Positive and Negative Syndrome Scale (PANSS): The PANSS was developed as a 30-item rating scale, which adapted 18 items from the Brief Psychiatric Rating Scale (BPRS) and 12 items from the Psychopathology Rating Schedule (PRS).¹¹⁹ The PANSS requires a 30 to 40 minute patient interview to gather information on which to assess the patient with regard to the presence and severity of psychopathology in the previous week. The PANSS instrument provides a complete definition of each item as well as detailed anchoring criteria for each of seven rating points: 1 = absent, 2 = minimal, 3 = mild, 4 = moderate, 5 = moderate–severe, 6 = severe, 7 = extreme. In the 30-item scale, seven items related to positive symptoms, seven items to negative symptoms, and 16 items to general psychopathology (as shown below). Finally, a composite scale may be derived by subtracting the negative score from the positive score. The 30 items of the PANSS are presented below:¹²⁰

Positive Scale

• P1. Delusions
  P2. Conceptual disorganization
  P3. Hallucinatory behaviour
  P4. Excitement
  P5. Grandiosity
  P6. Suspiciousness
  P7. Hostility

Negative Scale

• N1. Blunted affect
  N2. Emotional withdrawal
  N3. Poor rapport
  N4. Passive/apathetic social withdrawal
  NS. Difficulty in abstract thinking
  N6. Lack of spontaneity and flow of conversation
  N7. Stereotyped thinking

General Psychopathology Scale

• G1. Somatic concern
  G2. Anxiety
  G3. Guilt feelings
  G4. Tension
  G5. Mannerisms & posturing
  G6. Depression
  G7. Motor retardation
  G8. Uncooperativeness
  G9. Unusual thought content
  G10. Disorientation
  G11. Poor attention
  G12. Lack of judgment and insight
  G13. Disturbance of volition
  G14. Poor impulse control
  G15. Preoccupation
  G16. Active social avoidance

Kay et al. reported on psychometric testing of the PANSS in 101 in-patients with schizophrenia.¹¹⁹ Scores on all subscales were reported to exhibit a normal distribution, suggesting suitability for parametric statistical analysis. Further, the range of scores was less than the potential range, suggesting the lack of a ceiling effect. Internal consistency was demonstrated for the positive (α = 0.73), negative (α = 0.83), and general psychopathology (α = 0.79) subscales. Test-retest reliability was assessed three to six months later on a cohort of 15 patients who remained hospitalized; Pearson correlation coefficients were 0.80, 0.68, and 0.60 for the positive, negative, and general psychopathology subscales, respectively.¹¹ Peralta and Cuesta reported on the inter-rater reliability of the PANSS from a sample of 100 consecutively admitted patients with schizophrenia.¹²¹ Positive and negative scales showed good inter-rater reliability: interclass correlation coefficients (ICC) of 0.72 and 0.80, respectively. Inter-rater reliability was moderate for the general psychopathology scale; ICC = 0.56.

More recently, a number of investigators have conducted principal components analysis to expand the identification of discrete dimensions of schizophrenia beyond the focus on positive and negative symptoms. A number of similar five-factor models, including most or all of the original PANSS items, have been proposed and tested for reliability and validity.^122–126 One such model was proposed by Marder et al. and categorizes all original PANSS items into five dimensions: positive symptoms (eight items), negative symptoms (seven items), disorganized thought (seven items), uncontrolled hostility/excitement (four items), and anxiety/depression (four items).¹²⁶

It is unclear what degree of improvement in the PANSS total or subscale scores is clinically important. The relationship between change in PANSS score and long-term clinical outcomes has not been clearly identified;¹¹⁴ however, a 20% reduction of PANSS has been used as the criterion of response to antipsychotic treatment in many clinical trials.^36,39,42 In a comparison of PANSS to the Clinical Global Impression (CGI) tool, it is suggested that an absolute reduction of 15 in the total PANSS score corresponds to “minimally improved” on the CGI—Improvement (CGI-I) scale, and a reduction of the CGI—Severity (CGI-S) score by one severity step.¹²⁷ A reduction of 33 in the total PANSS score corresponds to “much improved” on the CGI-I scale. The above estimates are sensitive to baseline severity of illness to the extent that participants with a lower baseline severity of illness required smaller reductions in the PANSS to produce a particular improvement in the CGI. For this reason, it has been suggested that change in PANSS score has limited usefulness as a primary outcome, due to variability in baseline symptom intensity.^128,129 Instead, a standardized remission criterion that may be suitable for use in clinical practice and clinical trials has been proposed. Specifically, a score of ≤ 3 on all eight PANSS items (P1, P2, P3, N1, N4, N6, G5, and G9) for a period of at least six months is considered to represent remission of disease.^128,129

Brief Psychiatric Rating Scale (BPRS): Consists of 24 symptom constructs, each rated on a seven-point scale of severity ranging from “not present” to “extremely severe” (1 = not present; 2 = very mild; 3 = mild; 4 = moderate; 5 = moderately severe; 6 = severe; 7 = extremely severe). Research suggests that a reduction of at least 10 points correlates with “minimally improved” on the CGI-I and to a change in CGI-S score of one severity step.¹²⁷ Similar to PANSS, the clinical implications of BPRS change are not well established, although 20% reduction of BPRS was used as a criterion of response to antipsychotic treatment in many clinical trials.^46,51,65

Clinical Global Impressions (CGI): The CGI scale is a three-item scale used to assess overall severity and response to treatment of mental disorders. It is not specific to schizophrenia, although efforts to adapt the scale specifically to this disorder have been undertaken.¹³⁰ The three scale items include severity of illness (CGI-S), overall improvement (CGI-I), and an efficacy index. CGI-S is a seven-point scale that indicates how mentally ill the patient is at a given time: 1 = normal, not at all ill; 2 = borderline mentally ill; 3 = mildly ill; 4 = moderately ill; 5 = markedly ill; 6 = severely ill; 7 = among the most extremely ill patients.¹³¹ CGI-I is a seven-point scale that indicates symptom improvement after the treatment: 1 = very much improved; 2 = much improved; 3 = minimally improved; 4 = no change from baseline; 5 = minimally worse; 6 = much worse; 7 = very much worse).¹³¹ CGI-I is only measured after treatment. A CGI-I score of 1 or 2 was used as a response criterion in some trials.^36,40 The efficacy index incorporates the clinician’s assessment of therapeutic effect in relation to adverse events. The difficulty of combining the two concepts of efficacy and adverse events has led to criticism of this last item.¹³⁰ However, there is no total score for the CGI; rather, scores on the individual items are considered separately.

As the CGI can be administered quickly, it is well suited to clinical settings; however, there is little information regarding its reliability or validity. Rabinowitz et al. sought to validate the CGI-S via a comparison of PANSS and CGI-S scores from seven trials of risperidone in schizophrenia.¹³² CGI-S scores from the pooled trials corresponded to the following mean PANSS scores: 1 (normal) = PANSS 55.5; 2 (borderline ill) = PANSS 67.0; 3 (mildly ill) = PANSS 79.6; 4 (moderately ill) = PANSS 92.4; and 5 (markedly ill) = PANSS 99.7. Predefined measures of clinical improvement were a 20% reduction in the PANSS score and a 1-point decrease on the CGI-S. The sensitivities and specificities for the CGI-S to detect this level of improvement in the seven trials ranged from 64.5% to 89.6% and 65.7% to 82.8%, respectively. From this assessment, it appears that the CGI-S and PANSS are correlated and exhibit substantial agreement in detecting change.

Barnes Akathisia Rating Scale (BARS): The BARS is the most commonly used scale to measure antipsychotic-induced akathisia in clinical trials.¹³³ The BARS is a four-item scale that scores patients’ akathisia based on (i) brief observation by the clinician (ranked 0 to 3); (ii) patient report of awareness of restlessness (ranked 0 to 3); (iii) patient report of distress related to restlessness (ranked 0 to 3), which produces (iv) a global clinical assessment of akathisia.¹³⁴ The global clinical assessment contains five well-defined severity categories that are considered clinically relevant: 0 = absent; 1 = questionable; 2 = mild; 3 = moderate; 4 = marked; 5 = severe.¹³³ Inter-rater reliability for all four items, based on duplicate rating of 42 chronic in-patients and measured by Cohen’s kappa, were observation (0.74), awareness (0.83), distress (0.90), and global clinical assessment (0.96).¹³⁴ The BARS has been reported to correlate only weakly with motor activity measured by actometry, potentially due to the fact that actometry measures only actual movement, while the BARS also measures the subjective experience of awareness and distress.¹³⁵

Abnormal Involuntary Movement Scale (AIMS): The AIMS is a 12-item scale for assessing dyskinesias completed by the clinician or researcher. The first seven items pertain to abnormal movements in three specific anatomical sites: facial and oral movements (four items), extremity movements (two items), and trunk movements (one item).¹³⁶ The remaining items are global assessments (three items, including global severity, incapacitation, and patient awareness), and two items specific to dentition. Except for the items related to dentition, items are scored on a five-point scale; none (0), normal (1), minimal (2), mild (3), moderate (4), or severe (5). Inter-rater reliability in a sample of 38 outpatients with a history of dyskinesia was reported to be high; Pearson correlation coefficient = 0.87 for all items except those related to dentition.¹³⁷ However, inter-rater reliability is reported to be higher among experienced raters.¹³⁸ The validity of the AIMS has been established via comparisons to other similar instruments; the Extrapyramidal Symptom Rating Scale (ESRS) and the Schedule for the Assessment of Drug-Induced Movement Disorders (SADIMoD).^139,140 Gharabawi et al. examined associations between individually related and overall severity scores from the AIMS and ESRS via logistic regression.¹³⁹ R² values ranged from 0.30 (trunk movements) and 0.67 (lips and perioral area); R² value was 0.56 for global severity. Loonen examined associations between (i) total AIMS scores, (ii) total items excluding global and dental items, and (iii) four facial and oral movement items.¹⁴⁰ Spearman’s correlation coefficients between the active global dyskinesia subscale of the SADIMoD and the above AIMS scores were 0.76, 0.82, and 0.83, respectively. It is unclear what would constitute a meaningful change in the AIMS. However, the presence of tardive dyskinesia is accepted based on a rating of mild in two or more anatomical areas, or moderate or greater symptoms in one or more anatomical areas.^139,141,142

Simpson-Angus Scale (SAS): The Simpson-Angus Scale was developed in the 1960s to identify neuroleptic-induced parkinsonism. The scale contains 10 items: one measuring gait, six measuring rigidity, and three measuring glabella tap, tremor, and salivation.¹⁴³ Each item is scored on a five-point scale from 0 (complete absence) to 4 (extreme), and a total score is obtained by adding all item scores and dividing by 10 (the total number of items). Scores of up to 0.3 were considered to be within the normal range; however, it has recently been suggested that the upper limit of normal be raised to 0.65.¹⁴³ Inter-rater reliability of the SAS between two physicians in a trial of haloperidol containing 14 participants was determined; a correlation coefficient of 0.87 was reported.¹⁴⁴ In this same trial, SAS scores were significantly higher for participants treated with haloperidol compared with placebo, supporting the discriminant validity of the SAS.