HL7 Connector for External Applications (HCEA)

[Solution Deployment Guide] This Solution Deployment Guide describes how to install and use the default HCEA models in the HCEA Evaluation Kit. The models included allow you to extract generic data from common segments of all HL7 messages, as well as the specialized data from the following message types:

‰ ORU-RO1 ‰ RAS This guide also describes:

‰ How to customize the default models ‰ How to easily create your own models using the pre-defined templates and functions included in the Evaluation Kit

©2013 Datawatch Corporation. All other trademarks or registered trademarks are properties of their respective owners.

HCEA Deployment Guide

Table of Contents [Overview] ................................................................................................................................ 1  Sample HL7 file (input to the HCEA Solution) ................................................ 1  Sample HCEA TABLE Output (Partial) ........................................................... 1  Sample HCEA Summary Output: Clinic Summary (Partial) ............................... 2 

The HCEA Advantage .............................................................................................................................................2 

HCEA Requirements and Files ........................................................................................... 3  Requirements ............................................................................................................................................................. 3  Monarch Professional ................................................................................. 3  Panopticon Dashboard ............................................................................... 3 

Files.................................................................................................................................................................................... 3  Models ..................................................................................................... 3  Dashboards .............................................................................................. 4 

A Note About Input Files ....................................................................................................................................... 5 

Deployment Steps ................................................................................................................ 6  1. Unzip the files to the required folder......................................................... 6  2. Load your own HL7 files. ........................................................................ 6  3. View Dashboards. .................................................................................. 7  4. Customize as necessary. ........................................................................ 7 

Customization ........................................................................................................................ 8  Monarch Models in Relation to HL7 Conventions ................................................................................. 8  General Guidelines for Creating Your Own Custom Models ........................................................ 10  1. Create a model. .................................................................................. 10  2. Link the HL7 model. ............................................................................. 10  3. Define Templates for each new segmenT ................................................ 10  4. Extract the fields of a segment. ............................................................. 10  5. Extract the components from fields ........................................................ 10 

Contact Datawatch for Model Building Services ..................................................................................11 

Appendix A HL7 Terminology and Definitions .............................................................12  HL7 Definitions ..........................................................................................................................................................12  Message ................................................................................................ 12 

Segment .................................................................................................12  Delimiters ...............................................................................................12  Field.......................................................................................................13  Item number ...........................................................................................13  Component .............................................................................................13  Data Type ...............................................................................................14 

Appendix B Linking to an External Model ..................................................................... 15  Appendix C Defining a New Template............................................................................18  Appendix D User-Defined Functions (UDFs) .............................................................. 20  Appendix E Creating User-Defined Functions .......................................................... 22 

HCEA De eployment G Guide

[Ove erview w] The HL7 Connector for Exte ernal Applications (HCEA)), powered by y Datawatch’s Information n Optimization O Platform P (IOP P), is a messa age connecto r system that translates H HL7 message es (e e.g., the ORU U message type) into a forrmat that ma ay be easily u understood an nd utilized by y no on-HL7 syste ems.

SAMPLE S HL7 H FILE (INPUT ( TO O THE HCEA SOLU UTION)

SAMPLE S HCEA H TAB BLE OUTP PUT (PAR RTIAL)

1|P a g e

SAMPL LE HCEA SUMMAR S RY OUTPU UT: CLINIC C SUMMA ARY (PAR RTIAL) This samp ple chart shows the numb ber of patients treated by Physician Gro oups broken down by Gender.

The HCEA A Adva antage e HCEA A greatly simp plifies:

‰ Ex xtraction of diagnostic d infformation from m HL7 messa ages into tables and summ maries for us se by non-HL L7 systems

‰ Accurate deterrmination of pending orde ered tests forr improved re esource and logistics management m

‰ Report genera ation/visualiza ation for vario ous third-parrty non-hospiital systems (e.g., HMOs, government health departments s, insurance ccompanies)

‰ Time-motion analysis a to de etermine way ys to reduce o operation cossts and impro ove productivity

‰ Statistical com mparisons of department d performance p on a hospitall/hospital uniit basis ‰ Collection of data and/or id dentification of o target pop ulations for cclinical trials

2|P a g e

HCEA Deployment Guide

HCEA Requirements and Files Requirements MONARCH PROFESSIONAL HCEA requires Monarch Professional. You need Monarch Professional to view and use the HCEA Reports. If you do not have Monarch Professional installed, you can download a trial version here: http://www.datawatch.com/products/information-optimization-platform/item/200-free-trial.

PANOPTICON DASHBOARD The HCEA Evaluation kit also comes with pre-defined Panopticon dashboards. If you wish to use Panopticon for visual data analysis but do not have Panopticon Dashboards installed, you can download a trial version here: http://www.panopticon.com/EvaluationSoftware-Request.

Files MODELS The HCEA Evaluation Kit includes a standard HL7 model that extracts data from common segments of all message types. In addition, two models extract information from the following message types:

‰ ORU-R01 (Unsolicited Transmission of an Observation) ‰ RAS (Pharmacy/Treatment Administration) These models are ready to use straight out of the box. You can work with full tables as well as pre-defined reports and summaries. You can export the data to Excel or Access for use in Panopticon, or for further analysis. You can also use the initial models as a basis for creating your own models to handle other message types.

FILE

DESCRIPTION

HL7.xmod/xprj

Model and project for a standard HL7 file. This model extracts relevant fields only from the following HL7 segments:

‰ MSH ‰ PID ‰ ORC

3|P a g e

FILE

DESCRIPTION The model includes the following reports/summaries:

‰ Patient Summary by State ‰ Patient Summary by Clinic ‰ Clinic Patient Demographic ORU-R01.xmod/xprj

Model and project for standard ORU-R01 (Unsolicited Transmission of an Observation) messages. Aside from extracting data from the standard HL7 segments listed above, this model also extracts fields from the following segments:

‰ OBR ‰ OBX The model includes the following reports/summaries:

‰ Tests by Ordering Physician ‰ Results per Day RAS.xmod/xprj

Model and project for a standard RAS (Pharmacy/Treatment Administration) messages. Aside from extracting data from the standard HL7 segments listed above, this model also extracts fields from the following segment:

‰ RXA The model includes the following reports/summaries:

‰ Medicines per Provider ‰ Medicines Administered

DASHBOARDS The HCEA Evaluation Kit also includes Panopticon dashboards. The dashboard provides visualization for your HL7, ORU-R01, and RAS data. Like the models, the dashboards are ready to use straight out of the box. Simply export the data from your models and the dashboards will be populated automatically. FILE

DESCRIPTION

HCEA.exw

Dashboard definition file. The dashboard contains tabs for the following:

‰ HL7 ‰ ORU-R01 ‰ RAS

4|P a g e

HCEA De eployment G Guide

Sample Dashb board for ORU U-R01 data.

A Note e Abou ut Inpu ut Filess Note that the HCEA implem mentation currrently assum mes that all in nput files will be placed in the fo older: c:\HCEA\inp puts\ n include an ny tool or cap pability to mo ove your HL7 7 files to this The implementation does not fo older. Yo ou must man nually copy yo our own files to this folde r.

5|P a g e

Deployment Steps 1. Unzip the files to the required folder. Unzip the file HCEA.zip to the Information Optimization machine’s c:\ drive. After unzipping, your drive should have the following folders: FILE

DESCRIPTION

c:\HCEA\

Contains:

‰ all of the model (xmod) and project (xprj) files. The following models/projects are available: • • •

HL7.xmod/xprj ORU-R01.xmod/xprj RAS.xmod/xprj

‰ the Panopticon Dashboard definition c:\HCEA\inputs

Contains sample files. You will need to move the actual files from your HL7 application to this folder.

c:\HCEA\exports

Contains the output reports. You need to run Project Exports to produce your own reports.

2. Load your own HL7 files. To use the solution with your own files: 1.

In Monarch Professional, open a project in c:\HCEA:

‰ HL7.xprj ‰ ORU-R01xprj ‰ RAS.xprj 2.

Load your HL7 inputs and:

‰ View the data in Monarch table view ‰ View the data in Monarch summary view, or ‰ Run the project exports to produce Excel file(s) containing your data.

6|P a g e

HCEA Deployment Guide

To run exports, select Project Exports from the Exports section of the Home tab:

If you export your data, these will be available in: c:\HCEA\exports.

3. View Dashboards. To view the dashboards: 1.

Ensure that you have exported the data from your models using the Project Exports feature.

2.

Open the Dashboard definition: c:\HCEA\HCEA.exw.

3.

View your dashboards. Select a tab to view corresponding data for:

‰ HL7 ‰ ORU-R01 ‰ RAS To maximize your visualization, view your dashboards in Presentation mode. Select View > Presentation Mode on the Panopticon Designer menu to do so.

4. Customize as necessary. This Evaluation Kit only extracts fields from the following segments:

‰ ‰ ‰ ‰ ‰ ‰

MSH PID ORC OBR OBX RXA

If you wish to extract data from other segments or message types, you will need to customize the existing models. Customization steps are described in the next section.

7|P a g e

Cusstom mizatio on HL7 defines a series of electronic messages m to support s admiinistrative, lo ogistical, finan ncial, and clinical prrocesses, amo ong others. HL7 H messages use a huma an-readable ((ASCII) non-XML encoding syntax based d on segmentts (lines) and d one-charactter delimiterss. Refer to Appendix A A: HL7 H Terminollogy and Defiinitions for de efinitions and d terminology y used in this docum ment.

Mon narch Model M s in Re elation n to HL L7 Conv ventio ons Since we know how fie elds are sepa arated and the location of each specificc field, the Mo onarch models th hat come with h the Evaluattion Kit use extraction rule es that simply expand on the HL7 standards s. For example: ‰

Each segment is extracted e as a single line, with the 3-ch haracter segm ment ID used d as the literall trap.

In this example, we w identify alll PID segmen nts by using tthe segment ID text PID as the literall trap. ‰

Each field on the segment s is co omputed usin ng the followi ng formula: it(, >,,) Thus, to compute for the Patient Name field d (PID field # #5), we use the following formula: lspli it(PID,38,"| |",5) To de efine the 6th field, we use: lspli it(PID,38,"| |",6) And so s on.

‰

The segment fields are named using the following conve ention: - Thus, on a PID Segment: •

th he first field is PID-1

8|P a g e

HCEA Deployment Guide

•

the second field is PID-2

•

and so on

For easier extraction, User-Defined Functions (UDFs) have been created for several of the common HL7 fields and data types. For instance, UDFs have been defined to extract the Name components from an XPN (Extended Person Name) data type. Example: PID-5 is defined as the Patient Name, with data type XPN. If we extract PID-5, we will obtain a strange-looking string that looks like this Johnson^Tomas^V^Jr^Mr^^^ Since this field has an XPN data type, we can use the XPN-related UDFs to extract name components, or the real name values: To get

Use

Sample Value

Last Name

XPNFamily([PID-5])

Johnson

First Name

XPNGiven([PID-5]

Tomas

Name

XPNName([PID-5]

Johnson, Tomas V Jr

Refer to Appendix D: User-Defined Functions for a list of UDFs you can use. ‰

As described in previous sections, a standard HL7 model (HL7.xmod) is included with the Kit. This model only extracts standard fields from the following standard segments: • • •

MSH PID ORC

It also contains the standard UDFs for several data types. Each time you create a new model, you can simply link to HL7.xmod to include the template extraction rules for the segments above, or use the UDFs in that model. For instance: •

The ORU-R01 model links to the standard HL7 model. In addition new rules were added to extract data from the new segments OBR and OBX.

•

The RAS model links to the standard HL7 model. Ln addition, new rules were added to extract data from the new segment RXA.

As long as you know the number of fields within a segment and the field number of the specific field you want to include, you can easily create your own models for any message type and/or segment.

9|P a g e

General Guidelines for Creating Your Own Custom Models To define model for your own models for a specific message type:

1. CREATE A MODEL. You can either:

‰ Create a new model, or ‰ Use an existing model and save it as a new file.

2. LINK THE HL7 MODEL. Link the HL7 (HL7.xmod) model to include standard HL7 segments and fields. You only need to do this step if you created a new model. Steps for linking to an External Model are described in Appendix B: Linking to an Existing Model.

3. DEFINE TEMPLATES FOR EACH NEW SEGMENT Define new templates for each new segment required by the message type. Extract each segment you wish to include as a single line and then use the 3-character segment ID used as the literal trap. For example, to build the RAS model, we define the template for the RXA Segment. Details are described in Appendix C: Defining a New Template.

4. EXTRACT THE FIELDS OF A SEGMENT. Follow the conventions described earlier. You can easily extract fields using the LSPLIT function. Use the form: lsplit(,,,) Thus, to extract the Patient Name field (PID field #5), we use the following formula: lsplit(PID,38,"|",5)

5. EXTRACT THE COMPONENTS FROM FIELDS Use UDFs where applicable. UDFs have been defined for the following common data types: ‰

XPN (Extended Person Name)

‰

XAD (Extended Address)

‰

DTM (Date/Time)

‰

XON (Extended Composite Name and ID Number for Organizations)

10 | P a g e

HCEA Deployment Guide

‰

XTN (Extended Telecommunication Number)

‰

XCN (Extended Composite ID Number and Name for Persons)

‰

CWE(Coded with Exceptions)

Refer to Appendix D: User-Defined Functions for a complete list of UDFs. Refer to Appendix E: Creating User-Defined Functions for instructions on how you can create your own User-Defined Functions.

Contact Datawatch for Model Building Services If you want Datawatch to help you build your models, we provide Model Building Services. Visit our website for more details: http://datawatch.com/customer-support/services/monarchmodel-building

11 | P a g e

Appendix A HL7 Terminology and Definitions HL7 Definitions The following are standard HL7 definitions that are also used throughout this document.

MESSAGE A message is the entire unit of data transferred between systems in a single transmission. It is a series of segments in a sequence defined by message specifications. Example: MSH|^~\&|SurroundLab|ACME^33D0159701^CLIA|||201304291412||ORU^R01|6294132|P|2.3.1| PID|1|376824412|1431763|D2097296| JOHNSON^TOMAS^V^Jr^Mr^^^||19620330|M|||^^^^||917-449-3144|||||37682|| ORC|RE||||F||||||||SLAB||||||||Tomas & Ruden M.D.s^^^^^CLIA|201 East 10th Street^^New York^NY^10065… OBR|1||D2097296|22002022^Hemoglobin A1c^L|||201304180001|||||||201304182211||^ALLEN,KAMILA… OBX|1|NM|17856-6^Hemoglobin A1c/Hemoglobin.total^LN||5.9|Percent|4.1-5.6|H|||F|||201304260020|

This is a fragment of an HL7 Observation Record (ORU) message. It has 5 segments.

SEGMENT A segment is a logical grouping of data fields. Segments within a defined message may be required or optional, may occur only once, or may be allowed to repeat. Each segment is named and identified by a segment ID, a unique 3-character code. Example: PID|1|376824412|1431763|D2097296| JOHNSON^TOMAS^V^Jr^Mr^^^||19620330|M|||^^^^||917-449-3144|||||37682||

This is the PID Segment.

DELIMITERS Delimiter characters are used to separate segments, fields, and components in an HL7 message. Messages used in this guide and in the Evaluation Kit use the following delimiters: Delimiter Character

Usage

|

field separator

^

component separator

&

sub-component separator

~

repetition separator

\

escape character

12 | P a g e

HCEA Deployment Guide

FIELD A field is a string of characters that make up a specific data type. Each field is identified by the segment in which it is included as well as its position within the segment. Example: PID|1|376824412|1431763|D2097296|JOHNSON^TOMAS^V^Jr^Mr^^^||19620330|M|||^^^^||917-449-3144|||||37682||

The PID Segment has 39 fields. The first 14 are as follows: Field No.

Description

Type

Repeats

Required

Len

1 2

Set ID - PID Patient ID

SI CX

-

No No

4 0

3

Patient Identifier List

CX

(1)

Yes

250

4

Alternate Patient ID - PID

CX

(1)

No

0

5

Patient Name

XPN

(1)

Yes

250

6

Mother's Maiden Name

(1)

(1)

No

250

7

Date/Time of Birth

DTM

-

No

24

8

Administrative Sex

IS

-

No

1

9

Patient Alias

XPN

(1)

No

0

10

Race

CWE

(1)

No

705

11

Patient Address

XAD

(1)

No

250

12

County Code

IS

-

No

0

13

Phone Number - Home

XTN

(1)

No

250

14

Phone Number - Business

XTN

(1)

No

250

Note that fifth field is the Patient Name field. A field is bounded by the first delimiter, a | (pipe) character. A maximum field length is stated as normative information. Exceeding this listed length, however, is not considered an error.

ITEM NUMBER As displayed on the table above, each field is assigned a unique item number. Example: PID|1|376824412|1431763|D2097296|JOHNSON^TOMAS^V^Jr^Mr^^^||19620330|M|||^^^^||917-449-3144|||||37682||

The Patient Name field on a PID Segment has the item number: 5.

COMPONENT A component (or sub-field) is one of several logical groups of items that comprise the contents of a coded or composite field. A field is bounded by the second delimiter, a ^ (caret) character. Within a field having several components, not all components may have values. Example: PID|1|376824412|1431763|D2097296|JOHNSON^TOMAS^V^Jr^Mr^^^||19620330|M|||^^^^||917-449-3144|||||37682||

13 | P a g e

The Patient Name field has eight components: • • • • • • • •

family+last name (FN) given name (ST) middle initial or name (ST) suffix (e.g., JR or III) (ST) prefix (e.g., DR) (ST) degree (e.g., MD) (IS) name type code (ID) name representation code (ID)

The Patient Name field in the example only has five components with values: the family name, given name, middle initial, suffix, and prefix. The rest are blank or optional.

DATA TYPE A data type restricts the contents and format of the data field. Data types are given a 2- or 3letter code. Some data types are coded or composite types with several components. The applicable data type is listed and defined in each field definition. Example: PID|1|376824412|1431763|D2097296|JOHNSON^TOMAS^V^Jr^Mr^^^||19620330|M|||^^^^||917-449-3144|||||37682||

The Name field on a PID Segment has a data type of XPN (Extended Person Name).

14 | P a g e

HCEA De eployment G Guide

Appe A endix xB Linkin L ng to an a Extternal Mod del To o link to an external e mode el, use the Ex xternal Mode ls wizard. Wh hen you link to a model you ca an import tha at model’s tem mplates and functions, an nd use these like they are part of the cu urrent model. Steps: 1.

Select Extternal Models from the Data D Definittions group o of the Home ttab.

The Ex xternal Modells dialog disp plays.

2.

Select New w.

15 | P a g e

The External Model M dialog displays. d

3.

Select the HL7 mo odel from c:\ \HCEA. Provide a name an Expose func ctions nd click the E check kbox.

4.

Select OK on both h dialog boxes s.

5.

s from the Re eport tab, sele ect the Impo rt link. Select Templates

The Import Temp plate – Source e Models diallog displays.

6. Click on Default and a HL7 and select Next.. An Import T Template – S Select dialog b box 16 | P a g e

HCEA De eployment G Guide

displays.

7.

Click on Select S All and then Next. The Import Template – Type of Link dialog displays.

8.

Select No one and then select Finish.

17 | P a g e

App pendix C Defiining a New w Tem mplate e Follow the ese instructio ons to define a new template for a new w segment. Steps 1.

Load your report in Monarch Prrofessional.

2.

ou wish to ex xtract data frrom. Double-click on the segment yo w double-clic ck on the RXA A Segment. In this example, we

The Template T Defiinition dialog displays:

3.

Do the following: •

Use U the segm ment’s 3-chara acter code as s the identifie er for the seg gment.

•

Set S the role to: Detail.

•

Use U the segm ment’s 3-chara acter code as s the Templatte name

18 | P a g e

HCEA Deployment Guide

4.

Select OK.

5.

Define the fields.

19 | P a g e

Appendix D User-Defined Functions (UDFs) Use the following UDFs for the XPN (Extended Person Name) data type: Function

Description

XPNFamily

Extracts the Family name from a Name field (component no: 1)

XPNGiven

Extracts the Given name from a Name field (component no: 2)

XPNMiddle

Extracts the Middle name from a Name field (component no: 3)

XPNSuffix

Extracts the Suffix from a Name field (component no: 4)

XPNDegree

Extracts the Degree from a Name field (component no: 6)

XPNName

Creates a Full Name field by concatenating: XPNFamily, XPNGiven, XPNMiddle, and XPNSuffix. The result will have the form: FamilyName, GivenName MiddleName Suffix Example: Doe, John Martin Jr.

Use the following UDFs for the XAD (Extended Address) data type: Function

Description

XADAddress1

Extracts the first Address part from an Address field (component no: 1)

XADAddress2

Extracts the second Address part from an Address field (component no: 2)

XADState

Extracts the State from an Address field (component no: 4)

XADCity

Extracts the City from an Address field (component no: 3)

XADZip

Extracts the Zip Code from an Address field (component no: 5)

XADCountry

Extracts the Country from an Address field (component no: 6)

XADAddress

Create a full Address field by concatenating the individual address fields.

Use the following UDFs for the DTM (Date/Time) data type: Function

Description

DTMDate

Extracts the Date part from a Date+Time field.

DTMTime

Extracts the Time part from a Date+Time field.

20 | P a g e

HCEA Deployment Guide

Use the following UDF for the XON (Extended Composite Name and ID Number for Organizations) data type: Function

Description

XONName

Extracts the Name from a Name field (component no: 1)

Use the following UDF for the XTN (Extended Telecommunication Number) data type: Function

Description

XTNPhone

Extracts the Phone number from a Phone field (component no: 1)

Use the following UDFs for the XCN (Extended Composite ID Number and Name for Persons) data type: Function

Description

XCNID

Extracts the ID from a Name field (component no: 1)

XCNLast

Extracts the Last Name from a Name field (component no: 2)

XCNFirst

Extracts the First Name from a Name field (component no: 3)

XCNMiddle

Extracts the Middle Name from a Name field (component no: 4)

XCNSuffix

Extracts the Suffix from a Name field (component no: 5)

XCNPrefix

Extracts the Prefix from a Name field (component no: 6)

XCNDegree

Extracts the ID from a Name field (component no: 7)

Use the following UDFs for the CWE (Coded with Exceptions) data type: Function

Description

CWEIdentifier

Extracts the Identifier from a CWE field (component no: 1)

CWEText

Extracts the Text from a CWE field (component no: 2)

CWECodingSystem

Extracts the Coding System from a CWE (component no: 3)

21 | P a g e

App pendix E Crea ating Userr-Defiined F Functiions Monarch allows a you to o create your own customized function s, called user-defined fun nctions, which you u can then us se in calculate ed field, filterr and find exp pressions. In Monarc ch, with its ability to impo ort and link objects, you ccan easily sha are your userr-defined functions between models and othe er users. To make you HL7 mod dels easier to maintain, we recommend d that you: ser-Defined Functions for standard/gen neric segmen nts in the HL7 7 generic mod del Define Us (HL7.xmo od) Define Us ser-Defined Functions for specific segm ments of speccific message type in the m model for that mess sage type. Steps: 1.

On th he Table tab, Data group, select the drop-down d bu utton for the Functions to ool to displa ay the Functio ons menu:

2.

Click the New link k.

22 | P a g e

HCEA De eployment G Guide

The User-D Defined Func ction Definitio on dialog disp plays:

3.

In the Nam me field, ente er a name forr the user-de fined field. The name you u specify will be added to the t Functions s list on the Formula F tab of the Filter D Definition dia alog, the Formula tab of the Field Propertties dialog forr a Calculated d Field, and tthe expressio on builder of tthe Search Panel.

4.

In the Des scription field, enter a des scription of w hat the functtion will do. T This descriptio on will appear at the botto om of the Formula tab w hen configuring a filter, search, or calculated field express sion.

5.

Click the Add A button to o display the Edit Form di alog:

With this dialog, d you ca an specify the e parameterss and expresssion for the u user-defined 23 | P a g e

function. If you want to create different forms of the function, use the Add button to create a new form. 6.

Specify the Return Type for the function, either character, date/time or numeric.

7.

Use the Formula Tab to enter your expression. If you specified any parameters, they will be available in the left hand Parameters box, sorted alphabetically. Use the lsplit function again to extract the component from a segment field. Example: For instance, we have extracted the Patient Name from a PID Segment and named it PID5 whose data type is XPN. We want to define User-Defined Functions to extract the following components of PID-5: Component #

Description

1

family+last name (FN)

2

given name (ST)

3

middle initial or name (ST)

4

suffix (e.g., JR or III) (ST)

5

prefix (e.g., DR) (ST)

6

degree (e.g., MD) (IS)

7

name type code (ID)

8

name representation code (ID)

To define a User-Define Function that extracts first name (component #1), you can use the following expression: LSplit([PID-5],8,"^",1) To define a User-Define Function that extracts given name (component #2), you can use the following expression: LSplit([PID-5],8,"^",2) 8.

Once your expression is complete, you can either click OK to return to the User-Defined Function dialog, or use the Parameters Tab to test the expression. Click OK when you are satisfied with your function.

9.

Under the Categories heading, select the appropriate check box(es) to specify the desired category (or categories) for the user-defined function to appear under in the Formula tab of the Expression editor.

10. It may be very useful to add a comment to the user-defined function to explain its workings and purpose, especially if other users will be using it as well. To do so, click the Comment button to display the Documentation Comment dialog. 11. Click OK to save the changes and close the dialog (or click Cancel to close the dialog without saving any changes). „ EOF

24 | P a g e