The FPbase microcomputer system for managing community health screening and intervention data bases

The FPbase Microcomputer System for ManagingCommunity Health Screening and Intervention Data Bases

ANNLOUISE R. ASSAF, PhDSTEPHEN W. BANSPACH, PhDTHOMAS M. LASATER, PhDJAMES RAMSEYROBERT J. TIDWELLRICHARD A. CARLETON, MD

Dr. Assaf, Dr. Lasater, and Dr. Carleton are with thePawtucket Heart Health Program (PHHP) at Memorial Hospi-tal of Rhode Island, Pawtucket, RI, and are affiliated with theBrown University School of Medicine. Dr. Banspach, formerlyassociated with PHHP, is now with Centers for Disease Con-trol, National Center for Chronic Disease Prevention andHealth Promotion. Mr. Ramsey is Data Manager for PHHP.Mr. Tidwell is a former Program Associate.The project was supported in part by the National Institutes

of Health, National Heart, Lung, and Blood Institute, undergrant HL23629.

Tearsheet requests to Annlouise R. Assaf, PhD; PawtucketHeart Health Program, Memorial Hospital, 111 Brewster St.,Pawtucket, RI 02860; tel. (401) 729-2530; fax (401) 729-2494

Synopsis....................................

Health promotion and intervention projects atState and community levels need computerized data

bases to assist in making policy decisions and inoperating the projects. Computer data base systemsare used in entering, storing, retrieving, and analyz-ing information about health project activities andtheir participants in a timely and cost-effectivemanner. Computer support is essential for suchlabor-intensive tasks as post-screening followup ofparticipants, identifying subpopulations, and evalu-ating recruitment efforts and behavior change pro-grams.

The Pawtucket Heart Health Program developeda microcomputer software package, FPbase, forcommunity health project data base management.FPbase is described and is available for use byother organizations. FPbase incorporates formativeand process interactive data base activities and issuitable for use in operating intervention andscreening programs at State and local levels. Thesystem accommodates management of data forsocial marketing, evaluation, followup, andpromo-tional activities.

FPBASE IS A MICROCOMPUTER SOFTWARE SYSTEMfor managing data associated with operatingcommunity- or State-level health agency projectswith-intervention and screening programs.FPbase was developed at the Memorial Hospital

of Rhode Island's Division of Health Education.The system is used by the Pawtucket Heart HealthProgram (PHHP), a major community researchand demonstration study investigating communityintervention efforts related to cardiovascular dis-ease (1-3).FPbase responds to the needs of health agencies

and projects that provide the residents of a com-munity or State with intervention and screeningprograms on a limited budget. A computerizeddata base management system facilitates collectingand storing information about program partici-pants, providing timely and personalized followupmailings to persons identified as at high risk,directing promotional mailings to specific groups,and analyzing program participation rates.PHHP, funded by the Public Health Service

through the National Heart, Lurng, and BloodInstitute, is a research project designed to improvethe cardiovascular disease risk status of the resi-dents of an entire community relative to a socio-demographically similar comparison community (2,4-6). PHHP volunteers and staff members offerservices related to programs for changing healthbehaviors. Individuals, small groups, organizations,and the community at large are eligible for theservices (7, 8).The original FPbase was developed by the com-

bined efforts of PHHP staff members in the years1983-88. FPbase originally was a tool for bothformative and process evaluation. The name of thesystem reflects the initials of these two functions.As it was used and modified, many additionalfunctions were added, including such program aidsas mailing label services; followup reminders ofphysician referrals; and social marketing functions.The system consists of a specialized community

health intervention tracking program integratedwith the Paradox (A) relational data base manage-

November-Docmber 1M2, Vol. 107, No. 6 695

ment system. FPbase was tested, evolved, and usedon a mainframe computer by PHHP staff atMemorial Hospital of Rhode Island for the earlyyears. Advances in microcomputer technology, costefficiencies, and the need for registries that personswithout technical training in computers could usein tracking participants led to the development ofthe microcomputer version by the PHHP Evalua-tion Unit staff in 1988. The current version,FPbase, v. 1.73, came into use with PHHP in1988. The program is available to other organiza-tions and has been used in the New York StateHealthy Heart Programs.

System Capabilities

The system enables staff members of healthagencies, projects, and programs to collect andstore information about participants, to providetimely and personalized followup mailings to thoseidentified as at risk, to direct promotional mailingsto specific groups, and to analyze program partici-pation rates.Such comprehensive tasks are difficult or unma-

nagable without the use of a computerized database management system. As the numbers of pro-gram events and rates of participation increase, theneed grows for a system that can be used tomanage the information systematically, quickly,and economically, using nontechnical persons. Suchissues provided impetus for the development ofFPbase.FPbase is a data collection and distribution

system as well as a marketing and managementtool, allowing accurate data collection and quickanalysis for study of project-related issues.

Historical record keeping. The system maintainsrecords of participants and specific subpopulations,such as sex and age groups. Such differentiationsallow analysis of trends and determination of thereach and effectiveness of programs.

Followup mailings. Segments of the participatingpopulation can be identified to obtain additionaldata or to followup on previous findings.

Project planning. The system enhances the abilityto forecast the activity and success of future events.The availability of already collected data aboutsimilar events makes forecasting participation ratesof intended populations a practical planning option.

Management reports. Management reports can begenerated on a regular basis to reflect participationcounts and cross-tabulations by specific demo-graphic criteria.

Program accountability. Regular updates can beprovided on program progress, participation ratesat selected events, and long-term community partic-ipation rates. Management reports can show popu-lation participation rates and can cross-tabulatespecific demographic criteria. Program staff mem-bers, print and broadcast news media, and fundingagencies are audiences for the reports. Graphs ofquarterly data are used for study of the populationby sex and age groups and in determining programcoverage.

Process evaluation. The data base provides infor-mation for evaluating objectives and progress ofhealth education and health promotion activities.

Health promotion activities. The system enables aprogram to provide participants with information,announcements, and program updates. The systemhas the ability to provide mailing labels for selectedgroups with specific risk factors, such as smoking,obesity, and high blood cholesterol levels. This fea-ture is a cost-effective method of getting informa-tion to those who need it most. A program forlowering cholesterol levels of overweight women,for example, could be directed only to those meet-ing both criteria.The system can print letters automatically to be

sent to those with high-risk levels of blood pressureor cholesterol who do not return for remeasure-ment within a specified time period. One letterrecalls the circumstances of the first contact andprovides the blood pressure reading. The recipientis reminded to have a blood pressure check twice ayear and is given the times and location of thewalk-in clinic. The system provides a followupmailing to the recipient after the reading is ob-tained. This feature offers cost-effective referralfollowups to all screening programs.

Data Gathering

PHHP conducts outreach events at public gath-

696 Public Health Reports

erings and places and records contacts with partici-pants of intervention programs by using a registra-tion form, called a contact card. The one-pageform asks for basic information about the partici-pant, such as name, address, telephone number, dateof birth, and sex. Health-related questions asked are

1. Has a parent, brother, or sister had a heartattack or died of heart disease before age 55?

2. Do you have diabetes?3. Do you smoke?4. Have you ever been told you have high blood

pressure?5. Do you presently take blood pressure medica-

tion?6. Have you ever been told you have elevated

blood cholesterol?7. Do you presently take cholesterol-lowering

medication?8. Do you presently exercise vigorously for 20

minutes or more at least three times a week?9. How did you first hear about this event?

The data obtained, plus physiologic risk factorassessments, such as blood pressure and cholesterollevels, are supplemented with a precoded activityidentification number that identifies the type ofevent at which the encounter took place, the date,and the location of the event. The data for eachperson are identified by a unique alphanumericpersonal identification (PID) number assigned byFPbase at the time of data entry. The PID and theactivity identification number permit tracking par-ticipants by activity, generalizing data among activ-ities, and selecting participants by activity forfollowup by mail or telephone.

Participants in health screenings or programs areasked to complete the contact card as part of aregistration process. Early problems were the reluc-tance of staff members to require necessary infor-mation from participants and participants resistingcompleting the card. The problems lessened as staffmembers gained more understanding of the impor-tance and necessity of having respondents fill outthe contact card. After a participant fills out thecontact card, it is collected by volunteers or staffmembers and used to record risk factor assess-ments. The contact card provides background in-formation necessary to counseling and referralservices. Being able to identify behavior patternsrelated to cardiovascular risk factors permits pro-gram personnel to focus recommendations andreferral advice. After the event, the information isentered in the system.

Data Organization

FPbase has two major levels of data, the individ-ual level, focusing on the history of the specificperson, and the activity level, focusing on anactivity and cross-sectioning data around that activ-ity. Individual-level data includes name, address,and telephone number; socio-demographic charac-teristics; cardiovascular disease risk factor values;and program participation history. Activity-leveldata includes location, date, and program, with therisk factor and type of program. Each reportgenerated by FPbase is based on one of the twoviews of data.The output can be mailing labels for referral

reminders or directed mailings, reports, data filesfor statistical analyses, and sampling frames forfollowup.

Data Analysis

Case study. The following case study illustrates thesystem's operation in a cholesterol screening, coun-seling, and referral event (SCORE) that was held ata local jewelry manufacturing facility on March 15,1988. The 110 employees who attended the eventhad at entry into the program a mean blood choles-terol level of 204.3 milligrams per deciliter (mg perdl).A followup cholesterol SCORE was held there 2

months later. Of the 110 employees, 77 returned,along with 7 employees who had not attended inMarch. The 77 employees had a mean cholesterollevel of 201.56 mg per dl. The 7 new attendees hada mean blood cholesterol level of 206.57 mg per dl.The system was able to track each person's

baseline and followup cholesterol value and calcu-late the difference between the two measurements.The difference ranged from 45 mg per dl to minus95 mg per dl during the 2-month period. Thisinformation, in conjunction with demographic andself-reported risk factor data contained in theFPbase data registry, permitted individual counsel-ing of participants and aided the company inmapping out strategies for future health educationprograms.

Individual-level function. Choosing the individual-level of data allows the program to follow the in-tervention contacts with each participant. Thistechnique is an effective generalization tool usefulin determining if participants tend to enter a pro-gram only once or are repeat users of the interven-tion activities.

November-Decmber 1992, Vol. 107, No. 6 697

Data Base Operation

:~~~~~~. ...'. ..̂ ' E=. xS

....~ , .. ..

The technique allows program planners to note ifrepeat participants generalize from one risk factorto another. An example of generalization is shownby program statistics for 1985, picked as a typicalyear. In 1985, 8,251 persons joined their firstPHHP activity. Of them, 5,280 (64 percent) joineda second activity between 1985 and 1991, while2,971 (36 percent) did not. Of those joining asecond activity, 3,624 (44 percent) joined a screen-ing, counseling, and referral event; 911 (11 percent)joined a group event; 606 (7 percent) served as avolunteer; and 139 (2 percent) joined a self-helpactivity.Grouped by risk factors, this level of analysis

allows review of the number of persons whoengaged in multiple program offerings by order ofparticipation. The historical record of participantinvolvement is helpful in evaluating the dynamicsof the reach of programs within the community. Aparticularly important subpopulation that is identi-fied by this level of analysis is what PHHP refersto as new joiners, those participating in their firstPHHP program activity, rather then those whoparticipated in programs in the past. New joinersare identified by having the system search the database for previous entries for each new PID. If nomatch is made, the entry is labeled as a firstencounter. Each type of activity can be examinedin terms of socio-demographic characteristics andexpanding program reach.

Activity-level function. The activity-level functionpermits a review of participation rates across riskfactors and program intervention sites. By choosingthis analysis perspective, FPbase can compileproject information related to numbers, types ofprogram activities delivered, and sites. Activity-level function allows differentiation between typesof participants recruited for different activities atdifferent sites. It can be used to select participantsfrom a specific type of program or site for follo-wup evaluation of self-reported behavior change bymail or telephone.

Configuration. The software programs that makeup FPbase are the MS-DOS (B) operating system,the Paradox data base management system, FPbaseapplication programs, and FPbase data bases. Par-adox is the basis of FPbase. It stores data in theform of tables that are arranged in fields (columns)and records (lines). Reports may reflect data fromdifferent tables if the tables are related by a com-mon data element, such as the PID.The system uses a many-to-many multi-relational

data base scheme. The participants' personal infor-mation, such as name and address, is in a mastername file. The participants' measurements are in atransaction file, thereby simplifying confidentialitycontrols and using less computer storage by havingnames in only one place. Each time a participant isseen, a new transaction record is created for thatperson. Other files contain codes and informationon each activity. The linked data base can bequerried to provide a variety of reports.

Networking. The system functions equally well onstand-alone computers or on networks such as No-vell. When running on a network, data entry andreporting can take place simultaneously, using atemporary transaction file created at data entry.Transactions are posted to the system at the end ofthe day. In a multi-user setup, more than one per-son can enter and edit data at the same time. Witha suitable hardware configuration, authorized userscan access the system remotely to perform data en-try and access the data base to retrieve participantinformation for use in counseling and referral.

Report generation. FPbase generates reportsquickly on project activities, groups, and individualparticipants. The reports provide the program man-ager, program staff, and researchers with neededdata for evaluations and strategic planning. Thesystem creates standard reports, such as on bloodpressure and cholesterol levels, by sex, age group,participants requiring followup, and other criteria.If a modem or a laptop computer is being used, thesystem can generate an analysis report at the inter-vention site that gives frequencies and cross-tabulations of a participant's measurements andrisk factors.Each output option can be based on any one or

a combination of various types of data in the database. Options include several valuable educationalfeatures, such as generating a referral reminderletter for a person found to be at high risk with a

6S Public H"ah Rport

high level of blood cholesterol or blood pressure.Such a participant is referred to a physician imme-diately and is requested to return for a followupmeasurement at a future SCORE. If the participantdoes not return within a specified period of time,the system automatically generates a reminder letterof the physician referral that is individualized forthat respondent.Automatic reminders ensure that the important

followup responsibility of screeners will be metwithout extensive effort. The ability to maintainrecords of measured values of specific persons canbe valuable in counseling. A person's history canbe recalled from the data base and displayed whenthe counselor has access to the equipment.

Standard reports generated by FPbase can berequested and printed by the person in charge ofdata entry, or by other program personnel. Thesystem is menu-driven and does not require exten-sive training to use the report option. The ability tocreate mailing labels and to produce participantform letters is accessible through menus. For op-tional detailed analyses, FPbase data can be ex-ported to other software applications, such as SAS(C), SPSS (D), and BMDP (E). These softwareapplications can provide statistical analyses of theregistry data.

This procedure requires a staff person with thetechnical knowledge of data transfer and experiencewith statistical software. The procedure allowsquick compilation of site comparisons during anyperiod, of changes in cardiovascular disease riskfactor values, and of differences in socio-demographic or other variables among those at-tracted by specific programs at various sites, suchas schools, work sites, and organizations.For example, blood cholesterol level measure-

ment values of those who returned for a secondcholesterol measurement and those who did notmay be compared, together with differences inmeasured values in any period. The files containinformation gathered on each person and may becombined with other data sets, such as telephoneinterviews, household surveys, morbid and mortalevents, and other types of data from other sources.They may be combined with FPbase output so longas the same rules are used for forming the PID.Using Paradox's query by example (QBE) capa-

bilities, lists and counts of participants can begenerated based on any criteria, a technique that isuseful for identifying specific groups of partici-pants. For example, one can select all women ages40 through 59 years who smoke, have high bloodcholesterol levels, and are overweight. QBE can

provide reports that show how participants havegeneralized from one risk factor to another or fromone activity to another within risk factors.

System Requirements

FPbase operates on IBM AT- or 386-compatiblemicrocomputers with 1 megabyte of random accessmemory (RAM). System performance is improvedwith additional RAM. Up to 2 megabytes (MB) ofRAM can be accessed by the FPbase system. Theamount of hard disk storage space needed dependson the needs of the project and the number ofexpected program participants (each participantrecord takes up about 1 kilobyte of storage space).The operating system uses about 1 to 2 MB of harddisk space, depending on project needs for thereport or management options.

Personnel Requirements

FPbase is designed to be incorporated into aproject's daily operations. Operating the systemmay be made part of a clerical worker's jobresponsibilities. Depending on the applications ofthe system and the number of people using it, thelevel of staff expertise needed can be minimal.The volume of activities and the numbers of

program participants determines the number andtype of staff needed to run FPbase. The data entryand report functions can be handled by secretarialor clerical staff members. This activity involvestyping the contact card information into the systemand editing or updating existing information. Qual-ity control mechanisms are built into the softwareto prevent duplication of PIDs that can arisethrough use of other names, incorrect birth dates,or other sources of error. Another quality controlfeature is the ability to query the system formatching PID components following a namechange, as from marriage or divorce. The timespent updating the data base is directly related tothe number of program contacts. On an officenetworked system, data entry can take place atseparate work stations simultaneously.

Summary

FPbase is a powerful, comprehensive, and effi-cient mechanism for storing and retrieving informa-tion on participants in community health screeningand intervention programs, maintaining accurateinformation about community interventions, pro-viding staff members with detailed reports on

November-Docmber 1992, Vol. 107, No. 6 69f

program performance, and identifying populationsubgroups for health promotion direct mailings orcommunity telemarketing efforts.

References..................................

1. Assaf, A. R., et al.: The Pawtucket Heart Health Progra-m: evaluation strategies. R I Med J 70: 541-546 (1987).

2. Carleton, R. A., et al.: Pawtucket Heart Health Progra-m: an experiment in population-based disease prevention.R I Med J 70: 533-538 (1987).

3. McGraw, S. A., et al.: Methods in program evaluation:the process evaluation system of the Pawtucket HeartHealth Program. Evaluation Rev 13: 459-483 (1989).

4. Lasater, T. L., Lefebvre, R. C., and Carleton, R. A.: ThePawtucket Heart Health Program: community level pro-gramming for heart health. R I Med J 71: 63-67 (1988).

5. Carleton, R. A., and Lasater, T. M.: Primary preventionof coronary heart disease: a challenge for behavioralmedicine. Circulation 76 (supp. 1): 124-129 (1987).

6. Lefebvre, R. C., Lasater, T. L., Carleton, R. A., andPeterson, G. S.: Theory and delivery of health program-ming in the community: the Pawtucket Heart HealthProgram. Prev Med 16: 80-95 (1987).

7. Elder, J. P., et al.: Organizational and community ap-proaches to community-wide prevention of heart disease:the first two years of the Pawtucket Heart Health Pro-gram. Prev Med 15: 107-117 (1986).

8. Elder, J. P., et al.: The use of volunteers in massscreening for high blood pressure. Am J Prev Med2: 268-272 (1986).

Equipment

A. Paradox, v. 3.0 or later. Borland International, ScottsValley, CA, 1988.

B. MS-DOS, v. 3.0 or later. Microsoft Corporation, Red-mond, WA.

C. SAS, v. 6.01 or later. SAS Institute, Inc., Cary, NC.D. SPSS, v. 3.0 or later. SPSS, Inc., Chicago, IL.E. BMDP, v. 88.2. BMDP Statistical Software, Inc., Los

Angeles, CA.

Statement of Ownership, Management,and Circulation

(Required by 39 USC 3685; reported onUSPS Form 3526)

Title of publication: Public Health ReportsPublication number: DHHS Publication No. (PHS)

92-50193, USPS 324-990, ISSN 0033-3549Date of filing: October 1, 1992Frequency of issue: BimonthlyAddress of known office of publication and general

business office: Parklawn Building, Room 13C-26,5600 Fishers Ln., Rockville, MD 20857

Annual subscription rate: $12 domestic, $15 foreignPublisher: U.S. Department of Health and Human Ser-

vices, Public Health Service, Office of the AssistantSecretary for Health, Office of Health Communica-tions, Parklawn Building, Room 13C-26, 5600 FishersLn., Rockville, MD 20857

Editor: Marian P. TebbenOwner: U.S. Department of Health and Human Services,

Public Health Service, Office of the Assistant Secretaryfor Health

Known bondholders, mortgagees, and other securityholders: None

The purpose, function, and nonprofit status of thisorganization has not changed during the preceding 12months.

Actual number ofAverage number of copies of single

copies of each issue publishedissue in preceding nearest to filling

12 months dateA. Total number of copies,

net press run .......... 10,475 9,014B. Paid circulation:

Requested and countersales ................ 3,364 1,614

Mail subscription ...... 3,111 3,100C. Total paid 6,475 4,714D. Free distribution ....... 4,000 4,300E. Total distribution ........ 10,475 9,014F. Copies not distributed:

Office use ............. 0 0Returns from sales ..... 0 0

G. Total . ........... 10,475 9,014I certify that the above statements made by me are correctand complete.

MARIAN P. TEBBEN, Executive Editor

700 Public Health Reports