Evaluation of Packaged Ice Sold in Iowa

Executive Summary

The quality of packaged ice has been investigated by examination of
physical, chemical and microbiological characteristics of 22 ice samples
purchased at retail stores throughout Iowa. Only one sample exceeded a
primary health standard under the Safe Drinking Water Act (SDWA) and that
sample contained Klebsiella pneumoniae, a member of the total coliform
group of bacteria. Several samples of ice manufactured in convenience
stores had heterotrophic plate counts which exceeded the recommendations
(<500 CFU/mL) established by the Packaged Ice Association and all samples
failed to meet the package labeling recommendations of that organization.
While ice consumption does not represent an immediate threat to personal
or public health, the potential for disease transmission exists in an industry
which is voluntarily self-regulated.

Background

Ice has long been associated with food preservation by regulation of
temperature of foods to levels which restrict microbial growth. During the
past 50 years, ice has become an increasingly important aspect of the
recreational pursuits of Americans. The Packaged Ice Association estimates
that 86 million tons of ice are sold in the United States each year, or
approximately two pounds per person per day (6).
Ice is defined as a food in the U. S. Public Health Service Food Service
Sanitation Manual (PHS Publication 78-2081, 1976). As a manufactured
food, ice is covered by the Good Manufacturing Practices (GMP) Regulations
for foods contained in CFR 21, Chapter 1, Parts 20 and 110. These
regulations address the facilities where ice is manufactured, the quality of
source water and the sanitary practices of employees during ice production.
States are granted the option of regulating ice manufacture further within
their jurisdiction but most state laws only restate the federal GMP
regulations.

Despite the elaborate precautions taken to insure the quality and safety of
water and food, federal, state and local health officials have done little to
inspect or sample ice to insure its safety. In a recent survey sponsored by
the Association of Food and Drug Officials, 33% of responding states had no
information on the number of ice plants in their states, and 72% of states
did not know how many on-premises ice machines were packaging ice for
sale in retail stores (6). To understand the potential for exposure of the
population to disease from contaminated ice, consider that soft drinks served
in fast-food restaurants are 40% crushed ice, millions of drinks are sold daily
and most of the ice used is manufactured and distributed without regulatory
oversight. Clearly, ice is taken for granted by modern society.

The medical literature over the past ten years has documented ice-related
nosocomial transmission of Escherichia coli, Pseudomonas aeruginosa,
Legionella pneumophila, Salmonella enteritidis and Hepatitis A virus.
Pseudoepidemics of Mycobacterium gordonae have been attributed to
contaminated cold water lines feeding ice machines in hospitals.

The literature on illness associated with commercially manufactured ice is
sparce. In 1972, CDC investigated 98 cases of infectious hepatitis associated
with commercially produced ice pellets (3). The private well providing source
water for ice manufacture was contaminated by sewage from the tile
drainage field of an adjacent septic tank serving a house where several
people with known hepatitis had lived six weeks before the outbreak
occurred. In 1985, human illness was attributed to ingestion of carbonated
beverages served with contaminated ice (5). However, the 1987 outbreak of
viral gastroenteritis involving over 5000 people in Pennsylvania and
Delaware (4) was the crucial event which focused public health attention
upon the problems associated with ice manufacture, distribution and
consumption.

Investigations following complaints of illness associated with ice revealed
that sanitation during production and packaging of ice was a major problem.
Violations to existing federal regulations included improper facilities, poor
source water quality, microbial contamination, filth and foreign objects and
poor personal hygiene of employees. Investigators have documented the
presence of insect parts, fibers, glass, metal and plastic fragments and sand
in packaged ice (7,10).

By 1989, minimal voluntary standards in addition to GMP regulations for ice
manufacture were adopted by members of the Packaged Ice Association and
endorsed by the Association of Food and Drug Officials (1,8,9). Key features
of these standards require that:

• Ice manufacturing facilities must be maintained indoors in a room
separate from other non-manufacturing activity.
• Product ice must be regularly tested for total and fecal coliforms (MPN
<2.2/100 mL) and heterotrophic plate count (<500 CFU/mL).
• Packaging must be clearly labeled to show name of manufacturer,
location of the plant and production code to facilitate investigation of
complaints.

The Iowa Department of Agriculture and Land Stewardship published a
Notice of Intended Action in the Iowa Administrative Bulletin on February 21,
1990, proposing the creation of a new chapter in the Iowa Administrative
Code, pursuant to Iowa Code Chapter 159.5(11) and to implement Chapter
159.5(15). The general intent of the new rule was to impose drinking water
regulatory standards for physical, chemical and bacteriological quality upon
manufacturers of bottled and vended water and packaged ice in
conformance with Iowa Code Chapter 455B. It recommended that tests for
coliform bacteria be conducted every three months and that chemical and
physical testing occur annually to ensure that ice sold in Iowa and consumed
by the public be free of undesirable microorganisms.

A public hearing regarding the proposed rule was held on March 13, 1991. A
study bill (SSB 2) to regulate the commercial production, processing and
distribution of water products was introduced in the Iowa Senate on January
16, 1991, and assigned to the Senate Commerce Committee for
consideration. The legislature did not act on this proposed legislation prior to
its adjournment. The quality of packaged and vended ice consumed in Iowa
is still unregulated and thus is not periodically tested for purity.

The monitoring tests mandated in the proposed new rule noted above were
few in number and would have been conducted infrequently. The Packaged
Ice Association had previously established voluntary sanitary standards for
packaged ice, including annual physical and chemical testing, radiological
testing every four years, and monthly random sample testing for total and
fecal coliform bacteria and heterotrophic plate count in compliance with U. S.
Environmental Protection Agency drinking water regulations. Its members
are encouraged to implement process testing for ethylene glycol, propylene
glycol, lead, cadmium, zinc, chromium and nitrate, and to test quarterly by
random sample of finished product for glycols and chlorides. However, the
level of compliance with these voluntary sanitary standards is currently
unknown.

The quality of packaged ice sold in retail establishments in Iowa has not
been examined. We report results of a prospective study which examined
the physical, chemical and microbiological quality of packaged ice in Iowa.

Experimental Design

To insure a geographical representation of ice samples to be included in the
study, the telephone directories of several cities throughout Iowa were
reviewed to determine the number and location of commercial ice
manufacturing establishments and distributors. Fifteen manufacturers were
identified but it was determined that ice sold in many convenience stores
was produced and packaged on the premises. A list of manufacturers from
whom samples were obtained appears in Table 1. Sampling was performed
by Hygienic Laboratory staff during April 1991. All samples were purchased
at retail establishments as staff traveled during the course of their normal
duties. Ice was marketed in 5, 6, 7, 8, 8.5, and 16 lb. bags. Because 7-8 lbs.
of ice was required to produce the required total melted sample volume of 3
liters, two each of the 5, 6, and 7 lb. bags were purchased and combined for
melting and subsequent analytical procedures. All samples were placed in
insulated coolers and immediately transported to the laboratory where they
were stored at -20 C awaiting sample processing. The period of time
between purchase and sample processing never exceeded 72 hrs. and no
noticeable thawing occurred during transit or storage. To prepare samples
for analytical procedures, bags of ice were removed from the freezer, a
lower corner of the bag was wiped with 70% ethanol and a sterile scalpel
was used to aseptically cut a 3 inch opening which allowed ice to fall directly
into a previously sterilized 10 liter Nalgene carboy. A screw-cap was placed
on the carboy and samples were left at room temperature overnight for
melting. The resulting water remained cold after 12 hrs. at room
temperature. Water was mixed to insure sample homogeneity and poured
into specially prepared sample containers for microbiological, particulate,
inorganic and organic analyses. Samples were transported to the respective
analytical laboratories under standard conditions and analyzed within the
holding times specified by the analytical method. Samples for microbiological
analysis were processed first after melting and were tested within 2 hours.
Sample bags were saved and all information printed on the bags was
transcribed onto project worksheets.

Analytical procedures conformed to quality guidelines established by the
Association of Food and Drug Officials, the Food and Drug Administration,
the U.S. Environmental Protection Agency, Standard Methods for
Examination of Water and Wastewater and sound laboratory practice. The
list of analytes selected and methods used in this study is presented in Table
2. Primary and secondary federal standards established for drinking water
are listed with study results, however these standards are not legally binding
for ice producers. The quantitative detection limit of each method is
indicated by use of a less than (<) indication before the numerical result.
The heterotrophic plate count (HPC) was performed by the spread plate
method using 1 mL and 0.1 mL volumes of undiluted sample on pre-dried
were determined by the spread plate method using 1 mL of undiluted
sample on Sabouraud Dextrose agar containing choloramphenicol to inhibit
bacterial overgrowth. These plates were incubated for 5 days at 25 C.
Particulate analysis was performed by passing 300 mL of sample through a
0.4 m, 47 mm diameter polycarbonate membrane filter (Nucleopore). Direct
microscopic examination of the filters was performed at total magnifications
of 27x, 100x and 900x. Fibers and inanimate particulates were identified by
the UHL Senior Microscopist, Air Quality Section, using micro-chemical,
physical and morphological properties. Insect parts were confirmed by the
UHL parasitologist. Bacteria and mold identifications were performed in the
UHL Reference Bacteriology and Mycology Laboratories, respectively. Algae,
iron bacteria and hyphal elements of fungi were identified by the UHL
Environmental Microbiology Laboratory staff.

A telephone survey of ice producers was undertaken to determine the extent
of compliance with the voluntary guidelines prepared by the Packaged Ice
Association. Questions covered the source and quality of water used for ice
production and manufacturing and labeling practices.

Quality Assurance

The Hygienic Laboratory is certified by EPA for analytical testing of drinking
water under the Safe Drinking Water Act (SDWA). Samples were transported
to the laboratory in insulated coolers together with a field blank to control
for introduction of organic contaminants during transit or processing. New
Nalgene carboys were purchased for this study. These containers were filled
with water, autoclaved and emptied. This cycle was repeated 2-3 times and
water from the last cycle was examined for residues which could result in
false positive analytical tests.

Results and Discussion

A summary of information obtained from the telephone survey and directly
from ice bags is presented in Table 3. Five of 21 manufacturers were
members of the Packaged Ice Association and the membership status of one
manufacturer could not be determined. Private wells were used as source
water for ice production by four manufacturers and the remaining 18
producers used source water from a public water supply. Water treatment
methods varied from no additional treatment of municipal source water to
various combinations of softening, demineralization, filtration, chlorination
and in one case, treatment by reverse osmosis. Managers of convenience
stores were generally uncooperative and no information about water
treatment or production methods could be ascertained. Eleven of 22
manufacturers performed some analytical tests on their product to monitor
microbiological safety. The most common test employed was analysis for
total coliform bacteria but the frequency ranged from monthly to once every
two years. The quantity of ice per bag varied from 5 lbs to 16 lbs. Bags were
all made of plastic but there was considerable variation in the thickness and
strength of the bags. All bags had at least one hole and thinner bags had
numerous stretches and tears which could allow contamination of ice from
environmental surfaces. Bag closures included metal staples (13), twist-ties
(2) hard plastic clip (1), plastic tape (1), draw-string (1) and a heat sealed
bag (1). All product purchased was in the form of cubes except one which
was manufactured as “tubes”. Bags contained a variable amount of printing
on the bags from less than 10% of the total surface area to 100% of the
available area. There was no correlation between the amount of printing on
the bags and lead concentrations of samples. Labeling of bags was checked
against the voluntary guidelines of the Packaged Ice Association (name of
manufacturer, location of plant and production code). Two manufacturers
advertised membership in this organization on their labels. All bags from
large commercial suppliers contained the name and location of manufacture
except one. Those products produced on the premises of convenience stores
contained only the corporate address, not the location of manufacture. None
of the bags were identified with a production code which could facilitate
epidemiological follow-up during investigations of a disease outbreak.
Several manufacturers claimed their product was produced under sanitary
conditions, cited a municipal water source for their product or made other
claims about the superiority of their product (Table 3). Ironically, one
product claiming to be “untouched by human hands” contained
Acinetobacter lwoffi, an organism found on skin and in soil. Most general
claims are impossible to validate in a quantitative fashion and some claims
may violate truth in advertising laws, eg. “Lasts Longer”, “Holds the Cold”,
etc. Value conscience consumers should be aware of the variation in net
weight of ice bags while comparison shopping.

The microbiological quality of ice was evaluated by performing heterotrophic
plate counts (HPC), mold counts, and tests for total and fecal coliform
bacteria and Pseudomonas aeruginosa (Table 4). Total HPC counts ranged
from <1 to >54,000 colony forming units (CFU)/mL. Ice produced on the
premises by convenience stores demonstrated consistently higher bacterial
counts than ice produced by commercial ice companies and all ice produced
on the premises of convenience stores exceeded the upper limit of 500
CFU/mL recommended by the Packaged Ice Association. While no standards
have been developed for mold contamination of ice, mold counts paralleled
bacterial counts and served as an indication of unsanitary conditions at the
production site. One sample was positive for total coliform bacteria (MPN 2.2
organisms/100 mL) and all tests for fecal coliforms and Pseudomaonas
aeruginosa were negative. Klebsiella pneumoniae was identified as the
predominant organism in the total coliform positive sample. This organism
may be found in the intestinal tract of humans and animals and in soil,
vegetative matter and water. K. pneumoniae causes various infections in
humans and is considered a primary opportunistic pathogen.
The identifications of bacteria and fungi isolated from ice during this study is
shown in Table 5. Many organisms isolated from environmental sources grow
poorly on culture media used in identification of medically significant
organisms and are thus difficult to identify by the methods used at UHL. The
common sources and medical significance of these organisms is shown in
Table 6.

The results of particulate analysis is presented in Table 7. Algae, mold
spores and insect parts were common in ice produced by convenience
stores. Hair and glass were found in a preliminary ice sample purchased
from a convenience store as an internal control for the filtration procedure.
No other analytical tests were performed on this sample, therefore these
results were omitted from Table 7. The presence and nature of particulate
materials in ice reflects the conditions under which it was produced,
packaged and stored. Dust mites were the most common insect parts found.
The inorganic chemical analytes for this study were selected to provide an
indication of source water quality and to detect potential contamination
during secondary water treatment or ice manufacturing. No primary
inorganic SDWA standards were exceeded (Table 8). However, the
secondary standard for pH (acceptable range 6.5-8.5) was exceeded by
samples 28311, 29072, 29504 and 30244. Sample 28311 contained a level
of potassium higher than would be expected in a public drinking water
supply. This sample also contained the highest level of sulfate (30 mg/L) and
zinc (0.22 mg/L) encountered in the study. Detectable zinc concentrations
may result from use of galvanized materials (auger, bin, etc.) in production
or packaging of ice. Sample 29072 had an unexplained pH of 9.8 and
reportable levels of copper and lead. The lead level of 0.028 mg/L exceeds
the proposed screening level but does not violate current regulations.
Sample 30243 contained a chloride level of 22 mg/L yet remained well below
a level which would be detected by taste. Softening may contribute to the
detectable chloride concentrations of some samples. No MCL has been
established for chloride. This sample had an iron level of 0.21 mg/L which
may result from processing equipment.

The results of organic chemical analyses are presented in Table 9. The total
trihalomethane (THM) concentrations result from the action of chlorine
residuals on water containing organic material. Of the halogenated
hydrocarbons included in the screening test (chloroform, bromoform,
bromodichloromethane and chlorodibromomethane), only chloroform was
detected. All THM levels were below the MCL of 100 g/L as a yearly average
for drinking water. The absence of THM in most samples may indicate the
loss of volatile organic compounds during freezing of ice. 1,1,1-
Trichloroethane was observed in one sample, however the level detected is
well below the MCL for drinking water (200 g/L). This organic compound is
associated with cleaners and degreasers and may represent residual
contamination from mechanical equipment or storage of cleaning agents in
the room where ice is produced. Ethylene and propylene glycols were
included as a screen for refrigerants used in ice production. The analytical
method used was direct aqueous injection into the column of a gas
chromatograph equipped with a flame ionization detector. The detection limit
reported represents the concentration of the low standard used which varies
slightly between runs. No standards for glycols have been established for
drinking water and all ice samples were below the detection limit of the
method.

Conclusions

The quality of packaged ice sold in Iowa reflects the quality of source water
and the sanitary conditions during manufacture. Ice produced in
convenience stores was of consistently poorer microbiological quality than
ice produced by major commercial producers. No manufacturer fully
complied with the recommendations of the Packaged Ice Association
regarding use of a product code which would facilitate investigation of
disease outbreaks. The overall quality of ice is much like the quality of
bottled water (2). However, unlike bottled water, ice is not ingested in
quantities which represent a significant threat to personal or public health.
Nevertheless, the potential for disease transmission exists in an industry
which is voluntarily self-regulated.

Acknowledgements

The authors thank Robert Jensen, John Scott, Mark Hurt, Wesley Byrd,
Theresa Wakeford, and Ramona Schultz for collection of samples at sites
throughout the state. Fungal identifications were performed by Dr. Dennis
Gaunt and bacterial identifications were performed by Larry Holcomb. Mary
Richey assisted with fiber and particle identification and Mike Last confirmed
the presence of insect parts. The number of personnel involved in organic
and inorganic chemical analyses are too great to name but their participation
was invaluable for the timely completion of the study. Carilyn Wieland and
staff provided clerical support during preparation of the manuscript.

References
1. Association of Food and Drug Officials. 1989. Guidelines for the Inspection
and Enforcement of GMP Regulations for Handling and Manufacturing
Packaged Ice. JAFDO 53: 70-74.
2. Breuer, G. M., L. A. Friell, N. P. Moyer, G. W. Ronald and W. J. Hausler,
Jr. 1990. Testing of Bottled Waters Sold in Iowa. UHL Technical Report
91-1.
3. Centers for Disease Control. 1972. Infectious Hepatitis- Wickes, Arkansas.
EPI 71-130-2.
4. Centers for Disease Control. 1987. Outbreak of Viral Gastroenteritis -
Pennsylvania and Delaware. MMWR 36:709.
5. Dickens, D. L., H. L. Dupont and P. C. Johnson. 1985. Survival of
Bacterial Enteropathogens in the Ice of Popular Drinks.
JAMA 253:3141-3143.
6. Felix, C. W. 1989. Ice-the Forgotten Food. JAFDO 53:19-24.
Ladanzi, P. A. and S. M. Morrison. 1962. A Sanitary Survey of Ice. J. Milk
and Food Technol. 30: 253-258.
7. Packaged Ice Association. 1989. Sanitary Standards for Packaged Ice.
Packaged Ice Association. A Guide to the Sanitation of Packaged Ice
(Undated Brochure). Raleigh, North Carolina.
8. Young, T. R. and G. J. Barela. 1975. Sediment Testing of Ice Proves of
Value in Denver. J. Environ. Health 38:92-93.