Elsevier

Annals of Epidemiology

Volume 16, Issue 9, September 2006, Pages 696-700
Annals of Epidemiology

ICD-9-CM Coding of Emergency Department Visits for Food and Insect Sting Allergy

https://doi.org/10.1016/j.annepidem.2005.12.003Get rights and content

Purpose

Little is known about the role of International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes for identification of specific allergic reactions in the emergency department (ED).

Methods

Investigators in 10 EDs reviewed 1395 charts of consecutive patients presenting with food allergy (ICD-9-CM codes 693.1 and 995.60 to 995.69) and insect sting allergy (code 989.5). They also reviewed charts of patients with “unspecified” allergic reactions (codes 995.0 [other anaphylactic shock] and 995.3 [allergy, unspecified]) to identify additional patients with food or insect sting allergy.

Results

Of 406 patients with food allergy, 216 patients (53%) were coded as food allergy, whereas the remaining 190 patients (47%) were not. Of 394 patients with insect sting allergy, 341 (87%) were coded as insect sting allergy, whereas 53 patients (13%) were not. Characteristics of ICD-9-CM–identified compared with chart-review–identified patients differed for both food and insect sting allergy. ICD-9-CM–identified patients with food allergy were less likely to experience anaphylaxis.

Conclusions

Almost half the patients with food allergy would have been missed by using food-specific ICD-9-CM codes alone, whereas only 13% of patients with insect sting allergy would have been missed. Furthermore, characteristics of these allergy patients would have been biased by studying only patients identified by using the allergen-specific ICD-9-CM codes.

Introduction

Ascertainment of cases for epidemiologic studies performed in clinical settings often involves the use of International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes. Many studies examined the accuracy of identifying disease cases by ICD-9-CM code 1, 2, 3, 4, 5 and found that the ICD-9-CM classification system has several limitations, including diagnostic codes that do not completely encompass the condition of interest and conditions being distributed across multiple codes (3). These studies found that identification of soft tissue disorders of the neck and upper limbs (2), ischemic cerebrovascular disease (3), acute ischemic stroke (1), pneumococcal pneumonia (5), and cerebrovascular malformations (4) by using ICD-9-CM codes often yielded incomplete or inaccurate case lists and concluded that ICD-9-CM codes should not be the only source used to identify cases of disease.

In 1993, the ICD-9-CM system introduced codes to identify specific food allergies. To our knowledge, no study has examined the accuracy of ICD-9-CM coding for allergic reactions or anaphylaxis. Additionally, little is known about how frequently these specific codes are used to identify specific allergic reactions, such as food allergy or insect sting allergy, in the emergency department (ED). The objective of the current study is to determine the accuracy of using ICD-9-CM codes to identify ED visits for food and insect sting allergies and determine the potential bias that might be introduced by sole reliance on ICD-9-CM codes.

Section snippets

Methods

This retrospective cohort study was performed as part of the Multicenter Airway Research Collaboration, a division of the Emergency Medicine Network (EMNet) (6). Using a standardized protocol and data abstraction forms, investigators at 10 EDs examined two cohorts, one of ED visits for food allergy and one of ED visits for insect sting allergy, each during any 1-year period between January 1999 and November 2001. Sites reviewed charts of randomly selected patients who presented with ICD-9-CM

Food Allergy

Of 775 consecutive patients with an acute allergic reaction, 216 patients (28%) were coded as having a food-related allergic reaction. Of the remaining 559 patients with unspecified allergy, 190 patients (34%) were identified as having probable or definite food allergy based on chart review. Thus, only 53% (95% confidence interval, 48%–58%) of patients with a food-related allergic reaction were identified by specific ICD-9-CM codes. The number of patients identified by each ICD-9-CM code is

Discussion

To our knowledge, we present the first study assessing the accuracy of using ICD-9-CM codes to identify ED visits for food-related and insect sting-related acute allergic reactions. The use of ICD-9-CM codes in epidemiologic studies of acute allergic reactions and anaphylaxis requires that the codes accurately and completely identify cases to ensure valid conclusions. In our cohort of patients with food and insect sting allergy, specific ICD-9-CM codes identified only 53% of patients with food

References (11)

  • S. Clark et al.

    Multicenter study of emergency department visits for food allergies

    J Allergy Clin Immunol

    (2004)
  • S. Clark et al.

    Multicenter study of emergency department visits for bee sting allergy

    J Allergy Clin Immunol

    (2005)
  • L.B. Goldstein

    Accuracy of ICD-9-CM coding for the identification of patients with acute ischemic stroke: Effect of modifier codes

    Stroke

    (1998)
  • R. Buchbinder et al.

    Lack of concordance between the ICD-9 classification of soft tissue disorders of the neck and upper limbs and chart review diagnosis: One steel mill's experience

    Am J Ind Med

    (1996)
  • C. Benesch et al.

    Inaccuracy of the International Classification of Diseases (ICD-9-CM) in identifying the diagnosis of ischemic cerebrovascular disease

    Neurology

    (1997)
There are more references available in the full text version of this article.

Cited by (83)

  • Anaphylaxis among elderly emergency department patients in a large health system in New York

    2022, Annals of Allergy, Asthma and Immunology
    Citation Excerpt :

    It is unclear how much of these deficiencies are because of under-recognition and underdiagnosis vs undertreatment or some combination of these factors. Although older adults are less likely to be prescribed EAI after presenting to the ED,31 no difference has been reported in epinephrine-dispensing patterns between the sexes in the elderly population.46 There are sparse data on whether older adults are at increased risk of adverse cardiac events from epinephrine, and more research is needed to determine what risk factors might predispose the elderly to such adverse outcomes.31,34,51,56

View all citing articles on Scopus

S.C. is supported by grant no. T32 ES07069 from the National Institute of Environmental Health Services (Research Triangle Park, NC) and C.A.C. is supported by grant no. AI52338 from the National Institute of Allergy and Infectious Disease (Bethesda, MD). This project was supported by unrestricted grants from Dey Laboratories (Napa, CA) and the Food Allergy and Anaphylaxis Network (Fairfax, VA).

View full text