Module 4: Epidemiology Investigation

Epidemiology Division Analysis and Reporting Branch

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services Module Learning Objectives Develop a case definition Explain an epidemic curve Develop an initial hypothesis List three types of study design and a method of statistical analysis Calculate measures of association Interpret significance of data Determine if hypotheses are confirmed or rejected Perform group exercise Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 2 Estimated Annual US Foodborne Disease Burden

Foodborne Illnesses 47.8 million Hospitalizations 127, 839 Deaths 3, 037

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 3 AL Foodborne Outbreak (FBO) Team

Montgomery Surveillance Branch - Epidemiology Analysis and Reporting Branch - Epidemiology Bureau of Clinical Laboratories (BCL) Bureau of Environmental Services – Food, Milk and Lodging Throug hout t he State Environmentalists Field S ur veillance Staff (FSS) Others (in special situations)

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 4 Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 5 Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 6 AL Foodborne Outbreak (FBO) Definition

An incident in which two or more persons from different households experience a similar illness resulting from the ingestion of a common food. Exceptions: botulism and chemical poisoning (1 case = outbreak)

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 7 How Are Outbreaks Named?

The outbreak name identifies the state, the year, month, species/serotype, primary county, and sequence (order of outbreaks in the same county in the same month). The format is Alabama/year/month/species, serotype, or enzyypme pattern/h yphen/count y number/ and se quence. PHA03 Deli :

PHA08 Church Luncheon : 26

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 8 AL Foodborne Outbreaks

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 9 Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 10 County Numbers Autauga 01 Conecuh 18 Houston 35 Morgan 52 Baldwin 02 Coosa 19 Jackson 36 Perry 53 Barbour 03 Covington 20 Jefferson 37 Pickens 54 Bibb 04 Crenshaw 21 Lamar 38 Pike 55 Blount 05 Cullman 22 Lauderdale 39 Randolph 56 Bullock 06 Dale 23 Lawrence 40 Russell 57 Butler 07 Dallas 24 Lee 41 St. Clair 58 Calhoun 08 DeKalb 25 Limestone 42 Shelby 59 Chambers 09 Elmore 26 Lowndes 43 Sumter 60 Cherokee 10 Escambia 27 Macon 44 Talladega 61 Chilton 11 Etowah 28 Madison 45 Tallapoosa 62 Choct aw 12 FttFayette 29 Marengo 46 TlTuscaloosa 63 Clarke 13 Franklin 30 Marion 47 Walker 64 Clay 14 Geneva 31 Marshall 48 Washington 65 Cleburne 15 Greene 32 Mobile 49 Wilcox 66 Coffee16Hale33Monroe50Winston67 Colbert 17 Henry 34 Montgomery 51

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 11 Develop Initial Case Definition: Count ing Appl es and Oranges

StSet of crit itieria f or d ecidi ng w hthhether an individual should be classified as “ill” Objective criteria Outbreak-associated vs. normal background Value of routine surveillance data Primary vs. secondary cases

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 12 Outbreak Case Definitions

Begin general, but become increasingly specific as information is gathered Person, Place and Time association Clinical criteria Classify cases based on certainty Confirmed (()positive lab) Probable (no lab; symptoms) Suspect (no lab; some symptoms)

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 13 Revise Case Definitions

Ongoing throughout investigation PidfiiiPrecise definitions re duce potential for misclassification: Estimated 36% of enteric infections are foodborne; IlthithtIncomplete case history can haunt you; Incorrectly classifying individuals can bias results; Make it harder to detect true associations .

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 14 Misclassification Example: SldSalad Exposure

Not ill

Salad Yes (?)

Not ill

Salad No (?)

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 15 Misclassification Example: Exposure Date Incubation Period

Norovirus average incubation period: 36 hrs (range: 10-50 hrs) Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 16 Misclassification Example: Exposure Date Incubation Period

36912152124273033363942454851543 6 9 12 15 21 24 27 30 33 36 39 42 45 48 51 54

Norovirus average incubation period: 36 hrs (range: 10-50 hrs) Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 17 Make Epidemiologic Associations

PPlTiPerson, Place, Time Syyygystematically organize key information Develop initial hypothesis

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 18 Person: AL Outbreak Examples

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 19 Syypmptoms: Fre quency Table

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 20 Place: Common Exposure Location

Case 1 – rest. A, B, C, D Case 2 – rest. B, E, F Case 3 – rest. A, B, G, H, I Case 4 – rest. B,,,, D, J, K

What is the common exposure location?

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 21 Place: Common Exposure LLtiocation ((t)cont.) Restaurant Same restaurant Multipp,le restaurants, same chain or owner Multiple chains, common distributor Banquets, Birthday Parties, Wedding Receptions, Conferences, Basketball Game, School Field Trip, Summer Camp, etc. Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 22 Place

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 23 An Epidemic Curve Shows progression of outbreak over time

12 Fatal X axis = units of time (1/4 to 1/3 of incubation time) 11 Non-fatal Y axis = number of cases 10 9 There is an inherent ases 8 delay between the date CC 7 of illness onset and the 6 date the case is reported to Public 5 mber of Health uu 4 N 3 2 1 0 6/27 8/2 9/6 10/11 11/15 12/20 1/24 2/27 4/3 Date Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 24 Source: Centers for Disease Control and Prevention Salmonella Outbreak Investigations: Time line for Repor tin g Cases Patient Eats 1-3 Days Incubation Contaminated Food Time to contact with health Patient care system = 1-5 days Becomes Ill

Stool Sample Time to diagnosis = 1-3 days Collected

Shipping Time = 0-7 days Salmonella Identified Public Health Lab Receives Case Sample Serotyping and “DNA Confirmed fingerprinting” Time = 2-10 days As Part of Outbreak

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services Source: CDC 25 Point-- Source Exposure

10 9 8 7 Cases 6 5

ber of 4 mm 3

Nu 2 1 0 13579111315 Days

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 26 Ongoing Exposure 10 9 8 7 Cases 6 5

ber of 4 mm 3

Nu 2 1 0 13579111315 Days Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 27 Secondary Exposures 10 9 8

ases 7 CC 6 5

ber of 4 mm 3

Nu 2 1 0 13579111315 Days Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 28 Hypothesis

“An unproved theory ...tentatively accepted to explain certain facts or to provide a basis for further investigation”

Source: Webster’s New World Dictionary, 3rd Edition

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 29 HHthypotheses and QQtiiuestionnaires

Hypothesis generating

Hypothesis testing

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 30 Example of Hypothesis

Food item consumed at the Smith Wedding reception caused illnesses Data needed to test information What kind of food items were eaten? Who ate the food item? Who did not eat? How much did each person eat? Illness onset (()date and time)?

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 31 Developpyp Initial Hypothesis

Multippyple hypotheses ma y be compatible with data initially Helps clarify What is known What is missing Actions needed to gather missing information

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 32 Unknown Causative Agent

Review what is known about cases Symptoms, severity of disease Events attended or anyygthing unusual Foods consumed and methods of food pppreparation Identify most likely agent(s) Review references Consultation

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 33 Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 34 Known Causative Agent

Review what is known about the agent Typical signs and symptoms Modes of transmission Foods in past outbreaks Is this situation similar to other previously reported incidents? Be aware of recent food recalls

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 35 New Routes of Transmission are Being Identified

Example: E. coli O157:H7 in apple cider

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 36 Case Definition vs. Hypothesis

Case Definition Hypothesis Classify cases Describe vs. controls exposure to Includes test analytically Person Theory Place Do NOT include Time hypothesis in case definition! Symptoms

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 37 Group Exercise

Case Definition and Hypothesis

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 38 Questionnaire Design/ Interview Process

Questions should be clear, concise and asked in an uniform way Avoid open -ended questions, unless hypotheses generating questionnaire Include menu, when available, to avoid recall bias Questionnaire should take into account completeness of data collection, time involved, and relevance of questions Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 39 QQg/uestionnaire Design/ Interview Process --22

Include as many not-ill as possible; Review questionnaire to make sure there is no conflicting/missing information to avoid calling back; Stool and food samples are very helpful in solving the outbreak

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 40 Examples of Epidemiological Studies

1. Case Series 2. Cohort Studies 3. Case-Control Studies

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 41 1. Case Series

AittifAppropriate to use if: Small number of cases (< 5) Controls are unavailable (e. g., everyone ill) Provide person, place, and time associations

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 42 2. Cohort Studies

Groups of exposed and unexposed individuals can be easily identified Compare risk of illness by what was/was not eaten

Example: Church supper

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 43 3. Case-- Control Studies

Typically used in foodborne investigations when: Exposed group is very large or Exposed group is not easily identified Compare ill with non-ill Individuals to determine likelihood of having eaten specific foods Exampl es: very large we dding recep tion or sporadic E. coli cases

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 44 Measures (Calculations)

Measure of Occurrence Attack Rate Measure of Association Relative Risk (RR): Cohort studies Odds Ratio (OR): Case -Control studies

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 45 Attack Rate: Measure of Occurrence Attack Rate (AR) Expresses occurrence of a disease among a particular at-risk population for a limited period of time, often due to a very specific exposure. Can be event-specific or food-specific Number ill people who consumed item AR = Total number people consuming item

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 46 Measures of Association Between Exposure (Food Item) and Disease

Depends on type of study “Relative risk”  cohort studies Risk of developing disease given the exposure “Odds ratio”  case-control studies Odds of h avi ng th e exposure g iven the disease Computer programs greatly speed the calculations

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 47 Estimating Risks Associated with “Exposure”

Compare attack rates among exposed and unexposed

Relative Risk = Attack Rate (exposed) / Attack Rate (unexposed)

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 48 2 by 2 Table to Calculate Relative Risk Estimates magnitude of association between exposure and disease in exppgposed group relative to unexp pgposed group

Disease

Yes ( + ) No ( - ) total

Yes ( + ) aba + b a / (()a + b) Exposure RR = No ( - ) cdc + d c / (c + d)

total a+ca + c b+db + d a+b+c+d

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 49 Relative Risk -- Cohort Studies

Measurement of the risk of developing disease given a specific exposure (a food item) Compares attack rates between exposed and unexposed groups

Relative Risk = Attack Rate (()AR) among Exposed = a / (a + b) (RR) Attack Rate (AR) among Unexposed c /(/ (c + d)

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 50 Number of persons who ate Number of persons who did specified item not eat specified item Ill Not Ill Total Attack rate (%) Ill Not Ill Total Attack rate % Baked ham 29 17 46 63 17 12 29 59 Spinach 26 17 43 60 20 12 32 62 Mashed potato 23 14 37 62 23 14 37 62 Cabbage salad 18 10 28 64 28 19 47 60 Jello 16 7 23 70 30 22 52 58 Rolls 21 16 37 57 25 13 38 66 Brown bread 18 9 27 67 28 20 48 58 Milk 22 4 50 44 27 71 62 Coffee 19 12 31 61 27 17 44 61 Water 13 11 24 54 33 18 51 65 Cakes 27 13 40 67 19 16 35 54 Ice cream (()van) 43 11 54 80 31821 14 Ice cream (choc) 25 22 47 53 20 7 27 74 Fruit salad 42 6 67 42 27 69 61

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services Reference: CDC Epi-Ready Training Manual 51 Church Supper Cohort Study – Relative Risk

Van illa ice cream AR (exposed) ÷ AR (unexposed) 79. 6 ÷ 14. 3 = 565.6 Persons consuming vanilla ice cream were 5.6 times more likely to become ill than those who did not eat vanilla ice cream

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 52 Number of persons who ate Number of persons who did specified item not eat specified item Ill Not Ill Total Attack rate (%) Ill Not Ill Total Attack rate % Baked ham 29 17 46 63 17 12 29 59 Spinach 26 17 43 60 20 12 32 62 Mashed potato 23 14 37 62 23 14 37 62 Cabbage salad 18 10 28 64 28 19 47 60 Jello 16 7 23 70 30 22 52 58 Rolls 21 16 37 57 25 13 38 66 Brown bread 18 9 27 67 28 20 48 58 Milk 22 4 50 44 27 71 62 Coffee 19 12 31 61 27 17 44 61 Water 13 11 24 54 33 18 51 65 Cakes 27 13 40 67 19 16 35 54 Ice cream (()van) 43 11 54 80 31821 14 Ice cream (choc) 25 22 47 53 20 7 27 74 Fruit salad 42 6 67 42 27 69 61

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services Reference: CDC Epi-Ready Training Manual 53 Church Supper Cohort Study --RelativeRelative Risk

Chocolate ice cream AR (exposed) ÷ AR (unexposed) 53. 2 ÷ 74. 1 = 070.7 Persons consuming chocolate ice cream were 0.7 times as likely (or 30% less likely) to become ill than those who did not eat chocolate ice cream

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 54 2 by 2 Table to Calculate Odds Ratio Estimates magnitude of association between exposure and disease in diseased group (cases) relative to non-diseased group (control)

Disease

Y()Yes ( + ) N()No ( - ) ttltotal

Yes ( + ) aba + b a/c ad OR = = Exposure b/d bc No ( - ) cdc + d

total aac + c bbd + d a+b+c+d

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 55 Odds Ratio -- CaseCase--ControlControl Studies Measurement of the odds of having an exposure (specific food consumption) given the disease Estimates the Relative Risk derived from cohort studies

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 56 How Valid are Your Findings? Is the observed association between exposure and disease (expressed by the RR or OR) due to alternative explanations?

Bias: systematic error Selection Information (recall, interviewer, misclassification) Chance Sampling variability Sampling size

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 57 Examples of Bias

Random misclassification of cases vs. controls or exposed vs. unexposed biases OR or RR toward “1” Recall bias cases remember exposures better than do controls problem with retrospective studies may result in inflated OR

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 58 Evaluating the Role of Chance

P-value Pro ba bility a g iven assoc ia tion cou ld have occurred by chance alone “Statistically significant” defined as p ≤ 0.05 Consider all available evidence when interpreting p-values 95% Confidence Interval Range within which the true association lies, based on 95% assurance

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 59 PP--vvalues and Confidence Intervals Example 1 Table 16 Analysis of Food Item 2 ILL 5 / (5 + 10) ..3333 ITEM2 | + - | Total RR = = ------+------+------9 / (9 + 11) .45 + | 5 10 | 15 - | 9 11 | 20 ------+------+------Total | 14 21 | 35

• RR= 0.74 • 95% confidence limits for RR = 0.43 < RR < 2.79 Not significant because includes “1”

•P-value = 0.486 Not significant because more than “0.05”

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 60 PP--vvalues and Confidence Intervals Example 2 Table 15 Analysis of Food Item 1 ILL 11/(1111 / (11 + 4) .7333 ITEM1 | + - | Total ------+------+------RR = = + | 11 4 | 15 3 / (3 + 17) .15 - | 3 17 | 20 ------+------+------Total | 14 21 | 35

• RR= 4.89 • 95% confidence limits for RR = 1.65 < RR < 14.50 Significant because does not include “1”

•P-value = 0.000490 Significant because less than “0.05”

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 61 P-values and Confidence Intervals Example 2 (cont.)

Indicates that best estimate of increased risk of illness associated with consumption of food item 1 is 4.89 times; We are 95% confident that the true risk ratio is no less than 1.65 and no greater than 14.5 times the risk of those not consuming this item. Lack of inclusion of 1.0 in this interval demonstrates a significantly positive association.

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 62 Increase the Sample Size

Increase the number of controls (persons not ill)

Outbreak # Ill # Not Ill

12 3 14 7 ☺ 11 11 721

Controls Cases

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 63 Provide Denominator Among Exposed Example: AL1003JEG-26a (Church ) 15 Interviewed (11 ill, 4 not ill) 2 events (Banquet on Saturday and Luncheon on Sunday) 60-65 attended banquet 20-25 attended luncheon

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 64 PP--vvalue: Effect of Sample Size

Disease Exposure Ill Not ill Total ++-- Eat food + 441155 Did not eat food - 112233 553388

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 65 PP--vvalue: Effect of Sample Size (cont.)

Disease Exposure ill Not ill Total ++--

Ate food + 8 2 10 Did not eat food - 224466

10 661616

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 66 PP--vvalue: Effect of Sample Size (cont.)

Disease Exposure Ill Not ill Total ++-- Ate food + 12 331515 Did not eat food - 336699 15 992424

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 67 True Story

4 friends meet for dinner 3 eat chicken fried rice 1 chop suey 3 ill with vomiting & diarrhea 12- 18 hrs. after meal No samppgles available for testing Time temperature abuse of fried rice documented P-value = 0.2 Is this a foodborne outbreak? Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 68 The Significance of Significance

“Not significant” doesn’t necessarily mean “no association.” The measure of association (relative risk, odds ratio))g indicates the direction and strength of the association. The statistical test (p-value), indicates how likely it is that the observed association may have occurred by chance alone. No siggynificance may reflect no association in the source population but may also reflect a study size too small to detect a true association in the source

population. Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 69 Source: Field Epidemiology, 1996, Michael B. Gregg Interpret Results

Hypotheses confirmed or rejected? If rejected, develop new hypotheses Test new hyyypotheses by: More analysis with existing data Gather and analyze new data Final conclusions Was it foodborne?

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 70 Problems Can Interfere with Outbreak Solution

Questionnaire design Data collection (()ex., incomplete data) Sample size (ill and not ill) No clinical or food sample for testing Reporting time

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 71 Statistical Summary

RR is calculated in cohort studies RR compares rate of illness in exposed group to rate of illness in unexposed group OR is calculated in case-control studies OR = odds in favor of exposure among cases compared to the odds in favor of exposure among the controls RR and OR are both measures of association “Not significant” ≠ “No association” No significance may reflect lack of association but may also reflect a study size too small to detect true association Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 72 Statistical Take Home Message

If RR or OR = 1, then there is no association between consuming food item and getting sick; If RR or OR is less than 1, then food item is “protective”; If RR or OR is more than 1, then food item is “implicated”; If the 95% CI includes 1, then the result is not considered statistically significant (i.e. there is no statistical difference between the groups); The lower the p-value the more unlikely the results are due to chance

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 73 MultiMulti--statestate FBOs

Not all agents are created equal (e.g. S. typhimurium) Microbes have finggperprints too! CDC monitors a database (PulseNet) with microbial fingerprint information and notifies states that are part of multi-state outbreaks; Periodical conference calls with involved states and regulatory agencies to collaborate on outbreak investigation and response; More information from cases may be needed CDC and ADPH website, links, and media coverage

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 74 Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 75 National 2010 FBOs

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 76 Module Summary TkTake Home PPitoints Have a working Case Definition and Hypothesis; Bfibltftii(Be aware of possible routes of transmission (new and old); Questionnaire design is paramount; Interview as many people as possible including “not ill”; Avoid misclassification (exposure) and bias (e. g. recall, selection); Timeliness of reporting is key; Communication is extremely important; Outbreak investigation is a collaborative effort!

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 77 Module Learning Objectives

Developed a case definition Explained an epidemic curve Developed an initial hypothesis LitdthListed three types o f s tdtudy d es ign an d a method of statistical analysis Calculated measures of association Interpreted significance of data DtDeterm ine difhd if hypo theses are con firme d or rejected

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 78 Group Exercise Joe’s Thanksgiving Dinner Outbreak

Epidemiology Division Bureau of Clinical Laboratories Bureau of Environmental Services 79