This assignment provided an opportunity to collect product samples to assess the microbiological quality and safety of the apples and cider being inspected. Samples were collected for microbiological analysis of incoming whole apples, water used for washing, whole apples before pressing, finished apple cider before preservatives were added, and finished product with or without preservatives. Each sample consisted of four duplicate subsamples except for the finished product that consisted of ten duplicate subsamples. Each sample was also composited (a portion of each subsample was combined to form one composite sample). FDA collected and analyzed over 5000 subsamples for this assignment.
Each subsample was analyzed quantitatively by several methods to assess the quality of the product. These methods were: (1) aerobic plate counts (APC) reported in colony-forming units (CFU), (2) total coliforms (TC) reported as most probable numbers (MPN), fecal coliforms (FC) reported as MPN, and generic Escherichia coli (E. coli) reported as MPN. Fecal coliforms and generic E. coli are not pathogens but are used as indicators of fecal contamination. The subsamples were also analyzed for selected pathogens to assess the safety of the product. These analyses were for E. coli O157:H7, reported as positive or negative, and composite samples for Salmonella sp. also reported as positive or negative. The microbiological analytical methods utilized are those outlined in the Bacteriological Analytical Manual (BAM), 8th Ed, 1995.
FDA found one composite sample positive for a pathogen. A composite of four subsamples of the incoming apples at one firm was positive for Group B2 Salmonella sp. No finished products were positive for pathogens.
A summary of the inspectional findings at the firm with the positive Salmonella sp. sample provides insight into some of the issues faced by the cider industry in producing a safe product. FDA's inspection revealed the following.
- The firm has an annual production of 30,000 gallons. This places the firm in the top 25% by volume of the firms inspected. FDA had never inspected this firm previously, and did not have any record of any consumer complaints against the firm or its products.
- The firm produces other fresh fruit and vegetable juice products and juice blends. The firm also manufactures cut salad items such as trimmed broccoli, cut beans, cut mushrooms, and prepared carrots.
- The cider is sold retail to about 30 local markets and two distributors who sell it to hotels and restaurants. The firm has received two to three complaints annually on product taste and of diarrhea of unknown origin. No in-house investigations were made.
- The firm's apples are purchased mostly from intrastate orchards, but occasionally an interstate supplier is used, including foreign suppliers. The two most common sources of apples are a local intrastate grower (70%) and a commercial produce market in a nearby city. Apples from the produce market could come from anywhere in the country or the world.
- The apples are purchased in vendor-supplied and previously used cardboard boxes. The apples are transferred to large plastic bins and held under refrigerated storage until use.
- The overall sanitation at the firm exceeded what was found at most cider operations. The operation is located in a modern facility with stainless steel equipment. All employees wear white coats, plastic aprons, hair restraints, rubber gloves, and rubber boots. The facility is cleaned and sanitized at the end of each day, and rinsed again in the morning before production begins.
Analysis of sample results of apples under refrigerated storage yielded an APC "too numerous to count" in three of the four samples, and total coliform levels as high as 2,300 MPN/ml, for one lot sampled. The fecal coliform and E. coli MPN levels were below the level of detection.
The firm applies two controls to the raw apples prior to pressing.
- The apples are refrigerated at 33 -34° F from time of receipt until processing.
- The apples are sanitized, washed and wet brushed prior to being pressed. The sanitizing is done by soaking the apples for at least 10 minutes in water with chlorine concentrations of 200 PPM or more.
The sample of apples taken after sanitizing and washing but before pressing showed a median APC of 4000 CFU/ml, and a median total coliform level of 13 MPN/ml. Before and after comparisons can not be made because the apples sampled in the cooler and in production were from different shipments and suppliers (lots). The median levels for the finished product sample result, which again did not reflect the same lot as the raw apples sampled, were APC at 9850 CFU/ml, total coliform at 23 MPN/ml, and fecal coliform and E. coli at <3 MPN/ml.
The positive Salmonella sp. sample result on incoming apples illustrate that:
- microbial hazards can be present on the raw apples used for cider production;
- control measures should be applied to reduce the microbial hazards on the raw apples to acceptable levels; and
- sanitation controls should be applied to prevent microbial contamination on raw apples and shipping containers from being spread by employees and processing equipment into finished product.
A review of the analytical results of the other firms inspected by FDA showed that TC, FC, and E. coli counts were consistently low. There was generally little variation between results of different subsamples. E. coli O157:H7 was not detected. To facilitate the analysis of results, the APC, total and fecal coliform, and E. coli counts of each subsample was used to calculate a median for the sample, and these median levels were used in the analysis of results. For the purpose of the analyses, sample results that were below the level of detection of the test (<3/ml or g) or were positive but less than 3 were recorded as "3 MPN/ml" in order to be shown on the data plots.
The sample results, calculated as medians, were plotted logarithmically because of the wide range of the APC and total coliform counts. APC ranged from the hundreds to greater than one million. The sample results from all the firms were organized into sets with each set consisting of all samples taken at one point in the process and analyzed by one method. For example, all the results of samples of incoming apples analyzed for total coliform are one set. Each set of data were sorted and ranked in ascending order, and plotted. Each plot shows the distribution of one or more sets of results (i.e. APC, TC, FC, and generic E. coli) on one graph. The units of measure, shown on the Y-axis, for the APC are CFU's and are different than the units of measure for total and fecal coliform, and E. coli which are MPN's. The distribution of results, plotted on the X-axis, shows the cumulative number of samples that are less than or equal to a given microbial count. For example, on Table 12 the graph shows that there were 209 total samples in each set that were sorted in ascending rank order. On the plot of the total coliform sample results, 131 of 209 sample results were less than or equal to 3 MPN/ml, 208 of 209 fecal coliform samples were less than or equal to 3 MPN/ml, and all E. coli samples were less than or equal to 3 MPN/ml. On each graph APC counts are the top line plotted, the total coliform counts are the second line from the top, and the fecal coliform and generic E. coli plots are together at the bottom with a few counts at the end which are greater than 3 MPN.
Incoming Apples --- The overall microbial quality of the raw apples used by processors to produce cider is very good. Figure 12 shows that the APC and total coliform counts are normally low, and the fecal coliform and E. coli counts are consistently below the level of detection. There are, nevertheless, a few high counts for APC and total coliform that indicates the quality of these particular lots of apples was poor.
Figure 12. Microbial Quality of Incoming Apples Used for Cider (*)
(*) Generic E. coli levels are all less than or equal to 3 MPN/ml; and one fecal coliform count exceeded 3 MPN/ml
Spray-Wash Water --- Most firms used water from a public or private supply that was tested annually to wash apples at the time of receipt and/or just prior to pressing. In most instances the water was applied onto the apples by a spray-washer unit. The overall microbial quality of the water used for spraying is shown in Figure 13. Approximately half of the firms (80) had microbial levels less than or equal to 3 CFU or 3 MPN/ml for all test organism groups. Wash water with elevated APC and total coliforms levels was found at 30 firms. In 12 of the 30 firms, fecal coliforms and generic E. coli organisms were also found which indicates the water could be subject to fecal contamination. A linear regression analysis was performed on the log transformed total coliform data. It was found that a small but statistically significant (p = 0.0132, r = 0.25) positive association between levels of coliforms in the wash water and in the final cider product was evident, after adjusting for coliform levels in the incoming apple samples.
Figure 13. Microbiological Quality of Spray Wash Water.
Apples Before Processing --- The microbial quality of the apples sampled after washing but before processing is shown in Figure 14. These results show that the fecal coliform and E. coli levels are low. A few firms, shown at the right end of the graph, used apples with elevated APC and total coliform counts that can indicate poor quality.
Figure 14. Microbial Quality of Apples Prior to Chopping and Pressing.
Comparison of Apple Cider With and Without Preservatives Added --- Two samples were collected at 37 firms that added potassium sorbate and sodium benzoate preservatives to their apple cider. The samples were collected from the same lot before and after the preservatives were added in order to make comparisons. Figure 15 shows that the APC and total coliforms levels decreased after preservatives were added.
The fecal coliform and E. coli sample results are not shown because they were low. Six of the samples collected before the preservatives were added were positive for fecal coliforms and generic E. coli (range of 4 to 170 MPN/ml), and two of the samples were positive (3.6 MPN/ml) after the preservatives were added.
Figure 15. Comparison of the Microbial Quality of Cider Before and After Preservatives are Added.
Apple Cider Without Preservatives Added --- The results for the 169 finished product samples without preservatives were generally low as seen in Figure 16. Total and fecal coliforms were detected in 36% and 5% of the samples, respectively. In addition, there were 26 samples that were positive for E. coli. One sample had an EC median of 930 MPN/ml and an upper range of 4,600 MPN/ml. Overall, E. coli was detected in 7% of the finished product samples with and without preservatives.
Figure 16. Microbial Quality of Cider Without Preservatives.
The following summarizes the inspectional observations from the four firms with the highest generic E. coli median levels in the final product.
- Firm A with a median EC of 930 MPN/ml --- This small firm produces 2,000 gallons/year with apples from their own orchard. The orchard is spray-irrigated with surface water. Drops are not used, and the apples are refrigerated upon receipt at the processing facility. Before processing the apples are washed but not brushed nor sanitized. The apples are culled, but some rotten apples and leaves enter the chopper for pressing. Washed apples are placed in wire baskets and placed on the floor prior to chopping. Plastic pails used to collect the cider from the press and the jugs of cider when filled are also placed on the floor. The are no tables or other work surfaces. The employees who handle the bottom of the wire baskets, the plastic pails, and plastic jugs also handle apples without washing and sanitizing their gloves. Workers wear clean rubber boots, gloves and aprons in the processing room. The owner stated he is not concerned with cross contamination from the floor.
- Firm B with a median EC of 93 MPN/ml --- This small firm produces about 19,000 gallons/year with apples from their own and other local orchards. There is no irrigation and drops are not used. The apples are spray washed and brushed but not sanitized prior to chopping. Badly bruised apples are not discarded. Food grade rice hulls are added to the chopped apples to aid in pressing. The firm had good sanitation practices.
- Firm C with a median EC of 23 MPN/ml --- This small firm produces about 15,000 gallons/year with apples from their own orchards. The firm irrigates with tested well water. Drops are used. The apples are washed and brushed before chopping, but no sanitizer is used. Objectionable conditions included an employee eating while working at the sorting table, open doorways, and flying insects in the processing area and in an open surge tank of cider at the filler.
- Firm D with a median EC of 23 MPN/ml --- This small firm produces approximately 20,000 gallons/year with apples from its own orchards. The firm rarely irrigates and does not use drops. There are no restroom facilities in the fields. The apples are prewashed before storage and lightly spray washed before processing but not brushed or sanitized. Badly bruised apples are trimmed by hand and used. Insanitary conditions included cats in the production area, toilet facility was not clean and did not have hot water, open doorways, debris in production area, and press cloths dried outside hanging under a tree.
The purpose of the inspections was to ascertain the level of sanitation and the quality and safety of the product that existed on a particular day at each firm. The samples taken at different steps in the process were not intended to be used as the basis for a scientific study. As a result, the samples from a particular firm were not always collected from the same lot of apples at each step of the process due to the erratic nature of the production schedule at many firms. A more controlled and statistically based study is needed if acorrelation between microbial levels at different production steps is to be examined.
Given these limitations, an analysis was made with the available data of the correlation between the microbial levels in raw apples at the beginning of processing (Figure 12) and levels in the finished product (Figure 16). The analysis used the total coliform data because there were insufficient samples with elevated fecal coliform or generic E. coli levels. The analysis sought to determine whether:
- coliform levels in the finished product had decreased because sanitation controls were effective in reducing elevated microbial loads during processing; or
- coliform levels had increased because contamination was added during processing.
The analysis shows that, overall, cider processing neither decreased nor increased coliform levels between incoming ingredients and outgoing cider. This is shown graphically in Figure 17. Although there is large individual sample variation, the overall trend is that there are approximately as many samples showing a decrease in coliform levels in the end product as samples showing an increase. The exception is samples from the small number of firms that had poor quality wash water which resulted in poorer quality finished product. Thus, it is likely that any microbial hazards that are introduced at the beginning of processing will be carried through to the finished product.
These sample results are consistent with the inspection findings in two respects:
- it would not be expected that microbial contamination would be added during processing since the overall level of sanitation at the firms was good; and
- there were few sanitation controls or interventions applied during processing that would significantly reduce any elevated microbial loads on the incoming raw apples.
Figure 17. Log Plot of the Relationship Between Median Total Coliform Count in Raw Apples Before Pressing and the Corresponding Count in the Finished Product.