“Customer received 98 tablets instead of 100”
“This is the 3rd complaint for short fill this month”

Count variation complaints are one of the most common and serious issues in solid dosage bottle packing. They directly impact patient trust, regulatory standing, and cost.

In 2026, with serialization and AI inspection, there is zero tolerance for count errors. Let’s break down why they happen and how to prevent them permanently.

1. What is Count Variation in Bottle Packs?

Count Variation = The number of tablets/capsules in the bottle is different from the declared count on label.

Types:

• Short Fill: 98 tablets instead of 100. Most common complaint.
• Over Fill: 102 tablets instead of 100. Cost + regulatory risk.
• Mixed Count: 100 tablets but 2 are broken/chipped pieces counted as 1.

2. Top 8 Root Causes of Count Variation

Cause 1: Counting Machine Malfunction

Issue: Dust on sensor, faulty photoelectric eye, vibration settings wrong.

Result: Machine skips 2-3 tablets per bottle.

Cause 2: Product Related Issues

Issue: Excessive fines, broken tablets, sticking capsules, static charge.

Result: Tablets don’t fall properly. Chute jamming.

Cause 3: Bottle Related Issues

Issue: Wrong bottle size, neck finish variation, oval bottles.

Result: Bottle not aligned under chute. Partial filling.

Cause 4: Feeding & Singulation Problem

Issue: Vibratory bowl amplitude too high/low. Double feeding.

Result: 2 tablets counted as 1 or 1 tablet missed.

Cause 5: Human Error During Changeover

Issue: Wrong count recipe selected in HMI. Operator forgot to clear previous batch tablets.

Result: 60 count recipe used for 100 count product.

Cause 6: Lack of In-process Checks

Issue: IPC done every 1 hour instead of every 30 min. No checkweigher installed.

Result: 5000 bottles produced with error before detection.

Cause 7: Rejects Rework

Issue: Rejected bottles from previous shift manually refilled without re-counting.

Result: Overfill or underfill goes to market.

Cause 8: Post-Packing Handling

Issue: Bottles dropped during transportation. Cotton/Desiccant absorbed tablets.

Result: Customer receives short count but it happened in distribution.

3. How to Investigate a Count Variation Complaint: 5 Step Approach

1. Quarantine & Data Collection
   Hold remaining stock of same batch. Collect: Batch record, IPC data, machine log, CCTV footage, complaint sample.
2. Verify the Complaint
   Open complaint bottle in QA. Video record. Count manually 3 times. Check for broken pieces.
3. Check Retained Samples
   Open 20 bottles from retain. If 2/20 are short, it’s a batch issue. If 0/20, it may be isolated.
4. Review Machine Data
   Check L3 data: Was there any “short fill” alarm? Any reject bin activity? Any recipe change during batch?
5. Conclusion
   Classify as: Machine, Material, Method, Man, or Measurement error. This decides CAPA.

4. CAPA for Count Variation Complaints

5. Prevention Strategies: Building a Zero-Defect Line

Strategy 1: Technology

• Dual Counting: Primary sensor + Secondary weight check
• AI Vision: Detects broken tablets and rejects before filling
• Inline Checkweigher: 100% weight check. Triggers reject if ±2% deviation
• 360° Inspection: Verifies fill level in bottle

Strategy 2: Process Control

• IPC every 15 minutes for first 1 hour after startup
• Shift-wise reconciliation: Tablets fed vs Bottles filled x Count
• Lock HMI recipe with supervisor password

Strategy 3: People & Training

• Training on “Why 1 tablet matters” – Patient safety angle
• Second person verification for recipe change
• Reward system for zero-complaint months

6. Regulatory & Quality Impact

1. FDA 483: “Failure to establish adequate process controls to ensure label claim”
2. Recall Classification: Usually Class II. If for critical drug = Class I
3. Customer Impact: Loss of trust. Hospital tenders cancelled.
4. Cost: 1 recall = $2M to $10M. 1 complaint investigation = $5000

7. KPIs to Monitor

• Count Accuracy %: Target >99.99%
• Complaints per Million Bottles: Target <5 PPM
• IPC Failure Rate: Target 0%
• Reject Bin Reconciliation %: Must be 100%

8. FAQs

Q1: Is 1-2 tablet variation acceptable?
No. Label claim is a legal commitment. Even 99 instead of 100 is a deviation.

Q2: Who owns count variation – Production or QA?
Production owns prevention. QA owns detection and investigation. Both own CAPA.

Q3: Can serialization help?
Yes. Link count data to each SN. If complaint comes, trace exact time, operator, and machine.

9. Conclusion

Count variation is not a “small” complaint. It’s a symptom of a weak process.

You cannot inspect quality into the bottle. You must build it into the counting process.

The formula for 2026 is simple:

Right Machine + Right Process + Right People + Right Data = Zero Count Complaints

Invest in dual checks, train your team, and treat every complaint as a gift to improve.

Remember: The patient is counting on us to count correctly.