Global Assessment of Functioning (GAF) Scale (DSM - IV Axis V)
Note: This version of the GAF scale is intended for academic use only. Although it is based on
the clinical scale presented in the DSM - IV, this summary lacks the detail and specificity of the
original document. The complete GAF scale on page 32 of the DSM - IV should be consulted for
clinical use.
Code
Description of Functioning
91 - 100
Person has no problems OR has superior functioning in several areas
OR is admired and sought after by others due to positive qualities
81 - 90
Person has few or no symptoms. Good functioning in several areas. No
more than "everyday" problems or concerns.
71 - 80
Person has symptoms/problems, but they are temporary, expectable
reactions to stressors. There is no more than slight impairment in any
area of psychological functioning.
61 - 70
Mild symptoms in one area OR difficulty in one of the following: social,
occupational, or school functioning. BUT, the person is generally
functioning pretty well and has some meaningful interpersonal
relationships.
51 - 60
Moderate symptoms OR moderate difficulty in one of the following:
social, occupational, or school functioning.
41 - 50
Serious symptoms OR serious impairment in one of the following:
social, occupational, or school functioning.
31 - 40
Some impairment in reality testing OR impairment in speech and
communication OR serious impairment in several of the following:
occupational or school functioning, interpersonal relationships, judgment,
thinking, or mood.
21 - 30
Presence of hallucinations or delusions which influence behavior
OR serious impairment in ability to communicate with others OR serious
impairment in judgment OR inability to function in almost all areas.
11 - 20
There is some danger of harm to self or others OR occasional failure
to maintain personal hygiene OR the person is virtually unable to
communicate with others due to being incoherent or mute.
1 - 10
Persistent danger of harming self or others OR persistent inability to
maintain personal hygiene OR person has made a serious attempt at
suicide.
GAF Scoring Vignettes
The following vignettes come from an e-mail thread on FORUM in November 1997. The vignettes were
presented by John Simpson from the Brockton/West Roxbury VAMC. The e-mail has been edited to include
only the vignettes and discussion pertaining to the vignettes.
In early October 1997, the IPCC (Intensive Psychiatric Community Care) team at the VAMC in Brockton,
Massachusetts reviewed all 80 patients in the program and as a team discussed and assigned GAF scores to reflect
current functioning (e.g., during the past month). The following patients were then selected to illustrate specific GAF
ratings and to provide some guidance to program staff doing individual ratings in the future. The vignettes are
presented here in the hope they will be of interest and possible use to other VA clinicians faced with periodically
assigning GAF scores to chronic mentally ill outpatients.
The Brockton IPCC team claims no special expertise with the GAF Scale, although team members have been making
GAF ratings for years as part of data collection undertaken by the Northeast Program Evaluation Center (NEPEC).
Our experience has been that periodic team discussion helps substantially to reduce inter-rater variability and perhaps
helps control an apparent tendency to assign higher scores than are sometimes justified by a strict interpretation of
the GAF scale. We welcome any comments or questions concerning the following "anchors," and encourage other
programs to conduct similar exercises and make the results available for discussion and mutual edification.
GAF = 20:
Mr. A is a single veteran in his mid 60s who was admitted to the IPCC program about 2 years ago following lengthy
hospital stays and repeated failures to adjust to residential care placements. One residential care sponsor described
him as being very dependent and requiring constant supervision and attention. Mr. A was rehospitalized in 1992 when
he assaulted a residential care sponsor after she told him to take a shower because he had been incontinent of feces.
During that admission he developed somatic delusions about having cancer, his ADLs continued to be very poor, and
he began to smear feces. With IPCC support, he was discharged in Nov. 1995 and placed in a rest home because he
required a high level of care and supervision of his ADLs. Mr. A attends activities at the community-based IPCC day
program 3 days a week. His speech is tangential and irrelevant at times, but he will usually cooperate if given explicit
directions. Despite the fact that Mr. A has been able to live outside the hospital for the past 2 years, he remains very
dependent on the rest home and IPCC staff for all his needs, and smearing feces continues to be a problem.
CURRENT GAF RANGE 11-20 ("...OCCASIONALLY FAILS TO MAINTAIN MINIMAL PERSONAL HYGIENE, E.G.,
SMEARS FECES...")
GAF = 22:
Mr. B is a veteran in his early 40s with a diagnosis of schizoaffective disorder who has had multiple psychiatric
admissions, including 11 in the past two years for severe delusions (e.g., he believes he is a character from the Little
Rascals, that he has a girfriend who is a famous TV actress, and that he owns seven businesses). He has extremely
poor money management skills, e.g., he is unable to purchase food and has been evicted for failure to pay his rent.
He has a history of giving large sums of money away (i.e. $500-$1,000 at a time) to people (often drug users) he just
met off the streets and considers his "friends" for religious reasons. He refuses to participate in the community-based
IPCC day program and prefers to stay in bed much of the day. He is noncompliant with taking psychotropic
medications and attending outpatient appointments at the hospital. Recently, a conservator was appointed to handle
his funds. CURRENT GAF RANGE 21-30 ("BEHAVIOR IS CONSIDERABLY INFLUENCED BY DELUSIONS" AND
"SERIOUS IMPAIRMENT IN ... JUDGEMENT" AND "INABILITY TO FUNCTION IN ALMOST ALL AREAS (E.G.,
STAYS IN BED ALL DAY; NO JOB...)."
GAF = 25:
Mr. C is a single veteran in his late 50s who was admitted to the IPCC program on July 1, 1996 after thirty years of
continuous hospitalization at the BVAMC. He now lives in a residential care home and attends the IPCC communitybased day program five days a week. He needs lots of prompting to attend to ADLs and uses an assisted
transportation system to attend the day program. He hoards large amounts of money on his person and refuses to
open a bank account, because he believes "the banks are controlled by the Mafia." He continues to express bizarre
delusions of a religious and grandiose nature, believing that he is the Son of God and that he has the ability to
communicate with animals and extraterrestrial beings. CURRENT GAF RANGE 21-30 ("BEHAVIOR IS
CONSIDERABLY INFLUENCED BY DELUSIONS" AND "SERIOUS IMPAIRMENT IN ... JUDGEMENT")
GAF = 28:
Mr. D is a veteran in his mid 40s who was discharged 2 years ago after 18 years of hospitalization. He resides in a
residential care home and attends IPCC community-based day programming 5 days/wk. He is in a structured money
management program and receives concrete rewards for completion of daily ADLs. He is able to negotiate the city
transportation system and is very capable of accessing community resources. He is extremely delusional, believing
that his body is made out of glass or wood and believes that many people are his mother, including the queen mother
in England. He is also very thought-disordered and tangential; suspected brain damage contributes to speech
oddities. CURRENT GAF RANGE 21-30 ("BEHAVIOR IS CONSIDERABLY INFLUENCED BY DELUSIONS...")
GAF = 32:
Mr. E is a single veteran in his mid 50s who has a long history of schizoaffective illness characterized by frequent
mood swings. When in a manic phase, he exercises very poor judgment regarding financial matters and his behavior
is very inappropriate. When depressed he becomes sullen and withdrawn. He also experiences auditory hallucinations
at times. He had numerous and lengthy hospitalizations for his illness, although far fewer and much shorter lengths of
stay since entry into the IPCC program in 1989. He has not been able to hold a job in many years. His personal
hygiene is quite poor. He resides in a boarding home where he receives assistance with his medications. CURRENT
GAF RANGE 31-40 ("...MAJOR IMPAIRMENT IN SEVERAL AREAS SUCH AS WORK OR SCHOOL, FAMILY
RELATIONSHIPS, JUDGEMENT, THINKING OR MOOD")
GAF = 35:
Mr. F is a veteran in his early 50s who has had multiple admissions to the hospital. Two admissions during the past
year were for an increase in auditory hallucinations that were telling him to harm himself and others. He continues to
hear voices but is able to separate out that they are not real and will not act on them. In some areas he functions
independently, for example, he showers and changes clothes daily, attends mass daily, and has a group of church
friends with whom he has coffee. He is compliant with appointments and medications (even though he insists that
religion is better than the meds), and gets along well with the other day program members and with the residents at
his residential care home. At the community-based IPCC day program, he follows through with his work assignment
and attends groups, where he raises pertinent issues about community living and relates his personal experiences.
However, he is very religiously preoccupied and has delusions about electricity, telephones and vehicles that are
impairing, e.g., because of his delusions he will not ride in cars. As a result, he remains dependent on program staff
for some essential services. CURRENT GAF RANGE 31-40 ("SOME IMPAIRMENT IN REALITY TESTING" AND
"MAJOR IMPAIRMENT IN SEVERAL AREAS, SUCH AS WORK OR SCHOOL, FAMILY RELATIONSHIPS,
JUDGEMENT, THINKING, OR MOOD")
GAF = 45:
Mr. G is a veteran in his early 50s who has had multiple hospitalizations due to non-compliance with medications.
When he was living in his own condo he would stop taking meds and become very paranoid, with hallucinations. He is
now living in an apartment within a residential care home, where meals are provided and meds are supervised. He
continues to experience some symptoms of paranoia and will have an occasional hallucination in which someone is
calling his name. However, he drives his own car and attends the IPCC community-based day program, where he is in
charge of collecting lunch monies and keeping data for the program's point-based reward system. He usually keeps to
his own at the day program and what conversations he has tend to be short, but he visits his family twice a month.
CURRENT GAF RANGE 41-50 ("SERIOUS SYMPTOMS" AND "SERIOUS IMPAIRMENT IN SOCIAL,
OCCUPATIONAL ... FUNCTIONING")
GAF = 52:
Mr. H is a divorced in-country Vietnam Veteran in his late 40s with a dual diagnosis of schizoaffective disorder and
alcoholism, although he has been abstinent from alcohol for several years and has not been overtly psychotic since
entry into the IPCC program in 1990. However, he becomes very anxious in social situations or when confronted with
a stressful situation. He had not worked for several years until two years ago when he obtained part-time employment
bagging groceries at a supermarket. He had some significant difficulties coping with the job and at one point became
depressed and required admission to inpatient psychiatry. He currently is unemployed because he quit his part-time
job and attempted to work full time, but then quit working altogether when he felt unable to cope with the demands of
the new job. He resides in a rooming home and manages his own medications and funds. CURRENT GAF RANGE
51-60 ("MODERATE SYMPTOMS" AND "MODERATE DIFFICULTY IN SOCIAL, OCCUPATIONAL, OR SCHOOL
FUNCTIONING")
Comment/question
The GAF examples (anchors) were very helpful. The need to be as accurate as possible with our GAF scores has
taken on new importance with the mandate to provide these scores for all patients and their use in capitation formulas.
The information, examples, and opening of a dialogue provided by the Brockton IPCC team is very valuable in
promoting this aim. In this regard, their illustrative "anchors" raised some questions for me (perhaps because I also
work with the severely mentally ill):
1) To what degree must a patient's behavior be influenced by their delusions or hallucinations to warrant a GAF below
30? The examples of Mr. D & F raised this question. Mr. D has somatic delusions and delusions about others being
his mother (plus disordered thinking). From what I can tell from the information provided, however, they don't seem to
effect his behavior "considerably." I would be more inclined to give a GAF of 32-33.
The vignette about Mr. F raised the same question, but in the opposite direction. He also has delusions, but they are
noted to be impairing. His are about electricity, telephones and vehicles. On could conclude that if his life is
"considerably influenced," perhaps consistently significantly disrupted by these delusions that he might score more
closely to a 29-30 on the GAF.
2) To what degree must a patient's reality testing be intact to warrant a GAF above 40? This question was raised by
the Mr. G vignette. It notes that he has some symptoms of paranoia and an occasional AH of someone calling his
name. Given the description, I probably would not view him as having AH, but the "symptoms of paranoia" could be
viewed as impaired reality testing. This suggests the possibility of a lower GAF: 39-40.
3) To what degree must a patient's general functioning improve (moderate difficulty vs serious impairment) to warrant
a GAF above 50? The vignette about Mr. H raised this question. Mr. H tried to work, but ended up quitting because he
couldn't cope with the demands of the job. To me this suggested serious occupational functioning, and I would be
inclined to give him a GAF score closer to 49-50.
Response:
On behalf of the Brockton IPCC team, we appreciate the thoughtful comments concerning some of our GAF ratings.
The points made are excellent and have prompted a good deal of discussion amongst the team (hence the delay in
responding).
A general consideration is that our initial ratings were based on the team's clinical knowledge of the patients, and then
the vignettes were written to try to summarize the salient facts. We then SHOULD have rated the vignettes to make
sure they were fully consistent with our ratings of the patients, but unfortunately we did not. Hence, some of the
vignettes do not convey all the information used to make the ratings. This probably accounts for some of the apparent
discrepancies, as discussed below.
A second consideration is that we attempted to apply the following guideline suggested by Michael First, M.D.
("Mastering DSM-IV Axis V." Jrnl Prac. Psych. and Behav. Hlth., Nov. 1995, 258-259): "treat the GAF as if it were two
scales: one for symptom severity and another for level of functioning. [Then] ... make one rating for severity and a
second for level of functioning. The worst of the two can be used as the GAF." This is a useful approach in cases
where there is some discrepancy between the patient's symptomatology and level of functioning, e.g., a patient with
psychotic symptoms who nevertheless functions fairly independently. This rating rule also helps counter-act an
apparent tendency to try to balance symptomatology and functioning or even to discount the worse dimension and so
assign a patient a higher GAF rating than is probably justified. UNFORTUNATELY, we now realize that we did not
always follow this rule strictly, and this probably accounts for some of the additional discrepancies that were noted.
I will now address the specific questions in the order they were raised:
Mr. D (GAF = 28): THE QUESTION RAISED IS WHETHER THE DELUSIONS NOTED IN THE VIGNETTE
AFFECTED HIS BEHAVIOR "CONSIDERABLY" AS REQUIRED FOR A GAF RATING IN THE 21-30 RANGE. This is
one of the cases where we rated the patient, but failed to rate the vignette and so did not realize that some essential
information was missing from the vignette (which by itself probably does not justify a GAF below 30). The following
information about Mr. D is also relevant: his speech is almost unintelligible and his judgment is very poor ("SERIOUS
IMPAIRMENT IN COMMUNICATION OR JUDGMENT"). For example, he burned himself by somehow lighting his
pocket on fire, but didn't tell anyone and so the burns were only detected later by an IPCC staff member who was
assisting Mr. D in our supervised ADL program. In a separate instance, he was bitten by a dog but again failed to tell
anyone, so medical assistance was delayed. We believe these additional facts would support the assigned rating
(GAF = 28).
Mr. F (GAF = 35): THE QUESTION WAS WHETHER THE DELUSIONS NOTED "CONSIDERABLY INFLUENCED"
HIS LIFE, SUGGESTING A GAF SCORE IN THE UPPER END OF THE 21-30 RANGE WOULD BE MORE
ACCURATE. This was difficult to decide, but after much discussion the team agreed that Mr. F is influenced by his
delusions, but not "considerably." For example, he manages to get around his delusion-imposed limitation of not riding
in cars by walking or riding public transportation, and so generally functions adequately. This is obviously a judgment
call, and on reconsideration we would assign a lower score in the 31-40 range, namely GAF = 32.
Mr. G (GAF = 45): THE QUESTION IS WHETHER THE SYMPTOMS NOTED IN THE VIGNETTE (SOME PARANOIA
AND AN OCCASIONAL AUDITORY HALLUCINATION) INDICATE "SOME IMPAIRMENT IN REALITY TESTING"
AND SO INDICATE A GAF RATING IN THE 31-40 RANGE. Upon reconsideration, we agree that Mr. G's symptoms
meet the GAF criterion of "some impairment in reality testing," and so the GAF score should be revised downward,
e.g., to GAF = 38. [Note: this is a clear instance where the team gave too much weight to the patient's relatively high
level of functioning -- he drives his own car and successfully invests his own money -- and not enough to his psychotic
symptoms. However, the rule of rating symptoms and functioning separately, and then picking the lower rating, should
give the most weight to the psychotic symptoms and thereby lead to the lowered GAF score.]
Mr. H (GAF = 52): THE QUESTION IS WHETHER MR. H, WHO ENDED UP QUITTING HIS LATEST JOB, MEETS
THE GAF CRITERION OF "ANY SERIOUS IMPAIRMENT IN ... OCCUPATIONAL FUNCTIONING" FOR A SCORE
IN THE 41-50 RANGE. This is another instance where the information provided in the vignette was not adequate to
justify the rating assigned, but additional facts probably do justify the rating. In particular, Mr. H worked successfully
part-time for over two years, and then quit the part-time job in order to try full-time employment in a job which also
required a higher level of skill. He couldn't cope with this full-time position, so he quit that job and remains unemployed
because he did not go back to part-time work. However, his employment record suggests he is capable of holding
down a part-time job at an appropriate skill level, which we would rate as moderate rather than severe difficulty in
occupational functioning. This is another difficult judgment call, but we prefer to maintain our rating of GAF = 52.
IN CONCLUSION: We greatly appreciate the above insightful comments, which have led us to revise our GAF ratings
downward in two cases (Mr. F from 35 to 32; and Mr. G from 45 to 38) and to provide additional information needed to
justify the GAF ratings for Messrs. D and H. Some of these issues are difficult to resolve, and disagreements will likely
persist. As I stated in my initial comments, we claim no expertise in using the GAF, and are just trying to muddle
through like everyone else. In particular, we have NOT attempted to provide a sort of "gold standard" that other
respondents have mentioned. Despite these limitations, we hope this discussion has been fruitful and we look forward
to future exchanges of views.
Comment
Many thanks to Brockton staff for initiating this discussion. I note that the patient population at Brockton was CMI
OUTpatients. We have rated a population of INpatients 3 times this year, on a chronic ward that has about 40%
"STAR III" and 60% "STAR II" patients. The first time was on 3-31-97, and it came as a surprise to us...here we were
doing the annual census in the middle of the year! We kept the scores, used them as a foundation for the next rating
on 6-30-97. When we looked at that distribution of ratings, we knew some homework was needed. We had rated each
patient independently of the others, and we had used only scores divisible by 5...nothing in between. And we had
clearly assigned higher scores than were warranted (the tendency noted above by Brockton staff). We had not had
the benefit, however, of the sets of vignettes noted in this string.
So we got the original 1976 paper on the GAS: Endicott, J., et al. The Global Assessment Scale. Archives of General
Psychiatry, 1976(33), 766-771. We learned, "In order to determine the scale point within the ten-point interval, the
defining characteristics of the two adjacent intervals are examined to determine whether the subject is closer to one or
the other." (p. 766) We had found the "ladder" format of the DSM-IV's GAF, on p. 32, rather difficult to use...it
scrambles together the symptom severity and level of functioning characteristics.
It boggled my brain to think of three intervals at any one time. So I turned the scale from vertical to horizontal, in order
to have the perceptual/conceptual benefits of a Likert-type scale, and this helped us a lot. Only the words of the DSM
itself were used, but now displayed in a way that looks more like a continuum:
Horizontal GAF Rating Scale
On the page, the top row therefore has 5 cells with information, the next row has 2 cells (personal hygiene), the next
row has 3 cells (the major theme being communication), and the bottom row has 3 cells (level of psychosocial
functioning). We then used this form on the third rating occasion, the end of year census, in a treatment team meeting.
Each patient was rated individually, using the GAF characteristics, then we compared patients who seemed to have
similar scores. We think our rating this time was more valid and more reliable, but we know we still have much to
learn.
Comment
The new Directive 97-059 entitled, Instituting Global Assessment of Function (GAF) Scores in Axis V for Mental
Health Patients, was signed off by Dr. Kizer, dated November 25, 1997.