Surveillance

Who decides what diseases SHOULD be required notification?

CSTE. Council of State and Territorial Epidemiologists.

Flexibility

Can adapt to changing needs or operating conditions

What are the limitations of using aggregate data?

Cant report a mean, can't know specific ages or make new breakdowns. Can't provide a case-based analysis. May also be prone to ecologic fallacy (applying group level findings to individuals)

How is surveillance process different from developing an epi research study?

Case definition developed, as opposed to enrollment criteria Develop analytic approach instead of analysis plan (ongoing basis for surveillance, vs. one or two times for epi study) Ensuring use of the analysis and its interpretation

Data in developing countries sources

Case reports, sentinel sites, exit interviews, survey, focus groups, vital event registration, verbal autopsy

Things to consider when interpreting surveillance graphs

Change in definition or truly an epidemic? This is only Reported cases Natural history isn't applicable if there is an active vaccine, treatment. Are they testing in all areas? Why? Resistance? New Strain?

Describe some of the duties of registry workers

Classification, coding, state data exchange, death certificates, auditing, quality control, rapid case ascertainment

Clinically compatible case

Clinical syndrome generally compatible with disease, but no specific clinical criteria need to be met (unless they are noted in the case classification)

Quality

Completeness and validity

Overall approaches to HAI surveillance (3), with 2 directions

Concurrent or retrospective: 1. Hospital-wide 2. Targeted 3. Objective Based

3 Classifications of disease cases

Confirmed - confirmed for REPORTING Probable - classified as probable for reporting purposes Suspect - classified as suspected for reporting

Lab-confirmed case

Confirmed by one or more laboratory methods listed in the case definition under lab criteria for diagnosis. -other lab methods can be used in clinical diagnosis, but for reporting purposes must use lab methods given in case definition

What is an example of a highly sensitive surveillance system?

Congenital syphillis

Whats a problem with using illinois death registry as an endpoint for illinois diagnoses of cancer?

Could have died from other causes, but more importantly MANY of these patients will die in states outside of Illinois.

What are the 3 components of the process of cancer registry?

Data collection. Data quality control. Data dissemination

Challenges of surveillance in developing countries

Denominator difficult to establish, electricity, computer, analytic, dissemination/feedback, infrastructure

PVP

(Predictive Value Positive) the proportion of individuals who are identified by system AND truly have the disease

Chart method for determining if number of cases is an aberration

-Create grid of Months by years with case # -Look at the preceding, current, and following month for the past 5 years -Add up ALL those cases, and divide by the number of cells -That is your expected value -Then we use a computer program to see if statistically significant -But on test we can just make a judgement call

Types of surveillance reporting

-Individual case-based reports that include info on each person -Aggregated data on total number of persons (often stratified by gender or age) -Total number of cases

Establishing objectives for disease surveillance: what do you want to know?

-burden of disease -disease characteristics (drug susceptibility, clinical manifestations, incubation period -spatial distribution of disease -risk factors for disease or death -"outcomes", morbidity, hospitalization, mortality, severity -effectiveness of policy and/or programs

Name 4 time intervals of interest when reporting, analyzing surveillance data

-exposure to onset (incubation) -onset to diagnosis (delay to diagnosis) -diagnosis to reporting (reporting delays) -duration of hospitalization (severity)

Briefly describe the historical origins of surveillance

-quarantining sick passengers on ships -keeping vitals records -data on bubonic plague with weekly reports -census

What are the 6 issues that were identified in switching from case-based to hospital discharge in the pertussis database?

1. Can't trace contacts 2. Incubation period requires timeliness 3. Missing those who aren't hospitalized 4. Only certain hospitals 5. Missing key variables 6. No lab data 7. Who has access to this hospital? Can't calculate a rate

What are the 3 approaches to collecting cases for a cancer registry? Give short definition

1. Cohort - tracks everyone (expensive) 2. Registry - Registers new cases and tracks them for survival 3. Sampling - periodically surveys select populations

What are the Cancer Registry standards?

1. Completeness (all cases included) 2. Quality (information true & accurate) 3. Timeliness (Data ready to disseminate) 4. Confidentiality (patient privacy protected) 5. Data utilization

What are the uses of cancer registry data?

1. Describe burden, trend, distribution 2. Design & manage prevention / control 3. Develop research

Overall reasons (8) for conducting HAI surveillance

1. Establish baseline rates/distribution 2. Monitor for changes (subpopulations, sudden v gradual) 3. Generate hypotheses about risk factors 4. Institute and evaluate prevention & control meausres 5. Monitor compliance with institution policies 6. Monitor compliance with legislation 7. Guide treatment 8. Compare to other hospitals

What are the 4 steps of planning healthcare facility surveillance?

1. Goals & objectives 2. Case definitions 3. Resources/funding (including staff and data management) 4. Buy-in from staff

What are three types of cancer registries?

1. Hospital-based 2. Special-purpose (specific cancers) 3. Population-based (by geographic areas or populations, i.e. STATE registries)

Overall steps of Evaluation (5)

1. Purpose & objectives 2. Description of its operation 3. Documentation of how useful 4. Attributes 5. Estimate cost of system

What 2 things should you consider when deciding what to perform targeted surveillance on?

1. Significance of the event (severity, cost, spread, availability of prevention) 2. Institutional needs & characteristics

What are the ATTRIBUTES when evaluating a surveillance system? (9)

1. Simplicity 5. Sensitivity 6. Reliability 2. Flexibility 4. Acceptability 3. Quality 7. Representativeness 8. Timeliness 9. PVP SSR, FAQ, RTP

What are the steps when planning a surveillance system in a low resource or emergency setting?

1. determine health risks 2. prioritize health risks 3. determine effective interventions 4. Determine resources needed 5. Develop plans to prevent, control, treat 6. Develop a SYSTEM to do the above

3 things to consider when choosing appropriate analytic method

1. purpose of system 2. data structure 3. how is baseline determined?

What are some of the issues with hospital-based registries?

1. the cachment population is not defined 2. the incidence rate is not calculable 3. standardization varies 4. used for patient care not necessarily surveillance

Findings: Food poisoning study

193 report forms were filled out, after CDPH responded to tweets containing food poisoning term. 133 restaurants were inspected, a bunch either failed or passed with critical violations. (15% failed, 23% passed with serious/critical violations). With limitations, this appears to be a successful endeavor that did not overburden food inspectors in Chicago.

Class of diseases in Communicable Diseases Code

1a-immediate reporting (botulism) 1b-within 24 business hours (meningitis) 1c-within 7 days (salmonella)

Sensitivity

Ability to detect epidemics, proportion of true cases that are detected by the surveillance system

Reliability

Ability to do its job without failure (outages.. etc).

What are usual rationales for disease becoming notifiable?

Baseline incidence of disease is not well known Action needed in response and prompt notification would be beneficial Epi and natural history is not well described

Symptoms to collect during emergency

Bloody diarrhea Water diarrhea Suspect cholera Lower respiratory tract infection Measles meningitis

THM regarding worldwide surveillance in general (from pertussis lecture)?

Disparate, discrepant, not comparable, not reporting same things, not using same case definition

Describe the FRAMEWORK for evaluation process (not what the evaluation consists of)

Engage Stakeholders Describe Surveillance System (importance, use, resources) Plan Evaluation Gather evidence of attributes state, justify conclusions make recommendations Ensure use of findings

Why is evaluating a surveillance system important?

Ensure monitored efficiently and effectively Make recs for improving quality and usefulness Increase UTILITY and EFFICIENCY

What are the (many) uses of public health surveillance? Try to memorize at least 5

Estimate magnitude of health problems Explore the natural history of diseases Detect epidemics Document spread of disease (geographic... ) Identify research needs, test hypotheses Evaluate how well prevention or control (need info on program in graph) Monitor changes in infectious agents (% resistant...) Monitor isolation activities (must have patient isolation info on graph) Detect changes in health practice (must have info on graph) Facilitate planning

What do we consider when determining if a disease should be surveilled? (6)

Frequency Severity Cost Preventability / Actionable Communicability Public interest

How do you decide if a disease is of enough importance to surveil?

Frequency (incidence, prevalence) Cost of disease Communicability Morbidity, mortality, hosp, life lost Public Interest (approved by legislature) Preventability / Actionable (prophylaxis, isolation, quarantine, recall, antitoxin, mass immunization, terminal cleaning)

What are 6 things that are needed to keep the cancer registry going?

Funding Training Standards Experienced work force Connection with users /action Laws & rules

What are the advantages/disadvantages of objective based surveillance? (created with an intervention in mind)

Generally created with an intervention in mind Good: adaptable, management, focus on problems of your institution Bad: miss things in non-targeted areas, can't compare, may not have baseline rates

Discuss advantages / disadvantages of hospital wide surveillance

Good: coverage, quick identification, analyze subgroups and hospital units Bad: resource intensive, difficult to prioritize, lot of data

Discuss advantages / disadvantages of targeted surveillance

Good: flexible, less resource intensive, can focus on greatest risk Bad: May miss things in non-targeted area, can't compare to other hospitals, may not have a baseline rate

Surveillance during an emergency, what is most important?

High disease impact? Significant epidemic potential? Specific target of a control program? Lead to cost-effective public health action?

How does WHO recs for surveillance change based on country characteristics?

High vaccination should be case-based Low vaccination countries can do aggregate

Why might you want a really sensitive case definition ?

Highly communicable, spreadable disease

Who has the legal authority to require communicable disease reporting?

IDPH

I-NEDDS

Illinois national electronic disease surveillance system, where confirmed cases of reportable diseases are entered

Capture recapture assumptions

Individuals captured by both sources can be matched Likelihood of capture is independent by source Same study period and area are examined

Data smoothing (also called a moving average)

Instead of reporting monthly, take 5 month block averages (not mutually exclusive blocks, moving one month at a time) to more clearly define a trend over extended time

Representativeness

Is it getting everyone equally ? Systematically missing people?

Premise: Leprosy study

Leprosy is endemic in the Amazon. Give information about the surveillance system to WHO to increase comparability between nations. Conducted evaluation.

What are laboratory reportable diseases? Why are they this way?

MUST HAVE specimens!! Sent to the state public health laboratory. -increased lab technologies, freeze for future use, molecular fingerprint, outbreak prone

Calculate total N using capture recapture

N1*N2/x11

Definition of HAI

No present OR Detected on admission - must be on or after the 3rd day of admission

Are case definitions used to initiate treatment or public health actions?

No. These are REPORTING definitions, not clinical diagnoses. Do not need to use reporting case definitions to initiate treatment, public health actions, etc.

Steps for planning a surveillance system

Objectives Case definition Data source Collection mechanism Collection instruments Test methods Develop and test analytic approach Develop dissemination mechanism Ensure use of analysis and its interpretation Evaluation

Surveillance data limitations

Observational in nature not from a randomized study can't be used for formal CAUSAL purposes Problems with underreporting Some reportable diseases have many steps or stages of disease or reporting that hinder reporting Underrepresentiveness of reported cases -higher reporting in public health facilities -hospitalization more likely than outpatient cases -state and county differences

How do we define public health surveillance? (long and short)

Ongoing systematic collection and analysis of health related data, along with dissemination of these data to those responsible for prevention. Essential definition: Information for Action

Premise: Cholera study

Outbreaks of cholera have occurred for many decades. These usually are associated with rain, flooding, disrupted sanitation. The government has been implementing prevention and education measures. There are several very distinct geoclimate regions that have different rates of cholera. The purpose of this study was to evaluate the surveillance system to assess incidence, mortality, and to see if there was a seasonality pattern (predicted due to rain).

Active vs. Passive surveillance

Passive-information flows to you Active-out asking questions and seeking information

Epidemiologically linked case

Patient has had contact with one or more persons with disease/point source of infection Transmission is plausible Sometimes at least one contact in chain of transmission needs to be laboratory confirmed

What can cause a decrease in timeliness?

Patients recognition of disease Reliance on confirmatory lab testing Provider awareness of surveillance Provider knowledge of idsease Provider acceptance of surveillance Mode of data transfer

Digit attraction

People tend to report onset of disease on the first or the fifteenth of the month.... This contributes strongly to how the data is distributed

How does low specificity affect reporting?

Persons with other diseases may be included

What benefit does using probable or suspect cases give?

Probable - more sensitive, less specific

What actions can be taken in response to pertussis surveillance?

Prophylaxis to close contacts Vaccine Early treatment Outbreak detection

Acceptibility

Provider buy-in

Distinction between quarantine and isolation

Quarantine is for well people who may or may not develop disease. Isolation is for PATIENTS of disease. Cases.

Why might you want a really specific case definition?

RESOURCES, impact on patients, could possibly INFECT people who are well. Because actions have consequences, isolation, treatment etc. are disruptive and possibly expensive?

What is an example of a highly specific surveillance system?

Rabies

validity

Refers to whether the condition as reported reflects the true condition as it occurs

What are the two national cancer registries? Who funds them? Which is more complete state-wise?

SEER-NCI (surveillance epi and end results)-only 20 states. NPCR - CDC (national program of cancer registries). This one is more complete.

What is targeted surveillance?

Specifies either a: Body site Time-limited period Unit or ward of the hospital

Premise: Twitter Food Poisoning Study

Technology (social media) may allow the government to better engage with the public to track food poisoning and inspect restaurants to prevent this.

What government body has the authority to require notification of cases?

The State Legislature

Timeliness

The delay between any of the steps in surveillance OR time required to detect epidemic

Findings: Cholera study

There was an overall decline in cases. There was no significant seasonality detected, which was surprising. Males were more affected (occupational risk). The authors recommend continued and strengthened cholera surveillance in Uganda. Most important: sanitation, water control during flooding, fresh water supply, latrines, hygeine, refugee camps, water crossings, rivers, etc.

How does low sensitivity affect reporting?

Underreporting

capture-recapture

Used for counting total # of cases in a population using one or more incomplete lists Cases can be matched by unique identifiers Estimates the # of cases that are not captured by any source Another source could be hospital discharge data, etc.

Findings: Leprosy study:

Useful, most of the attributes were well-satisfied. Sensitivity and representativeness were difficult to assess. The main finding was that only 46% of household contacts were investigated, so their recommendations centered on strategies and trainings for workers for this purpose. Additionally, it was recommended to update those who abandoned treatment as defaulters in the system.

Reliability

Whether a particular condition is reported consistently by different observers

Can case definition change over time in different conditions?

Yes, outbreak vs. non-epidemic periods may affect criteria. Activism, technological advances/medical understanding, etc

Calculating PVP, PPV

add image

Who has the responsibility to report reportable diseases?

almost EVERYONE you can think of

Scan Statistic

determines whether number of cases is aberrant for a certain time period

Simplicity

ease of operation, amount/type of data, integration with other systems, etc.

Farr

father of vital statistics

Minimum data in an emergency

mortality=most important Morbidiy, popularion figures, nutrition, basic needs, vaccination

Name several general types of surveillance systems

notifiable disease reporting, vital stats, sentinel surveillance, registries, surveys, insurance data

PHIN

public health information network - standardized codes, ICD-9 etc

Why might pertussis be on the rise again?

waning immunity mutations ACELLULAR vaccine enhanced reporting