Troubleshooting Inconsistent Batch Performance in Pharmaceuticals

The core of pharmaceutical production is the concept of batch consistency. Every batch must conform with all predetermined specifications regarding quality, purity, potency, dissolution, uniformity, and stability of the product. Noncompliance of batches, even if they were produced in the same manner, raises questions regarding the reliability of the manufacturing process and its control over the operation.

Inconsistent batch behavior creates one of the biggest challenges to manufacturing staff including quality assurance and validation personnel. The variation may not result in an OOS situation but having variability in batches should be taken as a signal that some of the process parameters are not under control.

Investigations have shown that it is common for organizations to look first at the final test results instead of concentrating on the manufacturing process. It is essential to define not only the reason for the deviation in one batch but also understand what caused the process to create the variability. Meticulous investigation and use of all possible data are necessary for finding the root cause and fixing the issue.

Recognizing Inconsistent Batch Performance

There are many forms batch inconsistency can take. For example, even if all specifications for a batch are met, the batch can behave differently during the production of the product. In addition, some variability may only become evident during stability tests or through complaints from customers. Common signs of problems with batches include:

• Assay amounts that vary
• Dissolution profiles that are inconsistent
• Lack of uniformity in the content of the batch
• Not enough variation of the blend
• Variation in the weight of tablets
• Unstable compression forces
• Coating problems
• Differences in drying time
• Variability in granulation process
• Different yields

Before a recurring variation becomes an important quality problem, it should be investigated.

Understanding Past Production

In order to understand the limitations of the batch, one should compare it to other batches that were produced successfully in the past. One should analyze:

• Data from the last 10-20 batches
• Validation batches
• Engineering batches
• Scale-up batches
• Stability batches

Trend analysis helps disclose gradual changes of the processes that may be not visible when analyzing a single batch. It is important to compare:

• Time of production
• Equipment used in the manufacture
• Shifts of operators
• Conditions affecting production
• Raw materials used
• Utility performance
• Process parameters

Historical comparisons frequently narrow the investigation much faster than isolated testing.

Check Raw Material Consistency

Raw material variability is still one of the leading causes of poor performance in manufacturing. Even if the materials are qualified according to specifications, path of the production may be impacted by the differences between suppliers or production lots.

Check:

• API particle size distribution
• Moisture content
• Bulk density
• Flow properties
• Polymorphic form
• Excipient grade
• Supplier changes
• Trends of Certificates of Analysis

For instance, a small change in the API particle size can lead to a drop in the dissolution while higher moisture content in lactose may have a big impact on wet granulation. Do not assume that "in specifications" means "the same."

Review Critical Process Parameters

All the processes that have been validated in manufacturing will have some CPPs that impact the CQAs. Direct comparisons of actual manufacturing data are done against the defined ranges of operation.
Some examples of CPPs include:

Process	Critical Parameters
Blending	Mixing time, blender speed, fill level
Granulation	Binder addition rate, endpoint, impeller speed
Drying	Temperature, airflow, endpoint moisture
Milling	Screen size, rotor speed
Compression	Compression force, turret speed, feeder speed
Coating	Spray rate, inlet temperature, atomization pressure

Gaps in performance can still occur due to the combination of minor deviations occurring together even if the individual parameters remain within the defined operating range.

Checking the Equipment Efficiency

Mechanical equipment does not fail suddenly. The performance of such equipment declines in most cases steadily. Make sure to examine the equipment history for:

• Overall preventive maintenance
• Calibration status
• Repair reports
• Equipment alarm events
• Sensor failure events
• PLC events
• Vibration history
• Lubrication history

Be particularly concerned with those parts that have a direct influence on process steadiness:

• Load cells
• Temperature sensors
• Pressure sensors
• Spray nozzles
• Compression rollers
• Feeding systems

If there is a partially clogged spray nozzle, for instance, the granulation will be unstable even if the process parameters are correct.

Assess Environmental Conditions

The environment is significant in pharmaceutical processes, especially those of solid dosage manufacture. Consider the following:

• Temperature
• Humidity
• Pressure
• Ventilation
• HVAC Alarms

Various issues are included in this area:

• Excessive humidity may cause powder to stick.
• Low humidity causes an increase in electrostatic charge.
• Variations in temperature cause variations in viscosity.
• Variations in airflow cause variations in coating processes.

Environmental trends have to be always analyzed together with relevant manufacturing data.

Examine Human Factors

Performance of operators must be analyzed fairly, avoiding presumption of human mistakes being the main cause. Think about:

• Were Standard Operating Procedures followed properly?
• Did operators have training up to date?
• Was there any manual intervention?
• Were the changes of shifts influential on production?
• Were all significant observations recorded?

If dependency on operator decisions is too much, then there is something that needs to be improved in the production process.

Use Statistical Trend Analysis

It often happens that data uncovers patterns which initially are not obvious. Some statistical methods include:

• Control charts
• Process capability analysis (Cp/Cpk)
• Pareto analysis
• Regression analysis
• Histograms
• Trend charts

For example, the value of compression force may meet the standard, but it may show gradual increase for several batches indicating that machinery is getting worn. Statistical analysis converts solitary observations into relevant process knowledge.

Investigate Utility Performance

Utility performance has a direct impact on product consistency; however, it is often neglected in an investigation. Consider the following:

• Compressed air quality
• Purified water quality
• Steam pressure
• Chilled water temperature
• HVAC performance
• Stability of electrical supply

A momentary drop in compressed air pressure can affect the tablet compression process without triggering an alarm.

Investigate Utility Performance

Utilities affect production consistently, but their role is not always recognized in the analysis. The evaluation involves:

• Quality of compressed air
• Quality of water
• Pressure of steam
• Temperature of cooling water
• Performance of HVAC
• Stability of electricity

A brief disturbance in the operation of compressed air will change the process of tablet compression or spraying without causing the equipment alarm report.

Apply Structured Root Cause Analysis

Avoid making assumptions and jump to conclusions. Various tools can be used for investigation such as:

1. Fishbone Diagram

Look for the potential causes in several categories including.

• Materials
• Equipment
• Methods
• People
• Environment
• Measurement

This method allows for more thorough reasoning and less bias of investigator.

2. Five Whys Method

Example Issue: Variations in the hardness of the tablet.

1. Why was it so? Compression force was inconsistent.
2. Why was it so? The feeding frame delivered inconsistent powder flow.
3. Why was it so? Properties of powder flow might have been changed.
4. Why was it so? Moisture in granules changed.
5. Why was it so? End sensor of the drying unit should be adjusted or calibrated.

Review Process Validation Data

Validated processes can serve as a useful measuring rod for troubleshooting. Compare production batches with validated batches and look for:

• Trends in the critical process parameters (CPP)
• Results of the critical quality attributes (CQA)
• Data from sampling
• In-process controls used
• Conclusions from validation

If the commercial process consistently behaves differently from the validated process, then continuous process verification or revalidation may be needed.

Enhance Process Monitoring

Research usually shows that current monitoring systems identify deviations quite late in the process. Implement some of the below-mentioned methods:

• Monitoring the process in real-time
• Applying Statistical Process Control technique
• Using Process Analytical Technology
• Setting up advanced alarm limits
• Utilizing electronic batch trends

Early detection gives way for correcting the process earlier than it damages product quality.

Develop Effective CAPA

Corrective measures should restore the origin of the issue.
Examples:

• Reworking process parameters.
• Revising manufacturing standard operating procedures.
• Training employees.
• Modifying specifications of raw materials.
• Requalifying machinery.
• Revising preventive maintenance period.
• Strengthening supplier qualification.
• Adding more process monitoring.

Every CAPA should include verification of efficiency of the measures taken to rule out any deterioration in batch consistency.

Common Investigation Mistakes

There is a number of repeatable mistakes which reduce the efficiency of batch investigations. The following mistakes should be avoided:

1. Focusing exclusively on lab results.
2. Presuming that the last step of the process was responsible.
3. Not paying attention to historical batches.
4. Blaming workers without much evidence.
5. Investigating the departmental operation and not looking into the process as a whole.
6. Closing an investigation before validating.
7. Considering isolated variations as an accidental incident.

Investigation activity should be based on evidence and not on preconceptions.

Building a More Robust Manufacturing Process

To decrease variability in batch performance, process reliability must be integrated into the standard operations processes. The following methods are recommended:

• Ongoing verification of all processes.
• Review of processes every year.
• Regular review of CPPs and CQAs.
• Regular evaluation of vendor performance.
• Cross-disciplinary operations with Production, QA, QC, Engineering and Validation.
• Risk-influenced review after changes in the installation or formulation.
• Regular control of process compliance index with standards.

Organizations that follow continuous process monitoring allow to notice inconsistencies before they turn into more serious problems, providing higher level of manufacturing process reliability in the future.

Batch performance that is inconsistent is not generally attributable to one incident per se. More often than not, the inconsistency is related to the interaction of raw material differences, equipment status, process parameters, environmental impacts, and human factors. When it comes to resolving this issue, it is best to examine not just the final test results, but the entire process rather than just the finished product. This needs to be stressed and it is necessary to conduct an investigation that is systematic and uses data analysis techniques as mentioned above.

Based on my experience, the best pharmaceutical manufacturers are not the ones that have no process variations, but the ones that find out about them early enough and carry out a proper investigation of these phenomenon.

Share

Tweet

Share

Pin it

Share