A web application for the design of multi-arm clinical trials

Grayling, Michael J.; Wason, James

doi:10.1186/s12885-020-6525-0

Cited by 18 publications

(17 citation statements)

References 33 publications

(45 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We used https://mjgrayling.shinyapps.io/multiarm/ 27 to calculate that a total of 6200 participants would need to be randomized to detect a 20% reduction in the proportion of participants meeting the primary outcome with 84% marginal power 28 and 5% type 1 error rate, assuming a 20% risk of experiencing at least one swab test- or doctor-confirmed ARI in the no offer group at six months, 25% loss to follow-up, and a 1:1:2 ratio of participants randomized to higher-dose, lower-dose or no offer, respectively. Statistical analyses were performed according to intention-to-treat; adjustment for multiple testing for the primary outcome used Dunnett’s test 29 with a critical P-value threshold of 0.027.…”

Section: Methodsmentioning

confidence: 99%

Vitamin D Supplements for Prevention of COVID-19 or other Acute Respiratory Infections: a Phase 3 Randomised Controlled Trial (CORONAVIT)

Jolliffe

Holt

Greenig

et al. 2022

Preprint

View full text Add to dashboard Cite

BACKGROUND: Vitamin D metabolites support innate immune responses to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and other respiratory pathogens. Randomized controlled trials of vitamin D to prevent coronavirus disease 2019 (Covid-19) have not yet reported. METHODS: We randomly assigned 6200 U.K. adults to receive an offer of a postal finger-prick 25-hydroxyvitamin D (25[OH]D) test with provision of a 6-month supply of higher-dose vitamin D (3200 IU/d, n=1550) or lower-dose vitamin D (800 IU/d, n=1550) to those with blood 25(OH)D concentration <75 nmol/L, vs. no offer of testing or supplementation (n=3100). The primary outcome was the proportion of participants experiencing at least one swab test- or doctor-confirmed acute respiratory infection (ARI) of any cause at six months. Secondary outcomes included incidence of swab test-confirmed Covid-19. RESULTS: Of 3100 participants offered testing, 2958 (95.4%) accepted, and 2690 (86.8%) had 25(OH)D <75 nmol/L and were sent vitamin D supplements (1356 higher-dose, 1334 lower-dose). 76 (5.0%) vs. 87 (5.7%) vs. 136 (4.6%) participants in higher-dose vs. lower-dose vs. no-offer groups experienced at least one ARI of any cause (odds ratio [OR] for higher-dose vs. no-offer 1.09, 95% CI 0.82-1.46; lower-dose vs. no-offer 1.26, 0.96-1.66). 45 (3.0%) vs. 55 (3.6%) vs. 78 (2.6%) participants in higher-dose vs. lower-dose vs. no-offer groups developed Covid-19 (OR for higher-dose vs. no-offer 1.13, 0.78-1.63; lower-dose vs. no-offer 1.39, 0.98-1.97). CONCLUSIONS: Among adults with a high baseline prevalence of vitamin D insufficiency, implementation of a test-and-treat approach to vitamin D replacement did not reduce risk of all-cause ARI or Covid-19.

show abstract

Section: Methodsmentioning

confidence: 99%

Vitamin D Supplements for Prevention of COVID-19 or other Acute Respiratory Infections: a Phase 3 Randomised Controlled Trial (CORONAVIT)

Jolliffe

Holt

Greenig

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…It is also possible to obtain the same power with fewer experimental units, preferable due to ethical aspects for experiments with animals or humans. Sample size calculations might be performed using the web application developed by Grayling and Wason (2020), or by simulation (Colegrave & Ruxton, 2021).…”

Section: Discussionmentioning

confidence: 99%

Unequal sample sizes according to the square‐root allocation rule are useful when comparing several treatments with a control

2021

View full text Add to dashboard Cite

When investigating a specified group, generally a control, and k other groups, called treatment groups hereafter, often the control group is compared with each of the other groups. In this many-to-one situation, Dunnett's test (Dunnett, 1955) is recommended for normally distributed data. Generalised Dunnett's tests exist for other types of data, for example there are analogues for proportions and nonparametric approaches (Hothorn, 2016). However, here we focus on the original test proposed by Dunnett (1955). This test controls the experiment-wise Type I error rate.Dunnett's test is based on t test statistics. However, the pooled variance estimate is based on data from all groups and the correlation between the t statistics is considered (see, e.g., Bretz et al., 2011, pp. 71-75). Thus, the method is more powerful than a Bonferroni adjustment made after multiple two-sample tests (Bretz et al., 2011, p. 74).When applying Dunnett's test, appropriately chosen unbalanced sample sizes can give a more powerful test than a balanced design, even when there is no difference in variability between groups.Here, we seek to promote greater awareness of this. Groups of equal size are sometimes suggested very generally, see, for example,

show abstract

“…If a single‐step method such as Bonferroni is used, the power calculations can be adjusted by adjusting the significance level for each test. Accounting for the use of other multiplicity adjustment procedures, such as the Hochberg or fixed sequence procedures, can be more complex 19 …”

Section: Methodsmentioning

confidence: 99%

Power analysis for stepped wedge trials with multiple interventions

Sundin

Crespi

2022

Statistics in Medicine

View full text Add to dashboard Cite

Stepped wedge design (SWD) trials are cluster randomized trials that feature staggered, unidirectional cross‐over between treatment conditions. Existing literature on power for SWDs focuses primarily on designs with two conditions, typically a control and an intervention condition. However, SWDs with more than one treatment condition are being proposed and conducted. We present a linear mixed model for SWDs with two or more interventions, including both multiarm and factorial designs. We derive standard errors of the intervention effect coefficients, and present power calculation methods. We consider both repeated cross‐sectional and cohort designs. Design features, with a focus on treatment allocations, are examined to determine their impact on power.

show abstract

A web application for the design of multi-arm clinical trials

Cited by 18 publications

References 33 publications

Vitamin D Supplements for Prevention of COVID-19 or other Acute Respiratory Infections: a Phase 3 Randomised Controlled Trial (CORONAVIT)

Vitamin D Supplements for Prevention of COVID-19 or other Acute Respiratory Infections: a Phase 3 Randomised Controlled Trial (CORONAVIT)

Unequal sample sizes according to the square‐root allocation rule are useful when comparing several treatments with a control

Power analysis for stepped wedge trials with multiple interventions

Contact Info

Product

Resources

About