Summary description of the tool

Page Links:

About this tool

This tool is a web application that shows the amount of "heterogeneity" of tuberculosis incidence for different high-risk populations. Tuberculosis (TB) is a disease caused by bacteria that are spread from person to person through the air. TB usually affects the lungs, but it can also affect other parts of the body, such as the brain, the kidneys, or the spine. In most cases, TB is treatable and curable; however, people with TB can die if they do not get proper treatment.

"Heterogeneity" – or diversity – is the difference in the level of TB in a high-risk group (for example, people born outside the United States), from one U.S. state to the next. Users can select a high-risk group of interest, and then select a specific U.S. state to serve as the "benchmark" state. TB rates for all other states will then be compared to that state. Results are shown in graphs and based on inputs from the user. The graphs update automatically.

For the definitions of terms and abbreviations used in this application, see the GLOSSARY section of this summary. Additionally, common questions about this tool are answered in the FREQUENTLY ASKED QUESTIONS (FAQ) section below. Any further questions may be directed to ddowdy1@jhmi.edu

The charts shown in this tool are based on calculations from reported TB incidence data in the United States. They can be used to better understand the differences in TB rates, comparing each state to the next.

Depending on the requirements of TB controllers in each state, this tool can:

  1. Illustrate the magnitude of between-state differences in TB incidence among key populations;
  2. Highlight the successes of certain surveillance programs on comorbid conditions (for example, states with high TB incidence in a given population may produce better estimates of that population); and
  3. Provide insights into key population groups in each state that may require extra effort in TB control (for example, states with high TB incidence in a given population may want to focus more effort on that population)

For further details on the epidemiological factors included in the model, detailed methods, and main results see the paper, “Tuberculosis Incidence Among Populations at High Risk in California, Florida, New York, and Texas”; S.T. Cherng, S. Shrestha, S. Reynolds, A.N. Hill, S.M. Marks, J. Kelly, D.W. Dowdy. American Journal of Public Health: 108(S4), pp. S311–S314

The findings and conclusions described in this web application and linked journal article are those of the author(s) and do not necessarily represent the views of the U.S. Centers for Disease Control and Prevention. This web tool was funded by the CDC, National Center for HIV, Viral Hepatitis, STD, and TB Prevention Epidemiologic and Economic Modeling Agreement (NEEMA, # 5 NU38PS004646-05-00)

Back to top

Organization of the tool

This tool has two sections with interactive graphs and data tables: Incidence Plot (and Incidence Table), and the Mean Annual Average Plot (and Mean Annual Average Table). Each section can be accessed from the direct hyperlinks.

The Incidence Plot shows the “heterogeneity-reflective incidence” for each state, based on a user-selected or default risk factor. The heterogeneity-reflective incidence is calculated as the relative difference between each state’s mean annual TB incidence and a benchmark incidence, or the "minimum incidence." The "minimum incidence" is the lowest mean annual TB incidence among the four most populous states specific to the selected risk factor.

The Incidence Table shows the values used to calculate each state's heterogeneity-reflective incidence.

The Mean Annual Average Plot shows the mean annual TB cases by risk factor across U.S. states from 2010 to 2015.

The Mean Annual Average Table shows estimates of mean annual TB cases by risk factor across U.S. states.

Back to top

Incidence Plot

User options are shown in a dropdown menu. The user specifies one of eight risk factors, as follows:

  • Non-US–Born - Europe (default option): Non-US–Born persons who were born in Europe and live in the United States;
  • Non-US–Born - Asia: Non-US–Born persons who were born in Asia and live in migrated to the United States;
  • Non-US–Born - The Americas: Non-US–Born persons who were born in the Americas and live in the United States;
  • Non-US–Born - Africa: Non-US–Born persons who were born in the Americas and live in the United States.
  • Diabetic: persons with diabetes living in the United States.
  • Homeless: persons experiencing homelessness while living in the U.S.
  • Incarcerated: persons experiencing incarceration while living in the U.S.
  • HIV+: persons living with HIV* and/or AIDS in the U.S.

* Human Immunodeficiency Virus, or HIV, is a type of virus that is spread through certain bodily fluids. HIV infects the human immune system, the system in the body in charge of fighting off illness. In most cases, HIV is preventable and treatable; however, people with HIV can progress to AIDS and die if they do not get proper treatment.

The resulting bar-graph depicts the difference in mean annual tuberculosis incidence between U.S. states by risk factor during the years spanning 2010 and 2015.

Users can also examine each state’s information by either hovering their mouse arrow over the desired state's column in the graph or selecting the desired state from the dropdown menu following the Incidence Table.

Heterogeneity-Reflective Incidence

  • The lower striped portion of each bar reflects the 'minimum incidence', or the lowest mean annual TB incidence among the four most populous states specific to the selected risk factor
  • The upper cross-hatched pattern of each bar represents the 'heterogeneity-reflective incidence'. This measure reflects the overall incidence for each state that occurs in addition to the minimum incidence based on the selected risk factor.
  • The solid horizontal line reflects the TB incidence of the benchmark state. It can be used as a reference to compare incidences across states. Users can change the benchmark state from the default option (Florida), as described in the next section.

Changing the benchmark incidence

The solid horizontal line provides flexibility for the user to explore different states as benchmarks. The default options are either Florida or New York.

To change this so, the user may access the dropdown menu following the Incidence Table and change the benchmark state to move this solid line up or down along the left axis (TB incidence). This will allow the user to compare the magnitude of incidences across states.

Back to top

Incidence Table

A user can view the underlying data used to generate the Incidence Plot. The values in this table correspond to the risk factor and baseline state selected by the user in the dropdown menus.

The table lists the following variables:

  • State, representing the 50 U.S. States, and the District of Columbia;
  • Minimum Incidence, representing the benchmark incidence, or the lowest risk-factor-specific incidence among the four most populous U.S. states;
  • Heterogeneity-Reflective Incidence, representing the relative difference between the mean incidence in a selected state and the minimum incidence;
  • Total Incidence, representing the total number of incident TB cases per 100,000 people, per year, by state;
  • Selected benchmark value, representing the total incidence of the benchmark state selected by the user;
  • Total difference from selected benchmark, representing the difference between the total incidence and the selected benchmark value.

Note: States that report lower TB incidence than the minimum incidence will report zero heterogeneity-reflective incidence (third column) for that risk factor.

Back to top

Mean Annual Average Plot

Users can observe the mean annual TB cases across U.S. states based on the same risk factor selected for the Incidence Plot

Note: Users can click the checkbox below the graph to toggle the inclusion of the total annual TB cases for the entire U.S.

Back to top

Mean Annual Average Table

Users can view the underlying data used to generate the Mean Annual Average Plot. The values in this table correspond to the risk factor selected by the user in the dropdown menu for the Incidence Plot.

Further information on the tool

While many state health departments may be aware of which populations in their specific state are at highest risk for TB disease, showing the incidence within these key populations across all states provides a basis for measurable improvement.

Importantly, differences in resource allocation for TB control between states may mean that different states have different needs and priorities regarding the measurement of TB incidence by risk group.

User-friendly online tools can thus aid local jurisdictions in decision-making about resource allocation in TB control, and are complementary to the state TB epidemiology reports.

Back to top

Quantifying TB Heterogeneity, An Example

This example is intended for more technical readers who want to understand the calculations used.

State A's Tuberculosis Incidence
State A
Mean Total Population
(2011-2015)
Mean Total Cases
(2011-2015)
TB Incidence, per 100,000/year
1,406,299 123.6 123.6/1,406,299 * 100,000 = 8.79
State A's Risk Factor Statistics
State A
Mean Total Diabetics
(2011-2015)
Mean Total Cases in Diabetics
(2011-2015)
TB Incidence among Diabetics, per 100,000/year
109,691 41 41/109,691 * 100,000 = 37.38
State B's Tuberculosis Incidence
State B (Benchmark Comparison State)
Mean Total Population
(2011-2015)
Mean Total Cases
(2011-2015)
TB Incidence, per 100,000/year
6,705,586 200.6 200.6/6,705,586 * 100,000 = 2.99
State B's Risk Factor Statistics
State B (Benchmark Comparison State)
Mean Total Diabetics
(2011-2015)
Mean Total Cases in Diabetics
(2011-2015)
TB Incidence among Diabetics, per 100,000/year
532,423 14 14/532,423 * 100,000 = 2.63
The following three table rows contain a compact, readable version of the 3 equations that follow in the table
HRPstate,risk group=Max(0,Incidenceselected[state,risk group]-Incidencebenchmark[state,risk group]/Incidenceselected[state,risk group]) Eq. 1 from manuscript0.93 = (37.38 - 2.63) / 37.38
Cases specific to State 'A'=HRPstate,risk group * Casesstate,risk group Modified Eq. 2, selected between states38.13 = 41 * 0.93
HRCnational=Sum(All HRPstate,risk group * Casesstate,risk group Eq. 2 from manuscriptNo Data in this example
HRP state,risk group = Max(0,Incidence comparsion state,risk group - Incidence benchmark state,risk group)Eq. 1 from manuscript
Incidence selected state,risk group
Incidence selected state,risk group = 37.38, Incidence benchmark state,risk group = 2.63
0.93 = Max(0,37.38 - 2.63)
37.38
Cases specific to State 'A' = HRP selected state,risk group * Cases selected state,risk groupEq. 2 (state selected)
HRP selected state,risk group = 0.93, Cases selected state,risk group = 41
38.13 = 0.93 * 41
Therefore there are 38.13 cases of TB in diabetics specific to living in selected State 'A'
The following equation describes this calculation made for the entire United States
HRC national, risk group = Σ(HRPselected state,risk group * Cases selected state,risk group)Eq. 2 from manuscript

  • Mean Total Cases: mean (average) total number of new cases of TB disease by risk factor across six years (2010 to 2015) in the population of interest;
  • TB Incidence: mean annual number of incident (new) TB cases per 100,000 people, per year, by risk factor across six years (2010 to 2015);
  • State B (Comparison State): the benchmark state, or the state with the lowest risk-factor-specific TB incidence (minimum incidence), among the four most populous states (California, Texas, New York, Florida)
  • Risk group: the population with the specified risk factor;
  • Heterogeneity-Reflective Incidence, or Proportion (HRP): The relative difference in risk factor-specific TB incidence between a selected state and the benchmark state (as a proportion of the risk factor-specific TB incidence in the selected state).
  • Heterogeneity-Reflective Cases (HRC): The mean number of incident TB cases among people with the given risk factor in the selected state, multiplied by that risk group’s heterogeneity-reflective proportion.

Note: Heterogeneity is given as a proportion of incidence.

Back to top

Sources of Data

Non-U.S.–born people

The sizes of non-U.S.–born populations by region of birth were obtained from the American Community Survey (ACS) factfinder website using the dataset containing 5-year estimates from 2015 for the years 2011-2015 by state. Regions of birth were specified in the ACS as Europe (Northern, Western, Southern, and Eastern), Africa, the Americas (North, Central, South, Latin, and Mexico), and Asia (Western, South Central, and Eastern). Incident cases of TB among non-U.S. born individuals were reported by the National Tuberculosis Surveillance System (NTSS) using World Health Organization (WHO) regional designations: European, Southeast Asia, Eastern Mediterranean, Western Pacific, Americas, and African. For these analyses, we aggregated non-U.S. born regional categories from the NTSS data into the four regional categories that correspond to the ACS: Americas, Europe (European and Eastern Mediterranean), Africa, and Asia (Southeast Asia and Western Pacific).

People living with diabetes

Populations with diabetes were estimated from the Behavioral Risk Factor Surveillance System (BRFSS) using the U.S. Diabetes Surveillance System website from 2011-2015. Respondents were considered to have diagnosed diabetes if they responded "yes" to the question, "Has a doctor, nurse, or other health professional ever told you that you have diabetes?" Diabetes prevalence from each state was multiplied by the total estimated population in that state using U.S. census data.

People experiencing homelessness

Homeless populations were estimated for years 2012-2015 from Part 2 of the U.S. Department of Housing and Urban Development's (HUD's) annual homeless assessment report (AHAR) to the U.S. Congress. Based on the AHAR, one-year national estimate of sheltered homelessness cases and the estimated proportion of total sheltered homeless populations by state, we calculated the total cases of homeless people occurring in every state. Sheltered homelessness is defined as people who "used an emergency shelter or transitional housing program at any time from October 1 through September 30 of the following year." These data are associated with the calendar year corresponding to October 1st in our analyses, and exclude people served by victim service providers and individuals experiencing homelessness who never accessed a shelter program during the 12-month period. HUD estimates that approximately 35% of homelessness can be attributed to individuals experiencing homelessness outside of a sheltered setting. Assuming our homelessness denominators were to increase proportionally across all states, we would expect our estimates of incidence differences and percentage of heterogeneity attributable cases to decrease slightly both in risk-group specific estimates, and across the United States.

People experiencing incarceration

Incarcerated populations were estimated from the U.S. Department of Justice Corrections Statistical Analysis for the years 2011-2015. This population includes all correctional inmates in both prison and jail settings under the jurisdiction of state or federal correctional authorities, as reported by central respondents in each of the 50 state departments of corrections. These data through the 31st of December are reported annually.

People living with HIV

HIV-positive populations were estimated from the CDC National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention AtlasPlus database from 2010-2014. We used HIV positive case counts by state for individuals aged 13 years and older from the online tool.

Back to top

Glossary of Terms

Risk factor

Any characteristic or exposure of an individual that increases their likelihood of being infected and/or developing a disease. For example, risk factors include diabetes, homelessness, incarceration.

Risk Group

The population with a specified risk factor. For example, risk groups include diabetics, people experiencing homelessness, people experiencing incarceration.

Benchmark state

The state whose risk-factor-specific TB incidence is being used for comparison (against all other states).

Incidence (or incident cases)

Incident cases are new disease cases. Incidence refers to the number of new cases that develop in a particular population in a given period of time.

Heterogeneity

A measure of diversity, or variation within a population.

Heterogeneity-Reflective Incidence

A measure for quantifying heterogeneity in TB incidence from one state to the next. Heterogeneity-reflective incidence is calculated as the difference between the mean incidence for a particular risk group in a selected state and the mean incidence for the same risk group in the benchmark state, divided by the mean incidence in the selected state. Also referred to as Heterogeneity-Reflective Proportion of incidence

Heterogeneity-Reflective Cases

The mean number of incident TB cases for each risk factor in a selected state, multiplied by its respective heterogeneity-reflective proportion, or HRP.

Where can I find more information about TB?

General information and resources on tuberculosis can be found on the Centers for Disease Control’s Tuberculosis webpage.

Back to top

To return to the data, click here