beyond the first event and interindividual variations in event rates past the second event have been sparse. Low-density lipoprotein cholesterol concentrations and lipid-lowering therapy use were investigated.
This retrospective register-based study included adult patients with an incident cardiovascular event between 2004 and 2016 treated in the hospital district of southwest Finland. Patients were followed for consecutive cardiovascular events or cardiovascular death, low-density lipoprotein cholesterol and statin purchases. The timing of event recurrence was evaluated, and predictive factors were assessed.
A wide interindividual variation in cardiovascular event recurrence was observed, each additional event caused an increased risk, the median time of recurrence decreased from 7 to one year for the second and fifth event. Event rates increased correspondingly from 12 to 43/100 patient-years and were most pronounced in the first years following the previous event. The low-density lipoprotein cholesterol goal (<1.8 mmol/l) was reached by 18% in the year after the event and statin underuse was associated with an increased risk of recurrence. Six months after the index event high intensity statins were used by only 22% of the cohort.
The study provides new perspectives on individual risk assessment showing that event rates are not stable for all patients but increase 1.2-1.9-fold per consecutive event. The underuse of statins and poor adherence support the identification of these patients for intensified multifactorial preventive measures.

While the solid-state nanopore sensors have shown exceptional promise with their single-molecule sensitivity and label-free operations, one of the most significant challenges in the nanopore sensor is the limited analyte translocation event rate that leads to prolonged sensor response time. This issue is more pronounced when the analyte concentration is below the nanomolar (nM) range, owing to the diffusion-limited mass transport. In this work, we systematically studied the experimental factors beyond the intrinsic analyte concentration and electrophoretic mobility that affect the event rate in glass nanopore sensors. We developed a quantitative model to capture the impact of nanopore surface charge density, ionic strength, nanopore geometry, and translocation direction on the event rate. The synergistic effects of these factors on the event rates were investigated with the aim to find the optimized experimental conditions for operating the glass nanopore sensor from the response time standpoint. The findings in the study would provide useful and practical insight to enhance the device response time and achieve a lower detection limit for various glass nanopore-sensing experiments.

The PRESERVE trial used a 2 × 2 factorial design to compare intravenous saline solution with intravenous sodium bicarbonate solution and oral N-acetylcysteine with placebo for the prevention of 90-day major adverse kidney events and death (MAKE-D) and contrast-associated acute kidney injury (CA-AKI) among patients with chronic kidney disease undergoing angiography. In this ancillary study, we evaluated the predictive capacities of preangiography injury and repair proteins in urine and plasma for MAKE-D, CA-AKI, and their impact on trial design.
Longitudinal analysis.
A subset of participants from the PRESERVE trial.
Injury (KIM-1, NGAL, and IL-18) and repair (MCP-1, UMOD, and YKL-40) proteins in urine and plasma 1 to 2 hours preangiography.
MAKE-D and CA-AKI.
We analyzed the associations of preangiography biomarkers with MAKE-D and with CA-AKI. We evaluated whether the biomarker levels could enrich the MAKE-D event rate and improve future clinical trial efficiency through an online biomarker prognostic enrichment tool available at prognosticenrichment.com.
We measured plasma biomarkers in 916 participants and urine biomarkers in 797 participants. After adjusting for urinary albumin-creatinine ratio and baseline estimated glomerular filtration rate, preangiography levels of 4 plasma (KIM-1, NGAL, UMOD, and YKL-40) and 3 urine (NGAL, IL-18, and YKL-40) biomarkers were associated with MAKE-D. Only plasma KIM-1 level was significantly associated with CA-AKI after adjustment. Biomarker levels provided modest discriminatory capacity for MAKE-D. Screening patients using the 50th percentile of preangiography plasma KIM-1 or YKL-40 levels would have reduced the required sample size by 30% (∼2,000 participants).
Evaluation of prognostic enrichment does not account for changing trial costs, time needed to screen patients, or loss to follow-up. Most participants were male, limiting the generalizability of our findings.
Preangiography levels of injury and repair biomarkers modestly predict the development of MAKE-D and can be used to improve the efficiency of future CA-AKI trials.

Tyrosine-kinase inhibitors (TKIs) can be associated with vascular events (VEs). The expected VE rates in patients with chronic myeloid leukemia (CML) are unknown. The present study examined the event rates and mortality among elderly patients with and without CML.
Linked Surveillance, Epidemiology, and End Results cancer registry and Medicare claims data were used to identify patients aged ≥ 66 years with an incident (index) diagnosis of CML from 2004 to 2009. A comparison cohort of patients without cancer was matched 1:1 to the CML cohort. All patients were followed up from 12 months before the index diagnosis through death or December 31, 2010. The overall survival and rates of myocardial infarction (MI), stroke, pulmonary embolism (PE), and peripheral arterial disease (PAD) were analyzed.
A total of 1466 patients with CML (mean age, 78 years; average follow-up period, 25 months) were identified and matched 1:1 to a noncancer cohort (mean age, 78 years; follow-up period, 42 months). Compared with the noncancer patients, those with CML had greater mortality (63% vs. 23% died during the follow-up period; median survival, 23 vs. > 84 months) and greater rates of MI (33.0 vs. 11.9 per 1000 person-years), stroke (83.2 vs. 43.0), PE (6.6 vs. 2.6), and PAD (92.1 vs. 59.3; P < .01 for all). Of the 15% of CML patients with TKI claims, 97% had received imatinib. The event rates were not elevated for TKI-treated patients compared with the overall group of patients with CML.
Elderly patients with CML had greater mortality and greater rates of MI, stroke, PE, and PAD than did noncancer patients. The event rates were not elevated among the TKI-treated (primary imatinib) patients, suggesting that the VE risk in these patients with CML was driven primarily by the underlying factors associated with CML.

OBJECTIVE: The negative predictive value of a normal single-photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI) is very high. However, prognostic implication of a normal SPECT MPI in patients with known coronary artery disease (CAD) is not clear. Objective of this study was to evaluate the cardiac event rate in patients with known CAD who had a normal stress SPECT MPI.
METHODS: This prospective study accrued 428 consecutive patients with a history of CAD [revascularization or previous myocardial infarction (MI)] who had a normal stress (dynamic exercise or dipyridamole intervention) and rest Tc-99m-MIBI SPECT MPI. These patients were followed for 2-5 years (median: 3.1 years) for all-cause and cardiac mortality and non-fatal MI. Univariate and multivariate analyses were performed to identify predictors of outcome.
RESULTS: During a follow-up period, all-cause mortality was found in 60 patients (14%) and 41 (10%) died of cardiac reasons. Non-fatal MI was found in 77 (18%) patients. Annualized cardiac mortality and non-fatal MI rates were 2% and 3.6% respectively. Smoking, congestive heart failure (CHF) and failure to achieve 85% age predicted heart rate were found to be predictors for all-cause and cardiac mortality. Diabetes, dyslipidemia, smoking and limited functional capacity (<7 METS) were found to be predictors for non-fatal MI.
CONCLUSIONS: Patients with known CAD had higher cardiac event rates despite a normal stress SPECT MPI. Diabetes, dyslipidemia, smoking and limited functional capacity were the predictors for fatal and non-fatal cardiac events. A cost effective but comprehensive surveillance strategy is warranted.

In clinical trials with counts of recurrent event data, it is often of particular interest to test whether the experimental treatment reduces the event rate in comparison with a control. The sample size calculation for such a trial often assumes fixed follow-up time for each patient. In many trials, however, we follow all patients for a predetermined follow-up time after the end of the accrual period, in which case the follow-up time is variable. This article provides methods for sample size calculation for clinical trials with ordinary Poisson count data and variable follow-up time allowing for nonuniform accrual and early dropouts. We also generalize the sample size formula to count data with overdispersion.