Ascertainment and selection of study samples.
For 10 psoriasis therapies, the distribution of never heard of; heard of, not used; and used among members and nonmembers of the National Psoriasis Foundation.
Nijsten T, Rolstad T, Feldman SR, Stern RS. Members of the National Psoriasis FoundationMore Extensive Disease and Better Informed About Treatment Options. Arch Dermatol. 2005;141(1):19-26. doi:10.1001/archderm.141.1.19
Copyright 2005 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2005
Patient advocacy groups such as the National Psoriasis Foundation (NPF) serve as representatives of those affected by disease and provide information about the condition. Our objective was to assess the extent to which NPF members differ from nonmember patients with psoriasis in their knowledge and use of therapies.
Using random-digit dialing, we identified and interviewed patients with psoriasis in the general US population. Randomly selected NPF members were also interviewed.
Main Outcome Measures
Multivariate logistic regression models were used to estimate differences (odds ratios and 95% confidence intervals) in demographic and clinical characteristics and in awareness and use of therapies between members and others diagnosed as having psoriasis.
Of 601 individuals with psoriasis identified from the general population survey, 185 provided a second interview and were defined as nonmembers. We interviewed 289 randomly selected members of the NPF. Although members were significantly older and wealthier and had more extensive disease, they reported the disease to be significantly less of a burden and were more satisfied with therapy than others affected. Compared with nonmembers, members were significantly more likely to have heard of and used most of the 10 therapies assessed. However, the proportion of respondents who were aware of a therapy and who also used it did not differ between groups.
Members of the NPF are better informed and more satisfied with available treatment options than nonmember affected patients.
Psoriasis is a common, chronic disease that affects about 4.5 million Americans.1 The extent of disease and its impact on patients’ quality of life varies considerably among patients and over time. In other chronic diseases, it has been demonstrated that providing educational information about the disease and its treatments and providing links to peer-led support groups may increase the quality of care and reduce the burden of disease.2- 4 Moreover, utilization of therapies is likely to affect both the extent of disease and quality of life impairment due to psoriasis.5- 7
The mission of most nonprofit patient advocacy organizations such as the National Psoriasis Foundation (NPF) is to improve the lives of individuals and families affected by a disease. To achieve this goal, many patient advocacy groups provide patient education, encourage public awareness, lobby for insurance coverage, and advocate for or directly fund research.8,9 Their representatives also provide congressional testimony and serve on governmental and voluntary committees, including those at the National Institutes of Health. Although patient advocacy groups have become more prominent and numerous, the characteristics of the members and the effects of these organizations are not well documented.8
The present study examines how representative members of the NPF compare with all other persons affected by psoriasis and whether members are more knowledgeable about their disease and its treatments or more likely to utilize treatment. To assess these questions, we surveyed members of the NPF, an established, intermediate-size patient advocacy group for people affected by psoriasis, and patients with psoriasis identified from the general population.
In late 2001, the NPF commissioned 2 contract survey organizations to interview a stratified sample of the general population aged 18 years or older who were residents of the contiguous 48 United States to identify people diagnosed by a physician as having psoriasis.1 The organizations conducted their surveys independently, but used the same sampling procedures and questionnaires. In the general population, households to be contacted were selected by random-digit dialing and were called up to 3 times before being replaced by a next randomly selected telephone number. Members of the NPF were randomly selected from the organization’s files and called up to 6 times before contacting the next member on the list (Figure 1). In our primary analysis, data collected by both organizations were analyzed together. To assess consistency between the 2 survey organizations, we compared estimates of selected outcomes based on the 2 survey groups’ data.
The initial survey was designed to establish the prevalence, extent, and burden of psoriasis in the United States.1 Subjects who indicated during this survey that they had been diagnosed as having psoriasis by a physician were defined as general population sample and were asked if they would be willing to be called again to answer additional questions about their disease and its treatment. Those from the general population who participated in the second telephone interview were defined as nonmembers (Figure 1). The same interview was administered to a random sample of NPF members who were current donors and who also reported a psoriasis diagnosis (Figure 1). The response rate for almost all questions asked to those who agreed to participate exceeded 90% except for treatment satisfaction among nonmembers and for annual income among both groups.
In addition to questions about demographic characteristics, knowledge about and use of treatment, disease status, and medical access, we asked respondents to estimate the number of palms needed to cover all of their psoriasis at the time of the interview (little or none, 1-2, 3-10, or >10 palms). Extensive disease was defined as 3 or more palms of psoriasis. We also asked patients to use a 10-point Likert scale to rate the extent to which psoriasis was a problem in their everyday life (1, no problem to 10, very large problem) and global treatment satisfaction (1, extremely unsatisfied to 10, extremely satisfied).10 We defined a score of 8 or higher on these 2 scales as substantialproblem and very [dis]satisfied with treatment, respectively.10
In addition to questions identical to those asked in the first interview, subjects were asked in the second interview to rate their knowledge of and experience with specific therapies for psoriasis as never heard of, heard of but never used, have used, use currently, or do not know. To simplify analyses, responses were combined to create new, nonexclusive variables. We assumed that those responding do not know were not aware of that treatment, and we therefore included them in the never heard of category. We also assumed that those who indicated prior or current use of a treatment had heard of this treatment and combined them with respondents who indicated heard of, but never used to create the heard of category. Used was defined as prior or current use of a therapy. Used if heard of was calculated as the ratio of subjects who had used a treatment to those who had heard of it.
We compared the demographic and disease characteristics of the general population sample with those of the nonmembers. Within the general population sample, we compared the characteristics of the respondents who indicated willingness to participate in a second interview with those of respondents who did not want to be contacted again. We then compared the respondents from the general population sample who indicated willingness to participate but did not provide a second interview with those who were willing and provided a second interview.
The survey was granted exemption from institutional review board coverage by an independent institutional review board.
To account for differences in the distribution of demographic and disease characteristics between the entire general population sample and those who completed a second interview (nonmembers), we calculated inverse probability weights for each person from the general population sample who participated in the second interview using standard methods.11,12 We tested for statistical differences in continuous variables using a t test and categorical variables using a χ2 or Fischer exact test, as appropriate. We used univariate and multivariate logistic models designed for survey data to estimate odds ratios (ORs) and 95% confidence intervals (CIs) to assess demographic and socioeconomic differences between members and nonmembers as well as differences in knowledge and use of specific therapies.11
To identify potential confounding variables, we used the change-in-estimate method,13 including the variables and categories listed in Table 1. First, we used a univariate analysis to estimate the OR for a variable between members and nonmembers. If adding a second variable changed the OR of the first variable more than 10% in the bivariate analysis, it was included in a multivariate model to estimate the adjusted OR for the first variable. This method was also used to select confounding variables for the 3 therapy-related variables (heard of, used, and used if heard of). However, to limit the number of bivariate analyses from 120 to 36, we selected treatment paradigms for each of the therapy classes: calcipotriene for topical therapy, psoralen–UV-A (PUVA) for photo(chemo)therapy, and methotrexate for systemic therapy. If a variable listed in Table 1 changed the OR of all 3 therapies by more than 10% in a bivariate analysis, it was included in a multivariate model.13 We applied this method to heard of, used, and used if heard of as independent variables of the analysis, separately.
For the 3 selected standard therapies from each class of therapies (calcipotriene, PUVA, and methotrexate, respectively), age, number of physicians consulted, and/or extent of disease were significantly associated with each of the 3 treatment-related variables. Occasionally (in 4 of 36 cases), another variable such as income, duration of disease, or dermatologic care changed the OR of 1 of the treatments investigated by 10% or more, but never for all 3 therapies, and the change never exceeded 20% of the univariate OR (data not shown).
The selection of confounding variables was similar if PUVA was replaced by UV-B, methotrexate by cyclosporine, or calcipotriene by tazarotene (data not shown). Therefore, we assumed this procedure to be reasonably accurate and unlikely to have excluded important confounding variables.
All statistical tests were 2 sided, and an α level of less than.05 was considered significant. All statistical analyses were performed with Stata version 7.0 (Stata Corp, College Station, Tex) and Crunch version 4 (Crunch Software Corp, Oakland, Calif).
Of the 27 220 people aged 18 years or older contacted in the survey, 601 (2.2%) indicated that their condition had been diagnosed as psoriasis by a physician (Figure 1). Of that group, which became our general population sample, 328 respondents indicated willingness to be contacted again, and 185 of those completed the second interview (nonmembers) (response rate, 57.6%). Only 1 of these 185 persons (0.5%) knew of the NPF.
Compared with those 185 persons, the 143 persons who were willing but did not complete the second interview were significantly older (mean age, 46.7 vs 51.2 years; P = .01), less likely to be retired (27.0% vs 9.8%; P<.001), less likely to indicate extensive disease (≥3 palms, 49.7% vs 12.6%; P<.001), and more likely to be very satisfied with therapy (25.4% vs 54.6%; P<.001).
Compared with the 328 respondents indicating willingness to be contacted again, the 273 who did not wanted to be contacted again were more likely to be male (44.7% vs 37.2%; P = .06) and less likely to report extensive disease (≥3 palms, 22.3% vs 7.7%; P<.001), substantial problems in everyday life (15.2% vs 8.4%; P = .009), or that they were very dissatisfied with treatment (28.1% vs 17.6%; P = .003).
Compared with all persons identified from the general population sample, nonmembers (ie, those participating in the second interview) were significantly older and more likely to report 3 or more palms covered with psoriasis (15.6% and 29.7%, respectively; P<.001) and to rate problem in everyday life higher and global treatment satisfaction lower (Table 1).
Compared with nonmembers, members were significantly older, more often affluent (≥$75 000 annual income), and more likely to be married and retired (Table 1). After adjusting these variables for age, only high annual income (≥$75 000) was significantly more common for members than nonmembers (adjusted OR, 4.19; 95% CI, 2.45-7.14). More often, members reported a significantly greater extent of disease (Table 1). Members were twice as likely to report more than 10 palms affected by psoriasis (Table 1). Despite the greater extent of disease, the problem of psoriasis in everyday life was similar, and treatment satisfaction was higher among members than nonmembers (Table 1).
Age, annual income, extent of disease, dermatologic care, and number of physicians consulted each changed the OR of both reporting psoriasis to be a substantial problem and of being globally very satisfied with therapy by 10% or more in bivariate analyses and were therefore included in a multivariate model. After adjusting for these variables, members were as likely to report the disease to be a substantial problem (adjusted OR, 0.85; 95% CI, 0.40-1.80) but remained significantly more likely to be very satisfied with therapy (adjusted OR, 1.78; 95% CI, 1.05-3.01) than nonmembers. Members consulted significantly more physicians and were more likely to receive specialized care than nonmembers (Table 1). Even after adjusting for age and extent of disease, which both changed the OR for number of physicians consulted and active dermatologic care by at least 10%, members remained more likely to have consulted 3 or more physicians (adjusted OR, 1.94; 95% CI, 1.06-3.54) and to be currently under a dermatologist’s care (adjusted OR, 2.24; 95% CI, 1.44-3.49). After adjusting for age, duration of disease was similar among both groups (adjusted OR, 0.95; 95% CI, 0.78-1.15).
Survey participants were asked about tools they used to obtain information about psoriasis. The general population sample relied most on friends, family, books, and magazines and only infrequently (16%) consulted organizations for medical information (Table 2). In contrast, nearly 95% of members indicated that they utilized organizations, among which the NPF is likely to be most prominent, to access information about psoriasis. This was the highest percentage for any source for members (Table 2). Members’ greater use of the Internet, books, and magazines is consistent with their higher incomes (Table 2).
In response to questions concerning knowledge about and use of specific psoriasis therapies, members were significantly more likely than nonmembers to have heard of each of the 10 therapies assessed (Figure 2 and Table 3). For topical treatments such as calcipotriene and tazarotene, photo(chemo)therapy, and methotrexate, the odds that members had heard of these therapies were 5 to about 25 times higher than for nonmembers (Table 3). After adjusting for age, extent of disease, and number of physicians consulted, which substantially changed the OR for treatment awareness for calcipotriene, PUVA, and methotrexate, members remained significantly more likely than nonmembers to have heard of each of the 10 therapies (Table 3). The higher adjusted ORs of members being aware of 2 topical therapies such as calcipotriene and tazarotene and 2 systemic therapies such as PUVA and methotrexate were particularly striking (adjusted ORs ≥9) (Table 3).
Except for the immunomodulatory drugs such as cyclosporine, infliximab, and etanercept, members were significantly more likely to have used most therapies (odds ranged from 3 to 16) (Figure 2 and Table 4). Differences in usage between members and nonmembers were greatest for tazarotene, photo(chemo)therapy, and methotrexate. The use of calcipotriene, PUVA, and methotrexate significantly varied with age and extent of disease. However, the higher use for members changed little after adjusting for age and extent of disease (Table 4).
In contrast to marked differences in knowledge and overall use, the proportion of patients with psoriasis who had heard of a therapy and who also reported using it did not significantly differ between members and nonmembers for 7 of 10 therapies studied (Figure 2 and Table 4). After adjusting for extent of disease and number of physicians consulted, the likelihood of use once heard of was comparable between groups for all therapies (Table 4), except for UV-B, which members were more likely to use (adjusted OR, 4.45; 95% CI, 1.95-10.15).
Members of the NPF differ in several aspects from individuals diagnosed as having psoriasis from the general population. First, members are older, more affluent, and more likely to report extensive disease than individuals from the general population. Second, although only 1% of patients with psoriasis belong to the NPF, membership is not disproportionately composed of those overwhelmed by their disease or alienated from medical care. On the contrary, members reported disease of the same extent to be less of a burden and were less often dissatisfied with therapy than nonmembers. This may be owing to a fuller understanding of the disease, much greater knowledge and use of most available treatment options, peer-to-peer contacts, or, most likely, a combination of these and other factors.
Third, our data suggest that membership in the NPF is significantly associated with higher awareness of treatment options. Although these findings may in part reflect self-selection for membership (ie, well-informed patients are more likely to join the NPF), they may also demonstrate the effectiveness of the organization’s educational efforts. Education of affected individuals about their disease and treatment options is a key goal of patient advocacy groups such as the NPF, and prior studies have suggested that health advocacy group educational efforts and peer-to-peer support may improve care and patient satisfaction and reduce burden due to chronic conditions.2- 4
As in any cross-sectional study, our study demonstrates associations between variables. These associations do not prove causality. We are unable to determine from our data, nor would we be confident in doing so from other retrospective data, the extent to which members were aware of treatments before joining the NPF. In fact, it is possible that persons learned about psoriasis therapies from sources other than the NPF, and in learning about psoriasis treatment from other sources also learned about and were motivated to join this organization. The 6-fold higher percentage of members compared with the general population survey participants (93% vs 15.5%) who indicated that organizations were a source of information about psoriasis strongly suggests that the NPF played an important role in informing these members about their disease and its treatment.
Because more than three fourths of the nonmembers and two thirds of the members reported limited disease (<3 palms to cover psoriasis), it is not surprising that substantially more people in both groups were aware of topical than of systemic therapies, which are generally reserved for severely affected patients. However, of the 10 therapies we assessed, most nonmembers, including those with extensive disease, knew only about topical steroids. Members were more than 20-fold more likely to have heard of a topical vitamin D derivative, which, after topical steroids, is the next most used prescription drug for psoriasis.14 Knowledge about treatments for more severe psoriasis varied greatly between groups and among therapies. These differences were greatest for older therapies such as UV-B, PUVA, and methotrexate and least for therapies that were newer or (in 2 cases) were still awaiting US Food and Drug Administration approval for this indication. These findings may be owing to increased understanding of adverse effects of older therapies, physicians adopting newer treatments, the role pharmaceutical companies play in informing patients with psoriasis about their products,15 or other factors.
Although the NPF is comparable with other advocacy groups for patients with chronic conditions, such as diabetes mellitus16 and rheumatoid arthritis,17 only about 1% of individuals with psoriasis are aware of or are members of the NPF. This observation suggests that the NPF is not yet reaching most of its intended population, the 4.5 million Americans with psoriasis.8 Also, only a minority of members know about all standard psoriasis treatments, which suggests that its informational role could be improved for members as well.
Does treatment awareness lead to treatment use? Generally, differences in absolute use rates paralleled differences in knowledge, with the greatest differences arising between members and nonmembers for those treatments usually administered by dermatologists, such as UV-B. Members were 2- to 5-fold more likely to use 8 of 10 of the therapies we assessed. Alternatively, additional knowledge about treatment-related adverse effects could make patients more reticent to use certain treatments and lower treatment satisfaction. Our data suggest that this is not likely to be a common occurrence for NPF members.
Data from surveys of members of patient advocacy groups for psoriasis in Scandinavia and the Netherlands indicated that use of photo(chemo)therapy is comparable between American, Dutch, and Scandinavian members of specialty organizations.18,19 American members were more likely to use systemic therapies such as methotrexate, oral retinoids, or cyclosporine.18,19
However, in contrast to large differences in absolute use rates, the proportion of members and nonmembers who knew of a therapy and used it was comparable for 9 of 10 therapies (all but UV-B). This finding suggests that the added knowledge about psoriasis therapies among NPF members increases treatment use by creating greater treatment awareness but may not increase the probability that a person made aware of a therapy will actually use it. Therefore, informing patients about available treatment options may increase satisfaction with care and decrease burden of disease2- 4 without motivating patients to use therapies.
The strengths of this study include the representative nature of both samples, with nonmembers identified from a nationwide survey (>27 000 persons) and a random sample of all members being interviewed. Random-digit dialing is a widely used and well-accepted method for population-based health research, including in cross-sectional surveys such as ours.20 However, noncoverage of households without a telephone and potentially low and often unknown response rates may bias our results.20
The consistency of the results between the 2 independent survey organizations (data not shown) further suggests that sampling bias is unlikely to be a major problem and supports our combining the data collected by the 2 organizations. Because the survey was introduced as a “general health survey” and the first questions did not ask about psoriasis, substantial response bias is unlikely.
The differences in the characteristics of persons with psoriasis ascertained in the general population sample who did not participate in a second interview were similar to those observed in other studies.21- 23 To account for these differences, we constructed inverse probability weights and used statistical methods for surveys.11,12 Nevertheless, we cannot exclude the possibility that these 2 groups differed in treatment awareness, use, and satisfaction. The validity of the diagnosis for self-reported psoriasis has not been studied, but in other common chronic diseases, self-reported diagnosis appears to be reasonably accurate.24 When compared with physician evaluation, self-administered psoriasis severity questionnaires showed good agreement between physicians and patients, but a very different measure of severity was used.25,26 The number of palms affected by the disease (about 1% of total body surface) is derived from the “rule of nines” and has been accepted as a practical technique to assess involvement of body surface area but has never had formal validation testing.27 Our patient-oriented questions on treatment satisfaction and burden of disease are necessarily subjective but may be as sensitive as other measures of health status.28
In conclusion, our findings suggest that the demographic, clinical, and psychological characteristics of members of the NPF organizations differ from those of all other individuals affected. Also, joining this specialty organization is associated with increased knowledge about treatments without an apparent increase in the likelihood that a patient who is knowledgeable about a treatment will use it. Hopefully, this information improves both care of patients with psoriasis and the adjustment of these patients to their chronic condition. The low prevalence of membership, failure to reach the poor, and less satisfaction and less complete information among nonmembers about disease treatment suggest that there are substantial gaps in most patients’ knowledge not currently being filled by patient advocacy groups, physicians, or others. These gaps may impede patients’ availing themselves of effective therapies.
Correspondence: Robert S. Stern, MD, Department of Dermatology, Beth Israel Deaconess Medical Center, 330 Brookline Ave, Boston, MA 02215 (email@example.com).
Accepted for Publication: June 30, 2004.
Funding/Support: This work was funded in part by unrestricted grants from Amgen, Thousand Oaks, Calif, and Biogen, Cambridge, Mass; Public Health Service contract N01-AR-0-2246 from the National Health Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health, Department of Health and Human Services, Bethesda, Md (Dr Stern); and a grant from the Fund for Scientific Research, Flanders, Belgium (FWO-Vlaanderen) (Dr Nijsten).
Financial Disclosure: None.