Christine S. Ahn, MD

Airbags are lifesaving during motor vehicle accidents (MVAs), but their deployment has been associated with skin issues such as irritant dermatitis¹; lacerations²; abrasions³; and thermal, friction, and chemical burns.^4-6 Ocular issues such as alkaline chemical keratitis⁷ and ocular alkali injuries⁸ also have been described.

Airbag deployment is triggered by rapid deceleration and impact, which ignite a sodium azide cartridge, causing the woven nylon bag to inflate with hydrocarbon gases.⁸ This leads to release of sodium hydroxide, sodium bicarbonate, and metallic oxides in an aerosolized form. If a tear in the meshwork of the airbag occurs, exposure to an even larger amount of powder containing caustic alkali chemicals can occur.⁸

We describe a patient who developed a bullous reaction to airbag contents after he was involved in an MVA in which the airbag deployed.

Case Report

A 35-year-old man with a history of type 2 diabetes mellitus and chronic hepatitis B presented to the dermatology clinic for an evaluation of new-onset blisters. The rash occurred 1 day after the patient was involved in an MVA in which he was exposed to the airbag’s contents after it burst. He had been evaluated twice in the emergency department for the skin eruption before being referred to dermatology. He noted the lesions were pruritic and painful. Prior treatments included silver sulfadiazine cream 1% and clobetasol cream 0.05%, though he discontinued using the latter because of burning with application. Physical examination revealed tense vesicles and bullae on an erythematous base on the right lower flank, forearms, and legs, with the exception of the lower right leg where a cast had been from a prior injury (Figure 1).

Two punch biopsies of the left arm were performed and sent for hematoxylin and eosin staining and direct immunofluorescence to rule out bullous diseases, such as bullous pemphigoid, linear IgA, and bullous lupus. Hematoxylin and eosin staining revealed extensive spongiosis with blister formation and a dense perivascular infiltrate in the superficial and mid dermis composed of lymphocytes with numerous scattered eosinophils (Figures 2 and 3). Direct immunofluorescence studies were negative. Treatment with oral prednisone and oral antihistamines was initiated.

At 10-day follow-up, the patient had a few residual bullae; most lesions were demonstrating various stages of healing (Figure 4). The patient’s cast had been removed, and there were no lesions in this previously covered area. At 6-week follow-up he had continued healing of the bullae and erosions as well as postinflammatory hyperpigmentation (Figure 5).

Comment

With the advent of airbags for safety purposes, these potentially lifesaving devices also have been known to cause injury.⁹ In 1998, the most commonly reported airbag injuries were ocular injuries.¹⁰ Cutaneous manifestations of airbag injury are less well known.¹¹

Two cases of airbag deployment with skin blistering have been reported in the literature based on a PubMed search of articles indexed for MEDLINE using the terms airbag blistering or airbag bullae^12,13; however, the blistering was described in the context of a burn. One case of the effects of airbag deployment residue highlights a patient arriving to the emergency department with erythema and blisters on the hands within 48 hours of airbag deployment in an MVA, and the treatment was standard burn therapy.¹² Another case report described a patient with a second-degree burn with a 12-cm blister occurring on the radial side of the hand and distal forearm following an MVA and airbag deployment, which was treated conservatively.¹³ Cases of thermal burns, chemical burns, and irritant contact dermatitis after airbag deployment have been described in the literature.^{4-6,11,12,14,15} Our patient’s distal right lower leg was covered with a cast for osteomyelitis, and no blisters had developed in this area. It is likely that the transfer of airbag contents occurred during the process of unbuckling his seatbelt, which could explain the bullae that developed on the right flank. Per the Centers for Disease Control and Prevention, individuals should quickly remove clothing and wash their body with large amounts of soap and water following exposure to sodium azide.¹⁶

In 1989, the Federal Motor Vehicle Safety Standard No. 208 (occupant crash protection) became effective, stating all cars must have vehicle crash protection.¹² Prior to 1993, it was reported that there had been no associated chemical injuries with airbag deployment. Subsequently, a 6-month retrospective study in 1993 showed that dermal injuries were found in connection with the presence of sodium hydroxide in automobile airbags.¹² By 2004, it was known that airbags could cause chemical and thermal burns in addition to traumatic injuries from deployment.¹ Since 2007, all motor vehicles have been required to have advanced airbags, which are engineered to sense the presence of passengers and determine if the airbag will deploy, and if so, how much to deploy to minimize airbag-related injury.³

The brand of car that our patient drove during the MVA is one with known airbag recalls due to safety defects; however, the year and actual model of the vehicle are not known, so specific information about the airbag in question is not available. It has been noted that some defective airbag inflators that were exposed to excess moisture during the manufacturing process could explode during deployment, causing shrapnel and airbag rupture, which has been linked to nearly 300 injuries worldwide.¹⁷

Conclusion

It is evident that the use of airbag devices reduces morbidity and mortality due to MVAs.⁹ It also had been reported that up to 96% of airbag-related injuries are relatively minor, which many would argue justifies their use.¹⁸ Furthermore, it has been reported that 99.8% of skin injuries following airbag deployment are minor.¹⁹ In the United States, it is mandated that every vehicle have functional airbags installed.⁸

This case highlights the potential for substantial airbag-induced skin reactions, specifically a bullous reaction, following airbag deployment. The persistent pruritus and lasting postinflammatory hyperpigmentation seen in this case were certainly worrisome for our patient. We also present this case to remind dermatology providers of possible treatment approaches to these skin reactions. Immediate cleansing of the affected areas of skin may help avoid such reactions.

Airbags are lifesaving during motor vehicle accidents (MVAs), but their deployment has been associated with skin issues such as irritant dermatitis¹; lacerations²; abrasions³; and thermal, friction, and chemical burns.^4-6 Ocular issues such as alkaline chemical keratitis⁷ and ocular alkali injuries⁸ also have been described.

Airbag deployment is triggered by rapid deceleration and impact, which ignite a sodium azide cartridge, causing the woven nylon bag to inflate with hydrocarbon gases.⁸ This leads to release of sodium hydroxide, sodium bicarbonate, and metallic oxides in an aerosolized form. If a tear in the meshwork of the airbag occurs, exposure to an even larger amount of powder containing caustic alkali chemicals can occur.⁸

We describe a patient who developed a bullous reaction to airbag contents after he was involved in an MVA in which the airbag deployed.

Case Report

A 35-year-old man with a history of type 2 diabetes mellitus and chronic hepatitis B presented to the dermatology clinic for an evaluation of new-onset blisters. The rash occurred 1 day after the patient was involved in an MVA in which he was exposed to the airbag’s contents after it burst. He had been evaluated twice in the emergency department for the skin eruption before being referred to dermatology. He noted the lesions were pruritic and painful. Prior treatments included silver sulfadiazine cream 1% and clobetasol cream 0.05%, though he discontinued using the latter because of burning with application. Physical examination revealed tense vesicles and bullae on an erythematous base on the right lower flank, forearms, and legs, with the exception of the lower right leg where a cast had been from a prior injury (Figure 1).

Two punch biopsies of the left arm were performed and sent for hematoxylin and eosin staining and direct immunofluorescence to rule out bullous diseases, such as bullous pemphigoid, linear IgA, and bullous lupus. Hematoxylin and eosin staining revealed extensive spongiosis with blister formation and a dense perivascular infiltrate in the superficial and mid dermis composed of lymphocytes with numerous scattered eosinophils (Figures 2 and 3). Direct immunofluorescence studies were negative. Treatment with oral prednisone and oral antihistamines was initiated.

At 10-day follow-up, the patient had a few residual bullae; most lesions were demonstrating various stages of healing (Figure 4). The patient’s cast had been removed, and there were no lesions in this previously covered area. At 6-week follow-up he had continued healing of the bullae and erosions as well as postinflammatory hyperpigmentation (Figure 5).

Comment

With the advent of airbags for safety purposes, these potentially lifesaving devices also have been known to cause injury.⁹ In 1998, the most commonly reported airbag injuries were ocular injuries.¹⁰ Cutaneous manifestations of airbag injury are less well known.¹¹

Two cases of airbag deployment with skin blistering have been reported in the literature based on a PubMed search of articles indexed for MEDLINE using the terms airbag blistering or airbag bullae^12,13; however, the blistering was described in the context of a burn. One case of the effects of airbag deployment residue highlights a patient arriving to the emergency department with erythema and blisters on the hands within 48 hours of airbag deployment in an MVA, and the treatment was standard burn therapy.¹² Another case report described a patient with a second-degree burn with a 12-cm blister occurring on the radial side of the hand and distal forearm following an MVA and airbag deployment, which was treated conservatively.¹³ Cases of thermal burns, chemical burns, and irritant contact dermatitis after airbag deployment have been described in the literature.^{4-6,11,12,14,15} Our patient’s distal right lower leg was covered with a cast for osteomyelitis, and no blisters had developed in this area. It is likely that the transfer of airbag contents occurred during the process of unbuckling his seatbelt, which could explain the bullae that developed on the right flank. Per the Centers for Disease Control and Prevention, individuals should quickly remove clothing and wash their body with large amounts of soap and water following exposure to sodium azide.¹⁶

In 1989, the Federal Motor Vehicle Safety Standard No. 208 (occupant crash protection) became effective, stating all cars must have vehicle crash protection.¹² Prior to 1993, it was reported that there had been no associated chemical injuries with airbag deployment. Subsequently, a 6-month retrospective study in 1993 showed that dermal injuries were found in connection with the presence of sodium hydroxide in automobile airbags.¹² By 2004, it was known that airbags could cause chemical and thermal burns in addition to traumatic injuries from deployment.¹ Since 2007, all motor vehicles have been required to have advanced airbags, which are engineered to sense the presence of passengers and determine if the airbag will deploy, and if so, how much to deploy to minimize airbag-related injury.³

The brand of car that our patient drove during the MVA is one with known airbag recalls due to safety defects; however, the year and actual model of the vehicle are not known, so specific information about the airbag in question is not available. It has been noted that some defective airbag inflators that were exposed to excess moisture during the manufacturing process could explode during deployment, causing shrapnel and airbag rupture, which has been linked to nearly 300 injuries worldwide.¹⁷

Conclusion

It is evident that the use of airbag devices reduces morbidity and mortality due to MVAs.⁹ It also had been reported that up to 96% of airbag-related injuries are relatively minor, which many would argue justifies their use.¹⁸ Furthermore, it has been reported that 99.8% of skin injuries following airbag deployment are minor.¹⁹ In the United States, it is mandated that every vehicle have functional airbags installed.⁸

This case highlights the potential for substantial airbag-induced skin reactions, specifically a bullous reaction, following airbag deployment. The persistent pruritus and lasting postinflammatory hyperpigmentation seen in this case were certainly worrisome for our patient. We also present this case to remind dermatology providers of possible treatment approaches to these skin reactions. Immediate cleansing of the affected areas of skin may help avoid such reactions.

Airbags are lifesaving during motor vehicle accidents (MVAs), but their deployment has been associated with skin issues such as irritant dermatitis¹; lacerations²; abrasions³; and thermal, friction, and chemical burns.^4-6 Ocular issues such as alkaline chemical keratitis⁷ and ocular alkali injuries⁸ also have been described.

Airbag deployment is triggered by rapid deceleration and impact, which ignite a sodium azide cartridge, causing the woven nylon bag to inflate with hydrocarbon gases.⁸ This leads to release of sodium hydroxide, sodium bicarbonate, and metallic oxides in an aerosolized form. If a tear in the meshwork of the airbag occurs, exposure to an even larger amount of powder containing caustic alkali chemicals can occur.⁸

We describe a patient who developed a bullous reaction to airbag contents after he was involved in an MVA in which the airbag deployed.

Case Report

A 35-year-old man with a history of type 2 diabetes mellitus and chronic hepatitis B presented to the dermatology clinic for an evaluation of new-onset blisters. The rash occurred 1 day after the patient was involved in an MVA in which he was exposed to the airbag’s contents after it burst. He had been evaluated twice in the emergency department for the skin eruption before being referred to dermatology. He noted the lesions were pruritic and painful. Prior treatments included silver sulfadiazine cream 1% and clobetasol cream 0.05%, though he discontinued using the latter because of burning with application. Physical examination revealed tense vesicles and bullae on an erythematous base on the right lower flank, forearms, and legs, with the exception of the lower right leg where a cast had been from a prior injury (Figure 1).

Two punch biopsies of the left arm were performed and sent for hematoxylin and eosin staining and direct immunofluorescence to rule out bullous diseases, such as bullous pemphigoid, linear IgA, and bullous lupus. Hematoxylin and eosin staining revealed extensive spongiosis with blister formation and a dense perivascular infiltrate in the superficial and mid dermis composed of lymphocytes with numerous scattered eosinophils (Figures 2 and 3). Direct immunofluorescence studies were negative. Treatment with oral prednisone and oral antihistamines was initiated.

At 10-day follow-up, the patient had a few residual bullae; most lesions were demonstrating various stages of healing (Figure 4). The patient’s cast had been removed, and there were no lesions in this previously covered area. At 6-week follow-up he had continued healing of the bullae and erosions as well as postinflammatory hyperpigmentation (Figure 5).

Comment

With the advent of airbags for safety purposes, these potentially lifesaving devices also have been known to cause injury.⁹ In 1998, the most commonly reported airbag injuries were ocular injuries.¹⁰ Cutaneous manifestations of airbag injury are less well known.¹¹

Two cases of airbag deployment with skin blistering have been reported in the literature based on a PubMed search of articles indexed for MEDLINE using the terms airbag blistering or airbag bullae^12,13; however, the blistering was described in the context of a burn. One case of the effects of airbag deployment residue highlights a patient arriving to the emergency department with erythema and blisters on the hands within 48 hours of airbag deployment in an MVA, and the treatment was standard burn therapy.¹² Another case report described a patient with a second-degree burn with a 12-cm blister occurring on the radial side of the hand and distal forearm following an MVA and airbag deployment, which was treated conservatively.¹³ Cases of thermal burns, chemical burns, and irritant contact dermatitis after airbag deployment have been described in the literature.^{4-6,11,12,14,15} Our patient’s distal right lower leg was covered with a cast for osteomyelitis, and no blisters had developed in this area. It is likely that the transfer of airbag contents occurred during the process of unbuckling his seatbelt, which could explain the bullae that developed on the right flank. Per the Centers for Disease Control and Prevention, individuals should quickly remove clothing and wash their body with large amounts of soap and water following exposure to sodium azide.¹⁶

In 1989, the Federal Motor Vehicle Safety Standard No. 208 (occupant crash protection) became effective, stating all cars must have vehicle crash protection.¹² Prior to 1993, it was reported that there had been no associated chemical injuries with airbag deployment. Subsequently, a 6-month retrospective study in 1993 showed that dermal injuries were found in connection with the presence of sodium hydroxide in automobile airbags.¹² By 2004, it was known that airbags could cause chemical and thermal burns in addition to traumatic injuries from deployment.¹ Since 2007, all motor vehicles have been required to have advanced airbags, which are engineered to sense the presence of passengers and determine if the airbag will deploy, and if so, how much to deploy to minimize airbag-related injury.³

The brand of car that our patient drove during the MVA is one with known airbag recalls due to safety defects; however, the year and actual model of the vehicle are not known, so specific information about the airbag in question is not available. It has been noted that some defective airbag inflators that were exposed to excess moisture during the manufacturing process could explode during deployment, causing shrapnel and airbag rupture, which has been linked to nearly 300 injuries worldwide.¹⁷

Conclusion

It is evident that the use of airbag devices reduces morbidity and mortality due to MVAs.⁹ It also had been reported that up to 96% of airbag-related injuries are relatively minor, which many would argue justifies their use.¹⁸ Furthermore, it has been reported that 99.8% of skin injuries following airbag deployment are minor.¹⁹ In the United States, it is mandated that every vehicle have functional airbags installed.⁸

This case highlights the potential for substantial airbag-induced skin reactions, specifically a bullous reaction, following airbag deployment. The persistent pruritus and lasting postinflammatory hyperpigmentation seen in this case were certainly worrisome for our patient. We also present this case to remind dermatology providers of possible treatment approaches to these skin reactions. Immediate cleansing of the affected areas of skin may help avoid such reactions.

Molluscum contagiosum (MC) is a double-stranded DNA virus of the Poxviridae family, which commonly infects human keratinocytes resulting in small, umbilicated, flesh-colored papules. The greatest incidence of MC is seen in the pediatric population and sexually active young adults, and it is considered a self-limited disease in immunocompetent individuals.¹ With the emergence of the human immunodeficiency virus (HIV) and subsequent AIDS epidemic in the 1980s, a new population of immunocompromised individuals has been observed to be increasingly susceptible to MC with an atypical clinical presentation and a recalcitrant disease course.² Although the increased prevalence of MC in the HIV population has been well-documented, it has been observed in other disease states or iatrogenically induced immunosuppression due to a deficiency in function or absolute number of T lymphocytes.

We present a case of a patient with long-standing psoriasis on biologic therapy who presented with MC with a subsequent workup that revealed AIDS. This case reiterates the importance of MC as a potential indicator of underlying immunosuppression. We review the literature to evaluate the occurrence of MC in immunosuppressed patients.

Case Report

A 33-year-old man initially presented for evaluation of severe plaque-type psoriasis associated with pain, erythema, and swelling of the joints of the hands of 10 years’ duration. He was started on methotrexate 5 mg weekly and topical corticosteroids but was unable to tolerate methotrexate due to headaches. He also had difficulty affording topical medications and adjunctive phototherapy. The patient was sporadically seen in follow-up with persistence of psoriatic plaques involving up to 60% body surface area (BSA) with the only treatment consisting of occasional topical steroids. Five years later, the patient was restarted on methotrexate 5 to 7.5 mg weekly, which resulted in moderate improvement. However, because of persistent elevation of liver enzymes, this treatment was stopped. Several months later he was evaluated for treatment with a biologic agent, and after a negative tuberculin skin test, he began treatment with etanercept 50 mg subcutaneous injection twice weekly, which provided notable improvement and allowed for reduction of dose frequency to once weekly.

At follow-up 1 year later, the patient had continued improvement of psoriasis with approximately 30% BSA on a treatment regimen of etanercept 50 mg weekly injection and topical corticosteroids. However, on physical examination, there were multiple small semitranslucent papules with telangiectases on the chest and upper back (Figure 1). Biopsy of a representative papule on the chest revealed MC (Figure 2). The patient was subsequently advised to stop etanercept and to return immediately to the clinic for HIV testing. He returned for follow-up 3 months later with pronounced worsening of disease and a new onset of blurred vision of the right eye. Cutaneous examination revealed numerous large erythematous plaques with superficial scale and cerebriform surface on the chest, back, abdomen, and upper and lower extremities involving 80% BSA (Figure 3). Biopsy of a plaque demonstrated psoriasiform dermatitis with neutrophils and parakeratosis consistent with psoriasis. Extensive blood work was notable for reactive HIV antibody and lymphopenia, CD4 lymphocyte count of 60 cells/mm³, and an HIV viral load of 247,000 copies/mL, meeting diagnostic criteria for AIDS. Additionally, ophthalmologic evaluation revealed toxoplasma retinitis. Upon initiation of highly active antiretroviral therapy (HAART) and continued use of topical corticosteroids, the patient experienced notable improvement of disease severity with approximately 20% BSA.

Comment

Molluscum contagiosum is a common skin infection. Among patients with HIV and other types of impaired cellular immunity, the prevalence of MC is estimated to be as high as 20%.³ The MC poxvirus survives and proliferates within the epidermis by interfering with tumor necrosis factor–induced apoptosis of virally infected cells; therefore, intact cell-mediated immunity is an important component of prevention and clearance of poxvirus infections. In immunocompromised patients, the presentation of MC varies widely, and the disease is often difficult to eradicate. This review will highlight the prevalence, presentation, and treatment of MC in the context of immunosuppressed states.

HIV/AIDS
Molluscum contagiosum in HIV-positive patients was first recognized in 1983,² and its prevalence is estimated to range from 5% to 18% in AIDS patients.³ Molluscum contagiosum is a clinical sign of HIV progression, and its incidence appears to increase with reduced immune function (ie, a CD4 cell count <200/mm³).³ In a study of 456 patients with HIV-associated skin disorders, the majority of patients with MC had notable immunosuppression with a median survival time of 12 months. Thus, MC was not an independent prognostic marker but a clinical indicator of markedly reduced immune status.⁴

Molluscum contagiosum is transmitted in both sexual and nonsexual patterns in HIV-positive individuals, with the distribution of the latter involving primarily the face and neck. Although it may present with typical umbilicated papules, MC has a wide range of atypical clinical presentations in patients with AIDS that can make it difficult to diagnose. Complicated cases of eyelid MC have been reported in advanced HIV in both adults and children, resulting in obstruction of vision due to large lesions (up to 2 cm) or hundreds of confluent lesions.⁵ Giant MC, which appears as large exophytic nodules, is another presentation that has been frequently described in patients with advanced HIV. In these patients, the lesions often are too voluminous for conservative therapy and require excision.⁶ Atypical MC lesions also can resemble other dermatologic conditions, including condyloma acuminatum,⁷ nevus sebaceous of Jadassohn, ecthyma,⁸ and cutaneous horns,^9,10 as well as other bacterial and fungal infections in HIV-positive patients, such as cutaneous Cryptococcus neoformans,^11,12 disseminated histoplasmosis,¹³ and infections caused by Penicillium marneffei¹⁴ and Bartonella henselae.¹⁵ In most cases of MC in HIV-positive patients, diagnosis is dependent on the examination of biopsy specimens, which maintain the same histopathologic features regardless of immune status.

The management of MC in patients with HIV/AIDS is difficult. Molluscum contagiosum has shown no evidence of spontaneous resolution in patients with HIV, and treatment with one modality is often insufficient. Treatment is most successful when a combination approach is utilized with destructive procedures (eg, curettage, cryosurgery) and adjunctive agents (eg, retinoids, cantharidin, trichloroacetic acid). Imiquimod and cidofovir have been used off label for MC in AIDS patients.¹⁶ Imiquimod, which is used to treat genital warts, another cutaneous viral infection seen in patients with HIV, has demonstrated efficacy in treating MC.¹⁶ In a randomized controlled trial comparing imiquimod cream 5% to cryotherapy for MC in healthy children, imiquimod was slow acting but better suited than cryotherapy for patients with eruptions of many small lesions.¹⁷ For HIV patients, numerous reports have described successful treatment of disseminated or recalcitrant MC with topical imiquimod.^18-20 Cidofovir, an antiviral used to treat cytomegalovirus retinitis in patients with AIDS, is a promising antiviral agent against the poxvirus family. In a study of viral DNA polymerase genes of MC virus, cidofovir inhibited MC virus DNA polymerase activity.²¹ It has been used in both topical (1% to 3%) and intravenous form to successfully treat recalcitrant and exuberant giant MC.^6,22 However, the use of cidofovir is limited by its high costs, especially when compounded into a topical formulation.²³

From a systemic standpoint, numerous reports have shown that treating the underlying HIV by optimizing HAART is the most important first step in clearing MC.^24-27 However, a special concern regarding the initiation of HAART in patients with MC as well as a markedly impaired immune function is the development of an inflammatory reaction called immune reconstitution inflammatory syndrome (IRIS). This reaction is thought to be a result of immune recovery in severely immunosuppressed patients. During the initial phase of reconstitution when CD4 lymphocyte counts rise and viral load decreases, IRIS occurs due to an inflammatory reaction to microbial and autoimmune antigens, leading to temporary clinical deterioration.²⁸ The incidence has been reported in up to 25% of patients starting HAART, and 52% to 78% of IRIS cases involve dermatologic manifestations such as varicella-zoster virus, cytomegalovirus infections, genital warts, and MC.^29,30 In a cohort study of 199 patients, 2% of patients developed MC within 6 months of initiating HAART.³¹ In a case of exuberant MC lesions after beginning HAART, the lesions spontaneously resolved with the progression of immune reconstitution.²⁸

Malignancies
Patients with hematologic malignancies such as lymphoma and leukemia comprise another subset of patients at risk for atypical presentations of MC. Molluscum contagiosum has been described in patients with hematologic malignancies such as adult T-cell leukemia/lymphoma, multiple myeloma, chronic myeloid leukemia, acute lymphoblastic leukemia, lymphomatoid papulosis, and non-Hodgkin lymphoma. In a review of MC in children with cancer, 0.5% were diagnosed with MC.^32,33 Reports also have documented eruptive MC in the presence of solid organ cancers, including lung cancer.³⁴

In patients with malignancies, the differential diagnosis should include other common dermatologic conditions such as varicella, herpes simplex, papillomas, pyoderma, and cutaneous cryptococcosis, as well as MC. Similar to HIV-positive patients, the lesions of MC described in patients with malignancies do not tend to spontaneously resolve. In a report of a pediatric patient with acute lymphoblastic leukemia, MC presented as an ulcerated lesion without any classic features, requiring biopsy for definitive diagnosis. Only partial resolution was achieved with cryotherapy and crusting of the lesion in an attempt to slow the progression.³⁵ In a series of 5 children with hematologic malignancies and MC, little improvement was noted after treatment with surgical scraping, liquid nitrogen, and salicylic acid ointment 5%. Similar to patients with HIV, improvement of immune status and function help clear the disease, and patients who reach remission and discontinue chemotherapeutic agents have a higher rate of spontaneous resolution of previously recalcitrant MC lesions.³⁶

Transplant Patients
Molluscum contagiosum in transplant patients has features similar to patients with HIV/AIDS. In organ transplant recipients, there is an increased risk for cutaneous disease from iatrogenic immunosuppression or immunosuppression through infectious or neoplastic processes.³⁷ As in other immunocompromised populations, MC often has an atypical presentation in transplant patients with more extensive involvement and recalcitrant, rapidly recurring lesions.

In a review of 145 pediatric organ transplant recipients, MC was the fourth most common skin infection after verruca vulgaris, tinea versicolor, and herpes simplex/zoster. Affecting 7% of patients, the majority of patients demonstrated clinically typical lesions; however, the disease was difficult to eradicate if multiple lesions were present.³⁷ In other reports in adults, fulminant and giant MC have been described after renal and other solid organ transplants.^38,39 Molluscum contagiosum also has been reported to mimic other skin diseases in transplant patients including tinea barbae⁴⁰ and nodular basal cell carcinomas.⁴¹

The standard treatments are identical to those used in patients with HIV, including ablative methods via liquid nitrogen, electrocautery, cantharidin, trichloroacetic acid, and topical retinoids. Similar to MC in other immunocompromised states, treatment can be difficult and usually requires multiple modalities. For children, imiquimod cream 5% has been recommended due to high clearance rates (up to 92%) and the painless nature of the treatment.^42,43

Other Iatrogenic Immunosuppressive States
Immunosuppression through the use of steroids, chemotherapeutic agents, and biologic drugs often is the result of treatment of various diseases. In patients with psoriasis treated with systemic immunosuppressive agents, there are numerous reports that describe the appearance of eruptive MC in association with methotrexate, cyclosporine, and biologics. Methotrexate acts as an immunosuppressive agent by binding to dihydrofolate reductase, which inhibits DNA synthesis in immunologically competent cells.⁴⁴ It also may block host defense mechanisms against MC by suppressing the expression of serum inflammatory cytokines such as tumor necrosis factor α (TNF-α) and IFN-γ and suppressing the activity of TNF-α inducing apoptosis of virus-infected cells. Cyclosporine used in conjunction with methotrexate may exacerbate the insult to the immune system by inhibiting the production of IFN-γ.⁴⁵ Biologics are an emerging class of drugs that have demonstrated efficacy in moderate to severe psoriasis by inhibiting TNF-α or other inflammatory molecules. Several published reports have described eruptive or atypical MC in patients on biologic medications. In one case, within 2 weeks after initiation of infliximab, a monoclonal antibody against TNF-α, a patient developed an eruption of MC involving the entire body.⁴⁶ In another report, an anti–TNF-α agent for rheumatoid arthritis was associated with atypical MC with eyelid lesions.⁴⁷

There are other skin disorders treated with immunosuppressive agents that also have been associated with MC. In a patient with pemphigus vulgaris treated with prednisolone, pimecrolimus, and azathioprine, MC lesions were observed on the face and within healed pemphigus vulgaris sites.⁴⁸ Pimecrolimus and tacrolimus, corticosteroid-sparing agents, suppress cell-mediated immunity and inhibit inflammatory cytokines such as IL-2. The infection resolved with a gradual tapering of immunosuppressive therapy and 10 sessions of cryotherapy.⁴⁸ In a case of topical pimecrolimus for pityriasis alba, the patient developed biopsy-proven MC within 2 weeks of initiating treatment in the areas that were treated with tacrolimus.⁴⁹

In nontransplant patients with iatrogenic immunosuppression, MC treatment has not been documented to be as challenging as in patients with inherent immunosuppression. Most patients respond to either withdrawal of the drug alone or to simple ablative treatments such as cryotherapy.^45,46,48 This important difference is most likely due to the presence of an otherwise intact immune system.

Conclusion

This case describes the appearance of MC in a patient with psoriasis treated with a TNF-α inhibitor who was ultimately diagnosed with AIDS. Although atypical MC infections have been documented in patients with psoriasis undergoing treatment with biologics, it is thought to be more common for MC to occur in more remarkably immunocompromised states such as AIDS. Thus, the persistence and progression of MC in our patient despite discontinuation of etanercept suggested a separate underlying process. Subsequent workup led to the diagnosis of AIDS along with the opportunistic ocular infection of toxoplasmosis retinitis. This clinical sequence consisting of psoriasis treated with a biologic agent, development of MC, and subsequent diagnosis of AIDS is unique and clinically significant to dermatologists. The presentation of psoriasis in patients with HIV can be diverse with different levels of severity and atypical clinical features. In many cases, HIV is known to exacerbate the classic clinical presentation of psoriasis. However, there are other particular presentations of psoriasis in HIV patients that have been observed, which include a predilection for scalp lesions, palmoplantar keratoderma, flexural involvement, and higher levels of immunodeficiency.⁵⁰ Although tuberculin skin tests are required prior to initiating biologic therapy due to the potential for disease reactivation, there are no requirements for HIV antibody testing. In cases of severe recalcitrant psoriasis, an HIV test should be ordered during the workup to establish an early diagnosis so that an HIV-positive patient can avoid poor outcomes from either the disease processes, the use of certain therapeutic agents, or both. Furthermore, the benefit of avoiding possible harm to the patient and potential legal action outweighs the cost of performing surveillance HIV testing in this subset of patients. Thus, due to the potential additive immunosuppressive effect of HIV with biologic therapy, providers should always assess for risk factors and consider testing for HIV in all patients before initiating treatment with immunosuppressive agents such as biologics.

Molluscum contagiosum (MC) is a double-stranded DNA virus of the Poxviridae family, which commonly infects human keratinocytes resulting in small, umbilicated, flesh-colored papules. The greatest incidence of MC is seen in the pediatric population and sexually active young adults, and it is considered a self-limited disease in immunocompetent individuals.¹ With the emergence of the human immunodeficiency virus (HIV) and subsequent AIDS epidemic in the 1980s, a new population of immunocompromised individuals has been observed to be increasingly susceptible to MC with an atypical clinical presentation and a recalcitrant disease course.² Although the increased prevalence of MC in the HIV population has been well-documented, it has been observed in other disease states or iatrogenically induced immunosuppression due to a deficiency in function or absolute number of T lymphocytes.

We present a case of a patient with long-standing psoriasis on biologic therapy who presented with MC with a subsequent workup that revealed AIDS. This case reiterates the importance of MC as a potential indicator of underlying immunosuppression. We review the literature to evaluate the occurrence of MC in immunosuppressed patients.

Case Report

A 33-year-old man initially presented for evaluation of severe plaque-type psoriasis associated with pain, erythema, and swelling of the joints of the hands of 10 years’ duration. He was started on methotrexate 5 mg weekly and topical corticosteroids but was unable to tolerate methotrexate due to headaches. He also had difficulty affording topical medications and adjunctive phototherapy. The patient was sporadically seen in follow-up with persistence of psoriatic plaques involving up to 60% body surface area (BSA) with the only treatment consisting of occasional topical steroids. Five years later, the patient was restarted on methotrexate 5 to 7.5 mg weekly, which resulted in moderate improvement. However, because of persistent elevation of liver enzymes, this treatment was stopped. Several months later he was evaluated for treatment with a biologic agent, and after a negative tuberculin skin test, he began treatment with etanercept 50 mg subcutaneous injection twice weekly, which provided notable improvement and allowed for reduction of dose frequency to once weekly.

At follow-up 1 year later, the patient had continued improvement of psoriasis with approximately 30% BSA on a treatment regimen of etanercept 50 mg weekly injection and topical corticosteroids. However, on physical examination, there were multiple small semitranslucent papules with telangiectases on the chest and upper back (Figure 1). Biopsy of a representative papule on the chest revealed MC (Figure 2). The patient was subsequently advised to stop etanercept and to return immediately to the clinic for HIV testing. He returned for follow-up 3 months later with pronounced worsening of disease and a new onset of blurred vision of the right eye. Cutaneous examination revealed numerous large erythematous plaques with superficial scale and cerebriform surface on the chest, back, abdomen, and upper and lower extremities involving 80% BSA (Figure 3). Biopsy of a plaque demonstrated psoriasiform dermatitis with neutrophils and parakeratosis consistent with psoriasis. Extensive blood work was notable for reactive HIV antibody and lymphopenia, CD4 lymphocyte count of 60 cells/mm³, and an HIV viral load of 247,000 copies/mL, meeting diagnostic criteria for AIDS. Additionally, ophthalmologic evaluation revealed toxoplasma retinitis. Upon initiation of highly active antiretroviral therapy (HAART) and continued use of topical corticosteroids, the patient experienced notable improvement of disease severity with approximately 20% BSA.

Comment

Molluscum contagiosum is a common skin infection. Among patients with HIV and other types of impaired cellular immunity, the prevalence of MC is estimated to be as high as 20%.³ The MC poxvirus survives and proliferates within the epidermis by interfering with tumor necrosis factor–induced apoptosis of virally infected cells; therefore, intact cell-mediated immunity is an important component of prevention and clearance of poxvirus infections. In immunocompromised patients, the presentation of MC varies widely, and the disease is often difficult to eradicate. This review will highlight the prevalence, presentation, and treatment of MC in the context of immunosuppressed states.

HIV/AIDS
Molluscum contagiosum in HIV-positive patients was first recognized in 1983,² and its prevalence is estimated to range from 5% to 18% in AIDS patients.³ Molluscum contagiosum is a clinical sign of HIV progression, and its incidence appears to increase with reduced immune function (ie, a CD4 cell count <200/mm³).³ In a study of 456 patients with HIV-associated skin disorders, the majority of patients with MC had notable immunosuppression with a median survival time of 12 months. Thus, MC was not an independent prognostic marker but a clinical indicator of markedly reduced immune status.⁴

Molluscum contagiosum is transmitted in both sexual and nonsexual patterns in HIV-positive individuals, with the distribution of the latter involving primarily the face and neck. Although it may present with typical umbilicated papules, MC has a wide range of atypical clinical presentations in patients with AIDS that can make it difficult to diagnose. Complicated cases of eyelid MC have been reported in advanced HIV in both adults and children, resulting in obstruction of vision due to large lesions (up to 2 cm) or hundreds of confluent lesions.⁵ Giant MC, which appears as large exophytic nodules, is another presentation that has been frequently described in patients with advanced HIV. In these patients, the lesions often are too voluminous for conservative therapy and require excision.⁶ Atypical MC lesions also can resemble other dermatologic conditions, including condyloma acuminatum,⁷ nevus sebaceous of Jadassohn, ecthyma,⁸ and cutaneous horns,^9,10 as well as other bacterial and fungal infections in HIV-positive patients, such as cutaneous Cryptococcus neoformans,^11,12 disseminated histoplasmosis,¹³ and infections caused by Penicillium marneffei¹⁴ and Bartonella henselae.¹⁵ In most cases of MC in HIV-positive patients, diagnosis is dependent on the examination of biopsy specimens, which maintain the same histopathologic features regardless of immune status.

The management of MC in patients with HIV/AIDS is difficult. Molluscum contagiosum has shown no evidence of spontaneous resolution in patients with HIV, and treatment with one modality is often insufficient. Treatment is most successful when a combination approach is utilized with destructive procedures (eg, curettage, cryosurgery) and adjunctive agents (eg, retinoids, cantharidin, trichloroacetic acid). Imiquimod and cidofovir have been used off label for MC in AIDS patients.¹⁶ Imiquimod, which is used to treat genital warts, another cutaneous viral infection seen in patients with HIV, has demonstrated efficacy in treating MC.¹⁶ In a randomized controlled trial comparing imiquimod cream 5% to cryotherapy for MC in healthy children, imiquimod was slow acting but better suited than cryotherapy for patients with eruptions of many small lesions.¹⁷ For HIV patients, numerous reports have described successful treatment of disseminated or recalcitrant MC with topical imiquimod.^18-20 Cidofovir, an antiviral used to treat cytomegalovirus retinitis in patients with AIDS, is a promising antiviral agent against the poxvirus family. In a study of viral DNA polymerase genes of MC virus, cidofovir inhibited MC virus DNA polymerase activity.²¹ It has been used in both topical (1% to 3%) and intravenous form to successfully treat recalcitrant and exuberant giant MC.^6,22 However, the use of cidofovir is limited by its high costs, especially when compounded into a topical formulation.²³

From a systemic standpoint, numerous reports have shown that treating the underlying HIV by optimizing HAART is the most important first step in clearing MC.^24-27 However, a special concern regarding the initiation of HAART in patients with MC as well as a markedly impaired immune function is the development of an inflammatory reaction called immune reconstitution inflammatory syndrome (IRIS). This reaction is thought to be a result of immune recovery in severely immunosuppressed patients. During the initial phase of reconstitution when CD4 lymphocyte counts rise and viral load decreases, IRIS occurs due to an inflammatory reaction to microbial and autoimmune antigens, leading to temporary clinical deterioration.²⁸ The incidence has been reported in up to 25% of patients starting HAART, and 52% to 78% of IRIS cases involve dermatologic manifestations such as varicella-zoster virus, cytomegalovirus infections, genital warts, and MC.^29,30 In a cohort study of 199 patients, 2% of patients developed MC within 6 months of initiating HAART.³¹ In a case of exuberant MC lesions after beginning HAART, the lesions spontaneously resolved with the progression of immune reconstitution.²⁸

Malignancies
Patients with hematologic malignancies such as lymphoma and leukemia comprise another subset of patients at risk for atypical presentations of MC. Molluscum contagiosum has been described in patients with hematologic malignancies such as adult T-cell leukemia/lymphoma, multiple myeloma, chronic myeloid leukemia, acute lymphoblastic leukemia, lymphomatoid papulosis, and non-Hodgkin lymphoma. In a review of MC in children with cancer, 0.5% were diagnosed with MC.^32,33 Reports also have documented eruptive MC in the presence of solid organ cancers, including lung cancer.³⁴

In patients with malignancies, the differential diagnosis should include other common dermatologic conditions such as varicella, herpes simplex, papillomas, pyoderma, and cutaneous cryptococcosis, as well as MC. Similar to HIV-positive patients, the lesions of MC described in patients with malignancies do not tend to spontaneously resolve. In a report of a pediatric patient with acute lymphoblastic leukemia, MC presented as an ulcerated lesion without any classic features, requiring biopsy for definitive diagnosis. Only partial resolution was achieved with cryotherapy and crusting of the lesion in an attempt to slow the progression.³⁵ In a series of 5 children with hematologic malignancies and MC, little improvement was noted after treatment with surgical scraping, liquid nitrogen, and salicylic acid ointment 5%. Similar to patients with HIV, improvement of immune status and function help clear the disease, and patients who reach remission and discontinue chemotherapeutic agents have a higher rate of spontaneous resolution of previously recalcitrant MC lesions.³⁶

Transplant Patients
Molluscum contagiosum in transplant patients has features similar to patients with HIV/AIDS. In organ transplant recipients, there is an increased risk for cutaneous disease from iatrogenic immunosuppression or immunosuppression through infectious or neoplastic processes.³⁷ As in other immunocompromised populations, MC often has an atypical presentation in transplant patients with more extensive involvement and recalcitrant, rapidly recurring lesions.

In a review of 145 pediatric organ transplant recipients, MC was the fourth most common skin infection after verruca vulgaris, tinea versicolor, and herpes simplex/zoster. Affecting 7% of patients, the majority of patients demonstrated clinically typical lesions; however, the disease was difficult to eradicate if multiple lesions were present.³⁷ In other reports in adults, fulminant and giant MC have been described after renal and other solid organ transplants.^38,39 Molluscum contagiosum also has been reported to mimic other skin diseases in transplant patients including tinea barbae⁴⁰ and nodular basal cell carcinomas.⁴¹

The standard treatments are identical to those used in patients with HIV, including ablative methods via liquid nitrogen, electrocautery, cantharidin, trichloroacetic acid, and topical retinoids. Similar to MC in other immunocompromised states, treatment can be difficult and usually requires multiple modalities. For children, imiquimod cream 5% has been recommended due to high clearance rates (up to 92%) and the painless nature of the treatment.^42,43

Other Iatrogenic Immunosuppressive States
Immunosuppression through the use of steroids, chemotherapeutic agents, and biologic drugs often is the result of treatment of various diseases. In patients with psoriasis treated with systemic immunosuppressive agents, there are numerous reports that describe the appearance of eruptive MC in association with methotrexate, cyclosporine, and biologics. Methotrexate acts as an immunosuppressive agent by binding to dihydrofolate reductase, which inhibits DNA synthesis in immunologically competent cells.⁴⁴ It also may block host defense mechanisms against MC by suppressing the expression of serum inflammatory cytokines such as tumor necrosis factor α (TNF-α) and IFN-γ and suppressing the activity of TNF-α inducing apoptosis of virus-infected cells. Cyclosporine used in conjunction with methotrexate may exacerbate the insult to the immune system by inhibiting the production of IFN-γ.⁴⁵ Biologics are an emerging class of drugs that have demonstrated efficacy in moderate to severe psoriasis by inhibiting TNF-α or other inflammatory molecules. Several published reports have described eruptive or atypical MC in patients on biologic medications. In one case, within 2 weeks after initiation of infliximab, a monoclonal antibody against TNF-α, a patient developed an eruption of MC involving the entire body.⁴⁶ In another report, an anti–TNF-α agent for rheumatoid arthritis was associated with atypical MC with eyelid lesions.⁴⁷

There are other skin disorders treated with immunosuppressive agents that also have been associated with MC. In a patient with pemphigus vulgaris treated with prednisolone, pimecrolimus, and azathioprine, MC lesions were observed on the face and within healed pemphigus vulgaris sites.⁴⁸ Pimecrolimus and tacrolimus, corticosteroid-sparing agents, suppress cell-mediated immunity and inhibit inflammatory cytokines such as IL-2. The infection resolved with a gradual tapering of immunosuppressive therapy and 10 sessions of cryotherapy.⁴⁸ In a case of topical pimecrolimus for pityriasis alba, the patient developed biopsy-proven MC within 2 weeks of initiating treatment in the areas that were treated with tacrolimus.⁴⁹

In nontransplant patients with iatrogenic immunosuppression, MC treatment has not been documented to be as challenging as in patients with inherent immunosuppression. Most patients respond to either withdrawal of the drug alone or to simple ablative treatments such as cryotherapy.^45,46,48 This important difference is most likely due to the presence of an otherwise intact immune system.

Conclusion

This case describes the appearance of MC in a patient with psoriasis treated with a TNF-α inhibitor who was ultimately diagnosed with AIDS. Although atypical MC infections have been documented in patients with psoriasis undergoing treatment with biologics, it is thought to be more common for MC to occur in more remarkably immunocompromised states such as AIDS. Thus, the persistence and progression of MC in our patient despite discontinuation of etanercept suggested a separate underlying process. Subsequent workup led to the diagnosis of AIDS along with the opportunistic ocular infection of toxoplasmosis retinitis. This clinical sequence consisting of psoriasis treated with a biologic agent, development of MC, and subsequent diagnosis of AIDS is unique and clinically significant to dermatologists. The presentation of psoriasis in patients with HIV can be diverse with different levels of severity and atypical clinical features. In many cases, HIV is known to exacerbate the classic clinical presentation of psoriasis. However, there are other particular presentations of psoriasis in HIV patients that have been observed, which include a predilection for scalp lesions, palmoplantar keratoderma, flexural involvement, and higher levels of immunodeficiency.⁵⁰ Although tuberculin skin tests are required prior to initiating biologic therapy due to the potential for disease reactivation, there are no requirements for HIV antibody testing. In cases of severe recalcitrant psoriasis, an HIV test should be ordered during the workup to establish an early diagnosis so that an HIV-positive patient can avoid poor outcomes from either the disease processes, the use of certain therapeutic agents, or both. Furthermore, the benefit of avoiding possible harm to the patient and potential legal action outweighs the cost of performing surveillance HIV testing in this subset of patients. Thus, due to the potential additive immunosuppressive effect of HIV with biologic therapy, providers should always assess for risk factors and consider testing for HIV in all patients before initiating treatment with immunosuppressive agents such as biologics.

Molluscum contagiosum (MC) is a double-stranded DNA virus of the Poxviridae family, which commonly infects human keratinocytes resulting in small, umbilicated, flesh-colored papules. The greatest incidence of MC is seen in the pediatric population and sexually active young adults, and it is considered a self-limited disease in immunocompetent individuals.¹ With the emergence of the human immunodeficiency virus (HIV) and subsequent AIDS epidemic in the 1980s, a new population of immunocompromised individuals has been observed to be increasingly susceptible to MC with an atypical clinical presentation and a recalcitrant disease course.² Although the increased prevalence of MC in the HIV population has been well-documented, it has been observed in other disease states or iatrogenically induced immunosuppression due to a deficiency in function or absolute number of T lymphocytes.

We present a case of a patient with long-standing psoriasis on biologic therapy who presented with MC with a subsequent workup that revealed AIDS. This case reiterates the importance of MC as a potential indicator of underlying immunosuppression. We review the literature to evaluate the occurrence of MC in immunosuppressed patients.

Case Report

A 33-year-old man initially presented for evaluation of severe plaque-type psoriasis associated with pain, erythema, and swelling of the joints of the hands of 10 years’ duration. He was started on methotrexate 5 mg weekly and topical corticosteroids but was unable to tolerate methotrexate due to headaches. He also had difficulty affording topical medications and adjunctive phototherapy. The patient was sporadically seen in follow-up with persistence of psoriatic plaques involving up to 60% body surface area (BSA) with the only treatment consisting of occasional topical steroids. Five years later, the patient was restarted on methotrexate 5 to 7.5 mg weekly, which resulted in moderate improvement. However, because of persistent elevation of liver enzymes, this treatment was stopped. Several months later he was evaluated for treatment with a biologic agent, and after a negative tuberculin skin test, he began treatment with etanercept 50 mg subcutaneous injection twice weekly, which provided notable improvement and allowed for reduction of dose frequency to once weekly.

At follow-up 1 year later, the patient had continued improvement of psoriasis with approximately 30% BSA on a treatment regimen of etanercept 50 mg weekly injection and topical corticosteroids. However, on physical examination, there were multiple small semitranslucent papules with telangiectases on the chest and upper back (Figure 1). Biopsy of a representative papule on the chest revealed MC (Figure 2). The patient was subsequently advised to stop etanercept and to return immediately to the clinic for HIV testing. He returned for follow-up 3 months later with pronounced worsening of disease and a new onset of blurred vision of the right eye. Cutaneous examination revealed numerous large erythematous plaques with superficial scale and cerebriform surface on the chest, back, abdomen, and upper and lower extremities involving 80% BSA (Figure 3). Biopsy of a plaque demonstrated psoriasiform dermatitis with neutrophils and parakeratosis consistent with psoriasis. Extensive blood work was notable for reactive HIV antibody and lymphopenia, CD4 lymphocyte count of 60 cells/mm³, and an HIV viral load of 247,000 copies/mL, meeting diagnostic criteria for AIDS. Additionally, ophthalmologic evaluation revealed toxoplasma retinitis. Upon initiation of highly active antiretroviral therapy (HAART) and continued use of topical corticosteroids, the patient experienced notable improvement of disease severity with approximately 20% BSA.

Comment

Molluscum contagiosum is a common skin infection. Among patients with HIV and other types of impaired cellular immunity, the prevalence of MC is estimated to be as high as 20%.³ The MC poxvirus survives and proliferates within the epidermis by interfering with tumor necrosis factor–induced apoptosis of virally infected cells; therefore, intact cell-mediated immunity is an important component of prevention and clearance of poxvirus infections. In immunocompromised patients, the presentation of MC varies widely, and the disease is often difficult to eradicate. This review will highlight the prevalence, presentation, and treatment of MC in the context of immunosuppressed states.

HIV/AIDS
Molluscum contagiosum in HIV-positive patients was first recognized in 1983,² and its prevalence is estimated to range from 5% to 18% in AIDS patients.³ Molluscum contagiosum is a clinical sign of HIV progression, and its incidence appears to increase with reduced immune function (ie, a CD4 cell count <200/mm³).³ In a study of 456 patients with HIV-associated skin disorders, the majority of patients with MC had notable immunosuppression with a median survival time of 12 months. Thus, MC was not an independent prognostic marker but a clinical indicator of markedly reduced immune status.⁴

Molluscum contagiosum is transmitted in both sexual and nonsexual patterns in HIV-positive individuals, with the distribution of the latter involving primarily the face and neck. Although it may present with typical umbilicated papules, MC has a wide range of atypical clinical presentations in patients with AIDS that can make it difficult to diagnose. Complicated cases of eyelid MC have been reported in advanced HIV in both adults and children, resulting in obstruction of vision due to large lesions (up to 2 cm) or hundreds of confluent lesions.⁵ Giant MC, which appears as large exophytic nodules, is another presentation that has been frequently described in patients with advanced HIV. In these patients, the lesions often are too voluminous for conservative therapy and require excision.⁶ Atypical MC lesions also can resemble other dermatologic conditions, including condyloma acuminatum,⁷ nevus sebaceous of Jadassohn, ecthyma,⁸ and cutaneous horns,^9,10 as well as other bacterial and fungal infections in HIV-positive patients, such as cutaneous Cryptococcus neoformans,^11,12 disseminated histoplasmosis,¹³ and infections caused by Penicillium marneffei¹⁴ and Bartonella henselae.¹⁵ In most cases of MC in HIV-positive patients, diagnosis is dependent on the examination of biopsy specimens, which maintain the same histopathologic features regardless of immune status.

The management of MC in patients with HIV/AIDS is difficult. Molluscum contagiosum has shown no evidence of spontaneous resolution in patients with HIV, and treatment with one modality is often insufficient. Treatment is most successful when a combination approach is utilized with destructive procedures (eg, curettage, cryosurgery) and adjunctive agents (eg, retinoids, cantharidin, trichloroacetic acid). Imiquimod and cidofovir have been used off label for MC in AIDS patients.¹⁶ Imiquimod, which is used to treat genital warts, another cutaneous viral infection seen in patients with HIV, has demonstrated efficacy in treating MC.¹⁶ In a randomized controlled trial comparing imiquimod cream 5% to cryotherapy for MC in healthy children, imiquimod was slow acting but better suited than cryotherapy for patients with eruptions of many small lesions.¹⁷ For HIV patients, numerous reports have described successful treatment of disseminated or recalcitrant MC with topical imiquimod.^18-20 Cidofovir, an antiviral used to treat cytomegalovirus retinitis in patients with AIDS, is a promising antiviral agent against the poxvirus family. In a study of viral DNA polymerase genes of MC virus, cidofovir inhibited MC virus DNA polymerase activity.²¹ It has been used in both topical (1% to 3%) and intravenous form to successfully treat recalcitrant and exuberant giant MC.^6,22 However, the use of cidofovir is limited by its high costs, especially when compounded into a topical formulation.²³

From a systemic standpoint, numerous reports have shown that treating the underlying HIV by optimizing HAART is the most important first step in clearing MC.^24-27 However, a special concern regarding the initiation of HAART in patients with MC as well as a markedly impaired immune function is the development of an inflammatory reaction called immune reconstitution inflammatory syndrome (IRIS). This reaction is thought to be a result of immune recovery in severely immunosuppressed patients. During the initial phase of reconstitution when CD4 lymphocyte counts rise and viral load decreases, IRIS occurs due to an inflammatory reaction to microbial and autoimmune antigens, leading to temporary clinical deterioration.²⁸ The incidence has been reported in up to 25% of patients starting HAART, and 52% to 78% of IRIS cases involve dermatologic manifestations such as varicella-zoster virus, cytomegalovirus infections, genital warts, and MC.^29,30 In a cohort study of 199 patients, 2% of patients developed MC within 6 months of initiating HAART.³¹ In a case of exuberant MC lesions after beginning HAART, the lesions spontaneously resolved with the progression of immune reconstitution.²⁸

Malignancies
Patients with hematologic malignancies such as lymphoma and leukemia comprise another subset of patients at risk for atypical presentations of MC. Molluscum contagiosum has been described in patients with hematologic malignancies such as adult T-cell leukemia/lymphoma, multiple myeloma, chronic myeloid leukemia, acute lymphoblastic leukemia, lymphomatoid papulosis, and non-Hodgkin lymphoma. In a review of MC in children with cancer, 0.5% were diagnosed with MC.^32,33 Reports also have documented eruptive MC in the presence of solid organ cancers, including lung cancer.³⁴

In patients with malignancies, the differential diagnosis should include other common dermatologic conditions such as varicella, herpes simplex, papillomas, pyoderma, and cutaneous cryptococcosis, as well as MC. Similar to HIV-positive patients, the lesions of MC described in patients with malignancies do not tend to spontaneously resolve. In a report of a pediatric patient with acute lymphoblastic leukemia, MC presented as an ulcerated lesion without any classic features, requiring biopsy for definitive diagnosis. Only partial resolution was achieved with cryotherapy and crusting of the lesion in an attempt to slow the progression.³⁵ In a series of 5 children with hematologic malignancies and MC, little improvement was noted after treatment with surgical scraping, liquid nitrogen, and salicylic acid ointment 5%. Similar to patients with HIV, improvement of immune status and function help clear the disease, and patients who reach remission and discontinue chemotherapeutic agents have a higher rate of spontaneous resolution of previously recalcitrant MC lesions.³⁶

Transplant Patients
Molluscum contagiosum in transplant patients has features similar to patients with HIV/AIDS. In organ transplant recipients, there is an increased risk for cutaneous disease from iatrogenic immunosuppression or immunosuppression through infectious or neoplastic processes.³⁷ As in other immunocompromised populations, MC often has an atypical presentation in transplant patients with more extensive involvement and recalcitrant, rapidly recurring lesions.

In a review of 145 pediatric organ transplant recipients, MC was the fourth most common skin infection after verruca vulgaris, tinea versicolor, and herpes simplex/zoster. Affecting 7% of patients, the majority of patients demonstrated clinically typical lesions; however, the disease was difficult to eradicate if multiple lesions were present.³⁷ In other reports in adults, fulminant and giant MC have been described after renal and other solid organ transplants.^38,39 Molluscum contagiosum also has been reported to mimic other skin diseases in transplant patients including tinea barbae⁴⁰ and nodular basal cell carcinomas.⁴¹

The standard treatments are identical to those used in patients with HIV, including ablative methods via liquid nitrogen, electrocautery, cantharidin, trichloroacetic acid, and topical retinoids. Similar to MC in other immunocompromised states, treatment can be difficult and usually requires multiple modalities. For children, imiquimod cream 5% has been recommended due to high clearance rates (up to 92%) and the painless nature of the treatment.^42,43

Other Iatrogenic Immunosuppressive States
Immunosuppression through the use of steroids, chemotherapeutic agents, and biologic drugs often is the result of treatment of various diseases. In patients with psoriasis treated with systemic immunosuppressive agents, there are numerous reports that describe the appearance of eruptive MC in association with methotrexate, cyclosporine, and biologics. Methotrexate acts as an immunosuppressive agent by binding to dihydrofolate reductase, which inhibits DNA synthesis in immunologically competent cells.⁴⁴ It also may block host defense mechanisms against MC by suppressing the expression of serum inflammatory cytokines such as tumor necrosis factor α (TNF-α) and IFN-γ and suppressing the activity of TNF-α inducing apoptosis of virus-infected cells. Cyclosporine used in conjunction with methotrexate may exacerbate the insult to the immune system by inhibiting the production of IFN-γ.⁴⁵ Biologics are an emerging class of drugs that have demonstrated efficacy in moderate to severe psoriasis by inhibiting TNF-α or other inflammatory molecules. Several published reports have described eruptive or atypical MC in patients on biologic medications. In one case, within 2 weeks after initiation of infliximab, a monoclonal antibody against TNF-α, a patient developed an eruption of MC involving the entire body.⁴⁶ In another report, an anti–TNF-α agent for rheumatoid arthritis was associated with atypical MC with eyelid lesions.⁴⁷

There are other skin disorders treated with immunosuppressive agents that also have been associated with MC. In a patient with pemphigus vulgaris treated with prednisolone, pimecrolimus, and azathioprine, MC lesions were observed on the face and within healed pemphigus vulgaris sites.⁴⁸ Pimecrolimus and tacrolimus, corticosteroid-sparing agents, suppress cell-mediated immunity and inhibit inflammatory cytokines such as IL-2. The infection resolved with a gradual tapering of immunosuppressive therapy and 10 sessions of cryotherapy.⁴⁸ In a case of topical pimecrolimus for pityriasis alba, the patient developed biopsy-proven MC within 2 weeks of initiating treatment in the areas that were treated with tacrolimus.⁴⁹

In nontransplant patients with iatrogenic immunosuppression, MC treatment has not been documented to be as challenging as in patients with inherent immunosuppression. Most patients respond to either withdrawal of the drug alone or to simple ablative treatments such as cryotherapy.^45,46,48 This important difference is most likely due to the presence of an otherwise intact immune system.

Conclusion

This case describes the appearance of MC in a patient with psoriasis treated with a TNF-α inhibitor who was ultimately diagnosed with AIDS. Although atypical MC infections have been documented in patients with psoriasis undergoing treatment with biologics, it is thought to be more common for MC to occur in more remarkably immunocompromised states such as AIDS. Thus, the persistence and progression of MC in our patient despite discontinuation of etanercept suggested a separate underlying process. Subsequent workup led to the diagnosis of AIDS along with the opportunistic ocular infection of toxoplasmosis retinitis. This clinical sequence consisting of psoriasis treated with a biologic agent, development of MC, and subsequent diagnosis of AIDS is unique and clinically significant to dermatologists. The presentation of psoriasis in patients with HIV can be diverse with different levels of severity and atypical clinical features. In many cases, HIV is known to exacerbate the classic clinical presentation of psoriasis. However, there are other particular presentations of psoriasis in HIV patients that have been observed, which include a predilection for scalp lesions, palmoplantar keratoderma, flexural involvement, and higher levels of immunodeficiency.⁵⁰ Although tuberculin skin tests are required prior to initiating biologic therapy due to the potential for disease reactivation, there are no requirements for HIV antibody testing. In cases of severe recalcitrant psoriasis, an HIV test should be ordered during the workup to establish an early diagnosis so that an HIV-positive patient can avoid poor outcomes from either the disease processes, the use of certain therapeutic agents, or both. Furthermore, the benefit of avoiding possible harm to the patient and potential legal action outweighs the cost of performing surveillance HIV testing in this subset of patients. Thus, due to the potential additive immunosuppressive effect of HIV with biologic therapy, providers should always assess for risk factors and consider testing for HIV in all patients before initiating treatment with immunosuppressive agents such as biologics.

Skin diseases are highly prevalent in the United States, affecting an estimated 1 in 3 Americans at any given time.^1,2 In 2009 the direct medical costs associated with skin-related diseases, including health services and prescriptions, was approximately $22 billion; the annual total economic burden was estimated to be closer to $96 billion when factoring in the cost of lost productivity and pay for symptom relief.^3,4 Effective and efficient management of skin disease is essential to minimizing cost and morbidity. Nondermatologists traditionally have diagnosed the majority of skin diseases.^5,6 In particular, primary care physicians commonly manage dermatologic conditions and often are the first health care providers to encounter patients presenting with skin problems. A predicted shortage of dermatologists will likely contribute to an increase in this trend.^7,8 Therefore, it is important to adequately prepare nondermatologists to evaluate and treat the skin conditions that they are most likely to encounter in their scope of practice.

Residents, particularly in primary care specialties, often have opportunities to spend 2 to 4 weeks with a dermatologist to learn about skin diseases; however, the skin conditions most often encountered by dermatologists may differ from those most often encountered by physicians in other specialties. For instance, one study demonstrated a disparity between the most common skin problems seen by dermatologists and internists.⁹ These dissimilarities should be recognized and addressed in curriculum content. The purpose of this study was to identify and compare the 20 most common dermatologic conditions reported by dermatologists versus those reported by nondermatologists (ie, internists, pediatricians, family physicians, emergency medicine physicians, general surgeons, otolaryngologists) from 2001 to 2010. Data also were analyzed to determine the top 20 conditions referred to dermatologists by nondermatologists as a potential indicator for areas of further improvement within medical education. With this knowledge, we hope educational curricula and self-study can be modified to reflect the current epidemiology of cutaneous diseases, thereby improving patient care.

Methods

Data from 2001 to 2010 were extracted from the National Ambulatory Medical Care Survey (NAMCS), which is an ongoing survey conducted by the National Center for Health Statistics. The NAMCS collects descriptive data regarding ambulatory visits to nonfederal office-based physicians in the United States. Participating physicians are instructed to record information about patient visits for a 1-week period, including patient demographics, insurance status, reason for visit, diagnoses, procedures, therapeutics, and referrals made at that time. Data collected for the NAMCS are entered into a multistage probability sample to produce national estimates. Within dermatology, an average of 118 dermatologists are sampled each year, and over the last 10 years, participation rates have ranged from 47% to 77%.

International Classification of Diseases, Ninth Revision, Clinical Modification codes were identified to determine the diagnoses that could be classified as dermatologic conditions. Select infectious and neoplastic disorders of the skin and mucous membrane conditions were included as well as the codes for skin diseases. Nondermatologic diagnoses and V codes were not included in the study. Data for all providers were studied to identify outpatient visits associated with the primary diagnosis of a dermatologic condition. Minor diagnoses that were considered to be subsets of major diagnoses were combined to allow better analysis of the data. For example, all tinea infections (ie, dermatophytosis of various sites, dermatomycosis unspecified) were combined into 1 diagnosis referred to as tinea because the recognition and treatment of this disease does not vary tremendously by anatomic location. Visits to dermatologists that listed nonspecific diagnoses and codes (eg, other postsurgical status [V45.89], neoplasm of uncertain behavior site unspecified [238.9]) were assumed to be for dermatologic problems.

Sampling weights were applied to obtain estimates for the number of each diagnosis made nationally. All data analyses were performed using SAS software and linear regression models were generated using SAS PROC SURVEYREG.

Data were analyzed to determine the dermatologic conditions most commonly encountered by dermatologists and nondermatologists in emergency medicine, family medicine, general surgery, internal medicine, otolaryngology, and pediatrics; these specialties include physicians who are known to commonly diagnose and treat skin diseases.¹⁰ Data also were analyzed to determine the most common conditions referred to dermatologists for treatment by nondermatologists from the selected specialties. Permission to conduct this study was obtained from the Wake Forest University institutional review board (Winston-Salem, North Carolina).

Results

From 2001 to 2010, more than 700 million outpatient visits for skin-related problems were identified, with 676.3 million visits to dermatologists, emergency medicine physicians, family practitioners, general surgeons, internists, otolaryngologists, and pediatricians. More than half (52.9%) of all skin-related visits were addressed by nondermatologists during this time. Among nondermatologists, family practitioners encountered the greatest number of skin diseases (20.5%), followed by pediatricians (11.3%), internists (9.2%), general surgeons (3.4%), otolaryngologists (1.0%), and emergency medicine physicians (0.2%)(Table 1).

Benign tumors and acne were the most common cutaneous conditions referred to dermatologists by nondermatologists (10.6% and 10.1% of all dermatology referrals, respectively), followed by nonmelanoma skin cancers (9.7%), contact dermatitis (8.8%), and actinic keratosis (7.8%)(Table 2). The top 20 conditions referred to dermatologists accounted for 83.7% of all outpatient referrals to dermatologists.

Among the diseases most frequently reported by nondermatologists, contact dermatitis was the most common (12.0%), with twice the number of visits to nondermatologists for contact dermatitis than to dermatologists (51.6 million vs 25.3 million). In terms of disease categories, infectious skin diseases (ie, bacterial [cellulitis/abscess], viral [warts, herpesvirus], fungal [tinea] and yeast [candida] etiologies) were the most common dermatologic conditions reported by nondermatologists (Table 2).

The top 20 dermatologic conditions reported by dermatologists accounted for 85.4% of all diagnoses made by dermatologists. Diseases that were among the top 20 conditions encountered by dermatologists but were not among the top 20 for nondermatologists included actinic keratosis, seborrheic keratosis, atopic dermatitis, psoriasis, alopecia, rosacea, dyschromia, seborrheic dermatitis, follicular disease, and neoplasm of uncertain behavior of skin. Additionally, 5 of the top 20 conditions encountered by dermatologists also were among the top 20 for only 1 individual nondermatologic specialty; these included atopic dermatitis (pediatrics), seborrheic dermatitis (pediatrics), psoriasis (internal medicine), rosacea (otolaryngology), and keratoderma (general surgery). Seborrheic dermatitis, psoriasis, and rosacea also were among the top 20 conditions most commonly referred to dermatologists for treatment by nondermatologists. Table 3 shows the top 20 dermatologic conditions encountered by nondermatologists by comparison.

Comment

According to NAMCS data from 2001 to 2010, visits to nondermatologists accounted for more than half of total outpatient visits for cutaneous diseases in the United States, whereas visits to dermatologists accounted for 47.1%. These findings are consistent with historical data indicating that 30% to 40% of skin-related visits are to dermatologists, and the majority of patients with skin disease are diagnosed by nondermatologists.^5,6

Past data indicate that most visits to dermatologists were for evaluation of acne, infections, psoriasis, and neoplasms, whereas most visits to nondermatologists were for evaluation of epidermoid cysts, impetigo, plant dermatitis, cellulitis, and diaper rash.⁹ Over the last 10 years, acne has been more commonly encountered by nondermatologists, especially pediatricians. Additionally, infectious etiologies have been seen in larger volume by nondermatologists.⁹ Together, infectious cutaneous conditions make up nearly one-fourth of dermatologic encounters by emergency medicine physicians, internists, and family practitioners but are not within the top 20 diagnoses referred to dermatologists, which suggests that uncomplicated cases of cellulitis, herpes zoster, and other skin-related infections are largely managed by nondermatologists.^5,6 Contact dermatitis, often caused by specific allergens such as detergents, solvents, and topical products, was one of the most common reported dermatologic encounters among dermatologists and nondermatologists and also was the fourth most common condition referred to dermatologists by nondermatologists for treatment; however, there may be an element of overuse of the International Classification of Diseases, Ninth Revision code, as any presumed contact dermatitis of unspecified cause can be reported under 692.9 defined as contact dermatitis and other eczema, unspecified cause. The high rate of referrals to dermatologists by nondermatologists may be for patch testing and further management. Additionally, there are no specific codes for allergic or irritant dermatitis, thus these diseases may be lumped together.

Although nearly half of all dermatologic encounters were seen by nondermatologists, dermatologists see a much larger proportion of patients with skin disease than nondermatologists and nondermatologists often have limited exposure to the field of dermatology during residency training. Studies have demonstrated differences in the abilities of dermatologists and nondermatologists to correctly diagnose common cutaneous diseases, which unsurprisingly revealed greater diagnostic accuracy demonstrated by dermatologists.^11-16 The increase in acne and skin-related infections reported by nondermatologists is consistent with possible efforts to increase formal training in frequently encountered skin diseases. In one study evaluating the impact of a formal 3-week dermatology curriculum on an internal medicine department, internists demonstrated 100% accuracy in the diagnosis of acne and herpes zoster in contrast to 29% for tinea and 12% for lichen planus.^5,6

The current Accreditation Council for Graduate Medical Education guidelines place little emphasis on exposure to dermatology training during residency for internists and pediatricians, as this training is not a required component of these programs.¹⁷ Two core problems with current training regarding the evaluation and management of cutaneous disease are minimal exposure to dermatologic conditions in medical school and residency and lack of consensus on the core topics that should be taught to nondermatologists.¹⁸ Exposure to dermatologic conditions through rotations in medical school has been shown to increase residents’ self-reported confidence in diagnosing and treating alopecia, cutaneous drug eruptions, warts, acne, rosacea, nonmelanoma skin cancers, sun damage, psoriasis, seborrhea, atopic dermatitis, and contact dermatitis; however, the majority of primary care residents surveyed still felt that this exposure in medical school was inadequate.¹⁹

In creating a core curriculum for dermatology training for nondermatologists, it is important to consider the dermatologic conditions that are most frequently encountered by these specialties. Our study revealed that the most commonly encountered dermatologic conditions differ among dermatologists and nondermatologists, with a fair degree of variation even among individual specialties. Failure to recognize these discrepancies has likely contributed to the challenges faced by nondermatologists in the diagnosis and management of dermatologic disease. In this study, contact dermatitis, epidermoid cysts, and skin infections were the most common dermatologic conditions encountered by nondermatologists and also were among the top skin diseases referred to dermatologists by nondermatologists. This finding suggests that nondermatologists are able to identify these conditions but have a tendency to refer approximately 10% of these patients to dermatology for further management. Clinical evaluation and medical management of these cutaneous diseases may be an important area of focus for medical school curricula, as the treatment of these diseases is within the capabilities of the nondermatologist. For example, initial management of dermatitis requires determination of the type of dermatitis (ie, essential, contact, atopic, seborrheic, stasis) and selection of an appropriate topical steroid, with referral to a dermatologist needed for questionable or refractory cases. Although a curriculum cannot be built solely on a list of the top 20 diagnoses provided here, these data may serve as a preliminary platform for medical school dermatology curriculum design. The curriculum also should include serious skin diseases, such as melanoma and severe drug eruptions. Although these conditions are less commonly encountered by nondermatologists, missed diagnosis and/or improper management can be life threatening.

The use of NAMCS data presents a few limitations. For instance, these data only represent outpatient management of skin disease. There is the potential for misdiagnosis and coding errors by the reporting physicians. The volume of data (ie, billions of office visits) prevents verification of diagnostic accuracy. The coding system requires physicians to give a diagnosis but does not provide any means by which to determine the physician’s confidence in that diagnosis. There is no code for “uncertain” or “diagnosis not determined.” Additionally, an “unspecified” diagnosis may reflect uncertainty or may simply imply that no other code accurately described the condition. Despite these limitations, the NAMCS database is a large, nationally representative survey of actual patient visits and represents some of the best data available for a study such as ours.

Conclusion

This study provides an important analysis of the most common outpatient dermatologic conditions encountered by dermatologists and nondermatologists of various specialties and offers a foundation from which to construct curricula for dermatology training tailored to individual specialties based on their needs. In the future, identification of the most common inpatient dermatologic conditions managed by each specialty also may benefit curriculum design.

Skin diseases are highly prevalent in the United States, affecting an estimated 1 in 3 Americans at any given time.^1,2 In 2009 the direct medical costs associated with skin-related diseases, including health services and prescriptions, was approximately $22 billion; the annual total economic burden was estimated to be closer to $96 billion when factoring in the cost of lost productivity and pay for symptom relief.^3,4 Effective and efficient management of skin disease is essential to minimizing cost and morbidity. Nondermatologists traditionally have diagnosed the majority of skin diseases.^5,6 In particular, primary care physicians commonly manage dermatologic conditions and often are the first health care providers to encounter patients presenting with skin problems. A predicted shortage of dermatologists will likely contribute to an increase in this trend.^7,8 Therefore, it is important to adequately prepare nondermatologists to evaluate and treat the skin conditions that they are most likely to encounter in their scope of practice.

Residents, particularly in primary care specialties, often have opportunities to spend 2 to 4 weeks with a dermatologist to learn about skin diseases; however, the skin conditions most often encountered by dermatologists may differ from those most often encountered by physicians in other specialties. For instance, one study demonstrated a disparity between the most common skin problems seen by dermatologists and internists.⁹ These dissimilarities should be recognized and addressed in curriculum content. The purpose of this study was to identify and compare the 20 most common dermatologic conditions reported by dermatologists versus those reported by nondermatologists (ie, internists, pediatricians, family physicians, emergency medicine physicians, general surgeons, otolaryngologists) from 2001 to 2010. Data also were analyzed to determine the top 20 conditions referred to dermatologists by nondermatologists as a potential indicator for areas of further improvement within medical education. With this knowledge, we hope educational curricula and self-study can be modified to reflect the current epidemiology of cutaneous diseases, thereby improving patient care.

Methods

Data from 2001 to 2010 were extracted from the National Ambulatory Medical Care Survey (NAMCS), which is an ongoing survey conducted by the National Center for Health Statistics. The NAMCS collects descriptive data regarding ambulatory visits to nonfederal office-based physicians in the United States. Participating physicians are instructed to record information about patient visits for a 1-week period, including patient demographics, insurance status, reason for visit, diagnoses, procedures, therapeutics, and referrals made at that time. Data collected for the NAMCS are entered into a multistage probability sample to produce national estimates. Within dermatology, an average of 118 dermatologists are sampled each year, and over the last 10 years, participation rates have ranged from 47% to 77%.

International Classification of Diseases, Ninth Revision, Clinical Modification codes were identified to determine the diagnoses that could be classified as dermatologic conditions. Select infectious and neoplastic disorders of the skin and mucous membrane conditions were included as well as the codes for skin diseases. Nondermatologic diagnoses and V codes were not included in the study. Data for all providers were studied to identify outpatient visits associated with the primary diagnosis of a dermatologic condition. Minor diagnoses that were considered to be subsets of major diagnoses were combined to allow better analysis of the data. For example, all tinea infections (ie, dermatophytosis of various sites, dermatomycosis unspecified) were combined into 1 diagnosis referred to as tinea because the recognition and treatment of this disease does not vary tremendously by anatomic location. Visits to dermatologists that listed nonspecific diagnoses and codes (eg, other postsurgical status [V45.89], neoplasm of uncertain behavior site unspecified [238.9]) were assumed to be for dermatologic problems.

Sampling weights were applied to obtain estimates for the number of each diagnosis made nationally. All data analyses were performed using SAS software and linear regression models were generated using SAS PROC SURVEYREG.

Data were analyzed to determine the dermatologic conditions most commonly encountered by dermatologists and nondermatologists in emergency medicine, family medicine, general surgery, internal medicine, otolaryngology, and pediatrics; these specialties include physicians who are known to commonly diagnose and treat skin diseases.¹⁰ Data also were analyzed to determine the most common conditions referred to dermatologists for treatment by nondermatologists from the selected specialties. Permission to conduct this study was obtained from the Wake Forest University institutional review board (Winston-Salem, North Carolina).

Results

From 2001 to 2010, more than 700 million outpatient visits for skin-related problems were identified, with 676.3 million visits to dermatologists, emergency medicine physicians, family practitioners, general surgeons, internists, otolaryngologists, and pediatricians. More than half (52.9%) of all skin-related visits were addressed by nondermatologists during this time. Among nondermatologists, family practitioners encountered the greatest number of skin diseases (20.5%), followed by pediatricians (11.3%), internists (9.2%), general surgeons (3.4%), otolaryngologists (1.0%), and emergency medicine physicians (0.2%)(Table 1).

Benign tumors and acne were the most common cutaneous conditions referred to dermatologists by nondermatologists (10.6% and 10.1% of all dermatology referrals, respectively), followed by nonmelanoma skin cancers (9.7%), contact dermatitis (8.8%), and actinic keratosis (7.8%)(Table 2). The top 20 conditions referred to dermatologists accounted for 83.7% of all outpatient referrals to dermatologists.

Among the diseases most frequently reported by nondermatologists, contact dermatitis was the most common (12.0%), with twice the number of visits to nondermatologists for contact dermatitis than to dermatologists (51.6 million vs 25.3 million). In terms of disease categories, infectious skin diseases (ie, bacterial [cellulitis/abscess], viral [warts, herpesvirus], fungal [tinea] and yeast [candida] etiologies) were the most common dermatologic conditions reported by nondermatologists (Table 2).

The top 20 dermatologic conditions reported by dermatologists accounted for 85.4% of all diagnoses made by dermatologists. Diseases that were among the top 20 conditions encountered by dermatologists but were not among the top 20 for nondermatologists included actinic keratosis, seborrheic keratosis, atopic dermatitis, psoriasis, alopecia, rosacea, dyschromia, seborrheic dermatitis, follicular disease, and neoplasm of uncertain behavior of skin. Additionally, 5 of the top 20 conditions encountered by dermatologists also were among the top 20 for only 1 individual nondermatologic specialty; these included atopic dermatitis (pediatrics), seborrheic dermatitis (pediatrics), psoriasis (internal medicine), rosacea (otolaryngology), and keratoderma (general surgery). Seborrheic dermatitis, psoriasis, and rosacea also were among the top 20 conditions most commonly referred to dermatologists for treatment by nondermatologists. Table 3 shows the top 20 dermatologic conditions encountered by nondermatologists by comparison.

Comment

According to NAMCS data from 2001 to 2010, visits to nondermatologists accounted for more than half of total outpatient visits for cutaneous diseases in the United States, whereas visits to dermatologists accounted for 47.1%. These findings are consistent with historical data indicating that 30% to 40% of skin-related visits are to dermatologists, and the majority of patients with skin disease are diagnosed by nondermatologists.^5,6

Past data indicate that most visits to dermatologists were for evaluation of acne, infections, psoriasis, and neoplasms, whereas most visits to nondermatologists were for evaluation of epidermoid cysts, impetigo, plant dermatitis, cellulitis, and diaper rash.⁹ Over the last 10 years, acne has been more commonly encountered by nondermatologists, especially pediatricians. Additionally, infectious etiologies have been seen in larger volume by nondermatologists.⁹ Together, infectious cutaneous conditions make up nearly one-fourth of dermatologic encounters by emergency medicine physicians, internists, and family practitioners but are not within the top 20 diagnoses referred to dermatologists, which suggests that uncomplicated cases of cellulitis, herpes zoster, and other skin-related infections are largely managed by nondermatologists.^5,6 Contact dermatitis, often caused by specific allergens such as detergents, solvents, and topical products, was one of the most common reported dermatologic encounters among dermatologists and nondermatologists and also was the fourth most common condition referred to dermatologists by nondermatologists for treatment; however, there may be an element of overuse of the International Classification of Diseases, Ninth Revision code, as any presumed contact dermatitis of unspecified cause can be reported under 692.9 defined as contact dermatitis and other eczema, unspecified cause. The high rate of referrals to dermatologists by nondermatologists may be for patch testing and further management. Additionally, there are no specific codes for allergic or irritant dermatitis, thus these diseases may be lumped together.

Although nearly half of all dermatologic encounters were seen by nondermatologists, dermatologists see a much larger proportion of patients with skin disease than nondermatologists and nondermatologists often have limited exposure to the field of dermatology during residency training. Studies have demonstrated differences in the abilities of dermatologists and nondermatologists to correctly diagnose common cutaneous diseases, which unsurprisingly revealed greater diagnostic accuracy demonstrated by dermatologists.^11-16 The increase in acne and skin-related infections reported by nondermatologists is consistent with possible efforts to increase formal training in frequently encountered skin diseases. In one study evaluating the impact of a formal 3-week dermatology curriculum on an internal medicine department, internists demonstrated 100% accuracy in the diagnosis of acne and herpes zoster in contrast to 29% for tinea and 12% for lichen planus.^5,6

The current Accreditation Council for Graduate Medical Education guidelines place little emphasis on exposure to dermatology training during residency for internists and pediatricians, as this training is not a required component of these programs.¹⁷ Two core problems with current training regarding the evaluation and management of cutaneous disease are minimal exposure to dermatologic conditions in medical school and residency and lack of consensus on the core topics that should be taught to nondermatologists.¹⁸ Exposure to dermatologic conditions through rotations in medical school has been shown to increase residents’ self-reported confidence in diagnosing and treating alopecia, cutaneous drug eruptions, warts, acne, rosacea, nonmelanoma skin cancers, sun damage, psoriasis, seborrhea, atopic dermatitis, and contact dermatitis; however, the majority of primary care residents surveyed still felt that this exposure in medical school was inadequate.¹⁹

In creating a core curriculum for dermatology training for nondermatologists, it is important to consider the dermatologic conditions that are most frequently encountered by these specialties. Our study revealed that the most commonly encountered dermatologic conditions differ among dermatologists and nondermatologists, with a fair degree of variation even among individual specialties. Failure to recognize these discrepancies has likely contributed to the challenges faced by nondermatologists in the diagnosis and management of dermatologic disease. In this study, contact dermatitis, epidermoid cysts, and skin infections were the most common dermatologic conditions encountered by nondermatologists and also were among the top skin diseases referred to dermatologists by nondermatologists. This finding suggests that nondermatologists are able to identify these conditions but have a tendency to refer approximately 10% of these patients to dermatology for further management. Clinical evaluation and medical management of these cutaneous diseases may be an important area of focus for medical school curricula, as the treatment of these diseases is within the capabilities of the nondermatologist. For example, initial management of dermatitis requires determination of the type of dermatitis (ie, essential, contact, atopic, seborrheic, stasis) and selection of an appropriate topical steroid, with referral to a dermatologist needed for questionable or refractory cases. Although a curriculum cannot be built solely on a list of the top 20 diagnoses provided here, these data may serve as a preliminary platform for medical school dermatology curriculum design. The curriculum also should include serious skin diseases, such as melanoma and severe drug eruptions. Although these conditions are less commonly encountered by nondermatologists, missed diagnosis and/or improper management can be life threatening.

The use of NAMCS data presents a few limitations. For instance, these data only represent outpatient management of skin disease. There is the potential for misdiagnosis and coding errors by the reporting physicians. The volume of data (ie, billions of office visits) prevents verification of diagnostic accuracy. The coding system requires physicians to give a diagnosis but does not provide any means by which to determine the physician’s confidence in that diagnosis. There is no code for “uncertain” or “diagnosis not determined.” Additionally, an “unspecified” diagnosis may reflect uncertainty or may simply imply that no other code accurately described the condition. Despite these limitations, the NAMCS database is a large, nationally representative survey of actual patient visits and represents some of the best data available for a study such as ours.

Conclusion

This study provides an important analysis of the most common outpatient dermatologic conditions encountered by dermatologists and nondermatologists of various specialties and offers a foundation from which to construct curricula for dermatology training tailored to individual specialties based on their needs. In the future, identification of the most common inpatient dermatologic conditions managed by each specialty also may benefit curriculum design.

Skin diseases are highly prevalent in the United States, affecting an estimated 1 in 3 Americans at any given time.^1,2 In 2009 the direct medical costs associated with skin-related diseases, including health services and prescriptions, was approximately $22 billion; the annual total economic burden was estimated to be closer to $96 billion when factoring in the cost of lost productivity and pay for symptom relief.^3,4 Effective and efficient management of skin disease is essential to minimizing cost and morbidity. Nondermatologists traditionally have diagnosed the majority of skin diseases.^5,6 In particular, primary care physicians commonly manage dermatologic conditions and often are the first health care providers to encounter patients presenting with skin problems. A predicted shortage of dermatologists will likely contribute to an increase in this trend.^7,8 Therefore, it is important to adequately prepare nondermatologists to evaluate and treat the skin conditions that they are most likely to encounter in their scope of practice.

Residents, particularly in primary care specialties, often have opportunities to spend 2 to 4 weeks with a dermatologist to learn about skin diseases; however, the skin conditions most often encountered by dermatologists may differ from those most often encountered by physicians in other specialties. For instance, one study demonstrated a disparity between the most common skin problems seen by dermatologists and internists.⁹ These dissimilarities should be recognized and addressed in curriculum content. The purpose of this study was to identify and compare the 20 most common dermatologic conditions reported by dermatologists versus those reported by nondermatologists (ie, internists, pediatricians, family physicians, emergency medicine physicians, general surgeons, otolaryngologists) from 2001 to 2010. Data also were analyzed to determine the top 20 conditions referred to dermatologists by nondermatologists as a potential indicator for areas of further improvement within medical education. With this knowledge, we hope educational curricula and self-study can be modified to reflect the current epidemiology of cutaneous diseases, thereby improving patient care.

Methods

Data from 2001 to 2010 were extracted from the National Ambulatory Medical Care Survey (NAMCS), which is an ongoing survey conducted by the National Center for Health Statistics. The NAMCS collects descriptive data regarding ambulatory visits to nonfederal office-based physicians in the United States. Participating physicians are instructed to record information about patient visits for a 1-week period, including patient demographics, insurance status, reason for visit, diagnoses, procedures, therapeutics, and referrals made at that time. Data collected for the NAMCS are entered into a multistage probability sample to produce national estimates. Within dermatology, an average of 118 dermatologists are sampled each year, and over the last 10 years, participation rates have ranged from 47% to 77%.

International Classification of Diseases, Ninth Revision, Clinical Modification codes were identified to determine the diagnoses that could be classified as dermatologic conditions. Select infectious and neoplastic disorders of the skin and mucous membrane conditions were included as well as the codes for skin diseases. Nondermatologic diagnoses and V codes were not included in the study. Data for all providers were studied to identify outpatient visits associated with the primary diagnosis of a dermatologic condition. Minor diagnoses that were considered to be subsets of major diagnoses were combined to allow better analysis of the data. For example, all tinea infections (ie, dermatophytosis of various sites, dermatomycosis unspecified) were combined into 1 diagnosis referred to as tinea because the recognition and treatment of this disease does not vary tremendously by anatomic location. Visits to dermatologists that listed nonspecific diagnoses and codes (eg, other postsurgical status [V45.89], neoplasm of uncertain behavior site unspecified [238.9]) were assumed to be for dermatologic problems.

Sampling weights were applied to obtain estimates for the number of each diagnosis made nationally. All data analyses were performed using SAS software and linear regression models were generated using SAS PROC SURVEYREG.

Data were analyzed to determine the dermatologic conditions most commonly encountered by dermatologists and nondermatologists in emergency medicine, family medicine, general surgery, internal medicine, otolaryngology, and pediatrics; these specialties include physicians who are known to commonly diagnose and treat skin diseases.¹⁰ Data also were analyzed to determine the most common conditions referred to dermatologists for treatment by nondermatologists from the selected specialties. Permission to conduct this study was obtained from the Wake Forest University institutional review board (Winston-Salem, North Carolina).

Results

From 2001 to 2010, more than 700 million outpatient visits for skin-related problems were identified, with 676.3 million visits to dermatologists, emergency medicine physicians, family practitioners, general surgeons, internists, otolaryngologists, and pediatricians. More than half (52.9%) of all skin-related visits were addressed by nondermatologists during this time. Among nondermatologists, family practitioners encountered the greatest number of skin diseases (20.5%), followed by pediatricians (11.3%), internists (9.2%), general surgeons (3.4%), otolaryngologists (1.0%), and emergency medicine physicians (0.2%)(Table 1).

Benign tumors and acne were the most common cutaneous conditions referred to dermatologists by nondermatologists (10.6% and 10.1% of all dermatology referrals, respectively), followed by nonmelanoma skin cancers (9.7%), contact dermatitis (8.8%), and actinic keratosis (7.8%)(Table 2). The top 20 conditions referred to dermatologists accounted for 83.7% of all outpatient referrals to dermatologists.

Among the diseases most frequently reported by nondermatologists, contact dermatitis was the most common (12.0%), with twice the number of visits to nondermatologists for contact dermatitis than to dermatologists (51.6 million vs 25.3 million). In terms of disease categories, infectious skin diseases (ie, bacterial [cellulitis/abscess], viral [warts, herpesvirus], fungal [tinea] and yeast [candida] etiologies) were the most common dermatologic conditions reported by nondermatologists (Table 2).

The top 20 dermatologic conditions reported by dermatologists accounted for 85.4% of all diagnoses made by dermatologists. Diseases that were among the top 20 conditions encountered by dermatologists but were not among the top 20 for nondermatologists included actinic keratosis, seborrheic keratosis, atopic dermatitis, psoriasis, alopecia, rosacea, dyschromia, seborrheic dermatitis, follicular disease, and neoplasm of uncertain behavior of skin. Additionally, 5 of the top 20 conditions encountered by dermatologists also were among the top 20 for only 1 individual nondermatologic specialty; these included atopic dermatitis (pediatrics), seborrheic dermatitis (pediatrics), psoriasis (internal medicine), rosacea (otolaryngology), and keratoderma (general surgery). Seborrheic dermatitis, psoriasis, and rosacea also were among the top 20 conditions most commonly referred to dermatologists for treatment by nondermatologists. Table 3 shows the top 20 dermatologic conditions encountered by nondermatologists by comparison.

Comment

According to NAMCS data from 2001 to 2010, visits to nondermatologists accounted for more than half of total outpatient visits for cutaneous diseases in the United States, whereas visits to dermatologists accounted for 47.1%. These findings are consistent with historical data indicating that 30% to 40% of skin-related visits are to dermatologists, and the majority of patients with skin disease are diagnosed by nondermatologists.^5,6

Past data indicate that most visits to dermatologists were for evaluation of acne, infections, psoriasis, and neoplasms, whereas most visits to nondermatologists were for evaluation of epidermoid cysts, impetigo, plant dermatitis, cellulitis, and diaper rash.⁹ Over the last 10 years, acne has been more commonly encountered by nondermatologists, especially pediatricians. Additionally, infectious etiologies have been seen in larger volume by nondermatologists.⁹ Together, infectious cutaneous conditions make up nearly one-fourth of dermatologic encounters by emergency medicine physicians, internists, and family practitioners but are not within the top 20 diagnoses referred to dermatologists, which suggests that uncomplicated cases of cellulitis, herpes zoster, and other skin-related infections are largely managed by nondermatologists.^5,6 Contact dermatitis, often caused by specific allergens such as detergents, solvents, and topical products, was one of the most common reported dermatologic encounters among dermatologists and nondermatologists and also was the fourth most common condition referred to dermatologists by nondermatologists for treatment; however, there may be an element of overuse of the International Classification of Diseases, Ninth Revision code, as any presumed contact dermatitis of unspecified cause can be reported under 692.9 defined as contact dermatitis and other eczema, unspecified cause. The high rate of referrals to dermatologists by nondermatologists may be for patch testing and further management. Additionally, there are no specific codes for allergic or irritant dermatitis, thus these diseases may be lumped together.

Although nearly half of all dermatologic encounters were seen by nondermatologists, dermatologists see a much larger proportion of patients with skin disease than nondermatologists and nondermatologists often have limited exposure to the field of dermatology during residency training. Studies have demonstrated differences in the abilities of dermatologists and nondermatologists to correctly diagnose common cutaneous diseases, which unsurprisingly revealed greater diagnostic accuracy demonstrated by dermatologists.^11-16 The increase in acne and skin-related infections reported by nondermatologists is consistent with possible efforts to increase formal training in frequently encountered skin diseases. In one study evaluating the impact of a formal 3-week dermatology curriculum on an internal medicine department, internists demonstrated 100% accuracy in the diagnosis of acne and herpes zoster in contrast to 29% for tinea and 12% for lichen planus.^5,6

The current Accreditation Council for Graduate Medical Education guidelines place little emphasis on exposure to dermatology training during residency for internists and pediatricians, as this training is not a required component of these programs.¹⁷ Two core problems with current training regarding the evaluation and management of cutaneous disease are minimal exposure to dermatologic conditions in medical school and residency and lack of consensus on the core topics that should be taught to nondermatologists.¹⁸ Exposure to dermatologic conditions through rotations in medical school has been shown to increase residents’ self-reported confidence in diagnosing and treating alopecia, cutaneous drug eruptions, warts, acne, rosacea, nonmelanoma skin cancers, sun damage, psoriasis, seborrhea, atopic dermatitis, and contact dermatitis; however, the majority of primary care residents surveyed still felt that this exposure in medical school was inadequate.¹⁹

In creating a core curriculum for dermatology training for nondermatologists, it is important to consider the dermatologic conditions that are most frequently encountered by these specialties. Our study revealed that the most commonly encountered dermatologic conditions differ among dermatologists and nondermatologists, with a fair degree of variation even among individual specialties. Failure to recognize these discrepancies has likely contributed to the challenges faced by nondermatologists in the diagnosis and management of dermatologic disease. In this study, contact dermatitis, epidermoid cysts, and skin infections were the most common dermatologic conditions encountered by nondermatologists and also were among the top skin diseases referred to dermatologists by nondermatologists. This finding suggests that nondermatologists are able to identify these conditions but have a tendency to refer approximately 10% of these patients to dermatology for further management. Clinical evaluation and medical management of these cutaneous diseases may be an important area of focus for medical school curricula, as the treatment of these diseases is within the capabilities of the nondermatologist. For example, initial management of dermatitis requires determination of the type of dermatitis (ie, essential, contact, atopic, seborrheic, stasis) and selection of an appropriate topical steroid, with referral to a dermatologist needed for questionable or refractory cases. Although a curriculum cannot be built solely on a list of the top 20 diagnoses provided here, these data may serve as a preliminary platform for medical school dermatology curriculum design. The curriculum also should include serious skin diseases, such as melanoma and severe drug eruptions. Although these conditions are less commonly encountered by nondermatologists, missed diagnosis and/or improper management can be life threatening.

The use of NAMCS data presents a few limitations. For instance, these data only represent outpatient management of skin disease. There is the potential for misdiagnosis and coding errors by the reporting physicians. The volume of data (ie, billions of office visits) prevents verification of diagnostic accuracy. The coding system requires physicians to give a diagnosis but does not provide any means by which to determine the physician’s confidence in that diagnosis. There is no code for “uncertain” or “diagnosis not determined.” Additionally, an “unspecified” diagnosis may reflect uncertainty or may simply imply that no other code accurately described the condition. Despite these limitations, the NAMCS database is a large, nationally representative survey of actual patient visits and represents some of the best data available for a study such as ours.

Conclusion

This study provides an important analysis of the most common outpatient dermatologic conditions encountered by dermatologists and nondermatologists of various specialties and offers a foundation from which to construct curricula for dermatology training tailored to individual specialties based on their needs. In the future, identification of the most common inpatient dermatologic conditions managed by each specialty also may benefit curriculum design.