PIC/S has published PI 006-4, the revised Recommendations on Qualification and Validation, superseding the 2007 PI 006-3 and entering into force on 1 October 2026. Cleaning validation now occupies a full chapter, with HBEL-driven scope, explicit limits on visual inspection and a clear position against continued test-until-clean practice.
What happened
On 30 July 2026 PIC/S announced the publication of the revised PIC/S Recommendations on Qualification and Validation (PI 006-4). The document supersedes PI 006-3, dated 25 September 2007, and will enter into force on 1 October 2026.
The revision was prepared by a dedicated PIC/S Working Group, successively chaired by Norman Gray (UK / MHRA) and subsequently jointly chaired by Kevin O'Donnell (Ireland / HPRA) and Rebecca Dowd (USA / US FDA). The document's own revision history gives the reason concisely: an up-date according to the revised chapters 3 and 5 and Annex 15 of the GMP Guide.
PI 006-4 is a 49-page consolidation of the Validation Master Plan, pre-qualification and qualification stages, process validation, selected additional topics and cleaning validation. Cleaning validation is not an appendix here — it is Chapter 7, running from 7.1 Principles to 7.13 Maintaining a Validated Cleaning Process.
Why it matters
PI 006-3 was the reference text most cleaning validation programmes were originally built against. Nineteen years is long enough for a programme, its protocols and its SOPs to drift from the underlying expectations — not through negligence, but because documents age faster than the science around them. PI 006-4 is the first PIC/S restatement of cleaning validation expectations since that baseline, explicitly aligned with the Annex 15 revision of 2015 and with ICH Q9(R1) quality risk management.
PIC/S is not a legal instrument, and PI 006-4 is written as a guidance and training resource for GMP inspectors and industry. That framing matters less than it sounds: documents of this kind shape what inspectors ask about, and therefore what sites must be able to demonstrate.
What is particularly interesting
Several positions in Chapter 7 leave little interpretive room.
The extent of the cleaning programme is toxicologically driven. PIC/S states that it should depend on a toxicological risk assessment, for example based on HBEL values, and that this assessment should determine the operational and technical controls needed for a validated cleaning process.
Repeated "test until clean" is named as unacceptable. Continued cleaning failures and/or testing until clean are not acceptable. Where failure is intermittent, sites should implement remediation: improving procedures, reassessing risk, or introducing further mitigation such as equipment or facility dedication. Cleaning failures during and after validation must be recorded and investigated, including visual inspection failures and the initial second-person check after manual cleaning.
The limits of visual inspection are addressed directly. Visual inspection remains valid, with an acceptance criterion of "visually clean", but PIC/S notes it will likely not be appropriate for lower-HBEL products or where visual thresholds sit nearer to the required safety limit. Where maximum allowable carry-over limits fall below the limit of visual detection, cleaning verification after every change-over should be applied — of special concern for manual cleaning.
Sampling hierarchy. Swab sampling is generally more effective than rinse sampling and should be employed if feasible; rinse sampling averages results, so localised high-residue areas may not be identified. Recovery studies must represent the actual materials of construction and the actual method used in production.
Hold times and campaign length. Sites should hold evidence for dirty hold time, clean hold time and maximum campaign length, with campaign lengths normally defined by time and by number of batches. Non-contact parts from which residues can migrate — seals, flanges, mixing shafts, oven fans — should be assessed through a risk-based study using HBEL values.
Elsewhere, retrospective validation is stated to be no longer acceptable, OPV is described as having taken the place of periodic revalidation, and the number of process validation batches is expected to be scientifically justified rather than fixed — with PIC/S itself calling the current Annex 15 three-batch wording transitory.
What could we learn
Large organisations rarely lose cleaning validation capability in a single step. The common pattern is gradual: a detergent or supplier change nobody linked back to the original validation; a hold time studied once under older assumptions and then treated as permanent; a test-until-clean habit that solves today's batch and slowly normalises failure. Each decision is locally defensible; the aggregate is a programme whose documentation no longer reflects the science.
PI 006-4's positions largely describe what a mature, risk-based programme looks like when it is genuinely maintained rather than merely closed out. The practical question for most sites is not philosophical alignment, but whether the current evidence base still supports the conclusions their documents assert.
Wider context
PI 006-4 arrives alongside a joint EMA–PIC/S concept paper on the revision of Annex 15, with its stated intent to extend the annex's scope to active substances manufacturers and to reflect ICH Q9(R1). The direction across these documents is consistent: qualification and validation are lifecycle activities, with extent and depth set by scientific rationale and risk, and the validated state demonstrated continuously rather than certified once.
Validation takeaway
PI 006-4 reframes cleaning validation as a lifecycle position rather than a completed study: scope is derived from HBEL-based toxicological risk, visual inspection has recognised limits, and continued test-until-clean practice is expressly stated to be unacceptable. The practical task before 1 October 2026 is an evidence-led gap assessment, not a document rewrite.
Questions worth asking in our own programme
- Does the documented extent of our cleaning programme trace back to a current, product-specific toxicological risk assessment based on HBEL values, and when was that assessment last reviewed?
- Can we demonstrate evidence for dirty hold time, clean hold time and maximum campaign length — and are campaign limits defined by both time and number of batches?
- For products whose maximum allowable carry-over limits fall below the limit of visual detection, is cleaning verification performed after every product change-over, and is this documented as routine rather than exceptional?
- Have our recovery studies been performed on all relevant materials of construction using the same sampling technique, solvent and swab material that operators use in production?
- Do any cleaning deviations show a repeated test-until-clean pattern, and if so, what remediation has been evaluated — including the possibility of dedicated equipment for the worst-case products?
- Are supplier-driven variables such as starting-material impurity profiles, site water pressure and detergent formulation changes covered by our change control and revalidation triggers?
