In practice, atopic dermatitis is commonly diagnosed based on the clinical presentation and either a good response to glucocorticoid therapy or a positive intradermal test or serum test for allergen-specific IgE. Unfortunately, glucocorticoid therapy leads at least to temporary improvement in many skin diseases. Positive skin or serum test results have been documented in normal animals or patients affected with skin diseases other than allergies and are thus absolutely unsuitable to diagnose atopic dermatitis.

Thus, rather than confirming a suspicion of the disease with one or several laboratory test results, the clinician has to look at the history and clinical signs and rule out all differential diagnoses for that particular case to make the diagnosis. This requires a lot of intellectual, emotional and financial dedication of the owner and a thorough understanding of a logical workup by the attending clinician. In some patients, due to owner constraints, the clinician will have to resort to guessing the most likely diagnosis and performing trial therapy. There is nothing wrong with that approach as long as clinician and owner are aware of the limitations and risks involved in such an approach.

If, however, an owner is interested in the best long-term strategy, only a proper diagnosis will allow an efficient and efficacious treatment of the pet's skin condition, and a diagnostic workup needs to be recommended. Each differential diagnosis needs to be excluded with appropriate procedures and tests/treatments. The diagnosis is atopic dermatitis if all other diagnoses have been ruled out.

The primary clinical presentation of canine atopic dermatitis is pruritus. Secondary skin lesions result from self-trauma and include alopecia, crusting and lichenification. The most commonly affected areas are the face, feet, ventral abdomen/inguinal region, axillae, flexor surface of the tarsal joint/extensor surface of the carpal joint, pinnae and external ear canals. When the pinnae and external ear canals are affected, otitis externa is a common complication. Secondary infections with yeast or bacteria are frequently seen. Breed predispositions are reported, but may vary from location to location. However, German shepherd dogs, Golden and Labrador retrievers as well as West Highland White terriers are predisposed in most reports.

In cats, a number of cutaneous reaction patterns have been associated with atopic dermatitis. Miliary dermatitis, noninflammatory alopecia, head and neck pruritus, and eosinophilic granuloma all have been reported.

For most dogs with atopic dermatitis, food adverse reaction is a possible differential diagnosis. The only reliable way of identifying these patients at this point in time is an elimination diet. Superficial skin scrapings, Sarcoptes titres and trial therapy for scabies are all indicated for a dog with suspected atopic dermatitis. Typically, demodicosis in early stages and without a significant secondary infection is not associated with severe pruritus. However, any dog with pododermatitis and any dog with lesional pruritus should undergo deep skin scrapings to rule out demodicosis. Secondary infections are common and cytology is an important tool to identify organisms involved in clinically suspected infections. Appropriate antimicrobial therapy must be followed up with clinical and microscopic evaluation. Complete resolution of pruritus and lesions in an older dog will raise the possibility of hormonal disease with secondary infections. In a young dog this is unlikely, and stress, parasites or allergies are common causes for infections. If the pruritus persists in face of lesion resolution, allergies as the underlying cause of infections are most likely in the younger dog. If, however, lesions and pruritus persist although microorganisms are absent on cytology and ectoparasites are ruled out with skin scrapings and trial therapies, a biopsy should be considered as immune-mediated disease, neoplasias or unusual skin diseases need to be ruled out.

In the cat, miliary dermatitis is characterized by numerous, small, localized or generalized papules and crusted papules. It is the most common feline problem in small animal dermatology and the list of differential diagnoses is lengthy. Allergies such as fleabite hypersensitivity, atopy or food intolerance, ectoparasites such as Otodectes cynotis, Notoedres cati, or Cheyletiella blakei, infections with dermatophytes or bacteria, neoplastic diseases such as mast cell tumours, immune-mediated diseases such as pemphigus foliaceus and nutritional shortcomings such as essential fatty-acid deficiencies may all cause a papulocrustous dermatitis.

Noninflammatory alopecia may also be caused by a variety of conditions. It is important to distinguish between lack of hair growth and hair removal, and trichogram can be a very useful diagnostic tool. Baby suits and E-collars can also help determine this, but they take some time and not all cats (or owners!) will participate. If the cat licks the hair out, fleabite hypersensitivity, food intolerance and atopic dermatitis all may be involved. Very rarely do we see psychogenic alopecia based on a classic history and personality type. If the hair falls out, we may deal with dermatophytosis, anagen defluxion (where severe metabolic diseases or chemotherapy interfere with hair production and lead to structurally weak hair shafts that may break, leading to alopecia), telogen effluvium (where a severe stress 4–12 weeks prior to the hair loss pushed all the hair from the growth or anagen into the resting or telogen phase and now the new hair growth pushes out the old hair leading to clinical alopecia) or hormonal disease such as hypothyroidism or hyperadrenocorticism, both of which are extremely rare diseases in cats.

Head and neck pruritus in the cat may also be due to a number of different causes. Allergies such as fleabite hypersensitivity, atopy or food intolerance; ectoparasites such as Otodectes cynotis or Notoedres cati; infections with dermatophytes or bacteria; neoplastic diseases such as mast cell tumours; or immune-mediated diseases such as pemphigus foliaceus may all cause severe pruritus of the head. Thus, ectoparasite treatment trials and elimination diet are always indicated; a biopsy may be performed to rule out neoplastic or immune-mediated disease in some selected patients. Similarly, in cats with eosinophilic granuloma complex lesions, a flea control trial and elimination diet may be useful.

A logical workup will confirm or exclude the diagnosis of atopic dermatitis in most patients and allow us to choose the best long-term management for that particular patient. However, owner constraints are an unfortunate but regular feature of practice. If owners refuse the diagnostic workup we offer despite our efforts to explain the limits of educated guesses, we will have to rely on our clinical experience and a good guess in those patients.

Serum tests for allergen-specific IgE and skin testing have all been advocated as diagnostic aids, but neither is a good way to diagnose atopic dermatitis! It should be diagnosed based on the history and clinical signs as well as diagnostic trials to rule out diseases with similar clinical presentation.

There are only three reasons to identify the offending allergens in an atopic animal:

 Curiosity

 Some owners will simply want to know what their dog is allergic to without any effect on treatment of the condition.

 Avoidance - This is initially the most common reason mentioned by owners in support of further testing. However, it is extremely rare that we can avoid allergens that are airborne.

 Immunotherapy - Allergy shots or immunotherapy is the major reason for us to recommend identifying the offending allergens in an atopic animal.

Skin Testing

The allergens used in the test should be chosen according to their importance in the respective environment. A positive skin test result indicates only the presence of specific IgE on mast cells in this patient and has to be interpreted in light of the patient's history and environment. If one cannot perform at least 1–2 skin tests per week, evaluation of weak reactions may not be very reliable. In addition, allergens have a very short half-life (approximately 2–6 weeks), once diluted to the concentration used for skin testing. Thus, skin testing is financially viable usually only if at least 1–2 skin tests per week allow efficient usage of the diluted allergens. False-positive reactions are fairly uncommon with the exception of dust and storage mites which should be tested at lower concentrations than pollens. Ideally antihistamines should be discontinued 2 weeks prior to skin testing, any injectable glucocorticoids 6 weeks prior, glucocorticoid tablets 4 weeks prior and any topical steroid-containing creams, drops or ointments (such as ear or eye drops) 2 weeks prior to skin testing. One possibility in a dog with atopic signs and a negative skin test is that the offending allergen was not included in the test kit.

Allergen-Specific Serum IgE

These tests are convenient, and particularly the Fcε receptor-based test is very reproducible and reliable. There is increasing evidence that glucocorticoid administration influences serum testing as well, thus withdrawing those is recommended where possible. Sometimes, grouped allergen testing is utilised for financial reasons and due to the high crossreactivity in human medicine. If group testing is positive, then one, some or all of the allergens in that group may have contributed to the hypersensitivity and thus the clinical signs. If all the allergens in that group are included in the allergy shots, immunotherapy with non-relevant allergens may lead to development of new allergies in addition to treating the present ones. Thus, immunotherapy based on testing for grouped allergens is strongly discouraged! There is some evidence in human medicine that antibody production may be local (in skin disease produced by plasma cells in the dermis), and blood and tissue levels of antibodies may not correlate for that reason. Some dogs do not produce IgE despite the classical clinical signs. Serum IgE concentrations thus may not be elevated in every atopic dog.