Satisficing in Diagnosis¶

Why It Matters¶

The drive for diagnostic precision can delay treatment unnecessarily. Often, you don't need to know the exact pathogen causing pneumonia - you need to know "bacterial pneumonia vs. other" to decide on antibiotics. Satisficing in diagnosis corrects the systematic error of treating all diagnostic questions as requiring maximum precision when many require only adequate categorization for action.

The Key Move¶

Before pursuing diagnostic refinement, ask: "What decision does this diagnosis enable? What level of precision does that decision require?" If the treatment is the same for all conditions in a category, diagnose the category and treat, rather than pursuing subcategorization. If subcategorization would change management, pursue it. Match diagnostic precision to decision requirements, not to epistemic curiosity.

Classic Application¶

Pneumonia etiology. Pneumonia can be bacterial (many species), viral (many types), fungal, or atypical. Precise microbiological diagnosis requires culture and sensitivity, taking days. But treatment decision only requires "community-acquired bacterial pneumonia vs. other." Treat empirically for common bacteria with broad-spectrum antibiotics. If patient doesn't improve, then refine diagnosis. The initial decision doesn't require precision - it requires adequate categorization.

Surprising Application¶

User behavior classification. Users are experiencing errors. Precise diagnosis: which specific code path, input combination, and environment configuration causes the error? But treatment decision only requires: "server-side bug vs. client-side vs. user error." Each category has different response (deploy fix, update client, improve error message). Initially categorize broadly, then refine only if broad-category treatment fails. Diagnostic precision matches decision requirements.

Failure Modes¶

Under-diagnosing when precision is actually needed. Some treatments have narrow therapeutic windows or serious side effects, requiring precise diagnosis before use. The tool isn't "avoid diagnosis" - it's "match diagnostic effort to decision requirements." Also fails when satisficing on category leads to suboptimal treatment that wouldn't have been necessary with better diagnosis. The worst failure is using satisficing to justify intellectual laziness rather than efficiency - being incurious when curiosity would improve outcomes.

Go Deeper¶

Simon, H.A. (1956). "Rational Choice and the Structure of the Environment," Psychological Review, 63(2), 129-138; Djulbegovic, B. et al. (2011). "Rational Decision Making in Medicine," Medical Decision Making, 31(1), 20-34.