1: Introduction Vitalnet is a comprehensive, integrated system for analyzing health data. California Epigram is the Vitalnet module for analyzing California mortality data. The software is designed to be easily used without a user's guide. However, many users learn better from a written text, and all users will probably benefit from a general overview of what the software can do before using it. This user's guide describes Epigram Professional Version (VitalPro). The Web Version (VitalWeb) is described at "http://www.ehdp.com/". Either system will greatly ease and speed your work. This user's guide incorporates a tutorial. Carrying out the tutorial will help you quickly become an expert. All procedures you are expected to carry out as part of the tutorial are highlighted by a different text appearance, as shown in this example:

Sample Tutorial Step - Press 'A' to add California as one area set.

Here's how this user's guide is organized: Chapter 1: Introduction - Describes general characteristics of California Epigram, access methods, and confidentiality requirements. Chapter 2: Understanding Results - Explains the organization and content of Epigram tables (the results you get from using Epigram). Shows and explains typical tables. Chapter 3: Using the Epigram Interface - Describes how to use Epigram menus to select parameters, produce tables, and save output, with examples. Guides you step-by-step to produce your first table. Outlines available menus. Chapter 4: California Epigram Parameters - Describes allowed selections for age groups, area sets (counties and health service areas), ICD-10 sets (causes of death), race groups, years, and other options. Glossary - Defines terms related to analyzing mortality data and using Epigram. Citation for Epigram - Goldman, DA, California Vitalnet User's Guide: Data Warehouse Software for Analyzing California Mortality Data. Austin, Texas. 1998. Browse http://www.ehdp.com/vitalnet/ for more information about Epigram or to contact EHDP. Acknowledgements - We gratefully acknowledge the cooperation and assistance of staff from the California Department of Health Services and other users.

Advantages and Benefits Epigram makes it easy to analyze California mortality data. Fast - You get results in seconds or minutes. Depending on the analysis, alternate methods could easily require hours to weeks to complete. Flexible - A wide variety of tables may be produced. Set rows and columns however you want. Standard parameters may be selected and combined as needed. Efficient - You can make a whole series of tables with one keystroke (multi-tables). Bar graphs allow for quick scanning for trends. Easy to use - Operations are menu-driven. Scrolling windows are used to select items from lists. You don't need to know any special codes such as county FIPS codes. Well-documented - Although Epigram has been designed to be self-explanatory, it also includes extensive on-line help. Each menu has its own help screen. Also, you may select from a list of on-line help topics, providing advice on all topics related to Epigram. The on-line help complements the information included in this user's guide. Useful on-line reference materials are also provided, such as tables of standard populations used for age-adjustment. Finally, all output is fully documented. Integrates with other software - Tables may be saved to a log file in ASCII format for subsequent editing and printing with any editor. You may also save Epigram tables in tab-delimited format or as a dBASE III file, for easy importing into data analysis, spreadsheet, mapping, graphing or other presentation software. Or, tables may be saved as HTML for display in a web browser.

Data Within Epigram Epigram links and analyzes population, mortality, ICD-10 and geographic data. Geographic information - Epigram includes a database of California counties and Health Service Areas, linked to the population and mortality data sets. Population data - The population variables are age, county of residence, race, sex, and year. Population figures are used for calculating mortality rates. A separate Vitalnet interface, PopTrend, analyzes population data for demographic trends and to obtain denominators for analyzing other data sets. Mortality data - The Center for Health Statistics at the California Department of Health Services provides all California mortality data. For an estimate on when the next data file will be loaded, contact the Center. The mortality variables are age of deceased, cause of death, county of residence, race, sex, and year. Epigram analyzes by place of residence of the deceased. For example, if a Sacramento resident died in a motor vehicle collision in Los Angeles, Epigram would classify the death under Sacramento. Standard mortality reports usually use place of residence (the other system is "place of occurrence"). Epigram analyzes by underlying cause (the cause which initiated the sequence of events leading to death). For example, if a death certificate lists rheumatoid arthritis, myocardial infarction (MI), and cardiac arrhythmia secondary to MI, the underlying cause is MI. Standard mortality reports usually use underlying cause. ICD (International Classification of Diseases) information - The ICD is the standard system for classifying causes of illness and death. A complete database of the ICD system is included for selecting and combining causes of death to analyze. For example, you can easily determine the leading causes of death, or select any combination of ICD codes for analysis. The ICD-10 is used starting with 1999 mortality data. Mortality data for 1980 to 1998 use the ICD-9. Data linking - Four data sets (population data, mortality data, geographic data, and ICD data) are linked with each other, so the Epigram data warehouse is greater than the sum of its parts.

Access Methods Both local and remote access are available for California Epigram. Local access (PC's and local area networks) - If you are a California Department of Health Services employee, you will typically use California Epigram from the DHS local area networks. California Epigram may also be installed to a stand-alone IBM-compatible PC or laptop running any version of Windows. Execution speed - Epigram is fast. Analyses that might otherwise take hours to weeks are done in seconds or minutes. For those who access California Epigram on a PC network, the program execution speed depends on the type of PC you are using and the characteristics of the network you are working on. The program has been tested to work well on all IBM-compatible PC's, and simply runs faster on faster PC's. DHS Network Access DHS employees may access California Epigram from the DHS networks. Windows - The following assumes that you are a California Department of Health Services employee, and are using a PC attached to a DHS network. Using any version of Windows, click on the Epigram icon to start the program. The icon looks like a doctor's bag. If you do not have an icon, request that your network manager set things up so that Epigram can be run from your network. When you are done - After you finish using California Epigram, you will still be logged on to your own server, so you do not need to reboot your PC. If you have a problem accessing California Epigram from a DHS network, ask your local network manager for help first.

Confidentiality Policy California Epigram users must comply with confidentiality requirements. Confidentiality policy - Your use of California Epigram indicates your agreement to the following conditions: You will not try to use California Epigram results nor let anyone else use California Epigram results to learn the identity of a reported death, or for any purpose other than statistical analysis. If you discover the identity of a reported death, you will advise the Director of the Center for Health Statistics at the California Department of Health Services of the incident, will safeguard or delete the information that would identify the individual, will make no use of the knowledge, and will inform no one else of the discovered identity. Cell suppression - If you so desire, California Epigram can suppress cells that have fewer than a user-defined number of deaths. Cell suppression can increase the confidentiality of written reports in some cases. Cell suppression is described more fully later.

2: Understanding Results Review - California Epigram is a powerful tool for analyzing California mortality data. The program basically compares anything with anything, and makes just about any kind of output table you want. Chapter 1 explained how Epigram makes analyzing mortality data easy, listed the linked data sets used by California Epigram, described the access methods, and spelled out the confidentiality policy. What's in this chapter - Before jumping in and using the program (Chapter 3), it is advised to get a good understanding of the results of the program. This chapter explains the organization and content of Epigram tables (the results you get from using Epigram). This chapter explains the four sections of a table: Header - Basic analysis parameters. Data section - Numerical results. Bar graphs - Graphical results. Footnotes - Other analysis parameters. In addition, this chapter shows examples of actual California Epigram tables, to give you an idea of what is possible. You are given an opportunity to practice interpreting sample tables.

Typical Epigram Table A typical Epigram table has a header, data section, bar graphs, and footnotes. Deaths Tabulated by Age and Sex Years: 1999 Place of Residence: Imperial, San Diego ICD J40-J47: Chronic lower respiratory diseases Age Male Female Total ----------------------------------- Birth-19 3 2 5 20-39 3 1 4 40-59 25 36 61 60-99+ 521 593 1,114 ----------------------------------- Total 552 632 1,184 Horizontal Bar Graphs (X = 39.5 Deaths, x = 19.8): Age Male Female ------------------------------------------- Birth-19 20-39 40-59 x X 60-99+ XXXXXXXXXXXXX XXXXXXXXXXXXXXX ------------------------------------------- Analysis Footnotes: Unique ID, for Keeping Track of Analyses: 507PMAZC Output Produced: Mon May 07 14:58:58 2001, by California VitalNet Deaths Classified by: ICD-10, Underlying Cause, County of Residence Mortality Data Source: DHS Office of Health Information and Research

This is a typical Epigram table. It has 4 parts: The header documents the basic type of analysis that was done. The table analyzes deaths for 1999 for two counties: Imperial and San Diego. Deaths from chronic lower respiratory diseases (ICD J40-J47) are counted. The data section contains the results, organized into columns and rows. The columns and rows can be set however you want. In this example, there is one column for each sex and one row for each of four age groups. Verify for yourself that there were 36 deaths in women age 40-59, 552 deaths in males, and a total of 1,184 deaths. By the way, each result, such as 36, 552, and 1,184, is called a "cell". The horizontal bar graphs represent the results in a simple graphical format, with a legend. In this example, the legend indicates that each 'X' (big X) symbol represents 39.5 deaths. Although the bar graphs are optional, you should usually include them, because they make it so easy to scan and understand the results. The bar graphs show that the great majority of deaths from chronic lower respiratory diseases in the two counties were in the 60-99+ age group, for both sexes. The footnote documents less important aspects of the analysis, such as when the table was produced, and assigns a unique ID to the table for future reference.

Table #1 to Interpret Here is a sample table (footnotes omitted) for you to practice on. Try interpreting the table by filling in the spaces below. Then, check your answers with those on the next page. Death Rate (per 100,000) Tabulated by Area Set and Sex Age: 20-39 Years Years: 1999 Place of Residence: Hlth Serv Area 5, Hlth Serv Area 11, Hlth Serv Area 12 ICD B20-B24: Human immunodeficiency virus [HIV] disease Male Female Total Area Set Rate, Deaths Rate, Deaths Rate, Deaths ----------------------------------------------------------------- Hlth Serv Area 5 10.1 35 2.1 7 6.2 42 Hlth Serv Area 11 13.0 202 2.5 35 8.0 237 Hlth Serv Area 12 8.5 40 2.5 11 5.6 51 ----------------------------------------------------------------- Total 11.7 277 2.4 53 7.2 330 Horizontal Bar Graphs (X = 0.9 Deaths / 100,000, x = 0.4): Male Female Total Area Set Rate Rate Rate -------------------------------------------------------------------- Hlth Serv Area 5 XXXXXXXXXXXx XXx XXXXXXX Hlth Serv Area 11 XXXXXXXXXXXXXXX XXX XXXXXXXXX Hlth Serv Area 12 XXXXXXXXXX XXX XXXXXXx -------------------------------------------------------------------- Total XXXXXXXXXXXXXx XXX XXXXXXXXx

Header Basic result type (statistic): ___________________   Years analyzed: ___________________   Geographic areas analyzed: ___________________   ICD groups analyzed: ___________________   Age groups analyzed: ___________________       Data Section    Rows variable: ___________________   Columns variable: ___________________   How many females died in HSA 11: ___________________   Death rate for males in HSA 12: ___________________       Bar Graphs Death rate symbolized by big 'X': ___________________

Table #1 Interpreted Here is the same table, and a suggested interpretation. If you had a problem, try reading through the answers again or ask a local data analyst for help. Death Rate (per 100,000) Tabulated by Area Set and Sex Age: 20-39 Years Years: 1999 Place of Residence: Hlth Serv Area 5, Hlth Serv Area 11, Hlth Serv Area 12 ICD B20-B24: Human immunodeficiency virus [HIV] disease Male Female Total Area Set Rate, Deaths Rate, Deaths Rate, Deaths ----------------------------------------------------------------- Hlth Serv Area 5 10.1 35 2.1 7 6.2 42 Hlth Serv Area 11 13.0 202 2.5 35 8.0 237 Hlth Serv Area 12 8.5 40 2.5 11 5.6 51 ----------------------------------------------------------------- Total 11.7 277 2.4 53 7.2 330 Horizontal Bar Graphs (X = 0.9 Deaths / 100,000, x = 0.4): Male Female Total Area Set Rate Rate Rate -------------------------------------------------------------------- Hlth Serv Area 5 XXXXXXXXXXXx XXx XXXXXXX Hlth Serv Area 11 XXXXXXXXXXXXXXX XXX XXXXXXXXX Hlth Serv Area 12 XXXXXXXXXX XXX XXXXXXx -------------------------------------------------------------------- Total XXXXXXXXXXXXXx XXX XXXXXXXXx

Header Basic result type (statistic): Death rate (per 100,000)   Years analyzed: 1999   Geographic areas analyzed: Health Service Areas 5, 11, 12   ICD groups analyzed: ICD B20-B24 (HIV/AIDS)   Age groups analyzed: 20-39       Data Section    Rows used in this example: Row for each area set   Columns used in this example: Column for each sex   How many females died in HSA 11: 35   Death rate for males in HSA 12: 8.5 per 100,000       Bar Graphs Death rate symbolized by big 'X': 0.9 per 100,000

Table #2 to Interpret Here is another table. Once again, give your interpretation by filling in the spaces below. Then, check your answers with those on the next page. Age-Adjusted Death Rate (per 100,000) Age Adjustment Standard Population: 2000 US Tabulated by Race-Ethnicity and Sex Years: 1999 Place of Residence: California ICD C33: Malignant neoplasm of trachea ICD C34: Malignant neoplasm of bronchus and lung Male Female Total Race Rate, Deaths Rate, Deaths Rate, Deaths --------------------------------------------------------- Am Indian 36.0 26 20.3 19 27.0 45 Asian/PI 44.0 554 23.0 360 32.3 914 Black 94.9 633 46.4 423 66.2 1,056 Hispanic 28.9 550 13.2 327 20.0 877 White 66.2 5,794 44.6 5,063 53.6 10,857 --------------------------------------------------------- Total 59.6 7,557 37.6 6,192 46.9 13,749 Overall Unknown Age for Rows (Not Added to Row Totals): 2 Total Overall Unknown Age for Columns (Not Added to Column Totals): 2 Total Horizontal Bar Graphs (X = 6.3 Deaths / 100,000, x = 3.1): Male Female Total Race Rate Rate Rate ------------------------------------------------------------ Am Indian XXXXXx XXX XXXXx Asian/PI XXXXXXX XXXx XXXXX Black XXXXXXXXXXXXXXX XXXXXXXx XXXXXXXXXXx Hispanic XXXXx XX XXX White XXXXXXXXXXx XXXXXXX XXXXXXXXx ------------------------------------------------------------ Total XXXXXXXXXx XXXXXX XXXXXXXx

Header Basic result type (statistic): ___________________   Years analyzed: ___________________   Geographic areas analyzed: ___________________   ICD groups analyzed: ___________________       Data Section    Rows used in this example: ___________________   Columns used in this example: ___________________   How many Black males died: ___________________   Death rate for Hispanic females: ___________________   Overall death rate: ___________________       Bar Graphs Death rate symbolized by big 'X': ___________________

Table #2 Interpreted Here is the second table again, and our interpretation of the results. If you had difficulty, carefully read the answers again and consult with local data experts for more help. Age-Adjusted Death Rate (per 100,000) Age Adjustment Standard Population: 2000 US Tabulated by Race-Ethnicity and Sex Years: 1999 Place of Residence: California ICD C33: Malignant neoplasm of trachea ICD C34: Malignant neoplasm of bronchus and lung Male Female Total Race Rate, Deaths Rate, Deaths Rate, Deaths --------------------------------------------------------- Am Indian 36.0 26 20.3 19 27.0 45 Asian/PI 44.0 554 23.0 360 32.3 914 Black 94.9 633 46.4 423 66.2 1,056 Hispanic 28.9 550 13.2 327 20.0 877 White 66.2 5,794 44.6 5,063 53.6 10,857 --------------------------------------------------------- Total 59.6 7,557 37.6 6,192 46.9 13,749 Overall Unknown Age for Rows (Not Added to Row Totals): 2 Total Overall Unknown Age for Columns (Not Added to Column Totals): 2 Total Horizontal Bar Graphs (X = 6.3 Deaths / 100,000, x = 3.1): Male Female Total Race Rate Rate Rate ------------------------------------------------------------ Am Indian XXXXXx XXX XXXXx Asian/PI XXXXXXX XXXx XXXXX Black XXXXXXXXXXXXXXX XXXXXXXx XXXXXXXXXXx Hispanic XXXXx XX XXX White XXXXXXXXXXx XXXXXXX XXXXXXXXx ------------------------------------------------------------ Total XXXXXXXXXx XXXXXX XXXXXXXx

Header Basic result type (statistic): Age-adjusted rate   Years analyzed: 1999   Geographic areas analyzed: California   ICD groups analyzed: ICD C33-C34       Data Section    Rows used in this example: Row for each race   Columns used in this example: Column for each sex   How many Black males died: 633   Death rate for Hispanic females: 13.5 per 100,000   Overall death rate: 46.9 per 100,000       Bar Graphs Death rate symbolized by big 'X': 6.3 per 100,000

3: Using the Epigram Interface Review - Chapter 1 explained how Epigram makes it easy to analyze California mortality data, listed the linked data sets contained within California Epigram, discussed the access methods, and described the confidentiality policy. Chapter 2 explained the layout of an output table, and presented sample tables for discussion and interpretation. What's in this chapter - Finally, you will get to use the program! This is probably what you have been waiting for! You will learn to navigate the interface, select parameters, and produce a few tables like those you learned about in the previous chapter. The chapter also has a schematic overview of all California Epigram menus, for your reference. Access the program - A tutorial runs through this chapter. All procedures you are expected to do as part of the tutorial are highlighted in a different type style, as shown below. At this point, do the following:

Access and start California Epigram using one of the methods listed previously.

You will either click on a Windows icon, select a item from the DHS menus, dial a number with your modem, or telnet to an Internet address. If needed, get assistance from another California Epigram user. You will start at the Main Menu (shown on next page).

Epigram Main Menu You will constantly return to the Main Menu. After Epigram starts, you are presented with the Main Menu, similar to the following, shown running under Windows 98: The Main Menu is the "command center" - You move to submenus to modify parameters, but return to the Main Menu to make a table. The Main Menu lists all parameters currently selected, providing an overview. Highlighting a parameter - One of the parameters, such as "Outcome", is highlighted with a "light bar" that you move by pressing an arrow key. The parameter list will scroll when you reach the bottom:

Press the down arrow key to move the light bar down. Keep pressing to see the parameter list scroll down.

The prompt is worth reading - The Main Menu (and every other Epigram menu) has a prompt at the bottom with guidance on what to do next. In this situation you can press an arrow key, the ENTER key, one of two letters (P or Q), or '?' for help. Changing window appearance - The font size for the window running Vitalnet may be changed from the Windows toobar. The size "7 x 11", as shown, is a reasonable size. The window running Vitalnet may be toggled to full screen (and back) by pressing ALT-ENTER (hold down the ALT key and press the ENTER key).

Submenus Access a submenu by highlighting an item and pressing ENTER. Selecting parameter to change - To change one of the parameters, highlight the parameter (using the arrow keys) and press ENTER. Immediate changes - Simple parameters change immediately. Try this:

Highlight the color scheme parameter by using the arrow keys. Press ENTER a few times.

Submenu changes - For complex parameters, a submenu will appear and help you change the parameter. For example, to change the geographic selection, you would highlight the parameter "Areas", and press ENTER. Do the same:

Use arrows keys to highlight "Areas". Then, press ENTER to access the submenu.

Geographic Submenu Submenus guide you step-by-step in selecting parameters. Submenu with scrolling window - This is a typical Epigram submenu. It has a scrolling window with a list of items (currently list of selected counties). One or more of the items may be highlighted by using the arrow keys. Epigram uses scrolling windows for adding and deleting ICD groups and areas, and for choosing between various other items. Letters are action items - This submenu has a list of actions, such as " E  Delete all sets". Pressing the letter (such as 'E') carries out the action. Carrying out an action - Here is a simple geographic operation:

Press 'E' to delete all areas sets. You may be asked to confirm. Then, press 'A' to add some areas. Highlight "Hlth Serv Area 1", and press ENTER to add. Then, press ESCAPE key to return to previous menu. Note that an item is added to the scrolling window. Finally, press ESCAPE to return to the Main Menu.

Select Another Submenu Access another submenu by highlighting an item and pressing ENTER. Make sure you're back - You should now be back at the Main Menu. Note that "Areas" has been modified (Hlth Serv Area 1). If you are not back at the Main Menu, get assistance from a coworker familiar with using California Epigram. Here is how to access another submenu to modify age:

Use the arrow keys to highlight the age selection. Then, press ENTER to move to a submenu for modifying age.

Age Submenu Submenus help you quickly and easily change parameters. Submenu for a single range - Your screen should look similar to that shown above. This is the type of submenu used to modify a single range, such as the age. Here is how to modify the range and return to the Main Menu:

Press LEFT or RIGHT arrow key to change the low end. Next, press the UP or DOWN arrow key to change the high end. Or, press HOME or END key to select the maximum range. Keep modifying range until you have selected a desired age group, such as 20-39. When satisfied, press 'Z' to return to the Main Menu.

Produce a Table

Press 'P' at the Main Menu when you are ready to make a table.

Check that you at the Main Menu - You should now be back at the Main Menu. You have used the submenus to select a few parameters, and are now ready to produce your first table! Produce a table - From the Main Menu, you can press 'P' to produce output.

Press the letter 'P' to produce a table similar to the following:

Time required - After you press 'P' from the Main Menu, Epigram calculates the results. The amount of time required depends on which parameters are selected and what kind of computer you are using, but is typically seconds. When Epigram finishes calculating the results, the output table will appear in a new window, as shown above. Examining the table - You may browse the table with the arrow keys and other cursor keys (PgUp, PgDn, HOME, END). When you are finished examining the table, press the ESCAPE key (or 'Z'). Epigram will ask if you want to save the table to your log file (see next page). Depending on your access method and which menu options are selected, you may also be asked if you want to print the table or to save to an alternate format (tab-delimited, dBASE III, HTML).

Press the down arrow key and other cursor keys to examine the results. Press ESCAPE or 'Z' when you are finished viewing the results. For now, press 'N' when asked if you want to save or print the table.

Save Results to Log File Each table may be saved to your log file. Purpose of the log file - Epigram always has a log file ready for saving the results exactly as you see them on the screen. After you finish viewing a table, Epigram always asks if you want to save it to your log file. If you press 'Y' the table will be appended to the log file (added to the end) as ASCII text. At any time, you may close the current log file and open a new one with a new name. Log file location (local access) - If you are using California Epigram on a local area network or a stand-alone PC, the log file will be saved directly on a network disk or local disk. Epigram always tells you the location, such as "C:\EPIGRAM.001" Word processing hints - After you finish using Epigram, you may edit and print the log files using any word processor. Use a non-proportional font such as Courier to keep the columns aligned. If needed, adjust the page orientation, margins or font size so the text does not wrap to the next line. Directly printing the results - Local users (stand-alone PC or LAN PC) have the additional option of immediately printing an output table. After asking if you want to save to the log file, Epigram asks if you want to print the results. By the way, if you never want to print, you may disable printing from the Main Menu.

Formats for Exporting Results may be saved to a tab-delimited, database, or HTML file. Saving to a second format - After you finish viewing the results, Epigram may ask if you want to save to a tab-delimited, HTML, or dBASE III file. If you never want to save to these optional file formats, set the "Second Format" option on the Main Menu to "None". Tab-delimited file (best for spreadsheets) - As shown below, tab-delimited output has a tab between each output item, and each text item is surrounded by "double quotes". Bar graphs are not included in tab-delimited output since the purpose of tab-delimited output is usually to import into presentation software capable of more sophisticated graphics. Tab-delimited output is ideal for importing into spreadsheet software. Tab-delimited output can also be imported into other types of software, but dBASE III format is usually better. "Deaths" "Tabulated by Race-Ethnicity and Sex" "Age: 20-39 Years" "Years: 1999" "Place of Residence: Hlth Serv Area 1" "Causes of Death (ICD-10):" "ICD A00-Y99: All causes of death" "Race" "Male" "Female" "Total" "--------------------" "Am Indian" 10 4 14 "Asian/PI" 3 2 5 "Black" 5 1 6 "Hispanic" 16 3 19 "White" 159 90 249 "--------------------" "Total" 193 100 293 "Analysis Footnotes:" "Unique ID, for Keeping Track of Analyses: 507YWEGB" "Output Produced: Mon May 07 17:24:49 2001, by California VitalNet" "Deaths Classified by: ICD-10, Underlying Cause, County of Residence" "Mortality Data Source: DHS Office of Health Information and Research"

Example of Tab-Delimited Output dBASE III file (useful for many other applications) - The results may also be saved to a dBASE III database file. Database files are excellent for importing into almost any data analysis, graphics, spreadsheet, mapping or other presentation software. Field names are automatically imported along with the data. Suppressed cells are represented by the number "-1". dBASE III limitations - 1) Header information listing analysis parameters is not included. To help out, you may want to set the name of the dBASE file to the same name as the table ID, such as "405MFAQB.DBF" so you can refer to the table later. 2) No more than 128 output columns may be saved to a dBASE III file. This will rarely if ever present a problem, because an unlimited number of rows is allowed.

Explore the Data Epigram allows rapid exploratory data analysis. The analysis cycle - Exploratory data analysis means you systematically refine your analyses. Once you have made a table, look it over. If it meets your needs, save it to your log file. Possibly save the results to a tab-delimited file or a dBASE III file. Next, look over the parameters on the Main Menu, highlight the parameter you want to change, and press ENTER. A submenu will lead you through the modification process. After setting all parameters as desired, return to the Main Menu and produce another table by pressing 'P'. Try it:

At the Main Menu, modify a parameter or two. Then press 'P' to produce and view another output table. Try some of the sample analyses shown later.

Hints for avoiding misinterpretation - Double-check the header and footer of each table to verify it was the analysis you intended. Double-check the ICD codes to make sure they are the diseases and injuries of interest. Find the right balance between: Aggregation (easier to present, more stable rates). Stratification (more detail, shows differences). Acknowledge limitations in mortality data, including misdiagnosis of the cause of death, and changes in coding practices by certifiers. Acknowledge limitations in population data, including possible census miscounts, and difficulty projecting population estimates to years between censuses. Use confidence intervals to help decide if differences and trends are statistically significant. If uncertainties remain, contact staff from the Center for Health Statistics for advice.

Continue with Analyses Epigram lets you quickly carry out a wide variety of analyses. It's easy - Once you get the hang of it, using Epigram is a snap. If you have gotten this far, you should be able to carry out about any analysis you desire. Explore the menus. Don't be afraid to experiment and try out different options. Refer to the help files and to other sections of this user's guide. On-line help - Although Epigram has been designed to be as self-explanatory as possible, it also includes extensive on-line help. To access on-line help from any menu, press '?'. A help screen will appear, with information related to the current menu. After viewing the help screen, a scrolling list of help topics will appear, for your selection. Try it:

At the Main Menu, press '?' and explore the help system.

Quitting Epigram - The 'Z' or ESCAPE key always returns to the previous menu. Later, when you are done, you may press the 'Q' key from the Main Menu to quit Epigram. What other menus and options are available? - If you have done the examples in this chapter, you should have a good understanding of how to use Epigram menus to select parameters. However, you have just scratched the surface of the capabilities and power of Epigram. For your reference, the following pages list all of the California Epigram menus. The purpose of the list of menus is to help you become aware of what is available so that you can take fullest advantage of the power of the software. Do this:

Referring to the outline on the next pages, explore the menus to learn what is available.

Outlines of Submenus Using California Epigram, explore the menus listed below. Main Menu allows you to: First, design table layout: Statistic Menu Columns Menu Rows Menu Multi-Tables Menu Next, modify data variables: Age Groups Menu Area Sets Menu Cause of Death Menu Race Groups Menu Sex Menu Year Ranges Menu Other Settings from Main Menu When ready, carry out an action: Produce Table(s) Exit Epigram

Submenus for Designing Table Layout Using California Epigram, explore four menus for designing table layout.

1) Statistic Menu sets type of result: Age-adjusted death rates Crude death rates Death counts Standardized mortality ratios YPLL rates Years of potential life lost (YPLL) 2-4) Row / Column / Multi-Table Menus. Row / Column / Table for each selected: Only one row Age group Area set (may sort rows) ICD set (may sort rows) Race group Sex Year

Submenus for Modifying Data Variables, Other Settings Using California Epigram, explore menus for modifying variables.

Race Menu: Select one or more races to analyze Sex Menu: Select one or more sexes to analyze Year Ranges Menu: Select a standard set, such as 2-year Extend upper / lower end of year range Split year range into individual years Delete a year range Age Groups Menu: Select standard set, such as 10-year Extend upper / lower end of age group Split age group into separate groups Delete an age group Cause of Death (ICD) Menu: Add causes from simple menu (includes most common causes) Add from detailed menus (includes complete ICD system) Add an ICD range (includes complete ICD system) Show list of ICD codes, for display and printing Delete one / all ICD sets Combine ICD sets into new set Split an ICD set into components Area Sets Menu: Add areas / area sets (submenu) Add preset list of areas (submenu) Change area grouping type Show list of areas, possibly print Delete one / all area sets Combine area sets into new set Split an area set into components Other settings from Main Menu: Change log file for saving output Enable display of bar graphs Enable dBASE, delimited, HTML output Enable row / column percents in output Set color scheme Set confidence level, or turn off Set level of detail for unknowns / trends Set number of decimal places to show

More Practice Analyses For additional practice, and to gain more understanding of how California Epigram can speed and simplify data analysis, carry out the following sample analyses. Columns are by sex for each practice table, unless otherwise noted.

For each practice analysis, select the parameters as shown below. Then, press 'P' from the Main Menu to produce output.

1. Current leading causes of death for California:       ICD sets - NCHS rankable causes     Area sets - California     Statistic - Deaths     Row sort - Sorted high to low     Row for each - ICD set     Races - All     Table for each - One only     Years - Most recent year     2. 1999 HIV (ICD B20-B24) age-specific death rates for one county:       Statistic - Death rate     ICD sets - B20-B24     Years - 1999     Row for each - Age group     Area sets - Sacramento     Races - All     Table for each - One only     Ages - 10-year groups     3. 1999 lung cancer (ICD C33-C34) death rates ranked by area:       Statistic - Age-adjusted rate     Row for each - Area set     ICD sets - C33-C34     Adjustment standard - 2000 US     Table for each - One only     Races - All     Row sort - Sorted high to low     Area sets - All counties     Years - 1999

4: Epigram Parameters Review - In the previous chapters, you have learned what Epigram is and how to use it. Chapter 1 explained that Epigram analyzes California mortality data, listed the linked data sets, discussed access methods, and outlined confidentiality requirements. Chapter 2 described the layout and interpretation of an Epigram output table. Chapter 3 showed how to use the interface to select parameters and produce output tables. What's in this chapter - This chapter lists and explains the parameters and special options you may modify using the California Epigram interface. Epigram allows great flexibility in selecting and modifying parameters. The program allows you to mix and match parameters with unprecedented ease. For your benefit, Epigram prevents you from selecting incompatible parameters. Here are the parameters and options covered in this chapter: Age groups Causes of death (ICD sets) Cell suppression Confidence intervals Decimal places Sex Geographic selection (Area sets) Multi-tables Race groups Ranges of years Statistic Table columns Table rows Table row sort settings Trend analysis

Statistic [Outcome] The statistic is the basic type of number in a table. What is a statistic? - Every table has a statistic. The statistic (or outcome) is the basic type of data generated in an output table. Each statistic is defined in the glossary, and on-line help files give calculation methods. Reliability Indicator Statistic Number of Deaths Confidence Intervals -- Y Number of deaths Y Y Death rate Y Y Age-adjusted death rate per 100,000 Y Y Standardized mortality ratio (SMR) Y Y Years of potential life lost (YPLL) Y Y YPLL rate per 100,000 Reliability indicator - As shown in the table above, a reliability indicator (either number of deaths or confidence limits) is displayed along with most statistics. The number of deaths is normally displayed along with each rate and YPLL. Optionally, you may request that confidence intervals be displayed as a reliability indicator. See below for more information on confidence intervals.

Columns, Rows, Sorting Table columns and rows may be set however you want. Columns and rows - Every table has columns and rows. Columns and rows may be set to any of the following: Only one row / column (not set to a parameter) One row / column per selected age group One row / column per selected area set (rows may be sorted) One row / column per selected sex One row / column per selected ICD set (rows may be sorted) One row / column per selected race group One row / column per selected year Example of table rows and columns - A previous interface example shows the Main Menu with a column for each sex and a row for each race group selected. A previous output example shows the resulting output with separate columns for Male and Female, and a row for each race. Sorted rows - A table with a row for each selected area set or ICD set may be sorted: Unsorted rows: Area sets are in alphabetical order, such as Orange before Yuba. ICD sets are in alpha-numerical order, such as ICD C34 before ICD K74. Rows sorted from high to low by total data Rows sorted from low to high by total data

Multi-Tables You may automatically produce a series of tables. Deaths Tabulated by Race-Ethnicity and Sex Years: 1999 Place of Residence: Alameda ICD A00-Y99: All causes of death Race Male Female Total ------------------------------------- Am Indian 23 8 31 Asian/PI 566 460 1,026 Black 1,052 1,000 2,052 Hispanic 388 337 725 White 2,795 3,234 6,029 ------------------------------------- Total 4,824 5,039 9,863 *** Tables omitted to save space *** Deaths Tabulated by Race-Ethnicity and Sex Years: 1999 Place of Residence: Yuba ICD A00-Y99: All causes of death Race Male Female Total ------------------------------------ Am Indian 4 3 7 Asian/PI 12 9 21 Black 8 11 19 Hispanic 17 8 25 White 246 233 479 ------------------------------------ Total 287 264 551

Purpose of multi-tables - Suppose you want to make a separate table for each county. It would be tedious to select the first county, make a table, select the second county, make a table, etc. Multi-tables automates the production of such a series of tables into one operation. The multi-table setting may be one of the following: Only one table (the default) One table per selected age group One table per selected area set One table per selected ICD set One table per selected race group One table per selected year

Geo Selection [Area Sets] Every table has a geographic specification. Geographical Areas - California Epigram analyzes data to the county level. California has 58 counties, organized into 14 Health Service Areas (HSA's). Each HSA is a group of one or more counties. You may select any combination of areas - Epigram makes it easy to compare results between different groupings ("sets") of geographic areas (counties and regions), or limit analysis to specific geographic areas. One or more geographic areas may be combined into an area set, and compared with other sets. Selection is quick and easy - All geographic operations are grouped together on one submenu, to allow quick and easy selection of any combination of areas desired. A previous interface snapshot shows the main geographic menu. Predefined lists of area sets - Epigram also includes standard lists of geographic areas used in California. For example, you may automatically specify that you want one set for each county, or one set for each Health Service Area.

Causes of Death [ICD] Mortality tables always include causes of death. ICD codes - Epigram classifies causes of death with ICD codes. ICD stands for "International Classification of Diseases", the standard system for classifying causes of illness and death. Each disease or condition has a separate code (character followed by two digits) (for example, C53 for cervical cancer), or 4-digit ICD code (for example, G71.0 for muscular dystrophy). Data from 1980 to 1998 use the 9th revision of the ICD system (ICD-9). As of 1999, mortality data use ICD-10. ICD groups and sets - An ICD group is one or more consecutive ICD codes (for example, C00-D48 for neoplasms). An ICD set is one or more ICD groups combined. For example, ICD C50 for breast cancer and ICD C53 for cervical cancer may be combined into an ICD set. At the broadest level, the ICD-10 system has 19 major disease categories. At the most detailed level, there are about 7,600 four-digit codes. Selecting ICD groups - Epigram allows you to easily select and combine ICD groups to meet your analysis needs. All ICD operations are grouped together on one menu, to allow quick and easy selection of any combination of ICD groups desired. You may select ICD sets by entering the range (for example, enter C16 for stomach cancer, or B20-B24 for HIV/AIDS). Or, you may select from ICD selection menus, organized to allow you to narrow your focus without having to know the specific code. Or, you may select one of several standard ICD lists. For example, for easily calculating leading causes of death, select the NCHS list of 50 leading causes. California Epigram automatically formats leading cause reports, to save you the trouble of converting ICD codes such as "ICD E10-E14" and set numbers such as "Set #18" to descriptive terms such as "Diabetes" and "Heart Disease".

Other Variables Race Groups You may select one or more race groups to analyze. A separate submenu allows easy selection. When analyzing data by race, keep in mind that differences in health status between races may be due to socio-economic differences. Each death or population count is classified as American Indian, Asian / Pacific Islander, Black, Hispanic, or White, according to the following method: All who identify themselves as Hispanic are classified as Hispanic. The remainder are classified as Black, White, American Indian, or Asian / Pacific Islander. White also includes Other-Specified, Refused To State, and Unknown. American Indian also includes Eskimo and Aleut. Asian / Pacific Islander includes Asian-Unspecified, Asian-Specified, Chinese, Japanese, Korean, Vietnamese, Cambodian, Thai, Laotian, Hmong, Indian, Filipino, Hawaiian, Guamanian, Samoan, and Pacific Islander. Age Groups Epigram is extremely flexible and lets you combine and analyze age groups any way needed. You may select any continuous range of one or more age groups to analyze, such as 22-34. In addition, you may select a set of age ranges, such as birth-9, 10-19, 20-39, 40-64, and 65-99+. The program will let you know which age groups are available for use. Standard age groupings, such as 5-year, 10-year, and 20-year age groups are easily selected. Age groups may be combined in any way desired. California uses 1-year groups up to age 21, and 5-year groups thereafter, up to the 95-99+ group. Up to age 21, an age group may be any range of 1-year ages (for example, 3 or 11-17). For 22 and over, an age group may use any combination of 5-year groups (for example, 40-44 or 30-49). The highest group is currently 95-99+ (95 and over). Sex You may select Male, Female, or both sexes combined. Years There is currently only one year of data (1999) available, so results are always for 1999.

Other Settings Confidence Intervals Statistical reliability - Results that are based on a smaller number of events are less reliable than those based on a larger number. Confidence intervals allow you to estimate the statistical reliability of your results. Confidence interval definition - A confidence interval (also known as a confidence limits) is the range of values within which the true value of a variable is thought to occur, with a specified confidence level (95%, 90%, 80%, etc.). A higher confidence level (for example, 99%) is more stringent and results in a smaller interval than a lower confidence level (for example, 80%). Use the 95% confidence level unless you have a reason to do otherwise. Methods for confidence intervals - Epigram uses the Poisson distribution to calculate most confidence intervals, using the method described in "Scientific Tables", Diem and Lentner (ed), Geigy, 1970 (equations 802a and 802b on page 189). Confidence intervals for standardized mortality ratios are calculated using the method of Rothman and Boice described in "Research Methods in Occupational Epidemiology", Checkoway, Pearce, and Crawford-Brown, Oxford University Press, 1989, (equations 5.7 and 5.8 on page 127). Confidence intervals for trend analysis are calculating using Student's t-distribution.

Decimal Places Epigram also allows specification of the number of decimal places in your results (0, 1, 2, 3).

Cell Suppression You may suppress cells with low numbers of deaths. How cell suppression works - If desired, cells with low numbers of deaths may be blanked out with an asterisk (*). You set the level at which results will be suppressed. Bar graphs are not included if any cells are suppressed, because the bar graphs would not be accurate. Suppression of row/column totals - A row total will be suppressed if there is exactly one suppressed cell in the row or if the row total is too low. A row total will displayed if there is more than one suppressed cell in the row, unless the row total is too low. The same rules apply to column totals. Use of cell suppression - Cell suppression may be used for certain data tables to be released to the public where there is a concern that low numbers should not be published due to potential confidentiality issues. Concerns about reliability of small numbers are usually better addressed by using confidence intervals, or by including the number of deaths as a reliability indicator. Cell suppression in dBASE III files - Data saved to a dBASE III file has suppressed data indicated as the number "-1". Deaths Tabulated by Age and Sex Years: 1999 Place of Residence: San Francisco ICD B20-B24: Human immunodeficiency virus [HIV] disease Age Male Female Total ---------------------------------- Birth-9 * * * 10-19 * * * 20-29 6 * * 30-39 47 4 51 40-49 71 11 82 50-59 43 6 49 60-69 6 * * 70-99+ * * * ---------------------------------- Total 175 22 197 Analysis Footnotes: Unique ID, for Keeping Track of Analyses: 507YSPPZ Output Produced: Mon May 07 17:32:14 2001, by California VitalNet Deaths Classified by: ICD-10, Underlying Cause, County of Residence Suppress Cell [*]: If 3 or Fewer Deaths In Cell Suppress Row/Column Total [*]: If Exactly One Suppressed Cell in Row/Column Mortality Data Source: DHS Office of Health Information and Research

Glossary

Age-adjusted death rate - Deaths per 100,000 population, adjusted to a standard population (such as US 2000), by the direct method. Age-adjusted rates are often better for making comparisons than unadjusted rates, because they adjust for differences in age distribution between populations. An age-adjusted rate is a summary measure. Besides calculating overall age-adjusted rates, it is also recommended to compare age-specific rates.

Age-adjustment standard - A standard population for calculating an age-adjusted death rate. The 2000 US Census population is the recommended standard, but others are also available within California Epigram.

Area set - One or more areas combined, for example Orange and Los Angeles counties.

ASCII file - A text file, with only alphabetical, numerical, and punctuation characters, like you would see in normal text. Epigram can produce ASCII format output.

Bar graphs - A section of an Epigram table. Gives a simple graphical representation of the data. May be omitted from the output table.

Cause of death - Any condition which leads to or contributes to death and is classifiable according to the International Classification of Diseases (see ICD below).

Cell - A space for a single numerical result in a table, at a row-column intersection.

Cell suppression - An '*' is placed in cells with fewer deaths than a limit set by the user. Row/column totals with exactly one suppressed cell in the row/column are also suppressed. If more than one cell in the row/column is suppressed, the row/column total is displayed.

Columns - Vertical lists of data in an Epigram table. Each column may be one of the following: age group, area set, ICD set (cause of death), race, sex, or year.

Confidence interval (confidence limits) - A range of values within which the true value of a variable is thought to lie, with a specified level of confidence. The smaller the interval, the more reliable the result. If the 95% confidence intervals do not overlap, there is a statistically significant difference. Epigram uses a variety of methods to calculate confidence intervals.

Confidence level - The likelihood that the true value of a variable is within a confidence interval. For example, for confidence intervals at the 95% level, we are 95% certain that the actual value of the variable is within the interval.

Cursor key - Arrow, PgDn, PgUp, Home, or End key. Used to move through an output table, or to move a light bar to another option.

Data Warehouse - Software such as Epigram making large complex databases readily available for querying and analysis.

dBASE III format - A file format derived from the database software of the same name. Files in dBASE III format may be readily imported into almost any data analysis, graphing, mapping, or other presentation software. Suppressed cells are represented as the number '-1'.

Death rate - Deaths per 100,000 population. May be used to compare the burden of disease between different groups.

Denominator - The number on the bottom of a fraction. Population data are often referred to as "denominator data" as they are used to calculate rates.

ENTER key - A key on your keyboard. Sometimes called RETURN key. Often used to select an item or complete an operation.

ESCAPE key - A key on your keyboard, often on the upper left. Tells Epigram to return to a higher-level (previous) menu. Pressing 'Z' usually does the same thing.

Export - Produce output that can be read into other computer programs. Epigram exports ASCII text, HTML, tab-delimited files, and dBASE III files.

Footnote - Last part of an Epigram table. Lists other details of the analysis, such as the date and time produced, and the data sources. Each table is assigned a unique ID, listed in the footnote, to assist you in keeping track of and organizing analyses.

Header - First part of an Epigram table. Lists basic parameters you selected to define the table, such as years analyzed.

HSA - Health Service Area. Groupings of California counties. There are 14 HSA's.

ICD - International Classification of Diseases. A widely used system of classifying diseases and injuries (causes of death). Each disease or set of diseases has an ICD code or ICD group assigned to it. Epigram uses the ICD-9 system for data from 1980 to 1998. Starting with 1999 data, the ICD-10 is used.

ICD code - A letter followed by a number, representing a single disease or injury. For example, ICD E06 for thyroiditis.

ICD group - A range of continuous ICD codes, for example C00-D48 for neoplasms.

ICD set - One or more ICD groups or codes combined, for example ICD C50 (breast cancer) and ICD C53 (cervical cancer).

Import - Read information into a computer program. ASCII text, tab-delimited files, and dBASE III files from Epigram are easily imported into word processing, spreadsheet, data analysis, mapping, graphing, and other presentation software programs.

Internet - A network of networks, based upon standard protocols. California Epigram may be accessed from the Internet by using the telnet communications program, or from within a web browser.

Kermit - A file transfer protocol that may be used by remote users to download Epigram output files when finished using Epigram.

Leading causes - The ten ICD sets with the highest number of deaths, out of a standard National Center for Health Statistics list of 50 potential leading ICD-10 causes. Epigram makes it easy to select and rank the 50 cause list.

Least-squares - A standard method for fitting the best straight line to a set of points. Produces a Y-intercept and a slope defining the least-squares line.

Light bar - A highlighted area on the computer screen that you can move by pressing an arrow key or other cursor keys. The light bar highlights an item that you may select, add or delete.

Local area network - A computer networking product, such as Novell Netware or Windows NT. California Epigram may be accessed from a local area network.

Log file - A computer file for saving Epigram tables, and other Epigram output.

Main Menu - The top level Epigram menu. All parameters are summarized on the Main Menu, and you will always return to the Main Menu before producing a table.

Modem - A device for connecting to another computer over a phone line. California Epigram may be accessed by modem.

Multiple age groups - One age group for each row or column of a table, or one age group for each table in a set of multi-tables. Example: Birth-19, 20-39, 40-59, 60-99+.

Multiple cause mortality data - Data which include all causes of death listed on the death certificate. Contrast with "underlying cause mortality data".

Multiple causes of death - All diseases or injuries which led directly to death, or all circumstances of the accident or violence which produced the fatal injury.

Multi-tables - Epigram option to automatically produce a series of tables. There may be one table for each selected age group, area set, ICD set (cause of death), race, sex, or year.

Password - A personal identifier possibly required for telnet or modem access to California Epigram. Do not disclose your password to anyone else.

Place of occurrence mortality data - Data compiled by the location where the death occurred, without regard to the place of residence of the deceased. California Epigram does not currently analyze for place of occurrence.

Place of residence mortality data - Data compiled by the usual place of residence of the deceased, without regard to the location where the death occurred. Epigram analyzes mortality data by place of residence.

Population - The number of people living in an area.

Rows - Horizontal lines in an Epigram table. Each row may be one of the following: age group, area set, ICD set (cause of death), race, sex, or year.

Row sort settings - Area or ICD rows may be sorted in ascending (high to low) or descending (low to high) order, based on total data for each row.

Set - A combination of one or more things. An ICD set could include ICD C50 (breast cancer) and ICD C53 (cervical cancer). An area set could include Orange and Los Angeles counties.

Single age group - Only one age group (30-49, for example) is selected. A single age group is used for tables that do not have age columns or age rows.

Stand-alone PC - A desktop or laptop computer running off its own local hard disk. Epigram may be installed on a stand-alone PC.

Standardized mortality ratio (SMR) - The ratio of the expected number of deaths in a population to the observed number of deaths. The expected number of deaths is derived from applying a standard set of rates (usually state or national rates) to the population. SMR's are useful for assessing whether the mortality in a population is higher than expected.

Statistic - In Epigram, the basic type of number generated in a table: deaths, death rate, age-adjusted death rate, years of potential life lost (YPLL), or YPLL rate. Also called "outcome".

Submenu - A menu accessed from a higher level menu. A submenu helps select or modify a parameter listed on the higher level menu.

Table - A set of results produced by Epigram. A table has four parts: header (basic analysis parameters), data section (numerical results), optional bar graphs (simple graphical representation of the data), and footer (less important analysis parameters).

Tab-delimited format - A type of computer output that is readily imported into other software, especially spreadsheet software. Each output item is separated by a tab from surrounding items, and each output text item is surrounded by "double quotes".

Telnet - An Internet program for logging on to a remote computer. California Epigram may be accessed by telnet.

Underlying cause - The disease or injury that initiated the train of events leading directly to death, or the circumstances of the accident or violence which produced the fatal injury. A single underlying cause is assigned to each death.

Underlying cause mortality data - Data which include only the underlying cause of death listed on the death certificate. Contrast with "multiple cause mortality data".

UNIX - A widely used computer operating system. Epigram can run on UNIX. Remote access to California Epigram by modem and telnet is to a UNIX computer.

User ID - A unique user name possibly required for modem or telnet access to California Epigram. Do not disclose your user ID to anyone else.

VT100 - A widely used "terminal emulation". A terminal emulation specifies how to display information on a computer screen. Epigram telnet and modem access use vt100 emulation.

Windows - A very widely used set of PC operating systems, including Windows 3.1, Windows 95/98, and Windows NT. Epigram will run under any version of Windows.

World Wide Web - A very widely used part of the internet that may be easily accessed with a web browser. Epigram may be remotely accessed by clicking on a link from a web page.

Years of potential life lost (YPLL) - Sum of the years of life lost by persons who suffered early deaths. Early death is usually defined as death occurring before the age of 65 (the YPLL age limit). For example, death at age 40 (40.5) results in 24.5 YPLL to age 65. YPLL is a widely used measure of premature mortality.

YPLL age limit - The age used for calculating YPLL. The most common age limit is 65, but other age limits may be selected from within Epigram.

YPLL rate - YPLL per 100,000 population in the appropriate age category. For example, the YPLL rate up to age 65 is calculated as follows: (YPLL up to age 65) / (population for age group 0-64). YPLL rate is not commonly used.