Spreadsheet Change Without Validation in Batch Release
A manufacturer of infusion pumps uses a custom spreadsheet and a commercial electronic document management system to review and release batch records. In March, the quality manager modified the spreadsheet formula for averaging pump flow-rate test results to exclude the highest and lowest values. The change was made directly in the production version; no change request, risk assessment, or validation was performed. Training records show the quality manager attended a 30-minute webinar on spreadsheet controls. During the audit, the auditor samples 12 released lots: three were released using the new formula, and no trend or impact assessment exists. The company's software validation procedure requires validation for automated production and quality system software. The quality manager argues the spreadsheet is a tool, not a medical device, and therefore not subject to validation.
1. Evaluate the quality manager's argument. What MDQMS requirements apply to this spreadsheet?
Read the worked answer
The argument confuses medical device regulation with QMS software control. ISO 13485 4.1.6 requires validation of computer software used in the QMS before initial use and after changes, as appropriate. The spreadsheet calculates results used to release product, so it directly affects conformity. The change also engages 4.2.4 (document and change control), 7.5.1 (control of production), 7.6 where results depend on measurement equipment or software, and 8.2.6 (monitoring and measurement of product). Risk management under 7.1 should have assessed the effect of excluding highest and lowest values. The 30-minute webinar does not demonstrate competence for validating a release-critical calculation. The key point is not whether the spreadsheet is a medical device; it is whether the software is used in the QMS in a way that affects product conformity.
2. What evidence should the auditor examine to determine the severity of the finding?
Read the worked answer
Examine the spreadsheet's current and previous versions, the formula change history, access controls, and any backup or version log. Review the validation procedure and any validation record for the spreadsheet. For the three affected lots, compare the original test data with the calculated averages and determine whether any lot would have failed under the prior formula. Check whether the exclusion rule was authorized, whether specifications permit it, and whether the change was risk-assessed. Review training and competence records for the quality manager and any reviewer. Sample other spreadsheets used in release to see whether the issue is isolated. Severity increases if product release decisions were affected or if the change was concealed; it may be major if the QMS software control process is systemically absent. It may be minor if the impact is clearly bounded and correction is immediate, but that conclusion needs evidence.
3. Draft the core of a nonconformity statement and the audit trail you would record.
Read the worked answer
Core statement: The organization changed the formula in a spreadsheet used to calculate flow-rate results for batch release without a change request, risk assessment, or validation, contrary to its own software validation procedure and 4.1.6. Three lots were released using the changed formula, and no impact assessment was performed. Audit trail: interviewed the quality manager; reviewed the spreadsheet version and formula; reviewed the software validation procedure; sampled 12 batch records; identified three affected lots; confirmed no change request, risk assessment, or validation record; reviewed training records. The statement should identify the requirement, the objective evidence, and the product effect. Do not write 'spreadsheet not validated' without the release linkage. Recommend containment: evaluate the three lots and any other outputs from the changed formula before further release.