Quality by Design: CQAs, CPPs and the Design Space

Start from what the patient needs, work backwards

The old way of working was: develop a formulation, freeze a recipe, and treat any change as a threat. Quality by Design (QbD) starts from the opposite end. First you define what the product must do for the patient — its target profile — and from that you derive the measurable properties that matter, called [[critical-quality-attribute|critical quality attributes (CQAs)]]. For a tablet these might be its content uniformity, its hardness, and how it performs in a dissolution test.

Next you ask: which knobs on the process actually move those CQAs? Those knobs are the [[critical-process-parameter|critical process parameters (CPPs)]] — things like granulation water amount, drying temperature, or compression force. Most settings on a machine barely matter; QbD's job is to find the few that truly do, so attention goes where it counts.

The design space: a region, not a point

Once you know your CPPs and how they push the CQAs, you can map the combinations of settings that all give acceptable product. That multidimensional region is the [[design-space|design space]]. Inside it, you are guaranteed good medicine; you can move around freely to absorb the normal wobble of raw materials and equipment without leaving quality behind.

Simple 2-parameter design space (granulation example)
----------------------------------------------------
CQA target: tablet dissolution >= 80% in 30 min

Studied CPPs:
  Granulation water:    8% -- 14%  (of dry mass)
  Compression force:    8 kN -- 16 kN

Experiments (selected results):
  water 8%,  force 8 kN  -> 92% dissolved  PASS
  water 8%,  force 16 kN -> 84% dissolved  PASS
  water 14%, force 8 kN  -> 81% dissolved  PASS
  water 14%, force 16 kN -> 71% dissolved  FAIL (too dense + over-granulated)
  water 11%, force 12 kN -> 88% dissolved  PASS  <-- chosen set point

Design space (region that always passes):
  water 8%--13%  AND  force 8--14 kN
Set point sits inside, with room on all sides.
Normal-operating-range = a safe box drawn INSIDE the design space.

A toy design space: experiments map where dissolution stays acceptable, and the set point is placed comfortably inside.

This is liberating for regulators and makers alike. Moving within an approved design space is not considered a change, so the process can adapt without lengthy re-approval. And because you understand why each boundary exists, you can watch the right things in real time using process analytical technology (PAT) — sensors that measure CQAs as the batch runs, instead of waiting for the lab.