129 QC microbiology interview questions

Growth promotion testing (GPT) checks that each batch of culture medium can support the growth of the organisms it is meant to detect. Without it, a 'no growth' result could simply mean the medium was faulty. Inoculate each test medium with not more than 100 CFU of each specified organism and incubate for the shortest period the method allows. On solid media, the count must be within a factor of 2 (50–200%) of the inoculum control or a previously approved batch; in liquid media, growth must be clearly visible and comparable. GPT is done on every lot, whether prepared in-house or bought ready to use.

TAMC (total aerobic microbial count) uses Soybean-Casein Digest Agar incubated at 30–35 °C for 3–5 days. TYMC (total combined yeast and mould count) uses Sabouraud Dextrose Agar incubated at 20–25 °C for 5–7 days, as given in USP <61> and Ph. Eur. 2.6.12. The higher temperature favours bacteria and the lower temperature favours fungi. Remember that any fungi growing on SCDA are counted in the TAMC, and any bacteria growing on SDA are counted in the TYMC. Countable ranges are up to 250 colonies per plate for TAMC and up to 50 for TYMC.

Every time a culture is subcultured, there is a chance of mutation and selection, so after many passages the organism can drift away from the characteristics of the reference strain. It might grow faster, lose a biochemical property or become less sensitive to an inhibitor. Pharmacopoeias therefore allow no more than five passages from the original reference culture. Labs control this with a seed-lot system: a master stock, working stocks prepared from it, and a log recording the passage number on every culture used for GPT, method suitability or identification.

The common compendial strains are Staphylococcus aureus ATCC 6538, Pseudomonas aeruginosa ATCC 9027, Bacillus subtilis ATCC 6633, Candida albicans ATCC 10231, Aspergillus brasiliensis ATCC 16404 and Escherichia coli ATCC 8739. Clostridium sporogenes ATCC 19404 or 11437 is used for anaerobic media and the test for Clostridia, and Salmonella enterica serovar Typhimurium ATCC 14028 for the Salmonella test. Equivalent strains from other collections such as NCIM, MTCC, NCTC or NBRC can be used when listed as equivalent. Interviewers often ask for at least the first five with their numbers.

A selective medium contains substances that suppress unwanted organisms so that the target can grow, for example bile salts and crystal violet in MacConkey agar suppress most Gram-positive bacteria. A differential medium shows a visible difference between organisms that do grow, for example MacConkey agar separates lactose fermenters (pink colonies) from non-fermenters (colourless). Many media are both, such as MacConkey, mannitol salt and XLD agar. In USP <62>, selective media need growth promotion and inhibitory checks, and differential media also need an indicative check.

Most media are sterilised by autoclaving, typically at 121 °C for 15 minutes, though the actual cycle must be validated for the load size because larger volumes take longer to reach temperature. Heat-sensitive supplements, such as antibiotics or egg yolk, are filter-sterilised and added after the base medium has cooled to about 45–50 °C. To check sterilisation, incubate a sample from each batch as a sterility check, review the autoclave cycle record, and validate the cycle itself with load mapping, temperature probes and biological indicators placed in the coldest positions.

A negative control is a sample of sterile diluent or medium handled exactly like the test samples but with no product. It shows that the media, diluents, equipment and analyst handling did not introduce contamination during the test. If the negative control shows growth, the results from that session cannot be trusted. The test is normally invalidated, the cause investigated (media sterility, environment, technique) and the test repeated. Negative controls are expected in every EM session, every MLT run and every sterility test.

Each medium is formulated to work within a defined pH range, and pH affects both the growth of target organisms and the selectivity of selective agents. Autoclaving can shift pH, especially in media with sugars or phosphate buffers, so pH is checked after sterilisation and cooling, at room temperature (about 25 °C). The typical tolerance is ±0.2 units of the manufacturer's stated value. Out-of-range pH means the batch is rejected or adjusted only if the SOP allows it. Record the meter calibration and the actual reading.

Shelf life is set by a hold-time study. Prepare media and store them under the intended conditions, for example plates at 2–8 °C in sealed bags protected from light. At defined intervals, test samples for growth promotion, pH, appearance, and for plates, signs of drying or contamination. The shelf life is the longest time at which all results still meet acceptance criteria. Typical values are a few weeks for plates and longer for tubes and bottles. Ready-to-use media follow the supplier's expiry, supported by your own GPT on receipt.

Gram-positive bacteria have a thick peptidoglycan cell wall, which traps the crystal violet–iodine complex during Gram staining, so they stay purple after decolourisation with alcohol or acetone. Gram-negative bacteria have a thin peptidoglycan layer surrounded by an outer membrane containing lipopolysaccharide. The decolouriser dissolves this membrane and washes the dye out, so they take up the safranin counterstain and appear pink. The outer membrane is also why Gram-negative bacteria release endotoxin, which is why they matter so much in water and parenteral manufacturing.

There are four main methods. Active air sampling draws a measured volume of air, usually 1 m³, onto an agar plate and gives results in CFU/m³. Passive air sampling uses 90 mm settle plates exposed for up to 4 hours to catch organisms that settle by gravity. Surface monitoring uses 55 mm contact plates on flat surfaces and swabs on irregular ones. Personnel monitoring uses glove prints of all five fingers and contact plates on gown sites such as forearms and chest. Non-viable particle counting runs alongside to give real-time data.

An alert level is an early warning that conditions are drifting from normal. It is set by the site from its own historical data, often the 95th percentile, and reaching it calls for attention and closer trending, but not necessarily a full investigation. An action level is a limit which, when exceeded, requires a formal investigation, impact assessment and corrective action. Action levels for EM come from regulation, such as Annex 1 Table 6. Alert levels must sit below action levels and should be reviewed regularly as data accumulates.

Under EU GMP Annex 1 (2022), Table 6, the Grade B action limits are 10 CFU/m³ for active air samples, 5 CFU per 4 hours for 90 mm settle plates, 5 CFU per plate for 55 mm contact plates, and 5 CFU per glove for five-finger glove prints. For comparison, Grade A is 'no growth' for all sample types, Grade C is 100, 50 and 25 for air, settle and contact plates, and Grade D is 200, 100 and 50. Grade C and D have no glove print limit in the table.

Cleanroom surfaces are regularly disinfected, so traces of disinfectant remain on them. If a contact plate picked up both organisms and disinfectant residue, the residue could keep killing the organisms on the plate, giving a false low count. Neutralisers in the agar, usually lecithin and polysorbate 80 (Tween 80), inactivate common disinfectants such as quaternary ammonium compounds, phenolics and alcohols. Some sites add histidine or sodium thiosulfate for other agents. The neutralising capacity should be verified against the disinfectants actually used on site.

Report the result immediately and raise a deviation or excursion record as per SOP. Identify the isolate, to species level for Grade A and B. Check the sampling and lab handling first to rule out lab error. Then review the area: cleaning and disinfection records, HVAC and pressure data, particle counts, maintenance work, interventions and the people present. Look at trends for that location and organism across the site. Assess the impact on every batch processed in that area during the period, define root cause and CAPA, and confirm effectiveness with follow-up monitoring.

Both Grade A and Grade B are ISO 5 at rest, based on the limit of 3,520 particles per m³ at 0.5 µm. In operation, Grade A must stay at ISO 5, because it is the critical zone where product is exposed. Grade B in operation is allowed up to 352,000 particles per m³ at 0.5 µm, which corresponds to ISO 7. Grade C is ISO 7 at rest and ISO 8 in operation, and Grade D is ISO 8 at rest, with in-operation limits set by the site.

During long exposure, the agar in a settle plate dries out in the airflow, and dried agar recovers organisms poorly. Four hours is the maximum exposure time used in Annex 1 Table 6 for that reason. If a plate is exposed for less than 4 hours, the result is still compared with the 4-hour limit, not scaled down. If a process lasts longer than 4 hours, plates are changed so that the whole process is covered. Each site should validate, by weight loss and recovery studies, that its plates still recover organisms after the exposure time used.

Surface samples are taken at the end of operations or after a critical activity, not during production. Pressing a contact plate onto a critical surface during filling could leave agar residue or introduce contamination right where product is exposed. Sampling at the end still shows how clean the surface was during the work. After sampling, wipe the area with sterile disinfectant to remove agar residue, which could otherwise support microbial growth. The timing rule should be written in the SOP and the sampling time recorded on the worksheet.

A common scheme is a single medium, usually soybean-casein digest agar with neutralisers, incubated first at 20–25 °C for 3–5 days and then at 30–35 °C for 2–3 days. Starting at the lower temperature favours fungi, and the higher temperature then picks up bacteria. Some sites use two plates instead: one at each temperature, or SDA for fungi. Whatever scheme you choose, support it with a recovery study using compendial strains and your own environmental isolates, and document the justification in the EM programme.

Use a documented risk assessment rather than a grid. Consider points closest to exposed product and product-contact surfaces, airflow patterns from smoke studies, where operators work and intervene, transfer points such as pass boxes and airlocks, surfaces that are hard to clean or collect water, and the history of excursions. Mark each location on a layout drawing with its justification. Review the plan at least yearly and after any change to the room, equipment or process. A plan that never changes suggests the risk assessment is not being used.

Recovery time is how long a cleanroom takes to return to its at-rest particle level after a particle challenge, usually measured as the time for a 100:1 reduction in concentration. It shows how efficiently the HVAC system removes contamination generated by activity or a disturbance. Annex 1 recommends determining it during qualification. A common guidance value is 15–20 minutes. Rooms with poor recovery may have too few air changes, poor air distribution or obstructions, and should be investigated before use.

Typical skin organisms are Staphylococcus species such as S. epidermidis, S. hominis and S. capitis, Micrococcus luteus, Kocuria, Cutibacterium (formerly Propionibacterium) acnes and Corynebacterium. These are the most common isolates in most cleanrooms, because people shed millions of skin cells every hour. Finding them in Grade A or B usually points to gowning problems, poor aseptic technique, too many people or bad behaviour such as fast movement. Identification to species, and comparison with personnel monitoring isolates, helps link a recovery to a person or practice.

Start with provisional alert levels set below the action limits, based on data from similar rooms or a fixed fraction of the action limit, for example 50%. Document the rationale clearly as interim. Collect data through EM performance qualification and routine monitoring. Once you have enough results, ideally a full year covering seasonal changes, calculate data-driven alert levels such as the 95th or 99th percentile. For very clean areas with mostly zero counts, use frequency-based alerts instead. Review and update the levels at least once a year.

Treat it as an excursion, because Grade A requires no growth. Identify the isolate to species. Check plate handling, labelling and incubation to rule out lab error. Review the intervention log, video, particle data and personnel monitoring for the exposure period. Bacillus is a spore-former, so look at material transfer, packaging and cardboard, and at sporicide use and frequency, and check for a spore-former trend. Assess every batch filled during exposure, including sterility test results. Batch disposition depends on the risk assessment, not just a passing sterility test. Implement CAPA and monitor its effectiveness.

Start from the contamination control strategy and a risk assessment of the line. Use smoke studies and process mapping to place Grade A locations at the point of fill, stopper bowl, open containers and sites of critical interventions. Plan continuous particle monitoring and viable air sampling, with settle plates covering the whole fill. Define surface sampling at the end of each session and personnel monitoring after interventions and at exit. Set frequencies by grade and shift. Run an intensive baseline during qualification, then set data-driven alert levels and a monthly trend review process.

Two media are used. Fluid Thioglycollate Medium (FTM) is incubated at 30–35 °C and supports anaerobic bacteria in its lower layers and aerobic bacteria near the surface. Soybean-Casein Digest Medium (SCDM) is incubated at 20–25 °C and supports fungi and aerobic bacteria. Both are incubated for not less than 14 days, as required by USP <71>, Ph. Eur. 2.6.1 and IP. Each medium lot needs sterility and growth promotion checks. FTM contains resazurin, so if more than one third of the medium turns pink before use, it can be restored once by heating.

Membrane filtration is preferred whenever the product can be filtered, and it is essential for products with antimicrobial activity. The membrane retains any micro-organisms while most of the product passes through, and rinsing with a validated fluid washes away preservatives or antibiotics that would otherwise inhibit growth. It also allows larger sample volumes to be tested. Direct inoculation is used only for products that cannot be filtered, such as some ointments, suspensions and medical devices, and the product must then be neutralised or diluted enough not to inhibit growth.

Only when at least one of four conditions is clearly shown: microbiological monitoring of the sterility testing facility shows a fault; a review of the testing procedure shows a fault; the negative controls show growth; or, after identification, the growth can be unequivocally attributed to faults in the materials or technique used. Suspicion is not enough. If the test is invalidated, it is repeated with the same number of units. Regulators look very closely at invalidations, so the evidence must be documented in detail.

A membrane with a nominal pore size of not more than 0.45 µm is used, with proven retention of micro-organisms. Membranes are usually 47–50 mm in diameter. Cellulose nitrate membranes suit aqueous, oily and weakly alcoholic solutions, while cellulose acetate is used for strongly alcoholic solutions. In closed canister systems, the membrane is built into the canister. The membrane type, supplier and rinse conditions are fixed during method suitability, and a change of membrane supplier usually requires repeating method suitability.

The volume of product must not exceed 10% of the volume of the medium. For example, if a tube contains 100 mL of FTM, you can add at most 10 mL of product. This rule keeps the product from diluting the medium too much or inhibiting growth through its own ingredients. If the required quantity of product exceeds this ratio, use a larger volume of medium or divide the product between several containers. The quantities to test per container come from the pharmacopoeia table and the product specification.

For FTM, the usual organisms are Staphylococcus aureus ATCC 6538, Pseudomonas aeruginosa ATCC 9027 and Clostridium sporogenes ATCC 19404 or 11437. For SCDM, they are Bacillus subtilis ATCC 6633, Candida albicans ATCC 10231 and Aspergillus brasiliensis ATCC 16404. Each is inoculated at not more than 100 CFU into the final rinse (membrane filtration) or into the medium containing product (direct inoculation). Growth must be visibly comparable to a positive control without product, within 5 days. The same organisms are used for growth promotion of the media.

If the product makes the medium cloudy so that growth cannot be judged visually, transfer suitable portions of not less than 1 mL into fresh vessels of the same medium. USP <71> allows this between day 3 and day 7 and at the end of the 14-day incubation. Incubate the subcultures for not less than 4 more days. If they remain clear, there is no growth. Never assume turbidity is just product precipitate without this check, and consider membrane filtration if possible to avoid the problem altogether.

The number depends on batch size and product type, as set out in the pharmacopoeia tables. For injectable batches of up to 100 containers, test 10% or 4 containers, whichever is greater. For 101 to 500 containers, test 10 containers. For more than 500 containers, test 2% or 20 containers, whichever is less. For large-volume parenterals, test 2% or 10 containers, whichever is less. A separate table sets the minimum quantity per medium from each container. The marketing authorisation can require more units, never fewer.

Because a false positive is very costly. If organisms from the test environment get into the media, the batch could be wrongly rejected. Worse, frequent false positives tempt labs to invalidate tests, and an invalidation culture can hide real contamination in product. An isolator, or a Grade A zone with a Grade B background, keeps the risk of adventitious contamination very low. Monitoring the test environment and the analyst's gloves during every session also provides the evidence needed if a positive result is ever investigated.

A closed canister system is a pre-sterilised, sealed filtration unit, usually supplied as a pair of canisters with tubing and a needle or spike. The product is pumped through the membranes inside the canisters by a peristaltic pump, rinse fluid follows the same path, and then FTM and SCDM are added through the tubing. The canisters are sealed and incubated directly, so the membrane is never exposed to the room. This greatly reduces handling and the risk of false positives compared with open funnels and cutting membranes.

The sterility test samples only a small number of units, so it can only detect gross contamination. If 0.1% of units in a batch were contaminated and you tested 20, the probability of picking at least one contaminated unit is only about 2%. Even at 1% contamination, 20 units would detect it only about 18% of the time. That is why sterility assurance depends on validated sterilisation or aseptic processes, media fills, environmental monitoring and process controls. The test confirms the process; it does not prove the batch is sterile.

Every session should include negative controls of the media, rinse fluids and diluents handled exactly like the samples, to show the test itself did not add contamination. The test environment is monitored during the session: active air and settle plates in the isolator or Grade A zone, and glove prints of the analyst at the end. For isolators, record the decontamination cycle and glove integrity checks; for a LAF or BSC, record its qualification status and airflow. These records are the evidence needed if a positive ever has to be investigated.

The endotoxin limit is K divided by M. K is the threshold pyrogenic dose: 5 EU/kg for most parenterals and 0.2 EU/kg for intrathecal products. M is the maximum dose per kg of body weight given in one hour, using 70 kg for an adult. The maximum valid dilution (MVD) is (endotoxin limit × concentration of the test solution) divided by λ, the lysate sensitivity or lowest standard. Example: a 10 mg/kg dose gives a limit of 0.5 EU/mg; at 50 mg/mL with λ = 0.125 EU/mL, the MVD is 200.

The pharmacopoeias describe gel clot methods (a limit test giving pass or fail at the lysate sensitivity, and a semi-quantitative endpoint method), and photometric methods: kinetic turbidimetric, kinetic chromogenic and endpoint chromogenic. All use Limulus amebocyte lysate, which reacts with endotoxin through an enzyme cascade. Recombinant Factor C and recombinant cascade reagents are animal-free alternatives now described in Ph. Eur. and USP. Gel clot is the referee method if results disagree. Photometric methods give numbers, which help trending and investigations.

Endotoxins are lipopolysaccharides (LPS) from the outer membrane of Gram-negative bacteria. They are released when bacteria die or divide. The lipid A part causes the toxic effect: when injected, it triggers immune cells to release cytokines, causing fever, and at high doses septic shock. Endotoxins are very heat-stable and survive normal steam sterilisation, so a sterile product can still be pyrogenic. They are removed by dry heat depyrogenation, for example 250 °C for 30 minutes, or controlled by preventing Gram-negative growth in water and raw materials.

The BET is incubated at 37 ± 1 °C, the temperature at which the Limulus enzyme cascade works as intended. For the gel clot method, tubes are placed in a dry block or water bath and left undisturbed for 60 ± 2 minutes before reading; vibration during incubation can break a forming gel and give a false negative. For kinetic methods, the plate reader holds the plate at 37 °C throughout the run. Temperature equipment must be calibrated, and the actual temperature recorded for each test.

Lambda (λ) is the sensitivity of the test. In the gel clot method, it is the labelled sensitivity of the lysate, the lowest endotoxin concentration that will form a firm gel, for example 0.03, 0.06 or 0.125 EU/mL. It must be confirmed for each new lysate lot. In photometric methods, λ is the lowest endotoxin concentration on the standard curve. λ is used in calculating the maximum valid dilution, so a more sensitive lysate allows greater dilution to overcome product interference.

In kinetic methods, the positive product control (PPC) is the product spiked with a known amount of endotoxin, usually at the middle of the standard curve. The measured recovery of the spike must be between 50% and 200%. Below 50% indicates inhibition; above 200% indicates enhancement. Either way, the result for that sample is invalid. The usual fix is to dilute the product further, within the MVD, or treat it, and repeat. Persistent failures mean the interference testing needs to be revisited.

For photometric BET methods, the standard curve must have an absolute correlation coefficient (|r|) of at least 0.980, across at least three endotoxin concentrations run in replicate. The curve is usually plotted as log onset time against log endotoxin concentration for kinetic methods. A low correlation suggests pipetting errors, contaminated standards or reagent problems. The lowest standard defines λ for the method. Check also that the negative control gives a result below the lowest standard, otherwise the run is invalid.

After 60 ± 2 minutes at 37 °C, each tube is gently removed and inverted 180° in one smooth movement. If a firm gel has formed that stays intact at the bottom of the tube, the result is positive. If no gel has formed, or the gel collapses or breaks and runs down the tube, the result is negative. Viscous liquid is not a positive. Read every tube the same way and record each result. The test is valid only if the positive and negative controls behave as expected.

Endotoxin sticks firmly to glass and other surfaces and is not destroyed by autoclaving, so glassware used for BET or for parenteral products must be depyrogenated. Dry heat is the usual method, for example 250 °C for at least 30 minutes, or another validated time-temperature combination. Validation uses endotoxin indicators spiked with a known amount, usually at least 1,000 EU, placed in the coldest parts of the load, and must show at least a 3-log reduction. Depyrogenated items are stored covered to prevent recontamination.

For intrathecal products, the threshold pyrogenic dose K is 0.2 EU/kg, much lower than the 5 EU/kg used for most parenterals. The cerebrospinal fluid is very sensitive to pyrogens, so even small amounts of endotoxin can cause serious reactions. The lower K gives a much tighter endotoxin limit, which in turn reduces the maximum valid dilution. That can make interference harder to overcome, so a more sensitive lysate or a kinetic method with a lower standard is often needed for intrathecal products.

The enzyme cascade in LAL works best at a pH of about 6.0–8.0, or the range stated by the reagent manufacturer. The pH of the mixture of sample and lysate is what matters, not just the sample. Products outside this range can inhibit the reaction. Adjust pH with endotoxin-free hydrochloric acid, sodium hydroxide or a suitable buffer, all prepared in LAL reagent water. Dilution often brings pH into range on its own. Record the pH check during method validation and in routine testing if adjustment is needed.

Typical microbial action levels are 100 CFU/mL for Purified Water and 10 CFU/100 mL for Water for Injection, as given in USP <1231> and used widely in India under IP and Schedule M. These are action levels for process control, not pass/fail specifications for the water itself. WFI must also meet an endotoxin limit of less than 0.25 EU/mL. Both waters must meet conductivity (USP <645>) and TOC (USP <643>) requirements. Sites set their own lower alert levels from trend data.

R2A is a low-nutrient medium designed for water testing. Bacteria living in pharmaceutical water systems are adapted to very low nutrient levels and are often stressed by heat, ozone or chlorine. On rich media such as plate count agar, many of them grow poorly or not at all, giving false low counts. R2A, combined with longer incubation, for example 5–7 days at 30–35 °C, recovers these slow-growing organisms better. The chosen medium and incubation should be supported by a comparison study on your own water.

The TOC limit is 500 ppb (0.5 mg/L) of carbon for both Purified Water and Water for Injection under USP <643> and Ph. Eur. TOC measures organic contamination, which can come from feed water, resins, gaskets or biofilm. Online TOC analysers are common in modern systems and give continuous data. A rising TOC trend, even within limits, can be an early sign of biofilm, resin breakdown or a problem with RO membranes, so trend it along with microbial counts.

WFI loops are usually kept at about 65–80 °C because few organisms can grow at that temperature. This prevents biofilm from forming on pipe walls and keeps counts very low. The water is cooled only at points of use through sub-loops or heat exchangers, where the cooled section is kept short and sanitised regularly. Hot loops also make sanitisation simpler. Temperature should be recorded continuously, with alarms, because a drop in temperature for a long period can allow organisms to establish.

Ambient Purified Water is controlled by a combination of design and operation. Keep water circulating continuously with turbulent flow, typically above 1 m/s in the loop. Use UV lamps to reduce organisms in the circulating water. Sanitise periodically with hot water (around 80 °C) or ozone, at a frequency supported by trend data. Avoid dead legs, use smooth sanitary pipework with good slopes, and maintain the RO and EDI units. Monitor each point of use regularly and trend the results against alert and action levels.

A dead leg is a section of pipe where water does not flow, such as a branch to a valve that is rarely opened. Water stagnates there, warms up or cools down, and allows bacteria to attach and form biofilm. The biofilm can then release organisms and endotoxin into the main loop. Good design keeps dead legs as short as possible, often described as less than three pipe diameters from the main loop, and uses zero-dead-leg valves at points of use. Dead legs found during walk-downs should be removed.

Phase 1 lasts 2–4 weeks. Every point of use and critical point is sampled daily while operating, cleaning and sanitisation SOPs are developed, and water is not used for production. Phase 2 lasts another 2–4 weeks with the final SOPs, showing the system consistently produces water of the required quality; at the end, water can usually be released for use. Phase 3 is one year of routine monitoring that captures seasonal variation in feed water. After Phase 3, routine monitoring and trending continue for the life of the system.

For WFI, filter at least 100 mL through a 0.45 µm membrane and place the membrane on R2A or another validated medium. A large volume is needed because the action level is 10 CFU per 100 mL, so testing only 1 mL would not detect counts at that level. For Purified Water, 1 mL by pour plate is the minimum, but many sites filter larger volumes, such as 10 or 100 mL, for better sensitivity. Sample containers must be sterile, and samples tested within the validated hold time.

Gram-negative non-fermenting rods are the classic water-system organisms: Pseudomonas, Ralstonia, Burkholderia, Sphingomonas, Stenotrophomonas, Methylobacterium and Brevundimonas. They grow in low-nutrient water, form biofilms on pipe walls and release endotoxin. Finding the same species at several outlets, or repeatedly at one point, suggests biofilm in the loop. Burkholderia cepacia complex is a particular concern for non-sterile aqueous products, because it has caused product recalls. Identification of water isolates should therefore be routine, not only after excursions.

Sample in exactly the way production uses the water: from the same outlet, through the same hose or connection, after the same flush time. If production fills a vessel through a hose, sample through that hose, because contamination in the hose reaches the product. Do not sanitise the outlet with alcohol just before sampling unless production does the same. Use sterile containers, label them correctly, and deliver them to the lab promptly or refrigerate at 2–8 °C and test within the validated hold time.

A typical PW system starts with pre-treatment: a multimedia filter to remove suspended solids, a water softener to remove hardness, and an activated carbon filter or sodium metabisulphite dosing to remove chlorine, which would damage RO membranes. Reverse osmosis follows, often in two passes, then electrodeionisation (EDI) to polish the water further. A UV lamp reduces organisms before storage. The water is held in a storage tank with a vent filter and circulated through a distribution loop to points of use, usually with UV and periodic sanitisation.

TAMC, the total aerobic microbial count, is counted on Soybean-Casein Digest Agar at 30–35 °C for 3–5 days, and any fungi that grow on that medium are included in the TAMC. TYMC, the total combined yeasts and moulds count, is counted on Sabouraud Dextrose Agar at 20–25 °C for 5–7 days, and any bacteria that grow on SDA are included in the TYMC. If bacterial growth on SDA causes the TYMC to exceed its limit, SDA with antibiotics may be used, but this must be justified.

USP <62> and Ph. Eur. 2.6.13 describe tests for bile-tolerant Gram-negative bacteria, Escherichia coli, Salmonella, Pseudomonas aeruginosa, Staphylococcus aureus, Clostridia and Candida albicans. Which ones apply depends on the dosage form, as set out in USP <1111>. For example, absence of E. coli is required for oral products, absence of S. aureus and P. aeruginosa for cutaneous and nasal products, and absence of C. albicans additionally for vaginal products. Each test uses enrichment and selective media, followed by confirmation of suspect colonies.

Under USP <1111>, an acceptance criterion written as a power of ten is interpreted with a built-in tolerance for count variability. A criterion of 10¹ CFU means a maximum acceptable count of 20, 10² means a maximum of 200, and 10³ means a maximum of 2,000. So for a product with a TAMC limit of 10² CFU/g, a result of 180 CFU/g complies and 210 CFU/g does not. Specifications should state clearly whether they use this interpretation, especially for internal or customer-specific limits.

Buffered sodium chloride–peptone solution at pH 7.0, or phosphate buffer at pH 7.2, are the usual diluents, as listed in USP <61>. They keep organisms alive without allowing them to multiply during preparation. For poorly soluble or fatty products, polysorbate 80 is added, often at 1 g/L, and the sample may be warmed to not more than 40 °C, or in exceptional cases 45 °C, to help dispersion. Neutralisers can be added for antimicrobial products. The diluent and preparation are fixed during method suitability.

USP <61> describes three methods. Membrane filtration passes the prepared sample through a 0.45 µm membrane, which is then placed on agar; it suits antimicrobial products and large volumes. The plate count method includes pour plating, where 1 mL is mixed with molten agar, and surface spreading, where a smaller volume is spread on the agar surface. The most probable number (MPN) method estimates counts from growth in a series of dilution tubes; it is the least precise and is used only when no other method is suitable.

In MLT method suitability, the mean count recovered from the product inoculated with the test organism must be within a factor of 2 of the mean count from the control, which has the same inoculum in diluent without product. That means recovery between 50% and 200%. The comparison must be made against a parallel control plated at the same time, not against the theoretical inoculum. If recovery fails, use dilution, neutralisers, membrane filtration with rinsing, or a combination, and repeat until the method passes.

Prepare the product as for the count test and transfer a defined quantity, usually equivalent to 1 g or 1 mL, into Soybean-Casein Digest Broth, then incubate at 30–35 °C for 18–24 hours. Subculture 1 mL into 100 mL of MacConkey broth and incubate at 42–44 °C for 24–48 hours; the higher temperature favours E. coli over other coliforms. Streak onto MacConkey agar and incubate at 30–35 °C for 18–72 hours. Pink, lactose-fermenting colonies are suspect and must be confirmed by identification tests. No suspect colonies, or negative confirmation, means the product complies.

Enrich the prepared product in Soybean-Casein Digest Broth at 30–35 °C for 18–24 hours. Transfer 0.1 mL into 10 mL of Rappaport Vassiliadis Salmonella Enrichment Broth and incubate at 30–35 °C for 18–24 hours. Then streak onto Xylose Lysine Deoxycholate (XLD) agar and incubate at 30–35 °C for 18–48 hours. Typical Salmonella colonies are well developed and red, with or without black centres from hydrogen sulphide production. Suspect colonies are confirmed by identification tests. The test is usually done on 10 g or 10 mL of product.

Cetrimide agar is the selective medium for Pseudomonas aeruginosa. Cetrimide is a quaternary ammonium compound that inhibits most other bacteria, while P. aeruginosa tolerates it. After enrichment in Soybean-Casein Digest Broth at 30–35 °C for 18–24 hours, streak onto cetrimide agar and incubate at 30–35 °C for 18–72 hours. P. aeruginosa colonies are often greenish because of pyocyanin pigment and may fluoresce under UV. Growth of suspect colonies needs confirmation by identification tests before the product is declared non-compliant.

Mannitol salt agar is used for Staphylococcus aureus. Its high salt concentration, about 7.5% sodium chloride, inhibits most other bacteria, while staphylococci tolerate it. Mannitol and a phenol red indicator make it differential: S. aureus ferments mannitol, producing yellow colonies surrounded by yellow zones. After enrichment in Soybean-Casein Digest Broth, streak onto mannitol salt agar and incubate at 30–35 °C for 18–72 hours. Yellow colonies with yellow zones are suspect and are confirmed by identification tests, such as coagulase.

Steam sterilisation is validated with Geobacillus stearothermophilus spores, usually on carriers or in self-contained indicators with a population of about 10⁶ spores and a D-value at 121 °C of at least 1.5 minutes. This organism is highly resistant to moist heat, so killing it gives strong assurance that less resistant organisms are also killed. BIs are placed at the coldest, hardest-to-reach positions in the load, as identified by temperature mapping. After the cycle, they are incubated, typically at 55–60 °C, and must show no growth, while unexposed positive controls must grow.

F₀ is the equivalent time, in minutes, at 121.1 °C delivered by a steam sterilisation cycle, calculated with a z-value of 10 °C. It lets you compare cycles that run at different temperatures or have long heating and cooling phases, because time spent at other temperatures is converted into its lethal equivalent at 121.1 °C. For example, one minute at 111.1 °C contributes about 0.1 minute of F₀. Many cycles target an F₀ of at least 15 minutes for overkill, while product cycles may use lower F₀ with bioburden-based justification.

A media fill, or aseptic process simulation, replaces the product with a sterile growth medium, usually soybean-casein digest medium, and runs it through the aseptic process exactly as a real batch, including worst-case interventions, durations and personnel. The filled units are incubated, typically for 14 days, and inspected for growth. The target is zero contaminated units. Under EU GMP Annex 1 (2022), any contaminated unit means a failed simulation, which must be investigated and normally repeated before routine aseptic production continues.

Start with suspension tests to screen agents and concentrations. The core is a surface (coupon) test: inoculate representative surfaces such as stainless steel, glass, vinyl and epoxy with standard organisms and your own environmental isolates, including a spore-former and a mould. Apply the disinfectant at the in-use concentration for the contact time operators actually achieve, neutralise it, recover survivors and calculate log reduction against untreated controls. Common targets are 3 log for vegetative bacteria and 2 log for spores (USP <1072>). Also validate diluted solution expiry and confirm with in-situ EM data.

The D-value is the time, or radiation dose, needed under a given condition to reduce a microbial population by 90%, which is one log. For example, if a spore population has a D-value of 2 minutes at 121 °C, then 2 minutes at that temperature reduces 10⁶ spores to 10⁵, and 12 minutes reduces them to 10⁰. D-values are specific to the organism, the conditions and the medium the organism is in. They are used to choose biological indicators and to calculate how long a cycle must run for a target sterility assurance level.

The z-value is the temperature change needed to change the D-value by a factor of 10. If an organism has a z-value of 10 °C and a D-value of 2 minutes at 121 °C, its D-value would be about 20 minutes at 111 °C and about 0.2 minutes at 131 °C. The z-value lets you convert time at one temperature into lethal equivalent time at another. For steam sterilisation calculations such as F₀, a z-value of 10 °C is used by convention. Dry heat calculations commonly use 20 °C.

A sterility assurance level (SAL) of 10⁻⁶ means the probability that any single unit is non-sterile after sterilisation is no more than one in a million. It is not something you can measure by testing, because you would need millions of units. Instead, it is calculated from the starting bioburden and the lethality of the validated process, using D-values. For example, an overkill steam cycle that achieves a 12-log reduction of a resistant BI gives SAL 10⁻⁶ even for high bioburden. Terminally sterilised products normally target this level.

Depyrogenation is validated with endotoxin rather than spores, because the aim is to destroy endotoxin, which is much more heat-resistant than spores. Endotoxin indicators, vials or carriers spiked with a known amount (often 1,000 EU or more), are placed at the coldest positions in the load. After the cycle, the remaining endotoxin is measured, and at least a 3-log reduction must be shown. Bacillus atrophaeus spores are used separately when validating dry heat sterilisation, which needs lower temperatures than depyrogenation.

Vaporised hydrogen peroxide cycles for isolators, pass boxes and rooms are validated with Geobacillus stearothermophilus spores, usually around 10⁶ per carrier on stainless steel or Tyvek. This organism is among the most resistant to hydrogen peroxide. BIs are placed at worst-case locations: corners, glove fingers, behind equipment and inside transfer ports. A 6-log reduction is typically required at every location. Chemical indicators are used during cycle development to map distribution quickly, and residual peroxide is measured after aeration.

The Bowie-Dick test is a daily check, usually run at the start of the day in an empty chamber, for pre-vacuum steam sterilisers. A standard test pack containing an indicator sheet is placed in the chamber and a short cycle is run, typically 3.5 minutes at 134 °C. If air removal and steam penetration are good, the indicator changes colour evenly. An uneven colour, with a lighter area in the centre, shows residual air or a leak, and the steriliser must not be used until the cause is found and fixed.

OOS, out of specification, is a test result that falls outside a product or material specification, such as a failed MLT, a sterility test positive or an endotoxin result above the limit. It directly affects batch disposition. OOL, out of limit, is commonly used for environmental monitoring or utility results that exceed alert or action levels, such as an EM count above the Annex 1 limit. Both must be documented and investigated, but OOS investigations follow the formal phased approach, while OOL handling depends on the grade, level and risk.

ALCOA+ describes the qualities GMP data must have. Attributable: you can tell who recorded it and when. Legible: it can be read throughout its life. Contemporaneous: it was recorded at the time of the activity. Original: it is the first record, or a certified true copy. Accurate: it is correct and checked. The plus adds Complete (nothing missing, including repeats), Consistent (same practice, logical sequence), Enduring (lasts for the retention period) and Available (can be retrieved when needed). Regulators use it to judge data integrity.

The main reason is to bring in a sporicidal agent periodically, because routine disinfectants such as alcohols and quaternary ammonium compounds do not kill bacterial spores. EU GMP Annex 1 requires periodic use of a sporicide, with the frequency based on EM data and the contamination control strategy. The older idea that organisms become resistant to a disinfectant, so you must rotate between two similar agents, is not well supported for disinfectants used at proper concentrations. Using a sporicide on a planned schedule matters far more than alternating routine agents.

A deviation is any departure from an approved procedure, specification, standard or instruction, whether planned or unplanned. Examples include an incubator temperature excursion, a missed EM sample, or a step done out of sequence. It must be recorded promptly, assessed for impact on product quality, classified by risk (minor, major, critical), investigated in proportion to that risk, and closed with corrective and preventive actions where needed. Deviations are trended to spot recurring problems. Hiding or not recording a deviation is itself a serious GMP failure.

CAPA stands for corrective and preventive action. Corrective action removes the cause of a problem that has already happened, so it does not recur; for example, replacing a faulty incubator controller. Preventive action removes the cause of a potential problem before it happens; for example, adding temperature alarms to all incubators after one failure. Each CAPA should be linked to a confirmed root cause, have an owner and due date, and include an effectiveness check with a measurable target. Retraining alone is rarely an effective CAPA.

Change control is a formal system for managing changes to processes, equipment, facilities, materials, methods, computer systems and documents. A change is proposed, its impact on product quality, validation and regulatory filings is assessed, it is approved by the right functions including QA, then implemented and verified. This ensures that no change affects quality without evaluation. In microbiology, examples include changing a disinfectant, a medium supplier, a test method or EM locations. Emergency changes still need documentation, often retrospectively within a short period.

Draw a single line through the incorrect entry so the original remains readable, write the correct entry nearby, and add your signature or initials, the date and a short reason for the correction, such as 'transcription error'. Never use correction fluid, never overwrite, never scribble out an entry and never tear out pages. If a record is badly damaged, follow the SOP for replacing it with a certified copy. For electronic records, the system's audit trail must capture the change, the user, the time and the reason.

A contamination control strategy (CCS) is the documented, site-wide plan for controlling micro-organisms, particles and endotoxins, required by EU GMP Annex 1 (2022). It brings together facility and equipment design, utilities, raw materials, personnel and gowning, cleaning and disinfection, process controls, monitoring, validation and continuous improvement, and explains how they work together. It points to the detailed SOPs and reports rather than repeating them. It must be reviewed regularly and updated after significant changes or trends, such as repeated EM excursions.

Qualification proves that equipment, utilities or facilities are suitable and work correctly. It usually follows design qualification (DQ), installation qualification (IQ), operational qualification (OQ) and performance qualification (PQ). Validation proves that a process, test method, cleaning procedure or computer system consistently produces the intended result. In practice, qualified equipment is a prerequisite for validation: you cannot validate a sterilisation process in an autoclave that is not qualified. For example, an incubator is qualified, while an MLT method is validated through method suitability.

An audit trail is a secure, computer-generated, time-stamped record of actions on GMP data: who created, modified or deleted a record, when, what the old and new values were, and why. It must be enabled at all times and must not be editable by users. Audit trails are reviewed as part of data review, focusing on critical events such as changed results, reprocessing, deleted data or changes to methods. In micro labs, endotoxin readers, identification systems, particle counters and LIMS all need audit trails.