Deep Dive: DDO vs. PICO (Aligned with DEPTh + Design Triad)
🧠 Think "DEPTh" First—Not Framework
Before choosing DDO or PICO, classify the clinical challenge:
| Type | Description | Examples |
| Diagnosis | What feature helps us identify X? | What predicts scrub typhus? |
| Etiognosis | What causes or increases risk? | Does PM2.5 cause preterm birth? |
| Prognosis | What predicts outcome? | What predicts ICU mortality in sepsis? |
| Therapeutic | Does intervention improve outcome? | Does metronomic chemo increase OS? |
| Methodologic | How should we best measure/study? | Is this tool valid/reliable? |
🚦 Match framework to challenge type:
- Use DDO for diagnosis, etiognosis, prognosis, and most therapeutic cohort studies.
- Use PICO for randomized therapeutic trials.
🧩 Framework Dissection: DDO vs PICO
| Feature | DDO | PICO |
| Full Name | Domain–Determinant–Outcome | Patient–Intervention–Comparator–Outcome |
| Best for | Observational, causal, prognostic, diagnostic | Randomized therapeutic interventions |
| Structure | Who → What X → What Y | Who → Treatment A vs B → What Y |
| Analysis Mapping | Easy to plug into `Y = f(X | confounders)` | Direct for superiority/inferiority analysis |
| Flexibility | High: allows causal/predictive logic | Narrow: fits RCT-specific logic |
🔍 Secret Insight: PICO assumes a "treatment effect" structure. If you’re not randomizing, PICO can mislead your method logic.
🧪 Advanced Examples with CECS Mapping
🌱 Causal Inference (Etiognosis)
"Does high air pollution exposure increase preterm birth risk?"
- DEPTh: Etiognosis
- DDO:
- Domain: Pregnant women in polluted cities
- Determinant: PM2.5 exposure
- Outcome: Preterm delivery
- Occurrence Equation:Preterm = f(PM2.5 | Age, BMI, Comorbidities)
- Design:
- Method: Retrospective cohort
- Analysis: Multivariable logistic regression
🧬 Diagnostic Accuracy
"What symptoms help distinguish leptospirosis from other febrile illnesses?"
- DEPTh: Diagnosis
- DDO:
- Domain: Suspected febrile illness cases
- Determinants: Mud exposure, conjunctival suffusion, WBC count
- Outcome: Lab-confirmed leptospirosis
- Analysis: ROC curve, AUC, LR+, LR−
💉 Therapeutic RCT (PICO justified)
"Does metronomic chemo extend survival in unfit AML patients vs best supportive care?"
- DEPTh: Therapeutic
- PICO:
- P: Unfit AML patients
- I: Metronomic chemo
- C: Best supportive care
- O: Median survival at 6, 12 months
- Design: Pragmatic RCT
- Analysis: Kaplan–Meier curves, Cox model
🧠 Why Bad Questions Derail Good Research
🛑 Common Errors:
- PICO logic used for prognostic studies → misfit
- DDO variables unclear → domain/determinant conflated
- Outcome not operationalized → “better outcome” ≠ measurable endpoint
🔍 Secret Insight: Ask yourself, “Can I sketch an occurrence equation?” If not, the framework isn't ready.
✅ Key Takeaways
- DDO is your Swiss Army knife for most study types, especially observational, prognostic, and diagnostic.
- PICO is laser-focused: use when testing an intervention in RCT logic.
- DEPTh helps you align clinical logic to research logic.
- Research design flows from clinical challenge → structured question → objectives → design → analysis.
- A solid occurrence equation helps align all decisions downstream.