The Physician Global Assessment (PGA) is a single-item, physician-reported measure of overall plaque psoriasis severity. It was introduced in 1998 by a US Food and Drug Administration advisory panel as the preferred global measure for psoriasis trials and has since been used as a co-primary endpoint, alongside the PASI, in the registration programmes of most biologic and small-molecule therapies. The static form of the scale (sPGA), which grades the disease state at a single point in time without reference to previous visits, has become the standard; it is also referred to as the Investigator Global Assessment (IGA).

Several versions of the scale exist. The most widely used today is the 5-point scale, graded 0 (clear), 1 (almost clear), 2 (mild), 3 (moderate), and 4 (severe), in which the assessor selects the single grade whose descriptors for erythema, plaque thickening, and scaling best match the overall appearance of the lesions. The descriptor anchors implemented here follow the 5-point IGA described by Langley et al. (2015), which was developed with regulatory input and applies a stringent definition of almost clear (normal to pink coloration, no thickening, and no to minimal focal scaling). Nail involvement is not part of the assessment. Pascoe et al. (2015) demonstrated that a 5-point PGA of this type can be recorded routinely at every visit in a busy clinical setting and used to track outcomes over time.

An alternative, component-based 6-point PGA was used in the ustekinumab PHOENIX trials and related programmes. Induration, erythema, and scaling are each graded from 0 to 5 according to the average appearance of all lesions, the three grades are averaged, and the mean is rounded to the nearest whole number to give a final PGA of 0 (cleared) to 5 (severe). A multi-item PGA of this kind was validated by Cappelleri et al. (2013), with good test-retest reliability, internal consistency, and correlation with PASI. Because the labels and thresholds differ between the two versions, the version used should always be stated when a PGA score is reported.

In clinical trials, treatment success is conventionally defined as a PGA score of 0 or 1 (clear or almost clear) together with an improvement of at least 2 points from baseline. Robinson et al. (2012) showed that PGA 0/1 response rates and PASI 75 response rates are highly correlated across biologic trials, and suggested that the simpler PGA is well suited to community-based outcome measurement. The product of the PGA and the percentage of body surface area involved (PGA x BSA), proposed by Walsh et al. (2013), correlates strongly with PASI (r = 0.87) while remaining quick to record, and has been proposed as a practical composite for routine care. The PGA does not capture the extent of disease on its own, and the reliability of the simple 5-point scale has been less thoroughly characterised than that of the PASI.