Best Approaches to Identifying Patients With Acne Using ICD-10 Codes

A search for 1 ICD-10 code for acne or 1 ICD-10 code coupled with an acne prescription in administrative databases may yield a more accurate return than looking for 2 ICD-10 codes for acne in identifying patients with acne for the purposes of acne research, study findings published in the Journal of the American Academy of Dermatology suggests.

This retrospective study included 300 patients (age range, 12-40 years) who were evaluated in outpatient dermatology clinics between January 1, 2016, and June 30, 2019. All patients had ≥1 ICD-10 code for acne. Another 300 randomly selected patients from the same clinical population without an ICD-10 code for acne were included in a control cohort.

The investigators performed a focused chart review for each patient to identify whether a documentation of clinical acne was made. Clinical acne was defined as the presence of comedones, papules, pustules, or cysts in the physical exam or a discussion of acne in the clinical encounter note. The specificity, sensitivity, negative predictive value, and positive predictive values were compared in 3 classification strategies: 1 ICD-10 code for acne, ≥2 ICD-10 codes for acne within 6 months, and 1 ICD-10 code and a prescription for an acne medication within a 6-month period.

The positive predictive value (PPV), negative predictive value (NPV), sensitivity, and specificity for using 1 ICD-10 code for acne were 0.990, 0.977, 0.977, and 0.990, respectively. Similar findings were reported for the PPV, NPV, sensitivity, and specificity of using ≥2 ICD codes for acne within 6 months (1.000, 0.751, 0.615, and 1.000, respectively). For the use of 1 ICD-10 code and a prescription for an acne medication within 6 months, the PPV, NPV, sensitivity, and specificity were 0.993, 0.977, 0976, and 0.993, respectively.

Using 1 code, 2 codes, and 1 code plus a medication prescription were associated with decreases in the estimated sensitivities to 0.812, 0.670, and 0.804, respectively. This was found if the frequency of acne encounters was 15%, which the researchers explained is the general frequency of acne encounters in the community.

A limitation of this study included the retrospective reliance on administrative claims.

The ideal method for accurate case identification is both sensitive and specific when using administrative databases, the investigators wrote. Regarding their suggested search strategy, they believe that additional “studies should examine whether these findings generalize to acne patients seen outside of dermatology.”

Reference

Barbieri JS, Weiner DM, Kakpovbia E, Nagler AR. Validating the optimal classification approach using ICD-10 codes to identify dermatology patients with acne [published online July 22, 2020]. J Am Acad Dermatol. doi: 10.1016/j.jaad.2020.07.075