Presented at the 2009 North American Congress of Clinical Toxicology meeting, San Antonio, TX, September 2009.
Factors Associated With Difficulty Achieving Initial Control With Crotalidae Polyvalent Immune Fab Antivenom in Snakebite Patients
Article first published online: 16 DEC 2010
© 2010 by the Society for Academic Emergency Medicine
Academic Emergency Medicine
Volume 18, Issue 1, pages 46–52, January 2011
How to Cite
Yin, S., Kokko, J., Lavonas, E., Mlynarchek, S., Bogdan, G. and Schaeffer, T. (2011), Factors Associated With Difficulty Achieving Initial Control With Crotalidae Polyvalent Immune Fab Antivenom in Snakebite Patients. Academic Emergency Medicine, 18: 46–52. doi: 10.1111/j.1553-2712.2010.00958.x
This study was a secondary analysis of data that were originally collected in a study funded by a grant from Fougera (now Nycomed, Zurich, Switzerland). Rocky Mountain Poison and Drug Center also received grant funding to revalidate and reanalyze the data from Protherics (now BTG, West Conshohocken, PA). Rocky Mountain Poison and Drug Center has consultation and call center contracts with Nycomed and BTG. The authors are salaried employees of Rocky Mountain Poison and Drug Center and do not receive any direct funding from Nycomed or BTG.
Supervising Editor: Mark Mycyck, MD.
- Issue published online: 10 JAN 2011
- Article first published online: 16 DEC 2010
- Received April 6, 2010; revisions received June 1 and June 8, 2010; accepted June 9, 2010.
ACADEMIC EMERGENCY MEDICINE 2011; 18:46–52 © 2011 by the Society for Academic Emergency Medicine
Background: The prescribing information for Crotalidae Fab antivenom (FabAV) instructs clinicians to administer FabAV until initial control of the envenomation syndrome is achieved. Risk factors for difficulty achieving initial control are not known.
Objectives: The study aim was to identify factors present before administration of antivenom associated with difficulty achieving initial control.
Methods: The authors conducted a retrospective study of all patients presenting to any one of 17 centers and receiving FabAV from 2002 to 2004. Demographic and historical information, as well as data about nine specific venom effects, were collected prior to the administration of antivenom. An expert panel used standard criteria to determine if initial control was achieved. The patient group that had difficulty achieving initial control was compared to the group that achieved initial control, and adjusted odds ratios were calculated using stepwise logistic regression.
Results: A total of 247 patients were included in the final analysis. The majority of patients were envenomated on the upper extremity and were young males. A total of 203 patients (82.2%) achieved initial control. In univariate analysis, thrombocytopenia, bleeding, neurologic effects, and a severe bite were significantly associated with difficulty achieving initial control. After logistic regression, the presence of neurologic effects and thrombocytopenia remained significantly associated with difficulty achieving initial control. When both factors were present, the patient was 13.8 times more likely to have difficulty achieving initial control.
Conclusions: A number of factors were present before the administration of FabAV that were independently associated with difficulty achieving initial control of the envenomation syndrome. Predicting which patients will have difficulty achieving initial control has important ramifications for patient disposition and may provide insight into the mechanisms for lack of antivenom efficacy.
Envenomation by pit viper snakes (family Viperidae, subfamily Crotalinae, genera Crotalus, Agkistrodon, and Sistrurus) is a potentially serious medical condition in the United States. The American Association of Poison Control Centers received more than 2,900 reports of patients who sought emergency care for a crotaline snakebite in 2008. Approximately 60% of these patients received antivenom as part of their treatment.1 The only specific therapy for crotaline envenomation in the United States is an ovine Fab antivenom product (referred hereafter as FabAV) approved by the U.S. Food and Drug Administration in October 2000 (CroFab, BTG, West Conshohocken, PA). FabAV is a polyclonal antibody mixture that contains only the antigen-binding fragment of the whole IgG molecule.
Envenomation by crotaline snakes is manifested by local, systemic, and/or hematologic effects. Local venom effects include progressive swelling and pain extending from the bite site. Systemic effects may include hypotension, vomiting, abdominal pain, confusion, dyspnea, and tachycardia. Hematologic effects include thrombocytopenia, hypofibrinogenemia, and coagulopathy. The presence of any one of these effects (local, systemic, or hematologic) is typically considered an indication for antivenom therapy.2
The arrest of these venom effects after administration of antivenom is generally described as “initial control.” The FabAV product insert instructs clinicians to administer four to six vials of antivenom, with additional doses as needed until initial control of the envenomation syndrome is achieved. Once initial control is achieved, the manufacturer recommends administration of maintenance dosing of two vials every 6 hours.3 While most patients achieve initial control with one to two doses (each dose being four to six vials) of FabAV, there are numerous reports of difficulty achieving initial control.4–9 Predicting which patients might have difficulty achieving initial control has important implications on determining whether a hospital has sufficient antivenom stocks to treat a given patient. The objective of this study was to identify factors present before receiving antivenom associated with a subsequent difficulty in achieving initial control.
This was a secondary analysis of data collected in a multicenter retrospective chart review from 17 U.S. hospitals located throughout the range of geographically indigenous crotaline venomous snakes.
Study Setting and Population
All patients who received FabAV for treatment of a crotaline snakebite from 2002 to 2004 were eligible. The medical record had to include clinical signs and symptoms before the first dose of antivenom, a record of all antivenom doses administered, and an indication of whether or not initial control was believed to have been achieved at the initial time of treatment. The institutional review board at each participating institution approved this study.
An electronic search of pharmacy records was used to identify all patients treated with FabAV at each study hospital during the study period. The chart for each patient who received FabAV during 2004 was abstracted. Investigators from institutions with fewer than 10 medical records from 2004 abstracted all medical records from 2002 through 2004, to decrease a possible sampling bias. An investigator at each site obtained original medical records and abstracted data to a standardized form. Standardized definitions of key variables were provided. Abstractors received 1 hour of training, followed by practice abstraction of two sample medical records. After practice abstraction, both group and individual feedback were provided to the data abstractors, and the tool was modified to resolve areas of ambiguity. Deidentified forms were sent to the study coordinating center, where the data were entered into a secure database (Access 2003, Microsoft Corp., Redmond, WA). Data were quality checked, and individual feedback with specific queries was sent back to site investigators. Site investigators were blinded to the study hypothesis.
Venom effects prior to the receipt of antivenom were recorded and are described in Table 1. Progressive pain and swelling were not defined with uniform criteria, but were determined by each site abstractor. In addition, data on these other variables were also recorded: modified severity score prior to receiving antivenom, number of vials of antivenom given in the first dose, platelet count, international normalized ratio, fibrinogen, and time to treatment with antivenom from envenomation. The modified severity score is a simplified version of the snakebite severity score established and validated by Dart et al.10 The modified severity score (see Table 2) was used because it includes only elements that were likely to be found routinely in medical records documenting an envenomation. Like the snakebite severity score, the modified severity score appears to correlate with a clinician’s global impression of severity (data not shown). A severe bite was defined as a modified severity score of either 5 or 6.
|Progressive pain||Increasing pain|
|Progressive swelling||Increasing swelling|
|Coagulopathy||Fibrinogen <150 μg/mL, prothrombin time >20 sec, or INR >1.2|
|Thrombocytopenia||Platelet count <150 × 109/L|
|Significant or spontaneous bleeding||Sudden, unexplained bleeding from gums/nose, hematuria, petechiae, or increased bruising tendency|
|Respiratory||Respiratory rate >20 breaths/min, dyspnea, apnea, chest tightness, or respiratory distress|
|Cardiovascular||Heart rate >125 beats/min or systolic blood pressure <100 mm Hg|
|Neurologic||Apprehension, headache, weakness, paresthesia, confusion, or fasciculation of the bite site|
|Gastrointestinal||Nausea, vomiting, or diarrhea|
|No progressive pain or swelling||0|
|Progressive pain or swelling||1|
|Progressive pain and swelling||2|
|No systemic venom effects||0|
|1–2 systemic venom effects||1|
|3–5 systemic venom effects||2|
|No coagulopathy or thrombocytopenia||0|
|Fibrinogen <150 μg/mL, PT >20 sec, INR >1.2, and/or platelets <150 × 109/L||1|
|Fibrinogen <50 μg/mL, PT >50 sec, INR >2.0, and/or platelets <50 × 109/L||2|
For this study we defined initial control of the envenomation syndrome as the simultaneous occurrence of no progressive local tissue effects, no systemic venom effects, including no evident bleeding, and all coagulation studies and platelet count clearly trending toward normal. Two medical toxicologists with extensive experience in management of crotalidae envenomation (EL, TS) independently reviewed each case abstraction form to determine whether initial control of all components of the envenomation syndrome was achieved. Patients who failed to achieve initial control did not meet satisfactory criteria for initial control during the documented visit as determined by the panel. Because we could not confirm whether a patient failed to achieve initial control simply because he or she had not received sufficient antivenom, the group that failed to achieve initial control will hereafter be referred to as having difficulty achieving initial control. Discrepancies were resolved by consensus. These investigators were blinded to the study hypothesis.
Data were analyzed with SAS, version 9.1.3 (SAS Institute, Inc., Cary, NC). The primary outcome was achievement of initial control. The Wilcoxon test was used to determine if there was a difference in the number of vials of antivenom between the two groups. Potential factors that were analyzed were the nine venom effects, patient age, sex, number of antivenom vials administered in the first dose, time to treatment, and severity both as a dichotomous and as a continuous variable. Each variable was fit separately to derive individual odds ratios for difficulty achieving initial control. Factors that were identified as potentially significant (p-value < 0.1, using the Wald test) with regard to difficulty achieving initial control were analyzed with stepwise multivariable logistic regression to derive adjusted odds ratios. In the final model, statistical significance was defined as p < 0.05. Interaction was checked between remaining variables in the model.
Because initial control was determined retrospectively, and difficulty achieving initial control may have been due to insufficient treatment with antivenom, we also performed a subanalysis to validate our findings. Patients who required more than 12 vials to achieve initial control and who did not achieve initial control with more than 12 vials, were compared to those patients who achieved initial control with 12 or fewer vials. Patients who did not achieve initial control but were given 12 or fewer vials were not included in this subanalysis. Twelve vials were chosen as a cutoff because this would equal two doses of six vials.
A total of 247 patients met inclusion criteria and comprised the study cohort. Of these patients, 203 (82.2%, 95% confidence interval [CI] = 76.9% to 86.4%) achieved initial control (see Table 3 for the demographic characteristics and Figure 1 for the geographic distribution of the patients). Our study cohort was reflective of the demographics of U.S. snakebite victims.11 There was no difference (p = 0.4) between the group that achieved initial control and the group that did not in total number of vials of FabAV received during the entire course of their treatment. However, the median number of vials needed to achieve initial control was six, compared to the group that did have difficulty achieving initial control, which received a median of 11.5 vials. Many (40.1%) of the implicated snakes were rattlesnake species, but specific species were not stated in 118 (47.8%) of the cases. Table 4 shows the snake species involved.
|Demographic||Initial Control Achieved||Initial Control Not Achieved||Total|
|Patients, n (%)||203 (82.2)||44 (17.8)||247|
|Age, median yr (range)||27.0 (1.0–81.0)||38.5 (2.0–91.0)||29.0 (1.0–91.0)|
|Male||167 (82.3)||39 (88.6)||208 (83.4)|
|Female||36 (17.7)||5 (11.4)||41 (16.6)|
|Black/African American||7 (3.4)||2 (4.5)||9 (3.6)|
|Asian||5 (2.5)||0 (0)||5 (2.0)|
|White||137 (67.5)||24 (54.5)||161 (65.2)|
|Hispanic||19 (9.4)||7 (15.9)||26 (10.5)|
|Native American||5 (2.5)||1 (2.3)||6 (2.4)|
|Not recorded||30 (14.8)||10 (22.7)||40 (16.2)|
|Body||1 (0.5)||0 (0)||1 (0.4)|
|Upper extremity||125 (61.6)||34 (77.3)||159 (64.4)|
|Lower extremity||77 (37.9)||10 (22.7)||87 (35.2)|
|Onset of symptoms to treatment, median hours (range)||3.0 (0.5–182.0), n = 199||3.0 (0.5–61.0), n = 42||3.0 (0.5–182.0), n = 241|
|Total number of vials received, median (IQR)||10 (6–14)||11.5 (6–18)||10 (6–14)|
|Species||Initial Control Achieved||Initial Control Not Achieved||Total|
|Species not reported||50 (86.2)||8 (13.8)||58 (23.5)|
|Copperhead||49 (92.4)||4 (7.6)||53 (21.5)|
|Unknown rattlesnake||26 (76.5)||8 (23.5)||34 (13.8)|
|Other species||16 (66.7)||8 (33.3)||24 (9.7)|
|Southern Pacific||15 (79.0)||4 (21.0)||19 (7.7)|
|Western diamondback||13 (86.7)||2 (13.3)||15 (6.1)|
|Water moccasin||11 (100.0)||0 (0.0)||11 (4.5)|
|Timber||5 (62.5)||3 (37.5)||8 (3.2)|
|Pygmy||7 (87.5)||1 (12.5)||8 (3.2)|
|Mojave||4 (66.7)||2 (33.3)||6 (2.4)|
|Eastern diamondback||3 (75.0)||1 (25.0)||4 (1.6)|
|Red diamond||1 (50.0)||1 (50.0)||2 (0.8)|
|Sidewinder||0 (0.0)||2 (100.0)||2 (0.8)|
|Unknown pit viper||2 (100.0)||0 (0.0)||2 (0.8)|
|Speckled||1 (100.0)||0 (0.0)||1 (0.4)|
|Total||203 (82.2)||44 (17.8)||247|
The unadjusted odds ratio of each potential risk factor for difficulty achieving initial control is shown in Table 5. Thrombocytopenia, neurologic effects, bleeding, and a severe bite were significantly associated with difficulty achieving initial control. Progressive pain was the only dichotomous variable in which an odds ratio could not be calculated, but there was no significant difference between the two groups largely because almost every patient had progressive pain present.
|Variable||Odds Ratio (95% CI)||p-value|
|Patient age (yr)*||1.01 (0.99–1.03)||0.09|
|<6 yr old compared to all other ages||0.90 (0.3–2.6)||0.9|
|>65 yr old compared to all other ages||0.80 (0.1–3.2)||0.8|
|Male sex||1.70 (0.7–5.1)||0.3|
|Significant/spontaneous bleeding*||15.90 (2.0–326.7)||0.002|
|First dose vials||1.20 (0.9–1.5)||0.2|
|Time to treatment (hours)||0.99 (0.96–1.02)||0.9|
|Severe bite (Modified Severity Score 5 or 6)*||5.10 (2.1–12.3)||0.0002|
|Severity as continuous variable*||1.90 (1.3–2.6)||0.0002|
To evaluate their independent contribution to the risk for difficulty achieving initial control, all factors with p < 0.10 were modeled with stepwise logistic regression. After multivariable modeling, the presence of thrombocytopenia and neurologic effects before receipt of antivenom remained significantly associated with difficulty achieving initial control. Table 6 shows the adjusted odds ratios for these two factors. There was no significant interaction (p = 0.99) between the two terms in the final model. When both thrombocytopenia and neurologic effects were present, the patient had a 13.8-fold (95% CI = 4.6 to 41.7) risk for difficulty achieving initial control.
|Risk Factor||Odds Ratio (95% CI)|
|Neurologic effects||3.8 (1.8–8.1)|
In univariate modeling, thrombocytopenia, neurologic effects, coagulopathy, and severe bites were significantly associated (all p < 0.05) with patients who required more than 12 vials to achieve initial control (n = 24) or did not achieve initial control with more than 12 vials (n = 15), compared to patients who achieved initial control with at most 12 vials (n = 179).
Last, seven patients received more than 24 vials (range = 25–46) of antivenom during the entire course of their treatment. Six of these patients had either neurologic effects present (n = 4) and/or thrombocytopenia (n = 3). The seventh patient did not have platelets measured before receiving antivenom.
We also did a post hoc power analysis based on 80% power using two-sided testing at the 5% level of significance. Based on the 17.8% of 247 subjects who did not reach initial control, the current analyses are adequately powered to detect odds ratios in the range of 3 to 3.3 for risk factors occurring in at least 20% of the initial control group, provided correlation between multiple risk factors is weak (R2 < 0.30).
Achievement of initial control of envenomation syndrome has important clinical implications, because additional administration of antivenom is largely dependent on this. In addition, predicting which patients will have difficulty achieving initial control may influence decisions on patient disposition, as smaller hospitals may not have sufficient antivenom stocks and may need to either acquire more antivenom or transfer the patient. In our study, the presence of thrombocytopenia or neurologic venom effects prior to treatment with antivenom were significantly and independently associated with difficulty achieving initial control. Patients who had both factors present were 13.8 times more likely to have difficulty achieving initial control than patients who had neither factor present. To the best of our knowledge, no other study has previously examined this issue in North American snakebite patients.
There are a number of possible explanations for these findings. It is possible that these two venom effects are associated with larger venom loads and that patients are more likely to have difficulty achieving initial control because they are not receiving sufficient neutralizing antibody. We believe that this is not likely, because there are case reports of patients receiving extremely large amounts of antivenom without apparent control of venom effects.4–6 It may also be that these venom effects result from an irreversible cascade of events precipitated by the envenomation. Time to treatment from envenomation to receipt of antivenom did not differ significantly between the two groups, which suggests that if a cascade of events is an important factor, it begins very early.
Another potential explanation is that these are venom effects that are not well neutralized by FabAV, irrespective of the quantity administered. Possible mechanisms for antivenom difficulty in neutralizing the venom may be that the responsible venom component may be too small to be bound by immunoglobulin or may not be immunogenic in sheep or the venoms used to create FabAV may not contain the responsible components. In our study population, thrombocytopenia and neurologic effects were also two of the three (the third factor was coagulopathy) most common factors that subsequently were responsible for the patient not meeting criteria for achieving initial control, showing that these were also persistent effects following receipt of antivenom (data not shown).
If FabAV is not neutralizing these venom effects, then it may not be reasonable to attempt to completely control these two venom effects with antivenom. Giving more vials of antivenom is essentially an assumption that the antivenom efficacy is not enough (i.e., the antivenom dose is too small). If neutralization of these effects is not a dose-dependent issue, then administering more doses of the antivenom will potentially increase the frequency of adverse events such as serum sickness, as well as increase the total health care cost, without resulting in increased efficacy. Perhaps it is also not necessary to control these venom effects. The incidence of significant bleeding from venom-induced delayed coagulopathy is considered low,5,12 and conservative management of these patients has been advocated.5 However, there are case reports of significant bleeding, including a fatality,7,13,14 so ultimately it will be important to determine whether patients who have delayed significant bleeding are a unique subset and whether venom effects or individual patient factors are the primary determinants for bleeding risk.
Alternatively, the efficacy of FabAV may be improved from addition of components to neutralize those specific venom effects. Evidence for this comes from reports of clinical effects of specific snake species. FabAV presently contains venom components from four snake species: western diamondback (Crotalus atrox), eastern diamondback (C. adamanteus), Mojave (C. scutulatus), and cottonmouth (Agkistrodon piscovorus). A retrospective review of timber rattlesnake (C. horridus) envenomations showed that thrombocytopenia was resistant to treatment with antivenin crotalidae polyvalent (Wyeth, Madison, NJ), which suggested that the venom contained a component that was not neutralized by the antivenom.15 A protein called crotalocytin has been identified in timber rattlesnake venom that can cause platelet aggregation and induce platelet ATP release.16,17 Cases of neurotoxicity resistant to treatment with FabAV following envenomation by the Southern Pacific rattlesnake (C. helleri) have also been reported.9 An animal lethality model showed that venom from C. helleri required higher doses of FabAV for protection than venom from other species,18 although a small case series demonstrated efficacy of FabAV in humans envenomated by C. helleri.19 Future research will need to determine the pathophysiology for the difficulty in the antivenom in controlling these effects, as inability to bind the venom component may only be one explanation.20 However, if further research is able to identify specific venom components that are associated with these factors, then it may be prudent to add these venom components into the formation of the antivenom. Controlling these venom effects may result in substantial reduction of health care costs by reducing the need for additional vials of antivenom, close monitoring of patients with persistent laboratory abnormalities, or both.
The following factors may have limited the internal validity of the study. Some subjective judgment is required in the determination of initial control. Although we standardized this as much as possible using explicit data analysis rules, the panelists noted that this analysis required them to classify some cases as not achieving initial control when the actions of the treating physician suggested a satisfactory clinical response. Hence, although our primary outcome was the failure to achieve initial control, because we could not confirm that all cases of failure to achieve initial control failed despite adequate treatment with antivenom, we have ultimately described the outcome as difficulty achieving initial control. Since there was no difference between the two groups in total number of antivenom vials received, it is unlikely that undertreatment with antivenom was the primary reason for difficulty achieving initial control. Because the results of the subanalysis were largely similar to the results obtained with the original analysis, we believe the results to be valid. In addition, we believe that the results obtained have face validity, given the case reports of patients with refractory thrombocytopenia and neurotoxicity.
A second limitation was that not every patient had all labs measured. It is likely that the patients who did not have some labs measured were less clinically ill and were probably more likely to achieve initial control.
This study was also a secondary analysis of data that were not originally designed to answer the question that we posed. Last, we were unable to model all potential factors such as snake species and geographic variations. Venom components are known to vary between snake species and even have geographic differences within species.21 Given the small numbers of individual snake species identified, our study did not have sufficient power to differentiate those factors. The limit to external validity is that the data were collected at regional referral centers with experience in caring for snakebite patients. This may result in a different spectrum of illness severity or clinical approach to management.
We identified a number of factors present before the administration of Fab antivenom that were independently associated with difficulty achieving initial control of the envenomation syndrome, which included the presence of thrombocytopenia and neurologic effects. When both venom effects were present, the patient was 13.8 times more likely to have difficulty achieving initial control. Predicting which patients are more likely to have difficulty achieving initial control has important ramifications for patient disposition and may provide insight into the mechanisms for lack of antivenom efficacy.
The authors acknowledge Becki Bucher-Bartelson, Vikhyat Bebarta, Sean Bush, Richard Clark, Kirk Cumpston, J. Ward Donovan, James Garrison, John Haynes Jr., Greg Hendey, Christopher Holstege, Erica Liebelt, David Morgan, Marcus Willett, John Shepard, Vincent Speranza, William Richardson, Rutherford Rose, and Anthony Pizon.
- 3CroFab Prescribing Information. Brentwood, TN: Protherics Inc., 2006..