Discussion: (0 comments)
There are no comments available.
Data Disclosure, Performance, and Recommendations for Reform
View related content: Health Care
No. 2, June 2011
As the Food Safety Modernization Act, signed by President Barack Obama on January 4, 2011, is implemented, it is appropriate to take stock of food-borne illness (FI) in the United States. This Outlook conducts a careful review of federal FI statistics and finds that reporting and data disclosure are out of date and woefully incomplete. As important, it finds that the available evidence indicates no increase in either the number of outbreaks or their severity. Further, a key measure of the government response to such outbreaks has been deteriorating. In particular, there has been an increase in the percentage of outbreaks for which the Centers for Disease Control and Prevention (CDC) or other food-safety agencies are unable to report which food caused people to fall ill. Responding effectively to FI outbreaks depends on knowing what food item caused the outbreak; this information underpins both enforcement action and targeted public health measures like recalls and “do not eat” warnings. To control this trend toward deteriorating performance, the CDC should fully and promptly disclose all food-safety data (subject to the usual protection of patient privacy). The CDC should also explicitly adopt quantitative performance goals for responding to FI outbreaks.
Key points in this Outlook:
FI outbreaks–multiple illnesses linked to the same food–are the most visible aspect of unsafe foods, accounting for the overwhelming majority of headlines and public angst. Concerns over recent outbreaks drove Congress to pass the Food Safety Modernization Act in December 2010. Outbreaks hurt not only public health, but also food producers and distributors, by depressing consumption and disrupting markets even after an outbreak is declared over. Researchers with the US Department of Agriculture, for example, found that consumers responded to a prominent outbreak of FI linked to spinach by eating less spinach after the outbreak was declared over and by reducing spending on all leafy greens, at least in the short term. Finally, some FI outbreaks will surely continue under any foreseeable new requirements. Thus, a review of the government’s response to such outbreaks is both important and timely.
This Outlook analyzes data disclosure by the federal government and finds it out of date, incomplete, and inconsistent with best practices. The CDC publishes estimates of trends in the incidence of FI based on data from active population-based surveillance in ten states for all laboratory-confirmed infections with select enteric pathogens. But it does so without disclosing the underlying data or responding to requests for the data under the Freedom of Information Act.
The CDC has disclosed FI outbreak data from 1998 through 2009. This Outlook finds that the percentage of outbreaks not linked to a specific food item has been growing for about a decade. This growth is greater in large outbreaks–those that sicken more than a hundred people. Moreover, between the three years ending in 2000 and the three years ending in 2008, the percentage of all large outbreaks for which the CDC or other food-safety agencies were unable to identify an implicated food item increased by 25 percentage points. The data do not permit identification of a root cause of this trend. The CDC could improve its response to outbreaks by adopting quantitative performance goals. To facilitate the adoption of such goals, the CDC should publicly disclose data that have been kept restricted and unavailable and commit to achieving performance targets for responding to outbreaks.
Outbreaks of Food-Borne Illness
My analysis is based on OutbreakNet, the CDC’s food-borne outbreak online database, which was first posted in the fall of 2009 and updated on October 20, 2010. These data are the best source of information on FI outbreaks available to the public, although the CDC does not claim that they are comprehensive. Before analyzing these data, it is worth noting that the CDC maintains extensive additional data on FI in FoodNet, which it used as the basis of reports in 2010 on trends in incidence. It noted, “FoodNet surveillance data for 2009 show reductions in the incidence of STEC O157 and Shigella infections, but little or no recent progress for other pathogens. Of the four infections with Healthy People 2010 targets (Campylobacter, Listeria, Salmonella, and STEC O157), only the target for STEC O157 was met in 2009.” Interestingly, the CDC does not share the underlying data with the public:
Furthermore, although the CDC monopolizes access to the FoodNet data, its reports about these data are not timely. The CDC’s FoodNet website provides annual reports for only 2007 and earlier and reports only “preliminary” data for 2008 and 2009.
The CDC’s publicly available database on outbreaks, OutBreakNet, consists of information on
While these CDC data are the official government record, they deserve several major caveats. First, they come with a disclaimer that reporting state and local agencies can modify data “at any time, even months or years after an outbreak.” Indeed, in October 2010, the CDC changed the reported number of illnesses and outbreaks for 1998 and 1999, increased the number of reported illnesses for 2000, and lowered those for 2004, 2006, and 2007. Put differently, the data are never final and are subject to change without explanation several years after the outbreak is first reported. Second, a quick perusal of the first hundred outbreaks in the database suggests a lack of consistency. Numbers of deaths are sometimes reported as zeros and other times as blanks. Identified foods are sometimes described in terms as vague as “specialty/ethnic dishes,” “other food *****,” or even “deli tray” and other times specified as precisely as “oysters” or “roma tomato.” The location is sometimes listed as a blank rather than “unknown.” The month is not precisely defined, so it may be interpreted as either the month when the first reported patient got sick or the month when the outbreak was first officially reported as an outbreak. Finally, the newest data are old, ending about twenty-one months before the most recent release, even though they are clearly provisional and not final. Collectively these caveats suggest a lack of respect for, let alone compliance with, basic standards for sound management of statistical data. Notwithstanding these limitations, the data permit some unambiguous conclusions.
First, the total number of outbreaks has declined pretty regularly for a decade, although the trends in outbreaks affecting more than five (or twenty-five) people have been essentially flat for the same period, as shown in figure 1. The outbreaks most likely to have won national attention, those affecting more than one hundred people, show no evidence of any increase (and hint of a decline that does not reach statistical significance), despite press reports suggesting they have risen.
Second, trends in the harm to health from outbreaks are also roughly constant. Figure 2 shows numbers of illnesses, hospitalizations, and deaths, normalized relative to 1998 values. Hospitalizations and deaths are more volatile than illnesses, which are much more numerous. Hospitalizations have increased somewhat while deaths have fallen slightly. The number of illnesses shows no clear increase. An average of these harmful effects–with each weighted to reflect plausible severity of public health effects–shows no increase.
Third, government performance in identifying the food items that cause FI outbreaks has been deteriorating, at least for those outbreaks that involve enough sick people to make the identification of a responsible food item both valuable and practicable. As shown in figure 3, among outbreaks of more than five illnesses, the percentage for which there is no implicated food has been rising for years. Interestingly, this increase is larger for the bigger outbreaks. For outbreaks with more than one hundred illnesses, for example, the percentage without an identified food rose from around 25 percent in 1998, 1999, and 2000 to around 50 percent in 2007 and 2008. Put differently, this trend suggests that in 2007 and 2008, about six additional major outbreaks (those with more than one hundred illnesses) per year were managed poorly because the contaminated food was not identified. Further, the odds of an outbreak having no identified food in 2008 were greater for outbreaks of more than one hundred illnesses (53 percent) than for outbreaks of greater than twenty-five illnesses or five illnesses. This finding is surprising because both the value and the ease of identifying the responsible food item grow as the size of the outbreak increases.
Why has performance deteriorated? One approach to this puzzle is to assess what explanations food-safety agencies have offered. The most authoritative government estimate of trends in FI outbreaks is a presentation made by a CDC official at a March 30, 2010, public meeting convened by the Food and Drug Administration (FDA) called “Measuring Progress on Food Safety–Current Status and Future Directions.” A key part of that analysis used the CDC’s OutbreakNet database to show that the percentage of outbreaks for which there is an implicated food item has been falling. In particular, it declined from just over 60 percent for the three years ending in 2000 to just under 50 percent for the three years ending in 2008. But neither this CDC presentation nor later presentations by CDC and FDA officials at the same meeting explored why this deterioration is occurring, or what can be done to reverse it. While this presentation might be part of a broader acknowledgment by government and independent researchers of longstanding difficulties with these data and the related database, it is regrettable that the CDC has not addressed publicly the causes of the decline. I explore possible causes below.
In principle, agencies may identify responsible food items in fewer outbreaks if the costs of doing so are rising or the benefits are falling. Unfortunately, determining agency behavior is difficult or impossible with available data. Simple tabulations show that several possible explanations of the observed trend appear implausible.
As shown in figure 1, the deterioration is not likely a result of additional workload resulting from more outbreaks, since the number of outbreaks has been roughly constant (except the number of smallest outbreaks, which has been falling). Of course, resources available to food-safety agencies may have fallen, adversely affecting their ability to identify food items causing outbreaks. Exploring this explanation is beyond the scope of this Outlook, however, because it would require compiling information on food-safety budgets for federal, state, and local agencies, all of which are involved in responding to outbreaks.
The deterioration could be due to a shift in outbreaks toward private homes and away from institutions, since identifying responsible food items is generally harder in private homes. In fact, considering only those outbreaks linked to private homes, the percentage with more than five illnesses for which there is no identified food item rose from 30 percent during the first three years of data to 37 percent during the last three years. Figure 4 shows trends in the number of such outbreaks. Not only are they falling, but the two curves are converging, as the share of outbreaks linked to private homes without identified food items rises. Thus, a shift in outbreak location toward private homes is not likely to be contributing to the apparent decline in performance.
A closely related question pertains to outbreaks unrelated to institutions such as schools, cafeterias, restaurants, and hospitals. Identifying the food item responsible for an outbreak is generally easier if the outbreak has been linked to an institution, which would have records of what meal was served when (and even to whom). In fact, trends in the identification of food items in outbreaks not linked to institutions show the same deterioration. The percentage of outbreaks with more than five illnesses not linked to institutions, and for which there is no identified food item, rose by about 13 percentage points over the data period. It was 54 percent for the three years ending in 2000 and 67 percent for the three years ending in 2008. As shown in figure 5, the total number of such outbreaks fell, while such outbreaks without an identified food item also fell, but by less. These data suggest that shifts in the pattern of outbreaks away from institutions and toward homes and similar settings are not likely a cause of trends toward less identification of food items responsible for FI outbreaks.
Finally, the increased use of pulsed field gel electrophoresis (PFGE), also known as DNA fingerprinting, could complicate the identification of responsible food items if it identifies outbreaks much later than conventional methods. Unfortunately, the CDC database provides no information on the use of this technology in investigating different outbreaks. Since a primary benefit of PFGE, however, is the definitive determination of etiology, we may take knowledge of etiology as a proxy for use of PFGE. As shown in figure 6, the percentage of outbreaks with no identified food item increased among outbreaks with any or confirmed etiology. Thus, increased use of PFGE is likely not the cause of deterioration in identifying food items responsible for FI outbreaks.
Of course, other explanations exist. It is possible, for example, that longer lags between exposure and symptoms have hindered efforts to collect information about food consumption from victims and family members. It is also possible that the standards for naming a responsible food item have tightened. Estimating the impact of these factors on the decline in performance is a task for future work.
Lacking information to distinguish definitively among these explanations, I turn below to a prospective solution–an institutional commitment by the CDC and other agencies to accept quantitative measures of their performance responding to outbreaks of FI. Two aspects of recent performance suggest that such a solution may be promising.
As shown in figure 3, the deterioration in the identification of responsible food items has been greater for the larger outbreaks (those with more than one hundred illnesses) than for the smaller ones. This finding is counterintuitive, in that identification of a responsible food item should be easier and more valuable with larger outbreaks. Increasingly scarce resources should therefore result in declining performance in the smaller outbreaks first. The finding that deterioration has been greater for the larger outbreaks suggests a lack of management discipline–and a resulting misallocation of resources to other activities.
Other aspects of the outbreak data set support the hypothesis of inadequate management of responses to outbreaks. For many outbreaks with identified “vehicles of transmission,” the CDC named a food but called it “unspecified.” From 2006 through 2008, for example, among outbreaks of more than twenty-five cases of illness for which the CDC identified a vehicle of transmission, it used the word “unspecified” to describe the vehicle in more than 13 percent of the outbreaks and the word “salad” to describe the vehicle in 7.6 percent of the outbreaks. Describing a vehicle of transmission as “salad” is clearly less informative and less useful than describing it more precisely (for example, spinach). FI outbreaks merit a stronger administrative response than has been seen to date.
Toward Better Control of FI Outbreaks
Federal agencies are responding to FI outbreaks without the quantitative performance measures essential to sound management of large organizations. A compelling example of the importance of such measures is the user-fee program that funds, in large part, the FDA’s review of applications to market new drugs. Before submitting a request to Congress to reauthorize the program, the FDA agrees with the industry on a set of specific quantitative performance goals that it will try to meet with the user fees that the industry agrees to pay. In fact, such performance goals are so important that the FDA and the industry have used them every time they have sought reauthorization of the user-fee program–in 1997, 2002, and 2007. A second example involves a possible user-fee program for the FDA’s review of applications to market generic drugs. The generics industry expects that any user-fee program will provide measurable results and use significant performance metrics, and the FDA has not objected to this view and indeed supports negotiations that identify appropriate measures and levels of performance. Thus, quantitative performance goals can lead to improved efficiencies in program management, at least according to those being asked to pay for substantial increases in such programs.
By contrast, the Obama administration’s budgets for fiscal years 2011and 2012 specified no objective quantitative measures of the CDC’s response to FI, even though this is one of the CDC’s key functions. To justify its budget requests, the administration instead provided quantitative outcome measures that are simply reductions in incidence of disease. For example, one measure is reduction in incidence of salmonella, which caused 15 illnesses per 100,000 people in fiscal year 2009, substantially above the longstanding goal of 6.8 established in 2000 as part of the Healthy People 2010 initiative. Such illnesses come largely from foods regulated by federal agencies–for example, poultry, regulated by the Food Safety and Inspection Service (FSIS) of the US Department of Agriculture (USDA), and produce, regulated by the FDA. Thus, more protective regulations and enforcement by these agencies should lower the incidence of disease from salmonella. Of course, reductions in incidence could also come from the CDC’s improved communications about the importance of good kitchen practices and washing hands, especially after handling pets. But reduction in incidence is by itself not a CDC performance measure because it could also be attributed to actions by the FSIS or FDA.
FI experts have long agreed that improved response to FI outbreaks could come from adopting quantitative performance measures. In 2005, the Council of State and Territorial Epidemiologists, under a cooperative agreement with the CDC, recommended developing performance criteria for routine surveillance and outbreak investigations. It also found that
it is apparent that performance standards should be developed to create expectations of our foodborne disease surveillance system, and to evaluate the performance of the system systematically. Such a system of performance measures and evaluation will support improvements, both in timeliness and effectiveness of foodborne disease surveillance.
One constructive step toward implementing performance standards is to estimate the time needed to respond to evidence of a potential FI outbreak. A CDC presentation at the March 30, 2010, meeting included a figure that illustrates the time elapsed at different stages of the surveillance process. Unfortunately, this schematic does not identify the time elapsed under either typical conditions or best practices.
Some data necessary to estimate such times are available through the CDC’s FoodNet database, but the CDC severely limits access to them. CDC guidelines state, “Proposals for new studies may be initiated by individuals at CDC, any of the FoodNet sites, USDA, or FDA. All proposals must be reviewed by the FoodNet Steering Committee and a vote taken on the proposal.” It is not surprising that members of the FoodNet Steering Committee, having negotiated for themselves a quasi monopoly over publication rights from the data, would be reluctant to let outside researchers have access to them, let alone use them to evaluate the time elapsed for various surveillance steps. The broader question is why leadership of the CDC or Department of Health and Human Services, which purportedly adhere to the administration’s policies of full transparency in governance, do not immediately post all FoodNet data in their entirety, subject only to incorporation of appropriate protection of patient privacy. Public access to such data is essential in estimating how much time elapsed in each step of the outbreak response, either under typical conditions or with best practices.
There is, however, an additional problem, beyond the CDC’s failure to share government-funded data on food-borne illness. Before adopting performance standards for an activity like identifying a vehicle of transmission, one needs agreement on standards–for example, what precisely is meant by a “vehicle”? A CDC presentation illustrates the problem. In discussing how best to respond to outbreaks, it proposes distinguishing between simple foods (that is, consisting of a single class of commodities, like fruit) and complex foods (that is, consisting of multiple commodities, like meat loaf). One approach might be first to define a vehicle as one of three progressively more specific classes–a category of food (leafy greens, seafood, canned foods, or meat), a specific type of food (fresh uncut spinach), and a specific type of food from a specific retailer or distributor (brand X fresh spinach). This standardized definition could help the CDC quantitatively evaluate its performance in recent outbreaks.
There is a similar complication in identifying a vehicle of transmission: the CDC lacks a standard for identification. In principle, it could be either a statistically significant finding on a well-developed control study or PFGE results on food products that match those from stool samples of people sickened by the FI outbreak, or something else entirely. Without a standard for what constitutes identification, there is no basis for judging a response to an outbreak.
The view that the CDC needs to adopt quantitative performance standards for responses to FI outbreaks is not new, as made clear by reports by the Council of State and Territorial Epidemiologists. The apparent inertia or disinterest in such standards results from institutional incentives. No single organization with clout is committed enough to improved performance to take the steps necessary to drive formal adoption of quantitative performance standards.
A user-fee program tied to performance standards might be an effective catalyst for change. Enactment of the Prescription Drug User Fee Act in 1992 prompted the FDA to formally adopt a set of quantitative performance goals that contributed to improved performance in its process of reviewing applications to market new drugs. The act also contributed to a robust public discussion about how to measure and improve the FDA’s performance in new drug review. A user-fee program providing funds to facilitate effective responses to FI outbreaks may have similar effects.
This review of CDC data shows that agencies responsible for food safety are increasingly ineffective at identifying the food item responsible for FI outbreaks. Although this review has not identified the root cause of this deterioration in performance, it suggests that the deterioration does not appear to be uniquely or primarily the result of
Despite the lack of a clear culprit in the deterioration in identifying foods responsible for FI outbreaks, experts believe that adopting quantitative metrics would improve performance. One necessary first step would be to share the FoodNet database, subject only to ensuring adequate protection of patient privacy. A second would be to continue posting in a more timely way all data related to food items implicated in outbreaks. The current practice of posting preliminary data many months or years after the close of the reporting period is at odds with accepted standards. In addition, greater public access to these data is the best way to jumpstart the analysis necessary to develop quantitative performance measures that could be formally adopted as agency commitments. A third step would be for the head of the CDC to say publicly that it is committed to improved performance in responding to outbreaks and that appropriate performance measures will be proposed shortly.
Randall Lutter ([email protected]) is an adjunct scholar at AEI.
I am grateful to David Acheson, Bob Scharff, and Richard Williams for helpful discussions and am solely responsible for all the contents.
1. See Carlos Arnade, Linda Calvin, and Fred Kuchler, “Consumers’ Response to the 2006 Foodborne Illness Outbreak Linked to Spinach,” Amber Waves, March 2010, www.ers.usda.gov/amberwaves/march10/features/OutbreakSpinach.htm (accessed June 2, 2011).
2. See Centers for Disease Control and Prevention (CDC), “OutbreakNet: Foodborne Outbreak Online Database,” wwwn.cdc.gov/foodborneoutbreaks/Default.aspx (accessed June 2, 2011).
3. CDC, “Preliminary FoodNet Data on the Incidence of Infection with Pathogens Transmitted Commonly through Food–10 States, 2009,” Morbidity and Mortality Weekly Report 59, no. 14 (April 16, 2010): 418-422. Although the CDC typically issues annual reports on FoodNet in April, as of this writing there is no such report in April or May 2011.
4. For example, the journal Nature states, “An inherent principle of publication is that others should be able to replicate and build upon the authors’ published claims. Therefore, a condition of publication in a Nature journal is that authors are required to make materials, data and associated protocols promptly available to readers without undue qualifications in material transfer agreements.” See www.nature.com/authors/policies/availability.html. Data-disclosure policies for Proceedings of the National Academy of Sciences and Science are similar. See www.pnas.org/site/misc/iforc.shtml#vii and www.sciencemag.org/site/feature/contribinfo/prep/gen_info.xhtml#dataavail, respectively.
5. See the CDC’s FoodNet Reports, available at www.cdc.gov/foodnet/reports.htm.
6. The outbreak data should also be contrasted with data on selected FI from all causes collected through the CDC’s active surveillance system, FoodNet, which uses results of laboratory tests for ten states. FoodNet includes illnesses unrelated to outbreaks and from any cause, which may include exposure to pets. Using FoodNet, CDC researchers have reported, “In comparison with the first 3 years of surveillance (1996-1998), sustained declines in the reported incidence of infections caused by Campylobacter, Listeria, Salmonella, Shiga toxin-producing Escherichia coli (STEC) O157, Shigella, and Yersinia were observed. The incidence of Vibrio infection continued to increase. Compared with the preceding 3 years (2006-2008), significant decreases in the reported incidence of Shigella and STEC O157 infections were observed.” See CDC, “Preliminary FoodNet Data on the Incidence of Infection with Pathogens Transmitted Commonly through Food–10 States, 2009.”
7. The CDC updates OutbreakNet less frequently, for example, than the European Union updates statistics on its FI outbreaks. In December 2009, the European Union issued statistics on FI outbreaks for 2008 and did not describe its estimates as preliminary. See European Food Safety Authority, “Community Summary Report: Trends and Sources of Zoonoses and Zoonotic Agents and Food-Borne Outbreaks in the European Union in 2008,” EFSA Journal 8, no. 1 (April 26, 2010): 1496, www.efsa.europa.eu/en/scdocs/doc/1496.pdf (accessed March 8, 2011).
8. For the three years beginning in 1998, outbreaks affecting more than one hundred people averaged thirty-nine per year; for the three years ending in 2008, they averaged only twenty-nine per year.
9. For illustrative purposes, I construct a mortality equivalent index assuming that one thousand illnesses or one hundred hospitalizations count the same as one death. During the three years ending in 2000, this index had an average annual value of forty-six, while for the three years ending in 2008 it had an average annual value of forty-four.
10. The average annual number of outbreaks with more than one hundred illnesses was twenty-nine during the years 2006-2008. Figure 3 suggests that the performance deteriorated by nearly 25 percentage points from the first three years to the last three years of the period of data. Thus, the deterioration amounts to roughly six major outbreaks annually.
11. Presentations made at that meeting are available from the Food and Drug Administration (FDA) through [email protected]
12. See CDC, “OutbreakNet: Foodborne Outbreak Online Database,” wwwn.cdc.gov/foodborneoutbreaks (accessed January 15, 2010). The CDC’s FoodNet database, described earlier, does not focus on outbreaks and is not publicly available. I analyze this database, rather than one posted by the Center for Science in the Public Interest (CSPI), because the CSPI database includes only outbreaks for which both the food and the pathogen have been identified. See CSPI, “Outbreak Alert! Database,” www.cspinet.org/foodsafety/outbreak/pathogen.php (accessed March 22, 2010). A presentation by Caroline Smith DeWaal on differences in reporting of food-borne illness among states also mentions that reporting appears to be declining over time. See Caroline Smith DeWaal, “Beyond Attribution: State-by-State Outbreak Reporting” (Food Safety Education Conference, Atlanta, Georgia, March 23-26, 2010), http://cspinet.org/new/pdf/statereport2010.pdf (accessed March 9, 2011).
13. See, for example, a summary of difficulties attributing outbreaks to foods: Michael B. Batz, Michael P. Doyle, J. Glenn Morris Jr., John Painter, Ruby Singh, Robert V. Tauxe, Michael R. Taylor, and Danilo M. A. Lo Fo Wong, “Attributing Illness to Food,” Emerging Infectious Diseases 11, no. 7 (July 2005), www.cdc.gov/ncidod/eid/vol11no07/pdfs/Vol11No7.pdf (accessed March 10, 2011). See also CDC, “Program Response to the External Peer Review of Foodborne Illness Detection and Investigation in MultiState Outbreaks,” www.cdc.gov/foodsafety (accessed March 10, 2011).
14. FDA, “For Industry: Letters (PDUFA),” www.fda.gov/ForIndustry/UserFees/PrescriptionDrugUserFee/ucm149212.htm (accessed March 9, 2011); see especially the list of specific quantitative performance goals at FDA, “For Industry: Section A: PDUFA Reauthorization Performance Goals and Procedures for Fiscal Years 2008 through 2012,” www.fda.gov/ForIndustry/UserFees/PrescriptionDrugUserFee/ucm119243.htm (accessed March 9, 2011).
15. Generic Pharmaceutical Association, “Statement from GPhA President and CEO Kathleen Jaeger in Response to Remarks by FDA Commissioner Margaret Hamburg at the GPhA Annual Meeting,” news release, February 18, 2010, www.gphaonline.org/media/press-releases/2010/statement-gpha-president-and-ceo-kathleen-jaeger-response-remarks-fda-comm (accessed March 9, 2011).
16. In testimony to the House Committee on Agriculture, CDC director Lonnie King stated: “CDC leads federal efforts to gather data on foodborne illnesses, investigate foodborne illnesses and outbreaks and monitor the effectiveness of prevention and control efforts.” See Findings from CDC’s Foodbourne Diseases Active Surveillance Network (FoodNet), Before the Committee on Agriculture, 111th Cong. (May 14, 2009) (statement of Lonnie J. King, D.V.M., Director of the Centers for Disease Control and Prevention), www.hhs.gov/asl/testify/2009/05/t20090514c.html (accessed March 9, 2011).
17. See Department of Health and Human Services, “Fiscal Year 2012, Centers for Disease Control and Prevention, Justification of Estimates for Appropriation Committees,” 121, www.cdc.gov/fmo/topic/BudgetInformation/appropriations_budget_form_pdf/FY2012_CDC_CJ_Final.pdf (accessed June 6, 2011). Other pathogens have incidence rates that already meet (E. Coli O157:H7 and Listeria moncytogenes) or almost meet (Campylobacter) the incidence goals established in Healthy People 2010. The Department of Health and Human Services led the development of the Healthy People 2010 goals, which were finalized in 2000. See www.healthypeople.gov.
18. In 2002, an advisory committee of the Council of State and Territorial Epidemiologists developed minimum performance/capacity standards for food-borne disease surveillance. See Council of State and Territorial Epidemiologists, “Food Safety: Background and History,” www.cste.org/dnn/ProgramsandActivities/InfectiousDiseases/FoodSafety/tabid/250/Default.aspx (accessed March 9, 2011).
19. Craig Hedberg, Enteric Disease Timeline Study (Division of Environmental Health Sciences, School of Public Health, University of Minnesota, August 2005), www.cste.org/dnn/Portals/0/EDITS-final-report.pdf (accessed March 9, 2011).
20. See Christopher R. Braden, M.D., “Measuring the Effectiveness of Process; Linking to Outcome” (presentation, Public Workshop on Measuring Progress on Food Safety, Food and Drug Administration, Washington, DC, March 30, 2010), 5, http://packmgmt.s3.amazonaws.com/pdfs/FDA_20100330/4%20Food%20Saftey%20Metrics%20Mtg%2030MAR10%20v1.pdf (accessed June 2, 2011).
21. The CDC describes FoodNet as follows: “The Foodborne Diseases Active Surveillance Network (FoodNet) is the principal foodborne disease component of CDC’s Emerging Infections Program (EIP). FoodNet is a collaborative project of the CDC, ten EIP sites, the US Department of Agriculture (USDA), and the Food and Drug Administration (FDA).” See http://www.cdc.gov/FoodNet. Variables include the date symptoms first began, date of hospital admission, date specimen was received in the laboratory, date case was entered into the database, etc.
22. See CDC’s Emerging Infections Program, Foodborne Diseases Active Surveillance Network (FoodNet) Protocol Development and Publication Policy, available on file with the author.
23. See Patricia Griffin, M.D., “Attribution of Foodborne Illness to Food Commodities: An Approach Using Complex Outbreak Data” (presentation, FDA/CDC public meeting, “Measuring Progress on Food Safety–Current Status and Future Directions,” March 30, 2010).
There are no comments available.
1150 17th Street, N.W. Washington, D.C. 20036
© 2016 American Enterprise Institute for Public Policy Research