EAHP BOOST 2026 | SYNTHETIC DEMONSTRATION
Antimicrobial review within 72 hours: checked results

This dataset is entirely invented for teaching. Each row represents one fictional antimicrobial treatment episode. Before and After are hypothetical periods around an imagined stewardship process. Review_72h records whether review within 72 hours was documented (Yes/No); an empty cell means the outcome is unknown. Ward is a fictional broad ward category. Episode_ID is a synthetic teaching identifier, not a patient ID.

                             Before   After
All fictional episodes             30      30
Documented review (Yes)           14      21
No documented review (No)         14       7
Missing outcome                    2       2
Non-missing outcomes              28      28
Yes among non-missing          50.0%   75.0%

Calculation: 14 / (14 + 14) = 50.0%; 21 / (21 + 7) = 75.0%. The observed difference is 75.0% minus 50.0% = +25.0 percentage points. This is a 25 percentage-point difference, not a 25% relative increase.

The figures were independently checked by counting the CSV rows and fields. There are 60 records and 60 distinct synthetic episode IDs. There are no unexpected period or outcome values in this file.

Missing outcomes are excluded from each valid denominator, not treated as No. This is only a descriptive complete-case comparison. The data are simulated, not evidence of effectiveness. In a real project, assess the reasons for missingness, eligibility, repeated patients, other changes over time, case mix and other threats to validity. An observed before/after difference does not establish that the intervention caused it. Discuss design and analysis with a statistician and follow local governance.