Rather than adhering to the usual confines of standard methods for conducting trials, the LAGO design adopts an iterative approach similar to that used by engineers who tweak, tailor, and adjust a design in real time until an optimal model is achieved.
In LAGO trials, researchers propose an initial multi-component intervention package that is then implemented in pre-determined stages. Data from each stage are analyzed and used to reassess and revise the intervention in preparation for the next stage. This process can continue until an optimal intervention is found that maximizes effectiveness while minimizing cost. Unlike other adaptive study designs, the composition of the intervention in the later stages of a LAGO trial depends on the outcomes from previous stages.
The concept behind the LAGO design first arose following the discouraging results of the BetterBirth Study, a large-scale, cluster-randomized clinical trial that sought to reduce maternal and neonatal mortality rates in Northern India through the implementation of a Safe Childbirth Checklist. Despite enrolling nearly 160,000 mothers and implementing a comprehensive intervention package, the study’s results were disappointingly null, with only modest improvements in birthing practices and no significant reduction in deaths.
Although researchers could see that the BetterBirth trial was failing to generate improved survival rates as it progressed, there was nothing they could do, due to the strict principles of randomized clinical trials.
LAGO’s rigorous study design, coupled with its inherent flexibility and build-as-you-go approach, allows researchers to “leapfrog” over the ill-fitting paradigm of a randomized clinical trial to accelerate the development and implementation of impactful public health interventions, Spiegelman said.
“Given the vast numbers of preventable deaths around the world and gaping health disparities seen among high-, medium-, and low-income countries, we cannot afford to run null trials of proven interventions,” said Spiegelman, the founding director of YSPH’s Center for Methods in Implementation and Prevention Science (CMIPS).
“The culture, values, and methodologies of randomized clinical trials, although extremely useful for testing the effectiveness of new drugs and devices, are not necessarily optimal when you apply them to large-scale public health interventions,” Spiegelman continued. “Large-scale public health interventions typically don’t involve the evaluation of a single intervention such as a multivitamin versus a placebo. They often involve multi-layered and complex intervention strategies.”
Development of the LAGO design was spearheaded by Spiegelman, who worked in collaboration with Daniel Nevo, an associate professor of statistics at Tel Aviv University, and Judith Lok, an associate professor of mathematics and statistics at Boston University.
Tests of the new study framework are ongoing. A LAGO trial is about to be implemented in Uganda with the goal of optimizing strategies to integrate hypertension prevention, screening, and treatment into longstanding HIV clinics. A separate implementation trial called Medly seeks to optimize treatment and care of patients with heart failure in Uganda with a focus on using digital technology to enhance self-care. That trial is being launched in collaboration with the Uganda Heart Institute; Yale Dr. Jeremy I. Schwartz, MD is a principal investigator. LAGO is also being used to optimize a complex, multi-layered intervention being tested by LaRon Nelson, associate dean of global health at the Yale School of Nursing. The intervention aims to reduce racial disparities in HIV incidence among Black men who have sex with men in the American South by leveraging social media influencers and peer support and implementing stigma reduction measures.