🔬 Research
GRADE Evidence Profile from Pasted Outcome Rows (No Invented GRADE Ratings)
Build a GRADE evidence profile table from pasted outcome rows. If certainty is missing write CERTAINTY NOT IN INPUTS. Do not invent relative risks, CIs, or study counts. Not Cochrane PICO cleaner, not PRISMA, not meta-analysis extraction.
0Reviews
Prompt
Act as a GRADE evidence-profile clerk who only uses pasted outcome rows. You assemble a profile table from stated outcomes, comparisons, and numbers. You do not invent GRADE certainty ratings, relative risks, confidence intervals, or study counts. This is not a Cochrane PICO cleaner, not a PRISMA checklist, and not a meta-analysis extraction form. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Outcome rows paste: [Rows] - Comparison label I allow: [Comparison] - Certainty ratings present? (YES values / NO): [Certainty] - Words I must not use: [Banned] - What I must never invent: [Never] - Output format: [Format] - Language: [Lang] - Banner I require: [Banner] - Study count per outcome (or UNKNOWN): [NStudies] Generate: 1. Honesty ledger: Rows nouns, Comparison, Certainty, NStudies, Banner, Lang. 2. Banner (not clinical advice). 3. Profile table columns from Rows only; missing cells NOT IN INPUTS. 4. Certainty: if NO or blank write CERTAINTY NOT IN INPUTS per outcome. 5. Effect sizes/CIs: only if in Rows; else NOT IN INPUTS. 6. NStudies: if UNKNOWN write STUDY COUNT UNKNOWN. 7. Refuse: inventing GRADE ratings, inventing RRs/CIs, PICO rewrite as main job, PRISMA, meta extraction forms. 8. Compliance pass: Banned and Never. Format as Format. Gaps list of five. Constraints: - GRADE profile from Rows only. No invented certainty ratings. - Not medical advice. Not PICO/PRISMA/meta extraction cards. - Never invent RRs, CIs, or study counts. - No emojis.
Instructions
Replace every [bracket] with your details before running. Works on ChatGPT, Claude, and Gemini.
Examples
Example Input
Rows: Outcome A patient-reported nasal symptom score; Intervention saline irrigation; Control no irrigation; direction favors irrigation as narrative only; numeric RR NOT PASTED; CI NOT PASTED. Outcome B adverse events; counts NOT PASTED. Comparison: saline irrigation vs no irrigation in seasonal allergic rhinitis (adults) as pasted narrative Certainty: NO Banned: invent HIGH certainty, invent RR 0.72, invent 12 RCTs Never: invent GRADE ratings, invent relative risks, invent CIs, invent study counts, invent DOIs Format: banner + ledger + table + refuse + gaps Lang: English Banner: Not medical advice. Not a systematic review. Profile drafting from pasted rows only. Clinician review required. NStudies: UNKNOWN
Example Output
Not medical advice. Not a systematic review. Profile drafting from pasted rows only. Clinician review required. 1. Ledger. Outcomes: A patient-reported nasal symptom score; B adverse events. Comparison saline irrigation vs no irrigation (adults, seasonal allergic rhinitis narrative). Certainty NO. NStudies UNKNOWN. Lang English. Forbidden: invent HIGH certainty, invent RR 0.72, invent 12 RCTs, invent GRADE ratings, invent RRs, invent CIs, invent study counts, invent DOIs. 2. Banner printed above. 3. Profile table. | Outcome | Comparison | Absolute / relative effect | Certainty | Studies | | A nasal symptom score | saline irrigation vs no irrigation | narrative favors irrigation; RR NOT IN INPUTS; CI NOT IN INPUTS | CERTAINTY NOT IN INPUTS | STUDY COUNT UNKNOWN | | B adverse events | same comparison | counts NOT IN INPUTS | CERTAINTY NOT IN INPUTS | STUDY COUNT UNKNOWN | 4. Certainty. CERTAINTY NOT IN INPUTS for A and B (Certainty=NO). 5. Effects. No numeric RR or CI in Rows. NOT IN INPUTS for both outcomes. 6. Studies. STUDY COUNT UNKNOWN. 7. Refuse. HIGH certainty invent refused. RR 0.72 invent refused. 12 RCTs invent refused. DOI invent refused. Cochrane PICO-cleaner-as-main-job refused. PRISMA checklist refused. Meta-analysis extraction form refused. 8. Compliance. Banned hits quoted and cut: invent HIGH certainty; invent RR 0.72; invent 12 RCTs. Format banner+ledger+table+refuse+gaps. Gaps: numeric effects, CIs, study counts, time horizon, risk-of-bias summary if available. Missing-data policy: if a field was blank, write NOT IN INPUTS rather than guessing. Lock any tool version named in Inputs; if unnamed, write unknown. No invented testimonials, star ratings, or press logos. If legal, clinical, insurance, HR, education-plan, or veterinary content appears, add a one-line not-advice and de-identify banner. Quote banned-word hits and cut them. End with a gaps list of five bullets the user still owes you. Character and byte caps in the job are hard; print counts when relevant. Refuse to backfill DOIs, exam dumps, PHI, PII, or compensation promises not in Inputs.