Scatterplot of observed values for Psoriasis Area and Severity Index (PASI) and self-administered PASI (SAPASI).
Sampogna F, Sera F, Mazzotti E, Pasquini P, Picardi A, Abeni D, and the IDI Multipurpose Psoriasis Research on Vital Experiences (IMPROVE) Study Group. Performance of the Self-administered Psoriasis Area and Severity Index in Evaluating Clinical and Sociodemographic Subgroups of Patients With Psoriasis. Arch Dermatol. 2003;139(3):353-358. doi:10.1001/archderm.139.3.353
MichaelBigbyMDRosamariaCoronaDSc, MDDamianoAbeniMD, MPHPaoloPasquiniMD, MPHMoysesSzkloMD, MPH, DrPHHywelWilliamsMSc, PhD, FRCP
There is a need to evaluate severity of psoriasis with a simple, patient-assessed instrument.
To investigate whether the self-administered Psoriasis Area and Severity Index (SAPASI) could be used as a measure of severity in different clinical types of psoriasis.
Hospital-based cross-sectional study, with measures of clinical severity collected separately by dermatologists (PASI) and patients with psoriasis (SAPASI).
Part of a large project on clinical, epidemiological, emotional, and quality-of-life aspects of psoriasis (the IDI Multipurpose Psoriasis Research on Vital Experiences study), performed between February 21 and August 31, 2000, at the inpatient wards of the Istituto Dermopatico dell'Immacolata–Istituto di Ricovero e Cura a Carattere Scientifico (IDI-IRCCS), Rome, Italy.
The study population comprised 351 eligible patients with complete sets of information on PASI and SAPASI hospitalized at IDI-IRCCS with a diagnosis of psoriasis.
Main Outcome Measures
Correlation between PASI and SAPASI scores and analysis of variance on the difference between PASI and SAPASI scores in subsets of patients based on clinical and sociodemographic characteristics.
A high correlation between the 2 measures was observed (overall Pearson correlation coefficient, r = 0.69). The SAPASI values were higher and had a wider scattering than PASI values, and SAPASI was able to discriminate properly between clinical types and global severity as assessed by dermatologists.
The SAPASI scoring system is well understood and accepted by patients in different populations than previously tested, adding confidence in the validity of the instrument. It could be used as a severity measure for psoriasis even for "at-distance" follow-up. Some caution, though, is needed when using SAPASI strictly to estimate PASI measurements, especially for guttate psoriasis.
IN CLINICAL practice, as well as in clinical research, it is important to objectively and accurately measure disease severity using standardized methods, both to compare different patients with different disease characteristics (discriminative properties) and to monitor the disease course within individual patients (evaluative properties).1,2
The assessment of psoriasis severity from a clinical point of view is frequently performed using the Psoriasis Area and Severity Index (PASI).3- 5 This score is based on the quantitative assessment of erythema, desquamation, and induration of the plaques, combined with the skin surface area involved. Because PASI measures are time-consuming and need trained personnel, an alternative measure has been proposed, the self-administered PASI (SAPASI). A number of studies suggested that this patient-assessed outcome measure is reliable and valid.6,7 Recently, the relevance of patient-assessed measures has been acknowledged also in evidence-based medicine for the evaluation of treatment effectiveness. In fact, important systematic reviews have chosen as main outcome measures "changes in patient-rated symptoms" and "global severity as rated by patients."8
We report the results of a cross-sectional analysis on baseline data of a sample of patients hospitalized at the Istituto Dermopatico dell'Immacolata–Istituto di Ricovero e Cura a Carattere Scientifico (IDI-IRCCS) of Rome, Italy, included in our IDI Multipurpose Psoriasis Research on Vital Experiences (IMPROVE) study. This is the largest sample in which SAPASI was administered so far and the first study to compare severity among subgroups using this measure. We compared SAPASI scores with PASI scores to understand whether the SAPASI scoring system could be an easier and less expensive way of evaluating psoriasis in different clinical types of the disease with a wide range of severity. Also, we aimed at evaluating whether SAPASI is well accepted in different clinical settings by patients of a culture different from the one in which it was originally designed and tested.
This study was carried out within the framework of a large project on clinical, epidemiological, emotional, and quality-of-life aspects of psoriasis (IMPROVE study). It was performed between February 21 and August 31, 2000, at the inpatient wards of a large dermatological hospital (IDI-IRCCS). Although the hospitalization pattern is rapidly changing, in Italy, patients with a wide range of clinical presentations of psoriasis are hospitalized for diagnostic procedures and treatments that may not be widely available (eg, psoralen–UV-A), especially in the more disadvantaged central and southern regions.
During the study period, patients hospitalized at IDI-IRCCS with a diagnosis of psoriasis were contacted by our research staff. After the diagnosis of psoriasis was confirmed by the senior staff of the hospital, specifically trained dermatologists verified the inclusion criteria and explained the purposes and methods of the study to the patients. Inclusion criteria were age 18 years or older, absence of any severe mental or physical illness, at least 5 years of education, ability to read Italian, and first hospitalization for psoriasis since the date of the beginning of the study. All eligible patients who gave their written informed consent were recruited in the study. Ethical approval was granted by the institutional review board of IDI-IRCCS prior to the study commencing.
To minimize sources of variation, PASI readings were made during the first hours of hospitalization before any medication was administered. The dermatologists also collected information on sociodemographic variables, clinical history, and other factors of clinical interest (eg, clinical type and location of the disease, personal and family history of psoriasis and other diseases, seasonal changes, and symptoms). At the end of the visit, the physician rated the severity of the skin condition on a 5-point scale used in previous studies.9- 11 After the visit, patients were given several self-administered questionnaires including the SAPASI, and dermatologists explained to the patients how to record information.
The PASI and SAPASI scores were computed using their original formulas.3,6 To derive the PASI score, a numerical value of 0 to 6 (0, no involvement; 1, <10%; 2, 10%-29%; 3, 30% 49%; 4, 50%-69%; 5, 70%-89%, and 6, 90%-100%) is assigned to the area of involvement for the 4 body regions (ie, head, trunk, and upper and lower extremities). Then, for each region, erythema, desquamation, and induration of the plaques are rated according to the classic 5-point scale (0, no involvement; 1, slight; 2, moderate; 3, striking; and 4, exceptionally striking). For the computation of the proportion of the areas involved, we used a detailed approach, measuring the area of involvement using figures of known area printed on transparent sheets. The percentage of involved area for each body district was then calculated, using as the denominator the body surface area (BSA) computed on the basis of weight and height, according to the Mosteller formula.12,13 This approach was used to obtain more accurate measurements of the area affected by psoriasis and more detailed information on the number, size, and location of plaques to use for further specific statistical analyses (eg, elaboration of other indexes of severity).
The SAPASI is a structured instrument consisting of a line-drawing silhouette of the front and back of a body for patients to shade in the affected areas and of 3 visual analog scales for recording the erythema, induration, and scaliness of one's own average lesion. Based on the silhouette shading, one of us (E.M.), who had not seen the patients and was blind to PASI scores, assigned a numeric value of 0 to 6 using the same cut points of PASI. Similar to the PASI, scores on the SAPASI can range from 0 to 72.
An analysis of variance was first performed, separately for PASI and SAPASI, to test for differences in mean scores between subsets of patients based on sex, age, marital status, education, occupation, age at onset, duration and severity of psoriasis, clinical type, localization, and presence of symptoms. To evaluate the performance of SAPASI, we studied its correlation with PASI, measuring the Pearson correlation coefficient. Because both measures had skewed frequency distributions, we used the logarithm of the 2 scores (lnPASI and lnSAPASI) to comply with a basic assumption of the analysis of correlations (ie, normal frequency distribution of the measures to correlate). An analysis of variance was then applied to the difference between PASI and SAPASI scores of each individual to see if the difference between the scores changed significantly in subsets of patients based on the above-mentioned variables. Erythrodermic psoriasis and other rare forms (ie, generalized, inverse, follicular, and nail psoriasis), as well as the last 2 classes of the global severity assessment, were grouped together because of the small number of cases.
We also studied the correlation of the subcomponents of BSA, erythema, desquamation, and induration to see whether physicians and patients assessed in a different way each of the subcomponents in the 2 instruments. While BSAs are measured in PASI and SAPASI on the same scale, erythema, desquamation, and induration cannot be directly compared because SAPASI considers only an "average" lesion, whereas PASI measures each subcomponent in each of the 4 regions of the body. Therefore, to make the 2 measurements comparable, we computed average PASI scores for each subcomponent, weighting the site-specific scores by the classic proportions of BSA used in the PASI. For example, Average Induration = (0.1 × Ih) + (0.2 × Iu) + (0.3 × It) + (0.4 × Il), where I(x) represents site-specific induration for the head (h), upper extremities (u), trunk (t), and lower extremities (l). For all comparisons between PASI and SAPASI subcomponents, Spearman ρ correlation coefficient was computed. All statistical analyses were performed with SPSS, version 9.0 for Windows.14
During the study period, 636 patients were hospitalized at IDI-IRCCS with an admission diagnosis of psoriasis, 376 (59.1%) of whom were eligible and agreed to participate. Of these 376 patients, only 14 (3.7%) did not provide the full information needed to compute SAPASI scores, while some information needed to compute PASI scores was missing for another 11 patients. In our analyses, we considered only the 351 patients (88.6% of those eligible) with complete sets of information for both PASI and SAPASI.
The sociodemographic and clinical characteristics of the sample are summarized in Table 1. The distributions of both PASI and SAPASI scores were asymmetric. Their means were 8.5 and 14.7 and medians, 6.8 and 11.2, respectively. The interquartile range was 4.7 to 9.5 for PASI and 6.8 to 19.8 for SAPASI. To explore the relationship between crude PASI and SAPASI scores of each individual, a scatterplot was produced (Figure 1). The overall correlation coefficient between lnPASI and lnSAPASI was 0.69. Correlation coefficients remained high in almost all subgroups within each variable of interest. In all different subgroups of variables, mean values of SAPASI were higher than those of PASI.
Table 2 summarizes the results concerning the clinically most important findings. The PASI and SAPASI scores showed a similar pattern in most variables. However, clinical severity measured by PASI was higher in men than in women, whereas it was similar when self-evaluated using the SAPASI. Both PASI and SAPASI were highly associated with dermatologist-rated severity of psoriasis. They were also significantly different according to clinical type. Besides the rarest forms, psoriatic arthritis was clinically the more severe type when PASI was used, whereas in self-evaluation, guttate psoriasis showed a higher mean value. The significant difference in the scores in the evaluation of guttate clinical type was confirmed when analyzing the differences between PASI and SAPASI. Finally, the difference between the 2 scores significantly increased with physician-rated severity.
When particular localizations of the lesions were considered (eg, palm of the hands and genitals), PASI and SAPASI scores were significantly higher in the group with lesions compared with the group without such lesions, with the exception of localization on hands when measured with the SAPASI. As for symptoms, itching was not associated with higher scores. The differences between the scores were not significantly different in the subgroups of variables when considering symptoms and specific localization. The only notable exception is that the difference between PASI and SAPASI scores was significantly larger among people with lesions on the face. The correlation between the 2 scores was generally high.
In regard to the subcomponents of the 2 instruments, moderate to good and statistically significant correlations (P<.05) were found for each body area, with ρ = 0.55 for the head, ρ = 0.40 for the upper extremities, ρ = 0.68 for the trunk, and ρ = 0.49 for the lower extremities. The correlation coefficients between the average erythema, induration, and desquamation from the SAPASI and the computed average PASI severity measures, although statistically significant, were moderate to modest (ρ = 0.37 for erythema, ρ = 0.30 for desquamation, and ρ = 0.24 for induration).
In our study, the SAPASI measurements showed almost invariably, for the different subgroups of the various factors of interest, a high correlation with PASI scores. Also in previous studies6,7,15,16 it was suggested that SAPASI was a good predictor of PASI. An excellent correlation between the 2 scores was found, although with substantial differences in the estimation for individual patients.6 Separating the SAPASI into its subcomponents, Feldman et al6 found that BSA, erythema, desquamation, and induration were significantly correlated with the clinician-rated PASI subcomponents, with the exception of the evaluation of induration in women. Also in our study, induration was the SAPASI subcomponent that more poorly correlated with the corresponding subcomponent of PASI. As it has been pointed out by Exum et al,17 it should be investigated whether there is concordance between clinicians in the evaluation of PASI components to understand if the induration is in general a difficult component to evaluate.
Because average SAPASI scores are consistently higher than those of PASI, patients might tend to overestimate their state compared with the evaluation performed by the dermatologists. This is more evident as scores on the physician global severity scale increase (ie, there is a differential bias in SAPASI score compared with PASI score with regard to severity evaluated by the clinician). The same hypothesis could concern clinical type for which discrepancies in the 2 measures were observed. The differences observed in guttate psoriasis could reflect the different evaluation of the extension of this clinical form from the dermatologist's or from the patient's point of view. Patients tend to consider the whole body area on which small lesions are scattered as being affected and not only the actual surface of the lesions. This way of self-evaluating guttate psoriasis could also explain why SAPASI does not show the difference between men and women observed when using PASI. When we excluded the 52 guttate cases from the analysis, the P values observed for sex and age when using PASI were much more similar to those obtained when using SAPASI (P = .12 and P = .40, respectively) than those from the full set of data reported in Table 2. In our sample, guttate psoriasis is more frequent in women (22.6%) than in men (11.1%). Italy (e-mail: email@example.com).
It should also be considered that in SAPASI, patients evaluate an "average" psoriatic lesion instead of separately scoring lesional severity in each body area. This could explain in part the higher values of SAPASI if we imagine that patients might be biased toward their more severe lesions instead of evaluating an average lesion. However, when we compared the SAPASI subcomponent (erythema, desquamation, and induration) scores with those of the most severe lesion of the 4 PASI body areas instead of comparing them with the mean PASI value, the correlations did not substantially increase (data not shown).
It is more probable that a subjective component is present in self-evaluation. Probably, the evaluation of severity by the patients is influenced by the burden of the disease on their quality of life. In other words, it could be envisaged that when evaluating the severity of their disease, patients consider and include factors that contribute to severity but are not considered in PASI and SAPASI. For instance, patients with pruritus might emphasize their "reading" of 1 or more of the parameters that contribute to the SAPASI score.
Our results are useful because they show that SAPASI is well accepted by patients with different types of psoriasis who seem able to evaluate reliably the severity of their condition and accurately report the extent, redness, scaliness, and thickness of their psoriatic lesions. This observation implies that dermatologists could obtain measurements of psoriasis severity without even seeing their patients, using mail or via the Internet. For instance, we have started to give SAPASI forms with stamped self-addressed envelopes to patients at hospital discharge to obtain follow-up measurements. However, some caution is needed when using SAPASI strictly to estimate PASI measurements, especially for certain clinical types (ie, guttate psoriasis).
Finally, a complementary use of SAPASI and of patient-centered measures investigating the patient's emotional life and social functioning would provide a more comprehensive assessment of psoriasis.18 Such an approach is strongly needed to depict the complex biopsychosocial burden of this disease.19,20
Corresponding author and reprints: Damiano Abeni, MD, MPH, Clinical Epidemiology Unit, Istituto Dermopatico dell'Immacolata–Istituto di Ricovero e Cura a Carattere Scientifico, Via dei Monti di Creta 104, 00167 Rome, Italy (e-mail: firstname.lastname@example.org).
Accepted for publication October 31, 2002.
This study was partially supported by grant ICS-120.4/RF98.7 of the "Progetto Ricerca Finalizzata" and by "Progetto Ricerca Corrente, Dermatoepidemiologia, 2001-2002" of the Italian Ministry of Health, Rome.
We thank Maurizio Inzillo, who provided technical support in the preparation of the figures.
A cooperative effort of the Clinical Epidemiology Unit of the Istituto Dermopatico dell'Immacolata–Istituto di Ricovero e Cura a Carattere Scientifico (IDI-IRCCS) and the Archives of Dermatology
The IMPROVE Investigators Who Contributed to This Study
Massimo Alotto, MD, Gianluca Antonelli, MD, Simone Bolli, Rino Cavalieri, MD, Massimo Luca Chinni, MD, Marcello Fazio, MD, Giampiero Girolomoni, MD, Elisabetta Luchetti, Carmelo Franco Melchi, MD, Nidia Melo Salcedo, BA, Paola Moscatelli, MD, Paolo Piazza, MD, Orietta Picconi, DStat, Maria Antonietta Pilla, MD, Grazia Primavera, MD, Pietro Puddu, MD, Paolo Ruatti, MD, Giuseppe Ruggiero, MD, Valentina Salvatori, MD, Romeo Simoni, MD, Donatella Sordi, MD, and Albertina Tiago, MD.