Dermatology Online Journal

A simplified psoriasis area severity index (SPASI) for rating psoriasis severity in clinic patients
B Asher Louden¹, Daniel J Pearce MD¹, Wei Lang PhD², and Steven R Feldman MD PhD^1,2,3
Dermatology Online Journal 10 (2): 7

Departments of Dermatology¹, Public Health Sciences², and Pathology³, Wake Forest University Health Sciences, Winston-Salem, North Carolina.

---

Abstract

Introduction.—The Psoriasis Area and Severity Index (PASI) is the most widely used tool to assess psoriasis disease severity in clinical trials, although it can be exceedingly cumbersome for use in daily clinical practice. Because clinical trials rely on the PASI for inclusion criteria, having a PASI score on a clinic patient may be useful for determining if the patient has a level of disease severity similar to that of patients treated in clinical trials.

Purpose.—The purpose of this study is to assess a simplified measure of psoriasis disease severity that is more conducive to use in general dermatology practice, the simplified PASI (SPASI).

Methods.—We evaluate an area-weighted assessment of lesion severity composed of the sum of the average redness, thickness, and scaliness of all the psoriasis lesions multiplied by an estimate of total body surface area involved. The SPASI is mathematically derived from the PASI. The SPASI and PASI are not identical because of the categorical nature of area estimates used in the PASI. We use existing psoriasis-population data regarding the anatomical distribution of psoriasis lesions to create a simulated patient database. Monte Carlo analysis is then performed to determine the relation between the PASI and SPASI.

Results.—For a sample population with a mean PASI score of 12.8, the mean SPASI was 14.2. Correlation between the PASI and the SPASI was high (r = 0.90). Bland-Altman analysis showed no consistent bias between the PASI and the SPASI. When attempting to identify simulated patients with a PASI score of 12 (an inclusion criterion for many clinical trials for severe psoriasis), SPASI was 97 percent sensitive and 66 percent specific.

Discussion.—The SPASI is much less onerous than the PASI, requiring estimation of only four rather than sixteen independent variables. It provides a quick and practical estimate of disease severity similar to the PASI and can be used to communicate that patients have a level of disease severity similar or dissimilar to that of patients studied in clinical trials.

---

Introduction

Fredriksson and Pettersson created the PASI in 1978 as a method to evaluate the clinical efficacy of a new treatment for psoriasis [1]. When using the PASI, psoriatic plaques are graded based on three criteria: redness (R), thickness (T), and scaliness (S). Severity is rated for each index on a 0-4 scale (0 for no involvement up to 4 for severe involvement). The body is divided into four regions comprising the head (h), upper extremities (u), trunk (t), and lower extremities. In each of these areas, the fraction of total surface area affected is graded on a 0-6 scale (0 for no involvement; up to 6 for greater than 90 % involvement). The various body regions are weighted to reflect their respective proportion of body surface area (BSA). The composite PASI score can then be calculated by multiplying the sum of the individual-severity scores for each region by the weighted area-of-involvement score for that respective region, and then summing the four resulting quantities; mathematically this evaluation is as follows:

PASI = 0.1(R_h +T_h +S_h )A_h + 0.2(R_u +T_u +S_u )A_u
     + 0.3(R_t +T_t +S_t )A_t + 0.4(R_l +T_l +S_l )A_l
Where R_h, R_u, R_t, R_l = redness score of plaques on the head, upper extremities, trunk, and lower extremities, respectively (0-4) T_h, T_u, T_t, T_l = thickness score of plaques on the head, upper extremities, trunk, and lower extremities, respectively (0-4);
S_h, S_u, S_t, S_l = scaliness score of plaques on the head, upper extremities, trunk, and lower extremities, respectively (0-4); and
A_h, A_u, A_t, A_l = area of psoriatic involvement score for the head, upper extremities, trunk, and lower extremities, respectively (0-6).

The highest potential PASI score is 72; the lowest is 0. PASI scores are nearly continuous, with 0.1 increments within these values.

Since the initial use of the PASI, many have critiqued its utility [2, 3]. Although it has limitations, overall it has been a very practical measure of disease severity for clinical trials in patients with severe disease. In addition, many have attempted to revise the PASI and create new and improved methods of disease assessment. Some of the recent scoring systems include, but are not limited to, the self-administered psoriasis area and severity index (Sa-PASI)[4], dermatology index of disease severity [5], evaluation for prognosis with averaged PASI [6], Salford psoriasis index (SPI)[7], and psoriasis assessment severity score [8]. Despite the development of these more recent scales, the PASI score and percentage body surface area were the only measures recommended by an American Academy of Dermatology (AAD) consensus group to assess extent of psoriasis when planning treatment [9].

Although PASI has a defined role in the assessment of psoriasis patients in clinical trials, it is too complicated and time consuming to be practical in most clinical settings. For this reason, we aimed to derive a simplified PASI (SPASI) from the classic PASI model. For the SPASI, the clinician estimates the average redness, thickness, and scaliness of the lesions (on a 0-4 scale) and the area of involvement on a 1-6 scale. We hypothesize that rating psoriasis on the basis of total-body mean redness, thickness, scaliness and global BSA involvement should result in a number similar to PASI scores.

Methods

The PASI provides an area-weighted assessment of the severity of psoriasis. For each of four main body areas, the mean redness, thickness, and scaliness of the psoriasis are estimated and summed. The total is multiplied by an estimate of percentage area of involvement. Then these totals for all four body areas are summed to create the final score.

A simpler way to assess this same quantity would be for the clinician to estimate the mean redness, thickness and scaliness of the psoriasis for the whole body and multiply this by the whole body area of involvement (Fig. 1). That this result is similar to the PASI can be seen mathematically:

(1) PASI = 0.1(R_h+T_h+S_h)A_h + 0.2(R_u+T_u+S_u)A_u + 0.3(R_t+T_t+S_t)A_t + 0.4(R_l+T_l+S_l)A_l
(2) PASI = 0.1R_hA_h + 0.1T_hA_h + 0.1S_hA_h + 0.2R_uA_u + 0.2T_uA_u + 0.2S_uA_u + 0.3R_tA_t + 0.3T_tA_t + 0.3S_tT_t + 0.4R_lA_l + 0.4T_lA_l + 0.4S_lA_l
(3) PASI = (0.1R_hA_h +0.2R_uA_u + 0.3R_tA_t + 0.4R_lA_l) + (0.1T_hA_h + 0.2T_uA_u + 0.3T_tA_t + 0.4T_lA_l) + (0.1S_hA_h + 0.2S_uA_u + 0.3S_tA_t + 0.4S_lA_l)
(4) 0.1A_h + 0.2A_u + 0.3A_t + 0.4A_l = BSA involved.
(5) Multiplying and dividing each term in the step 3 above by the affected area
(4) yields: PASI = [(0.1R_hA_h + 0.2R_uA_u + 0.3R_tA_t + 0.4R_lA_l) _ (0.1A_h + 0.2A_u + 0.3A_t + 0.4A_l) + (0.1T_hA_h + 0.2T_uA_u + 0.3T_tA_t + 0.4T_lA_l) _ (0.1A_h + 0.2A_u + 0.3A_t + 0.4A_l) + (0.1S_hA_h + 0.2S_uA_u + 0.3S_tA_t + 0.4S_lA_l) _ (0.1A_h + 0.2A_u + 0.3A_t + 0.4A_l)] x BSA involved
(6) Mean redness, thickness, and scaliness for the whole body are given by the following: Mean redness = (0.1R_hA_h + 0.2R_uA_u + 0.3R_tA_t + 0.4R_lA_l) _ (0.1A_h + 0.2A_u +0.3A_t + 0.4A_l), Mean thickness = (0.1T_hA_h + 0.2T_uA_u + 0.3T_tA_t + 0.4T_lA_l) _ (0.1A_h + 0.2A_u +0.3A_t + 0.4A_l) Mean scaliness = (0.1S_hA_h + 0.2S_uA_u + 0.3S_tA_t + 0.4S_lA_l) _ (0.1A_h + 0.2A_u +0.3A_t + 0.4A_l)
(7) therefore PASI = (mean redness + mean thickness + mean scaliness) x BSA involved ≈ SPASI

The SPASI is only approximately the same as the PASI because the PASI utilizes area scores that are evaluated categorically (0-6 scale) and not continuously. Thus, we cannot assume that (0.1A_h + 0.2A_u + 0.3A_t + 0.4A_l) = total BSA involved measured on the same categorical scale. For example, if there is involvement only in the lower extremities, and it is less that 10 percent of the lower extremities, the PASI area in the calculation will be 0.4A_l = 0.4x1 = 0.4; whereas the SPASI area score would be less than 10 percent BSA = 1.0. To adjust for this difference, we looked at SPASI using two separate area score scales: one using the classic PASI area score, and one utilizing an additional area score category.

Monte Carlo analysis was used to assess how close SPASI scores would match PASI scores. This technique performs a simulation many times in order to calculate a distribution of likely results [10]. Limitations of such an analysis is that it lacks external validity and that the SPASI would need to be repeated with real patients. Three distinct sample patient populations (N = 1000 for each) were created using Excel 2000 (Microsoft; Redmond, Wash.). For each patient in each population, a random number generator was used to assign a score from 0 to 4 for redness, thickness, and scaliness in each of the four anatomic regions (0 for no involvement, 1 for mild, 2 for moderate, 3 for severe, and 4 for most severe). Regional BSA involvement was assigned for each anatomic region used in the PASI formulation. Regional involvement in population 1 was assigned using a random number generator (range 0-75) while accounting for population statistical data. In population 2, regional involvement was assigned randomly from 0 to 100 while also accounting for population statistical data. In population 3, involvement was assigned randomly from 0 to 100. These random numbers representing percent involvement by region were then converted to categorical scores using the PASI scoring system: 0 for no involvement, 1 for up to 10 percent, 2 for 10-30 percent, 3 for 30-50 percent, 4 for 50-70 percent, 5 for 70-90 percent, and 6 for greater than 90 percent. Sample calculations to illustrate the components of these measures is presented in Figure 1.

To account for population statistical data in sample populations 1 and 2, we utilized data presented in Fleischer et al. on the anatomic distribution of psoriatic skin involvement [11]. In accordance with the sample population of live patients in Fleischer et al., populations 1 and 2 were created such that 75 percent of these sample patients had psoriasis on the head, either the face or scalp, 69 percent had involvement of the upper extremities, 60 percent had involvement of the trunk, and 78 percent had involvement of the lower extremities. To incorporate this data in the model, we used the RAND Excel command to generate a random number (range 0-1) for each patient 1-1000. We then used the RANK command to convert these random numbers into unique integers ranging from 1 to 1000. Then, when assigning regional involvement of the head, for example, if the unique random number for a patient were less than 250, the regional percent involvement score would be assigned a 0. If the unique random number was greater than 250, the regional involvement score would be assigned randomly as designated for that sample population, 0-75 for population 1 and 0-100 for population 2. The unique random number thresholds when assigning involvement for the upper extremities, trunk, and lower extremities were 310, 400, and 220 respectively.

Calculations for SPASI were completed according to the assumptions given in the mathematical derivation above. Values for mean redness, thickness, scaliness, and percent involvement were rounded to the nearest whole number to best approximate use in a clinical scenario. The categorical scores for percent involvement were used in the final calculation step when multiplying the sum of the severity indices by this involvement score. Bland-Altman analysis was used to determine the agreement between SPASI and PASI [12].

For each of the three populations, a second iteration of the model was performed. An additional 0.5 score for area of involvement was added only in the 0-1 range; this score was assigned for 0-5 percent. A score of 1 was assigned for involvement from 5 to 10 percent. All other values remained the same. This assignment is subsequently referred to as area scoring system 2.

Results

The resulting PASI and SPASI scores would be as follows:

Calculations		PASI				SPASI
		Head	Upper extremities	Trunk	Lower extremities	Total body
1	Redness	2	2	2	2	2
2	Thickness	2	2	2	2	2
3	Scaliness	2	2	2	2	2
4	Sum of rows 1, 2 and 3	6	6	6	6	6
5	Affected area (%)	5	12	35	20	21.4
6	Area Score	1	2	3	2	2
7	Area weights	0.1	0.2	0.3	0.4	-
8	Row 4 * row 6 * row 7	0.6	2.4	5.4	4.8	-
9	PASI (Sum of row 8)	13.2				12

The mean PASI score in population 1 was 12.8 (Table 1). Mean SPASI using the classic area scoring system was 14.2. Mean SPASI using area scoring system 2 was 14.1. Correlation between the PASI and the SPASI was high (population 1, r = 0.90; population 2, r = 0.93; population 3, r = 0.93, Fig. 2). Bland-Altman analysis showed good agreement between the two methods of assessing psoriasis disease severity. With increasing severity of the sample population, agreement between the PASI and SPASI improves (table 1).

The absolute difference between PASI and SPASI scores was largely independent of disease severity (Fig. 3), with a mean difference between the PASI and the SPASI of -1.64 for population 1. The percentage difference between the PASI and SPASI was higher for patients with low areas of involvement (low PASI scores) and decreased with increasing PASI score (Fig. 4).

We also looked at degree of correlation for simulated patients with PASI scores greater than 12 because this is an inclusion criterion frequently used in clinical studies of patients with severe psoriasis. In population 1 (mean PASI 12.8), there was an 82 percent concordance for scores either both greater than or both less than 12. For the most severe population, population 3, there was 96 percent concordance.

Discussion

Despite its limitations, the PASI has been an extraordinarily useful tool. Having a quantitative measure of disease severity has facilitated clinical trials of new agents for psoriasis. PASI has led to new treatments that have advanced our ability to care for patients with severe disease. The development of these agents, some of which set new standards of high cost of dermatologic treatment, has led to new challenges as well. Third-party payers increasingly desire objective standards to determine who is a candidate for these treatments. The AAD and the National Psoriasis Foundation have attempted to answer this challenge [9, 13]. These organizations propose that the decision to implement systemic therapy ultimately be left to the judgment of the dermatologist, based on their assessment of the extent of disease and its impact on patients' lives. We agree with this approach.

Nevertheless, payers wish to be assured that only appropriate candidates are treated with expensive systemic treatment regimens. Such assurances may frequently be required by insurers. Dermatologists prescribing a new, expensive systemic agent may be able to provide objective support for their decision in part by demonstrating that the patient's disease severity is within the range of patients who had been treated with the new agent in clinical trials. Providing the patient's PASI score can do this, but the PASI is often too cumbersome to use in clinical practice. In all likelihood, the busy clinician roughly estimates, rather than vigorously calculates, the PASI.

For purely practical reasons, we have attempted to create a simplified PASI to address this need. In attempting to simplify PASI, we sought to remain within the framework of the current model using one area of involvement and three disease severity dimensions. In addition, because patients have been assessed with PASI for approximately 25 years, complete abandonment of PASI would result in a loss of correlation between past and future measures of disease severity. Our SPASI is conceptually identical and mathematically very similar to the original PASI score.

One limitation of the SPASI is it is relatively insensitive to change when there is less than 10 percent body-surface-area involvement or when disease is localized to one region; this, however, is not a major drawback when the instrument is used to assess patients with severe psoriasis. To improve the SPASI, we added a 0.5 area score for areas greater than 5 percent. We could further increase the sensitivity of the SPASI in patients with such mild to moderate disease by breaking up the more than 10 percent category further (as well could be done for the PASI). This adjustment would add to the complexity of the model and is not pertinent as the practical use of both the PASI and the SPASI is primarily for patients with extensive disease. The use of the SPASI, like the PASI, is not appropriate as a criterion for disabling disease involving less than 10 percent BSA (i.e., palmoplantar psoriasis), as such patients may have markedly diminished quality of life despite only very low PASI scores.

Another theoretical limitation of the SPASI is that it assumes a clinician can estimate the average redness, scaliness, and thickness of lesions for the entire body. How well dermatologists are capable of doing this is to some extent an open question. That a dermatologist can successfully integrate the severity components of lesions over an area and estimate a mean severity is not a new assumption, however. The PASI itself also requires this thoroughness, although the integration is done separately for each of four body areas. To the extent that the PASI has demonstrated validity, the ability of dermatologists to assess mean severity appears reasonable. Moreover, assessments of mean redness, thickness, and scaliness of lesions over the whole body are components of the self-administered PASI. This validated measure is performed by untrained patients and correlates with PASI scores in diverse psoriasis populations [14, 15, 16]. Proper validation of the SPASI in a real patient population is needed to support potential use as a surrogate or replacement for the PASI. Theoretically, however, the SPASI could provide an approximation of the PASI and may serve a much needed function in the treatment of psoriatics.

The Center for Dermatology Research is funded by a grant from Galderma Laboratories, L.P. There are no conflicts of interest to disclose.

References

1. Fredriksson T, Pettersson U. Severe psoriasis--oral therapy with a new retinoid. Dermatologica 1978; 157(4):238-244.

2. Marks R, Barton SP, Shuttleworth D, Finlay AY. Assessment of disease progress in psoriasis. Arch Dermatol 1989; 125(2):235-240.

3. Van de Kerkhof PC. On the limitations of the psoriasis area and severity index (PASI). Br J Dermatol 1992; 126(2):205.

4. Fleischer AB, Jr., Rapp SR, Reboussin DM, Vanarthos JC, Feldman SR. Patient measurement of psoriasis disease severity with a structured instrument. J Invest Dermatol 1994; 102:697-699.

5. Faust HB, Gonin R, Chuang TY, Lewis CW, Melfi CA, Farmer ER. Reliability testing of the dermatology index of disease severity (DIDS). An index for staging the severity of cutaneous inflammatory disease. Arch Dermatol 1997; 133(11):1443-1448.

6. Sugai J, Ozawa A, Kawakubo Y, Iizuka M, Miyahara M, Ohkido M. New method for determining prognosis of patients with psoriasis (E-PAP). J Dermatol Sci 1998; 16(2):165-169.

7. Kirby B, Fortune DG, Bhushan M, Chalmers RJ, Griffiths CE. The Salford Psoriasis Index: an holistic measure of psoriasis severity. Br J Dermatol 2000; 142(4):728-732.

8. Harari M, Shani J, Hristakieva E, Stanimirovic A, Seidl W, Burdo A. Clinical evaluation of a more rapid and sensitive Psoriasis Assessment Severity Score (PASS), and its comparison with the classic method of Psoriasis Area and Severity Index (PASI), before and after climatotherapy at the Dead-Sea. Int J Dermatol 2000; 39(12):913-918.

9. Callen JP, Krueger GG, Lebwohl M, McBurney EI, Mease P, Menter A, Paller AS, Pariser DM, Weinblatt M, Zimmerman G. AAD consensus statement on psoriasis therapies. J Am Acad Dermatol 2003; 49(5):897-899.

10. Mochal T. Use Monte Carlo analysis for sophisticated scheduling. Tech Republic . 7-29-2002. 11-22-0003. Ref Type: Electronic Citation

11. Fleischer AB, Jr., Feldman SR, Rapp SR, Reboussin DM, Exum, ML, Clark AR, Rajashekhar V. Disease severity measures in a population of psoriasis patients: the symptoms of psoriasis correlate with self-administered psoriasis area severity index scores. J Invest Dermatol 1996; 107(1):26-29.

12. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1(8476):307-310.

13. National Psoriasis Foundation. When Are Patients Candidates for Phototherapy or Systemic Treatments (Including Biologics)? National Psoriasis Foundation . 1-1-2003. 11-22-0003. Ref Type: Electronic Citation

14. Feldman SR, Fleischer AB, Jr., Reboussin DM, Rapp SR, Exum, ML, Clark AR, Nurre L. The self-administered psoriasis area and severity index is valid and reliable. J Invest Dermatol 1996; 106(1):183-186.

15. Fleischer AB, Jr., Feldman SR, Dekle CL. The SAPASI is valid and responsive to psoriasis disease severity changes in a multi-center clinical trial. J Dermatol 1999; 26(4):210-215.

16. Sampogna F, Sera F, Mazzotti E, Pasquini P, Picardi A, Abeni D. 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. The psoriasis area and severity index (PASI). Br J Dermatol 1992; 126(2):205.

© 2004 Dermatology Online Journal