Erika Cheung testified for the prosecution about Theranos laboratory operations, Edison's limited testing capability, recurring quality-control problems, proficiency testing, and her escalation of concerns. Cross-examination challenged the scope of her expertise and personal knowledge, while redirect clarified the continuing role of quality control after validation.
Erika Cheung
Former Theranos laboratory associate who worked in assay development and clinical testing.
About
Erika Cheung testified for the prosecution about her work in Theranos research and development and its clinical laboratory from October 2013 until approximately April 2014. She described Edison devices as processing one assay for one patient at a time and supporting a limited menu of tests, while the laboratory also relied on modified third-party analyzers, conventional equipment, and outside laboratories.
Her direct examination addressed continuing Edison validation, recurring quality-control failures, differences between written procedures and laboratory practice, manual removal of data points, device variability, and proficiency testing. Cheung testified that predicate-method results rather than Edison-generated results were provided to regulators during her tenure. She also described raising concerns with laboratory and company leaders, declining to process a hepatitis C sample on Edison because of reagent-stability concerns, leaving after Balwani questioned her qualifications and directed her to process samples without question, and later reporting concerns to CMS. An email to Holmes was admitted only to show notice, not for the truth of its contents.
On cross-examination, the defense tested the scope of Cheung's expertise and personal knowledge. She acknowledged that she largely ran experiments and compiled data for others to analyze, and that she lacked familiarity with several regulatory standards, policies, documents, and company initiatives. She also agreed that quality-control materials were not patient samples, calibration and bias correction were ordinary laboratory activities, and unresolved failures prevented patient results from being released. Cheung nevertheless maintained that actual practices sometimes differed from written procedures and disagreed with management's account that the written procedure had been followed during the November 2013 vitamin D incident.
On redirect, Cheung explained that validation was followed by continuing quality-control monitoring. She testified that recurring failures could require recalibration, occupy laboratory personnel around the clock, and delay patient results, and she described Edison processing as a multistep workflow rather than a minimally trained, highly automated process.
Trial Record (37)
No proceedings match this filter.