|
||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||
• How do I use CTSonline?
• How can I download my CTSonline results?
• How can I modify the format of the tables?
• I’m interested in CTS estimates that are not available using the CTSonline menus. How can I obtain those estimates?Technical information on CTSonline estimates
• How were the cases where survey respondents did not answer a question treated in calculating the estimates?
• Why are some estimates in CTSonline considered statistically unreliable?
• What exactly is the CTSonline feature that does approximate significance tests?More information on the Physician Survey system
• What is the CTS Physician Survey?
• What exactly do the Physician Survey estimates in CTSonline represent?
• For results that are disaggregated by selected subgroups (such as physician age), it would be useful to know how large each subgroup is. Where can I find that information?
• How are primary care physician (PCP) and specialist defined?
• How are the categories for area of residence defined?
• How is HMO penetration defined?
• How can I see estimates for basic demographic characteristics of physicians?
• What is the recommended citation for Physician Survey estimates from CTSonline?More information on the Household Survey system
• What is the CTS Household Survey?
• What exactly do the Household Survey estimates in CTSonline represent?
• How are the categories for Body Mass Index (BMI) defined?
• What is the definition of children with special health care needs?
• How are the categories for race/ethnicity defined?
• How are the categories for income defined?
• How are the categories for area of residence defined?
• How are the categories for health status defined?
• How are the categories for insurance coverage defined?
• What is the recommended citation for Household Survey estimates from CTSonline?How do I use CTSonline? CTSonline was designed to be very quick and easy to use. At its simplest, you select the topic that you want to see in your table and then click on a tab that immediately displays the table with estimates for that topic for each survey year. You can also select from a set of menus to indicate how you would like to disaggregate the results. For the Physician Survey, you can disaggregate the results by type of physician (PCP or specialist) and by a variety of other factors (such as physician age and gender). For the Household Survey, you can disaggregate the results by age (elderly, children, nonelderly adults) and by a variety of other factors (such as race/ethnicity, income, health status, and health insurance coverage).
How can I download my CTSonline results? Whenever you request a table in CTSonline, beneath the table you will see a link that you can click to get a version of the table that is easy to download. That version lists easy step-by-step instructions for saving the table as a Microsoft Excel file or a Microsoft Word file. Results can then be copied from Excel or Word into other software packages as needed.
How can I modify the format of the tables? Currently within CTSonline there is no way to adjust the format of the tables, such as combining response categories or changing the order of the rows and columns. If you would like to modify a table, you should download it and then edit it within a word processing or spreadsheet package.
I’m interested in CTS estimates that are not available using the CTSonline menus. How can I obtain those estimates? Although CTSonline does allow you to request a wide variety of customized tables, there are limits to the types of tables that it can generate due to the fact that it is based on a set of pre-calculated estimates. You can send requests for estimates that you would like to see added to CTSonline to ctsonline@hschange.org. In addition, you are encouraged to obtain the survey data files so that you can do your own data analysis.
Technical information on CTSonline estimates
How were the cases where survey respondents did not answer a question treated in calculating the estimates? The treatment of those cases varies with the survey question. For some measures, values for those missing cases were imputed and then included in calculation of the CTSonline estimates. For other measures, there was no imputation, and so those cases were excluded. Information about which approach was taken is provided in the footnotes that are displayed along with each table.
Why are some estimates in CTSonline considered statistically unreliable? An estimate in CTSonline is considered not statistically reliable if the relative standard error exceeds 30% and/or the estimate is based on fewer than 100 observations. Those cases are noted in the tables. In most of those cases, the estimates are reported because they can be calculated directly from the other estimates in the same row of the table.
What exactly is the CTSonline feature that does approximate significance tests? Along with each table is provided a link to a window that allows you to do an approximate test of whether the difference between two estimates is statistically significant. To do the test, you click on the link and then enter the two estimates and their associated standard errors. Those values are used to calculate a Z-statistic and p-value for the test of whether the difference between the estimates is zero. Please note, however, that this test will not yield exact results because of the complex sample design of the CTS surveys, which incorporates stratification, clustering, and oversampling. If you would like exact tests of whether two estimates differ, you need to obtain the survey data files.
More information on the Physician Survey system
What is the CTS Physician Survey? The CTS Physician Survey is a nationally representative telephone survey of physicians in the contiguous U.S. who provide direct patient care for at least 20 hours per week, excluding federal employees, specialists in fields in which the primary focus is not direct patient care (such as radiology, anesthesiology, and pathology), and foreign medical school graduates who are only temporarily licensed to practice in the U.S. It has been conducted four times beginning in 1996-97. The majority of the sample is clustered in 60 communities ("sites"), while the first three rounds included a smaller supplemental sample drawn from the entire contiguous U.S. There are approximately 12,000 physicians interviewed in each of the first three rounds of the survey, and about 6,600 in the fourth round. For more detailed information on the Physician Survey, visit the HSC web site to see a list of HSC Technical Publications. For each round of the survey, there are two user’s guides and two codebooks (one version for the public use data file and one version for the restricted use data file) that provide information on the survey as well as technical information on the variables and the data files. If you’re interested in additional detail, you can refer to the survey methods report and the survey instrument from each round of the survey.
What exactly do the Physician Survey estimates in CTSonline represent? The Physician Survey sample consists of physicians in the contiguous U.S. providing direct patient care for at least 20 hours per week, excluding federal employees, specialists in fields in which the primary focus is not direct patient care, and foreign medical school graduates who are only temporarily licensed to practice in the U.S. The sample has been assigned weights that adjust for systematic nonresponse and allow the sample to be used to make nationally representative estimates. Therefore, the estimates provided in CTSonline are nationally representative for this group of physicians for the time period of the surveys (1996-97 for Round 1, 1998-99 for Round 2, 2000-01 for Round 3, and 2004-05 for Round 4).
For results that are disaggregated by selected subgroups (such as physician age), it would be useful to know how large each subgroup is. Where can I find that information? Click here to see a table that shows the proportion each subgroup is within the full physician population for each year of the survey.
How are primary care physician (PCP) and specialist defined? In the CTS Physician Survey, a physician is considered a primary care physician (PCP) if most of his/her patient care time is spent in a primary care specialty (family practice, general practice, internal medicine, general pediatrics, internal medicine/pediatrics, adolescent medicine, or geriatric medicine). All other physicians are placed in the specialist category.
How are the categories for area of residence defined? Even though the estimates from the CTS Physician Survey are nationally representative, the sample is concentrated in 60 communities throughout the U.S. For CTSonline, physicians are grouped into three categories based on the size of the community in which they practice. Among the 60 communities, 51 are metropolitan areas, which are divided into large metropolitan areas (populations of more than two million) and small metropolitan areas (populations of less than two million). [If you are familiar with the sample design of the CTS, you will notice that these definitions of large and small differ from those used when communities were being selected for inclusion in the CTS sample.] The remaining 9 communities are all nonmetropolitan areas. Each community is defined by a specific set of counties, and that definition is the same for all years of the survey. The metropolitan areas generally conform to the metropolitan statistical areas (MSAs) defined by the Office of Management and Budget at the time the CTS was designed. The nonmetropolitan areas conform to the nonmetropolitan portions of the economic areas defined by the Bureau of Economic Analysis at the time the CTS was designed. For the metropolitan areas, information on population size came from the Census Bureau county population estimates for July 1996, 1998, 2000, and 2003. Click here to see a list of which communities are in which size category.
How is HMO penetration defined? The sample for the CTS Physician Survey is concentrated in 60 sites across the U.S., and so the level of HMO penetration for each physician is simply the level for the site in which the physician practices. Estimates from the CTS Household Survey were used to measure the HMO penetration rate in each site. The level of HMO penetration was calculated as the percentage of insured individuals who report being enrolled in an HMO. The site was considered to have "low" penetration if that rate was less than 40%, "medium" penetration if 40% to 55%, and "high" penetration if greater than 55%. Click here to find the appropriate technical documentation for the HMO penetration categories in each round.
How can I see estimates for basic demographic characteristics of physicians? Estimates of physician age, gender, race, Hispanic origin, specialty, and whether medical school was in the U.S. are not currently available in CTSonline tables. However, you can view a summary table with those estimates (for each year, aggregated over all physicians). If there are any demographic estimates that you would like to see added to CTSonline, send your suggestions to ctsonline@hschange.org.
What is the recommended citation for Physician Survey estimates from CTSonline? Center for Studying Health System Change, CTSonline (www.hschange.org), CTS Physician Survey (survey years), Date. [For the survey years, use 1996-97 and/or 1998-99 and/or 2000-01 and/or 2004-05.]
More information on the Household Survey system
What is the CTS Household Survey? The CTS Household Survey is a nationally representative telephone survey of the civilian noninstitutionalized population in the contiguous U.S. (To ensure proper representation, the survey also includes interviews with households that have no telephones.) The survey has been conducted five times: 1996-97, 1998-99, 2000-01, 2003, and 2007. For the first three surveys, the majority of the sample was clustered in 60 communities (“sites”), with a smaller supplemental sample drawn from the entire contiguous U.S. For the 2003 survey, the entire sample was clustered in 60 communities. In 2007 the survey did not include the 60 community clusters, it was only a national sample. The sample for each of the first four years includes 47,000 - 60,000 people; 2007 includes about 18,000 people. For more detailed information on the Household Survey, visit the HSC web site to see a list of HSC Technical Publications. For each round of the survey, there are two user’s guides and two codebooks (one version for the public use data file and one version for the restricted use data file) that provide information on the survey as well as technical information on the variables and the data files. If you’re interested in additional detail, you can refer to the survey methods report and the survey instrument from each round of the survey.
What exactly do the Household Survey estimates in CTSonline represent? The Household Survey sample consists of civilian noninstitutionalized individuals living in the contiguous U.S. (i.e., excluding Alaska and Hawaii). The sample has been assigned weights that adjust for systematic nonresponse and allow the sample to be used to make nationally representative estimates at either the family level or the person level; CTSonline has person-level estimates only. Therefore, the estimates provided in CTSonline are nationally representative for people in the civilian noninstitutionalized population for the time period of the surveys (1996-97, 1998-99, 2000-01, 2003, and 2007).How are the categories for Body Mass Index (BMI) defined? BMI is an indicator of someone’s weight adjusted for their height. It is calculated from survey respondents’ reports of their weight and height: BMI = { (weight in lbs.) / [ (height in inches) 2 ] } * 703. The BMI categories are defined by specific BMI values. For adults age 21 and over, the category “Underweight” is defined by BMI less than 18.5, the category “Normal” is defined by BMI of 18.5 – 24.9, the category “Overweight” is defined by BMI of 25.0 – 29.9, and the category “Obese” is defined by BMI 30.0 or above.
What is the definition of children with special health care needs? Children with special health care needs are identified through a series of survey questions, most of which come directly from the Children with Special Health Care Needs (CSHCN) Screener, which was developed through the Child and Adolescent Health Measurement Initiative. For CTSonline, a child is considered to have a special health care need if he/she has a medical, behavioral, or other health condition that has lasted or is expected to last for at least 12 months and causes any of the following: 1) child needs or uses medicine prescribed by a doctor, 2) child needs or uses more medical care, mental health, or educational services than is usual for most children of the same age, 3) child is limited or prevented in any way in his/her ability to do the things most children of the same age can do, 4) child needs or gets special therapy, such as physical, occupational, or speech therapy, or 5) child has any kind of emotional, developmental, or behavioral problem for which he/she needs or gets treatment or counseling.
How are the categories for race/ethnicity defined? Race/ethnicity was determined in the CTS Household Survey by two separate survey questions. The first addressed Hispanic ethnicity and was the same for all years of the survey: "Do you consider yourself to be of Hispanic origin, such as Mexican, Puerto Rican, Cuban, or other Spanish background?" The second addressed race and changed slightly between the 2000-01 and 2003 surveys; the respondent was asked to report a single race in the 1996-97, 1998-99, and 2000-01 surveys, whereas the respondent was allowed to report multiple races in the 2003 and 2007 surveys. The responses to the race and ethnicity questions were combined to create the four race/ethnicity categories used in CTSonline. Anyone answering "yes" to the first question was placed in the Hispanic category, regardless of his/her answer to the second question. For everyone else, only the responses to the second question were used to determine which of the remaining three categories was applicable (White non-Hispanic, Black non-Hispanic, or Other non-Hispanic). Because of the change in the question about race starting in the 2003 survey, the race categories for 2003 and 2007 are not exactly identical to the race categories for the prior years, although they are quite close. Starting in 2003, the "White non-Hispanic" category consists of only those non-Hispanic people who reported the single race White. Similarly, starting in 2003, the "Black non-Hispanic" category consists of only those non-Hispanic people who reported the single race Black. All other non-Hispanic people, including those who reported White in combination with some other race or Black in combination with some other race, were placed in the "Other non-Hispanic" category.How are the categories for income defined? Information on income in the CTS Household Survey comes from a question that asks for the family’s total income from all sources during the calendar year prior to the interview. The three income categories in CTSonline (low, middle, and high) are defined using the ratio of family income to the Federal Poverty Level (FPL) for the applicable family size. Someone is considered to be in a family with low income if his/her family’s income is less than 200 percent of the FPL. Middle income is defined as income 200 percent to 400 percent of the FPL, and high income is greater than 400 percent of the FPL.
How are the categories for area of residence defined? 1996-97, 1998-99, 2000-01, and 2003 CTS Household Surveys: even though the estimates are nationally representative, the sample is concentrated in 60 communities throughout the U.S. For CTSonline through 2003, people are grouped into three categories based on the size of the community in which they live. Among the 60 communities, 51 are metropolitan areas, which are divided into large metropolitan areas (populations of more than two million) and small metropolitan areas (populations of less than two million). [If you are familiar with the sample design of the CTS, you will notice that these definitions of large and small differ from those used when communities were being selected for inclusion in the CTS sample.] The remaining 9 communities are all nonmetropolitan areas. Each nonmetropolitan community is defined by a specific set of counties, and that definition is the same for the first four rounds of the survey. The metropolitan areas generally conform to the metropolitan statistical areas (MSAs) defined by the Office of Management and Budget (OMB) at the time the CTS was designed. The nonmetropolitan areas conform to the nonmetropolitan portions of the economic areas defined by the Bureau of Economic Analysis at the time the CTS was designed. For the metropolitan areas, information on population size came from the Census Bureau county population estimates for July 1996, 1998, 2000, and 2003. Click here to see a list of which communities are in which size category.
2007 CTS Household Survey: the sample no longer includes 60 communities, and the OMB replaced MSAs by CBSAs (Core Based Statistical Areas). Each CBSA must contain at least one urban area of 10,000 or more population. CBSAs with at least one urbanized area of 50,000 or more inhabitants are defined as metropolitan statistical areas; urban clusters with at least 10,000 but less than 50,000 population are called micropolitan statistical areas. For 2007 CTSOnline we used Census 2007 CBSA population counts to define large (greater than 2 million) and small (less than 2 million) metropolitan areas.How are the categories for health status defined? The CTS Household Survey includes a question that asks, “In general, would you say your health status is excellent, very good, good, fair, or poor?” Each adult responds to this question for himself/herself. For each child in the sample, an adult family member provides his/her own assessment of the child’s health status. For CTSonline, people are divided into two categories: 1) fair or poor health, and 2) good, very good, or excellent health.
How are the categories for insurance coverage defined? The definition of the insurance coverage categories in CTSonline is based on a series of questions in the CTS Household Survey designed to determine the type(s) of health insurance plan(s) by which each person is covered.
The survey first asks about private coverage, which includes employment-related health insurance (i.e., plans from current or former employers or unions) and insurance bought individually (either directly from a health plan or through a professional association or trade group). Note that private insurance refers only to comprehensive health plans, which means it excludes single-service plans such as dental plans. After determining private insurance coverage, the survey asks whether someone is covered by Medicare, followed by questions about Medicaid and any other state-specific health insurance programs, including SCHIP (State Children’s Health Insurance Program). Next, coverage by military health insurance (CHAMPUS, CHAMP-VA, TRICARE, VA, or other) is determined, and then coverage by the Indian Health Service (IHS). Lastly, the series of questions concludes by asking whether the person is covered by some other type of health insurance not specifically mentioned in the survey. People who do not have any health insurance coverage according to this series of questions are classified as uninsured.
To create the six insurance coverage categories used in CTSonline, we first separated people into two groups based on age: elderly (age 65 or greater) and nonelderly (age less than 65). The elderly group was divided into those covered by Medicare and those not covered by Medicare. Note that those without Medicare are not necessarily uninsured; they might have insurance coverage from other sources, such as a current or former employer. (Although it would be useful to be able to differentiate among the types of insurance coverage for the elderly who do not have Medicare, that group is too small to separate into subgroups for CTSonline.) Within the nonelderly group, anyone not covered by any type of insurance was classified as uninsured. Anyone covered by a private comprehensive insurance plan was classified as privately insured (except for the small group of people reporting both private insurance and Medicare). Among the people with no private insurance, those who reported coverage by Medicaid or other state plans were placed into their own separate category (except for the small group of people also reporting Medicare coverage). All the remaining people in the nonelderly group (i.e., those with Medicare, military insurance, IHS, or other insurance) were classified as nonelderly with other insurance.
What is the recommended citation for Household Survey estimates from CTSonline? Center for Studying Health System Change, CTSonline (www.hschange.org), CTS Household Survey (survey years), Date. [For the survey years, use 1996-97 and/or 1998-99 and/or 2000-01 and/or 2003 and/or 2007.]
