Biostatistics & Population Health

Outbreak investigation: epidemic curve and case definition

Clinical Overview and When to Suspect an Outbreak

— A clinician, lab, or school nurse reports ≥2 cases of the same illness with epidemiologic linkage (shared meal, daycare, hospital ward, nursing home, water source)

— Surveillance data show case counts exceeding the endemic threshold (often defined as mean + 2 SD of prior comparable periods)

— A single case of a notifiable/eradicated condition (measles, polio, anthrax, smallpox, viral hemorrhagic fever, botulism) — one case is an outbreak

— Unusual clustering by time, place, or person

— 1. Prepare for fieldwork

— 2. Establish existence of an outbreak

— 3. Verify the diagnosis

— 4. Construct a working case definition

— 5. Find cases systematically and record information (line list)

— 6. Perform descriptive epidemiology (epi curve, person/place/time)

— 7. Develop hypotheses

— 8. Evaluate hypotheses (analytic studies: cohort or case-control)

— 9. Refine hypotheses and conduct additional studies

— 10. Implement control/prevention; communicate findings

Definition: An outbreak is the occurrence of cases of a disease in excess of what is normally expected in a defined community, geographic area, or season. "Epidemic" is synonymous but often implies larger scale; "cluster" is an aggregation of cases grouped in place and time without yet exceeding expected baseline.

When to suspect an outbreak (Step 3 trigger phrases):

Step 3 management: When notified of a possible outbreak, the physician's immediate duties are (1) report to the local/state health department per state law, (2) preserve clinical specimens, (3) isolate patients if transmission risk exists, and (4) begin a line list before memory degrades.

10 classic CDC outbreak investigation steps (memorize order — frequently tested):

Board pearl: Steps 2 and 3 (confirm outbreak exists; confirm diagnosis) precede case definition — never jump to interventions before verifying the diagnosis isn't a lab artifact or improved reporting masquerading as a true increase in incidence.

Presentation Patterns and Key History — Recognizing the Outbreak Signal

— Cluster of patients with similar GI symptoms hours after a wedding, cruise, or buffet → foodborne

— Multiple hospital roommates with new diarrhea on antibiotics → C. difficile nosocomial outbreak

— Several daycare children with watery diarrhea and vomiting in winter → rotavirus/norovirus

— Pneumonia cluster among hotel guests near a cooling tower → Legionella

— Febrile rash in unvaccinated school cohort → measles

— Bloody diarrhea after a petting zoo or undercooked beef → STEC O157:H7

— Who: age, occupation, immunization status, immune status, shared exposures

— What: symptom complex, severity, duration; ask about exact onset date and time (critical for epi curve)

— When: date/time of symptom onset for each case → builds the epi curve

— Where: geographic location, travel, restaurants, water sources, animal contact

— Why/exposures: food history (72 hr for most foodborne, longer for Listeria, hepatitis A), sick contacts, sexual contacts, IV drug use, healthcare exposures

— Short (1–6 h): preformed toxins — S. aureus, B. cereus emetic

— 8–16 h: C. perfringens, B. cereus diarrheal

— 1–3 days: norovirus, Salmonella, Shigella, Campylobacter

— 3–8 days: STEC, Yersinia

— Weeks: hepatitis A (15–50 d), Listeria (up to 70 d in pregnancy)

How outbreaks present to the Step 3 examinee:

Key history elements (the "5 W's" of outbreak intake):

Incubation period as a history tool:

Board pearl: If the stem provides symptom onset times for multiple patients sharing a meal, compute the median incubation period — it narrows pathogen identity faster than stool studies and is the highest-yield first analytic step on the exam.

Key distinction: "Cluster" describes geographic-temporal grouping without epidemiologic confirmation of excess; "outbreak" requires demonstrating incidence above baseline. Use the right word in your H&P note.

Descriptive Epidemiology — Person, Place, Time, and the Physical "Exam" of an Outbreak

— Age and sex distribution (e.g., outbreak hitting only women aged 20–40 → suspect a salon, cosmetic injectable, or obstetric exposure)

— Occupation (farmworkers → zoonoses; healthcare workers → nosocomial)

— Vaccination/immunocompromise status (measles in unvaccinated; PCP in HIV)

— Race/ethnicity and socioeconomic indicators if relevant to access or exposure

— Spot map plots case residences or exposure sites — John Snow's Broad Street pump remains the prototype

— Clustering around a single venue → point-source suspicion

— Diffuse spread along transit lines, schools, or households → propagated suspicion

— Hospital floor plans for nosocomial outbreaks (room, ward, shift, shared equipment)

— Build the epidemic curve (next chunk)

— Note secular trends, seasonality, day-of-week patterns

— AR = (ill exposed) ÷ (total exposed) × 100

— Compute AR for each food item; the item with the highest AR among eaters AND lowest AR among non-eaters is the likely vehicle

— Use the relative risk in cohort designs of closed outbreaks (wedding, cruise)

Outbreak investigations replace the physical exam with descriptive epidemiology: characterizing cases by person, place, and time.

Person:

Place:

Time:

Attack rate (AR):

CCS pearl: When the simulated case involves a suspected nosocomial outbreak, "order" a chart review/line list, contact infection prevention, place suspected cases on appropriate transmission-based precautions, and notify public health — all before sending advanced microbiology.

Board pearl: Person-place-time characterization is mandatory before hypothesis testing. Skipping descriptive epi and jumping to case-control analysis is a classic distractor on exam questions about next-best-step in an outbreak.

Key distinction: Spot maps show place; epidemic curves show time; line lists organize person-level data. The exam will ask which tool answers which question.

The Case Definition — Construction and Components

— Clinical criteria: specific signs, symptoms, lab findings (e.g., "fever ≥38.5°C plus ≥3 loose stools in 24 h")

— Person: demographic restrictions (age, sex, occupation) if relevant

— Place: geographic restriction (e.g., "attended the May 12 banquet at Hotel X" or "resident of ward 4B")

— Time: onset within a defined window (e.g., "symptom onset between May 12 and May 20, 2024")

— Confirmed: clinical criteria + definitive lab confirmation (culture, PCR, serology)

— Probable: clinical criteria + supportive but not definitive lab data, or strong epi link

— Suspect: clinical criteria alone, often with looser symptom thresholds

— Early in an investigation → use a sensitive (broad) case definition to capture all possible cases and avoid missing the outbreak's scope

— For hypothesis testing/analytic phase → tighten to a specific definition to reduce misclassification bias that would dilute exposure–disease associations toward the null

— Including exposure in the case definition (circular — you can't then test whether exposure causes disease)

— Using lab confirmation only (misses true cases without testing → underestimates AR)

— Failing to specify time window (allows contamination by background endemic cases)

A case definition is a standardized set of criteria used to decide whether an individual should be classified as having the condition under investigation. It is the analytic backbone of every outbreak investigation.

Four required components (memorize):

Tiers of certainty (CDC convention):

Sensitivity vs specificity tradeoff:

Board pearl: The same investigation often uses two case definitions simultaneously — a sensitive one for case-finding and a specific one for the analytic study. Exam favorite distinction.

Common pitfalls (testable):

Step 3 management: When public health asks the inpatient team to identify additional cases, apply the current case definition consistently to all chart reviews; document each case as confirmed/probable/suspect on the line list.

Confirming the Diagnosis and Building the Line List

— Review clinical records and original lab results (don't trust verbal reports)

— Re-test specimens at a reference lab when feasible

— Visit 1–2 patients personally if possible — confirms clinical syndrome and uncovers exposures missed in charts

— Rule out pseudo-outbreaks from lab contamination, new test introduction, or improved surveillance sensitivity

— One row per case

— Columns include: unique ID, demographics, symptom onset date/time, symptoms, lab results, case classification (confirmed/probable/suspect), exposures, outcome

— Updated continuously; drives both the epi curve and the analytic study

— Stool for bacterial culture, ova/parasites, C. difficile toxin/PCR, norovirus PCR

— Blood cultures if febrile or septic

— Serum acute/convalescent pairs for serology (3–4 weeks apart)

— Food/water samples — work with public health labs; preserve refrigerated, not frozen, when culture intended

— Environmental swabs for nosocomial investigations (sinks, scopes, ventilator circuits)

— PFGE historically, now largely replaced by whole-genome sequencing (WGS) for organism subtyping

— PulseNet (CDC) connects isolates nationally — explains how geographically dispersed cases get linked to one product

Verify the diagnosis before expanding the investigation:

The line list — the operational database of every outbreak:

Specimen handling (high-yield):

Molecular epidemiology:

Board pearl: A single isolate matched by WGS to a national PulseNet cluster transforms a local case into a multistate outbreak and shifts jurisdiction to CDC/FDA — frequently the answer to "what is the next best step?"

CCS pearl: In the simulated environment, order stool studies and blood cultures before empirical antibiotics when feasible; for suspected STEC, avoid antibiotics and antimotility agents (HUS risk).

Key distinction: Confirming the diagnosis (step 3) is about individual cases; confirming the outbreak (step 2) is about whether incidence exceeds baseline. Both must occur — neither substitutes for the other.

The Epidemic Curve — Construction and Interpretation

— X-axis units = ¼ to ⅓ of the suspected incubation period (e.g., for norovirus ~24–48 h incubation, use 6–12 h bins; for hepatitis A ~30 d, use weekly bins)

— Use onset date, not diagnosis or report date

— Stack bars by case classification or by subgroup when relevant

— Mark the probable exposure period and any interventions with vertical reference lines

— Pattern of spread (point-source vs continuous vs propagated vs intermittent)

— Probable time of exposure by counting back one incubation period from the peak (and from the earliest case for the minimum incubation)

— Outliers — early cases may be index/source cases; late cases may be secondary spread

— Effect of interventions — does the curve fall after a control measure?

— Identify median onset date; subtract the median incubation period → estimated exposure date

— Identify earliest onset; subtract minimum incubation → start of exposure window

— Identify latest onset (excluding secondary cases); subtract maximum incubation → end of exposure window

An epidemic curve ("epi curve") is a histogram plotting number of new cases (y-axis) by date or time of symptom onset (x-axis). It is the single most informative graphic in outbreak investigation.

Construction rules:

What the curve tells you (4 inferences):

Estimating exposure window from the curve (high-yield calculation):

Board pearl: If the exam gives you a list of onset dates and asks "when did exposure most likely occur?", compute the median onset and subtract the median incubation period — this is essentially always the intended approach.

Step 3 management: Update the epi curve daily during an active outbreak; a curve that continues to rise after intervention suggests the source is not yet controlled or a secondary mode of transmission exists.

Key distinction: Time on the epi curve = symptom onset, not date of report or date of testing. Using report date flattens and right-shifts the curve, obscuring the source.

Patterns of Epidemic Curves — Point-Source, Continuous, Propagated, Intermittent

— All cases exposed at the same brief time

— Curve: steep upslope, sharp peak, gradual downslope; entire outbreak spans roughly one incubation period

— Classic example: church potluck Salmonella; banquet Staph toxin

— All onsets fall within minimum-to-maximum incubation of exposure

— Source persists, exposing people over an extended period

— Curve: plateau rather than peak; duration exceeds one incubation period

— Example: contaminated municipal water supply; ongoing Listeria from a single processing plant

— Source exposes people in irregular pulses

— Curve: irregular peaks without periodicity; reflects when source was active

— Example: a contaminated ice machine used episodically; intermittent fecal contamination of a well

— Curve: successive peaks roughly one incubation period apart, each peak larger than the last until herd immunity, intervention, or susceptible exhaustion intervenes

— Example: measles, pertussis, COVID-19 household clusters, norovirus on a cruise ship (mixed point-source + propagated)

— Initial point-source followed by propagated person-to-person transmission

— Common with norovirus (food vehicle then secondary spread), shigellosis, hepatitis A

Point-source (common-source, single exposure):

Continuous common-source:

Intermittent common-source:

Propagated (person-to-person):

Mixed:

Board pearl: A propagated curve with peaks ~14 days apart in unvaccinated children + Koplik spots = measles; report immediately, institute airborne precautions, and post-exposure prophylaxis (MMR within 72 h or IG within 6 d) for contacts.

Key distinction: Point-source outbreaks span one incubation period; continuous-source outbreaks span more than one; propagated outbreaks span multiple incubation periods with discrete waves.

Step 3 management: Identifying the curve pattern dictates the control strategy — remove the vehicle (point/continuous), interrupt person-to-person transmission (propagated), or both (mixed).

Analytic Epidemiology — Testing the Hypothesis

— Classic: wedding, banquet, cruise, daycare class

— Calculate attack rates among exposed and unexposed for each item

— Compute relative risk (RR) = AR_exposed ÷ AR_unexposed

— The vehicle typically shows high RR, high AR in exposed, low AR in unexposed, and accounts for most cases

— Compare exposures of cases vs matched controls

— Compute odds ratio (OR) as the measure of association

— Useful for community-wide outbreaks (e.g., multistate produce contamination)

— Chi-square or Fisher's exact for categorical exposures

— 95% CI excluding 1.0 → statistically significant

— Beware multiple comparisons when testing dozens of food items (Bonferroni or focus on biologically plausible items)

— 100 attendees: 60 ate potato salad (45 ill, AR 75%); 40 did not (5 ill, AR 12.5%) → RR = 6.0

— 100 attendees: 50 ate cake (20 ill, AR 40%); 50 did not (30 ill, AR 60%) → RR 0.67 (not the vehicle)

Once descriptive epi suggests a vehicle or mode, analytic studies quantify the association.

Retrospective cohort study — used when the exposed population is well-defined and enumerable:

Case-control study — used when the source population is large or unknown:

Statistical testing:

Worked example (frequent exam format):

Board pearl: The exposure with the highest RR (or OR) that also accounts for the largest fraction of cases is the likely vehicle. A high RR for a rarely consumed item may still miss the true vehicle.

Key distinction: Cohort → relative risk (closed, defined outbreak). Case-control → odds ratio (open, large population). Picking the wrong design wastes time and biases results.

Step 3 management: Even before analytic confirmation, implement reasonable control measures based on descriptive epi if delay would cause harm — public health protection trumps statistical purity.

Special Populations — Elderly, Immunocompromised, and Healthcare-Associated Outbreaks

— High-risk for respiratory (influenza, RSV, COVID-19, Legionella), GI (norovirus, C. difficile), and skin (scabies, MRSA) outbreaks

— CDC LTCF outbreak threshold: influenza-like illness in ≥2 residents within 72 h on the same unit triggers investigation

— Control: cohort ill residents, restrict communal activities, exclude ill staff, antiviral prophylaxis for unaffected residents during influenza outbreaks (oseltamivir for ≥2 weeks and 7 days past last case)

— CLABSI clusters → review insertion practices, chlorhexidine bathing, line maintenance

— Surgical site infection clusters → trace by surgeon, OR, instrument set, scrub tech

— Duodenoscope-related CRE/CRKP outbreaks → review reprocessing

— C. difficile surges → hand hygiene with soap and water (alcohol doesn't kill spores), bleach-based environmental cleaning, antimicrobial stewardship

— Lower symptom threshold for case definition (afebrile cases common)

— Atypical pathogens (CMV, PJP, mold) may cluster around construction or HVAC failures

— Adjust antimicrobial doses (e.g., oseltamivir for CrCl <60); avoid nephrotoxins when possible

— STEC infection in elderly carries higher mortality; supportive care, no antibiotics

Long-term care facilities (LTCFs):

Hospital outbreaks:

Immunocompromised cohorts:

Renal/hepatic considerations in outbreak management:

CCS pearl: When an inpatient on your service develops a new HAI matching a known unit cluster, the immediate orders are: contact + appropriate precautions, notify infection prevention, save specimens, review shared exposures (equipment, staff, room), and document on the unit line list — all before changing antibiotics empirically.

Board pearl: In an LTCF respiratory outbreak, start empirical oseltamivir for all symptomatic residents while awaiting testing, and prophylaxis for asymptomatic contacts — don't wait for confirmation; transmission is too fast.

Special Populations — Pregnancy, Pediatrics, Schools, and Vulnerable Communities

— Listeria monocytogenes: deli meats, soft cheeses, melons, sprouts; incubation up to 70 days; risk of fetal loss, neonatal sepsis; treat with ampicillin

— Hepatitis E in outbreaks (developing-country travel) carries 20% mortality in pregnancy

— Measles, varicella, parvovirus B19, Zika — congenital sequelae; immune-status screening of exposed pregnant patients is mandatory

— Influenza in pregnancy → empirical oseltamivir regardless of trimester

— Norovirus and rotavirus in daycare; hand hygiene, exclusion until 48–72 h asymptomatic

— Measles, mumps, pertussis in undervaccinated communities — propagated curves

— Hand-foot-mouth (coxsackie A) — supportive only; school exclusion criteria vary

— Head lice and scabies — direct contact; permethrin; treat household

— Notify school health and local health department for vaccine-preventable diseases

— Measles: MMR within 72 h; IG within 6 d for high-risk (pregnant, immunocompromised, <12 mo)

— Varicella: VZIG within 10 d for high-risk exposed

— Hepatitis A: vaccine ± IG depending on age and risk, within 14 d

— Pertussis: azithromycin for close contacts regardless of vaccination

— Meningococcal: rifampin, ciprofloxacin, or ceftriaxone for close contacts

— Homeless shelters → TB, hepatitis A, shigellosis outbreaks; vaccination campaigns

— Correctional facilities → respiratory and bloodborne pathogens

— Refugee/migrant populations → screen per CDC guidelines

Pregnancy-specific outbreak considerations:

Pediatric/school outbreaks:

Post-exposure prophylaxis (PEP) — high yield:

Vulnerable populations:

Key distinction: Isolation = separating those who are ill; quarantine = restricting those exposed but not yet ill. Step 3 stems test this terminology.

Board pearl: A pregnant patient exposed to varicella with negative IgG → administer VZIG (or IVIG if VZIG unavailable) within 10 days. Live MMR/varicella vaccines are contraindicated in pregnancy.

Complications and Adverse Outcomes of Outbreaks

— STEC O157:H7 → HUS (5–10% of cases), especially children; avoid antibiotics and antimotility agents

— Salmonella → bacteremia, endovascular seeding, reactive arthritis

— Shigella → seizures in young children, HUS (S. dysenteriae type 1)

— Campylobacter → Guillain-Barré syndrome (1 in 1000)

— Measles → pneumonia, encephalitis, SSPE years later

— Influenza → bacterial superinfection (S. aureus, S. pneumoniae), ARDS

— Norovirus → dehydration, AKI in elderly

— Listeria → meningoencephalitis, fetal loss

— Surge demand on EDs and ICUs

— Staff illness depleting workforce (healthcare worker absenteeism)

— Supply chain shortages (PPE, antivirals, vaccines)

— Misinformation and erosion of public trust

— Economic loss to implicated industries (the "spinach scare" effect)

— Premature closure of investigation before source eliminated → continued exposure

— Incorrect implication of a food/product → economic harm and missed real source

— Recall bias in case-control studies inflating spurious associations

— Failure to detect secondary cases through person-to-person spread

Direct medical complications depend on the pathogen:

Public health and system complications:

Investigation pitfalls that produce adverse outcomes:

Step 3 management: For confirmed STEC infection in an outbreak, hydrate aggressively with isotonic IV fluids, monitor CBC, creatinine, LDH, haptoglobin, and peripheral smear for HUS through day 7–10; do not give antibiotics, loperamide, or NSAIDs.

Board pearl: New-onset ascending weakness 1–3 weeks after a Campylobacter outbreak gastroenteritis → Guillain-Barré. Diagnose with LP (albuminocytologic dissociation) and NCS; treat with IVIG or plasma exchange; monitor FVC for respiratory failure.

Key distinction: HUS from STEC = non-immune microangiopathic hemolytic anemia + thrombocytopenia + AKI; TTP shares triad plus fever and neuro changes, with ADAMTS13 <10%. Treatments diverge sharply (supportive vs PLEX).

Escalation, Reporting, and Public Health Coordination

— Each state maintains a list of notifiable conditions; nationally reportable conditions are reported via CDC's NNDSS

— Immediate (phone) reporting required for: measles, rubella, pertussis outbreaks, meningococcal disease, anthrax, botulism, plague, smallpox, viral hemorrhagic fevers, novel influenza A, SARS-CoV/MERS-CoV/SARS-CoV-2 clusters, foodborne outbreaks, hospital outbreaks of unusual pathogens

— Routine (written/electronic) reporting within days for: hepatitis A/B/C, HIV, TB, salmonellosis, shigellosis, syphilis, gonorrhea, chlamydia, Lyme disease

— Clinician → local/county health department → state health department → CDC

— Foodborne with implicated product → FDA (most foods) or USDA-FSIS (meat, poultry, egg products)

— Waterborne → EPA collaboration

— Bioterrorism suspicion → FBI + CDC's Emergency Operations Center, 770-488-7100

— Cluster of unexplained infections on a single unit → activate infection prevention, hospital epidemiology, and outbreak response team

— Pathogen of public health concern (e.g., Candida auris, CRE, novel respiratory virus) → notify state lab for confirmatory testing

— Surge exceeding capacity → activate hospital incident command system (HICS)

Mandatory reporting (US):

Chain of escalation:

Hospital-level escalation triggers:

CCS pearl: In the simulated case, "call public health department" is a billable action — invoke it early for any cluster of reportable disease or any unusual organism. Missing the call costs points just like missing a stat troponin.

Board pearl: HIPAA permits disclosure of protected health information to public health authorities without patient authorization for disease surveillance, investigation, and intervention. The "minimum necessary" standard still applies, but consent is not required.

Step 3 management: Document the date/time of public health notification, the recipient, and the information shared; this becomes part of the legal record of the outbreak response.

Differential Diagnosis — Distinguishing True Outbreaks from Look-Alikes (Same Category)

— A new laboratory test (more sensitive PCR) → apparent increase in disease that is actually improved detection

— Change in case definition broadening criteria → more "cases" without true incidence change

— New reporting requirement or active case-finding → uncovers previously missed endemic cases

— Contaminated reagent, water bath, or bronchoscope yielding spurious positive cultures

— Clue: all "cases" lack clinical illness; isolates are identical by molecular typing; cluster confined to one lab

— Influenza, RSV, norovirus, rotavirus have predictable seasonality

— Compare current incidence to same period in prior years, not to summer baseline

— Long-term increases (e.g., antimicrobial resistance trends, opioid overdose deaths) may exceed historical baseline without representing a discrete outbreak

— Cases reported in a single batch after a holiday may falsely appear as a temporal cluster

— Influx of a susceptible population (refugees, students) raises absolute case counts without changing per capita risk

Before declaring an outbreak, exclude other explanations for an apparent increase in cases. Same-category mimics (epidemiologic phenomena):

Pseudo-outbreak from surveillance artifact:

Pseudo-outbreak from laboratory contamination:

Seasonal or cyclical increase:

Secular trends:

Reporting delays and batch reporting:

Demographic shift:

Board pearl: Always ask "compared to what?" Establishing the expected baseline (mean ± 2 SD over the prior 3–5 comparable periods) is step 2 of outbreak investigation. A "spike" that lies within historical variability is not an outbreak.

Key distinction: Pseudo-outbreak = apparent increase from artifact (lab, surveillance, reporting); true outbreak = real excess of disease. Mismanaging a pseudo-outbreak wastes resources and damages credibility.

Step 3 management: When suspicion of pseudo-outbreak arises (e.g., clustered positive M. tuberculosis cultures without clinical TB), review lab logs, sequence isolates, and audit specimen handling before initiating mass treatment or contact tracing.

Differential Diagnosis — Outbreak vs Other Public Health Phenomena (Other Categories)

— Unusual pathogen for the geography (anthrax in an urban office worker)

— Single source affecting people with diverse demographics across a wide area

— Atypical clinical presentation (inhalational anthrax instead of cutaneous)

— Animal die-off preceding human cases

— Multiple simultaneous outbreaks of different agents

— Cluster among first responders or postal workers

— Carbon monoxide cluster in a building → multiple patients with headache, nausea after furnace failure; mimics viral illness

— Lead poisoning in a neighborhood near a smelter; mimics nonspecific GI/neuro complaints

— Chemical contamination of water (PFAS, nitrates) producing chronic effects

— Cluster of symptoms (dizziness, nausea, fainting) without identifiable biological or chemical agent; rapid onset and resolution; often in schools or workplaces; diagnosis of exclusion after thorough environmental and biological evaluation

Endemic disease: baseline level present continuously in a population (malaria in sub-Saharan Africa; chickenpox in pre-vaccine era US). Not an outbreak unless incidence rises above expected.

Hyperendemic: persistently high incidence beyond what is typical for other regions; still constant rather than excess.

Pandemic: outbreak crossing international boundaries with global spread (1918 H1N1, 2009 H1N1, 2019 SARS-CoV-2). Same investigative principles, vastly larger scale and political dimension.

Sporadic disease: isolated, irregular cases without clear temporal or geographic clustering. A single rabies case is sporadic; multiple linked cases would be an outbreak.

Bioterrorism vs natural outbreak (testable red flags for intentional release):

Environmental/toxic exposure mimics:

Mass psychogenic illness:

Board pearl: Inhalational anthrax in a previously healthy adult mail handler with mediastinal widening on CXR = bioterrorism until proven otherwise. Notify FBI and CDC immediately; start ciprofloxacin or doxycycline plus one or two additional agents.

Key distinction: Pandemic = geographic spread; outbreak = excess incidence. A disease can be pandemic and not an outbreak in a given locale (if local incidence is at expected pandemic baseline).

Control Measures and Secondary Prevention

— Recall contaminated food product; close implicated restaurant pending remediation

— Disinfect water supply; boil-water advisory; switch source

— Repair HVAC, replace cooling tower water for Legionella

— Remove and reprocess contaminated medical equipment

— Hand hygiene reinforcement; PPE; isolation/quarantine

— Environmental cleaning (bleach for C. difficile and norovirus; alcohol effective for most enveloped viruses)

— Cohorting ill patients and dedicated staff

— Travel restrictions or venue closures when proportionate

— Vaccination campaigns (MMR catch-up, hepatitis A, meningococcal)

— Chemoprophylaxis (oseltamivir, rifampin, doxycycline depending on agent)

— Education on hand hygiene, safe food handling, sexual practices, vector avoidance

— Food safety legislation, HACCP (Hazard Analysis Critical Control Points) audits

— Improved surveillance systems (syndromic surveillance, ESSENCE, BioSense)

— Vaccination program strengthening (close immunization gaps revealed by the outbreak)

— Antimicrobial stewardship to prevent C. difficile recurrence

— Engineering controls (negative-pressure rooms, water system design)

Control measures should target the weakest link in the chain of transmission: agent, reservoir, mode of transmission, portal of entry, susceptible host.

Targeting the agent or source:

Targeting transmission:

Targeting the host:

Long-term/secondary prevention:

CCS pearl: When you discharge a patient who was part of a foodborne outbreak, counsel on hand hygiene, exclude from food handling and healthcare work until symptom-free for 48 h (longer for specific pathogens per CDC: 24 h for norovirus is insufficient — use 48 h or longer per local rules), and provide written return precautions.

Board pearl: Food handlers with typhoid (Salmonella Typhi), STEC O157, Shigella, Hepatitis A, or norovirus require documented clearance (typically negative stool cultures or specified asymptomatic period) before returning to work — single negative isn't always enough for Salmonella Typhi.

Follow-Up, Monitoring, and Final Report

— Continue active surveillance for ≥2 maximum incubation periods after the last case to declare the outbreak over

— Monitor for secondary spread and recurrence

— Track the epi curve daily; falling curve after intervention supports control measure efficacy

— Audit hand hygiene compliance, vaccination uptake, environmental cleaning logs

— Re-culture environmental sources to confirm remediation

— STEC patients → CBC, creatinine, urinalysis through day 7–10 for HUS

— Hepatitis A contacts → LFTs and clinical follow-up at 2 and 6 weeks

— Measles contacts → 21-day quarantine for unvaccinated; monitor for fever/rash

— TB exposure → TST or IGRA at baseline and 8–10 weeks

— Bloodborne exposure (needlestick during outbreak response) → HIV, HBV, HCV testing per OSHA timelines

— Background, methods, descriptive epi (with final epi curve and spot map), analytic findings, environmental investigation, control measures, recommendations, references

— Distribute to participating agencies, affected facilities, and (often) publish for broader learning (MMWR)

— What worked, what failed, gaps in surveillance, communication breakdowns, resource shortfalls

— Generate corrective action plan with assigned owners and deadlines

Ongoing case-finding:

Monitoring control measures:

Patient-level follow-up:

Final report (CDC convention) includes:

System-level after-action review:

Board pearl: Declaring an outbreak "over" requires no new cases for ≥2 maximum incubation periods. For measles (incubation up to 21 d), that's 42 days; for Ebola, 42 days as well (2 × 21 d).

Step 3 management: During follow-up of an outbreak patient, document return-to-work/return-to-school clearance, vaccination updates triggered by the exposure, and any chronic sequelae (HUS-related CKD, GBS-related neuropathy) requiring long-term monitoring.

Key distinction: Process measures (hand hygiene rates, vaccination coverage) vs outcome measures (new case counts). A successful response shows improvement in both, but outcome measures are definitive.

Ethical, Legal, and Patient Safety Considerations

— HIPAA explicitly permits disclosure to public health authorities (45 CFR 164.512(b))

— Physicians who fail to report a notifiable disease may face licensure action and civil liability

— Patients cannot opt out of reportable disease notification; counsel them about the process up front

— States hold primary police-power authority to impose isolation/quarantine; federal authority (CDC under 42 USC 264) covers interstate and international situations

— Must use least restrictive means consistent with public health goals

— Subjects retain rights to due process, including review by a neutral decision-maker if detention is prolonged

— Investigational vaccines/therapeutics deployed under EUA still require disclosure of investigational status, alternatives, risks/benefits, and the right to refuse

— Research conducted during outbreaks (e.g., serosurveys) requires IRB review even when expedited

— Children and decisionally impaired adults require surrogate consent; assent from older minors when feasible

— Avoid naming outbreaks by ethnicity, nationality, or geography (per WHO 2015 guidance)

— Ensure vulnerable populations (homeless, undocumented, incarcerated) receive equitable access to testing, treatment, and PEP without immigration-status repercussions

— When transferring a patient who is part of an outbreak (e.g., LTCF to hospital, hospital to home), communicate outbreak status, isolation requirements, and pending cultures in the handoff

— Failure to communicate active outbreak exposure on transfer is a sentinel event in many systems

— Investigators must disclose ties to implicated industries; recusal may be necessary

Mandatory reporting vs confidentiality:

Isolation and quarantine — legal authority:

Informed consent edge cases:

Equity and stigma:

Patient safety/transitions of care:

Conflicts of interest:

Board pearl: A patient with active pulmonary TB refusing treatment can be legally detained under state public health statutes when they pose a transmission risk and less restrictive measures (DOT, education) have failed — this is the most-tested ethics scenario in outbreak epidemiology.

Step 3 management: Document the public health rationale, least-restrictive-means analysis, and patient counseling in any case involving compelled isolation/quarantine.

High-Yield Associations and Rapid-Fire Facts

— Single sharp peak, all onsets within one incubation → point-source

— Plateau → continuous common-source

— Successive peaks one incubation apart → propagated

— 1–6 h vomiting → S. aureus, B. cereus emetic toxin

— 8–16 h diarrhea → C. perfringens

— 12–72 h → most bacterial gastroenteritis

— 15–50 d hepatitis → HAV

— Weeks to months → Listeria, TB

— Fried rice → B. cereus

— Reheated meat/gravy → C. perfringens

— Eggs/poultry → Salmonella

— Undercooked ground beef → STEC, Campylobacter

— Raw shellfish → Vibrio spp., norovirus, HAV

— Unpasteurized dairy → Listeria, Brucella, Campylobacter

— Deli meats, soft cheese, melons → Listeria

— Sprouts → Salmonella, STEC

— Hotel/cruise water systems → Legionella

— Attack rate = ill ÷ at risk

— Relative risk = AR_exposed ÷ AR_unexposed

— Odds ratio = (a×d) ÷ (b×c)

— Sensitivity vs specificity in case definitions — broad early, narrow for analytics

— Local health department first; CDC EOC 770-488-7100 for emergencies

— FoodNet, PulseNet, NORS, NHSN for specific surveillance streams

Epi curve patterns → suspected etiology:

Incubation period clues:

Vehicle clues:

Calculations to memorize:

Reporting hotlines/agencies:

Board pearl: Norovirus is the most common cause of foodborne illness outbreaks in the US and the leading cause of healthcare-facility GI outbreaks. Hand hygiene with soap and water beats alcohol gel for both norovirus and C. difficile spores.

Key distinction: Incidence (new cases per population per time) is what outbreaks measure; prevalence (existing cases) is not the right metric for acute outbreak surveillance.

Rapid fact: The Snow cholera investigation (1854) is the prototype natural-experiment cohort study and the conceptual ancestor of every spot-map analysis.

Board Question Stem Patterns

— "30 wedding guests develop diarrhea, with onsets clustering 36 h after the reception. Which pathogen is most likely?" → median incubation 1–3 d points to Salmonella, Shigella, or norovirus; correlate with food vehicle.

— Histogram shows one sharp peak spanning ~24 h → point-source; multiple peaks ~14 days apart → propagated (think measles).

— Suspected outbreak in nursing home → notify local health department and infection prevention, not "order MRI" or "transfer all patients."

— Early in investigation, choose the broader/more sensitive definition; for analytic study, choose the specific definition. Watch for choices that build exposure into the case definition (wrong — circular).

— Given a 2×2 table per food item, calculate RR; the item with the highest RR (and broad consumption among cases) is the vehicle.

— Measles exposure of unvaccinated child <72 h → MMR; pregnant exposed → IG; pertussis household contacts → azithromycin; meningococcal close contacts → ciprofloxacin (adults) or rifampin.

— Foodborne cluster, suspected bioterrorism, novel respiratory virus → immediate phone report; HIPAA permits without consent.

— Bloody diarrhea after petting zoo → do not prescribe antibiotics or loperamide; hydrate, monitor for HUS.

— Cluster of positive AFB cultures with no clinical disease → suspect lab contamination, sequence isolates before treating.

— Inhalational anthrax in mail handler → notify FBI/CDC; ciprofloxacin or doxycycline + adjuncts.

Stem 1 — Compute the exposure window:

Stem 2 — Identify epi curve pattern:

Stem 3 — Choose the next best step:

Stem 4 — Case definition design:

Stem 5 — Vehicle identification by attack rate:

Stem 6 — Post-exposure prophylaxis:

Stem 7 — Reporting obligations:

Stem 8 — STEC management trap:

Stem 9 — Pseudo-outbreak:

Stem 10 — Bioterrorism red flag:

Board pearl: When the question asks for the single best next step in a possible outbreak, the answer is almost always a public health action (notify, isolate, line list, case definition) rather than an advanced diagnostic test. Step 3 rewards systems thinking.

Key distinction: Watch for stems that test terminology precision (isolation vs quarantine; outbreak vs cluster; incidence vs prevalence; RR vs OR). Lexical accuracy wins the point.

One-Line Recap

Outbreak investigation in one sentence: Confirm an outbreak by demonstrating incidence above expected baseline, verify the diagnosis, build a sensitive case definition anchored in person/place/time/clinical criteria, plot the epidemic curve by symptom onset to infer point-source vs continuous vs propagated transmission, generate hypotheses from descriptive epi, test them with a cohort (RR) or case-control (OR) study, implement control measures targeting agent–transmission–host, and report to public health throughout.

— Case definition = clinical criteria + person + place + time, stratified into confirmed/probable/suspect; broad for case-finding, narrow for analysis

— Epi curve x-axis = symptom onset date in bins of ¼–⅓ the incubation period; pattern reveals transmission mode and exposure window (subtract median incubation from median onset)

— Point-source = one sharp peak within one incubation; continuous = plateau; propagated = serial peaks one incubation apart

— Cohort study → relative risk (closed outbreak); case-control → odds ratio (community/multistate outbreak); the vehicle has highest RR/OR and accounts for most cases

— Step 3 reflex: suspected outbreak → notify local health department, start line list, institute isolation/precautions, preserve specimens — before advanced workup

— Don't miss: STEC → no antibiotics (HUS risk); measles/varicella PEP within 72 h–10 d; food handlers with Typhi/STEC/Shigella/HAV/norovirus require documented clearance; HIPAA permits public health reporting without consent

Highest-yield recap bullets:

Board pearl: The exam's favorite "next best step" answer in any outbreak vignette is the public health action (report, define, list, curve, isolate) — not the imaging study or the empirical antibiotic.