OSU Epi: Analysis Overview, Secondary Data, and Secondary Data

what are things that make a research question good?

• specific

• measurable

• feasible

describe research question flexibility

• research question can be adjusted only if

  • depends on data you have access to

  • measurements you assessed

  • sample you have access to

what are steps for analysis

• data cleaning

• descriptive analysis

• analytical analysis

• subgroup analysis

• sensitivity analysis

differentiate between primary data and secondary data

• primary

  • data that you collect

  • primary purpose is research

• secondary

  • data that is collected by someone else

can datasets be linked

• yes, if there is an adequate linking variable

What is surveillance

• the process used to collect, manage, analyze, interpret, and report information on the status of diseases, conditions and populations

what is the goal of surveillance

• describe where health problems are occuring and who is affected

• monitor occurence of a disease over time within a population

• gather data which can help predict changes in occurance of disease over time

• lay the groundwork for possible future causal studies

what are some examples of conditions that are monitored

• exposures and outcomes

  • infectious diseases

  • reproductive health

  • injuries

  • environmental hazards

  • health risk behaviors

  • cancer

describe the design of a surveillance system

• can either…

  • collect all cases

  • collect representative sample cases

what is the denominator in target population surveillance?

represents the total population at risk of developing the disease

what are the 3 things all surveillance systems should have

• denominator of target at risk population

• defined geographic area

• defined time period

what is the case definition

• the criteria for an event to be included in the surveillance system.

describe the ethics of confidentiality

• tension between confidentiality vs protecting personal and population health

what is passive surveillance

• routine notifiable diseases from labs, hospitals, healthcare providers without actively searching for them

describe active surveillance

• local or state health departments or researchers initiate the collection of specific cases of disease from laboratories, physicians, hospitals, clinics, or healthcare providers

describe syndromic surveillance

• NOT based on clinically confirmed cases

• collected health-related data on symptoms and clinical signs

  • identifies potential outbreaks before they happen

describe lab-based surveillance

• uses diagnostic laboratories as basis for real-time surveillance

  • cons

    • may not be representative of target population

    • may not provide enough epidemiologically useful information

what is registry based surveillance

• listings of all occurrences of a disease in a defined area

  • focus on in-depth monitoring of specific populations and diseases

what is survey based surveillance

• ongoing surveys

  • periodic and representative

  • ex: NHANES

describe electronic health records surveillance

• collection of patient data from digital medical records for public health monitoring

  • diagnosis and procedures

describe record linkage studies

• combines data from multiple sources

  • ex: link cohort with cancer registries

describe smoothing and standardizing in descriptive epidemiology

• standardizing

  • makes comparisons between populations more accurate by adjusting for potential confounders

• smoothing

  • addresses instability of estimates in small areas wtih sparse data

what are some challenges in surveillance data

• lag between diagnosis and reporting

• difference between place of reporting and residence

what are biases in surveillance data

• confounding

• missing data

  • limitations in fully describing data

• selection bias

• information bias

• the gravest threat