Side Effects of Allergy Injections: A Clinical Evidence Review

The number needed to harm (NNH) for allergy injections varies by reaction severity. Derived from AAAAI/ACAAI National Surveillance data (Bernstein 2010; 54 million injection visits 2008-2016): NNH for any systemic reaction is approximately 100 injection visits, reflecting the 0.1% per-visit rate. NNH for Grade 3 severe reaction (laryngeal edema or severe bronchospasm) is approximately 300,000 visits, based on the 0.3 per 10,000 visit rate. NNH for Grade 4 anaphylaxis is approximately 160,000 visits. NNH for fatal reaction is approximately 9 million visits in the most recent surveillance period. These NNH figures are context for informed consent discussions and risk-benefit assessment, particularly when comparing SCIT to lower-risk immunotherapy modalities.

The WAO Subcutaneous Immunotherapy Systemic Reaction Grading System, published by Cox et al. in JACI 2010 and endorsed by AAAAI, ACAAI, SLAAI, and APAAACI, classifies systemic reactions by the most severe organ system involved. Grade 1 involves one organ system only — cutaneous (generalized urticaria, flushing), upper respiratory (rhinitis, sneezing), or conjunctival. Grade 2 involves more than one organ system, or lower-respiratory asthma responding to bronchodilator, or gastrointestinal symptoms. Grade 3 covers severe asthma with FEV1 or PEF drop of 40% or more not responding to bronchodilator, or upper-airway laryngeal or uvular edema. Grade 4 is anaphylaxis with respiratory failure or cardiovascular collapse. Grade 5 is death. The 2024 WAO update by Turner et al. aligned Grades 3-5 with CoFAR v3.0 clinical anaphylaxis criteria warranting epinephrine.

Approximately 3% of allergy injection systemic reactions are classified as severe (WAO Grade 3) and less than 0.5% reach Grade 4 anaphylaxis, based on Year 1 AAAAI/ACAAI surveillance data (Bernstein 2010; 8.1 million injection visits, 8,502 systemic reactions reported). Translated to absolute incidence: Grade 3 reactions occur at approximately 0.3 per 10,000 injection visits (1 per 300,000 visits). Grade 4 anaphylaxis occurs at approximately 1 per 160,000 visits (Bernstein/Epstein 2020, Allergy Asthma Proc). The majority of systemic reactions — 74% — are Grade 1 (cutaneous only), which are self-limited and rarely require epinephrine. These proportions have been consistent across multiple surveillance years from 2008 to 2017.

Antihistamine premedication reduces the frequency and size of large local reactions (LLRs) but does not reduce systemic reaction incidence. Reimers et al. (JACI 2000) demonstrated benefit of antihistamine premedication for local reactions in a controlled study, but the effect on systemic reaction rates was not significant. The 2011 AAAAI/ACAAI Practice Parameter Third Update does not recommend routine antihistamine premedication for conventional SCIT because masking of early cutaneous warning signs may delay recognition of a progressing systemic reaction. Premedication is more consistently recommended for cluster and rush build-up protocols, where per-injection systemic reaction rates are significantly higher and pre-injection H1 plus H2 antihistamine plus leukotriene antagonist combinations have shown benefit in reducing but not eliminating systemic reactions.

Evidence-graded risk factors for severe allergy injection systemic reactions include: uncontrolled asthma (Level B evidence per Practice Parameter — present in 88% of SCIT fatalities in Bernstein 2004 survey); prior systemic reaction history (Level B — Roy 2007 found fourfold higher rate of subsequent SRs); beta-blocker therapy (Level C — impairs epinephrine efficacy, complicates treatment); pollen season dosing (Level C — Epstein 2013 Year 3 surveillance found practices that always reduced doses during peak season had 44% fewer Grade 2-3 SRs versus 65% in non-reducing practices); cluster or rush build-up protocols (Tversky 2022, JACI: 3.3-fold higher per-injection SR rate versus conventional); and new vial transitions. Mast cell disorders with elevated baseline tryptase represent an emerging high-risk category.

Delayed systemic reactions — defined as beginning more than 30 minutes after injection — account for approximately 15% of all systemic reactions in AAAAI/ACAAI surveillance (Epstein 2011, Year 2 analysis; Epstein 2019, 54-million-visit update). Grade-specific delayed proportions: 15% of Grade 1, 10% of Grade 2, and 12.5% of Grade 3 reactions begin after the standard 30-minute observation window. A single-center 10-year analysis (Larenas-Linnemann 2017, JACI Pract; 57,102 injections) reported a notably higher 52.8% of systemic reactions occurring after 30 minutes, with seven beginning 90 or more minutes post-injection. This outlier suggests practice-level variation in timing distributions. The 30-minute window captures the great majority of severe reactions but not all; EAACI recommends 60 minutes for rush protocols, unstable asthma, and high-degree hypersensitivity patients.

Fatal allergy injection reaction rates have declined substantially across three major surveillance periods. Lockey et al. (JACI 1987) estimated approximately 1 fatal reaction per 2.8 million injection visits for the period 1945-1984, based on 30 IT and skin-testing deaths. Reid et al. (JACI 1993) reported 17 deaths in 1985-1989, approximately 1 per 2 million injection visits. Bernstein et al. (JACI 2004) identified 41 fatalities from 1990-2001, approximately 1 per 2.5 million visits or 3.4 deaths per year. Epstein 2019 (JACI Pract; 54.4 million visits 2008-2016) reported approximately 0.8 deaths per year, roughly 1 per 9 million injection visits — a 3.75-fold reduction since the 1990-2001 period. This improvement is attributed to systematic adoption of the AAAAI/ACAAI Practice Parameter: pre-injection asthma screening, peak-season dose reduction, and the 30-minute observation mandate.