# Streamline in-process controls without risk.

Sampling and measuring every 30 minutes "because that's how it's always been done" bloats the batch record and the teams' workload — without reducing risk when the process is capable and stable. The Sinfony method **scales in-process control (IPC) frequency to real risk**, with evidence to back it, and no compromise on compliance. Lessons learned from a pharmaceutical site in Morocco.

0 OOS/OOT

over 3 years, on the parameters checked in-process: the historical evidence that justifies streamlining the frequency — not a hunch.

Sinfony lessons learned — pharmaceutical site in Morocco (3-year review).

An **in-process control** (IPC) checks a parameter during production. Multiplying these controls feels reassuring, but once the process is proven **capable and stable**, most of them no longer add any information: they lengthen the batch record and tie up the teams. Streamlining doesn't mean loosening — it's about **concentrating the effort where the risk requires it**.

We test at the frequency of habit, not the frequency of risk.

IPC frequencies frozen for years, never tied to the process's actual capability or to the deviation history: that's operator and quality workload, a batch record that swells, a review that drags on — for near-zero added value when nothing drifts. Over-control isn't safety, it's waste in disguise.

## Justify the streamlining, step by step.

Every decision is tied to verifiable data — never to an isolated average.

1

### Historical analysis

Review of quality reports over 3 years and use of the OOS/OOT history.

2

### Product / process review

Product criticality and selection of the "worst case" scenarios to challenge first.

3

### Machine capability

The equipment's ability to hold the critical parameters within specification (Cpk).

4

### Risk analysis (FMEA)

Assessment of the risks and of the machine safeguards actually in place.

5

### New frequency (AQL)

A streamlining decision backed by an AQL sampling plan, documented and traceable.

[The proof through capability (Cpk) →](/en/blog/process-capability-cpk-streamline-controls/) [The AQL sampling plan →](/en/blog/aql-sampling-plan-batch-record/)

## Fewer controls, same command.

At a pharmaceutical site in Morocco, the proof was in the data.

30 min → 1 h

control frequency streamlined, with evidence to back it (2 h in phase 2).

0 OOS/OOT

over 3 years on the parameters concerned: a stable process.

High Cpk

a highly capable, well-centered process: drift unlikely between two controls.

A cascading benefit: fewer samples means **workload given back** to operators and quality, a **shorter batch record** that's therefore reviewed faster, and a **release lead time** that tightens — all without degrading compliance, since the decision is traced and justified.

## Proportion the control, everywhere.

[Batch record

### Reduce batch release time

A leaner batch record is reviewed faster: the same fight, on the flow side.

Read the article →](/en/blog/reduce-batch-release-time-batch-record/) [Quality control

### Streamline incoming controls

Same logic, on the raw-materials side: control proportioned to risk.

Read the article →](/en/blog/reduce-qc-controls-risk-analysis/) [Consulting

### Operational performance

Assess the workload and de-risk your flows, from €5k.

See the approach →](/en/consulting/operational-performance/)

## Streamlining IPC, in practice.

What is an in-process control (IPC)? +

An IPC (in-process control) is a check performed during production — weight, hardness, volume, etc. — to make sure the process stays within specification. Its frequency should depend on the risk and the stability of the process, not on a frozen habit.

Is streamlining controls compatible with GMP? +

Yes, as long as the decision rests on a documented risk analysis (history, process capability, FMEA) and a justified sampling plan. You don't remove the control: you adjust its frequency to the real risk, keeping traceability and rollback criteria.

What is the justification based on? +

On five building blocks: the review of the quality history and of OOS/OOT over several years, product criticality, machine capability (Cpk), an FMEA of the risks and safeguards, then an AQL sampling plan. The decision is tied to verifiable raw data (batch records, logbooks, deviations), never to an isolated average.

What does it save? +

Workload given back to operators and quality, a shorter batch record that's reviewed faster, and a tightened release lead time — without degrading compliance. In our lessons learned, the frequency went from every 30 minutes to every hour, on a process with zero OOS/OOT in three years.

## Stop over-controlling what doesn't drift.

Let's analyze your IPC and the workload you could streamline, with evidence to back it — no compromise on compliance.