10.59. Fast Alcohol Screening Test (FAST)¶
10.59.1. History and guide¶
Original: Hodgson et al. (2002) 1.
10.59.2. Sources¶
10.59.3. Notes on the method of administration and scoring¶
This is slightly trickier than it appears at first glance – there are a couple of administration guidelines that differ very slightly, but the scoring method is consistent.
Development
Hodgson et al. 1 developed the FAST from the AUDIT, seeking a quick version. They had the aim that one question serves as a first filter. AUDIT Q3 was found to be good for that. They modified it to use a threshold of ≥8 units for men. They found that AUDIT Q5, Q8, and Q10 were collectively the best for a second-stage filter, with minor modifications.
The four FAST questions are therefore approximately as follows:
FAST Q1: How often ≥6 units (female) or ≥8 units (male) on a single occasion in the last year? = AUDIT Q3, approximately, but not quite.
FAST Q2: … failed to do what was normally expected… = AUDIT Q5.
FAST Q3: … unable to remember… the night before… = AUDIT Q8.
FAST Q4: … relative/friend/doctor/health professional… concerned = AUDIT Q10.
The test is referred to as a two-stage screening test (e.g. Abstract).
Administration
There seems to be a slight difference in administration instructions between different sources of the test.
The original paper 1 says (p62): “AUDIT Question 3 [= FAST Q1]… served as the best first filter… If the response is ‘never’… then there is no hazardous use. If the response is ‘weekly’/’daily or almost daily’… then there is probably hazardous use… this one questions classifies… 66%… with an accuracy of 97%. Only the 34% of patients who responded ‘less than monthly’ or ‘monthly’ to [AUDIT] Question 3 [FAST Q1] need to be asked further questions”.
The FAST manual says that “a response must be obtained for each question… In the case of self-completion, a staff member… should check that all questions have been answered” (2, page 6), although it is not necessary to score all (depending on Q1). There is no suggestion that one should not answer Q2-Q4 if certain answers are given to Q1. The paper version of the test simply says “For the following questions, please circle the answer…”
Public Health England says that “the test consists of only 4 questions… which are asked in 2 stages” 4. It clarifies that one should “only answer the following questions [Q2-Q4] if the answer above [to Q1] is Never (0), Less than monthly (1) or Monthly (2). Stop here if the answer is Weekly (3) or Daily (4)” (5, page 1).
I presume that PHE trimmed the administration procedure a little for speed, without any change to the scoring methodology (see below).
Scoring
Hodgson et al. 1 considered two scoring methods (p64):
Both methods decide “FAST negative” if the response to Q1 is “Never”.
Both methods decide “FAST positive” if the response to Q1 is “weekly” or “daily or almost daily”.
Then:
Method 1 (not preferred): FAST negative if (FAST) Q2 and Q3 are “Never” and Q4 is “No”; FAST positive for all others (i.e. positive for any hint of a problem).
Method 2 (preferred): all questions are scored 0-4 and the result is FAST positive if the total score for all four questions is ≥3.
In more detail: questions Q1-Q3 have five options and are scored 0-4; question Q4 has three options, scored 0, 2, and 4.
The original paper prefers Method 2 1. The FAST manual uses Method 2 2. The PHE scoring method is Method 2 5. So that’s clear.
CamCOPS
CamCOPS marks all four questions as mandatory, following the FAST manual, and following the principle of collecting more data (given also some ambiguity) in a patient-completed questionnaire. Note also that the original validation involved a questionnaire in which participants saw and answered all four questions.
Reviewed 2019-11-01. The scoring method was correct; see fast.cpp in the
source code and
camcops_server.tasks.fast.Fast.is_positive(). Explanatory text added
here.
10.59.4. Intellectual property rights¶
Believed to be in the public domain.
Made freely available with encouragement to copy by the UK NHS Health Development Agency and the UK National Institute for Clinical Excellence 3 and subsequently by Public Health England 4 5 along with the AUDIT.
Developed from the AUDIT (q.v.)
Footnotes
- 1(1,2,3,4,5,6)
Hodgson R, Alwyn T, John B, Thom B, Smith A (2002). The FAST Alcohol Screening Test. Alcohol and Alcoholism 37: 61-66. http://www.ncbi.nlm.nih.gov/pubmed/11825859
- 2(1,2,3)
Hodgson R, Alwyn T, John B, Smith A, Newcombe R, Morgan C, Thom B, Hodgson R, Waller S [2006, by PDF date]. Fast screening for alcohol problems: manual for the FAST Alcohol Screening Test [v3, by PDF document title]. http://www.dldocs.stir.ac.uk/documents/fastmanual.pdf
- 3(1,2)
UK National Institute for Health and Care Excellence. Document previously at http://www.nice.org.uk/niceMedia/documents/manual_fastalcohol.pdf (e.g. 2013), now gone (2019-11-01). See https://web.archive.org/web/2013*/http://www.nice.org.uk/niceMedia/documents/manual_fastalcohol.pdf.
- 4(1,2,3)
Public Health England (2017). Alcohol use screening tests. https://www.gov.uk/government/publications/alcohol-use-screening-tests; https://www.gov.uk/government/publications/alcohol-use-screening-tests/guidance-on-the-5-alcohol-use-screening-tests; https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/684828/Fast_alcohol_use_screening_test__FAST__.pdf.
- 5(1,2,3,4)
Public Health England (2017). Fast alcohol screening test (FAST). https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/684828/Fast_alcohol_use_screening_test__FAST__.pdf