Table I: Temperature data interpretation for a heat distribution study during exposure phase.Function Minimum temperature Run Number 1 2 3 4 1 2 3 4 1 2 3 4 T5 120.43 121.71 121.35 121.85 123.27 121.91 122.37 122.16 121.76 121.84 121.88 122.00 121.87 0.10 T0 121.87 121.81 121.91 122.09 122.02 122.13 122.22 122.44 121.95 121.97 122.05 122.21 122.05 0.12 Values within the Run Thermocouple Numbers for Heat Distribution T1 T2 T3 T4 T6 121.85 121.37 121.79 121.71 121.53 121.82 121.77 121.92 121.81 121.39 121.92 121.84 121.75 121.71 121.53 122.07 122.01 122.02 122.10 121.66 122.02 121.96 122.16 122.14 121.92 122.02 121.98 122.15 122.20 122.02 122.14 122.08 122.30 123.45 122.00 122.42 122.34 122.35 122.46 122.16 121.93 121.87 121.99 121.98 121.73 121.94 121.89 122.06 122.02 121.72 122.03 121.96 122.15 122.18 121.77 122.19 122.11 122.12 122.22 121.89 122.02 121.96 122.09 122.10 121.84 0.12 0.11 0.06 0.11 0.12 T7 121.62 121.58 121.62 121.29 122.05 122.07 122.12 122.83 121.82 121.85 121.85 121.94 121.84 0.08 T8 121.63 121.78 121.70 121.83 122.03 122.25 122.16 122.39 121.81 122.03 121.93 122.01 121.95 0.09 Values across the Run 120.43 121.39 121.35 121.28 123.27 122.24 123.44 122.82 121.89 121.92 121.98 122.08

Maximum temperature

Mean temperature

Mean of the four runs Standard deviation for the means t value

1 2 3 4 1 2 3 4

—

0.07

0.10

0.28

0.01

0.02

0.71

0.69

0.30

0.09 0.10 0.13 0.12— 0.59 0.15 0.002

One thermocouple must always be located next to the autoclave temperature probe, which is usually placed in the chamber drain. The reason for this practice is that, theoretically

, the“cold spot” of a chamber is in the drain. Depending on various chambers’ geometry, sometimes the location of the cold spot in the chamber drain is not confirmed in practice. The OQ phase consists of a verification of the equipment functionality. A series of checks and tests should be performed after the unit is installed. The proper operation of alarms and safety devices also must be tested. For a prevacuum cycle involving an air-removal phase, the attainability, effectiveness, and consistency in time of vacuum conditions must be tested. Chamber mapping consists of heat distribution studies throughout the chamber for each sterilizing temperature used. Heat distribution studies determine during processing whether the temperature is uniform and reproducible throughout the empty chamber and localize the cold spots within the chamber. These studies also confirm that the system works within the specified limits throughout the entire range of operational parameters. The sterilizing temperatures are programmed into the autoclave controls. For each sterilizing temperature being validated, the empty chamber should be monitored three times during a period of one hour, and the reproducibility of the results obtained must be compared. The comparative readings of the reference temperature sensor, the thermocouple near the temperature sensor, and the temperature displayed by the autoclave recorder should be 1 C. During the exposure phase, a difference no greater than 2.5 C between the temperature of100

the chamber’s coolest spot and the mean chamber temperature is acceptable (8). Performance qualification (PQ) phase. PQ represents the confirmatory phase of the validation program and consists of tests performed with the autoclave chamber under loaded conditions. During the PQ phase, which is sometimes referred to as process validation, the following objectives must be attained (1):● demonstration of the uniformity and effectiveness of the process in inactivating or removing microorganisms to the required safety level● demonstration of the reproducibility of the process—through the use of sufficient cycles● demonstration of the compatibility of the process with the items to be sterilized—through the assessment of the influence of the sterilization process on the products. The PQ phase consists of studies that use thermocouples inserted into the articles being sterilized and studies that use BIs in operationally fully loaded autoclave conditions (2). Heat penetration studies are considered the most critical component of the entire validation program. These studies are intended to find areas in the loads that are difficult to penetrate or heat. When selecting the monitoring sites, one must take into account the cold spots previously found during the monitoring of the empty chamber. Heat penetration studies must demonstrate the reproducibility of a cycle in relation to the loads and the effectiveness of the killing effect throug

hout the chamber and load. Selecting loading configurations must be performed very carefully and should be based on worst-case situations. The maxiwww.phar

Pharmaceutical Technology

OCTOBER 2002

Table II: Temperature data interpretation for a heat penetration study during exposure phase.Function Minimum temperature Run Number T10 1 120.26 2 120.84 3 120.62 4 120.81 1 123.10 2 122.35 3 122.83 4 122.91 1 121.67 2 121.53 3 121.63 4 121.58 121.60 1 2 3 4 1 2 3 4 0.06 T11 120.91 121.35 120.95 121.48 122.33 121.92 122.27 122.14 121.68 121.63 121.62 121.77 121.67 0.06 Values within the Run Thermocouple Numbers for Heat Penetration T12 T13 T14 T15 T16 T17 121.05 120.76 120.69 120.86 120.57 120.67 121.04 120.80 121.08 120.73 120.80 120.87 121.18 120.96 120.86 120.92 120.45 121.15 121.31 121.33 121.19 121.38 121.27 121.20 122.20 122.31 122.66 122.34 122.49 122.52 122.11 122.31 121.99 122.40 122.10 122.17 122.01 122.04 121.91 122.09 121.85 121.71 122.26 122.19 122.32 122.20 122.37 122.27 121.69 121.61 121.64 121.60 121.53 121.55 121.55 121.53 121.56 121.57 121.56 121.63 121.54 121.53 121.45 121.54 121.33 121.52 121.81 121.81 121.84 121.83 121.82 121.84 121.64 121.62 121.62 121.63 121.56 121.63 0.12 0.13 1.66 0.13 0.2 0.14 T18 120.91 121.07 120.22 121.14 122.28 122.04 122.40 122.30 121.54 121.64 121.33 121.81 121.58 0.19 T19 121.15 121.14 120.87 121.23 121.92 121.98 121.84 122.25 121.56 121.63 121.39 121.82 121.60 0.17 Values across the Run 120.26 120.73 120.21 120.81 123.10 122.40 122.83 122.90 121.61 121.58 121.49 121.79

Maximum temperature

Mean temperature

Mean of the four runs Standard deviation for the means t value

—

0.17

0.57

0.83

0.4

0.69

0.71

0.69

0.83

0.98

1.04 0.04 0.10 0.07— 0.35 0.007 0.00002

mum number of bottles in the loads for the most-critical or used products should be decided. Ideally, three consecutive validation runs should be performed per loading configuration. During testing, if some configurations cannot be successfully sterilized because they are too big, the worst case in regard to the size of the loading configuration should be redesigned. If the results are still unsatisfactory, then the cycle should be redesigned, which usually can be achieved by modifying the exposure time. Heat penetration results must show the slowest heating points in the load and must ensure that the minimum time and temperature requirements are met. The success of a qualified cycle depends on the determination of the F0 value measured inside the item located at the coldest spot. It should be confirmed that a minimum prescribed F0 value is delivered consistently throughout the autoclave chamber. Microbiological validation, which consists of tests involving BIs, should be performed concomitant with heat penetration studies to verify independent of the temperature data results that the minimum F0 value is met at the coldest spot of the load. For terminal moist heat sterilization, heat penetration studies must demonstrate an

SAL of 10 6 or higher. The BIs should be placed in operational, maximum loads at the locations presumed to be the least accessible to the sterilizing agent. One should place the BIs next to the thermocouples and inside the item being tested. They must not be placed directly on surfaces or outside the loads because the heating process is rapid in those areas and the BIs may get killed faster. The number of BIs used per load depends on the size and the complexity of the load.102

For statistical significance, tests with BIs that are performed during the PQ phase should be conducted at least three times per loading configuration. After sterilization, the BIs should be incubated for seven days and checked daily for growth. As a part of the documentation, information provided by the manufacturer concerning the spore lot number (i.e., D and z values for the current lot) must be available. Demonstrating the integrity of the product after sterilization represents another aspect of process validation. The product must be tested for possible physical and chemical changes that may have occurred during sterilization. In this regard, the pH of solutions and the physical condition and appearance of goods before and after sterilization must be verified. For culture media, tests for growth promotion that are recommended by USP should prove whether the culture media show growth following the sterilization process. Qualification–validation report. The final stage of the validation program requires the documentation of all acquired data. The qualification–validation report summarizes the overall results of validation. It includes the calibration certificates for calibrating instrumentation, calibration records, and methods for calibrating the measuring instruments, gauges, and recorders as well as the accuracy verification data of thermocouples (4). Test data such as high–low temperature ranges, average temperatures during exposure time, minimum and maximum F0 values achieved for every load configuration, run date and time, and autoclave records must be included. The qualification–validation report is not complete unless it contains evidence checks for the availability of an instrumentwww.phar

Pharmaceutical Technology

OCTOBER 2002

Table III: Accumulated F0 value interpretation for a heat penetration study.Std. Mean Std. Error F0 Dev. of per per the Run Run Mean24.56 24.35 23.49 25.87 24.57 0.59 0.53 0.96 0.55 1.08 0.18 0.16 0.31 0.17

Run Function No. T10 F0 1 25.822 3 4 23.95 23.85 24.68 24.58

T1125.04 24.81 24.55 26.05 25.11

T1224.92 24.01 24.26 26.13 24.83

Thermocouple Number T13 T14 T15 T1623.86 23.50 23.96 26.14 24.36 24.50 24.20 23.03 26.05 24.45 24.35 23.91 24.42 26.43 24.78 23.76 24.25 22.02 25.65 23.92

T1724.42 24.80 24.08 26.32 24.91

T1824.42 24.90 22.03 25.15 24.13

T1924.51 25.12 22.69 26.06 24.60

95% Confidence Interval24.14* 23.97* 22.80* 25.47* 24.98* 24.73* 24.17* 26.26*

t Values across the Run— 0.4 0.0

09 0.00008

Mean of the four runs Standard deviation for the means t value

24.23** 24.90**

0.90

0.66

0.95

1.20

1.24

1.12

1.50

0.99

1.43

1.42

—

0.38

0.71

0.78

0.87

0.79

0.49

0.64

0.62

0.98

* The confidence interval per run was calculated considering t for 0.05 (95%) confidence level 2.262 (N-1 9 degrees of freedom). ** The confidence interval across runs was calculated considering the critical value for 95% level of confidence 1.96.

logbook, the standard operating procedures (SOPs) used with the autoclave, procedures for preventive and unscheduled maintenance, and recalibration programs. The validation report also should include the user’s training records. Training should begin only after the validation has been accomplished to ensure that users will implement already-validated loading configurations. The qualification–validation report must include the drawings of all loads tested. The location of the thermocouples and BIs must be specified in each drawing. Any specification deviations that were encountered during validation activities and the procedures that followed their discovery should be reviewed. After the validation activities are completed, all data and documents that were accumulated must be revised, approved, and certified by both the owner of the autoclave and the contractor in cases when the validation was performed with third parties or by the manufacturer. Once the autoclave is qualified and the loading configurations are validated, the following documents must be present in the laboratory at all times:● diagrams of various loading configurations stamped with their effective dates● a list of the determined lag times for each loading configuration● temperature–pressure recorder printouts

●

user, calibration, and maintenance SOPs.

When requalification–revalidation activities are required. Once a unit has been installed and qualified, it does not need requalification, unless the unit is reinstalled in another location or has undergone major modifications. Only revalidation is performed periodically. Generally, the extent of requalification–revalidation depends on the nature of the changes that occur and how they affect the already-validated sterilization cycles. The responsibility for deciding which validation activities should be reperformed must be assigned. Validation frequency must be stipulated in the validation documents. The need for possible revalidation should be assessed every 12–24 months, and revalidation should occur whenever major repairs have been performed. Any modifications to control systems should be evaluated to confirm that process conditions delivered to the load are comparable with those originally validated. However, the validation should be repeated whenever new sterilization cycles, loading configurations, or major changes in the sterilization procedure are introduced. Change control. A system for change control should be implemented that establishe

s when the qualification–validation process should be repeated (1). The nature of the change must be assessed to determine the potential effects on the cycles and load. Changes that could affect sterilization conditions are not allowed to be introduced without documenTable IV: Sterility assurance levels (SALs) attained for each run. tation. The change-control procedure must follow a F0 Values Representing change-control SOP. the Calculated Lower Changes in any of the following items can invaliLimit of the 95% date an autoclave validation: pressure gauge, temperaRun Confidence Interval SAL ture gauge, control panel programming, temperature1 24.14 10–10.09 control valve, steam trap, and steam source. Changes 2 23.97 10–9.98 introduced by preventive or unplanned maintenance 3 22.80 10–9.2 procedures must be evaluated and documented in re4 25.47 10–10.98 gard to a decision for revalidation.Pharmaceutical TechnologyOCTOBER 2002

104

www.phar

Figure 1:Representation of the autoclave chamber (overhead view). Bis the number of the bioindicator ampul,and T is the thermocouplenumber. Heat-distribution thermocouples are marked with a numberand letter indicating their position in the chamber:u upper position,m middle position,b bottom position,r under the rail,

d in the chamber drain. T8was placed in the piping heading to thepressure gauge. The dotted rectangles represent the two rails of theautoclave chamber. The circles represent the 50 bottles,some of themcontaining thermocouples and bioindicator ampuls.

During routine use,personnel should ensure that the processspecifications established during qualification testing are fol-lowed and remain valid.The critical parameters governing theprocess must be monitored routinely to confirm that previouslydetermined conditions are achieved.

Estimation of probability of nonsterility of a load duringvalidation of the steam sterilization process

Qualification–validation activities were performed for a steamsterilizer,model 2020,from Amsco AHSC-South Africa (Pty)Ltd.The following example demonstrates the validation ofasaturated steam–vented cycle used with a load containing 50250-mL bottles filled with 200-mL trypticase soy broth (TSB).TSB is a heat-sensitive medium that is prone to alteration dur-ing the sterilization process.Therefore,the parameters for steri-lization had to be carefully chosen.The microbiology labora-tory referred to in this article uses the sterilized broth for microbiological testing ofnonsterile products and stab cultures.Materials and methods.As recommended by pharmaco-poeias (2,9,10),the reference sterilization conditions for liquidmedia (121 C for an exposure of15 min) were tested.A pre-liminary experiment was performed to determine the lag-timecharacteristic for this loading configuration.A simulation loadcomprising 50 bottles filled with 200 mL oftap water was usedfirst.

To monitor the temperatures attained in the chamber underloaded conditions,9 thermocouples were placed in various lo-cations within the chamber,and another 10 were placed in themiddle ofthe liquid contained in the bottles.The placement ofthe thermocouples in the chamber and the loading configu-ration are presented in Figure 1,which shows the overhead viewofthe autoclave chamber.

Timing the lag time began when the thermocouple placed inthe drain next to the sterilizer temperature probe reached thesterilizing temperature.Timing stopped when the last thermo-couple placed in the load reached the sterilizing temperature.The heat penetration studies were conducted concomitantwith the BI studies.An ampul containing a population of 106spores ofBacillus stearothermophiluswas placed next to eachthermocouple inserted into the load.In the loading diagram(see Figure 1) the positioning ofthe thermocouples and BIs isindicated by the letters T and B.For the ampuls,the numbermarked on each ofthe 50 bottles was assigned.After insertioninto the thermocouples,the sterilized ampuls were incubatedwith the positive controls (nonsterilized ampuls) for seven daysat 55–60 C and were checked daily for growth.As a means toverify the sterilization efficiency,5 ofthe 50 bottles containingTSB were incubated after sterilization for five days at 32 C.Thebroth was checked daily for growth.

To check the effectiveness ofthe culture medium after its steri-lization,the TSB was tested in accordance with the USP GrowthPromotion Test (11).The broth was tested for growth in thepresence ofBacillus subtillus,ATCC 6633,spores and incubatedat 37 C for 48 h.

Another means for verifying the integrity ofthe physicalproperties ofthe broths and buffers is to measure their pH val-ues before and after sterilization.For TSB a pH of7.3 0.2should be obtained after its preparation.

Results.For the simulation load,a lag time of 13 min wasmeasured.The TSB has a slightly higher viscosity than tap water.A lag time of15 min was considered sufficient to be added tothe required exposure time of15 min.

Temperature profiles recorded by using thermocouples in-serted into a simulation load have indicated that during the ex-posure phase the temperature ofthe load is slightly below theset temperature of121 C.To achieve the prescribed steriliza-tion temperature of121 C inside the load,the set temperatureofthe PLC was raised to 122 C.Therefore the experiments forthe validation loads were performed with the new set parame-ters of122 C and 30 min.As an example,Figure 2 shows themiddle portion ofthe temperature profiles achieved within theautoclave chamber,in the chamber drain,and inside the loadduring the second run.

Figure 2:Representation of the middle portion of the temperatureprofiles in the chamber recorded during the second run.

The data acquired by heat distribution thermocouples dur-ing four runs are summarized in Table I.The data acquired byheat penetration thermocouples introduced into the bottlesfilled with broth during four runs are summarized in Table II.When observing the mean temperature values per run cal-culated for heat distribution (see Table I),one will note that po-sition 4 is the hottest spot in the chamber,and positions 6 and7,situated under the rail,can be considered as the coldest spotswithin the chamber.Analysis ofthe data with the ttest indi-cated no statistically significant differences in mean tempera-ture values at a 95% level ofconfidence between the drain lo-cation and any ofthe arbitrarily chosen points in the chamber,with the exception ofthermocouple Tthe chamber farther from the steam inlet.3,which was situated onthe top corner ofThisfact is not critical because loads will not be placed in the topcorner.

The heat-penetration temperature data (see Table II) indi-cate positions 11 and 17 as the hottest spots and positions 14and 19 as the coldest spots in the chamber.The temperaturedata recorded at various points in the load proved not to be sta-tistically different from each other at a 95% level ofconfidence.Temperature readings taken every 30 s were used for the com-putation ofthe F0value (8).The value was calculated with

[1]

in which t (t 100 C attained during the total set time ofn t0).The sterilizing effects for temperaturesthe cycle,includ-ing the entire exposure phase and portions ofthe heating andcooling phases,were considered in the calculations.The timeat which the autoclave temperature probe reached the steriliz-ing temperature was considered as tthe cycle started.The lethality values (see Table III) were ana-0,and therefore the timinglyzed to determine the 95% confidence interval for the meanF0value per individual run.

The SAL achieved was determined for each ofthe four runs.The calculated lower limit ofthe 95% confidence interval wasintroduced in Equation 4 ofPart I ofthis article indicated byEP(9) for the calculation ofthe F0value.In rearranging thatequation,one can define a probabilistic term Pas the absolutevalue ofNtexpressed as

[2]

To calculate the probabilities ofnonsterility ofthe product,theequations obtained were solved for Nt.

As indicated in the certificate ofconformity provided by themanufacturer for the B.stearothermophilusspores,D6spores per vial.The results of121 was con-sidered 1.5 min and N0as 10theNtcalculated at the lower limit ofthe 95% confidence intervalare shown in Table IV.It can be concluded that,when repro-ducible sterilization conditions are applied to the studied load,a 95% confidence level exists that the probability ofa single or-ganism’s survival will be 10 9.2or better.

Discussion

The previously described approach is based on a worst-case assumption that a population of106spores per item exists andmust be sterilized and that the microorganisms in the biobur-den have the same heat resistance as the spores ofB.stearother-mophilus.However,achieving an SAL of10 12is not necessaryfor this situation because the raw materials used for the prepa-ration ofthe broth are pure and possess a low bioburden.An adequate SAL also was indicated by the results ofthe ex-periment using BIs.After sterilization,which was followed byseven days ofincubation at 55–60 C,all the spores containedin the ampuls were killed.The positive controls showed growthafter the incubation period.

The visual examination ofthe broth after five days ofincu-bation indicated no spore growth.A bottle containing unsteri-lized TSB was incubated as a positive control,and the mediumturned turbid after the incubation period.This confirmed theconclusion that the sterilization process was efficient.

The effectiveness ofthe culture medium after its steriliza-tion was also confirmed by the growth promotion test.Aftersterilization,the broth,spiked with B.subtillusand incubatedfor 48 h at 37 C,showed a positive spore growth.The fact thatthe medium turned turbid after the prescribed incubation pe-riod confirmed that its stability and chemical composition werenot altered through the sterilization process and that themedium still supports growth when spiked with a suitable microorganism.

Four pH measurements ofthe sterilized broth were performed.All pH values ranged between 7.15 and 7.17,which confirmedthe preservation ofthe initial physical and chemical propertiesofthe broth after sterilization.Also,a visual test performed bythree independent observers confirmed no color changes or pre-cipitation phenomena in the broth after sterilization.

It can be concluded that,for the studied load,the study’s ob-jectives for the PQ phase have been fulfilled.Monitoring thephysical parameters attained during the sterilization phase re-vealed an even distribution oftemperatures in the chamber dur-