Table 1.

 Rationale for monitoring initial response to drug treatment

Aims and scenario	Target	Examples of monitoring
Maximise benefits
Therapy has rapid effect on patient centred outcome	Patient centred outcome	Observation of size and appearance of wound infection after empirically chosen antbiotics Monitoring for resolution of symptoms after treatment for depression
Therapy has slow effect on patient centred outcome	Surrogate outcome	Monitor cholesterol after statins Monitor blood pressure after antihypertensives
Minimise harms
Therapy has rapid effect on patient centred outcome	Patient centred outcome	Monitor for cough after angiotensin converting enzyme inhibitors
Therapy has slow effect on patient centred outcome	Surrogate outcome	Monitor drug levels after renal transplant patient on cyclosporin-A Monitor white cell count after clozapine in patient with psychosis
Maximise compliance
Compliance with therapy uncertain	Adherence	Use directly observed treatment strategy to ensure compliance with anti-tuberculosis treatment Count patient’s pills at follow-up visit to estimate how many have been taken if patient is suspected to be non-compliant Measure levels of anticonvulsant drugs to assess compliance with treatment after breakthrough seizure