REVIEW ARTICLE
Online appendix - 1Evaluation of the risk factors for the rate of allergic reaction after bee and vespid sting in adults – a systematic review of the literature
Appendix 1: Scientific research that has studied a prognostic factor that would reliably show more severe systemic allergic reactions (anaphylaxis), or milder or major local allergic reactions after a wasp, bee or hornet sting.
Author, year of 
publication
The objective of 
the study
Research 
epidemiological 
basis
Methods of data collection Observed population/
environment/period
Results Main findings
Parker et al. 
[1982]
(23)
Identify risk 
factors for 
a systemic 
reaction using 
an insect sting 
provocation test.
Prospective 
observational 
cohort study.
A provocation test was performed on 21 
patients, 5 of whom had a history of a large 
local reaction after a previous sting, 16 of 
whom had a history of a systemic reaction to 
the sting. No one received immunotherapy. 
Just before
the provocation test (bees, wasps, hornets), 
skin tests were performed and sIgE and sIgG 
antibodies were measured.
• IgE: 
• positives were expressed as the percentage 
of sIgE-bound serum of the non-sensitized 
person (negative control) and the levels, 
that were higher than 200%, were 
evaluated as positive 
If several provocative tests were performed 
on one patient (with different insects (wasp 
and bee), the next was performed at least 60 
minutes after the first.
21 patients (10 
women, 11 men).
Mayo, United States of 
America.
From June 1979 until 
October 1980.
21 subjects, 5 with a history of LLR and 16 with a history of SR at the previous sting.
In the LLR group, 7/21 skin tests were positive before the provocation test, and in the SR group, 30/34 skin tests were 
positive.
sIgEs were at 63% to 789% for LR and at 66% to 801% for SR.
sIgG antibodies ranged from <5 to 332 U/ml, levels were similar in both groups.
55 provocation tests (bee stings 14, wasps 11, hornets 8): 7 systemic reactions were induced, which were only in the 
SR history group.
The article describes each patient individually with a systemic reaction, but does not provide an assessment of the 
systemic reaction compared to the previous – first – systemic reaction. There is also no description of factors (sex, 
age, other factors) that would predict a recurrent systemic response (small observed population).
For LLR, 1/5 of the patients had a negative skin test and negative sIgE; 1/5 of patients had a positive skin test and 
positive sIgE; 2/5 had a negative skin test and positive sIgE; and 1/5 positive skin test and negative sIgE.
For SR, 4/16 patients had a positive skin test and negative sIgE; 0/16 a negative skin test and positive sIgE; 1/16 a 
negative skin test and negative SIgE; and 11/16 a positive skin test and positive sIgE.
7 individuals (3 in the LR history group and 4 in the SR history group) who tolerated the first provocation test repeated 
it with venom a total of 27 more times. Of these 7, skin tests were now positive at 10/15 in the LLR group and at 10/12 
in the SR group. sIgE levels ranged from 103% to 737% for the LR group and 85% to 2389% for the SR group before the 
second provocation test. IgG antibodies were the same in both groups. All other provocation tests (27 in total) were 
without a systemic response.
There is no in-vitro or in-vivo test 
that can predict with certainty the 
clinical immune status of insect 
stings for
sensitized patients. It is still the 
most reliable method
for referral to immunotherapy 
provocation test.
Blaauw et al. 
[1984]
(24)
To present 
experience 
with diagnostic 
tests to identify 
allergy to insect 
venoms.
Prospective 
observational 
cohort study.
106 patients underwent an insect sting 
provocation test: skin tests, sIgE, sIgG level, 
and sIgG / sIgE ratio were measured.
86 patients with a positive history, skin tests 
and sIgE were eligible for immunotherapy 
according to the defined criteria.
The severity of SR was assessed according to 
the Müller’s grading system.
106 patients.
Helmond, 
Netherlands.
From May 1979 until 
August 1983.
86 patients had a history of SR and a positive skin test and sIgE for insect venom. 29 of these patients responded with 
a systemic response and were candidates for immunotherapy. Of the 29, 39% responded to wasp venom provocation 
and 61% to bee venom provocation. This is due to the fact that 80% of the patients in the group of 86 patients were 
beekeepers.
The article does not describe whether the reactions after the provocation test were more severe or lesser with respect 
to the Müller’s grading system, nor other risk factors.
A comparison of diagnostic data from 29 responders and 57 non-responders from 86 patients showed that a 
provocative test could provide evidence for an insect sting allergy that manifests itself with a severe systemic reaction 
and requires immunotherapy. 57 patients without a reaction had a provocation test repeated after six months to see if 
the allergy increased over time. No one reacted with serious symptoms.
In the group of the other 20 patients, 2 responded with SR, including one who had negative sIgE and one with 
negative skin tests. Both patients responded to the provocation test with a more severe systemic response than after 
the first sting (grade IV).
The study defined the importance 
of provocative tests in assessing a 
severe systemic reaction after an 
insect sting.
Kampelmacher 
et al. [1987]
(25)
Clarify doubts 
about an insect 
sting provocation 
test that clinical 
data, sIgE and 
skin tests, and a 
positive medical 
history for SR 
are reliable tools 
in diagnosing 
and deciding on 
immunotherapy.
Prospective 
observational 
cohort study
90 patients (51 males, 39 females) with 
previous SR for bee stings (15) or wasp stings 
(74) or both (1) were admitted to the hospital 
intensive care unit
for a provocation test with a resting. The 
severity of SR was assessed according to the 
Müller’s grading system.
20 of them had a confirmed allergic disease 
(allergic rhinitis, food allergy, drug allergy).
90 patients with a 
previous systemic 
reaction.
Utrecht, Netherlands.
Summer of 1983 and 
1984.
Of the 90 patients who underwent the provocation test, 25 (28%) responded with recurrent SR, of which 5 (20%) had 
allergic disease (allergic rhinitis, food allergy, drug allergy).
65 (72%) responded to the provocation test with a large local reaction only, of which 15 (23%) patients had allergic 
disease (allergic rhinitis, food allergy, drug allergy).
A comparison of the assessment of the severity of the systemic reaction before and after the provocation test showed:
• lower severity in 80 patients,
• increased severity in 2 patients,
• same severity in 8 patients.
They did not demonstrate a statistically significant association between the severity of the first reaction and the 
severity after the provocation test.
The interval between the last SR and the provocation test (> 1 year and <1 year) did not show a statistically significant 
difference between SR and non-responders.
There was no statistically significant difference in age, sex, and the presence of other allergic diseases among patients 
who did not respond to the provocation test.
However, a negative skin test and/or sIgE does not preclude a recurrent systemic reaction.
The long-lasting value of the 
provocation test has proven to be 
good as the results of the repeated
provocation test were the same. 
Skin tests and sIgE are useful as 
diagnostic tools in determining the 
insect in question. The provocation 
test provides information on the 
probability of a recurrent systemic 
reaction.
2 - Online appendix Zdrav Vestn | september – oktober 2020 | Letnik 89 | https://doi.org/10.6016/ZdravVestn.2973
PUBLIC HEALTH (OCCUPATIONAL MEDICINE)
Author, year of 
publication
The objective of 
the study
Research 
epidemiological 
basis
Methods of data collection Observed population/
environment/period
Results Main findings
Pucci et al. 
[1994]
(26)
To determine 
whether a 
short interval 
(<2 months) 
between two 
stings affects the 
development of 
allergies (SR) for 
insect stings.
A retrospective 
observational 
cohort study.
The study compared the time interval 
between repeated insect stings in 120 allergic 
patients who experienced the first insect 
sting with no systemic reaction (73 for wasps 
and 47 for bees). They were interested in 
whether the short interval between the first 
and resting was a predictive factor for the 
severity of the systemic response.
Skin tests and sIgE were measured. Patients 
were divided into two groups (resting < 2 
months and > 2 months).
120 patients with a 
systemic reaction to 
insect stings and 100 
healthy controls.
Ancona, Italy.
There was a significant difference in the distribution of the time interval between the two stings in both groups (p = 
0.0001). In 71 of the 120 allergic patients, the provocative sting that provoked SR was essentially the second sting after 
some previous sting that did not provoke a systemic reaction and occurred less than two months before the systemic 
reaction. There were only four individuals in the control group who experienced a sting less than two months after the 
resting. The mean age was lower (34 ± 12 years) in the 71 patients who responded to SR < 2 months after resting than 
in the other 49 patients (43 ± 14 years). Therefore, age 34 ± 12 years is associated with an increased risk of SR reaction. 
A comparison of the severity of the allergic reaction after the resting did not show that the severity of the first reaction 
predicts the severity of the subsequent reaction.
Sex and a history of atopy were not associated with the severity of the allergic reaction after the resting. Skin tests 
and sIgE did not predict the severity of the resting reaction.
The type of insect was not associated with the severity of the resting reaction.
The short interval between the two 
stings is indicated as an important 
risk factor for the development of 
SR on insects.
Van der Linden 
et al. [1994]
(27)
Identify a 
predictor of an 
anaphylactic 
reaction after an 
insect sting.
Prospective 
observational 
cohort study.
324 patients with a history of SR-anaphylaxis 
for wasp stings (272) or bees (52) were invited 
to the hospital for a provocation test. They 
considered sIgE measurement, skin tests, 
severity of previous reaction, sex, age, atopic 
constitution (history of eczema, hay fever, 
asthma, allergic reaction to drugs, X-ray 
contrast agents or food), histamine skin test 
results, location and number of previous 
stings, time from previous SR in association 
with the clinical severity of the reaction after 
hospital provocation sting test. The severity 
of SR was assessed according to the Müller’s 
grading system.
324 patients with a 
history of SR, Utrecht, 
Netherlands.
From 1982 to 1990.
Repeated SR after provocation test with a sting:
Bee: In 14 (27%) patients, SR was assessed according to the Müller’s grading system the same as in the previous 
reaction. In most patients, SR was less severe or non-existent after the provocation test.
Wasp: In 31 (11%) patients, SR was assessed according to the Müller test after the provocation test as in the previous 
reaction. In most patients, SR was less severe or non-existent after the provocation.
None of the 324 patients had a more severe reaction after the provocation test than the previous SR.
Patients sensitized to bee stings were more likely to have a severe relapse (grade III or IV) (31%) compared to those 
sensitized to wasp stings (13%).
Neither skin tests nor sIgE and sIgG4 were significantly correlated with the rate of allergic reaction in both the wasp 
and bee sting provocation test, taking into account the overall population.
The severity of the previous reaction, sex, atopy, histamine skin tests, placement of the previous sting, time between 
first reaction and the provocation test, and the number of previous stings – did not correlate with the severity of the 
reaction after the provocation test.
The severity of this reaction was significantly related to age. Patients with severe SR were significantly older on both 
the bee and wasp venom provocation tests.
The time interval between provocation and the onset of the reaction was related to the severity of the reaction. 
Patients with a more severe reaction developed symptoms and signs more quickly in the provocation test with both 
wasp and bee stings.
Logistic regression results:
Wasps: the elderly and sIgE levels, only in the population of these elderly patients, pose a higher risk of a recurrent 
severe reaction.
The severity of the first reaction poses an increased risk of a recurrent severe reaction.
Bees: the elderly and sIgE levels do not pose a greater risk of a recurrent severe reaction. The severity of the first 
reaction poses a higher risk of reacting again with a severe reaction.
They concluded that the 
current criteria for assessing 
hypersensitivity to insect venom 
(skin tests, sIgE, sIgG4) do not 
predict the incidence and severity 
of SR after insect sting provocation.
Bjornsson et 
al. [1995]
(28)
To identify 
patients at risk 
for a systemic 
allergic reaction 
after an insect 
sting.
A retrospective 
observational 
cohort study.
The study was conducted in 1,815 patients, 
of whom 52% were men and 48% were 
women, aged 20 to 44 years. The prevalence 
of sensitization to bee and wasp venom was 
assessed by measuring sIgE. Atopy and the 
presence of other allergic diseases were 
also considered. From 1,815 patients, 1,399 
patients were selected from the general 
population.
1.815 patients in 3 
areas in Sweden.
Between 1991 and 
1992
Among the subjects, 9.3% had sIgE for bee or wasp venom, 1.5% had SR on bee or wasp stings, and 0.6% had sIgE for 
both venoms.
Sensitization to bee or wasp venom was positively associated with atopy (hay fever, asthma, allergic rhinitis, skin 
eczema) (OR = 2.0; 95% CI = 1.4-2.8, p <0.0001), with male sex (OR = 1.8; 95% CI = 1.3-2.5, p <0.001) and age (OR = 2.0; 
95% CI = 1.4-2.8, p <0.001), and negatively with life in the northernmost part of Sweden (OR = 0.4 95% CI = 0.3-0.7, p 
<0.001).
Atopy has not been identified as a risk factor for a systemic reaction.
They found that the prevalence 
of allergies in Sweden is not as 
high as in other countries. People 
with atopy have a higher risk of 
becoming sensitized, but do not 
develop a systemic reaction more 
often than non-atopic people.
REVIEW ARTICLE
Online appendix - 3Evaluation of the risk factors for the rate of allergic reaction after bee and vespid sting in adults – a systematic review of the literature
Author, year of 
publication
The objective of 
the study
Research 
epidemiological 
basis
Methods of data collection Observed population/
environment/period
Results Main findings
Blaauw et al. 
[1996]
(29)
To determine 
whether an 
insect sting 
provocation test 
in a hospital 
can be used 
as a criterion 
for initiating 
immunotherapy.
A retrospective 
observational 
cohort study.
They included patients with a history 
of severe systemic reaction after bee or 
wasp stings, with a positive skin test and 
positive sIgE, in whom a provocation test 
was performed, which was negative. These 
patients were followed for 3 years after a 
negative provocation test.
With 479 patients, a new insect sting 
provocation test was performed. Patients 
with a negative response to the first 
provocation were asked about their 
experiences with stings in the natural 
environment. Surveys were sent to them 
and they were called by phone for additional 
information.
The doctor determined the grade according 
to Müller’s grading system.
479 patients (136 
sensitized with bee 
venom and 343 with 
wasp venom), The 
Netherlands.
Monitoring patients 
from 1979 to 1994.
Immunotherapy was advised for all patients with a history of severe reaction and a systemic reaction after the first 
provocation test.
Bees: Of the 136 patients, a positive response with SR 60 (44%) and a negative response to the first provocation were 
76 (56%). The SR grade after the provocation test was significantly different from the previous stings. A statistically 
significantly higher proportion of patients with a history of severe SR (III or IV) changed to a milder form of SR (II or I) 
upon provocation.
Of the 60 patients who responded to the SR provocation test, 56 received immunotherapy, and the remaining 4 were 
beekeepers who refused immunotherapy. 2 stopped beekeeping and 2 experienced natural unresponsive stings with 
SR. All 56 patients with VIT underwent another provocation test six months later. 51 subjects were protected and 5 
received further immunotherapy with a higher dose. Of these, 3 became protected, 1 stopped immunotherapy due to 
side effects, and 1 did not become desensitized. VIT for bees was successful in 96.4% of patients.
Of the 76 patients sensitized to bee venom with a negative first provocation test, 41 (53.9%) experienced a 
subsequent natural sting; 6 (14.6%) had a mild SR (I). A more severe SR (III) than the first in the anamnesis was 
reported by one patient. Characteristics of patients who did not resting after a negative provocation test and those 
who experienced a natural resting after a negative provocation were not statistically different in age, severity of SR in 
history, sIgE, and interval between the sting and the provocation test. However, there were several men in the group 
with a resting (beekeepers).
Wasps: Of the 343 patients, 59 (17.2%) had a positive SR response and 284 (83%) had a negative response to the first 
wasp venom test. The SR grade after provocation was significantly different from the previous stings. A statistically 
significant higher proportion of patients with a history of severe SR (III or IV) changed to a milder form of SR (II or I) 
upon provocation.
Of the 59 patients with SR, 58 went for VIT. 1 patient resigned without giving reasons. 3 patients withdrew due to 
severe adverse reactions.
All VIT patients underwent a provocation test again after six months. 50 were protected, 5 patients continued higher-
dose VIT, and 3 became protected. 1 patient refused further VIT due to side effects and 1 did not become desensitized. 
Thus, wasp VIT was successful in 91.4% of patients.
Of the 284 wasp-sting-sensitized patients with a negative first provocation test, 127 (44.7
%) later experienced a natural sting; 13 (10.2% had SR again, of which 9 (7.1%) SR had a mild (I, II) and 4 (3.1
%) patients had severe SR (III, IV). Of the 127 patients who experienced a natural wasp resting, 5 responded with 
the same severity as assessed by Müller’s grading system and the remaining 122 with a milder reaction than in their 
previous history.
Without the provocation test as a selection criterion for immunotherapy, the proportion of patients treated 
unnecessarily was calculated to be 45% for those sensitized to bee venom and 74% for those sensitized to wasp 
venom. However, with a negative provocation test as a selection criterion for immunotherapy discontinuation, 14.6% 
of bee-sensitized patients and 10.2% of wasp-sensitized patients remain at risk for a systemic reaction at a later sting.
Immunotherapy for bee or 
wasp venom is justified only 
after a positive response to the 
provocation test.
Annila et al. 
[1996]
(30)
Assess the 
prevalence and 
type of reaction 
after bee stings 
or wasps and to 
further identify 
potential 
risk factors 
for systemic 
reactions in 
beekeepers.
A retrospective 
observational 
cohort study.
An online questionnaire was sent to all 
members (274) of the local beekeeping 
association. 191 beekeepers (27 females, 
164 males) met the inclusion criteria. 
The presence of atopy was assessed by 
anamnesis (allergic rhinitis, allergic asthma, 
atopic dermatitis).
The severity of SR was assessed according to 
the Müller’s grading system.
191 beekeepers, 
Tampere, Finland.
September 1993.
A systemic reaction to bee stings was experienced by 50 (26%) beekeepers (13 (I), 15 (II), 15 (III), 7 (IV)) and 73 
(38%) beekeepers experienced a large local reaction. The association between patients' history and SR severity 
was statistically insignificant. 18 of them reacted with SR in the last season and 32, 2 years or more ago. After their 
last SR in the anamnesis, each subject was stung by a bee between 1 to 360 times. Beekeepers with a history of 
SR were statistically significantly younger and beekeepers less time than those without SR. 50 SR-responders had 
more apiaries and fewer stings in 1993 than the other 144, but this difference was not statistically significant. Body 
constitution and age at onset of beekeeping were not associated with SR severity. 45 (90%) beekeepers with SR used 
protective gloves. 97 (69%) of 141 non-responders used protective gloves. The difference was statistically significant. 
Sex and smoking were not statistically significantly associated with SR. 24 (48%) of those who reacted with SR, 39 (28
%) of other beekeepers had a history of atopic disease, which was statistically significantly associated with SR. During 
work in the apiary, 54% of those who reacted systemically and 23% of the others had nasal and ocular symptoms. 
Those who reacted systemically were younger and beekeepers for a shorter time. The results of
multiple-logistic regression had shown that the risk of a systemic reaction is 4 times higher in the presence of nasal 
and ocular symptoms, during work in the apiary, and 2 times higher for beekeepers who have been beekeeping for 
less than 15 years.
The SR reaction after the eighth sting was experienced by 2%, or a large local reaction by 13% of beekeepers.
Of the 50 systemic responders, 47 did not have a wasp sting. Three beekeepers responded with a systemic reaction to 
wasp stings. All three also had a history of a systemic reaction following a bee sting.
The occurrence of systemic and 
large local reactions after bee 
stings is high in beekeepers. A 
history of atopy is associated with 
a systemic reaction. Both; the 
presence of symptoms in the nose 
and eyes during work in the apiary 
and a history of beekeeping for 
less than 15 years, significantly 
increases the risk of a systemic 
reaction.
4 - Online appendix Zdrav Vestn | september – oktober 2020 | Letnik 89 | https://doi.org/10.6016/ZdravVestn.2973
PUBLIC HEALTH (OCCUPATIONAL MEDICINE)
Author, year of 
publication
The objective of 
the study
Research 
epidemiological 
basis
Methods of data collection Observed population/
environment/period
Results Main findings
Anilla et al. 
[1997]
(31)
Evaluate the 
usefulness of 
a structured 
questionnaire 
in addition to 
sIgE values 
in predicting 
beekeepers' 
systemic 
response to bee 
stings.
A retrospective 
observational 
cohort study.
Participants in previous studies completed 
a questionnaire on potential risk factors for 
a systemic response. They measured sIgE 
and prick skin tests before the beekeeping 
season. From the history of stings during the 
season, they assessed reactions to bee stings. 
A new questionnaire regarding reactions 
to stings during the season was sent to 
beekeepers after the end of the beekeeping 
season.
The severity of SR was assessed according to 
the Müller’s grading system.
78 beekeepers (7 
females, 71 males), 
Tampera, Finland.
The first 
questionnaires were 
sent in September 
1993.
Second 
questionnaires were 
sent in October 1994.
In 1994, 11 (14%) of 78 patients responded to SR after a bee sting. 1 was excluded due to a toxic reaction. 9 (90%) of 10 
had pre-season sIgEs higher than 1.0 kU/L, which meant there was a statistically significant
connection with SR. Before 1994, 8 (80%) patients had SR, which again meant a statistically significant association 
with sIgE. The presence of nasal and respiratory symptoms during work in the apiary was statistically significantly 
higher in the responders than in the non-responders. Positive skin tests, history of atopy, and large local reactions 
were more common in beekeepers with SR in 1994 than in others, but the differences were not statistically significant. 
Respondents also had more beehives and fewer stings per year, were younger, and beekeepers less time. An 
important risk factor is indicated: the presence of pre-seasonal serum sIgE at concentrations greater than 1.0 kU/l 
increased the risk of a systemic reaction by 12-fold. The risk was 10 times higher in the presence of nasal and/or 
respiratory symptoms when working in the apiary. If the beekeeper worked in the apiary for less than 8 years, the risk 
was 9 times higher, and with the anamnesis indicating a previous systemic reaction, the risk of a recurrent systemic 
reaction was 8 times higher.
The use of a more accurate 
patient history in combination 
with laboratory tests, especially 
measured sIgE, can significantly 
improve the reliability of the risk 
assessment for a systemic reaction 
to bee stings.
Golden et al. 
[1997]
(32)
To determine 
the natural 
course of poison 
sensitization by 
observing the 
rate of increase 
or decrease of 
sensitization in 
healthy adults 
at a follow-
up of 5 to 10 
years. Clinical 
significance of 
these results 
includes the 
frequency 
of systemic 
reactions to 
sting during the 
observation 
period.
Prospective 
observational 
cohort study.
520 volunteers, adult, light industry workers 
in the suburban area.
Two follow-up visits were planned. Three 
visits were planned for each subject. The first 
visit was scheduled after five years for all 520 
subjects. The second visit was scheduled 2 to 
3 years after the first and the third visit 5 to 
10 years after the first.
122 subjects did not respond to the invitation 
to another follow-up. 398 subjects responded 
at least once.
520 healthy 
volunteers, Baltimore, 
United States of 
America.
Of the 520 subjects identified at the beginning of the study, 122 did not respond to further visits. The follow-up survey 
was conducted on 398 volunteers (375 early visits and 205 later visits).
Of the 375 subjects who came for the first (early) visit, 87 had positive skin tests at baseline. 58 (67%) had positive 
skin tests after 2.5 years and 29 (33%) had negative skin tests after that time. At the second (later) visit (median of 6.8 
years), 11 (20%) of the 54 patients still had positive skin tests and 43 (80%) had negative skin tests. Despite the fact 
that skin tests became negative, sIgE remained positive at 11 (38%) of 29 at the first visit and 13 (30%) of 43 at the 
second visit.
Skin tests were negative at the first visit (early visit) in 288 of 375 patients and in 151 of 205 patients at the second 
(later) visit. At the first visit, 23 (8%) of 288 responders that had a negative skin test reacted with a positive skin test. 
By the second visit, 9 (6%) of 151 responders that had a negative skin test reacted with a positive skin test.
Overall, in 398 patients with at least one visit after an average of 4 years, skin tests changed from positive to negative 
in 44 (45%) patients, and with 98 patients, from negative to positive in 27 (8.7%) of 309 patients. For these 27 patients, 
the explanations are as follows: 8 had positive sIgE at baseline and 18 were stung by an insect between visits. In 
7 patients, there was no reason for a change from negative to positive skin tests. However, with patients who had 
positive skin tests during follow-up visits, we cannot attribute sensitization to intermediate stings between visits 
because the frequency of those stings was the same as with those patients whose skin tests became negative.
The risk of SR due to a sting was assessed in the correlation of skin tests and elapsed time. In patients with a positive 
skin test, 65 of them experienced stings and 11 (17%) responded with SR. There was no statistically significant 
difference in SR frequency between patients who experienced stings earlier than 4 and a half years after the start of 
the study, or later than 4 and a half years. There were 120 stings without a reaction in the negative skin test group.
The risk of SR in patients who had a positive skin test and a negative history of SR existed in 17% of patients. 
Therefore, 17% is statistically significantly less than 50% with those with a history of anaphylactic reaction, but 
statistically significantly higher than 2–5% with those who received VIT.
Asymptomatic sensitization to 
insect venom is common but 
transient and disappears at a rate 
of 12% per year. However, the 
risk of a systemic reaction with 
a negative history after resting is 
significant in adults with a positive 
skin test (17%) and increases if 
the positive skin tests persist for 
several years.
REVIEW ARTICLE
Online appendix - 5Evaluation of the risk factors for the rate of allergic reaction after bee and vespid sting in adults – a systematic review of the literature
Author, year of 
publication
The objective of 
the study
Research 
epidemiological 
basis
Methods of data collection Observed population/
environment/period
Results Main findings
Golden et al. 
[2001]
(33)
To determine the 
frequency and 
characteristics 
of patients with 
a negative skin 
test and systemic 
reaction after an 
insect sting.
Prospective 
observational 
cohort study.
The prevalence of a negative skin test 
in patients with a history of systemic 
insect post-sting reactions  was assessed 
prospectively. They also measured sIgE. In 
patients who signed consent, the results 
after repeated provocative insect stings were 
analysed.
307 patients, 
Baltimore, United 
States of America.
Of the 307 patients eligible for the study who had a positive history with a provocation sting test, 208 (68%) tested 
positive for insect venom (concentration above 1µg/mL).
In 36 (36%) of 99 patients with a positive history and a negative skin test, the sIgE results were low-level positive (1–3 
ng/mL), or the repeated skin tests were positive. SIgE was high-level positive (4–243 ng/mL) in 7 (7%) of those 99 
patients. 56 (57%) of these 99 patients had negative sIgE results.
A provocation sting test was performed in 51 of these 99 patients.
A provocation sting test was performed in 37 of 43 patients with a negative skin test and positive sIgE and in 14 of 56 
patients with a negative skin test and negative sIgE. 11 patients with a negative skin test responded with a systemic 
reaction to the provocation sting test: 2 had negative sIgE and 9 positive sIgE (7 with mild systemic reaction, 4 with 
moderate systemic reaction). All responders were allergic to wasp stings.
In patients with a positive skin test, a sting provocation test was performed in 141 of 196 patients. 30 (21%) responded 
with a systemic reaction.
The frequency of systemic reactions was 21% in patients with a positive skin test and 22% in patients with a negative 
skin test. 24% in patients with positive sIgE and 14% with negative sIgE.
The anamnesis of previous SRs was assessed as a possible factor in the reaction rate. There was no significant 
difference in the severity of the reaction in the history between patients with positive and negative skin test. In both 
groups, the history of mild SR was 25%; moderate in 55% and severe in 20% of patients. In patients with a negative 
skin test, 6 responded with a SR who had a history of mild SR, 4 with a history of moderate SR, and 1 with a history of 
severe SR.
In the two groups, there was no statistically significant difference in elapsed time since the last SR.
The response of skin tests after an 
insect venom sting may also be 
negative in those who will react 
with a systemic reaction.
Skin tests may also be negative in 
those with a history of systemic 
reaction after an insect sting and 
may be associated with positive 
serological tests for specific 
antibody (sIgE) venom. Skin 
tests for insect venom should be 
repeated during serological tests 
(sIgE).
Celikel et al. 
[2006]
(34)
To define the 
characteristics 
of reactions after 
insect stings and 
risk factors for 
the development 
of a systemic 
reaction in 
beekeepers in 
Turkey.
A retrospective 
observational 
cohort study.
A dedicated questionnaire was sent to 
1,250 beekeepers in 7 cities in Turkey. 494 
completed (39.6%) questionnaires were 
returned. The questionnaire contained the 
following questions: demographic data, 
history of bee stings, type of reaction, 
emergency room visit in the last 12 months, 
atopic diseases, smoking, drug and food 
allergies.
494 beekeepers (489 
men, 5 women) from 
7 selected Turkish 
places.
Between December 
2004 and June 2005.
444 beekeepers (89.9%) with a history of bee sting exposure in the last 12 months. 29 (6.5%) beekeepers experienced 
a systemic reaction, of which 9 (2%) had an anaphylactic reaction.
28 (5.7%) beekeepers had a history of emergency room visits, but only 5 in the last 12 months.
55% of beekeepers reported more than 100 stings in the last year. When monitoring the systemic response with 
respect to age and time of beekeeping in the logistic regression model, the results were significantly related to the 
systemic response: seasonal rhinitis (OR = 4.4; 95% CI = 1.2–11.5), persistent rhinitis (OR = 4.6; 95% CI = 1.2–18.2), food 
allergy (OR = 7.0; 95% CI = 2.0–25.0), asthma (OR = 8.0; 95% CI = 2.5–25.6) any other atopic disease (OR = 3.3; 95% CI = 
1.2–8.7) and two or more concomitant atopic diseases (OR = 10.9, 95% CI = 3.5–33.8).
The incidence of systemic reaction 
in Turkish beekeepers is low, which 
may be due to the protective effect 
of frequent bee stings. The risk 
of a systemic reaction increases 
approximately 3-fold in the 
presence of one atopic disease and 
11-fold in the presence of two or 
more atopic diseases compared to 
beekeepers without atopic disease.
Guenova et al. 
[2009]
(35)
To examine 
serum tryptase 
concentration 
as a diag-nostic 
parameter 
to assess an 
individual's risk 
and its impact 
on the severity 
of the allergic 
reaction in the 
elderly.
Prospective 
cohort 
observational 
research.
Serum tryptase was measured in 274 patients 
allergic to bee or wasp venom, with a history 
of systemic reaction, positive skin test, and 
positive sIgE who visited the dermatology 
department in Tübingen, Germany.
The SR level was assessed according to 
Müller.
274 patients (153 
men, 121 women), 
Tübingen, Germany.
Between 2004 and 
2006.
54 patients had a history of grade I SR, 106 of grade II, 66 of grade III and 48 of grade VI according to Müller’s grading 
system. Basal tryptase levels were elevated in 30 (10.9%) patients (> 11.4 µg/l), of whom only 4 (7.5%) were diagnosed 
with mastocytosis.
The severity of SR according to Müller was statistically significantly increased with elevated basal tryptase levels (p 
= 0.0003). The mean basal tryptase was 4,274 µg/l in grade I responders, while the mean basal tryptase in grade IV 
responders was 7.18 4 µg/l. Basal tryptase concentrations above 11.44 µg/l were measured in only 7% of grade I and 
21.4% of grade IV responders.
The risk of severe SR was statistically significantly increased in the elderly. Patients with grade I reaction were on 
average 38.9 years old, patients with IV. degree 51 years and a half.
Sex was not statistically significantly associated with basal tryptase levels.
The article did not assess whether recurrent SR was more severe or lesser than the first SR in the anamnesis.
The results confirmed that serum 
tryptase levels are a risk factor 
for severe systemic reactions to 
insect stings. There is evidence that 
serum tryptase levels increase with 
age and are thereby an indicator 
of increased mast cell load or 
reactivity,
which may be partly the source 
for the reported severe allergic 
reactions in the elderly. Because 
these patients are at higher risk 
for severe anaphylactic reactions, 
they are potential candidates for 
immunotherapy.
6 - Online appendix Zdrav Vestn | september – oktober 2020 | Letnik 89 | https://doi.org/10.6016/ZdravVestn.2973
PUBLIC HEALTH (OCCUPATIONAL MEDICINE)
Author, year of 
publication
The objective of 
the study
Research 
epidemiological 
basis
Methods of data collection Observed population/
environment/period
Results Main findings
Blum et al. 
[2010]
(36)
Assess the 
impact of total 
serum IgE and 
other potential 
risk factors 
on severity of 
allergic systemic 
reaction after 
insect sting.
Retrospektivna 
opazovalna 
kohortna 
raziskava
Of the 1,002 patients in the retrospective 
analysis who had an allergy to insect stings 
for more than 5 years, 865 reported a 
systemic allergic reaction, most commonly 
after a bee or wasp sting, and 137 a large 
local reaction or toxic reaction.
Of the 865, 758 had cIgE, sIgE, and basal 
tryptase levels available for analysis, along 
with a history of atopy (skin prick tests for 
the 14 most common allergens), age, and sex 
correlated with the severity of the allergic 
reaction.
The severity of SR was assessed according to 
Müller’s grading system.
1,002 patients, Bern, 
Switzerland.
From January 2003 to 
December 2007.
In the group of 865 patients with a systemic reaction, 256 (33.6%) had a reaction after a wasp sting, 212 (28.1%) after 
a bee sting, and 290 (38.3%) failed to identify the insect that stung them. Of the 758 patients, 547 had a positive skin 
test with bee venom and 651 with wasp venom.
Of the 758, 519 responded with a severe systemic reaction (grade III or IV) and 239 with a mild systemic reaction 
(grade I or II). 61% had a double-positive skin test for wasp and bee venom, and 65% had a double-positive sIgE.
They were unable to show a statistically significant effect of cIgE on the severity of the reaction (grade IV) with logistic 
regression models using categorized cIgE values. There was only a slight tendency for lower cIgE levels in patients 
with grade IV response rate and higher cIgE levels in patients with grade III reaction rate.
There was also no statistically significant increase in sIgE in patients who responded with grade IV in the use of 
categorized sIgE values with logistic regression models. However, a statistically significant association of higher sIgE 
levels and the response of patients with grade IV response was found in the values for bee venom sIgE when using 
bivariate analysis with the Wilcoxon test.
Statistical analysis showed a positive association of SR severity with older age (> 50 years).
The severity of the systemic allergic reaction was statistically significantly associated in patients with basal tryptase 
levels > 11.4 µg/l (p = 0.0001). However, higher basal tryptase concentrations were statistically significantly associated 
with age.
Atopy affected the severity of the reaction. Severe SR (grade III) was reported statistically significantly several times 
in atopic patients. Atopy was strongly associated with higher cIgE levels. Atopy was not statistically significantly more 
common in elderly patients.
The association of severe systemic 
response with lower cIgE is also 
related to age. The presence of 
cardiovascular disease and higher 
basal tryptase levels also have a 
significant impact on the elderly, 
and both are risk factors for a more 
severe reaction.
Richter et al. 
[2011]
(37)
To identify the 
factors that 
predict systemic 
response 
in British 
beekeepers and 
to study visits 
to a specialist 
after a systemic 
reaction 
triggered by a 
bee sting.
A retrospective 
observational 
cohort study.
A questionnaire was published in the 
monthly magazine and on the website 
of the British Beekeepers Association. 
Information included: demographics, district 
of residence, history of allergy to bee venom 
in the family, number of years of beekeeping, 
number of stings per year, time from start of 
beekeeping to first reaction, prophylactic use 
of antihistamines, current atopic diseases, 
cardiovascular and respiratory diseases, 
severity of systemic reaction, examination by 
an allergist, prescribed adrenaline injection 
and immunotherapy.
852 beekeepers, 
United Kingdom.
852 beekeepers responded, of which 63% were men, most of them aged 51 to 60 years. 28% of all involved 
experienced a large local reaction and 21% a systemic reaction. Factors characteristic of beekeepers with a systemic 
reaction were: female sex, positive family history of insect venom allergy, more than two years of beekeeping before 
the systemic reaction, and taking antihistamines before the start of beekeeping. 44% of beekeepers with a systemic 
reaction visited the emergency room, 16% were examined by a specialist and only 8% have an adrenaline self-injector 
with them at all times.
The analysis identified some new 
factors related to the systemic 
response: the levels of emergency 
room and specialist allergist visits 
and adrenaline supplies were low, 
indicating the need to educate 
beekeepers and doctors and other 
healthcare professionals.
REVIEW ARTICLE
Online appendix - 7Evaluation of the risk factors for the rate of allergic reaction after bee and vespid sting in adults – a systematic review of the literature
Author, year of 
publication
The objective of 
the study
Research 
epidemiological 
basis
Methods of data collection Observed population/
environment/period
Results Main findings
Sturm et al. 
[2014]
(38)
To highlight 
the importance 
of clinically 
insignificant 
sensitization 
by deliberately 
provoking a 
sting test and 
monitoring 
serological 
changes over a 
period of two 
years.
Prospective 
observational 
cohort study
A sting provocation test was performed 
in 94 subjects (44 women) with hitherto 
insignificant (no systemic reaction) 
sensitization (presence of sIgE for 
insect venom). In subjects with double 
sensitization, a provocation test with both 
points was performed on the same day. 
After the provocative test, they repeated 
serological tests. Clinical outcome was 
correlated with sIgE, skin tests, and basophil 
activation test. SIgE levels were monitored 
after three hours, one week, four weeks, and 
one year.
The systemic response was assessed 
according to Ring and Messemer.
94 subjects, Graz, 
Austria.
48 (51.1%) subjects had atopy, 27 (28.7%) had a history of large local reaction.
Of the 94 provocations, 41 subjects have been stung by a bee, 16 by wasps and 37 by both insects in the past.
Only 6 (5.3%) subjects had a systemic reaction (2 per bee sting, 2 per wasp sting, and 1 on both stings), 4 subjects had 
a grade I reaction, 1 subject grade II reaction and 1 subject grade IV reaction. There is no comparison in the article 
regarding the degree of severity of the reaction after the provocation test or the previous one in the anamnesis. 41 
(43.6%) subjects had a large local reaction after the provocation test, 16 (39%) of them had a large local reaction 
already in the anamnesis. Compared to the general population, there was a 9.5-fold higher risk of large local reaction, 
but nothing greater for a systemic reaction. Subjects with a history of LLR did not have a statistically significant higher 
risk of SR after provocation than subjects without a history of LLR.
Of the 37 subjects with low cIgE levels, 1 (2.7%) reacted with SR; of the 29 subjects with moderately high cIgE levels, 
1 (3.4%) reacted with SR; of the 28 subjects with high cIgE levels, 3 (10.7%) responded with SR. It follows that there is 
no statistically significant difference between the different levels of cIgE and SR.
sIgE and skin tests after provocation with a systemic reaction for the wasp sting were positive for all subjects.
sIgE after a provocation test with a systemic reaction for bee stings were positive in only 1 of 3. The frequency of 
positive serological tests after negative provocation was not statistically related to SR after provocation.
A large local reaction was less common after bee stings, in 26 (33.3%) of 78 subjects, and in wasp stings in 29 (54.7%) 
of 53 subjects, but the difference was not statistically significant.
Three hours after the provocation sting test, sIgE levels were reduced, but in none of the 89 subjects were the results 
negative. After 1 week, sIgE levels increased (2.2-fold for wasp venom and 2.7-fold for bee venom) and 4 weeks after 
provocation, by more than 3.5-fold for both venoms. (0.2 to 34.0). In systemic responders, only 3 of 5 cases could be 
analysed, but the increase in sIgE was similar.
To assess the clinical relevance of this increase, 18 subjects were selected for repeated provocation test after one 
year (16 bee venom, 1 wasp venom, 1 both venoms). Again, 50% had a large local reaction and none had a systemic 
reaction. The increase in sIgE was similar to that after the first provocation, 4 weeks after the sting.
cIgEs were not statistically significantly reduced in either systemic response or large local response agents.
Although sensitization to insect 
venoms is common, the risk of a 
systemic reaction for sensitized 
persons is low. An increase in sIgE 
after a provocation test is not a 
true indicator of conversion to 
symptomatic sensitization.
Currently available tests cannot 
distinguish between a symptomatic 
sensitization, large local reaction, 
and systemic reaction.
Legend: LLR - large local reaction, SR - systemic reaction, cIgE - total IgE, sigE - specific IgE, sIgG - specific IgG, cIgG - total IgG, OR - odds ratio, 95% CI- 95% confidence interval, VIT - Venom Immunotherapy.