A protocol for quantitative measurement of light intensity levels in food service operations.

Introduction

Although the United States has probably some of the strictest food safety regulations in the world, foodborne-disease outbreaks occur every year. It is estimated that foodborne outbreaks cause 76 million illnesses, 325,000 hospitalizations, and 5,000 deaths each year in the United States (Mead et al., 1999). Factors that contribute to foodborne outbreaks in the United States can be broken down into three categories: 1) time and temperature, 2) poor hygiene practices by food handlers, and 3) cross-contamination (Moore, 2002). Improper holding temperatures, poor personal hygiene, contaminated equipment, inadequate cooking, and food from unsafe sources accounted for 37 percent, 19.3 percent, 15.8 percent, 10.8 percent, and six percent, respectively, of the outbreaks reported to the Centers for Disease Control and Prevention (CDC) from 1993 to 1997. Other factors accounted for 11.1 percent of the outbreaks (Olsen, Mackinon, Goulding, Bean, & Slutsker, 2000).

Although inadequate lighting was not one of the contributing factors for the outbreaks reported to CDC (Olsen, MacKinon, Goulding, Bean, & Slutsker, 2000), it is an important food safety element and might indirectly contribute to foodborne outbreaks. The authors believe that adequate lighting is essential to ensure that all food contact surfaces and utensils are properly cleaned and sanitized, and that they are free of food particles. Adequate lighting also is essential for monitoring employee hygiene and food-handling practices, for proper inspection of equipment and premises of food facilities, for proper storage of foods in refrigeration units, for detecting cross-contamination, and in general for food safety, good house-keeping, and the productivity and safety of food handlers.

To minimize food safety risk to consumers, health officials at federal, state, and local levels enforce food safety standards through laws, regulations, inspections, and compliance actions (Taylor & Hoffman, 2001). Inspections have not, however, totally prevented the occurrence of outbreaks associated with food service operations. Of the outbreaks reported to CDC from 1993 to 1997, the majority (48 percent) were caused by foods eaten in food service facilities such as delicatessens, cafeterias, and restaurants. Foods consumed at private residences, school cafeterias, picnic sites and campsites, and other places accounted for 22 percent, 3 percent, 2 percent, and 25 percent of the outbreaks, respectively (Olsen, MacKinon, Goulding, Bean, & Slutsker, 2000).

Current inspection practices have many limitations and may not always predict the likelihood of outbreaks. Some of these limitations include a lack of periodic inspections of food service facilities as a result of manpower shortages and the fact that inspections are largely a visual process and do not always accurately determine compliance with code requirements (Taylor & Hoffman, 2001). In addition, food service facilities are inspected only a few times a year and seldom during weekends and evenings, and hazardous situations may not occur at the time of inspections, or may be missed during inspections (Bryan, 2002).

According to a recent study, fecal bacteria were found in different sites in 50 percent of restaurants that passed visual inspection for cleanliness (Kassa, Harrington, Bisesi, & Khuder, 2001). A previous study also concluded that visual inspections were not correlated with results of microbiological examinations of restaurant equipment, utensils, and employee hands (Tebbutt, 1991; Tebbutt & Southwell, 1989). In addition, results of visual inspections and interpretations of code requirements may differ with different sanitarians or in different jurisdictions (Bryan, 2002).

To date only one published study has showed that routine inspections of food service facilities could successfully identify restaurants at increased risk for foodborne outbreaks (Irwin, Ballard, Grendon, & Kobayashi, 1989). Two other studies have reported foodborne-disease outbreaks in frequently penalized food service operations (Kassa, 2001; Luby, Jones, & Horan, 1993). A study conducted in Miami-Dade County, Florida, in 1995, however, found that inspections of restaurants conducted in the county did not reliably identify restaurants that were at increased risk for foodborne-disease outbreaks (Cruz, Katz, & Suarez, 2001).

[FIGURE 1 OMITTED]

Quantitative rather than qualitative or visual-inspection methods not only are important for obtaining more accurate results, but also can minimize the variation in inspection outcomes. Periodic microbiological evaluations of high-risk food service operations, in addition to visual inspections, have been recommended for minimizing risks of foodborne outbreaks (Kassa et al., 2001). A preliminary survey of some of the local health departments in northwest Ohio showed that most use qualitative or visual inspections rather than quantitative methods to ensure compliance with most food code requirements. Lighting is one area in which quantitative methods can be applied, and inadequate lighting could be an indirect cause of foodborne outbreaks. The goal of this study is to develop a protocol for quantitative measurement of lighting levels in food service operations.

In Ohio, the minimum enforceable light intensity level is 10 foot-candles (ft-c) measured 30 inches above floor level for walk-in refrigeration units and at least 50 ft-c at food preparation surfaces (Ohio Department of Health, 2001). The same light intensity levels have been specified in the federal Food Code (Food and Drug Administration 2001). These codes not only do not indicate how these light intensity standards were determined, but also do not specify the methods by which light levels should be assessed.

Visual appraisal of light intensity levels lacks both precision and accuracy, because the results may vary from individual to individual and are not always reproducible even when the same individual is making the measurement a second time (Rea, ed., 1993). In addition, individuals can make only qualitative assessments about whether lighting intensity levels are sufficient for performing particular tasks. Such qualitative assessments may not always be accurate. As a result, in the lighting industry qualitative visual-appraisal methods are rarely used now; instead, measurements are made with physical instruments (Rea, ed., 1993).

The three primary objectives of the study reported here were 1) to appraise light levels in retail food service facilities with a light meter and determine whether the facilities complied with regulations, 2) to compare the inspection reports of food inspectors with results of the quantitative measurements, and 3) to examine the utility of current codes and make recommendations for the future.

Materials and Methods

The study was conducted during June, July, and August of 2002. The participating health department has 16 health districts. Eleven of the 16 health districts (68.8 percent) were included in this study. The lighting evaluations were made during standard unannounced inspections of food service operations conducted by the district sanitarians. Light intensities inside of walk-in coolers and in food preparation areas were measured in 57 food service operations with a factory-calibrated TESTO 545 Luminous Light Intensity Meter.

Walk-in Coolers

To measure lighting intensity levels inside walk-in coolers, the authors first drew to scale floor areas and the positions of shelving racks, and noted the locations and types of light fixtures (incandescent or fluorescent) (Figure 1). The accessible floor area inside the walk-in cooler was divided into rectangular areas of approximately 4 square feet. Light intensity levels were measured 30 inches above the center of each floor rectangle. The average for the floor was then calculated and recorded. In addition, light intensity levels were measured at the center and corners of each shelving unit. The measurements were taken within the interior of each shelving unit at heights of 24 (lower shelf area) and 60 inches (upper shelf area) above the floor area. The averages for the lower and upper shelving areas were then calculated and recorded. The majority of the 57 food service operations had only one walk-in cooler, but where more than one walk-in cooler existed, only one was evaluated (chosen on the basis of usage level).

Food Preparation Counters

Light intensity levels were measured at 185 food preparation counters in the 57 food service operations. Each food preparation counter was divided into rectangular areas of approximately 4 square feet. Light intensity levels were measured at approximately 4 inches above the center of each rectangular area. The average light level for each food preparation counter was then calculated and recorded. The number of food preparation counters evaluated per restaurant varied from a minimum of one (100 percent of the restaurants) to a maximum of five (7 percent of the restaurants). About 63 percent and 89 percent of the restaurants had three and two food preparation counters, respectively.

Statistical Analysis

The distribution light levels of the walk-in coolers and food preparation counters were examined visually with relative-frequency histograms and cumulative-probability distribution charts. The raw data were positively skewed on the histogram, and the plot of the data points against the cumulative-probability scale was curvilinear. A histogram of the log-transformed data appears normally distributed, and a log-probability cumulative-distribution plot approximates a straight line ([r.sup.2] > .9). Following these visual tests, the data were confirmed as log normally distributed by a W-test (p < .05). Because the data were log normally distributed, geometric mean (GM) and geometric standard deviation (GSD)--rather than arithmetic mean and standard deviation--were used to describe central tendencies and variability. Inferential statistics, such as t-tests, were performed on the log-transformed values.

Results

Walk-in Coolers

Of the 57 walk-in coolers, 12 had fluorescent lightbulbs, 43 had incandescent lightbulbs, and two had lightbulbs of both types (and were excluded from the comparative portion of the study). Of the 12 walk-in coolers with fluorescent lightbulbs, seven had more than one light fixture, and five had only one fixture. Similarly, of the 43 coolers with incandescent lightbulbs, nine had more than one light fixture, and 34 had only one light fixture. Geometric-mean light levels did not differ significantly between coolers fitted with one fixture and those fitted with two or more fixtures (p > .05). The lack of difference is likely due to the fact that coolers with two or more fixtures were generally bigger than those with one light fixture so the additional fixtures were needed.

The descriptive statistics for the cooler light levels are shown in Table 1. Typical light levels (GM) at all locations in coolers with fluorescent lighting were significantly higher than the corresponding light levels in coolers with incandescent light fixtures (p < .05). The geometric mean light intensity levels were significantly higher at center floors than at lower shelving in coolers with light fixtures of either type (p < .05).

Table 2 shows the percentage of coolers that met the standard of 10 ft-c and half the standard (5 ft-c). Although the height of the light level assessment is specified by the regulations (30 inches), the location in the coolers where the light levels should be evaluated is not specified. If the center of the cooler is considered, 75 percent of the coolers with fluorescent lighting met the standards, compared with only 7 percent of the coolers with incandescent lighting. All of the coolers with fluorescent light had at least 5 ft-c (50 percent of the standard) at the center location, while less than one-third of those with incandescent lighting met half the standard. Coolers with fluorescent lighting were significantly more likely to meet the standard than were incandescently lit units (Chi-square test, p < .001). For both types of fixtures, light levels drop off greatly at shelving locations (Figure 2). Light levels on shelving are important because these locations are where food service workers need to be able to see to determine if proper food service practices are observed. More will be said on this subject in the discussion.

[FIGURE 2 OMITTED]

Food Preparation Counters

The GM light intensity level for all 185 food preparation counters in the 57 restaurants was 38.7 ft-c (GSD: 1.92) (Table 3). Of these restaurants, 13 (22.8 percent) met the standard (50 ft-c) at all the food preparation surfaces evaluated. At 27 of the food service operations (47.4 percent), none of the food preparation counters evaluated met the standard. The GM light intensity levels for restaurants with 100 percent and 0 percent pass rates were 71.6 ft-c (GSD: 1.37) and 23.4 ft-c (GSD: 1.68), respectively (Table 3). The remaining 17 restaurants (29.8 percent) had light intensity levels exceeding 50 ft-c at some food preparation counters and less than 50 ft-c at others. Out of the 185 food service preparation counters sampled in all 57 restaurants, 76 (41 percent) had mean light intensity levels greater than 50 ft-c.

Visual Inspections

Of the 57 restaurants inspected by the district health inspectors, 19 (33.3 percent) received citations for violations in walk-in coolers or at food preparation counters. The total number of violations in these restaurants was 24, of which 21 were for violations in walk-in coolers and three for violations at food preparation counters. Of the 24 violations, only five pertained to lack of adequate lighting--four in walk-in coolers, and one at a food preparation counter. Other violations were improper thawing of potentially hazardous food, contamination and cross-contaminations of food, lack of cleanliness, and lightbulbs not covered with shatterproof shields.

Discussion

Effectiveness of Qualitative Inspections

Qualitative visual inspections failed to identify the majority of the walk-in coolers and food preparation surfaces that did not meet the light intensity standards. While 45 out of the 57 restaurants (79 percent) failed to meet the standard when light intensity level was measured at the center of the coolers, only three (5 percent) of the restaurants received citations for insufficient light levels inside coolers as a result of qualitative assessment by visual inspection. Similarly, visual inspections failed to identify food service operations that did not meet light intensity standards at food preparation counters. Of the 27 food service operations that had no food preparation counters meeting the standard, only one was cited for insufficient lighting as a result of qualitative assessment by visual inspection. Although violations of lighting standards are not by themselves predictors of the risk of food-borne-disease outbreaks, systematic quantitative measurement of lighting levels might be necessary in food service operations with poor housekeeping practices, in establishments with a history of repeated violations in critical aspects of food safety for which adequate lighting levels are essential, and in restaurants with large volumes of foods stored in refrigeration units that have inadequate space.

Fluorescent Versus Incandescent Lighting

Among walk-in coolers fitted with fluorescent-light fixtures, 75 percent exceeded the standard at center floor, compared with a 7 percent pass rate for incandescent lighting. Clearly, fluorescent lighting improved the compliance rate. Away from the center of the cooler, only 25 percent and 8 percent of coolers equipped with fluorescent lighting exceeded the standard at upper shelving and lower shelving respectively. The pass rate is even lower, however, for coolers fitted with incandescent-light fixtures. Chi-square tests revealed that walk-in coolers with fluorescent-light fixtures are significantly more likely to meet or exceed the standard than coolers with incandescent-light fixtures (p < .001). Food service operators should preferentially select walk-in coolers equipped with incandescent lighting. Walk-in cooler manufacturers should consider making fluorescent lighting the standard for all units and providing a retrofit of existing incandescent units to convert to fluorescent lighting.

Gaps in the Regulations

Neither the Ohio nor the FDA food code specified the locations where light levels should be assessed in walk-in coolers. The location is important because light intensity levels may not be uniform throughout walk-in coolers. In this study the authors found that light intensity levels were higher in the center of coolers than in areas where food items were stored. In the majority of the restaurants sampled, food items requiring refrigeration were stored on shelves, along the sides of walk-in coolers/freezers, and--rarely--in center floor. Light intensity levels, however, were much lower at the upper or lower shelves and were higher at center floor. Overall, 12 of the 57 food service operations (21 percent) passed the standard at center floor, but when light was measured at upper and lower shelving, only four (7 percent) and one (1.8 percent), respectively, passed the standard. The shelves are precisely where illumination is needed for identification of improper storage practices.

The Illuminating Engineering Society of North America (IESNA) has established procedures for selecting illumination levels for specific activities based upon factors important for visual performance (Rea ed., 1993). These factors are 1) type of activities; 2) characteristics of task in terms of contrast, size, and time required to complete task; 3) age of the person performing the task; 4) importance of speed and accuracy in the task; and 5) reflectance of the task surface. The illumination regulations of the food codes address little of this. Nor do the codes specify how light intensity levels should be measured. Local health departments may therefore rely on cursory visual or qualitative assessments rather than on instruments or quantitative methods for assessing light intensity levels in food service operations.

It is also not clear that the recommended light intensity levels are adequate for the tasks described in the food codes. IESNA recommends a range of light intensity levels for various types of activities. For example, the recommended light intensity levels for the performance of occasional visual tasks range from 10 to 20 ft-c. The current code standard for walk-in coolers is 10 ft-c. The location of the lighting level, however, needs to be at the location of the visual task, which is at the shelving location. Current codes, while providing a quantitative standard for lighting, are simplistic in their approach to adequate illumination. Furthermore, these codes lack specificity in describing how compliance should be assessed.

While area-based light intensity guidelines are easy to enforce, additional guidelines based on the nature of the visual task, contrast, duration, speed of the task, and age of workers are much preferable in assuring food safety and good housekeeping, as well as for increasing the productivity and safety of food-handlers. This study is only a preliminary one, and additional research is required to assess light intensity levels in food service operations on the basis of factors recommended by IESNA.

Conclusion and Recommendation

The state and FDA food regulations specify quantitative illumination standards in walk-in coolers and at food preparation counters in food service operations. Many local health departments rely on visual assessment of light intensity levels to enforce food code standards. Such visual inspections are qualitative in nature and are not always accurate. Quantitative assessment with light meters provides more accurate results and is reproducible. The authors recommend that all health departments use standard light meters to determine light intensity levels in food service operations. Since walk-in coolers fitted with fluorescent-light fixtures have a higher passing rate than walk-in coolers fitted with incandescent-light fixtures, the authors also recommend that manufacturers consider using fluorescent-light fixtures when manufacturing walk-in coolers intended for food storage purposes.

The authors found that inspectors pay little attention to lighting standards during inspections. Although violations of lighting standards are not critical risks for food-borne disease, if inadequate lighting levels are suspected, the authors recommend that inspectors conduct systematic quantitative measurements of lighting levels as necessary and that they not simply rely on visual assessment. Making this kind of measurement becomes important if inadequate lighting levels are suspected as causes for critical food safety violations, such as poor housekeeping, improper storage of foods in walk-in coolers, and contaminated equipment. Continual, though informal, inspections of food service operations are needed to ensure that hygienic food storage and preparation practices are maintained and, whenever possible, inspection should include quantitative measurements of lighting levels to ensure compliance with the standards.

Current Ohio and FDA food code standards are very simplistic and lack adequate guidelines for enforcement. Since light levels are normally lower at the sides and corners than at the center floor of walk-in coolers, the codes should specify the location and the number of light measurements per location. Furthermore, the lighting standards should be based on current scientific knowledge and guidelines established by the Illuminating Engineering Society of North America.

TABLE 1

Geometric Mean (GM) and Geometric Standard Deviation (GSD) Light
Intensity Levels at Different Locations in Walk-in Coolers with
Fluorescent and Incandescent Light Fixtures

                  Fluorescent (n = 12)

       Center Floor    Upper Shelf   Lower Shelf

GM      15.27 ft-c      7.11 ft-c      3.97 ft-c

GSD      2.08 ft-c      2.34 ft-c      2.32 ft-c

Range  6.4-85.5 ft-c  2.1-45.7 ft-c  1.2-29.9 ft-c

                  Incandescent (n = 43)

       Center Floor    Upper Shelf    Lower Shelf

GM       3.43 ft-c      1.53 ft-c      0.544 ft-c

GSD      2.01 ft-c      2.88 ft-c       3.00 ft-c

Range  1.0-16.7 ft-c  0.22-10.1 ft-c  0.03-4.9 ft-c

Note: At each location, the mean light levels were significantly higher
(p < .05) for walk-in coolers with fluorescent light fixtures than for
walk-in coolers with incandescent light fixtures. The minimum
enforceable standard for walk-in coolers is 10 foot-candles.

TABLE 2

Percentage of Walk-in Coolers with Light Levels That Met or Exceeded
100% or 50% of the Standard, by Location Within Cooler

                                                Fluorescent
                                                  (n = 2)

                                                Center Floor

[greater than or equal to]10 ft-c                    75%

<10 ft-c, but [greater than or equal to]5 ft-c      100%

                                                  Fluorescent (n = 2)

                                                Upper Shelf  Lower Shelf

[greater than or equal to]10 ft-c                   25%           8%

<10 ft-c, but [greater than or equal to]5 ft-c      58%          33%

                                                Incandescent
                                                  (n = 43)

                                                Center Floor

[greater than or equal to]10 ft-c                    7%

<10 ft-c, but [greater than or equal to]5 ft-c      27%

                                                 Incandescent (n = 43)

                                                Upper Shelf  Lower Shelf

[greater than or equal to]10 ft-c                    2%          0%

<10 ft-c, but [greater than or equal to]5 ft-c      19%          0%

Note: The standard for walk-in coolers is 10 ft-c.

TABLE 3

Geometric Mean (GM) and Geometric Standard Deviation (GSD) of Light
Intensity Levels at Food Preparation Surfaces

                      All Restaurants (n = 57)

Number of food                  185
preparation surfaces

GM                           38.7 ft-c

GSD                          1.92 ft-c

Range                      2.9-196.8 ft-c

                      Restaurants with 0%  Restaurants with 100%
                      pass rate (n = 27)    pass rate (n = 13)

Number of food                77                    37
preparation surfaces

GM                         23.4 ft-c             71.6 ft-c

GSD                        1.68 ft-c             1.37 ft-c

Range                    2.9-46.0 ft-c        50.5-196.8 ft-c

Note: The minimum enforceable standard at food preparation surfaces is
50 ft-c; only 41.1% of the food preparation surfaces sampled at 57
restaurants met or exceeded this standard.

Acknowledgments: The authors thank the Faculty Research Committee of Bowling Green State University for funding the study. They also would like to thank the Toledo-Lucas County Health Department for granting them permission to conduct the study.

REFERENCES

Bryan, F.L. (2002). Where we are in retail food safety, how we got to where we are, and how did we get there? Journal of Environmental Health, 65(2), 29-36.

Cruz, M.A., Katz, D.J., & Suarez, J.A. (2001). An assessment of the ability of routine restaurant inspections to predict food-borne outbreaks in Miami-Dade County, Florida. American Journal of Public Health, 91(5), 821-823.

Irwin, K., Ballard, J., Grendon J., & Kobayashi, J. (1989). Results of restaurant inspections can predict outbreaks of foodborne illness: The Seattle-King County Experience. American Journal of Public Health, 79(5), 586-590.

Kassa, H. (2001). An outbreak of Norwalk-like viral gastroenteritis in a frequently penalized food service operation: A case for mandatory training of food handlers in safety and food hygiene. Journal of Environmental Health, 64(5), 9-12.

Kassa, H., Harrington B., Bisesi, M. & Khuder S. (2001). Comparisons of microbiological evaluations of selected kitchen areas with visual inspections for preventing potential risk of foodborne outbreaks in food service operations. Journal of Food Protection, 64(4), 509-513.

Luby, S.P., Jones, J.L., & Horan, J.M. (1993). A large salmonellosis outbreak associated with a frequently penalized restaurant. Epidemiology and Infection, 110(1), 31-39.

Mead, P.S., Slutsker, L., Dietz, V., McCaig, L.F., Brese, J.S., Shapiro, C., Griffin, P.M., & Tauxe, R.V. (1999). Food-related illness and death in the United States. Emerging Infectious Diseases, 5(5), 607-625.

Moore, G. S. (2002). Living with the Earth: Concepts in environmental health (2nd ed.). Boca Raton, FL: Lewis Publishers.

Ohio Department of Health. (2001). Ohio uniform food safety code. Columbus, Ohio. http://www.odh.state.oh.us/Rules/Final/Chap3717_1/Fr3717_1_1st.htm (10 Aug. 2002).

Olsen, S.J., MacKinon, L.C., Goulding, J.S., Bean, N.H., & Slutsker, L. (2000). Surveillance for foodborne disease outbreaks--United States, 1993-1997. Morbidity and Mortality Weekly Report, 49(SS01), 1-51.

Rea, M.S. (Ed.). (1993). Lighting handbook (8th ed.). New York: Illuminating Engineering Society of North America.

Taylor, M.R., & Hoffman, S.A. (2001). Redesigning food and technology. Issues in Science and Technology Online. http://www.nap.edu/issues/17.4/p_taylor.htm (15 Nov. 2002).

Tebbutt, G.M. (1991). Development of standardized inspection in restaurants using assessments and microbiological sampling to quantify the risks. Epidemiology and Infection, 107, 393-394.

Tebbutt, G.M., & Southwell, J.M. (1989). Comparative study of visual inspections and microbiological sampling in premises manufacturing and selling high-risk foods. Epidemiology and Infection, 103, 475-486.

Food and Drug Administration (2001). Food code, Washington, DC. http://www.cfsan.fda.gov/~dms/fc01-toc.html (10 Aug. 2002).

Hailu Kassa, Ph.D., M.P.H., M.S.O.H., R.S.

Charles Keil, Ph.D., C.I.H.

Kenneth W. Fent

Corresponding Author: Hailu Kassa, Assistant Professor, Department of Public & Allied Health, 223 Health Center, Bowling Green State University, Bowling Green, Ohio 43403. E-mail: [e-mail].