In the 1800s Joel R. Poinsett, the American Ambassador to Mexico, introduced the poinsettia to the United States. The plant gained immense popularity and became known as the Christmas flower, Christmas Star, and of course, the poinsettia, to acknowledge Ambassador Poinsett. However, since 1920 the poinsettia has been one mankind's most maligned and feared plants. In that year, Rock reported the unconfirmed fatality of a 2-year-old Hawaiian child who ingested a portion of a poinsettia plant the previous year. However, that publication was not the infamous one that catapulted the poinsettia into the limelight as a sinister plant. That dubious distinction can be traced to Arnold in his 1944 book entitled Poisonous Plants of Hawaii. ^[2]

The forward to the Arnold book provides an excellent example of how misconceptions about the toxic nature of plants are propagated and perpetuated. Arnold stated, ``Even a superficial study of the poisons found in various parts of plants is sufficient to convince one that they are the most dangerous known.'' Furthermore, Dr. Arnold indicated that ``... there are no poisons which even approach in strength and in violence those found in plants... .'' These statements became a self-fulfilling prophecy that led to the authoritative discussion of indigenous Hawaiian plants and their respective toxicities and formed the basis for the paranoia that has surrounded the poinsettia.

In the section of the book that specifically discusses the poinsettia, Arnold stated, ``The two-year-old child of an Army officer at Fort Schafter died from eating a poinsettia leaf in 1919.'' He further stated, ``It causes intense emesis and catharsis, and delirium before death.'' In another section of the book he said, ``... Poinsettia, are highly poisonous when ingested... .'' Although Dr. Arnold later admitted that the information regarding the poinsettia-related fatality had never been confirmed and was merely hearsay, ^[3] his original account became legendary and led to the dissemination of information about the extremely toxic nature of the poinsettia plant.

The lay press sensationalized the poinsettia as a botanical villain of astronomical proportions. Activist citizen groups demanded that poinsettias carry a caution label to alert consumers about their inherent toxicity. In 1980 a county health officer in North Carolina prohibited poinsettias in nursing homes because of their toxicity. Even such an authority as Kingsbury, who authored Deadly Harvest, has given credence to the poinsettia's lethality: ``... poinsettia has been responsible for deaths among children'' and ``... poinsettias,...do not bear warning labels, yet most persons are unaware of their potentially lethal nature.'' According to the Society of American Florists, even the Food and Drug Administration supported the issue by stating in a 1970 press release, ``One poinsettia leaf can kill a child.'' However, the morbidity and mortality issues have been neither scientifically validated in humans nor adequately addressed.

To paraphrase Paracelsus: ``Everything is poisonous. The amount dictates the expression of that toxicity.'' The same can be said of the poinsettia. Literature reports and laboratory research make it apparent that there is little toxicity associated with casual exposure to the species. However, this has never been conclusively and irrefutably confirmed in a large human cohort. The objective of this research project was to confirm that the poinsettia is not associated with either significant morbidity or any mortality.

METHODS

Through a data grant from the American Association of Poison Control Centers (AAPCC), an electronic search was undertaken of poison exposure cases reported to the AAPCC Toxic Exposure Surveillance System (TESS) by participating poison information centers for the years 1985 to 1992 to extract all cases that involved exposures to a plant. Those data were provided to the investigators on electromagnetic computer tape, stored in a relational database system, and then electronically analyzed using a DEC 5000 workstation to extract and form a subset of all cases involving exposure to Euphorbia pulcherrima for computer analysis. The data were analyzed using the parameters inherent to the AAPCC TESS database. Specifically, the data were searched for frequency of exposure by age group, sex, reason for exposure, route of exposure, outcome, the pattern of referral to health care facilities, and treatment. Specific symptomatology was not reported to AAPCC TESS during the study period and is not included in the analysis. The morphologic portion of the plant and the amount ingested or involved in the exposure were not part of AAPCC TESS and, therefore, are not included in the analysis. The data were analyzed by using descriptive statistics.

The outcomes were defined by AAPCC TESS as no effect, minor effect, moderate effect, major effect, fatality, not followed nontoxic exposure, not followed potentially toxic exposure, unrelated effect, and unknown. Those abbreviated and paraphrased outcome definitions are as follows: no effect, the patient developed no symptoms as a result of the exposure; minor effect, the patient exhibited some symptoms as a result of the exposure, but the symptoms were minimally bothersome to the patient; moderate effect, the patient exhibited symptoms as a result of the exposure that were more pronounced, more prolonged, or more of a systemic nature than minor symptoms and were not life-threatening; major effect, the patient exhibited some symptoms as a result of the exposure and the symptoms were life-threatening or resulted in a significant residual disability or disfigurement; fatality, death resulted as a consequence of the exposure; not followed nontoxic exposure, the patient was not followed because the exposure was assessed as nontoxic; not followed potentially toxic, the patient was not followed, but the exposure was assessed as potentially toxic; unrelated effect, the patient became symptomatic, but the effect was probably unrelated to the exposure; unknown, the outcome of the patient was unknown.

RESULTS

The analysis of AAPCC TESS for the 8-year period yielded 849,575 plant exposures. Poinsettia exposures accounted for 22,793 cases. Males were involved in 49.1% of cases, compared to 49.3% in females, with 1.6% being unknown. Children accounted for 93.3% of the exposures, with 77.3% occurring in children younger than 2 years of age, 13.3% in children 2 to 5 years of age, 2.1% in children 6 to 12 years old, 0.6% in adolescents 13 to 17 years old, 5.5% of exposures were experienced by adults, and the age was unknown in 1.2% of the reports. Exposures reported during the months of December, January, and February were responsible for 74.9% of all cases

The majority (98.9%) of the exposures were accidental and only 0.9% had an intentional etiology. The reason for the exposure was unknown in 0.2% of cases. Ingestion (94.5%) was the primary route of exposure, followed by 4.8% of the cases being dermal exposures (Table 1) .T1

There were no poinsettia-related fatalities in the 22,793 cases. The outcome in 92.4% of the exposures was no effect

673

Most patients (96.1%) were not treated in a health care facility. A small number (1.6%) of poinsettia-exposed individuals did not contact a poison center and self-referred or were referred by another health care professional to a health care facility for treatment. Poison centers were responsible for 0.7% of the treatment referrals. The site of treatment was unknown in 1.6% of patients. The ultimate disposition of patients who had not been referred by a poison center and were managed in a health care facility was that 74.1% were treated and released, 9.3% were admitted for medical care, 3.8% were admitted for psychiatric care, and 12.8% were lost to follow-up. Comparatively, among patients who were referred for treatment by a poison center, 36.8% were treated and released, 2.7% were admitted for medical care, 13.6% refused referral, and 46.9% were lost to follow-up.

No therapy was used in the management of poinsettia exposure patients in 34.6% of cases, and 62.7% received some type of decontamination (miscellaneous therapies were implemented in 2.7% of patients). Within the decontamination subset of patients, dilution was the most prevalent form of intervention (89.6%), followed by irrigation of external surfaces (6.3%) and emesis induced by syrup of ipecac (2.2%). Other therapies accounted for the remaining 1.9% of decontamination procedures. For example, activated charcoal was used in the decontamination of only 29 patients. Gastric lavage was used 7 times in the 22,743 poinsettia exposures.

The effect of therapy versus no type of therapy in patients who were initially asymptomatic was also tabulated; these data are reflected in Table 2 .T2

DISCUSSION

It is apparent from these data that poinsettia exposures have good outcomes, just as we thought. This conclusion was based on the analysis of 22,793 exposures to Euphorbia pulcherrima, otherwise known as the poinsettia plant. These data represent the largest compilation of human exposures to

In contrast to the normal age distribution of poisoning exposure reports, in which approximately 56% of the exposures occur in children younger than 5 years of age, 90.6% of the poinsettia exposures involved that age group. ^[8] Most likely, this is due to the attractiveness of the poinsettia's foliage to a curious child and to the location of the plants, which are frequently within the reach of a child. Not surprisingly, nearly 75% of the exposures occurred during the 3-month period (December through February) that coincides with the Christmas holidays, when the poinsettia is used to decorate homes and businesses.

Since the majority of the poisoning exposure reports involved children, it is not surprising that 98.9% of poinsettia exposures were accidental. The poinsettia has no chemical abuse potential and, despite its notoriety as a poisonous plant, it is not used as a homicidal or suicidal agent with any frequency. However, the poinsettia was used by 16 individuals for abuse purposes and by 27 people as a suicidal agent.

Excessive hand-to-mouth activity by young children is probably responsible for the majority of pediatric exposures. Accordingly, ingestions (94.5%) accounted for the majority of poinsettia exposures, followed by a small percentage (4.8%) of dermal exposures. Those who experienced a dermal exposure had a greater incidence of minor and moderate toxicity. In the dermal exposure category 82.4% of the patients had an outcome of ``no effect'' or ``not followed, nontoxic,'' compared with 92.8% in the ingestion group. This does validate as a standard outcome the occasional reports that describe oral or dermal toxicity ^{[9] [10]} ; however, it represents a rather insignificant issue because only 6 of the 1,129 dermal exposure patients experienced a moderate outcome. Following dermal exposure to the latex from a poinsettia, skin irrigation and cleansing should prevent or diminish the risk of minor irritation.

AAPCC TESS reports a known patient outcome only when a follow-up call has been placed to determine the patient's final disposition. If it is the opinion of the Specialist in Poison Information that the exposure is insignificant from either a quantity or toxin perspective, the case may be documented as ``not followed, nontoxic.'' When the data from this category are combined with the documented ``no effect'' data, 92.4% of the patients suffered no adverse effect from the poinsettia exposures. There were no fatalities and only one case was classified as resulting in a major effect. Based on a review of the computerized record of the patient with the only reported major effect, it is apparent that the case was incorrectly coded. A 13-month-old girl ingested some portion of the poinsettia plant, was initially symptomatic, received no therapy, and was not known to be admitted to the hospital. This is totally inconsistent with the reported outcome since, by definition, cases resulting in a major effect must have a life-threatening problem, a severe disfigurement, or a disability that results in hospitalization.

The number of self-referrals, meaning that a poison information center was not consulted prior to this decision, was low (1.6%) compared with the normal pattern of

674

As expected, there was no correlation between patient outcome and therapeutic intervention in asymptomatic patients. While the data in Table 2 gives the appearance that the induction of emesis provides an improved outcome, this is misleading. Combining the ``no effect'' and ``not followed, nontoxic'' categories as an indicator of a nontoxic outcome demonstrates that the ``no therapy'' group had a nontoxic incidence of 97.3%. This is compared to 96.7% in the category which combined all types of decontamination and 94.4% in the ipecac-induced emesis group. The induction of emesis, the use of dilution, and even skin decontamination appear to be of little or no value.

The limitations of the study are somewhat overshadowed by the large sample size. However, these data are from poison information centers and there is only verbal, not visual or professional, identification of the plants. The poinsettia has very characteristic features and is well known by the general public. Therefore, the number of inaccurate identifications was probably minimal. A further limitation was that the data allow neither quantification of the amount of plant material nor identification of which morphologic portion of the plant was responsible for the exposure. Nevertheless, the influence of this limitation is probably minimal because animal research, which supports our human low toxicity data, utilized poinsettia leaves, bracts, and flowers, as well as homogenates of the plant in oral doses of up to 50 g/kg. ^{[3] [5] [6]}

CONCLUSIONS

The analysis of 22,793 poinsettia exposure cases has confirmed that the poinsettia is associated with neither significant morbidity nor any mortality despite the fact that there is a high exposure incidence among children. The majority of patients can be treated without referral to a health care facility. Gastrointestinal decontamination using dilution or emesis induced by syrup of ipepac has no impact on patient outcome and is not recommended. Dermal exposures to the poinsettia produce a slightly higher incidence of minor irritation than is observed among those who ingest the poinsettia. Although skin irritation is uncommon and minor when it does occur, dermal decontamination is easy to accomplish and may prevent the irritation.

REFERENCES