From the Guidelines
The primary recommendation for selecting initial treatment in a waterfall trial is to choose a single primary outcome that best captures the overall efficacy of a treatment, balancing clinical and statistical issues, as this decision greatly affects how the trial will be perceived. When designing a waterfall trial, it is crucial to select an initial treatment that has shown strong evidence of efficacy in preclinical data and mechanism of action, and then systematically work through alternative options if the initial treatment fails 1. This approach involves beginning with a treatment that has demonstrated the strongest evidence of efficacy in laboratory studies or early clinical trials for the specific condition being studied.
The choice of primary outcome is vital, as it will greatly impact the trial's interpretation, and a single P-value from hypothesis testing of the primary endpoint can never replace the need for a trial’s interpretation to rely on the totality of evidence, including secondary and safety outcomes 1. In the context of clinical practice, initial drug selections should be based on trial evidence of treatment efficacy, combined with recognition of compelling indications for use of an agent from a specific drug class, as well as the individual patient’s lifestyle preferences and traits 1.
Key considerations in selecting initial treatment in a waterfall trial include:
- Choosing a treatment with strong preclinical data and mechanism of action
- Systematically working through alternative options if the initial treatment fails
- Selecting a single primary outcome that balances clinical and statistical issues
- Considering the individual patient’s lifestyle preferences and traits
- Recognizing compelling indications for use of an agent from a specific drug class.
From the FDA Drug Label
The FDA drug label does not answer the question.
From the Research
Waterfall Trial Overview
- A waterfall trial is a type of clinical trial design where patients are initially treated with a standard therapy, and those who do not respond are then switched to an alternative therapy.
- The primary recommendation for selecting initial treatment in a waterfall trial is not explicitly stated in the provided studies.
Study Findings
- A study published in the Journal of the National Cancer Institute 2 discussed the use and misuse of waterfall plots, highlighting the variability in criteria used to generate them and the impact of measurement error.
- Another study published in JAMA network open 3 assessed the accuracy of waterfall plot representations of response rates in cancer treatment, finding that waterfall plots often overestimate response rates.
- A study published in BMC medical research methodology 4 compared individual trials and systematic reviews, suggesting that the most precise trial can provide similar estimates of effects to those of the meta-analyses.
Clinical Trial Design
- The ALLHAT trial, mentioned in a study published in Nederlands tijdschrift voor geneeskunde 5, compared the effects of different blood-pressure lowering drugs, but did not specifically discuss waterfall trial design.
- A study published in Trials 6 compared trialists' choice of primary outcome with what patients and health professionals want, finding that patients and healthcare professionals agreed with the choice of primary outcome made by trial teams only 28% of the time.
Key Considerations
- When interpreting results of waterfall plots in clinical trials, caution should be exercised due to the potential for variability in criteria used to generate them and measurement errors 2.
- The choice of primary outcome in a trial is crucial, as it determines the required number of participants and is used to judge the effectiveness of the intervention 6.