Home

Reminder

Treatment is not a decision:
it is a trial with checkpoints.

The principle Before starting treatment, always ask whether standard therapy is likely to fail. After starting, you must reassess at a fixed time point. The response determines whether you continue or escalate.
The reasoning Diseases behave differently in different patients. Some respond to standard therapy; others do not. Starting the same treatment without anticipating individual risk factors and reassessing response risks delay, complications, and harm.
The clinical mindset You are not just treating. You are testing a hypothesis: "Will this patient respond to this treatment?" The follow-up check is the answer to that question.
The Core

Management follows a structured sequence in many common diseases, though the specific steps vary by context and disease behavior.

The Framework

01
Anticipate
Assess clinical features that suggest the patient is at risk of non-response, atypical disease course, or the need for modified first-line therapy.
02
Treat
Based on the risk, begin appropriate therapy, selecting agent, dose, and route to fit the clinical context.
03
Reassess
Evaluate response at a predefined time point against the criterion defined before treatment started.
04
Decide
Continue if response meets the predefined criterion; escalate, de-escalate, or reframe the diagnosis if it does not.
01

Clinical Examples

Worked examples: the sequence applied

Three diseases, one framework

Disease
Nephrotic Syndrome (Child)
→
Anticipate → Treat → Reassess → Decide
Assess for atypical features predicting steroid resistance →
If no atypical features: Start steroids → 4–6 weeks: Urine protein in remission? → If no remission: investigate further and consider referral.
If atypical features: Investigate further and consider referral.
Atypical features: onset before age 1 or after age 12, macroscopic hematuria, hypertension, or impaired renal function at presentation — each one raises the chance that steroids alone will not work.
Disease
Acute Lymphoblastic Leukemia (Child)
→
Anticipate → Treat → Reassess → Decide
Risk stratification → Chemotherapy → Day 28 MRD? → If positive: Intensify
MRD (minimal residual disease): the fraction of leukemic cells still detectable after induction chemotherapy, measured because it predicts relapse risk far more reliably than a normal-looking bone marrow film alone.
Disease
Tuberculosis (Adult)
→
Anticipate → Treat → Reassess → Decide
Check MDR risk → Anti-TB therapy → 2 months: Sputum AFB → If positive: modify
MDR risk: prior TB treatment, known contact with a drug-resistant case, or poor response so far — these raise suspicion of multidrug-resistant disease before standard first-line therapy is even started.
02

Common mistakes

No anticipation

Starting treatment without assessing individual risk factors that may predict non-response or require a modified plan.

Before
Problem: Wrong starting plan leads to early failure.

No reassessment

Continuing treatment without checking response.

After
Problem: Delay in recognizing treatment failure harms patients.
Safety

The most dangerous mistake is not choosing the wrong treatment. It is failing to recognize that the treatment is not working.

Timeline Principle

Every disease has an expected response window. Knowing this is essential. If the patient has not improved within that window, assume failure until proven otherwise.

The Core Questions

Before
Will this patient fail standard treatment?
After
Did the treatment work within the expected time?
Decision
Should I continue or escalate treatment?
Common Errors

What Goes Wrong in Practice

Identifying the wrong step as violated Not all non-response is a Reassess failure. Sometimes the error happened at Anticipate, before treatment started.
Defining a time point without defining a criterion Knowing when to reassess is incomplete: you must also specify what success looks like before you measure.
Treating "no worse" as equivalent to "responding" Stability is not response: the framework requires a positive definition of improvement, not just absence of deterioration.
Defaulting to escalation when non-response may signal wrong diagnosis Before intensifying treatment, ask whether the diagnosis is correct; escalating the wrong treatment compounds the original error.
Applying the framework as a linear algorithm in emergency settings In time-critical situations, Treat precedes full Anticipate: the sequence is compressed, not abandoned.
In Short

Effective treatment requires three commitments before starting:
1. a hypothesis about likely response,
2. a specific criterion that defines success, and
3. a fixed time point at which that criterion will be measured,
so that non-response is recognized before it becomes irreversible.
Then, decide what to do next.

Key Takeaway
Good clinicians do not just start treatment.
They anticipate risk, monitor response against predefined criteria, and adapt early.

Management is not a single decision. It is a cycle:
Anticipate → Treat → Reassess → Decide.

Do not wait passively Response must be measured Non-response is information