In the data reviewed between the 18-month period of January 2019 to July 2020, 1,939,172 FSHE products were sold by CW to approximately 646,391 consumers. A total of 431 AEs were reported by 304 unique individuals, with some individuals reporting more than one AE (e.g., headache and nausea). Of these, 421 AEs (97.68%) were attributed to oral product use and 10 (2.32%) were attributed to topical product use, two of which were related to accidental oral ingestion of topical products. Of the 431 AEs, seven (1.62%) were classified as serious, with no AE categorized as a Grade 4 or 5 (i.e., life threatening or fatal). All seven SAEs were reported in the context of oral product use. The estimated percentage of consumers who reported at least one adverse event was 0.05%. The percentage of AEs and SAEs per total products sold was 0.02% and 0.0004%, respectively.

When examining the SAEs in more detail, two were categorized as gastrointestinal disorders, two were categorized as nervous system disorders, two were categorized as psychiatric disorders, and one was categorized as an eye disorder. The strength of association between all seven SAEs and product use was determined quantitatively by SafetyCall. In this assessment, five of the seven SAEs were unlikely to be associated with product use. One SAE was rated as possibly associated with product use and another SAE was rated as probably associated with product use.

Qualitative review of the SAEs revealed that two of seven SAEs occurred in the context of higher doses of FSHE products with one of the two being attributed to a possible drug interaction between a pharmaceutical medication and the trace amounts of THC found in FSHE products. Of note, many of the consumers reporting SAEs had complex medical histories and were taking multiple medications in combination with the FSHE products, which increases the potential for adverse reactions [25]. For example, multiple of the SAEs reported herein may be related to concurrent use of FSHE products and medications with known drug interactions, such as clonazepam and citalopram [25]. In one event, an 80-year-old woman was taking higher than recommend doses of the product, leading to the potential ingestion of up to 8 mg of THC, which is known to interact with gabapentin, a medication she was taking at the same time [27, 28]. The use of medications with known contraindications is especially important to consider when the SAEs reported are not consistent with adverse events typically associated with CBD or cannabinoid use [23–26]. In another example in which complex medical histories and potential medication interactions can be taken into consideration, a consumer with a significant medical history who was also taking gabapentin experienced an eye disorder. There is currently no evidence to support that cannabinoids interfere with the transmission of visual information from the retina [33, 34]. Moreover, one review suggests potential neuroprotective effects of cannabinoids concerning the pathophysiology of retinal ischemia [33]. In sum, all of these consumers had increased risk of experiencing an adverse event due to their medical history, concurrent medication use, or the use of FSHE products at higher than recommended doses. This is important information for the general public which underscores the importance of consulting a medical professional when adding FSHE products into a complex regimen of medications or for someone with a significant medical history.

This is the first attempt at reporting adverse events related to CW FSHE products. Unlike prescription medication or other products which are tied directly to specific consumers, consumption of dietary supplements is not. Products may be purchased and unused, purchased and shared by multiple people, or one consumer may use multiple products. This study addressed this issue by utilizing a calculation to estimate the total individuals who purchased CW products during the 18-month period. This calculation used DTC data on number of products sold and number of unique DTC purchasers to extrapolate a total number of products sold per person. While this approach was novel, with such large numbers (367,000 people purchased 1.08 million products directly from CW), it can be assumed that these consumers are representative of the larger consumer base that bought CW products across all channels (e.g., bricks and mortar retail). This calculation strategy is supported by research indicating that online and in-store shoppers are similar in their motivations and characteristics [35]. Further, this study also included the approach of measuring AEs based on products sold. This method is commonly used in postmarketing surveillance studies [22] and has been utilized by another study that examined the safety of other FSHE products [13]. While these methods for determining approximate number of consumers have been utilized in other studies and it can be assumed that the DTC purchasers are indeed representative of the larger consumer base, it should still be noted that these conclusions are based on estimations. A full-scale clinical trial would need to be conducted in order to arrive at the most accurate percentages of AEs to consumers and products purchased. This is outside the scope of this study, which was conducted in the context of postmarketing surveillance.

A limitation to this dataset, and many postmarketing surveillance studies, is the lack of certain details which would make AE reporting more accurate (e.g., reason for product use, other medication use, dosage). As observed in the SAEs reported above, consumers do not always capture their doses accurately or at all. There are also other potential confounding factors which were not reported or reported with missing details that complicate the ability to determine causality (e.g., underlying medical conditions, concomitant use of prescription medications or other dietary supplements, dietary changes, life stressors).

Additionally, a limitation to the methodology broadly is the use of self-report for adverse events as opposed to clinician-reported AEs. Self-reported AEs are known to have distinct issues. Under-reporting is a common limitation when relying on consumer self-report [7, 35]. Consumers may not report AEs due to a lack of awareness and recognition of what constitutes an AE, or they may incorrectly recall details surrounding the events [36–38]. Despite these inherent limitations, consumer self-reported AEs are seen as complementary to clinician-reported AEs, and it is considered valuable to collect both types of reports to capture potential safety issues [10, 11].

A final limitation to this study is the lack of a gold standard in assessing adverse event causality in the context of dietary supplement use. When examining the causality of an AE in the context of pharmaceuticals, certain methods are employed to help with standardization, such as the Naranjo algorithm [39]. No such standardization exists for evaluating dietary supplements and thus, the association assessment by SafetyCall was utilized for this publication. Future research should explore methods of adapting existing assessments from the pharmaceutical industry to apply to dietary supplements to reduce potential bias and increase replicability of results.

Despite the limitations to the dataset and the drawbacks related to self-reporting, postmarketing surveillance data and consumer participation in monitoring adverse events are important pieces to the puzzle of tracking product safety. The merits of self-reporting AEs are relied upon internationally, as both the European Medicines Agency and the United States Food and Drug Administration have encouraged consumers to report suspected AEs through their reporting databases [38, 40]. This shift in valuing self-reported AEs has produced tangible results, as consumer reports have led to labelling changes and drug withdrawals from the market [41, 42]. Overall, this movement supports the idea of consumers as “vigilant partners” in their own care, as they have the ability to decrease the risk of commercially available products [12, 43].