How this instrument works
The Crohn's Disease Activity Index comes from the National Cooperative Crohn's Disease Study, published by Best, Becktel, Singleton, and Kern in 1976. The investigators tracked 112 patients through 187 clinic visits, recording eighteen candidate variables at each one, then fit a regression equation to the subset that best tracked a clinician's own judgment of disease severity. What survived is eight terms, each carrying a different weight depending on how strongly it predicted that judgment.
Six of the eight terms come straight from a seven-day symptom diary the patient keeps at home: a daily count of liquid or soft stools, a daily pain rating from 0 to 3, and a daily general-wellbeing rating from 0 to 4, each summed across the week. The remaining terms come from the clinic visit itself — a count of extraintestinal complications, whether the patient is taking an antidiarrheal drug, whether the examiner feels an abdominal mass, the day's hematocrit compared with an expected value, and current weight compared with the patient's usual weight.
That mix is deliberate. A score built only from symptom-diary items could be talked up or down by how a patient chooses to describe a week. Anchoring part of the total to hematocrit and weight — figures a clinic visit measures rather than a patient reports — keeps the index from being gamed by wording alone. The 1976 paper itself defines only two thresholds directly: 150 or below as quiescent disease, and above 450 as extremely severe. The now-common four-tier reading — remission at or below 150, mild-to-moderate from 150 to 219, moderate-to-severe from 220 to 450, severe above 450 — is a convention later clinical trials layered on top, not a line drawn in the original paper.
- Enter the 7-day sum of liquid or soft stools and the 7-day sum of the daily 0–3 pain rating, both from the symptom diary.
- Enter the 7-day sum of the daily 0–4 general wellbeing rating from the same diary.
- Set Complications (0–6 of the listed categories), Antidiarrheal use, and Abdominal mass from the same-day exam.
- Set Sex, then enter current Hematocrit — the deficit is measured against 47% for men or 42% for women.
- Enter current Weight and Usual/standard weight, then read CDAI; the working block shows every term.
Worked example — 14 stools, moderate pain, one complication
A patient logs 14 loose stools and a pain sum of 7 over the week, with a wellbeing sum of 14, one complication, no antidiarrheal use, and no palpable mass. Stool term: 2 × 14 = 28. Pain term: 5 × 7 = 35. Wellbeing term: 7 × 14 = 98. Complications: 20 × 1 = 20. A same-day hematocrit of 42% in a man gives a deficit of 47 − 42 = 5, worth 6 × 5 = 30. At 70 kg against a usual weight of 75 kg, the weight factor is (1 − 70/75) × 100 = 6.67. Summed: 28 + 35 + 98 + 20 + 0 + 0 + 30 + 6.67 = 217.7, mild-to-moderate disease by the later convention.
The same eight terms can land anywhere from 0 to well past 450. No symptoms, a normal hematocrit for the reference sex, and weight exactly at the usual figure sum to 0 — deep remission. Push every input toward its severe end instead — 30 stools, a pain sum of 15, a wellbeing sum of 20, three complications, antidiarrheal use, a definite mass, a hematocrit of 35% against an expected 47%, and 15 kg under a 75 kg baseline — and the identical arithmetic totals 507.0, deep in the range the 1976 paper called extremely severe.
Questions
What does a CDAI score actually mean?
The 1976 paper directly defines only two points on the scale: 150 or below is quiescent disease, and above 450 is extremely severe. The four-tier reading widely quoted today — remission ≤150, mild-to-moderate 150–219, moderate-to-severe 220–450, severe >450 — is a convention later clinical trials adopted to standardize enrollment criteria, not a threshold the original derivation study itself specified. Treat the middle bands as trial custom rather than a rule stated in Best et al.
Why does the formula include hematocrit and weight, not just symptoms?
A score assembled purely from a symptom diary could drift with how generously or harshly a patient rates their own week. Hematocrit and weight are measured at the clinic visit, not self-reported, so anchoring a third of the point weighting to them keeps the total from being pushed around by diary wording alone. It is also clinically sensible: active Crohn's disease often drives anemia and unintentional weight loss alongside the symptoms patients feel day to day.
What counts as a complication in the CDAI?
Six categories, one point each toward the complications count before the ×20 weighting: arthritis or arthralgia; iritis or uveitis; erythema nodosum, pyoderma gangrenosum, or aphthous stomatitis; anal fissure, fistula, or abscess; any other fistula; and fever above 37.8°C during the week being scored. A patient can score anywhere from 0 to 6 on this term alone.
Is CDAI still used in Crohn's disease trials today?
Yes — it remains the most common enrollment and endpoint measure in Crohn's disease drug trials, largely for continuity with decades of prior data using the same instrument. It has real limitations: two of its heaviest-weighted terms are self-reported wellbeing and pain, and it correlates only loosely with what an endoscope actually shows in the bowel, which is why many trials now report an endoscopic score alongside it rather than instead of it.
Does a low CDAI mean the bowel has healed?
Not necessarily. CDAI was built to track a clinician's overall severity impression, not mucosal appearance, and studies comparing the two find meaningful numbers of patients in clinical remission by CDAI who still show active inflammation on colonoscopy. Symptom relief and endoscopic healing are related but distinct outcomes — a favorable CDAI is reassuring but not proof the gut lining has recovered.
Can the hematocrit deficit term go negative?
Yes. If a patient's hematocrit is above 47% (men) or 42% (women), the deficit term is negative and the 6× multiplier subtracts from the total rather than adding to it. The same is true of the weight factor if current weight exceeds the usual/standard figure entered — both terms are designed to swing either direction, not just contribute upward.
References
- Best et al. 1976, Gastroenterology — original CDAI derivation study (PubMed)
- NIHR HTA report — CDAI calculation table adapted from Best et al. (NCBI Bookshelf)
Read this first: This instrument computes a screening figure from population formulas — it is not a diagnosis, and it cannot see the whole picture a clinician can. Use it to inform a conversation, not to replace one.