In reply to “Interpretation of ‘An analysis of energy-drink toxicity in the National Poison Data System’”

To the Editor:

We welcome the opportunity to address the concerns expressed by Drs. Barker and Seger¹ regarding our article, “An analysis of energy-drink toxicity in the National Poison Data System (NPDS).”²

The authors¹ expressed concerns about our use of the words outcome, adverse effects, and toxicity. We used NPDS data to describe the clinical effects (CEs) associated with single-product, energy-drink poison center exposures.² The NPDS defines three clinical characteristics, reported signs, symptoms, and clinical findings, collectively as CEs describing the patient's exposure response.^3,4 The NPDS Medical Outcome categories indicate the severity of these effects and whether or not they are consistent with the agent's exposure profile.^3,4

According to NPDS coding guidelines, the Medical Outcome is the “final determination made by the Specialist in Poison Information based upon all the information available at the conclusion of a case.” Of the 11 NPDS Outcomes (no effect is coded as 0), we only analyzed calls with one of the five categories of known medical outcomes: No effect, minor, moderate, major, and death.² The remaining six categories indicate cases in which a medical outcome is unknown, the exposure is not responsible for the CEs, or death has been only indirectly reported.^3,4 Thus, exposures in these categories have no confirmed outcomes and were excluded from the analysis as stated in the article.²

The authors¹ correctly note that outcomes were unknown for 345 children, but saying that the incidence of adverse effects could be as low as 14% if these children were not followed due to lack of toxicity is speculation. As previously stated, unknown cases were excluded because of lack of a known medical outcome, not because of lack of toxicity, so inclusion of these would be misleading.

Assessing the ingested amount of a liquid and its ingredients, as in many energy-drink calls, is difficult. Determining the range of amounts and ingredients ingested in terms of clinical characteristics is an important research goal and one that NPDS data analysis alone might not answer.

Our data support the statement that “most calls for alcoholic energy-drink ingestion resulted in advice to seek medical care”.² Advice to seek medical care occurred in 76.7% of alcoholic energy-drink compared to 26.4% for non-alcoholic exposures.² Our study was not designed to determine causality.² We reported only “a greater proportion of alcoholic energy-drink cases … were referred to a healthcare facility.” Our ICU admission estimate of 10% is based on 94 usable reports from healthcare facilities; we did not include missing data (n = 47), lost-to-follow-up (n = 28), or non-arrivals (n = 13).²

The authors¹ did not take issue with our conclusion “the effects of caffeine in energy-drinks, where it may be combined with a number of other substances, is still relatively under-studied in vivo.”² Poison center data offer a novel view of reported energy-drink exposures across the United States. The true incidence of all poisoning exposures is unknown,⁵ and it is an area for further study as the authors suggest. Energy-drinks merit further study, especially in children and people with medical conditions that put them at increased risk for stimulant toxicity.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.