Protocol Design

case-control-study-planner

Design a structured case-control study framework with explicit source population logic, control selection rules, matching decisions, exposure measurement planning, and bias-control checkpoints.

86100Total Score
Core Capability
87 / 100
Functional Suitability
11 / 12
Reliability
9 / 12
Performance & Context
7 / 8
Agent Usability
15 / 16
Human Usability
6 / 8
Security
12 / 12
Maintainability
11 / 12
Agent-Specific
16 / 20
Medical Task
24 / 25 Passed
86Case-control study for postoperative pulmonary complications after major abdominal surgery
5/5
86Hospital-based case-control study on antibiotic exposure and Clostridioides difficile infection
5/5
85Community-based matched case-control study on occupational exposure and rare lymphoma
5/5
84Biomarker measured after diagnosis — validity of using as etiologic exposure
4/5
84Complex pharmacoepidemiology case-control with multiple exposure windows and confounders by indication
5/5

Veto GatesRequired pass for any deployment consideration

Skill Veto✓ All 4 gates passed
✓
Operational Stability
System remains stable across varied inputs and edge cases
PASS
✓
Structural Consistency
Output structure conforms to expected skill contract format
PASS
✓
Result Determinism
Equivalent inputs produce semantically equivalent outputs
PASS
✓
System Security
No prompt injection, data leakage, or unsafe tool use detected
PASS
Research Veto✓ PASS — Applicable
DimensionResultDetail
Scientific IntegrityPASS
No fabricated references, DOIs, PMIDs, statistical values, or clinical data detected.
Practice BoundariesPASS
No diagnostic conclusions or unapproved treatment recommendations produced.
Methodological GroundPASS
Strong source-population discipline and matching overmatching warnings enforced
Code UsabilityN/A
No code generated; Category 2, design planning only

Core Capability87 / 100 — 8 Categories

Functional Suitability
Description is very brief (one sentence); does not adequately communicate the full scope and trigger contexts of the skill
11 / 12
92%
Reliability
Proceeds with conditional language when inputs are unclear; no active minimum-clarification protocol defined
9 / 12
75%
Performance & Context
Strong score (7/8); minor gaps noted.
7 / 8
88%
Agent Usability
Strong score (15/16); minor gaps noted.
15 / 16
94%
Human Usability
Very brief description reduces discoverability; users may not recognize this skill without knowing epidemiology terminology
6 / 8
75%
Security
Full marks (12/12); no significant issues detected.
12 / 12
100%
Maintainability
Strong score (11/12); minor gaps noted.
11 / 12
92%
Agent-Specific
Trigger precision low due to brief description; no dedicated progressive disclosure of reference modules in description
16 / 20
80%
Core Capability Total87 / 100

Medical TaskExecution Average: 85 / 100 — Assertions: 24/25 Passed

86
Canonical
Case-control study for postoperative pulmonary complications after major abdominal surgery
5/5 ✓
86
Variant A
Hospital-based case-control study on antibiotic exposure and Clostridioides difficile infection
5/5 ✓
85
Variant B
Community-based matched case-control study on occupational exposure and rare lymphoma
5/5 ✓
84
Edge
Biomarker measured after diagnosis — validity of using as etiologic exposure
4/5 ✓
84
Stress
Complex pharmacoepidemiology case-control with multiple exposure windows and confounders by indication
5/5 ✓
86
CanonicalPass
Case-control study for postoperative pulmonary complications after major abdominal surgery

5/5 assertions passed.

Basic 35/40|Specialized 51/60|Total 86/100
A1Case-control design fit explicitly assessed before proceeding with design
A2Source population defined such that both cases and controls arise from same population
A3Matching strategy decision explicitly justified or rejected with overmatching risk noted
A4Exposure timing and recall bias risks explicitly addressed
A5No fabricated event rates, registry details, or guideline endorsements
Pass rate: 5 / 5
86
Variant APass
Hospital-based case-control study on antibiotic exposure and Clostridioides difficile infection

5/5 assertions passed.

Basic 34/40|Specialized 52/60|Total 86/100
A1Incident vs prevalent case distinction addressed
A2Exposure window defined with temporal alignment to outcome
A3Bias-control matrix present with at least selection, recall, information, and confounding rows
A4Odds ratio vs risk ratio distinction addressed
A5No clinical utility claims made beyond study scope
Pass rate: 5 / 5
85
Variant BPass
Community-based matched case-control study on occupational exposure and rare lymphoma

5/5 assertions passed.

Basic 34/40|Specialized 51/60|Total 85/100
A1Rare outcome justification for case-control stated
A2Matching strategy assessed with analytic consequences stated
A3Conditional logistic regression recommended when individual matching is proposed
A4Recall bias for occupational exposure explicitly flagged
A5No post-outcome measurement accepted as valid baseline exposure
Pass rate: 5 / 5
84
EdgePass
Biomarker measured after diagnosis — validity of using as etiologic exposure

4/5 assertions passed.

Basic 34/40|Specialized 50/60|Total 84/100
A1Post-diagnosis biomarker correctly flagged as invalid etiologic exposure without qualification
A2Reverse-timing distortion risk explicitly named
A3Alternative study framing recommended when biomarker timing is post-outcome
A4No case-control design rejected outright without offering viable alternative
A5Primary recommendation given even when design fit is weak
Pass rate: 4 / 5
84
StressPass
Complex pharmacoepidemiology case-control with multiple exposure windows and confounders by indication

5/5 assertions passed.

Basic 35/40|Specialized 49/60|Total 84/100
A1Confounding by indication explicitly identified as major bias risk
A2Multiple exposure windows handled with explicit timeline structure
A3Primary statistical analysis line stated (conditional logistic or logistic regression)
A4Feasibility and interpretation limits separated from design recommendations
A5No fabricated drug prevalence, event rates, or guideline positions
Pass rate: 5 / 5
Medical Task Total85 / 100

Key Strengths

  • Strong source-population and sampling-frame discipline enforced throughout — prevents fundamental case-control design errors
  • Matching hard rules are excellent: no default matching, overmatching risk explicitly flagged, analytic consequences always stated
  • Five reference modules cover question-fit, case/control definition, matching/exposure, bias/analysis, and output style comprehensively
  • Explicit prohibition on using post-outcome biomarkers as etiologic exposures — a common error in clinical research