The determination of the recovery rate is a preliminary step, necessary during a cleaning validation exercise

Determining the recovery rate validates sampling techniques by accurately measuring the amount of contaminant that can be recovered. This study makes it possible to demonstrate that if a residue is present on the surface of the equipment, the latter can be quantified adequately by combining the analytical method and the sampling methodology.

These studies provide a scientific basis for the use of sampling and analysis methods in a cleaning validation exercise.

The study of the recovery rate of a chemical residue is generally performed in the laboratory on different coupons (glass, stainless steel, plastic ...) depending on the validation needs. It consists in depositing a quantity of defined chemical residue and applying the chosen sampling technique.
Determining the recovery rate is a regulatory requirement (1).

Two sampling methods are mainly used in cleaning validation.
Sampling by swabbing, the recommended technique which targets the equipment surfaces that are most difficult to clean, and rinse water sampling which gives an overall view of the residual contamination of an item of equipment.
The sample will then be analyzed in accordance with the analytical method identified.
During the recovery study, the aim is really to simulate the sampling conditions carried out during a cleaning validation exercise (same solvent, same type of swab, same methodology ...)
It should be noted that for a rinse water sample, it is often difficult to replicate the laboratory rinsing procedure. Here “worst case” conditions should be utilized, while trying to reproduce the rinsing conditions as closely as possible.
A method for simulating “worst case” conditions may for example consist in placing the coupon at the bottom of a beaker and applying gentle agitation for a period similar to the rinsing time.

The recovery rate is a combination of several performance elements: the recovery efficiency (relationship between the quantity retrieved and the quantity deposited), the analytical efficiency (relationship between the quantity retrieved and the quantity measured) and in the case of swabbing the desorption efficiency of the swab.

A recovery rate must be established for all the surfaces sampled. For this a study may be conducted on each type of surface to be sampled. An alternative to this is to demonstrate equivalence in terms of percentage recovery between several types of surface. The study can then be conducted on a single surface from the group of surfaces judged equivalent.

Another approach is to exclude from the study sampled surfaces constituting less than a percentage (such as 1% or 2%) of the total surface of the equipment (example of a piece of equipment consisting of 98% stainless steel and 2% PTFE). In this case, the recovery rate used will be the lowest recovery rate of all other types of surfaces studied. (2).

In 2010, Destin A. Leblanc, President of Cleaning Validation Technologies, explained that the recovery rate should decrease proportionately when the quantity of residue increased.
To illustrate this, he took the example of a snow shovel:
“If I have permission to pick up as much snow as possible from a pavement with a single pass of a spade, it is likely that with only 7 centimeters of snow, my sample (a spade) will retrieve a relatively large quantity of snow.
This value could be between 60 to 80% of the snow present on the area sampled.
On the other hand, if I were to use the same spade and the same sampling procedure on a pavement covered with 70 cm of snow, I could only collect 30 to 40 % of the snow present on the area sampled”. The saturation of the swab is one of the reasons that would explain this phenomenon (3).

However, Destin A. Leblanc specifies that some studies have demonstrated a linear recovery rate for different quantities of residue deposited. Ultimately, the assumption is that the reduction in recovery rate would be only observable in a study conducted over a wide range.
If recovery rate studies were performed at significantly different levels (by factors of at least 10 or 20), then the percentage of recovery could diminish. However, there is little value in recovery rate studies involving multiple levels of contaminant. For this reason, it seems preferable to perform a recovery rate study on a single quantity of deposited residue, preferably close to the acceptance limit.
Performing several tests for the same amount of residue deposited seems more sensible because they will give a better idea of ​​the variability of the sampling process.(4).

For sampling by swabbing it seems preferable to take account of inter-operator variability and to perform the study with two people. Rinse sampling is not significantly operator-dependent, hence the study can be conducted with one person.

The FDA specifies that industry must challenge its analytical methods in order to obtain an acceptable recovery rate. A recovery rate below 50% generally calls the sampling method into question.

Two options are generally offered to manufacturers:

The first option consists in using the recovery percentage to correct the analytical result measured. For example, if the recovery rate is 78%, then the analytical value obtained will be corrected by dividing it by 0,78.
The second option consists in requiring that the recovery percentage be within a certain range so that no correction is required. For example, if the percentage is between 75% and 100% no correction factor will be needed(5).

It should be noted that recovery studies may not be necessary for some residues which are known to be easily soluble in water, such as sodium hydroxide and phosphoric acid used as cleaning agents.

Should we also determine a recovery rate to quantify the bioburden? Due to the variability inherent in the germ count, it does not seem appropriate to determine a recovery rate in the same way for a chemical residue. In the sense of the USP 1227, the microbial recovery rate consists of conducting a study on the absence of inhibition of bacterial growth by the sampling method. (6).

Sophie BOURGOIS - ASPEN Our Lady Of Bondeville


(1) GMP, Part II, Chapter 12.74
(2) Technical report no 49 - Points to Consider for Biotechnology Cleaning Validation
(3) Destin A. Leblanc, Cleaning Validation Technologies, Cleaning Memos October 2010
(4) Destin A. Leblanc, Cleaning Validation Technologies, Cleaning Memos September 2013
(5) Destin A. Leblanc, Cleaning Validation Technologies, Cleaning December Memories 2012
(6) USP 1227 "Validation of Microbial Recovery" from Pharmacopoeial ArticlesFDA - GUIDETO INSPECTIONSVALIDATION OF CLEANING PROCESSES