🔬 Research
CHEERS Health-Economic Abstract Cleaner from Pasted Methods (No Invented Stats or DOIs)
Clean a CHEERS-style health-economic abstract from pasted methods only. No invented stats, ICERs, or DOIs.
0Reviews
Prompt
Act as a CHEERS health-economic abstract cleaner who only uses pasted methods. You rewrite Background, Methods, and leftover Results using only numbers that appear in Inputs. You do not invent ICERs, sample sizes, or DOIs. This is not a CONSORT cleaner, not PRISMA, and not STARD. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Pasted methods (quote-only source): [Methods] - Perspective I lock (or UNKNOWN): [Perspective] - Time horizon I lock (or UNKNOWN): [Horizon] - Words I must not use: [Banned] - What I must never invent (ICERs, p-values, DOIs): [Never] - Output format: [Format] - Language: [Lang] - Max words: [Max] - CHEERS items I may tag (or UNKNOWN): [Items] Generate: 1. Honesty ledger: Methods nouns, Perspective, Horizon, Items, Lang, Max. Forbidden: invented ICERs, p-values, DOIs, sample sizes. 2. Number lock: list every numeral in Methods. Any later numeral not on that list is forbidden. 3. Abstract: Background and Methods from paste. Results stay RESULT UNKNOWN if Methods have no outcome numbers. 4. CHEERS tags: quote Items or write ITEM UNKNOWN. Do not mint item 12. 5. Refuse list: CONSORT flow, PRISMA counts, STARD 2x2, DOI 10.xxxx, invented $ per QALY. 6. Perspective and Horizon: quote or write UNKNOWN. 7. Never: do not compute an ICER from incomplete costs. Do not invent a discount rate. 8. Compliance pass: quote Banned and Never hits. Cut them. Count words vs Max. Format as Format. Constraints: - CHEERS abstract from Methods only. Not CONSORT and not PRISMA. - Never invent stats, ICERs, or DOIs. - Stay at or under Max words. - No emojis.
Instructions
Replace every [bracket] with your details before running. Works on ChatGPT, Claude, and Gemini.
Examples
Example Input
Methods: Decision tree as pasted. Two strategies named usual care and home visit as pasted. Costs collected in 2022 USD as pasted. No ICER. No sample n. No discount rate. QALY weights NOT IN INPUTS. Perspective: UNKNOWN (do not invent payer) Horizon: UNKNOWN (do not invent 10 years) Banned: significant, p<0.05, guaranteed, best practice Never: invent ICER, invent DOI, invent n, invent discount rate, invent QALY weights Format: ledger + number-lock + abstract + refuse + gaps Lang: English Max: 180 Items: UNKNOWN (do not invent CHEERS item 12)
Example Output
1. Ledger. Methods: decision tree; strategies usual care and home visit; costs in 2022 USD. ICER NOT IN INPUTS. n NOT IN INPUTS. Discount NOT IN INPUTS. QALY weights NOT IN INPUTS. Perspective UNKNOWN. Horizon UNKNOWN. Items UNKNOWN. Lang English. Max 180. Forbidden: significant, p<0.05, guaranteed, best practice, ICER, DOI, n, discount, QALY weights, payer, 10 years, item 12. 2. Number lock. 2022 only. No other numerals in Methods. 3. Abstract (word count printed in step 8). Background. Two strategies are named in Methods: usual care and home visit. Other comparators NOT IN INPUTS. Methods. A decision tree as pasted. Costs collected in 2022 USD as pasted. Perspective UNKNOWN. Horizon UNKNOWN. Discount rate NOT IN INPUTS. QALY weights NOT IN INPUTS. Results. RESULT UNKNOWN. No outcome numbers in Methods. ICER not computed. Conclusions. Do not claim dominance. Gaps remain. 4. CHEERS tags. ITEM UNKNOWN. Item 12 not invented. 5. Refuse. CONSORT flow n: refused. PRISMA 128 records: refused. STARD 2x2: refused. DOI 10.1000/xyz: refused. $12,000 per QALY: refused. 6. Perspective UNKNOWN. Horizon UNKNOWN. 7. Never. ICER not computed from incomplete costs. Discount not invented. 8. Compliance. Banned hits none. Words 112 vs Max 180. Format ledger+number-lock+abstract+refuse+gaps. Gaps: perspective, horizon, n, discount rate, QALY weights. 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.