Objectifying the conjunctival provocation test: photography-based rating and digital analysis

• Published in: International archives of allergy and immunology Vol. 163; no. 1; p. 59
• Main Authors: Dogan, S, Astvatsatourov, A, Deserno, T M, Bock, F, Shah-Hosseini, K, Michels, A, Mösges, R
• Format: Journal Article
• Language: English
• Published: Switzerland 2014
• Subjects: Adult; Allergens - administration & dosage; Allergens - immunology; Clinical Trials as Topic; Complex Mixtures - administration & dosage; Complex Mixtures - immunology; Conjunctivitis, Allergic - complications; Conjunctivitis, Allergic - diagnosis; Conjunctivitis, Allergic - immunology; Conjunctivitis, Allergic - therapy; Female; Humans; Image Interpretation, Computer-Assisted; Immunotherapy - methods; Male; Middle Aged; Monitoring, Immunologic - instrumentation; Monitoring, Immunologic - standards; Photography - instrumentation; Pollen - chemistry; Regression Analysis; Rhinitis, Allergic, Seasonal - complications; Rhinitis, Allergic, Seasonal - diagnosis; Rhinitis, Allergic, Seasonal - immunology; Rhinitis, Allergic, Seasonal - therapy; Severity of Illness Index
• ISSN: 1423-0097
• DOI: 10.1159/000355333

Patients with allergic rhinoconjunctivitis are susceptible to both nasal and ocular symptoms. The conjunctival provocation test (CPT) is an established diagnostic procedure used in allergic rhinoconjunctivitis, particularly to document a patient's current reactivity to allergens. To date, there are no international guidelines defining the CPT. No approved evaluation method exists for interpreting CPT results. This paper aims to establish the digital analysis of macroimages as an objective, validated and standardized method for interpreting CPT results.
In a clinical immunotherapy trial with 155 patients, treatment progress was documented based on the CPT. Local investigators used a symptom score to grade tearing, reddening and the patients' subjective perception of symptoms (mucosal irritation). A central observer rated conjunctival hyperemia via digital photography. Digital image analysis software was utilized to determine conjunctival hyperemia.
Spearman's correlation between the local investigators' and the central observer's ratings was r = 0.729 (p < 0.001); the percentage of total agreement was 48% (based on 739 photos). Digital image analysis (based on 48 photos) had a high percentage of total agreement with the central observer's ratings (69%) but a low percentage of total agreement with the investigators' ratings (38%). The corresponding correlations were r = 0.264 and 0.064, respectively.
Photography-based rating by a central observer may represent a valuable supplement to the local investigator's assessment for making an objective evaluation of CPT results. Digital image analysis possesses the potential of being an objective evaluation method compared to the wide-spread subjective evaluation by the investigators.