Method Verification Requirements for an Advanced Imaging System for Microbial Plate Count Enumeration

Historic Background to the Development of the Plate Count Method

The pioneering German bacteriologist Robert Koch developed the plate technique to isolate pure cultures of individual disease-causing bacterial species that revolutionizing medicine with the germ theory of disease (1). In addition, the technique resulted in the ability to quantify bacterial numbers. The enumeration of bacteria in air, water, soil, and clinical samples became possible using the plate count method.

The evolution of the plate count went through a number of interesting steps. In 1875, Joseph Schroeter studied pigment colonies of Serratia marcescens (red) and Chromobacterium violaceum (purple) on cut slices of potatoes, showing pigment production was a conservative characteristic of these species. Louis Pasteur earlier introduced broth culture that did not isolate pure cultures of microorganisms. In his seminal 1881 paper, Koch described adding gelatin to the broth culture to achieve a semi-solid media that would enable the isolation of single colonies and thus obtain pure cultures. Gelatin was quickly found to be an unsuitable gelling agent, as it liquefied at above ambient incubation temperatures required for growth by human pathogens, and it was replaced by the seaweed derivative used in Asian cooking, agar-agar. The final step was the introduction of the now familiar, flat, double-sided dishes, that is, the Petri plate.

The British surgeon Joseph Lister was the first to obtain a pure culture in a liquid medium using a limiting dilution method. His publication on the isolation of a pure culture of the milk-souring bacterium Lactobacillus lactis, by what we now call serial dilution, was motivated by showing that a single bacterial species could be responsible for a single disease (2). The dilution at each step is constant, resulting in a geometric progression of the cell concentration in a logarithmic fashion. A 10-fold dilution that is typically used by microbiologists for each step is called a log-dilution.

In the now-familiar standard pour plate count, a 1 milliliter (mL) aliquot of a serial dilution of the test sample is placed in the bottom of a sterile Petri plate, and 10–15 mL of molten agar is spread over the surface and mixed with the inoculum in the dish. This is allowed to solidify, incubated for 2–5 days at above ambient temperature, and the resulting colonies counted. To determine the plate count, multiplying the count by the dilution will give the colony-forming units (CFU) per mL. As discussed by Sutton (3), the commonly accepted countable range of colonies on a standard plate is 25 to 250, with a lower range for fungal counts on plates and all counts on smaller membrane filters (MF). This range would both prevent overcrowding and provide statistically valid counts. Sutton recommended that the limit of detection should be considered 1 CFU while the limit of quantification is 25 CFU per plate.

Plate Count Method Limitations

Although the plate count has a long history of use and is a simple, low-cost, and reliable microbiological technique, it does have disadvantages that include the following:

• Uneven distribution of the colonies within the semi-solid media used in the pour plate method make it more difficult to count the individual colonies.

• Colonies may have different sizes, colors, margins, and shapes with similar transparences as the media, i.e., low-contrast colonies, making it more difficult to count the individual colonies.

• A single colony-forming unit must go through sufficient cell divisions during an extended incubation time to produce a visible colony on the plate.

• Viable microorganisms in a test sample may lack the capacity to grow and divide on microbiological culture media due to their unique nutritive and physiological requirements.

• A colony-forming unit on a plate may be derived from a single cell or a cluster of cells, so it may under-represent the number of viable cells.

• A microbial cell may go through a long lag phase, adjusting to the culture media before entering a log phase, increasing the required incubation time for visualization.

• At higher density, colonies may converge and become confluent, suppressing the growth of other surrounding colonies due to competition for both space and nutrients and the secretion of antibiotic agents and/or metabolic products.

• The density of the colonies may fall outside a statistical countable range or be too numerous to count. Microbiologists have a tendency to use higher dilutions that may give rise to lower, less reliable counts to avoid lengthy counting.

• Spreaders may obscure other colonies, resulting in an undercount.

• Manual plate counting is time-consuming, tedious, and prone to error, especially with higher counts.

• Although the human eye is well adapted to count colonies on the plates, well-trained and experienced microbiologists are needed to achieve the most accurate results.

An early precursor of the Growth Direct™ System was the Frost's Little Plate Method for counting bacteria in milk. This is a modified plate count method where a sterile slide was substituted for a Petri plate and 50 uL of undiluted milk was applied to the slide. Next a drop of molten agar is added and the two were thoroughly mixed, spread over a 4 sq. cm. area, allowed to solidify, and incubated at 37 °C for 4–8 hours (hr). The slide is treated with acetic acid and stained with methylene blue, and 5–20 fields are examined with a compound microscope for micro-colonies to obtain a count of the number of bacteria per mL of milk. The advantages of the method were the shortened incubation period, its lower cost, portability of the method, and the permanency of the record when the slide is dried, fixed, and stored (4).

The shortened incubation period due to counting micro-colonies is common to both the Frost's Little Plate Method and the Growth Direct™ System. For a micro-colony to be consistently detectable early in the incubation period using the advanced imaging system of the Growth Direct™ System, it will typically need to consist of 128 to 258 bacterial cells generated by 7 to 8 cell divisions or generations versus in excess of 10⁵ cells from 18 generations for visual detection (see Table I). Detection time for large-volume cells, like yeast, is much quicker, requiring as few as 8 cells in the micro-colony. As the time to result is the critical difference between the traditional plate count and the Growth Direct™ System, the authors will examine the relationship between colony size and conducting the count in more detail (5). The effect of incubation time on the number of organisms in an individual colony in a microbial enumeration can be determined theoretically as follows:

1. The initial inoculum is 1 CFU on a plate or membrane.

2. The length of the lag phase, L, is 1 hr.

3. The population doubling time, Td, is 20 min or 0.3 hr.

4. Synchronous growth occurs as in the following equation: N = N₀ × 2^(T–L)/Td, in which N is the CFU at time T; N₀ is 1 CFU initially; L is the length of the lag phase, in hours; and Td is the doubling or generation time.

5. Since for E. coli N₀ = 1 CFU, L = 1 hr, and Td = 0.3 h, we can determine the cell count at 1, 2, 3, 4, 5, 6, 7, 8, 9, 12, 18, and 24 hr (see Table I).

A sample calculation for L = 2 hr is N = 2^(2–1)/0.3 or 7 CFU.

NOTE—The lag and generation times were selected as an example in Cundell (5) and may not be pertinent to all situations.

The literature supports the view that a visible colony of 0.2 mm in diameter for range of microbial species has at least 10⁵ viable cells (6).

Technical Background of the Advanced Imaging System Used in the Growth Direct™ System

Counting individual microbial cells or micro-colonies concentrated on an MF has been a strategy successfully employed by automated plate readers in place of a visual count. The most significant post–Second World War advancement in microbial enumeration was the widespread introduction of membrane filtration techniques to count microorganisms. The German filter manufacturer Sartorius-Werke AG first developed MFs commercially. Prior to the war, MFs were primarily used for sterile filtration of air or liquids. In response to the need to determine water quality after wartime bombing, the German Hygiene Institutes used MFs for enumerating coliforms. In 1947, the German microbiologist Muller published a method for counting coliforms on MFs using lactose–fushsin broth (7). Clark et al., in 1951, introduced the membrane filtration method into the United States (8). The MF technique was established in the late 1950s as an alternative to the most probable number (MPN) procedure for microbiological analysis of water samples. The MF technique offers the advantage of isolating discrete colonies of bacteria from large sample volumes, whereas the MPN procedure only indicates the presence or absence of an approximate number of organisms as indicated by turbidity in the medium. The MF technique was accepted by the U.S. Federal Environmental Protection Agency for microbiological testing of potable water in the 11th edition of Standard Methods for the Examination of Water and Wastewater published by the APHA/AWWA (American Public Health Association and American Water Works Association).

The enumeration of individual bacterial cells or micro-colonies captured on MFs using epifluorescence microscopy and flow cytometry have been aided by the use of vital stains like DNA-binding fluorochromes, for example, acridine orange and 4,6 diamidino-2-phenyl-indole; esterase indicator dyes, for example, 6-carboxyfluoroscein diacetate; or redox indicator, for example, 5-cyano-2, 3-ditolyi tetrazolium chloride (9) and adenosine triphosphate (ATP) bioluminescence (10).

The colony density on conventional plates with standard incubation times as recommended in an early publication by Breed and Dotterrer (11) is <10 CFU/cm². This density avoids undercounting due to the convergence of adjacent colonies as they increase in size. For a 90 mm petri dish with an effective surface area of 58–64 cm², colony densities 5 and 10 CFU per cm² would result in a count of 290 and 580 colonies, respectively. For a 47 mm MF with reduced effective surface area of 17–25 cm², a colony density of 5 or 10 CFU/ cm² would result in a count of 150 to 300 colonies, respectively. As micro-colonies take up less space on the membrane, are less likely to converge and compete for nutrients than visual colonies, the Growth Direct™ System will have a greater dynamic counting range, that is, up to 1000 to 10,000 micro-colonies depending on colony morphology and doubling rate.

General Description of the Growth Direct™ System

The Growth Direct™ System (Rapid Micro Biosystems Inc., Lowell, MA; see Figure 3) is an automated, rapid microbial enumeration platform suitable for product testing, environmental monitoring, and water monitoring that integrates digital imaging, robotic cassette handling, incubation, and software control. Bar-coded cassettes containing the media and the membrane on the surface of the media are introduced into the system via a carrier. After being picked up by the robotic arm, the bar code is read to determine the sample identification and incubation parameters, and the cassette is then loaded into the built-in incubator (see Figure 1). To build up an image time series that differentiates growing micro-colonies from debris, cassettes are periodically removed from the incubator every 4 hours, imaged, and returned to the incubator by the robotic system. Before imaging, a low-powered heater removes any condensation that may have formed on the optical cassette lid.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Comparison of the Growth Direct™ System count and the mean visual growth. Cassette counts for representative microorganisms from the same growth cassette for TSA and R2A media. N = 410 cassettes for R2A and 350 for TSA.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Picture of Growth Direct™ System and Bioburden Consumable.

For imaging, the cassette is held on a computer-controlled stage with X, Y, and Z motion. Focusing is achieved using a non-contact laser distance sensor. The membrane in a cassette is illuminated with optically filtered, columnated blue light (wavelength 450–500 nm) from multiple light-emitting diodes (LEDs). Incident light induces fluorescence in the target micro-colonies, which is collected by a high-collection efficiency, non-magnifying lens system. The emitted fluorescent light is transmitted through a green emission filter (510–560 nm) and collected by a charge-coupled device (CCD) digital camera. To analyze the entire working area of the membrane, the imager acquires an image of the total membrane surface at three z-axis positions for a total of three superimposed images with defined x-y co-ordinates for all fluorescent events on the membrane surface. Accurate and consistent placement of the membrane for each image allows tracking of changes to the signal over time (see Figure 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Schematic of the detection of the autofluorescence of cells in micro-colonies.

Onboard, automated image analysis by custom software incorporates functions for background smoothing, object finding, and enumerating fluorescent objects that grow in size and intensity over time. Because the membrane and media display high-spatial frequency optical noise, which can vary with time, the software applies a background correction algorithm to smooth out the noise and separate signal from background. Clusters of neighboring pixels whose intensity significantly exceed the local background are classified as objects. The software monitors key morphological parameters of the objects including their position, intensity, size, and other characteristics. Algorithms assess changes in these characteristics over time to identify the objects as either growing micro-colonies or non-growing debris (12, 13). Instrument software and algorithms are subject to change control and revalidation, so the results generated will be expected to be the same irrespective of the software version employed.

Universality of Blue Light Excitation and Green Light Emission for the Detection of Micro-colonies

The intrinsic fluorescence of microbial cells where flavins fluoresce in the green region (510–560 nm) when excited by blue light (450–500 nm) is a universal characteristic of microorganisms with the intensity directly related to cell volume and number of cells, that is, the larger the microbial cells and the greater their quantity, the greater the fluorescence intensity. See Table II for the relationship between the wavelength and the color of visible light. A major advantage of using intrinsic fluorescence is that light emission does not require any reagents and the MF may be sequentially excited by blue light to help differentiate growing micro-colonies from autofluorescent debris. Because blue light has antibacterial activity, there is a potential danger of reducing the count. However, the literature reports that a 15–30 min exposure to 400 nm wavelength blue light doses of intensities of 54 to 108 joules per square centimeter (J/cm²) is necessary for up to a 5-log reduction in buffered saline (14). In contrast, the blue light excitation in the Growth Direct™ System is 1 J/cm² for 3 × 400 milliseconds, so it highly unlikely that the micro-colony detection method will reduce the cell numbers during the 4 hr time-series imaging steps. Studies have shown that the light-sensitive bacterium Pseudomonas fluorescens does not show any inhibitory effects from the blue light on growth (unreported results).

What is the reduced time-to-result seen with the Growth Direct™ System compared to the traditional plate count? Example results are presented for three model organism: the yeast Candida albicans and two bacteria, Escherichia coli (large cell size) and Brevundimonas diminuta (short cell size) (Table III). The time at which 50% of the autofluorescence micro-colonies were first detected and the 95% confidence interval is found in the second column. The time at which colonies become visible is based on the detection of 0.5 mm diameter colonies with the naked eye is found in the third column, while the overall reduction in incubation time to get the result using the Growth Direct™ System is found in the fourth column. The percent reduction in incubation time ranged from 55% to 90% (13).

In Table IV, the intrinsic cell fluorescence was determined by calibrated light microscopy images; the fluorescence per cell was measured on the Growth Direct™ System using calibrated beads; and the fluorescein equivalents at time of detection is based on a scanning electron microscopy estimates of cell numbers and the time-to-results given in Table III.

Selection of Media and Membrane Filters for Microbial Enumeration for Bioburden Determination and Environmental Monitoring

With the Growth Direct™ System, the MFs capturing the viable microorganisms are placed on standard microbiological growth media used for drug product, environmental monitoring, and water testing. They are then incubated at the recommended temperatures, but the incubation times are reduced because the enumeration is based on micro-colony detection using advanced imaging and not visible colonies (Table V). It should be borne in mind that all micro-colonies detected by advanced imaging would, on further incubation, become visible colonies. The membranes employed in the system are 0.45 micron, black mixed cellulose ester membranes. As with the traditional MF plate count, the porosity of this membrane favors good filtration rates, bacterial retention, and superior wicking of the media from the underlying plates, supporting colony development compared to 0.22-micron membranes used for sterile filtration (15). The membranes are stained black to quench the autofluoroscence of the cellulose esters of the membrane. This maximizes the signal-to-noise ratio to aid enumeration.

Advantages of the Growth Direct^TM System over the Plate Count

The application of robotic, advance imaging and computerization of the data offers advantages over the traditional plate count method that include the following:

• The automation of the incubation and the reading of micro-colonies growing on a membrane placed on standard semi-solid microbiological culture media.

• Bar coding to identify and track individual cassettes.

• Time to results up to 50% faster than the traditional plate count that relies on visual inspection by analysts.

• Elimination of time-consuming, tedious, and prone-to-error manual plate counting and recording.

• Superior counting performance compared to manual plate reading, especially with higher counts where microbiologists often underestimate the count.

• Improved data integrity and information management.

The reader should note that improved data integrity and information management is currently a hot compliance topic with the U.S. Food and Drug Administration, which recently published Draft Guidance for Industry: Data Integrity and Compliance with CGMP in April 2016.

Regulatory and Compendial Guidance for the Validation of Alternative Microbiological Test Methods

What is the USP position on alternative microbiological methods? USP40/NF35 General Notices 6 Testing Practices and Procedures provides guidance of the use of automated and alternative test methods. 6.20 Automated Procedures states, “Automated and manual procedures employing the same basic chemistry are considered equivalent”. The statement is equally true for procedures employing the same basic microbiology such as a plate count and the Growth Direct™ System. Furthermore, 6.30 Alternative and Harmonized Methods and Procedures states, “Alternative methods and/or procedures may be used if they have advantages in terms of accuracy, sensitivity, precision, selectivity, or adaptability to automation or computerized data reduction, or in other specialized circumstances. Such alternative procedures and methods shall be validated as described in the USP40/NF35 general chapter Validation of Compendial Procedures <1225> and must be shown to give equivalent or better results”. However, the authors understand that the General Notices is being updated to be restated in terms of microbial methods as they are validated as described in <1223> Validation of Alternative Microbiological Methods, not USP <1225>.

Why would the Growth Direct™ System not be considered an alternative method? USP40/NF35 <1223> states the following: “There are commercially-available enhancements to growth-based methods that allow colonies on solid media to be read more quickly, with substantially less incubation time, than is possible using only the unaided eye …. In the implementation of these enhanced methods for the detection of colony growth, only the detection capability of the method requires verification.” This statement supports the view that the Growth Direct™ System is not an alternative method requiring method validation.

Similarly the PDA Technical Report No. 33 (Revised), dated September 2013 and accessed from the PDA website, states the following: “Some alternative or rapid technologies may be considered automated traditional or compendial microbiological methods, especially when the results are in CFUs. These technologies may be qualified for their intended use without the need for demonstrating certain method validation requirements as specified in Section 5.0 of the Technical Report. For these technologies, at least accuracy and precision assessments should be performed, in addition to method suitability and equivalence/comparability studies.” The view expressed in USP <1223> was fully supported in this industry practice document.

Ph. Eur. 5.16 Alternative Methods for the Control of Microbiological Quality does discuss growth-based methods using the presence of endogenous autofluorescent molecules and metabolites such as reduced nicotinamide adenine dinucleotide phosphate (NADPH) and flavo-proteins within microorganisms. The revised chapter continues to view a technology like the Growth Direct™ System within the framework of alternative microbiological methods. If we take the risk-based approach as recommended in 5.1.6, then a method verification strategy can be justified.

All analytical equipment, including that used for microbiological testing methods, is subject to industry-standard instrument qualification requirements. The revised chapter <1223> cites USP40/NF35 <1058> Analytical Instrument Qualification for general guidance in this area. This latter chapter includes the development of user requirement specifications and the well-known elements of method qualification such as installation qualification (IQ), operational qualification (OQ), and performance qualification (PQ). The authors support the standard IQ and OQ approaches to the Growth Direct™ System but, as outlined in the next section, not the standard PQ. A simple method verification approach would be substituted for the standard PQ approach.

Review of the Validation Parameters and Their Appropriateness for the Growth Direct^TM System

The validation parameters for an alternative microbiological method, as listed in USP <1223>, are found in Table VI.

For the first six validation parameters—that is, accuracy, precision, specificity, limits of detection, quantitation, and linearity—the results will essentially be the same for a plate count and the Growth Direct™ System because both detected and enumerated the same colonies on the same membrane. The linearity results between the two may deviate, showing higher counts recorded with the Growth Direct™ System when there is large number of CFUs on a membrane. The major differences between the two procedures will be the time-to-results, the automation of the incubation, the electronic capture of the CFU count, reading of the number of CFUs, and the improved reliability of the count using an advanced imaging system instead of the human eye of the analyst. Keep in mind that the micro-colonies enumerated by the Growth Direct™ System will, on further incubation, develop into visible colonies.

The operating or dynamic range for the Growth Direct™ System will extend beyond the 250 CFU per plate because micro-colonies can occupy more space on the membrane without crowding each other out. The algorithms in the image analysis software for background smoothing, autofluorescence object finding, and counting fluorescent objects that grow in size and intensity represent the ability to differentiate between microbial colonies and autofluorescent particulates. For the validation parameters of robustness, repeatability, and ruggedness, Rapid Micro Biosystems, Inc. can provide data meeting these requirements to both the regulators and the purchasers of the Growth Direct™ System.

Recommendations for the Method Verification of the Growth Direct™ System

What do the authors recommend for method verification? The simplest approach is to plate representative microorganisms as found in USP40/NF35 <61> and <62> in the countable range of 25 to 50 CFU per membrane, enumerate the CFUs using the Growth Direct™ System, and continue to incubate for 3 days for bacteria and 5 days for fungi and have multiple experienced microbiologists count the colonies using the standard visual inspection procedure. According to the literature, the counts using different microbiologists should not differ 90% of the time by more than 10% (2004 APHA Standard Methods for the Examination of Dairy Products). Alternatively, the acceptance criterion may be no significant difference between the Growth Direct™ System and visual inspection counts using standard analysis of variance methods. The levels of replication recommended by Rapid Micro Biosystems, Inc. to be employed in method verification are 6 to 10 different microbial species including the organisms recommended in USP <61> and appropriate in-house microbial isolates, run as six replicates conducted in three independent runs with the plates visually read by at least three analysts. This testing is conducted in place of the typical PQ.

The use of a simple paired t-test would be used to compare the colony counts using the advanced vision system of the Growth Direct™ System and visual inspection counts. For best results, n should be >30. Note: The Growth Direct™ System counts may be slightly higher because of the merging of adjacent colonies later in the incubation and the tendency of microbiologists to undercount plates with high number of colonies, but higher Growth Direct™ System counts will not be a compliance issue, as equivalence is usually stated as good or better than the compendial method.

If we let X = the Growth Direct™ System count and y = the visual inspection count of the sample plate, then to test the null hypothesis that the true mean difference between the counting methods is zero, the procedure is as follows:

1. Calculate the difference (di = y_i – x_i) between the counts of each method, making sure you distinguish between positive and negative differences.

2. Calculate the mean difference, d cap.

3. Calculate the standard deviation of the differences, Sd, and use this to calculate the standard error of the mean difference, SE(d cap) = Sd/√n.

4. Calculate the t-statistic, which is given by T = d cap/SE(d). Given the null hypothesis, this statistic follows a t-distribution with n–1 degrees of freedom.

5. Use tables of the t-distribution to compare your value for T to the tn–1 distribution. This will give the P-value for a one- or two-tailed paired t-test. If P < 0.05 there is strong evidence that on average the difference is not significant.

Another excellent tool to compare the data generated by an automated plate reader and visual inspection is the box plot or box and whisker diagram, which is a standardized way of displaying the distribution of data based on the five number summary: minimum, first quartile, median, third quartile, and maximum.

Additional guidance for the statistical analysis to establish equivalence between microbiological enumeration methods can be found in International Standards Organization (ISO) Standard 17994:2014 and the 2013 PDA Technical Report No. 33 (Revised).

Errors Associated with Colony Counting Procedures

According to Jarvis (16), the errors common to all microbial enumerations are (1) sampling and dilution errors, (2) errors in pipetting aliquots of diluted samples, (3) distribution error, (4) counting and recording errors, and (5) calculation errors. For the purpose of comparing the performance of the Growth Direct™ System to a manual plate count, the discussion will be confined to counting errors as the other errors equally apply to both methods.

The ability to count accurately the number of colonies on a plate depends on the ability of the microbiologist to discriminate colonies of different size and shape from debris and imperfections in the agar or membrane. This ability will vary by individual microbiologists, their state of health, and the level of outside distractions. The counting efficiency will be lower as the number of plates counted increases and the number of colonies increases on the plate. In a classic study of the counting efficiency of six analysts, Fruin et al (17) showed that only 51% of the colony count lay within 5% of the real count as determined by photographs of the plates (termed photocount), and only 81% lay within 10%. Furthermore, as pointed out by Jarvis (16), analysts consistently undercount the plates, especially plates with high counts. The percentage within the photocount was significantly different in the 10 to 100 ranges than the three other ranges (Table VII).

An earlier study (18) reported that the analyst variation when conducting the standard plate count for milk using the method described in the 1967 APHA Standard Methods for Examination of Dairy Products demonstrated 18.2% amongst-analyst variation compared to 10% between-analyst variation, and 7.7% within-analyst variation compared to 5% in the Standard Methods.

These findings bring in to question the ability of microbiologists to count large numbers of plates accurately compared to the automated Growth Direct™ System, leading to potential data integrity issues especially when the counts are close to the alert/action levels for the microbial enumeration.

Additional microorganisms frequently isolated during product testing, and environmental and water monitoring, may be added to the method verification study. Examples may be Burkholderia cepacia and Bacillus cereus for product testing, Micrococcus luteus and Staphylococcus epidermidis for environmental monitoring, and B. cepacia and Ralstonia picketti for water monitoring. Either facility isolates or ATCC cultures may be used for this purpose.

Method Suitability Tests

The method suitability testing for individual pharmaceutical ingredients and drug products as required in USP <61> must be met prior to routine testing. This would demonstrate the recovery of the challenge microorganisms in the presence of any product residues after the rinsing of the MF with a diluent such as buffered saline or peptone water containing the appropriate neutralizing agents. This requirement is independent of method validation or verification. There should no more than a 2-fold difference between the recovery with and without the product, that is, between 50% and 200% recovery.

Cassettes used for air, surface, and personnel monitoring in a pharmaceutical facility usually contain neutralizing agents for commonly used disinfectants. The recovery of microorganisms especially from facility and equipment surfaces with residual disinfectant would need to be validated. Pharmaceutical-grade water would not require suitability testing because these waters contains no ingredients that would inhibit the recovery of the challenge organisms.

Conflict of Interest Declaration

Tony Cundell works as a consulting microbiologist to the pharmaceutical industry, sterile compounding pharmacies, and microbiological testing equipment manufacturers. He is a member of the 2015–2020 U. S. Pharmacopeia Microbiology Expert Committee; however, this publication represents the opinions of its authors and not the positions of the committee.