Global Assessment of Functioning Scale

Click Image to Download PDF

Under the new permanent disability schedule effective January 1, 2005, “Psychiatric impairment shall be evaluated by the physician using the Global Assessment of Function (GAF) 1.” This represents a significant change from the way permanent psychiatric disability had been rated in the past for Workers’ Compensation psychiatric claims in California. Previously, psychiatric disability was rated using the 8 Work Functions of permanent disability.

Although other body parts are evaluated and rated using the American Medical Association (AMA) Guides to the Evaluation of Permanent Impairment, 5th Edition 2, (hereinafter, “the Guides”), psychiatric impairment is not addressed in a similar manner. According to the Guides, “Percentages are not provided to estimate mental impairment in this edition of the Guides. Unlike cases with some organ systems, there are no precise measures of impairment in mental disorders. The use of percentages implies a certainty that does not exist.” The authors of the Guides note that percentages are likely to be used inflexibly by adjudicators, who then are less likely to take into account the many factors that influence mental and behavioral impairment. “After considering this difficult matter, the Committee on Disability and Rehabilitation of the American Psychiatric Association advised Guides contributors against the use of the percentages in the chapter on mental and behavioral disorders of the fourth edition, and that remains the opinion of the authors of the present chapter (p. 361).”

As a result, an alternative method for assessing permanent psychiatric impairment was selected, The Global Assessment of Functioning Scale. This scale is found on Axis V of the Multiaxial Assessment system advanced by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition 3. Each axis, of the five axis system, was intended to provide distinct and valuable information about the mental disorder and functioning level of a patient. Axis I focused on clinical disorders, Axis II, personality disorders and mental retardation, Axis III, general medical conditions, Axis IV, psychosocial and environmental problems and Axis V, global assessment of functioning using the GAF. The GAF is based on the assumption that the level of current functioning in psychiatric populations can provide meaningful information about treatment needs and planning, and serve as a reliable predictor of outcome status and social and occupational functioning.

The GAF is a 100 point scale that ranges from 1 (persistent danger of severely hurting self or others, persistent inability to maintain personal hygiene) to 100 (superior functioning in a wide range of activities). There are ten 10 point categories. A score of zero reflects inadequate information. There are two central sources of information that are evaluated by the clinician. The first concerns psychiatric symptoms (e.g., depression, sleep impairment, auditory or visual hallucinations) and the second concerns social/leisure and occupational functioning. (School functioning is also included but will not be discussed here given that it is generally not the focus of impairment in the Workers’ Compensation system). Both sources of information need to be evaluated in deriving the final GAF score. A single GAF score is then produced which serves as the final rating.

A central weakness concerning the GAF is that it asks the physician to rate and incorporate concepts that do not necessarily covary. That is, an individual may have legitimate residual mild psychiatric symptoms but may be functioning well in their employment setting and working full-time. Research suggests that physicians generally rate the psychiatric symptom portion of the GAF accurately. However, physicians often fail to adequately assess social and occupational functioning (Hilsenroth et al., 2000) 4. Clearly, social and occupational functioning should be adequately assessed in order to produce a valid and reliable GAF score.

Moreover, according to the DSM-IV-TR 5, the physician is required to take the lower of the ratings between psychiatric symptoms and functional impairment. The DSM-IV-TR states, “It should be noted that in situations where the individual’s symptom severity and level of functioning are discordant, the final GAF rating always reflects the worse of the two (p. 45).” For example, an injured worker may have developed an industrially related depressive condition with bonafide mild symptoms. This individual may have received psychiatric and psychological treatment and returned to work full-time without restrictions. The injured worker’s depression may persist at a mild level but it may not interfere with their ability to perform their job duties because they may have worked in the same position for 20 years and have a good social support network at home and at the office. According to the guidelines, the physician would still need to take the lower of the two ratings in determining the final GAF score for this patient.

The issue of the reliability and accuracy of GAF ratings by mental health professionals has been examined in the literature. Hilsenroth et al., 2004 found that mental health practitioners are capable of yielding good inter-rater reliability coefficients (r = .86). This suggests that clinicians reviewing the same data and material can reliably produce similar GAF scores.

However, other research has indicated significant problems with physician ratings on the GAF. Bates et al., (2002) 6 examined the impact of training on GAF ratings by clinicians. They presented two vignettes to 31 mental health professionals at a VA Medical Center. The clinicians first rated the two vignettes without training and then rated them again after a GAF training session. These vignettes were rated independently prior to the study so that an objective criterion could be used. They found that mental health practitioners used numerous incorrect strategies to rate the GAF including rating the scale based on the highest or lowest level of functioning, the average level of functioning, the least severe symptoms, the most severe symptoms and the worst of the combined symptoms which is the correct strategy. They found that 90% of the mental health practitioners used incorrect strategies and only 10% used the correct strategy. The most common strategy was to average symptom severity and functioning. Following the GAF training, 64% of clinicians used the correct rating strategy; however, 34% continued to use an incorrect strategy. Most importantly, the studies revealed that the incorrect ratings resulted in an 8-19 point increase in the final GAF score relative to the predetermined ratings.

Research in general has suggested that diagnostic accuracy improves when decision tree, actuarial models are utilized (Meehl, 1954) 7. Research examining the reliability of ratings using a computer based decision tree, GAF model, indicates that clinicians yield more reliable scores than do more traditional decision making approaches, particularly if the information to be analyzed is complex and comprehensive, as is often seen in Workers’ Compensation psychiatric claims (Woldoff, 2004) 8.

One actuarial decision tree method for rating the GAF was developed by Michael First, M.D. and Multi-Health Systems (2001) 9. Dr. First served as the Editor for the DSM-IV Text and Criteria as was the principle editor for the GAF as it is cited in both the DSM-IV and the DSM-IV-TR. As such, it is this author’s opinion that the actuarial model prepared by Dr. First can significantly increase the reliability and validity of ratings generated by mental health practitioners. This is based on the assumption that an actuarial method is superior to a nonactuarial approach, and that the principle editor of the scale being evaluated was the professional who developed the decision tree criteria. It is on this basis that this author selected the actuarial analysis (described below) for determining GAF scores. However, other actuarial, decision tree models are available in the literature (Yamauchi et al., 2001) 10.

The decision tree model developed by Dr. First recommends several distinct steps. It is recommended that psychiatric symptoms are assessed first and given a GAF score. Then the patient’s social and occupational functioning are assessed and a final GAF score is determined. Dr. First’s actuarial model recommends the following steps: First, the clinician is asked to assess whether there are dangerous or psychotic symptoms present. Five main areas are examined in this regard. (1) Danger to self or other (suicidal ideation or intent), (2) violence and aggression, (3), impaired judgment, (4) disturbed communication and (5) impaired reality testing (e.g., auditory or visual hallucinations). In essence the determination of the severity of psychiatric symptoms begins at the bottom of the scale (most severe) and works up the scale toward least severe symptomatology. The determination of each 10 point category reflects the extent and persistence of the symptomatology. For example, a severely depressed patient who is in persistent danger of severely hurting themselves would receive a GAF score between 1 and 10. If the patient was in some danger of intentionally hurting themselves the GAF would be between 11 and 20 and if the patient had minimal to no danger of hurting themselves but was preoccupied with suicidal thoughts the GAF would be between 21 and 30. Less and less preoccupation and clinical distress would lead to higher GAF scores across all the dangerous and psychotic categories.

It is important to note that passive suicidal ideation should not in and of itself produce significantly low GAF scores. It is quite common for mild to moderate depressed patients to report that, “Sometimes I think I would be better off if I was gone or with G-d” or “I think about suicide at times but I would never do it because I am a religious person and I have children.” Although these statements clearly need to be explored by the mental health specialist, the absence of distress or clinical behaviors suggestive of active suicidal ideation would suggest that such thoughts are passive and not consistent with a GAF score below 50. Ideally, a close review of the medical record would assist in determining the severity of the suicidal ideation. It would be important to know if the patient had been reporting active suicidal ideation to other physicians or disciplines. Was the patient considered a high-risk suicide patient and if so what suicide protocols were initiated by the treating physician? Was the patient psychiatrically hospitalized? Did they require a more intensive psychiatric intervention (i.e., an increase in psychotropic medications, psychotherapy or both)? Supportive collateral data should be present to justify the rendering of a low GAF score based on complaints of suicide. A similar analysis could be performed with patients who allege psychotic ideation or violence.

It is also important to note that a patient who is reporting active suicidal ideation or has an active psychotic process in likely to not be stable and may therefore would not be permanent and stationary (i.e., reached maximal medical improvement). Additional treatment protocols may be required before maximal medical improvement is achieved. Attorneys and adjusters should remember that a GAF score is often reported on Axis V of a psychiatric report, regardless of whether the patient has reached maximal medical improvement. Clearly, only a GAF score from a permanent and stationary report would be used in determining permanent psychiatric impairment.

The next phase of the assessment is to examine nondangerous, nonpsychotic symptoms. Symptoms associated with the previous categories would generally be associated with a GAF score below 40. The categories associated with nondangerous and nonpsychotic symptoms would generally be related to GAF scores between 41-70. The breakdown of the categories is as a follows: Severe symptoms: 41-50, Moderate symptoms: 51-60, Mild symptoms: 61-70. Within the current Workers’ Compensation rating system in California GAF scores of 70 or higher do not yield any permanent disability ratings. This is because GAF scores above 70 are typically considered to be transient and/or expected reactions to psychosocial stressors.

The DSM-IV-TR defines “Severe” as “Many symptoms in excess of those required to make the diagnosis, or several symptoms that are particularly severe, are present, or the symptoms result in marked impairment in social or occupational functioning.” Moderate symptoms are defined as “Symptoms or functional impairment between ‘mild’ and ‘severe’ are present”, and mild symptoms are defined as “Few, if any, symptoms in excess of those required to make the diagnosis are present, and symptoms result in no more than minor impairment in social and occupational functioning”.

Thus, for example, a patient with a chronic back injury with a persistent mild sadness, depression with slightly diminished sleep, appetite and libido would likely have a GAF symptom rating score in the 61-70 range. If the depression was at a moderate level of severity, the GAF would fall somewhere between 51 to 60 and so forth. It is also important to note that the final determination of the symptom severity portion of the rating will depend on the mix of current symptoms. For example, if a patient has moderate depressed mood as defined by clinical presentation and psychological test data (i.e., MMPI-2) but reports mild problems in sleep, appetite, libido, etc., then the patient’s GAF would be in the moderate symptom category but perhaps at the higher end of the moderate range (e.g., GAF = 58).

Once a GAF rating score has been established for the symptom severity, the next phase of the analysis is to assess the patient’s social and occupational functioning. It is important to note that the Final GAF score will not be any higher than the symptom severity score. It can only be lower. For example, if the patient’s GAF score for the symptom severity portion of the assessment is a 58 and an analysis of their functioning suggests a GAF of 62, the final GAF score would be a 58 because, as noted above, you are required to take the lower of the two ratings. Conversely, if the patient’s symptom severity suggests a GAF score of 62 and their social and occupational functioning indicates a GAF of 55, then the final GAF score will be a 55. Thus, the symptom severity GAF score will serve to represent the highest score the GAF can be. The functional assessment score can only lower the final GAF score.

Just as with the symptom severity assessment, the mental health professional begins at the bottom of the scale and focuses on the most significant functional disturbance first. The lowest category concerns impairment in basic hygiene skills. The presence of persistent deficits in one’s ability to maintain personal hygiene would warrant a GAF score of 1-10 whereas occasional failure to maintain personal hygiene would warrant a score of 11-20. For example, a psychotic patient with persistent delusional thinking may not attend to basic hygiene needs. They may not shower, change their clothes or because of their delusion and preoccupation, they may soil their clothes. In this examiner’s view, their mental disorder is producing marked functional impairment.

It is important to remember that the GAF excludes a rating of functional impairment based on objective, physical limitations. That is, the mental health professional should only assess to what extent the mental disorder itself produces functional impairment. For example, consider a patient who had a left hemisphere stroke that produced a dense paralysis of the right arm and leg, which in turn produced a persistent inability to maintain personal hygiene. In addition, assume the patient also developed a moderate reactive depression in association with his physical problems. In this example, it would be inappropriate for the mental health professional to give a very low functional rating score (e.g., GAF = 8 ) because the patient’s impaired hygiene functioning is related to actual physical limitations associated with the stroke. Unless the mood disturbance directly contributed to the functional impairment it should be deferred to other disciplines such as neurology who would rate the functional impairment within the central nervous system section (i.e., Chapter 13) of the AMA Guides.

It contrast, it could be reasonably argued that a mental diagnosis such as Pain Disorder Associated with both Psychological Factors and a General Medical Condition could produce functional impairment above and beyond any direct physical limitations from a medical condition. For example, a patient with a verifiable L4-L5 disc herniation could have functional limitations that exceed those expected on an objective medical basis alone. A psychiatric examination may yield evidence of marked pain behaviors (in the absence of malingering or embellishment) which result in avoidant behavior, deconditioning and reduced functional activities. In this example, the patient may be able to engage in a variety of functional activities but he does not secondary to his mental condition. This excess disability potentially would fall within the domain of psychiatry and the GAF rating process.

After the assessment of basic hygiene, the next level of impairment in functioning is an inability to function in almost all areas (GAF: 21-30) covering the domains of occupational (e.g., unable to work), social (e.g., avoids all contact with others) and leisure (e.g., unable to concentrate or engage in previous hobbies). If this level is not applicable, the next higher level would be whether there is major impairment in several areas of functioning (GAF: 31-40). The central distinction between the previous level and this level is the pervasiveness and severity of functional impairment across domains. The next level on the scale is serious (severe) impairment (GAF: 41-50) in social, occupational or leisure functioning. If this level is not applicable, then one would move up the scale to determine if the patient had moderate (GAF: 51-60) or mild (GAF: 61-70) functional impairment in these domains. Again, GAF scores above 70 are not considered clinically significant and do not yield any permanent psychiatric disability ratings. As with psychiatric symptoms, the determination of what functional GAF score is selected will depend on the combination of functional impairments found. For example, if the patient presents with moderate impairment in occupational functioning (51-60) but has mild impairment in social and leisure functioning, the score would fall at the higher end of the moderate range. Once a GAF score is determined for the patient’s functional status, the final GAF score can be established. As noted above, the mental health professional is required to take the lower of the two ratings between psychiatric symptoms and functional impairment.

The following example may be helpful in clarifying the GAF assessment process. Mr. Mechanic is a 47 year old, Caucasian, auto mechanic with 12 years of education. He was 46 at the time of the injury. He worked for his company for 7 years and was an excellent employee. He had no past medical, psychiatric or alcohol history. He has a good marriage and a stable family. Mr. Mechanic lifted a portion of an engine block at work and sustained a significant L4-L5 disc herniation. He experienced significant lower back pain. He received conservative treatment which included anti-inflammatories and physical therapy. Conservative treatment provided only temporary relief. He then received a series of 3 epidural injections which again provided only brief, temporary relief. Nine months after the injury, the patient’s orthopedist recommended surgical intervention and the patient agreed. Following surgery, the patient reported only modest gains in pain and functionality. He continued to have chronic back pain and functional limitations in basic and instrumental activities of daily living.

It was at this point that the patient began to experience depressed mood. He had tried conservative and more invasive interventions but continued to experience chronic pain and functional limitations. His depression was characterized by sadness, tearfulness, and impaired neurovegetative symptoms (sleep, appetite and libido). He was less motivated to engage in pleasurable activities and he was avoiding friends and family. Much of his day was spent at home watching television and worrying about his health, finances and vocational future. He denied suicidal ideation or intent. He was prescribed antidepressant medication and participated in psychotherapy for 4-5 months. At that time, he was found to have reached maximum medical improvement both from an orthopedic and psychological perspective.

An AME psychological examination was conducted and the patient reported mild to moderate depressed mood. He continued to have mild insomnia, neurovegetative symptoms and attentional difficulties secondary to his thoughts and concerns about his future and his ability to earn a viable income. He was more socially active with friends and family but was occasionally isolative. He was performing his basic activities of daily living and most of his instrumental activities as well although he now preferred that his wife manage the household finances due to his lack of motivation and attentional difficulties.

In this example, the patient’s GAF score could reasonably be argued to be a 62. His symptoms and functional impairments fell largely in the mild range although the presence of mild to moderate depressed mood would pull the GAF score closer to the moderate range. A GAF score of 62 would produce a permanent psychiatric disability rating of 22% in this case.

The RAND Study (2003) 11 examined various injured body regions and found that psychiatric injuries resulted in one of the largest average percentage losses in long term income relative to the other body parts. As a result, psychiatric injuries were given an FEC rank of 8 which yields a 40% (i.e., 1.4) FEC adjustment. This score is then modified again for occupation and age at time of injury. As a result, minor changes in the GAF score can result in substantial increases in the final permanent disability psychiatric ratings for a particular patient. Table A provides four different permanent disability rating scores for Mr. Mechanic based on five point incremental changes on the GAF.

Table A

GAF Rating WPI FEC Occ Adjust Age Adjust Final Pd%
GAF= 65 (mild Sx) 8 11 14 15 PD= 15%
GAF= 60 (mild to Mod) 15 21 26 28 PD= 28%
GAF= 55 (Mod Sx) 23 32 38 40 PD= 40%
GAF= 50 (Serious Sx) 30 42 48 50 PD=50%

In the above example, Mr. Mechanic reported mild to moderate depressed mood. Assume that his Psychological testing was valid and there was no evidence of embellishment or suboptimal effort. In principle, it could be argued that Mr. Mechanic suffered from moderate depressed mood, not mild (as evidenced by psychological tests such as the MMPI-2) and that the GAF score should therefore be in the moderate category (i.e., GAF = 51-60) with the combination of mild symptoms and functioning leading to perhaps a GAF score of 58. Ultimately, the rendering of a GAF score is a clinical determination based on the judgment of the physician. Given that the GAF scale is a subjective scale that requires the mental health specialist to integrate complex and sometimes discrepant information, it is likely that psychiatric claims will come under greater scrutiny by applicant and defense counsel with more intense cross examinations conducted given the potential for such claims to yield significant permanent disability awards.

REFERENCES 1. Schedule for Rating Permanent Disabilities. (January 2005). Labor and Workforce Development Agency. Department of Industrial Relations. Division of Workers’ Compensation. 2. Guides to the Evaluation of Permanent Impairment, Fifth Edition. (2001). American Medical Association. AMA Press. 3. American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. Washington, DC: American Psychiatric Association. 4. Hilsenroth, M.J., Ackerman, S.J., Blagys, M.D., Baumann, B.D., Baity, M.R., Smith, S.R., Price, J.L., Smith, C.L., Heindselman, T.L., Mount, M., & Holdwick, D.I (2000). Reliability and Validity of the DSM-IV Axis V. American Journal of Psychiatry, 157, 1858-1863. 5. American Psychiatric Association (2000). Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision. Washington, DC: American Psychiatric Association. 6. Bates, L.W & Shaw, XB. (2002). Effects of brief training on application of the Global Assessment of Functioning Scale. Psychological Reports, 91, (3 Pt I): 999-1006. 7. Meehl, RE. (1954). Clinical versus statistical prediction: A theoretical analysis and a review of the evidence. Minneapolis: University of Minnesota Press. 8. Woldoff, S. (2004). Reliability of The Global Assessment of Functioning Scale. Doctoral Dissertation. Drexel University. 9. First, M. (2001). GAF Report for The Global Assessment of Functioning Scale. (2001). Multi-Health Systems. Toronto, Canada. 10. Yamauchi, K., Yutaka, O., Baba, K., & Ikegami, N. (2001). The Actual Process of Rating the Global Assessment of Functioning Scale. Comprehensive Psychiatry, 42, (5), 403-409. 11. The Evaluation of California’s Permanent Disability Rating Schedule, Interim Report (December 2003). The RAND Institute for Civil Justice.

About the Author: David B. Freeman, PhD, QME, is an assistant clinical professor in the department of psychiatry at UCLA Medical Center and the Clinical Director for Comprehensive Psychological Services. He is a board certified, UCLA trained neuropsychologist who regularly lectures on topics such as GAF, mild traumatic brain injury, PTSD and the assessment of malingering in Workers’ Compensation and civil litigation. Dr. Freeman’s main office is located at 16530 Ventura Blvd., Suite 200 Encino, CA 91436.


No comments yet.

Leave a Reply