- Email: [email protected]
Medical school dermatology curriculum: Are we adequately preparing primary care physicians?
Medical school dermatology curriculum: Are we adequately preparing primary care physicians? Nina K. Hansra, MD,a Patricia O’Sullivan, MS, EdD,b Cynthia L. Chen, MD,a and Timothy G. Berger, MDa San Francisco, California See related commentary on page 36 Background: There is a lack of information regarding the dermatology learning needs of primary care physicians and residents. Objective: To determine dermatologic topics that primary care physicians consider important and to determine primary care residents’ ratings of the teaching adequacy of these topics in the undergraduate medical curriculum. Methods: Primary care physicians and residents were surveyed regarding the importance and teaching adequacy of 17 dermatologic content areas. Results: Ninety-two primary care physicians identified 13 dermatologic content areas important for their practices. Two hundred fifty-two primary care residents identified 8 adequately taught topics and 9 inadequately taught topics. Limitations: Internal medicine and family medicine physicians and residents from only 10 regions were surveyed. Conclusion: Seventeen content areas can be divided into 3 categories: dermatologic topics that are important and adequately taught, topics that are unimportant, and a group of important, yet inadequately taught content areas. This latter group should be further integrated into dermatology curricula at U.S. medical schools. ( J Am Acad Dermatol 2008;61:23-9.)
INTRODUCTION A needs assessment can help identify discrepancies between current and ideal approaches in an undergraduate medical school curriculum,1 revealing the present state of the curriculum, inadequately taught knowledge areas, and stakeholders’ priorities in those areas.2,3 It can also help justify the need for dedicating limited From the Department of Dermatologya and Office of Medical Education,b University of California, San Francisco. Funding sources: None. Conflicts of interest: None declared. Reprints not available from authors. Correspondence to: Timothy G. Berger, MD, Department of Dermatology, University of California, San Francisco, Box 0316, 1701 Divisadero St, 340, San Francisco, CA 94143-0316. E-mail: [email protected]. Published online May 1, 2009. 0190-9622/$36.00 ª 2009 by the American Academy of Dermatology, Inc. doi:10.1016/j.jaad.2008.11.912
Abbreviations used: PCP: primary care physician UCSF: University of California, San Francisco
curricular time to a specific content topic.2 A significant number of US medical school graduates pursue careers in primary care specialties,4 and as outpatient internists and family physicians, they dedicate 4.1% and 6.2% of their respective visit time to dermatologic disease.5 No needs assessment has been conducted for a U.S. medical school dermatology curriculum. In Britain, however, researchers have employed a modified Delphi technique to determine dermatology learning objectives for medical students. Based on survey responses from a multidisciplinary panel of 66 physicians, researchers identified 53 (out of 145) learning objectives that should be included in United Kingdom’s undergraduate medical 23
24 Hansra et al
J AM ACAD DERMATOL JULY 2009
curricula.6 The American Association of Medical Colleges published recommendations for an undergraduate clinical skills curriculum in various disciplines.7 These recommendations were based on a review of recommendations of societies teaching in the core clinical areas, which did not include dermatology. A needs assessment involves establishing the general needs of the field and the learners’ needs.1 General needs of dermatologists are well recognized and include dermatologic emergencies, inflammatory diseases of the skin, skin cancer, acne, leg ulcers, cutaneous infections, cutaneous complications of HIV infection, cutaneous infections in immunocompromised individuals, urticaria, cutaneous manifestations of systemic disease, blisters in the skin, and drug eruptions.8 However, the undergraduate learners’ needs within dermatology have not yet been well established. We designed a needs assessment of primary care (internal medicine and family medicine) community physicians to determine the dermatologic needs in their practice. Similarly, we surveyed primary care residents to determine the adequacy of the undergraduate dermatology clinical curricula. Residents can address their recent undergraduate preparation based on their primary care training experiences. The purpose of this study was to examine the adequacy of undergraduate dermatology clinical education in the context of topics that primary care physicians consider important in practice.
METHODS Between 2005 and 2007, we conducted a clinical dermatology needs assessment by surveying community-based primary care (internal medicine and family medicine) physicians about dermatologic topics important in their practice. Residents from internal medicine, family medicine, pediatrics, and emergency medicine responded to questions about the adequacy of their medical school education in dermatology. For the purposes of this paper, we included data from the internal medicine and family medicine resident subsets only, as we sought to compare them with responses from internal medicine and family medicine community physicians. Participants Community-based primary care physicians (PCPs) practicing in Fresno and San Francisco were surveyed. The two regions were selected as a result of their disparate characteristics: Fresno, an agricultural area in California’s Central Valley with a sunny climate and very few dermatologists, is more medically underserved compared to San Francisco.9
Eligibility criteria for community PCPs included those who self-identified as primary care internists or family medicine physicians in either the San Francisco or Fresno, California Yellow Pages. Approval from the University of California, San Francisco (UCSF) Committee for Human Research was obtained. Resident participants were solicited from all years of training beginning in August 2005. These residents came from UCSF and UCSF Fresno, Brooke Army Medical Center, Harvard University, Northeastern Ohio Universities College of Medicine, University of Alabama, Texas Tech University Medical Center, and University of Massachusetts. All internal medicine and family medicine residents training at participating institutions were eligible. The UCSF and UCSF Fresno, Brooke Army Medical Center, Harvard University, Northeastern Ohio Universities College of Medicine, University of Alabama, Texas Tech University Medical Center, and University of Massachusetts institutional review boards approved this study. Survey instruments We designed a written survey to identify dermatology education needs of primary care physicians. Dermatologic content areas were chosen after a review of teaching recommendations from the British Association of University Teachers of Dermatology,8 two previously reported dermatology undergraduate education studies,6,10 and a standard text used by many U.S. medical students.11 We categorized content areas according to standard dermatology texts used by dermatologists12-15 in order to ensure content validity. Community physicians and residents responded to the same 17 dermatologic content items; however, the question stem differed for the two groups. We asked community physicians: ‘‘How important are the following knowledge areas for your current practice using a three-point scale (not too important, somewhat important, or very important)?’’ (available at www.eblue.org as an online-only Appendix) We asked residents: ‘‘In preparation for residency, was your medical school training in the following areas inadequate, appropriate, or excessive?’’ We asked residents to compare their dermatology knowledge to asthma and diabetes, two commonly taught subjects in medical school curricula. Residents responded to a single question asking them to rate their preparedness to diagnose and treat ‘‘common dermatologic diseases’’ on a 5-point Likert scale (strongly disagree [1] to strongly agree [5]). We also asked both groups for demographic information: PCPs indicated their affiliation (internal medicine or
Hansra et al 25
J AM ACAD DERMATOL VOLUME 61, NUMBER 1
family medicine), geographic location (San Francisco or Fresno), and dermatologic procedures they performed. Residents gave their departmental affiliation (internal medicine or family medicine), year of training, and indicated whether or not they had completed an undergraduate dermatology rotation. Survey distribution Community physicians were recruited by telephone calls to their offices. Surveys were faxed to their offices, and physicians were asked to return responses by fax or postal mail. Surveys were deidentified. Internal medicine and family medicine residents were recruited during didactic sessions (ie, morning report, noon conferences). Residents surveyed at each particular session represented those who were assigned to duties at that site during the months of surveying and were present on the day of surveying. If they could not complete the survey during the meeting, they were asked to fill out an electronic or Web-based survey. Paper-and-pencil surveys did not request any identifying information, and electronic surveys were de-identified. Data analysis Returned surveys were tabulated. Data were entered by local site coordinators into a secure, central UCSF electronic database managed with Microsoft Access 2000 and Microsoft Excel 2000. Missing data that were ‘‘not applicable’’ were coded as such; unexplained missing data were left blank. We coded the community PCP ratings of the importance of 17 dermatologic content areas. In coding responses, we gave a response of ‘‘not too important’’ or ‘‘somewhat important’’ a value of 0, and a response of ‘‘very important’’ a value of 1. We coded resident responses to the teaching adequacy of the 17 dermatologic content areas similarly. We gave a response of ‘‘inadequate’’ a value of 0, and a response of ‘‘appropriate’’ or ‘‘excessive’’ a value of 1. Chi-square analyses were used to determine if there were significant differences between San Francisco PCPs and Fresno PCPs in dermatologic content and procedural areas. Chi-square analyses were also used to determine if there were significant differences between residents who completed a dermatology rotation as a medical student and those who did not. Those dermatologic content areas rated as ‘‘very important’’ by more than 50% of PCPs were considered very important to include in our proposed dermatology curriculum. Those content areas that more than 30% of primary care residents considered
‘‘inadequately taught’’ were deemed inadequately taught for the purposes of this manuscript.
RESULTS Respondent characteristics Ninety-five community internal medicine (n = 60) and family medicine (n = 35) physicians in San Francisco and Fresno completed the handwritten survey. Of the 73 San Francisco PCP offices contacted, 50 PCPs from 35 offices completed at least one part of the survey. Of the 53 Fresno PCP offices contacted, 47 PCPs from 26 offices completed at least one part of the survey. Two hundred fifty-two residents (internal medicine [n = 150] and family medicine [n = 102]) from over 75 different medical schools were surveyed from UCSF (29%, 57/200), UCSF Fresno (75%, 61/81), Brooke Army Medical Center (54%, 40/74), Harvard University (16%, 28/172), Northeastern Ohio Universities College of Medicine (10%, 17/171), University of Alabama (3%, 4/157), Texas Tech University Medical Center (33%, 19/57), and University of Massachusetts (21%, 27/131). Dermatologic topics important in primary care practice Forty-five San Francisco PCPs and 47 Fresno PCPs responded to the dermatologic content segment of the questionnaire. The percentage of PCPs who rated these content areas as very important are listed in Table I. The comparison between those physicians practicing in San Francisco and those in Fresno showed significant differences (P \.05) for 5 of the 17 topics (see Table I). Melanoma/moles and nonmelanoma skin cancer/sun damage were rated very important by Fresno PCPs. Herpes simplex/zoster, alopecia, and HIV dermatology were rated very important by San Francisco PCPs. Forty-five Fresno PCPs and 50 San Francisco PCPs responded to questions about performing dermatologic procedures in their practice. Eighty-seven percent of Fresno PCPs performed cryotherapy compared to 64% of their San Francisco counterparts (P = .011). Indications for performing cryotherapy for Fresno and San Francisco physicians included warts (100% and 100%, respectively), actinic keratoses (90% and 72%, respectively), and benign skin lesions (82% and 44%, respectively). Similarly, 87% of Fresno PCPs performed skin biopsies (ie, punch and/or shave), whereas only 20% of San Francisco PCPs reported doing these procedures (P \.0001). A higher percentage of PCPs in both Fresno and San Francisco performed skin biopsies to rule out cancer (100% and 80%, respectively) than to evaluate rashes (77% and 50%, respectively).
26 Hansra et al
J AM ACAD DERMATOL JULY 2009
Table I. Percentage of PCPs who rated dermatologic topics as ‘‘very important’’ to their practice
Dermatologic topic
Melanoma/moles* Atopic dermatitis/ contact dermatitis Non-melanoma skin cancer/sun damage* Bacterial fungal skin infections, including in immunocompromised patients Herpes simplex/zoster* Leg ulcers/wound care Urticaria/hives Cutaneous drug eruptions Psoriasis/seborrhea Acne/rosacea Warts Infestations (ie, scabies, lice) Viral exanthems Vasculitis/purpura Connective tissue disease (ie, lupus) Alopecia* HIV dermatology*
% All PCPs % Fresno % SF P (n = 92) (n = 47) (n = 45) Value
87 80
96 81
78 80
.01 .92
78
87
69
.03
76
77
76
.91
74 70 68 68
64 79 66 64
84 60 71 73
.02 .05 .60 .33
65 62 58 55
68 57 51 55
62 67 64 62
.56 .36 .19 .50
54 45 40
53 45 42.6
56 44 38
.82 .98 .64
38 17
28 6
49 29
.04 .004
PCPs, Primary care physicians; SF, San Francisco. *Topics that were rated significantly different (ie, P \ .05) between Fresno PCPs and SF PCPs.
Adequacy of teaching of dermatologic topics Two hundred fifty-two residents responded to the dermatologic content items. Table II presents the percentage of residents who rated these content areas as inadequately taught. Asthma and diabetes were considered better taught than the dermatologic content areas. Two hundred thirty-nine residents indicated whether or not they completed an undergraduate clinical dermatology rotation, which ranged from 2 to 4 weeks in length. The comparison between those residents who completed a dermatology clinical rotation and those who did not revealed significant differences for 7 of the 17 content areas (see Table II). Residents who completed a rotation felt significantly more prepared in alopecia, cutaneous drug eruptions, warts, non-melanoma skin cancer/sun damage, acne/rosacea, psoriasis/ seborrhea, and atopic dermatitis/contact dermatitis. Overall, 37% of primary care residents rated their medical school curriculum as having adequately prepared them to diagnose ‘‘common dermatologic
diseases.’’ Twenty-eight percent of primary care residents rated their medical school curriculum as having adequately prepared them to treat ‘‘common dermatologic diseases.’’ Fifty-one percent of primary care residents who completed a dermatology rotation rated their medical school curriculum as having adequately prepared them to diagnose common dermatologic diseases compared to 25% of those who did not complete a rotation (P \.001). Forty-two percent of primary care residents who completed a rotation rated their medical school curriculum as having adequately prepared them to treat common dermatologic diseases compared to 16% of those who did not complete a rotation (P \.001).
DISCUSSION This needs assessment of dermatology curricula in U.S. medical schools yields several important conclusions. According to primary care residents, dermatology is not as adequately taught as curricular areas such as asthma and diabetes. Less than 40% of primary care residents feel their medical school curriculum adequately prepared them to diagnose and treat common skin disorders. This may reflect the amount of curricular time devoted to dermatology in comparison to other content areas. There are several important content areas of dermatology that residents reported as adequately taught. These include melanoma/moles, atopic dermatitis/contact dermatitis, non-melanoma skin cancer/sun damage, herpes simplex/zoster, urticaria/hives, psoriasis/seborrhea, acne/rosacea, and warts (Fig 1, green area). There were numerous topics that PCPs identified as very important in their practices, but which residents deemed inadequately taught in their medical school dermatology curricula. These include skin infections, leg ulcers/wound care, cutaneous drug eruptions, infestations, and viral exanthems (Fig 1, red area). These topics should be prioritized for improvement in dermatology curricula in medical schools. These may represent areas where dermatology and other specialties might both be teaching parts of the content, but not in a coordinated manner. For example, skin infections may be included in microbiology, infectious disease, and dermatology curricula. Likewise, cutaneous drug eruptions may be considered in both pharmacology and dermatology. The departments that share responsibility in teaching these content areas would ideally coordinate their teaching, perhaps by jointly developing learning objectives and dedicating teaching time in both their curricula to adequately cover the topic. Joint teaching time during which a dermatologist shares a lecture with a microbiologist or pharmacologist might be an ideal way to
Hansra et al 27
J AM ACAD DERMATOL VOLUME 61, NUMBER 1
Table II. Percentage of residents who rated dermatologic topics as ‘‘inadequately’’ taught in medical school % Residents reporting inadequately taught Dermatologic topic
Asthma Diabetes HIV dermatology Alopecia* Cutaneous drug eruptions* Vasculitis/purpura Leg ulcers/wound care Viral exanthems Connective tissue disease (ie, lupus) Infestations (ie, scabies, lice) Bacterial/fungal skin infections, including in immunocompromised patients Warts* Non-melanoma skin cancer/sun damage* Acne/rosacea* Psoriasis/seborrhea* Melanoma/moles Atopic dermatitis/contact dermatitis* Urticaria/hives Herpes simplex/zoster
All residents
Residents who completed rotation (n = 118)
Residents who did not take rotation (n = 121)
P Value
6 7 68 60 46 42 38 36 36 35 33
7 7 68 48 36 44 36 31 33 30 25
7 8 69 72 53 46 45 40 36 40 40
.96 .66 .90 \.001 .007 .73 .15 .18 .59 .10 .29
29 26 26 25 21 18 18 13
18 19 15 13 15 9 18 8
38 34 37 31 25 21 21 13
.001 .008 \.001 .001 .066 .014 .57 .16
*Topics that were rated significantly different (ie, P \ .05) between residents who completed a rotation and residents who did not complete a rotation.
cover the topic adequately, especially given that clinicopathologic correlation and case-based learning improve knowledge retention in medical school. Content areas judged to be less important by resident physicians can be added to curricula if time permits. These include vasculitis/purpura, skin signs of connective tissue disease, alopecia, and HIV dermatology (Fig 1, yellow area). These areas are less frequently encountered in outpatient primary care practice. Some of these areas would be of greater importance for physicians practicing in certain venues. Hospitalists, for example, should be well aware of the evaluation of vasculitis and purpura, and rheumatologists should be well versed in the skin signs of connective tissue disease. These topics may be best reserved for teaching at the postgraduate level. As expected, when topics were considered unimportant, they were never rated as inadequately taught. Ratings of teaching adequacy of content areas were influenced by whether the residents completed an undergraduate dermatology clinical rotation. Residents who completed a dermatology elective felt significantly more prepared to diagnose and treat common skin disease compared to residents who did not complete a rotation. Skin infection teaching was judged to be inferior for more residents who did not take a clinical dermatology rotation (40% felt that it
was inadequately taught) when compared to those who completed a dermatology rotation (25% felt that it was inadequately taught) (Fig 1, dashed arrow). While non-melanoma skin cancer/sun damage, psoriasis/seborrhea, acne/rosacea, and warts were reported as adequately taught, these topics were inadequately taught for those who did not complete a dermatology rotation (Fig 1, solid arrows). If nonmelanoma skin cancer/sun damage, psoriasis/seborrhea, acne/rosacea, and warts are not addressed appropriately in the core curriculum, then it would be important to require, or at least highly encourage, medical students who anticipate a career in primary care to complete a clinical rotation in dermatology. San Francisco and Fresno PCPs value similar dermatologic content areas with certain exceptions. We recommend tailoring the dermatology curriculum at the postgraduate level to the needs of geographic regions in which trainees intend to practice. San Francisco PCPs place greater importance on herpes simplex/zoster, alopecia, and HIV dermatology than Fresno PCPs. Since San Francisco has a large immunocompromised population, including many HIV-infected individuals,16 this emphasis is not surprising. Fresno PCPs place greater value on melanoma/moles and non-melanoma skin cancer/sun damage. This is also not surprising given
28 Hansra et al
J AM ACAD DERMATOL JULY 2009
Fig 1. Dermatologic content areas in order of level of importance (determined by PCPs) and separated by teaching adequacy (determined by residents).
that Fresno sees higher rates of sun exposure compared to San Francisco. Most schools teach students about dermatologic procedures, but do not expect students to learn how to perform them.10 Since many PCPs perform cryotherapy and skin biopsies, these procedures should become a part of the medical school or postgraduate dermatology curriculum. As more than 60% of PCPs perform cryotherapy, it should ideally be taught in medical school or, at the very least, in all primary care residency training programs. The Association of American Medical Colleges has similarly recommended that these procedures be taught in medical school.7 Procedural training is especially important for residents planning to practice in medically and dermatologically underserved areas. Eighty-seven percent of PCPs in Fresno perform skin biopsies compared to 20% of PCPs in San Francisco. PCPs in underserved areas not only need to know how to perform the procedures correctly, but also to whom to send the tissue for evaluation and how to interpret biopsy results. Clayton, Perera, and Burge6 identified dermatologic content areas for inclusion in medical school curricula in the United Kingdom, and these overlap significantly with important topics identified by our study. One notable exception is purpura, which Clayton et al identified as important in medical school training, whereas our study did not. An important difference between our study and their
recommendations is that the British group surveyed physicians across several disciplines, especially dermatologists, and our study population consisted of community PCPs. A recent multicenter survey of the dermatology faculty responsible for teaching students at U.S. medical schools found that acne, seborrheic dermatitis, and warts were the 3 most frequently taught content areas.10 Again, these recommendations may not represent the actual needs of PCPs in practice. We argue that the most important stakeholders in determining an undergraduate dermatology medical school curriculum are PCPs, who are the first line in the diagnosis and treatment of skin conditions. In the current era of a critical shortage of dermatologists,17 teaching future PCPs when and how to appropriately refer is essential and should become a part of the medical school dermatology curriculum.6 This study has significant limitations. First, we only conducted a needs assessment of clinical dermatologic topics. We did not address the teaching of the basic structure and function of the skin, description of skin lesions, and performing a complete skin examination. We believed these topics were basic areas of knowledge that should always be included in the medical school curriculum. Second, we also only surveyed internal medicine and family medicine physicians from a few geographic regions. We felt internal medicine and family medicine
J AM ACAD DERMATOL
Hansra et al 29
VOLUME 61, NUMBER 1
physicians were the most important stakeholders in an undergraduate dermatology medical curriculum. The proposed curriculum might need to be enhanced or modified for physicians practicing in other specialties. Another limitation was the arbitrary cutoff values established for categorizing topics. These were chosen with the view that if half of PCPs thought a content area was important, it should be taught to all medical students. Similarly, if approximately one third of residents thought a content area was inadequately taught, the content area should be included and further emphasized in the undergraduate medical school curriculum. We support a core undergraduate clinical dermatology curriculum. This proposed curriculum would ensure that medical students receive exposure to the ‘‘very important’’ areas of clinical dermatology. Since PCPs perform dermatologic procedures in their offices, these skills should be introduced in medical school. Further, we recommend that certain dermatologic content areas be included or strengthened in current medical school dermatology curricula, others continued to be taught at their currently effective level, and some minimized or removed from the curriculum to allow for the addition of the currently under-taught and very important content areas (see Fig 1). The proposed dermatology curriculum might require slight modifications based on the needs of the communities in which residents intend to practice. Future smaller scale needs assessments may be warranted to help determine these regional needs. We acknowledge the contributions of the local site coordinators in obtaining institutional review board approval, distributing surveys, and logging data from each of their respective sites: Arin Isenstein, MD, at UCSF Fresno; Capt Patrick McCleskey, MC, USAF, at Brooke Army Medical Center; Stephanie Hu, MD, at Harvard Medical School, Boston, MA; Matthew Molenda, MD, at Northeastern Ohio Universities College of Medicine; Jamaiya Havel, MD, at Texas Tech University Medical Center; and Megan Seibert, MD, at University of Alabama School of Medicine. We thank Arianne Teherani, PhD, for providing her guidance in the development of our needs assessment survey. We thank Priya Rajendran, MD, for aiding in the conception of a needs assessment and piloting the first survey. We thank Jeanette Waller, MD, for helping to secure funding and recruiting local site
coordinators for the multicenter study. We thank Melissa Kanchanapoomi, BA, for aiding with data analysis. We thank Lee Zane, MD, MAS, for his vision and constructive criticism of our study. REFERENCES 1. Kern D, Thomas P, Howard D, Bass E. Curriculum Development for Medical Education: A Six Step Approach. Baltimore: Johns Hopkins University Press; 1998. 2. Laidlaw T, MacLeod H, Kaufman D, Langille D, Sargeant J. Implementing a communication skills programme in medical school: needs assessment and programme change. Med Educ 2002;36:115-24. 3. Witkin B. Assessing needs in educational and social programs: Using information to make decisions, set priorities, and allocate resources. San Francisco: Jossey-Bass; 1984. 4. Jeffe D, Andriole D, Hageman H, Whelan A. The changing paradigm of contemporary U.S. allopathic medical school graduates’ career paths: analysis of the 1997-2004 national AAMC Graduation Questionnaire database. Acad Med 2007;82: 888-94. 5. Fleischer A, Herbert C, Feldman S, O’Brien F. Diagnosis of skin disease by nondermatologists. Am J Manag Care 2000; 6:1149-56. 6. Clayton R, Perera R, Burge S. Defining the dermatological content of the undergraduate medical curriculum: a modified Delphi study. Br J Dermatol 2006;155:137-44. 7. The AAMC project on the clinical education of medical students: clinical skills education. 2005. 8. Davies E, Burge S. Audit of dermatological content of U.K. undergraduate curricula. Br J Dermatol 2009;160:999-1005. 9. Medically underserved areas/medically underserved populations: US Department of Health and Human Services: Health Resources and Services Administration; 2008. 10. McClesky P, Gilson R, Devillez R. Medical student core curriculum in dermatology. Chicago: Dermatology Teachers Exchange Group; 2007. 11. Marks J, Miller J. Lookingbill and Marks’ Principles of dermatology. Philadelphia: Saunders; 2006. 12. Bolognia J, Jorizzo J, Rapini R. Dermatology. St Louis: Mosby; 2007. 13. James W, Berger T, Elston D. Andrews’ Diseases of the skin: Clinical dermatology. Philadelphia: Saunders; 2005. 14. Wolff K, Goldsmith L, Katz S, Gilchrest B, Paller A, Leffell D. Fitzpatrick’s Dermatology in general medicine. McGraw-Hill Professional; 2007. 15. Burns D, Breathnach S, Cox N, Griffiths C, editors. Rook’s textbook of dermatology. Hoboken, NJ: Wiley-Blackwell; 2004. 16. Holmberg S. The estimated prevalence and incidence of HIV in 96 large US metropolitan areas. Am J Public Health 1996;86: 642-54. 17. Resneck J, Kimball A. The dermatology workforce shortage. J Am Acad Dermatol 2004;50:50-4.
29.e1 Hansra et al
J AM ACAD DERMATOL JULY 2009
APPENDIX. Survey for community-based primary care physicians
INTRODUCTION A needs assessment can help identify discrepancies between current and ideal approaches in an undergraduate medical school curriculum,1 revealing the present state of the curriculum, inadequately taught knowledge areas, and stakeholders’ priorities in those areas.2,3 It can also help justify the need for dedicating limited From the Department of Dermatologya and Office of Medical Education,b University of California, San Francisco. Funding sources: None. Conflicts of interest: None declared. Reprints not available from authors. Correspondence to: Timothy G. Berger, MD, Department of Dermatology, University of California, San Francisco, Box 0316, 1701 Divisadero St, 340, San Francisco, CA 94143-0316. E-mail: [email protected]. Published online May 1, 2009. 0190-9622/$36.00 ª 2009 by the American Academy of Dermatology, Inc. doi:10.1016/j.jaad.2008.11.912
Abbreviations used: PCP: primary care physician UCSF: University of California, San Francisco
curricular time to a specific content topic.2 A significant number of US medical school graduates pursue careers in primary care specialties,4 and as outpatient internists and family physicians, they dedicate 4.1% and 6.2% of their respective visit time to dermatologic disease.5 No needs assessment has been conducted for a U.S. medical school dermatology curriculum. In Britain, however, researchers have employed a modified Delphi technique to determine dermatology learning objectives for medical students. Based on survey responses from a multidisciplinary panel of 66 physicians, researchers identified 53 (out of 145) learning objectives that should be included in United Kingdom’s undergraduate medical 23
24 Hansra et al
J AM ACAD DERMATOL JULY 2009
curricula.6 The American Association of Medical Colleges published recommendations for an undergraduate clinical skills curriculum in various disciplines.7 These recommendations were based on a review of recommendations of societies teaching in the core clinical areas, which did not include dermatology. A needs assessment involves establishing the general needs of the field and the learners’ needs.1 General needs of dermatologists are well recognized and include dermatologic emergencies, inflammatory diseases of the skin, skin cancer, acne, leg ulcers, cutaneous infections, cutaneous complications of HIV infection, cutaneous infections in immunocompromised individuals, urticaria, cutaneous manifestations of systemic disease, blisters in the skin, and drug eruptions.8 However, the undergraduate learners’ needs within dermatology have not yet been well established. We designed a needs assessment of primary care (internal medicine and family medicine) community physicians to determine the dermatologic needs in their practice. Similarly, we surveyed primary care residents to determine the adequacy of the undergraduate dermatology clinical curricula. Residents can address their recent undergraduate preparation based on their primary care training experiences. The purpose of this study was to examine the adequacy of undergraduate dermatology clinical education in the context of topics that primary care physicians consider important in practice.
METHODS Between 2005 and 2007, we conducted a clinical dermatology needs assessment by surveying community-based primary care (internal medicine and family medicine) physicians about dermatologic topics important in their practice. Residents from internal medicine, family medicine, pediatrics, and emergency medicine responded to questions about the adequacy of their medical school education in dermatology. For the purposes of this paper, we included data from the internal medicine and family medicine resident subsets only, as we sought to compare them with responses from internal medicine and family medicine community physicians. Participants Community-based primary care physicians (PCPs) practicing in Fresno and San Francisco were surveyed. The two regions were selected as a result of their disparate characteristics: Fresno, an agricultural area in California’s Central Valley with a sunny climate and very few dermatologists, is more medically underserved compared to San Francisco.9
Eligibility criteria for community PCPs included those who self-identified as primary care internists or family medicine physicians in either the San Francisco or Fresno, California Yellow Pages. Approval from the University of California, San Francisco (UCSF) Committee for Human Research was obtained. Resident participants were solicited from all years of training beginning in August 2005. These residents came from UCSF and UCSF Fresno, Brooke Army Medical Center, Harvard University, Northeastern Ohio Universities College of Medicine, University of Alabama, Texas Tech University Medical Center, and University of Massachusetts. All internal medicine and family medicine residents training at participating institutions were eligible. The UCSF and UCSF Fresno, Brooke Army Medical Center, Harvard University, Northeastern Ohio Universities College of Medicine, University of Alabama, Texas Tech University Medical Center, and University of Massachusetts institutional review boards approved this study. Survey instruments We designed a written survey to identify dermatology education needs of primary care physicians. Dermatologic content areas were chosen after a review of teaching recommendations from the British Association of University Teachers of Dermatology,8 two previously reported dermatology undergraduate education studies,6,10 and a standard text used by many U.S. medical students.11 We categorized content areas according to standard dermatology texts used by dermatologists12-15 in order to ensure content validity. Community physicians and residents responded to the same 17 dermatologic content items; however, the question stem differed for the two groups. We asked community physicians: ‘‘How important are the following knowledge areas for your current practice using a three-point scale (not too important, somewhat important, or very important)?’’ (available at www.eblue.org as an online-only Appendix) We asked residents: ‘‘In preparation for residency, was your medical school training in the following areas inadequate, appropriate, or excessive?’’ We asked residents to compare their dermatology knowledge to asthma and diabetes, two commonly taught subjects in medical school curricula. Residents responded to a single question asking them to rate their preparedness to diagnose and treat ‘‘common dermatologic diseases’’ on a 5-point Likert scale (strongly disagree [1] to strongly agree [5]). We also asked both groups for demographic information: PCPs indicated their affiliation (internal medicine or
Hansra et al 25
J AM ACAD DERMATOL VOLUME 61, NUMBER 1
family medicine), geographic location (San Francisco or Fresno), and dermatologic procedures they performed. Residents gave their departmental affiliation (internal medicine or family medicine), year of training, and indicated whether or not they had completed an undergraduate dermatology rotation. Survey distribution Community physicians were recruited by telephone calls to their offices. Surveys were faxed to their offices, and physicians were asked to return responses by fax or postal mail. Surveys were deidentified. Internal medicine and family medicine residents were recruited during didactic sessions (ie, morning report, noon conferences). Residents surveyed at each particular session represented those who were assigned to duties at that site during the months of surveying and were present on the day of surveying. If they could not complete the survey during the meeting, they were asked to fill out an electronic or Web-based survey. Paper-and-pencil surveys did not request any identifying information, and electronic surveys were de-identified. Data analysis Returned surveys were tabulated. Data were entered by local site coordinators into a secure, central UCSF electronic database managed with Microsoft Access 2000 and Microsoft Excel 2000. Missing data that were ‘‘not applicable’’ were coded as such; unexplained missing data were left blank. We coded the community PCP ratings of the importance of 17 dermatologic content areas. In coding responses, we gave a response of ‘‘not too important’’ or ‘‘somewhat important’’ a value of 0, and a response of ‘‘very important’’ a value of 1. We coded resident responses to the teaching adequacy of the 17 dermatologic content areas similarly. We gave a response of ‘‘inadequate’’ a value of 0, and a response of ‘‘appropriate’’ or ‘‘excessive’’ a value of 1. Chi-square analyses were used to determine if there were significant differences between San Francisco PCPs and Fresno PCPs in dermatologic content and procedural areas. Chi-square analyses were also used to determine if there were significant differences between residents who completed a dermatology rotation as a medical student and those who did not. Those dermatologic content areas rated as ‘‘very important’’ by more than 50% of PCPs were considered very important to include in our proposed dermatology curriculum. Those content areas that more than 30% of primary care residents considered
‘‘inadequately taught’’ were deemed inadequately taught for the purposes of this manuscript.
RESULTS Respondent characteristics Ninety-five community internal medicine (n = 60) and family medicine (n = 35) physicians in San Francisco and Fresno completed the handwritten survey. Of the 73 San Francisco PCP offices contacted, 50 PCPs from 35 offices completed at least one part of the survey. Of the 53 Fresno PCP offices contacted, 47 PCPs from 26 offices completed at least one part of the survey. Two hundred fifty-two residents (internal medicine [n = 150] and family medicine [n = 102]) from over 75 different medical schools were surveyed from UCSF (29%, 57/200), UCSF Fresno (75%, 61/81), Brooke Army Medical Center (54%, 40/74), Harvard University (16%, 28/172), Northeastern Ohio Universities College of Medicine (10%, 17/171), University of Alabama (3%, 4/157), Texas Tech University Medical Center (33%, 19/57), and University of Massachusetts (21%, 27/131). Dermatologic topics important in primary care practice Forty-five San Francisco PCPs and 47 Fresno PCPs responded to the dermatologic content segment of the questionnaire. The percentage of PCPs who rated these content areas as very important are listed in Table I. The comparison between those physicians practicing in San Francisco and those in Fresno showed significant differences (P \.05) for 5 of the 17 topics (see Table I). Melanoma/moles and nonmelanoma skin cancer/sun damage were rated very important by Fresno PCPs. Herpes simplex/zoster, alopecia, and HIV dermatology were rated very important by San Francisco PCPs. Forty-five Fresno PCPs and 50 San Francisco PCPs responded to questions about performing dermatologic procedures in their practice. Eighty-seven percent of Fresno PCPs performed cryotherapy compared to 64% of their San Francisco counterparts (P = .011). Indications for performing cryotherapy for Fresno and San Francisco physicians included warts (100% and 100%, respectively), actinic keratoses (90% and 72%, respectively), and benign skin lesions (82% and 44%, respectively). Similarly, 87% of Fresno PCPs performed skin biopsies (ie, punch and/or shave), whereas only 20% of San Francisco PCPs reported doing these procedures (P \.0001). A higher percentage of PCPs in both Fresno and San Francisco performed skin biopsies to rule out cancer (100% and 80%, respectively) than to evaluate rashes (77% and 50%, respectively).
26 Hansra et al
J AM ACAD DERMATOL JULY 2009
Table I. Percentage of PCPs who rated dermatologic topics as ‘‘very important’’ to their practice
Dermatologic topic
Melanoma/moles* Atopic dermatitis/ contact dermatitis Non-melanoma skin cancer/sun damage* Bacterial fungal skin infections, including in immunocompromised patients Herpes simplex/zoster* Leg ulcers/wound care Urticaria/hives Cutaneous drug eruptions Psoriasis/seborrhea Acne/rosacea Warts Infestations (ie, scabies, lice) Viral exanthems Vasculitis/purpura Connective tissue disease (ie, lupus) Alopecia* HIV dermatology*
% All PCPs % Fresno % SF P (n = 92) (n = 47) (n = 45) Value
87 80
96 81
78 80
.01 .92
78
87
69
.03
76
77
76
.91
74 70 68 68
64 79 66 64
84 60 71 73
.02 .05 .60 .33
65 62 58 55
68 57 51 55
62 67 64 62
.56 .36 .19 .50
54 45 40
53 45 42.6
56 44 38
.82 .98 .64
38 17
28 6
49 29
.04 .004
PCPs, Primary care physicians; SF, San Francisco. *Topics that were rated significantly different (ie, P \ .05) between Fresno PCPs and SF PCPs.
Adequacy of teaching of dermatologic topics Two hundred fifty-two residents responded to the dermatologic content items. Table II presents the percentage of residents who rated these content areas as inadequately taught. Asthma and diabetes were considered better taught than the dermatologic content areas. Two hundred thirty-nine residents indicated whether or not they completed an undergraduate clinical dermatology rotation, which ranged from 2 to 4 weeks in length. The comparison between those residents who completed a dermatology clinical rotation and those who did not revealed significant differences for 7 of the 17 content areas (see Table II). Residents who completed a rotation felt significantly more prepared in alopecia, cutaneous drug eruptions, warts, non-melanoma skin cancer/sun damage, acne/rosacea, psoriasis/ seborrhea, and atopic dermatitis/contact dermatitis. Overall, 37% of primary care residents rated their medical school curriculum as having adequately prepared them to diagnose ‘‘common dermatologic
diseases.’’ Twenty-eight percent of primary care residents rated their medical school curriculum as having adequately prepared them to treat ‘‘common dermatologic diseases.’’ Fifty-one percent of primary care residents who completed a dermatology rotation rated their medical school curriculum as having adequately prepared them to diagnose common dermatologic diseases compared to 25% of those who did not complete a rotation (P \.001). Forty-two percent of primary care residents who completed a rotation rated their medical school curriculum as having adequately prepared them to treat common dermatologic diseases compared to 16% of those who did not complete a rotation (P \.001).
DISCUSSION This needs assessment of dermatology curricula in U.S. medical schools yields several important conclusions. According to primary care residents, dermatology is not as adequately taught as curricular areas such as asthma and diabetes. Less than 40% of primary care residents feel their medical school curriculum adequately prepared them to diagnose and treat common skin disorders. This may reflect the amount of curricular time devoted to dermatology in comparison to other content areas. There are several important content areas of dermatology that residents reported as adequately taught. These include melanoma/moles, atopic dermatitis/contact dermatitis, non-melanoma skin cancer/sun damage, herpes simplex/zoster, urticaria/hives, psoriasis/seborrhea, acne/rosacea, and warts (Fig 1, green area). There were numerous topics that PCPs identified as very important in their practices, but which residents deemed inadequately taught in their medical school dermatology curricula. These include skin infections, leg ulcers/wound care, cutaneous drug eruptions, infestations, and viral exanthems (Fig 1, red area). These topics should be prioritized for improvement in dermatology curricula in medical schools. These may represent areas where dermatology and other specialties might both be teaching parts of the content, but not in a coordinated manner. For example, skin infections may be included in microbiology, infectious disease, and dermatology curricula. Likewise, cutaneous drug eruptions may be considered in both pharmacology and dermatology. The departments that share responsibility in teaching these content areas would ideally coordinate their teaching, perhaps by jointly developing learning objectives and dedicating teaching time in both their curricula to adequately cover the topic. Joint teaching time during which a dermatologist shares a lecture with a microbiologist or pharmacologist might be an ideal way to
Hansra et al 27
J AM ACAD DERMATOL VOLUME 61, NUMBER 1
Table II. Percentage of residents who rated dermatologic topics as ‘‘inadequately’’ taught in medical school % Residents reporting inadequately taught Dermatologic topic
Asthma Diabetes HIV dermatology Alopecia* Cutaneous drug eruptions* Vasculitis/purpura Leg ulcers/wound care Viral exanthems Connective tissue disease (ie, lupus) Infestations (ie, scabies, lice) Bacterial/fungal skin infections, including in immunocompromised patients Warts* Non-melanoma skin cancer/sun damage* Acne/rosacea* Psoriasis/seborrhea* Melanoma/moles Atopic dermatitis/contact dermatitis* Urticaria/hives Herpes simplex/zoster
All residents
Residents who completed rotation (n = 118)
Residents who did not take rotation (n = 121)
P Value
6 7 68 60 46 42 38 36 36 35 33
7 7 68 48 36 44 36 31 33 30 25
7 8 69 72 53 46 45 40 36 40 40
.96 .66 .90 \.001 .007 .73 .15 .18 .59 .10 .29
29 26 26 25 21 18 18 13
18 19 15 13 15 9 18 8
38 34 37 31 25 21 21 13
.001 .008 \.001 .001 .066 .014 .57 .16
*Topics that were rated significantly different (ie, P \ .05) between residents who completed a rotation and residents who did not complete a rotation.
cover the topic adequately, especially given that clinicopathologic correlation and case-based learning improve knowledge retention in medical school. Content areas judged to be less important by resident physicians can be added to curricula if time permits. These include vasculitis/purpura, skin signs of connective tissue disease, alopecia, and HIV dermatology (Fig 1, yellow area). These areas are less frequently encountered in outpatient primary care practice. Some of these areas would be of greater importance for physicians practicing in certain venues. Hospitalists, for example, should be well aware of the evaluation of vasculitis and purpura, and rheumatologists should be well versed in the skin signs of connective tissue disease. These topics may be best reserved for teaching at the postgraduate level. As expected, when topics were considered unimportant, they were never rated as inadequately taught. Ratings of teaching adequacy of content areas were influenced by whether the residents completed an undergraduate dermatology clinical rotation. Residents who completed a dermatology elective felt significantly more prepared to diagnose and treat common skin disease compared to residents who did not complete a rotation. Skin infection teaching was judged to be inferior for more residents who did not take a clinical dermatology rotation (40% felt that it
was inadequately taught) when compared to those who completed a dermatology rotation (25% felt that it was inadequately taught) (Fig 1, dashed arrow). While non-melanoma skin cancer/sun damage, psoriasis/seborrhea, acne/rosacea, and warts were reported as adequately taught, these topics were inadequately taught for those who did not complete a dermatology rotation (Fig 1, solid arrows). If nonmelanoma skin cancer/sun damage, psoriasis/seborrhea, acne/rosacea, and warts are not addressed appropriately in the core curriculum, then it would be important to require, or at least highly encourage, medical students who anticipate a career in primary care to complete a clinical rotation in dermatology. San Francisco and Fresno PCPs value similar dermatologic content areas with certain exceptions. We recommend tailoring the dermatology curriculum at the postgraduate level to the needs of geographic regions in which trainees intend to practice. San Francisco PCPs place greater importance on herpes simplex/zoster, alopecia, and HIV dermatology than Fresno PCPs. Since San Francisco has a large immunocompromised population, including many HIV-infected individuals,16 this emphasis is not surprising. Fresno PCPs place greater value on melanoma/moles and non-melanoma skin cancer/sun damage. This is also not surprising given
28 Hansra et al
J AM ACAD DERMATOL JULY 2009
Fig 1. Dermatologic content areas in order of level of importance (determined by PCPs) and separated by teaching adequacy (determined by residents).
that Fresno sees higher rates of sun exposure compared to San Francisco. Most schools teach students about dermatologic procedures, but do not expect students to learn how to perform them.10 Since many PCPs perform cryotherapy and skin biopsies, these procedures should become a part of the medical school or postgraduate dermatology curriculum. As more than 60% of PCPs perform cryotherapy, it should ideally be taught in medical school or, at the very least, in all primary care residency training programs. The Association of American Medical Colleges has similarly recommended that these procedures be taught in medical school.7 Procedural training is especially important for residents planning to practice in medically and dermatologically underserved areas. Eighty-seven percent of PCPs in Fresno perform skin biopsies compared to 20% of PCPs in San Francisco. PCPs in underserved areas not only need to know how to perform the procedures correctly, but also to whom to send the tissue for evaluation and how to interpret biopsy results. Clayton, Perera, and Burge6 identified dermatologic content areas for inclusion in medical school curricula in the United Kingdom, and these overlap significantly with important topics identified by our study. One notable exception is purpura, which Clayton et al identified as important in medical school training, whereas our study did not. An important difference between our study and their
recommendations is that the British group surveyed physicians across several disciplines, especially dermatologists, and our study population consisted of community PCPs. A recent multicenter survey of the dermatology faculty responsible for teaching students at U.S. medical schools found that acne, seborrheic dermatitis, and warts were the 3 most frequently taught content areas.10 Again, these recommendations may not represent the actual needs of PCPs in practice. We argue that the most important stakeholders in determining an undergraduate dermatology medical school curriculum are PCPs, who are the first line in the diagnosis and treatment of skin conditions. In the current era of a critical shortage of dermatologists,17 teaching future PCPs when and how to appropriately refer is essential and should become a part of the medical school dermatology curriculum.6 This study has significant limitations. First, we only conducted a needs assessment of clinical dermatologic topics. We did not address the teaching of the basic structure and function of the skin, description of skin lesions, and performing a complete skin examination. We believed these topics were basic areas of knowledge that should always be included in the medical school curriculum. Second, we also only surveyed internal medicine and family medicine physicians from a few geographic regions. We felt internal medicine and family medicine
J AM ACAD DERMATOL
Hansra et al 29
VOLUME 61, NUMBER 1
physicians were the most important stakeholders in an undergraduate dermatology medical curriculum. The proposed curriculum might need to be enhanced or modified for physicians practicing in other specialties. Another limitation was the arbitrary cutoff values established for categorizing topics. These were chosen with the view that if half of PCPs thought a content area was important, it should be taught to all medical students. Similarly, if approximately one third of residents thought a content area was inadequately taught, the content area should be included and further emphasized in the undergraduate medical school curriculum. We support a core undergraduate clinical dermatology curriculum. This proposed curriculum would ensure that medical students receive exposure to the ‘‘very important’’ areas of clinical dermatology. Since PCPs perform dermatologic procedures in their offices, these skills should be introduced in medical school. Further, we recommend that certain dermatologic content areas be included or strengthened in current medical school dermatology curricula, others continued to be taught at their currently effective level, and some minimized or removed from the curriculum to allow for the addition of the currently under-taught and very important content areas (see Fig 1). The proposed dermatology curriculum might require slight modifications based on the needs of the communities in which residents intend to practice. Future smaller scale needs assessments may be warranted to help determine these regional needs. We acknowledge the contributions of the local site coordinators in obtaining institutional review board approval, distributing surveys, and logging data from each of their respective sites: Arin Isenstein, MD, at UCSF Fresno; Capt Patrick McCleskey, MC, USAF, at Brooke Army Medical Center; Stephanie Hu, MD, at Harvard Medical School, Boston, MA; Matthew Molenda, MD, at Northeastern Ohio Universities College of Medicine; Jamaiya Havel, MD, at Texas Tech University Medical Center; and Megan Seibert, MD, at University of Alabama School of Medicine. We thank Arianne Teherani, PhD, for providing her guidance in the development of our needs assessment survey. We thank Priya Rajendran, MD, for aiding in the conception of a needs assessment and piloting the first survey. We thank Jeanette Waller, MD, for helping to secure funding and recruiting local site
coordinators for the multicenter study. We thank Melissa Kanchanapoomi, BA, for aiding with data analysis. We thank Lee Zane, MD, MAS, for his vision and constructive criticism of our study. REFERENCES 1. Kern D, Thomas P, Howard D, Bass E. Curriculum Development for Medical Education: A Six Step Approach. Baltimore: Johns Hopkins University Press; 1998. 2. Laidlaw T, MacLeod H, Kaufman D, Langille D, Sargeant J. Implementing a communication skills programme in medical school: needs assessment and programme change. Med Educ 2002;36:115-24. 3. Witkin B. Assessing needs in educational and social programs: Using information to make decisions, set priorities, and allocate resources. San Francisco: Jossey-Bass; 1984. 4. Jeffe D, Andriole D, Hageman H, Whelan A. The changing paradigm of contemporary U.S. allopathic medical school graduates’ career paths: analysis of the 1997-2004 national AAMC Graduation Questionnaire database. Acad Med 2007;82: 888-94. 5. Fleischer A, Herbert C, Feldman S, O’Brien F. Diagnosis of skin disease by nondermatologists. Am J Manag Care 2000; 6:1149-56. 6. Clayton R, Perera R, Burge S. Defining the dermatological content of the undergraduate medical curriculum: a modified Delphi study. Br J Dermatol 2006;155:137-44. 7. The AAMC project on the clinical education of medical students: clinical skills education. 2005. 8. Davies E, Burge S. Audit of dermatological content of U.K. undergraduate curricula. Br J Dermatol 2009;160:999-1005. 9. Medically underserved areas/medically underserved populations: US Department of Health and Human Services: Health Resources and Services Administration; 2008. 10. McClesky P, Gilson R, Devillez R. Medical student core curriculum in dermatology. Chicago: Dermatology Teachers Exchange Group; 2007. 11. Marks J, Miller J. Lookingbill and Marks’ Principles of dermatology. Philadelphia: Saunders; 2006. 12. Bolognia J, Jorizzo J, Rapini R. Dermatology. St Louis: Mosby; 2007. 13. James W, Berger T, Elston D. Andrews’ Diseases of the skin: Clinical dermatology. Philadelphia: Saunders; 2005. 14. Wolff K, Goldsmith L, Katz S, Gilchrest B, Paller A, Leffell D. Fitzpatrick’s Dermatology in general medicine. McGraw-Hill Professional; 2007. 15. Burns D, Breathnach S, Cox N, Griffiths C, editors. Rook’s textbook of dermatology. Hoboken, NJ: Wiley-Blackwell; 2004. 16. Holmberg S. The estimated prevalence and incidence of HIV in 96 large US metropolitan areas. Am J Public Health 1996;86: 642-54. 17. Resneck J, Kimball A. The dermatology workforce shortage. J Am Acad Dermatol 2004;50:50-4.
29.e1 Hansra et al
J AM ACAD DERMATOL JULY 2009
APPENDIX. Survey for community-based primary care physicians