SCOFF Questionnaire: Score It at Therapy Intake

Authored by Amara Collins, a clinical psychologist who writes about continuity of care: how a screen from one intake has to still be readable at the next visit.

She said “stress eating.” Partner is in the waiting room. You have five yes/no questions and about eight minutes before this hour turns into a thought record about control.

That is the job of the SCOFF questionnaire at therapy intake. You already chose a brief eating-disorder screen. You score it in this visit, you treat two or more yes answers as suspicion rather than a diagnosis, and you leave a chart note the next clinician can read without reopening the form. If you are still choosing which brief measure belongs on the caseload at all, start with the scales for mental health matrix. Do not stack this screen on top of a second eating-disorder inventory in the same hour.

Educational content for licensed therapists, not clinical, medical, or legal advice. Screening is not diagnosis. Obtain the items from the original paper and follow its terms. Adapt cutoff, collaterals, and disposition to your setting and supervision.

How you score SCOFF in this hour

You are not running a population program. You are sitting with one client who already made it into therapy. The screen still has to do three things before they leave:

1. Name a load you can compare later, on this same screen.
2. Catch a positive screen without pretending it is a DSM diagnosis.
3. Leave a written next step: an interview, a medical loop, or a specialist path.

Morgan, Reid, and Lacey built five yes/no items to raise suspicion that an eating disorder might exist, then send you to a proper assessment. In their 1999 BMJ paper, they set the threshold at two or more positive answers. In a specialist clinic sample of 116 women aged 18 to 40 with anorexia nervosa or bulimia nervosa, and 96 women without an eating disorder, that cutoff gave 100% sensitivity (95% CI 96.9% to 100%) and 87.5% specificity (79.2% to 93.4%). They said the false-positive rate of 12.5% was an acceptable trade-off for that sensitivity. They also said the tool was designed to raise suspicion, not to diagnose.

Say how you asked it. The original sample heard the questions orally. Luck and colleagues later asked the same five items aloud in a primary care practice, in a separate room, in about two minutes (Luck et al., 2002). A form completed with a partner watching is not the same form completed alone. Write who was in the room.

Do not rebuild the stems from the acronym. Open the paper. Ask from a copy you are allowed to use.

What the five items screen for

The stems live on the 1999 BMJ paper. Open that paper for the SCOFF questionnaire item wording and scoring key.

The five yes/no items cover self-induced vomiting, loss of control over eating, a recent large weight loss (one stone, which is 14 lb, in three months), feeling fat when others say you are too thin, and food dominating life. Each yes is one point. The total runs from 0 to 5.

That is a narrow net. It was built around anorexia nervosa and bulimia nervosa in young women. It is not a body-image worksheet, not an EDE-Q, and not a medical exam. If the hour needs the fuller eating-disorder interview after a flag, move to eating disorder assessment. Stay here when the job is today’s total.

Score, cutoff, and what you do with the number

Write the total, the cutoff you used, and whether it met that line. A later reader cannot audit a bare “positive.”

Score	What it means	Next step	Limits
0 or 1	Below the published cutoff	Interview anyway if signs are already in the room	A quiet total does not clear medical risk, BED, ARFID, or men
2 or more	Positive screen	Same-visit interview, medical check, specialist path	Not a diagnosis. Original clinic sample overstates later accuracy

Two or more yes answers is the cutoff in both the 1999 derivation study and the 2002 primary-care study. Do not silently swap 2 and 3 across visits and call the series a trend.

A flagged total is a reason to talk, not a medication order and not a code. If the next document is a team plan rather than today’s screen, use the eating disorder treatment plan template. This visit still has to close before they reach the parking garage.

Download the mental health rating scales guide to keep the measure, cutoff, and next-step chart line ready for the next visit.

What later studies did to that 100% sensitivity

The original number is a specialist-clinic number. Do not quote it as if it were a primary-care waiting room.

Luck and colleagues asked sequential women aged 18 to 50 in two southwest London general practices. At the same cutoff of two or more, sensitivity was 84.6% (54.6% to 98.1%) and specificity was 89.6% (86.3% to 92.9%). Positive predictive value was 24.4%. Negative predictive value was 99.3%. The screen caught every case of anorexia nervosa and bulimia nervosa in that sample, and seven of nine eating disorders not otherwise specified. Of 328 women without an eating disorder, 34 screened positive.

Cotton, Ball, and Robinson compared the SCOFF with another four-item screen in mixed student and primary-care samples. At two or more abnormal answers, sensitivity was 78% (62% to 93%) and specificity was 88% (84% to 93%) (Cotton et al., 2003).

Kutz and colleagues pooled 25 validation studies. Overall sensitivity was 0.86 (0.78 to 0.91) and specificity was 0.83 (0.77 to 0.88). The highest sensitivity sat in case-control studies of young women with anorexia nervosa and bulimia nervosa. Sensitivity fell when studies included more men, included binge eating disorder, or recruited from large community samples. They concluded there is not enough evidence to use the SCOFF for the full range of DSM-5 eating disorders in primary care and community settings (Kutz et al., 2020).

The USPSTF 2022 evidence review, for detecting any eating disorder among adults at a cut point of 2 or more, reported pooled sensitivity of 84% (74% to 90%) and pooled specificity of 80% (65% to 89%) across 10 trials (n = 3684). At a cut point of 3 or more, sensitivity dropped to 69% (56% to 80%). The Task Force still issued an I statement for screening asymptomatic adolescents and adults: enough accuracy in adult women to describe the tool, not enough evidence that screening people with no signs or symptoms improves outcomes (USPSTF, 2022). Therapy intake is not that asymptomatic program. You still do not treat the total as a diagnosis.

Who the screen misses

A total under two is not a medical all-clear. If bradycardia, syncope, rapid weight loss, or restriction is already in the room, you assess that, cutoff or not.

Watch the populations the later papers name:

• Men, including muscularity-oriented eating the five items were not written to catch. USPSTF called accuracy inadequate in men and adolescents.
• Binge eating disorder. The original items were built before DSM-5 BED. Kutz found lower sensitivity when BED was in the sample.
• ARFID and other presentations without weight-and-shape overvaluation. Feeling fat and food dominating life are the wrong net.
• Large community samples, where the original 100% sensitivity does not hold.
• Anyone whose partner, parent, or interpreter is sitting in the chair. Write that. Rescore later if you need a private total.

NICE is blunt: do not use screening tools (for example, SCOFF) as the sole method to determine whether or not people have an eating disorder (NICE NG69). That sentence belongs in the chart logic, not as decoration.

Next step after a positive screen

The next step is a clinical move, not a restatement of the band.

In the R sheet, three yes answers sent you to a same-visit interview of restriction and compensation, a PCP booking, and a specialist-path conversation. It did not send you to a thought record. Regular eating, self-monitoring, and family refeeding sit on eating disorder therapy techniques after medical stability and training are in place. They are not the first move on a flagged intake.

Same-day, in the room:

• Interview restriction, bingeing, compensation, weight-and-shape overvaluation, and medical red flags. That interview is the eating disorder assessment, not a second pass of the five items.
• Check what you can check today: recent loss, fainting, chest pain, compensatory behavior, who is managing weight and labs. Same-day medical contact if the picture is acute.
• Name the path out loud: hold in outpatient therapy with a medical co-clinician, refer to a specialist team, or both.
• Write the score, the cutoff, the next step, and the limit you are carrying (who this tool was built on).

Do not start trauma processing on a first flagged total. Do not hand them a food log as the only follow-up. Stabilize the frame, then write.

What the chart note has to carry

Weak: “SCOFF done.”

Stronger, and short enough to sign:

• Scale name, and that you used a copy of the 1999 paper
• Mode, and who was in the room
• Score out of 5, the cutoff of 2 or more, and whether it met that line
• The action you took today
• The limit you are not pretending away
• The next date the same screen will be repeated, if you are repeating it

Here is the line from the sheet at the top, written so a covering clinician can use it on Thursday:

That sentence is the chart note. The rest of the progress note can hold the week and the plan. Repeat this screen only when you need a comparable total on the same five items. An EDE-Q from another client on your caseload is not the comparison.

Blank intake sheet

The filled sheet is the pattern. Use the blank fields when you are charting a real intake. Fictional initials only on the example.

Do not hand the sheet to them as a quiz. Do not paste the BMJ stems into the blanks.

Where the next visit finds the score

The total, the cutoff, and the next step have to sit in one record or Thursday starts from rumor. You still pick the screen, read the paper, run the medical pathway, and sign what is true.

References

• Morgan, J. F., Reid, F., & Lacey, J. H. (1999). The SCOFF questionnaire: Assessment of a new screening tool for eating disorders. BMJ, 319(7223), 1467-1468. DOI
• Luck, A. J., Morgan, J. F., Reid, F., O’Brien, A., Brunton, J., Price, C., Perry, L., & Lacey, J. H. (2002). The SCOFF questionnaire and clinical interview for eating disorders in general practice: Comparative study. BMJ, 325(7367), 755-756. DOI
• Cotton, M. A., Ball, C., & Robinson, P. (2003). Four simple questions can help screen for eating disorders. Journal of General Internal Medicine, 18(1), 53-56. PMC full text
• Kutz, A. M., Marsh, A. G., Gunderson, C. G., Maguen, S., & Masheb, R. M. (2020). Eating disorder screening: A systematic review and meta-analysis of diagnostic test characteristics of the SCOFF. Journal of General Internal Medicine, 35(3), 885-893. DOI
• National Institute for Health and Care Excellence. (2017). Eating disorders: Recognition and treatment (NICE Guideline NG69). NICE NG69
• US Preventive Services Task Force. (2022). Screening for eating disorders in adolescents and adults: US Preventive Services Task Force recommendation statement. USPSTF recommendation