Journal of the American Academy of Dermatology
Volume 62, Issue 1 , Pages 61-66, January 2010

Pediatric teledermatology: Observations based on 429 consults

  • Tina S. Chen, MD

      Affiliations

    • Department of Dermatology, University of California, Irvine, California
    • ,
  • Marc E. Goldyne, MD, PhD

      Affiliations

    • Department of Dermatology, University of California, San Francisco, California
    • ,
  • Erin F.D. Mathes, MD

      Affiliations

    • Department of Dermatology, University of California, San Francisco, California
    • Department of Pediatrics, University of California, San Francisco, California
    • ,
  • Ilona J. Frieden, MD

      Affiliations

    • Department of Dermatology, University of California, San Francisco, California
    • Department of Pediatrics, University of California, San Francisco, California
    • ,
  • Amy E. Gilliam, MD

      Affiliations

    • Department of Dermatology, University of California, San Francisco, California
    • Department of Pediatrics, University of California, San Francisco, California
    • Corresponding Author InformationReprint requests: Amy E. Gilliam, MD, Palo Alto Medical Foundation, 795 El Camino Real, Clark Bldg, Palo Alto, CA 94301.

Accepted 26 May 2009. published online 19 November 2009.

Article Outline

Background

Store-and-forward teledermatology is an emerging means of access for patients with skin disease lacking direct access to dermatologists.

Objectives

We sought to examine the patient demographics, diagnostic concordance, and treatment patterns in teledermatology for patients younger than 13 years.

Methods

We conducted a descriptive retrospective cohort study involving 429 patients.

Results

Diagnoses were concordant in 48% of cases, partially concordant in 10%, and discordant in 42%. Management recommendations were concordant in 28% of cases, partially concordant in 36%, and discordant in 36%. Primary care providers tended to underuse topical steroids and overuse topical antifungals and systemic antibiotics. Only 1.4% and 6.0% of patients required repeated teledermatology consultation and in-person dermatology consultation, respectively.

Limitations

Limitations were the inability to generalize the data from the population studied and the chances of error and bias in teledermatology diagnoses.

Conclusions

Store-and-forward teledermatology can improve diagnostic and therapeutic care for skin disease in children who lack direct access to dermatologists.

Key words: diagnosis, management, pediatric, store-and-forward, teledermatology

 

Capsule Summary

 


•Store-and-forward teledermatology is an emerging means of access for patients with skin disease lacking direct access to dermatologists, and can improve diagnostic and therapeutic care for skin disease in children.

•Agreement in diagnosis between the primary care provider and the teledermatologist was 48%, which is similar to previously reported studies. Agreement in management was found in only 28% of cases.

•Most commonly referred conditions mirrored those seen in general pediatric dermatology outpatient clinics, except for an increased number of teledermatology referrals for disorders of pigmentation.

•This study supports the conclusion of earlier reports that dermatologists are more likely than pediatricians to prescribe both topical steroids and higher potency topical steroids.

Telemedicine is the practice of health care delivery using digital data shared over public or private computer networks. There are two basic technologies: live-interactive and store-and-forward telemedicine. The former uses a real-time, broad-bandwidth audio-video link-up between the referring providers and the specialists; the latter requires the referral center to upload relevant text and image data onto their computer and forward them through encrypted e-mail to a specialist who makes diagnostic and therapeutic recommendations based on the data received. For dermatology, both of these techniques have been shown to be equally effective.1

The use of teledermatology as a branch of telemedicine has increased during the last several years. Currently, there are at least 115 programs in the United States; 47 provide teledermatology services in 32 states, and 4 of these are in California.2 Because the distribution of dermatologists is skewed toward metropolitan areas, many rural regions have no access to a dermatologist. Consequently, teledermatology offers a potentially significant means for improving the delivery and quality of skin care to underserved areas.3 Studies have shown that between 18.5% and 31% of clinic visits can be avoided when store-and-forward teledermatology consultation is available.4 Teledermatology may also decrease the cost of care in rural communities without a local dermatologist.5

Although a number of published studies have compared the abilities of dermatologists and nondermatologists to diagnose and treat skin disease,6, 7, 8, 9 few have focused on dermatoses affecting children10, 11, 12 and there are even fewer studies of teledermatology that have focused on the pediatric age group.13

To acquire a better understanding of potential issues that might arise in using teledermatology to care for children with skin disease, we performed a retrospective study of the electronic medical records of 429 children, aged 12 years or younger, referred from rural primary care facilities in California using store-and-forward technology to a teledermatology practice located in the San Francisco bay area.

Back to Article Outline

Methods 

Design, setting, and patients 

We undertook a retrospective cohort study using the electronic medical records of pediatric patients in a private store-and-forward teledermatology practice in the San Francisco bay area. At each referral site, a medical provider had completed a standardized electronic referral form and acquired digital images of the skin lesions. The referral form included a provisional diagnosis in 405 of the 429 study patients. These files were sent by encrypted e-mail for review by the teledermatologist, who generated an electronic consultation report providing a diagnosis (or differential diagnosis) and recommendations for additional tests (eg, biopsy) and therapy. Software (Second Opinion Software, LLC, Torrance, CA) was used for all referrals and consultations, and this software was responsible for encrypting the patient files. In Second Opinion software, all images and patient data are saved in a format that is not recognized by other graphics or word-processing applications. The use of this software requires a password, and passwords are provided only to those providers involved in patient care.

In all, 429 patients aged 12 years or younger were referred for teledermatology consultation between January 1, 2002, and May 1, 2006. All patients with complete referral and/or consultation data were included. For patients who had more than one teledermatology consult for the same condition, only the initial consultation was included. The study was approved by the committee on human research at the University of California, San Francisco.

Data collection 

Provisional and consultant diagnoses and all medications previously used and subsequently recommended by the consultant were recorded. Age, sex, referral location, reason for consultation, and the chronicity of the skin problem were tabulated for each patient. Specific requests for follow-up teledermatology consultation or a live visit to a dermatologist were also recorded and tabulated.

Diagnostic concordance 

The provisional (referral) and consultant diagnoses were tabulated. Patients were excluded from this analysis if they lacked a provisional diagnosis (2.1%) Concordance between provisional and consultant diagnoses was categorized as follows:

•A – “Agree” included cases for which provisional and consultant diagnoses were identical or an acceptable nonspecific term was used to represent the dermatologic condition involved. For instance, the provisional and consultant diagnoses were considered to be in agreement when “eczema” referred to atopic dermatitis and “birthmark” referred to a melanocytic nevus, nevus anemicus, nevus depigmentosus, nevus spilus, nevus of Ito, or halo nevus.

•P – “Partial” included cases where at least one diagnosis was concordant, eg, either more than one consultation diagnosis was made, or the correct diagnosis was mentioned in the differential diagnosis.

•D – “Disagree” included cases where the provisional diagnosis was either incorrect or not recorded.

Management concordance 

Recommended management by the referring provider and consultant, including the number and names of medications, if any, were recorded. Management data were available for 383 patients. The agreement in management between the referring provider and teledermatologist was categorized as follows:

•A – “Agree” included cases where the medications used and recommended were identical in name, strength, and vehicle, or when observation was recommended by both providers (eg, benign birthmarks).

•P – “Partial” included cases where therapies were categorically similar but differed in name, strength, or vehicle (eg, if the referring provider recommended hydrocortisone 2.5% cream and the consultant recommended triamcinolone 0.1% ointment), or when therapies in addition to the earlier management were recommended.

•D – “Disagree” included cases where topical or systemic agents differing from those previously used were recommended (eg, a topical antifungal instead of a topical steroid preparation).

All calculations were performed using software (Excel, Microsoft, Redmond, WA) and all percentages were rounded to the nearest tenth.

Back to Article Outline

Results 

Patient demographics 

The mean age was 5.9 ± 3.9 years (SD); the age range was birth to 12 years, 11 months. Male [212 (49.4%)] and female [217 (50.6%)] patients were equally represented. Nearly two-thirds [272 (63.8%)] were referred to the teledermatologist for treatment, whereas one-third [141 (32.9%)] were referred for both diagnosis and treatment; for 13 (3%) patients, there was no expressed indication. In 6% of cases, the consultant recommended an in-person visit to a dermatologist; in 1.4% of cases, the consultant recommended a follow-up teledermatology evaluation.

Duration of disease before consultation 

The majority of patients (58.3%) were referred for teledermatology consultation within a year of onset of the skin condition and the mean duration of known disease was 4.8 ± 4.0 months (SD). More than a quarter (27.9%) of patients were referred more than 24 months after disease onset; the majority of these had conditions that were present at birth or shortly thereafter (eg, a nevus or hemangioma) (Fig 1).

Agreement of diagnosis and management 

Of consultant diagnoses, 48% were in agreement with the referring providers' diagnoses, whereas 10% were in partial agreement. There was disagreement in 42% of cases. Of those cases that were in agreement, the provisional and consultant diagnoses were identical in 62% of cases, and the provisional diagnoses were nonspecific in 38% of cases (Fig 2).

Of the 383 patients for whom we had treatment data, there was agreement in 28% of cases, partial agreement in 36%, and disagreement in 36%. Of note, in 76% of cases (82 of 108 patients) where there was agreement on treatment, no therapy was needed (eg, for an uncomplicated hemangioma or benign nevus). Among cases where there was partial agreement, the consultant recommended additional therapies in 29%. Of the partial agreement cases, 34% involved an alteration in the type, strength, or vehicle of topical steroid used. In the remainder of cases with partial agreement, one of the medications recommended by the referring provider was in agreement with the consultant's recommendations (eg, if the referring provider had treated the patient with several medications including a topical steroid and the consultant recommended the sole use of a topical steroid).

Diagnoses 

Table I compares teledermatology diagnoses to previously published pediatric outpatient series. The most common reason for referral was atopic dermatitis (14.2% of patients), including nummular eczema, followed by the evaluation of nevi (13.9%), which included congenital melanocytic nevus, mole, nevus of Ito, halo nevus, hairy nevus, and nevus spilus. The next most common conditions were verruca vulgaris (10.4%) and molluscum contagiosum (9.7%), respectively.

Table I. Pediatric dermatology diagnoses: relative frequencies in current study compared with published frequencies
This studyPublished results∗
DiagnosesRank%Rank%†
Atopic dermatitis (eczema)114.2115.7-22.3
Nevi‡213.9N/A
Verruca vulgaris310.424.5-13.3
Molluscum49.761.9-6.0
Pigment alteration (postinflammatory pigment alteration, vitiligo)58.4N/A
Dermatitis (contact, irritant, or unspecified)65.852.1-6.1
Hemangioma74.5N/A
Pityriasis alba83.6N/A
Alopecia areata92.9N/A
Insect bites101.9N/A
Psoriasis101.9N/A
ImpetigoN/A 37.8-8.0
Tinea capitisN/A 44.8-7.1
Seborrheic dermatitisN/A 73.4-4.0
ScabiesN/A 82.8-3.6
Pityriasis roseaN/A 92.4-3.6
Granuloma annulareN/A 103.6
Other dermatophytesN/A 103.6

N/A, Not available.

∗Acne was not included because of age range of study patients.

†Ranges14, 15, 16 ranked according to highest percentage found.

‡Different types of nevi include: congenital nevus, mole, nevus of Ito, halo nevus, hairy nevus, and nevus spilus.

Most commonly misdiagnosed skin conditions 

There was 100% discordance between the referring provider and the teledermatologist for the consultant diagnoses of tinea versicolor (5/5), seborrheic dermatitis (4/4), pityriasis rosea (3/3), xerosis (3/3), and lichen striatus (2/2). Other common pediatric dermatology conditions with diagnostic discordance in more than 50% of cases included consultant diagnoses of allergic contact, irritant contact, and unspecified dermatitis; insect bites; pityriasis alba; eczema (primarily nummular); pigment alteration (eg, postinflammatory pigment alteration); and psoriasis.

Consultant diagnoses when provisional diagnosis was “rash” or “dermatitis” 

Referring providers used the nonspecific terms “rash” and “dermatitis” for a number of distinct skin diseases. These terms were used approximately 30% of the time to refer to atopic dermatitis or nummular eczema and 14.0% of the time for allergic contact, irritant contact, and unspecified dermatitis. They were also used to refer to insect bites, psoriasis, folliculitis, seborrheic dermatitis, molluscum, and different types of nevi, among other skin conditions.

Most common other medications used by referring provider when teledermatologist suggested sole use of topical steroid 

Among patients (131) for whom the teledermatologist recommended a topical steroid preparation alone, almost 50% had not been previously treated with medication, and topical antifungals had been used in 23%. Of those who had received earlier therapy, 36% had not been treated with a topical steroid, but rather with topical and oral antifungal agents, topical calcineurin inhibitors, anti-itch therapy (eg, antihistamines), topical scabicides, and topical and oral antibiotics (Table II).

Table II. Most common other medications used by referring provider when teledermatologist suggested sole use of topical steroid
RankMedicationsNo.Proportion
1Topical antifungal15/6622.7%
2Topical calcineurin inhibitor7/6610.6%
3Anti-itch therapy6/669.1%
4Oral antifungal4/666.1%
4Topical scabicide4/666.1%
6Oral antibiotic3/664.5%
7Topical antibiotic2/663.0%
Various combinations of above medications25/6637.9%

Of the 66 patients for whom consultant suggested sole use of topical steroid, 24 were not previously treated with topical steroid and 42 had been treated with topical steroid in addition to one of the listed superfluous medications.

Diagnostic discordance in patients with earlier topical antifungal therapy 

Among patients who were treated with medications, 20.5% had been prescribed topical antifungal preparations by the referring provider. The most common consultant diagnoses for these patients were atopic dermatitis/eczema (5), psoriasis (3), contact dermatitis (2), and pityriasis alba (2).

Earlier use of systemic antibiotics 

Before referral, 21 patients were prescribed systemic antibiotics to treat their skin disease, cephalexin in 57.1% of cases and amoxicillin or amoxicillin/clavulanate in 19.0% of cases. Less commonly used antibiotics included oral trimethoprim/sulfamethoxazole, ciprofloxacin, erythromycin, and intramuscular ceftriaxone. Systemic antibiotics were used for a number of conditions where they might be indicated, including impetigo, folliculitis, and atopic dermatitis (where infection can flare disease) but in 33% of cases they were prescribed erroneously for contact dermatitis, postinflammatory pigment alteration, alopecia areata, candidiasis, insect bites, seborrheic dermatitis, and tinea corporis.

Back to Article Outline

Discussion 

This study of the store-and-forward teledermatology consultations of 429 pediatric patients underscores the feasibility and usefulness of teledermatology as a vehicle through which the primary care provider can gain timely access to needed specialty advice. The most common teledermatology diagnoses paralleled those seen in a general outpatient setting.14, 15, 16 A notable exception was disorders of pigmentation including melanocytic nevi, postinflammatory pigment alteration, and vitiligo, which were referred more frequently than might be expected in a general outpatient setting. This may reflect a lack of confidence among primary care providers in recognizing benign versus potentially consequential pigmented lesions and congenital or acquired alterations in skin pigmentation in children.

The diagnostic concordances (ie, between a nondermatologist and the consulting teledermatologist) are similar to those reported by others using store-and-forward teledermatology.17, 18 They are lower than the published concordance values obtained when comparing diagnoses between office dermatologists and teledermatologists (64%-88%) as specialists would be expected to make similar diagnoses more frequently than nonspecialists compared with specialists (the situation in this study).19, 20, 21, 22, 23 The use of nonspecific terms such as “rash” or “dermatitis” as provisional diagnoses by the referring nondermatologist providers was frequently applied across a spectrum of nosologically unrelated skin diseases including insect bites, folliculitis, molluscum, and nevi. This deficiency in diagnostic skill may be improved through use of store-and-forward teledermatology, because the technology allows for relatively rapid feedback to the referring provider, who can juxtapose images of a skin disease with the dermatologist's diagnosis analogous to a point-of-care dermatology atlas.

Our study found an even lower concordance in therapeutic approaches between the referring providers and the teledermatologist. Recommended management, whether because of additional medications or a completely new suggested regimen, differed from the referral protocol 72% of the time; this finding agrees with studies by Marcin et al17 and Bijl et al,18 who reported 79% and 57% management changes, respectively. The use of topical steroids in particular was an area of major discordance. Approximately one third (36%) of patients for whom the teledermatologist recommended the sole use of a topical steroid had not previously been treated with a topical steroid but rather with medications such as topical antifungals, topical and oral antibiotics, anti-itch therapy (ie, antihistamines), topical calcineurin inhibitors, and topical scabicides. Of patients for whom the sole use of a topical steroid was recommended, 64% had previously been treated with a topical steroid in conjunction with combinations of the aforementioned nonsteroidal agents. Previous reports have emphasized that dermatologists are more likely than pediatricians to use both topical steroids and higher potency steroids for skin conditions such as atopic dermatitis.16 Our data support the concept that “steroid phobia” (on the part of the referring provider, parents, or both) continues to be an issue in treating skin disease.16 The unnecessary use of topical antifungal agents and systemic antibiotics to treat a variety of dermatologic conditions of noninfectious origin is also noteworthy. In addition to providing specialty care, a store-and-forward dermatology consultation, by way of its iterative nature, could potentially modify current primary care services and reduce these management errors.

Limitations of the study include the inability to generalize our data because of the limited scope of the referring sites (nearly all of which were in the Central Valley of California), and the use of diagnoses and management recommendations by a single dermatologist as the gold standard for comparison. Moreover, as reported concordance rates of dermatologist diagnoses made in person versus by store-and-forward technology are 64% to 88%,19, 20, 21, 22, 23 the chance of error in the consultant diagnoses is also a limitation to our study. There is, however, evidence reporting similar outcomes in terms of improvement, no improvement, or worse outcome regardless of whether patients were treated via store-and-forward teledermatology or by conventional in-person dermatology visits.24 Lastly, we cannot rule out bias in the consultant's diagnoses, as the teledermatologist was aware of the provisional diagnoses of 405 patients. It is possible the consultant, even subconsciously, may be more (or less) likely to agree (or disagree) with the referring diagnoses.

Our study is one of the first to examine the use of pediatric teledermatology. Most problems were easily addressed through teledermatology consultation and in only 6.0% of patients was an in-person dermatology consultation deemed necessary by the consulting provider; in even fewer cases (1.4%) a specific recommendation for repeated teledermatology consultation was made. These results suggest that store-and-forward teledermatology can be effective both in providing appropriate care for skin disease and in reducing the need for referrals for live dermatology consultation.25 The implications of these observations for increasing specialty access, reducing the costs of medical care, and providing more timely consultation and intervention are worthy of continued exploration.

Back to Article Outline

References 

  1. Whited JD. Teledermatology research review. Int J Dermatol. 2006;45:220–229
  2. Telemedicine Information Exchange. Available from: http://tie.telemed.org/programs_t2/programs_t2.asp?ID=3&FIELDNAME=AppID. Accessed September 4, 2009.
  3. Resneck J, Kimball AB. The dermatology workforce shortage. J Am Acad Dermatol. 2004;50:50–54
  4. Burgiss SG, Julius CE, Watson HW, Haynes BK, Buonocore E, Smith GT. Telemedicine for dermatology care in rural patients. Telemed J. 1997;3:227–233
  5. Norton SA, Burdick AE, Phillips CM, Berman B. Teledermatology and underserved populations. Arch Dermatol. 1997;133:197–200
  6. Ramsey DL, Fox AB. The ability of primary care physicians to recognize common dermatoses. Arch Dermatol. 1981;117:620–622
  7. Federman D, Hogan D, Taylor JR, Caralis P, Kirsner RS. A comparison of diagnosis, evaluation, and treatment of patients with dermatologic disorders. J Am Acad Dermatol. 1995;32:726–729
  8. Cassileth BR, Clark WH, Lusk EJ, Frederick BE, Thompson CJ, Walsh WP. How well do physicians recognize melanoma and other problem lesions?. J Am Acad Dermatol. 1986;14:555–560
  9. Gerbert B, Maurer T, Berger T, Pantilat S, McPhee SJ, Wolff M, et al. Primary care physicians as gatekeepers in managed care–primary care physicians' and dermatologists' skills at secondary prevention of skin cancer. Arch Dermatol. 1996;132:1030–1038
  10. Burch JM, Krol A, Weston WL. Sarcoptes scabiei infestation misdiagnosed and treated as Langerhans cell histiocytosis. Pediatr Dermatol. 2004;21:58–62
  11. Nguyen CM, Burch JM, Fitzpatrick JE, Peterson SL, Weston WL. Giant cell fibroblastoma in a child misdiagnosed as a dermatofibroma. Pediatr Dermatol. 2002;19:28–32
  12. Romano C, Gianni C, Papini M. Tinea capitis in infants less than 1 year of age. Pediatr Dermatol. 2001;18:465–468
  13. Heffner VA, Lyon VB, Brousseau DC, Holland KE, Yen K. Store-and-forward teledermatology versus in-person visits: a comparison in pediatric teledermatology clinic. J Am Acad Dermatol. 2009;60:956–961
  14. Findlay GH, Vismer HF, Sophianos T. The spectrum of pediatric dermatology: analysis of 10,000 cases. Br J Dermatol. 1974;91:379–384
  15. Tunnessen WW. A survey of skin disorders seen in pediatric general and dermatology clinics. Pediatr Dermatol. 1984;1:219–222
  16. Schachner L, Ling NS, Press S. A statistical analysis of a pediatric dermatology clinic. Pediatr Dermatol. 1983;1:157–164
  17. Marcin JP, Nesbitt TS, Cole SL, et al. Changes in diagnosis, treatment, and clinical improvement among patients receiving telemedicine consultations. Telemed J E Health. 2005;11:36–43
  18. Bijl D, Van Sonderen E, Haaijer-Ruskamp FM. Prescription changes and drug costs at the interface between primary and specialist care. Eur J Clin Pharmacol. 1998;54:333–336
  19. Zelickson BD, Homan L. Teledermatology in the nursing home. Arch Dermatol. 1997;133:171–174
  20. Krupinski EA, LeSueur B, Ellsworth L, Levine N, Hansen R, Silvis N, et al. Diagnostic accuracy and image quality using a digital camera for teledermatology. Telemed J. 1999;5:257–263
  21. High WA, Houston MS, Calobrisi SD, Drage LA, McEvoy MT. Assessment of the accuracy of low-cost store-and-forward teledermatology consultation. J Am Acad Dermatol. 2000;42:776–783
  22. Barnard CM, Goldyne ME. Evaluation of an asynchronous teleconsultation system for diagnosis of skin cancer and other skin diseases. Telemed J E Health. 2000;6:379–384
  23. Pak HS, Harden D, Cruess D, Welch ML, Poropatich R. National Capital Area Teledermatology Consortium. Teledermatology: an intraobserver diagnostic correlation study, part 1. Cutis. 2003;71:399–403
  24. Pak H, Triplett CA, Lindquist JH, Grambow SC, Whited JD. Store-and-forward teledermatology results in similar clinical outcomes to conventional clinic-based care. J Telemed Telecare. 2007;213:26–30
  25. Knol A, van der Akker TW, Damstra RJ, de Haan J. Teledermatology reduces the number of patient referrals to a dermatologist. J Telemed Telecare. 2006;12:75–78

 Funding sources: None.

 Conflicts of interest: None declared.

PII: S0190-9622(09)00675-6

doi:10.1016/j.jaad.2009.05.039

Journal of the American Academy of Dermatology
Volume 62, Issue 1 , Pages 61-66, January 2010