Adherence

The Design and Recruitment paper (1998) describes both adherence measurement approaches:

MEMS (Medication Event Monitoring System) is an electronic cap that fits on a medication bottle and records the date and time of each opening. It is considered the gold standard for adherence measurement in clinical trials because it captures actual dosing behavior, not just whether pills are missing from the bottle.

The MEMS data was collected at the Cincinnati/Columbus site. The Design and Recruitment paper states:

“At one clinical center, the MEMS was used to assess compliance. Results were comparable to pill counts (data not shown).”

— TLC Design and Recruitment, 1998, p. 325

This is the only mention of MEMS data in any TLC publication. The raw data was never published, and no subsequent paper provided the comparison.

The phrase “data not shown” is a red flag in clinical trial reporting. If MEMS adherence was truly comparable to pill counts, why not show the data? The comparison would strengthen confidence in the reported adherence figures.

Pill counts are known to overestimate adherence. Participants may dump pills before returning bottles (“pill dumping”), or pills may be lost or shared with others. MEMS captures the actual pattern of bottle openings, revealing whether doses were taken on schedule.

If MEMS adherence was substantially lower than pill counts indicated, several conclusions follow:

• The reported adherence figures would overestimate actual medication-taking behavior
• The per-protocol treatment effect (efficacy in adherent participants) would be diluted
• The intention-to-treat null result would be less informative about whether chelation works
• The null result would be more informative about whether this protocol, as actually implemented, worked

Without the MEMS data, we cannot distinguish between two interpretations: (1) succimer does not improve cognition even when taken as prescribed, or (2) succimer was not taken as prescribed, and the null result reflects implementation failure rather than treatment failure.

Based on pill counts alone, TLC reported high adherence rates. The Safety and Efficacy paper (2000) stated that >80% of participants took ≥80% of prescribed doses.

However, administering succimer to toddlers (ages 12–33 months at enrollment) presents significant challenges:

• Succimer capsules must be opened and mixed with food or drink
• The medication has an unpleasant sulfur taste and odor
• Dosing required three times daily for the first 5 days, then twice daily for 14 days
• Multiple courses were required for most participants

Whether caregivers actually achieved the reported >80% adherence rate — and whether children actually ingested the medication after it was mixed — cannot be confirmed without the MEMS data.