Overview

The therapeutic index (TI) is a numerical expression used in pharmacology and toxicology to describe how safe a drug is relative to its effective dose. It compares a dose that produces an undesirable or harmful effect in part of a population with the dose that produces the intended therapeutic effect. In broad terms, a higher TI indicates a wider safety margin between effective and harmful doses.

How it is calculated

There are several related ways to express the TI. In classical toxicology the ratio is often given as the median lethal dose (LD50) divided by the median effective dose (ED50): TI = LD50 / ED50. In clinical contexts practitioners generally use a toxic dose (TD50) rather than an LD50, so TI = TD50 / ED50. Each component is a statistical measure that represents the dose at which a certain percentage (commonly 50%) of subjects show the specified effect.

The therapeutic index is one of several metrics used to represent drug safety. Others include the therapeutic window or therapeutic range, which refers to the concentration range in blood associated with efficacy without unacceptable toxicity, and various safety margins that compare doses producing rare adverse effects to doses producing near-universal efficacy. These related measures can be more useful clinically because they focus on probabilities relevant to individual patients rather than population medians.

Clinical importance and monitoring

TI is particularly important for drugs with a narrow safety margin, where small changes in dose or blood levels can cause toxicity. Examples commonly cited as narrow therapeutic index drugs include digoxin, lithium, warfarin, certain anticonvulsants (e.g., phenytoin), some antibiotics (e.g., aminoglycosides), and theophylline. For such agents, therapeutic drug monitoring, dose adjustments based on blood concentrations, and careful attention to interactions and patient factors (age, kidney or liver function) are standard practice.

Limitations and interpretation

The TI has important limitations. It is derived from population statistics and does not predict outcomes for an individual patient. Measures such as LD50 are usually obtained in animal studies and cannot be ethically determined in humans. Interindividual variability, drug interactions, comorbidities and formulation differences can all change the effective or toxic doses in a given person. Consequently, clinicians use TI alongside clinical judgment, monitoring, and other safety metrics rather than as a sole determinant of dosing decisions.

Notable facts

  • The concept of comparing beneficial and harmful doses is rooted in toxicology’s long-standing principle that the dose determines the poison.
  • A large TI suggests a broader margin of safety, but does not eliminate the risk of rare or idiosyncratic adverse reactions.
  • Regulatory guidance and prescribing practice pay special attention to medications with narrow therapeutic indices to reduce the risk of harm.