EAHP BOOST 2026 | DEMONSTRATION DATA AND CODE

FILES
synthetic-review-data.csv: 60 invented antimicrobial treatment episodes; no real patient data.
review-summary-vba.txt: a readable, previously prepared example macro.
expected-results.txt: counts, percentages and interpretation to check against.

QUESTION
In this fictional example, what proportion of episodes had a documented antimicrobial review within 72 hours in the Before and After periods? Describe the difference in percentage points. Review_72h is Yes, No or blank (unknown). Exclude blank outcomes from the proportion and report their number. This is not a causal analysis.

LIVE AI PROMPT
I am practising with a fictional dataset in desktop Excel. The headers are Episode_ID, Period, Ward, Review_72h. Period is Before or After; Review_72h is Yes, No or blank. Write VBA that reads the active worksheet, leaves the source unchanged, creates a clearly named new summary sheet only if one does not already exist, and reports Yes, No, blank, other outcome values, the valid denominator (Yes + No), the Yes proportion, and After-minus-Before in percentage points. Also report rows with a missing ID or unrecognised period. Do not delete or overwrite anything. Explain how to inspect, run and validate the macro. Use no patient data. Do not claim a before/after difference proves causation.

Ask the AI to explain its assumptions and show how it handles missing data, then review the code. The live output will vary; check its calculations independently. Do not paste confidential information into an AI service. Follow your organisation's approved data and tool policies.

HERE'S ONE I PREPARED EARLIER
1. Open the CSV in desktop Excel. Save a copy as an Excel Macro-Enabled Workbook (.xlsm), retaining the data on the first sheet. CSV cannot store macros. Excel for the web cannot run VBA; use the checked results and manual counts there instead.
2. In Excel, open the VBA editor (Windows: Alt+F11; Mac: Tools > Macro > Visual Basic Editor, if available). In the workbook's project, Insert > Module. Paste the content of review-summary-vba.txt into the module. Save the .xlsm copy.
3. Select the data sheet, run SummariseSyntheticReviews from the Macros dialogue, and compare each output to expected-results.txt. The macro creates a new sheet named Review summary and does not overwrite an existing one. Rename or remove a prior summary yourself before rerunning.
4. Cross-check the denominator manually: count Yes and No separately, then divide Yes by their sum. Missing data must be reported separately. Investigate any unexpected values and missing IDs. A missing result is not a No.

Macro availability and security settings differ by institution. Only run code you have reviewed and are permitted to run in a disposable, synthetic workbook. The prepared macro is a teaching example, not a validated clinical, research or production tool. It checks expected headers and counts unrecognised outcomes/periods and missing IDs; it does not check duplicate IDs or validate whether data are representative. Installation and execution have not been verified here. Test it in your own Excel environment using a separate synthetic workbook. If macros are blocked, review the code and use the checked results instead.