The Primary Care Provider and the Care of Skin Disease: Title and subTitle BreakThe Patient's Perspective FREE

PATIENTS WITH skin disorders are extremely common. Approximately 6%of visits to all physicians entail a problem of the skin, hair, or nails;however, only approximately 40% of these patients are seen by dermatologists.¹ In primary care settings, the proportion of patientvisits involving dermatologic complaints is even higher, where up to one fourthof visits involve skin disorders.² As manyhealth care systems have adopted a "gatekeeper" system, whereby a patientmust first see his or her primary care provider (PCP), who then determineswhether the patient is in need of specialty care, information is clearly neededregarding the quality of dermatologic care delivered by PCPs.

Several studies^{3 - 6} have evaluated how PCPs compare with specialists in the diagnosis and/or treatmentof disease. Although PCP care may be comparable to that of specialists forcertain disorders (eg, hypertension and type 2 diabetes mellitus), such isnot the case for skin disease. Researchers^{7 - 14} have demonstrated that dermatologists provide greater accuracy in diagnosisand more appropriate treatment of skin disorders compared with PCPs.

Another important aspect of quality of care is patient satisfaction,since ultimately, patients are the purchasers of health care resources. Littleis known, however, about patients' preferences for the evaluation and treatmentof skin disorders. One previous study¹⁵ conductedin a dermatology practice suggested that patients were more satisfied withdermatologist care for their skin disease compared with PCP care, and thatdirect access to dermatologists was important. However since the study participantswere enrolled in a dermatologist's office, a referral bias in favor of dermatologiccare could have been introduced, as the study sample likely included individualswho were "PCP failures." In contrast, patients successfully treated by theirPCPs for a given skin disorder—and who may have a more favorable opinionof the care provided by PCPs—would have a lower probability of beingreferred for specialty evaluation. To test this hypothesis, we surveyed patientsin dermatology and primary care clinics and determined their confidence inPCPs' and dermatologists' abilities to diagnose and treat skin disease.

The West Haven Veterans Affairs Medical Center, West Haven, Conn, providescare for more than 29 000 veterans who are older (average age, 61 years)and predominantly male (91%). Patients are enrolled in primary care, and referralto subspecialty clinics must be made by patients' PCPs. Primary care clinicsare staffed by internal medicine attending physicians, nurse practitioners,physician assistants, and internal medicine residents (who work under thesupervision of the attending physicians). Dermatology clinics are staffedby dermatology attending physicians and dermatology residents, who work undertheir supervision. The attending physicians and residents in the dermatologyand primary care clinics are affiliated with the Yale University School ofMedicine, New Haven, Conn.

Eligible patients were those veterans in the waiting rooms of the primarycare and dermatology clinics awaiting scheduled appointments. Subjects wererecruited on 14 clinic days during a 2-month period. A trained research assistantadministered the study questionnaires to a convenience sample of patientsbefore their primary care or dermatology clinic appointment.

All participants were asked to rate their level of confidence in theirPCP's ability to: (1) care for any rashes or skin conditions that might develop,(2) diagnose skin cancer, (3) perform skin biopsies, (4) "freeze" skin lesionswith liquid nitrogen, and (5) perform cutaneous surgery. Study subjects wereasked identical questions regarding their confidence in a dermatologist, andthey were asked if they had ever been treated by dermatologists and/or PCPsfor skin conditions and about their satisfaction with the treatment.

Possible responses to questions regarding patient confidence were "stronglyagree," "agree," "neither agree nor disagree," "disagree," and "strongly disagree."For analysis, the responses "strongly agree" and "agree" were combined, whilethe 3 remaining categories were also combined into an "all others" category.Subjects who had received previous treatment for a skin problem were askedto rate how satisfied they were with previous care. The responses "very satisfied"and "somewhat satisfied" were combined, while the 3 remaining responses ("neithersatisfied nor dissatisfied," "somewhat dissatisfied," and "very dissatisfied")were combined.

Subjects were also asked where they preferred to have their skin problemscared for and should they develop a skin problem, whether they preferred tohave direct access to dermatologists without having to be seen by their PCP.

Demographic information, including participants' age, sex, level ofeducation, and self-reported health status, was obtained at the interview.

To assess for potential differences between the groups, we used a χ² or Fisher exact test for categorical variables and t tests for dimensional variables. P<.05(2-tailed) was considered significant.

The study protocol was approved by the local investigational reviewboard.

A total of 311 patients were asked to participate: 177 in the primarycare clinic and 134 in the dermatology clinic. Of these, 237 patients completedthe questionnaire, for an overall participation rate of 76%. Rates of participationdid not vary by clinic site: 133 (75%) of the 177 general medicine clinicpatients (group 1) agreed to participate, compared with 100 (75%) of the 134dermatology clinic patients (group 2). Subjects' average age was 65.7 years(65.8 years for group 1 vs 65.6 years for group 2), and 95% were male (group1 vs group 2, 94% vs 96%). There was no difference between the groups foreducational level or perception of self-rated health status (data not shown).

Subjects (N = 237) reported confidence in their PCP's ability to takecare of any rashes or skin conditions that develop, to diagnose skin cancer,to perform skin biopsies, to freeze skin lesions with liquid nitrogen, andto perform surgery on their skin. When the responses of groups 1 and 2 werecompared (Table 1), group 2 subjectswere less likely than group 1 participants to be confident in their PCP'sability to take care of any skin condition that might develop, to diagnoseskin cancer, to perform skin biopsies, and to perform cutaneous surgery.

Patients reported confidence in a dermatologist's ability to take careof any skin conditions that might develop, to diagnose skin cancer, to performskin biopsies, to freeze lesions with liquid nitrogen, and to perform cutaneoussurgery. No statistical difference was detected between groups 1 and 2 forany of these outcomes (Table 2).Subjects were significantly more confident in a dermatologist's ability toperform these procedures or measures than a PCP's ability (P<.001 for each comparison).

Thirty-five percent of group 1 (47/135) and 40% of group 2 (40/100)participants reported having been treated by their PCP for a skin condition(P = .41). Among this subgroup, 68 (78%) reportedsatisfaction with the previous treatment of their skin problem by their PCP.There was no significant difference in satisfaction with care by their PCPbetween the 2 groups (81% for group 1 and 75% for group 2; P = .50). In addition, among group 1 subjects, there was no differencein satisfaction with the care rendered by their PCP, whether they were alsoenrolled in a dermatology clinic (P = .33).

A total of 90 group 1 and 78 group 2 participants had previously beentreated by dermatologists for skin conditions. There was no difference inreported satisfaction between groups for the care rendered by dermatologists(92% vs 90%; P = .80). Patient satisfaction withdermatologist care was higher than that rendered by PCPs (P<.001).

Subjects were asked whom they would prefer to take care of their skinproblems (Table 3). Of the 234respondents, 181 preferred dermatologists, 29 preferred their PCP, 6 preferredother physicians, and 18 had no preference. Of the 134 group 1 respondents,26 preferred PCP care for their skin problems, while only 3 of the 100 group2 participants preferred PCP care. Similarly, 91 group 1 patients preferredthat a dermatologist care for their skin problems, while 90 group 2 patientspreferred a dermatologist.

Ninety-one (66%) of the 137 group 1 patients vs 79 (79%) of the 100group 2 patients agreed with the statement that they would prefer to see adermatologist directly without having to see their PCP should they developa skin problem (P<.001).

We looked to see if the type of PCP influenced patient satisfaction.Of the patients who had known PCPs, 151 were cared for by attending physiciansin general internal medicine, 23 by internal medicine house staff, and 52by nurse practitioners or physician assistants. For all measures other thanconfidence in PCP's ability to perform a biopsy (P= .03), no significant differences were detected between attending physiciansand other providers (house staff, nurse practitioners, and physician assistants).

The results of our survey further contribute to the growing literatureregarding the quality of care for patients with skin disorders. Previous studieshave demonstrated that dermatologists were superior to PCPs for the diagnosisand treatment of common dermatoses,^{7 - 12} the ability to recognize lesions suggestive of malignant neoplasms,^{13 ,16} and the appropriate use of topicalantifungal therapy.¹⁷

We found that although patients were more likely to have confidencein a dermatologist than their PCP for the care of skin disorders, they werestill confident in the abilities of their PCP. This was true not only fordiagnosis and treatment but for the performance of more invasive procedures,such as skin biopsy, the application of liquid nitrogen, and cutaneous surgery.This finding has important quality-of-care ramifications, since PCPs are beingtrained in these techniques.¹⁸ Despite evidencesuggesting that the care of skin disease may be better served by dermatologists,^{7 - 17} patients may not be entirely averse to having their PCP assume a greater rolein the delivery of this care.

We also found that among those who had been treated previously for askin disorder by their PCP, there was a high satisfaction rate, although thiswas significantly less than the satisfaction rate of subjects treated by dermatologists(78% vs 90%). In addition, no significant difference in satisfaction was detectedwhether they were questioned in the dermatology or primary care clinic. Forprevious treatment, we therefore could not confirm the potential bias introducedin the study of Owen et al,¹⁵ the first tolook at patient opinion with respect to who cares for their skin disease.However, we found that primary care patients were more likely to express confidencein their PCP and prefer PCP treatment of their skin problems than those questionedin the dermatology clinic.

Owen et al¹⁵ reported that 64% of patientsin their study who had been treated by a PCP for a skin condition were satisfiedwith the care rendered, which is comparable to, but slightly lower than, the78% we found. Possible explanations for the lower PCP satisfaction rate foundin their study include the finding that 11% of participants had previouslyseen 2 or 3 physicians for their condition, which raises the possibility thattheir patients had either higher expectations or more severe illness thanwe encountered. Furthermore, the skills of the university-affiliated physiciansin our study might be different from those of the PCPs in the study of Owenet al. Interestingly, while the physicians in our study were all trained ininternal medicine, there was a high percentage of family physicians and generalpractitioners among the PCP group in their study. Previous research¹⁹ has suggested that family physicians and generalpractitioners are superior to internists when tested on diagnostic abilitieswith respect to skin disease.

How can we reconcile the findings of previous studies that suggest thatPCPs are inferior to dermatologists in the diagnosis and treatment of skindisorders with the high rate of satisfaction and confidence in PCP care foundin our study? It is possible that since most skin conditions are not lifethreatening and often are self-limiting, they may spontaneously improve regardlessof the therapy rendered. Alternatively, patients may harbor positive feelings,in general, toward their PCP, who is involved in such broadly encompassingmedical endeavors as chronic disease management, disease prevention, cancerscreening, patient education, and overall coordination of care, and with whompatients have developed a longitudinal relation.

Our study has several possible limitations. Because our study populationconsisted largely of older, male veterans, our results may not be generalizableto other health care systems. In addition, patient satisfaction with dermatologycare may be different in other systems due to the reliance of dermatologyresidents in training at our facility. However, in our study, with the exceptionof only one measure, the performance of skin biopsies, patient confidencein their PCP was no different whether the PCP was an attending physician.Whether these findings could be replicated in traditional managed care organizationsshould be the subject of future investigation.

Patient satisfaction is clearly an important measure of quality of care.Despite patient confidence in and satisfaction with PCP care of skin problems,patients reported a higher degree of confidence in and satisfaction with dermatologists.Furthermore, most patients prefer having direct access to a dermatologistfor the care of their skin disease. The greater patient satisfaction engenderedby direct access may provide dividends to managed care systems by making themmore attractive to health care consumers.

Future studies should also evaluate whether the outcomes of care, financialand clinical, rendered by PCPs and dermatologists differ, which would furtherhelp shape health care policy. Until these studies are completed, managedcare organizations should consider incorporating patient preferences and expectationsinto policy as logistical constraints permit.

Accepted for publication September 12, 2000.

This study was supported by Career Development Awards from the Departmentof Veterans Affairs Health Services Research and Development Service, Washington,DC, and the Robert Wood Johnson Foundation, Princeton, NJ (Dr Reid); a grantfrom Bristol-Myers Squibb, Wallingford, Conn (Dr Feldman); and the DermatologyFoundation's Clinical Career Development Award in Health Care Policy (Dr Kirsner).

We thank Sue Kancir, RN, for her assistance with data analysis and JoanneHart for her assistance with manuscript preparation.

Corresponding author: Daniel G. Federman, MD, Veterans Affairs ConnecticutHealth Care System (11 ACSL), 950 Campbell Ave, West Haven, CT 06516 (e-mail: FEDERMAN.DANIEL_G+@WEST-HAVEN.VA.GOV).