Smoking Cessation

Design and testing of an interactive smoking cessation intervention for inner-city women

Anna M. McDaniel 1–3,5

, Gail R. Casper 4 , Sondra K. Hutchison

1

and Renee M. Stratton 2,3

Abstract

The purpose of this study was to design and test the usability of a computer-mediated smoking cessation program for inner-city women. Design and content were developed consistent with principles of user-centered design. Formative and summative evaluation strategies were util- ized in its testing. The summative evaluation was designed to test usability in a naturalistic envir- onment. A sample of 100 women who receive care at an inner-city community health center participated in the study. Average time for com- pleting the computer program was 13.9 minutes. Participants reported a high level of satisfaction with usability of the program. Standardized in- struments to measure cognitive processes of change related to smoking were completed at baseline and at 1 week. Participants reported a decrease in favorable attitudes toward smok- ing (P 5 0.014) and an increase in cognitive change processes at follow-up (P 5 0.037). These results indicate that interactive computer tech- nology is acceptable to, and potentially useful for, promoting smoking cessation in low-income women.

Introduction

Routine smoking cessation intervention by health

care providers as brief as 3 minutes can significantly

increase quit rates (Fiore et al., 2000). However, studies have shown that as few as one in four

smokers are advised to quit smoking by their health

care providers (Jaen et al., 1998; Thorndike et al., 1998). The most common reason for this omission

in primary care is the restricted time available for

preventive health teaching (McBride et al., 1997). Multimedia instructional technology is an innova-

tive and efficient method to deliver smoking cessa-

tion information in the primary care setting, and

may increase motivation to quit.

Computerized patient education is becoming

more common in clinical settings, yet questions

about patient acceptance and degree of efficacy with

naı̈ve users persist. Systematic reviews of computer-

based patient education have shown that interactive

computer interventions are effective in increasing

knowledge (Lewis, 2003). Patient receptivity to

computerized education is reported to be high across

diverse medical conditions and age groups (Revere

and Dunbar, 2001). Although access to computers

among low-income populations has increased (Na-

tional Telecommunications and Information Ad-

ministration, US Department of Commerce, 2000),

little is known about the acceptability and efficacy of

interactive technology for influencing behavior

change in this group (Stanley, 2003).

Computer technology has been used to enhance

smoking cessation by incorporating tailored infor-

mation into traditional self-help materials, which

have been shown to be more effective in promoting

smoking cessation than non-tailored educational

Schools of 1 Nursing and

2 Informatics, Indiana University,

Indianapolis, IN 46202, 3Walther Cancer Institute,

Indianapolis, IN 46202 and 4School of Nursing, University

of Wisconsin, Madison, WI 53792, USA 5 Correspondence to: A. M. McDaniel;

E-mail: [email protected]

HEALTH EDUCATION RESEARCH Vol.20 no.3 2005

Theory & Practice Pages 379–384

Advance Access publication 2 November 2004

Health Education Research Vol.20 no.3, � Oxford University Press 2004; All rights reserved doi:10.1093/her/cyg135

by guest on A pril 19, 2014

http://her.oxfordjournals.org/ D

ow nloaded from

resources (Shiffman et al., 2000; Etter and Perneger, 2001; Prochaska et al., 2001). These computer interventions were used to provide written feedback

to smokers over time, but were not presented in

real-time (i.e. printed reports were mailed to par-

ticipants at a later date) and thus did not capitalize

on the ‘teachable moment’ of the primary care

encounter.

Usability testing is a widely used technique to

assess the utility of a system or product to accom-

plish a goal from the perspective of the end-user

(Wichansky, 2000). Important features of an in-

teractive program that contribute to usability in-

clude such factors as ease of learning, efficiency of

use, user satisfaction and utility for which the

application is intended (US Department of Health

and Human Services, 2004). Usability testing is

a standard process in the information technology

industry (Nielsen, 1993), but may be overlooked in

non-commercial software development for con-

sumer health education. The purpose of this study

was to design and test the usability of an interactive

computer-mediated smoking cessation program for

inner-city women.

Methods

Program design and content

The program design and content were developed

using principles of user-centered design (Nielsen,

1993) and the Persuasive Health Message frame-

work (Witte et al., 2001). According to this frame- work, messages must address constant and transient

factors to effectively motivate health behavior

change. Constant factors consist of health threat balanced with an efficacy message, structure and

organization of the message, and demographic/

cultural profile of the targeted audience. Transient factors vary by audience, and include salient beliefs

about the perceived health threat and response

efficacy.

In designing the content of the program mes-

sages, we were guided by the findings of our focus

groups with women who were demographically

similar to the target audience (McDaniel et al., 2002). According to these women, key issues about

smoking cessation were: lack of social support, low

self-efficacy for quitting, stressors associated with

expectations of women and concerns about weight

gain. Participants expressed a preference for learn-

ing about diagnosis and treatment options from

health care professionals, but wanted information

about instrumental behavior change such as going

on a diet or exercising from ‘real people’.

Program interface and navigation

The program was designed to deliver tailored smok-

ing cessation messages in a relevant and acceptable

format, using Macromedia Director 7.0 software.

All screen text was simultaneously presented in

audio format to increase comprehension in lower

literacy participants. The program was navigated by

using a touch-screen monitor. Based on data the user

entered, algorithms derived from the scientific litera-

ture on smoking cessation and behavior change

determined program content and progression

through a series of customized output screens.

Navigation within the program was based upon

the principles of the Transtheoretical Model of

Behavior Change (TTM), which proposes that long-

term smoking cessation is a sequential process in-

volving five stages of change: precontemplation,

contemplation, preparation, action and mainten-

ance, as determined by the individual’s readiness

to quit (Prochaska and DiClemente, 1983; Prochaska

et al., 1991). Because the goal of this intervention was to motivate readiness to quit among women

who were current smokers, the program navigated

through only the first three stages of change. Within

each branch, the user received stage-appropriate

messages about the health effects of smoking and

strategies for quitting that addressed individual

smoking motives and concerns.

Program evaluation

Formative and summative evaluation strategies

were utilized in the usability test design. Experts

in smoking cessation and women’s health reviewed

program content throughout development. The

A. M. McDaniel et al.

380

by guest on A pril 19, 2014

http://her.oxfordjournals.org/ D

ow nloaded from

design team created a prototype of the program for

analysis by health care professionals familiar with

intended users. Formative feedback was incorpor-

ated into the design of the final program.

A summative evaluation strategy tested usability

in a naturalistic environment. Institutional Review

Board approval was obtained to recruit participants

from women at a neighborhood community health

center primarily serving the medically indigent

while awaiting their clinic appointments. The com-

puter and touch-screen monitor were contained

within a portable cabinet to provide mobility within

the clinic. Participants who met inclusion criteria

were taken to a private area to complete all study

procedures. After giving informed consent and

completing paper-and-pencil baseline measures,

participants were instructed on how to launch the

interactive computer program. A trained research

assistant observed the participants during use of the

program and recorded a summary statement of each

use. After finishing the program, participants com-

pleted a usability survey. Participants were con-

tacted by telephone after 1 week to assess changes

in cognitive measures and selected behavioral

indicators of the utility of the program.

Usability of the interactive program was assessed

using a 14-item investigator-developed tool mea-

suring satisfaction with usability of the program

including ease of using the program (three items),

efficiency (one item), comfort with using a com-

puter (three items), and acceptability of a compu-

terized learning format (seven items). In addition,

two open-ended questions were included to obtain

individual user preferences. Cognitive measures of

the utility of the program were assessed using two

well-established instruments based on the precepts

of the TTM: the Decisional Balance Scale (Velicer

et al., 1985) and the Processes of Change–Short Form (POC-SF) (Prochaska et al., 1988). See Table I for information on the study scales.

Data analysis plan

Descriptive statistics were used to examine demo-

graphic and smoking variables. Relationships

among participant characteristics and satisfaction

with program usability were examined using t-tests

and one-way analysis of variance. Repeated meas-

ures analysis of variance was used to detect changes

in cognitive factors associated with smoking from

baseline to follow-up. A v2 analysis was used to detect change in participants’ stage of change

for quitting.

Results

Sample

A non-probability sample of 110 women smokers

was recruited from 228 eligible women approached

in the clinic. Time constraint was the most fre-

quently cited reason for refusal to participate. The

final sample consisted of 100 women (91% of those

enrolled) between the ages of 18 and 71, with

a mean age of 41.5 years (SD = 12.4). Two subjects

left the clinic before completing the computer

program and eight were lost to follow-up. The

majority (68%) of the participants were Caucasian.

Educational level was typical of an inner-city

population, with 35 participants reporting that they

did not complete high school. Participants reported

a long history of smoking (M = 22.8 years, SD = 12.5) with a mean age of smoking onset at 16.7

years (range 9–36 years).

Usability

Observational data

Initially, we observed several problems with the

navigation and interface of the interactive program.

Occasionally, some content inappropriate to the

user was being delivered onscreen. The program

was modified by re-specification of the control

algorithms to ensure appropriate messages were

delivered to the remaining 85 users; however, the

data from the first 15 participants were retained in

order to minimize bias of the results.

Average time required to complete the program

was 13.9 minutes (range 7.2–88.8 minutes). Field

notes revealed that the outlier (88.8 minutes)

occurred due to interruptions while the participant

completed the program. No one failed to complete

An interactive smoking cessation intervention for inner-city women

381

by guest on A pril 19, 2014

http://her.oxfordjournals.org/ D

ow nloaded from

the program; only one subject requested assistance

with using the touch-screen monitor.

Usability survey results

Participant rating of satisfaction with usability of

the program was high. The mean score on this

measure was 60.2 (SD = 6.3) with a possible range

of 14–70 (actual range 40–70). There were no rela-

tionships between usability scores and age, health

status or prior computer experience. Minority par-

ticipants were significantly more satisfied with

usability of the computer program than Caucasian

participants. Satisfaction with usability was signifi-

cantly higher in women with at least a high school

education. Stage of change was significantly related

to usability scores, with women who were only

thinking about quitting (precontemplation stage)

being least satisfied with using the program. See

Table II for comparison of significant relationships

with satisfaction with usability.

Content analysis of the response to open-ended

questions revealed several positive themes: ease of

use (n = 52), quality of information (n = 18), use of technology (n = 15) and ‘fun’ (n = 6). Four par- ticipants noted that the best feature was the non-

threatening approach to smoking cessation. One

woman responded that she liked the privacy afforded

by the interactive program.

Changes in cognitive measures

After using the program, participants reported

a significant decrease in favorable attitudes toward

smoking, as measured by the Pros of Smoking

subscale of the Decisional Balance Scale, regard-

less of stage of change (RMANOVA, F = 6.31, P = 0.014). There was no significant difference in

negative perceptions of smoking (i.e. ‘Cons’ sub-

scale). There was a significant increase in scores on

Table I. Psychometric properties of study measures

Instrument No. items Response format Content/concept Reliability a

Usability survey

(McDaniel et al., 2002) 14 five-point Likert-type satisfaction with

program usability

Cronbach’s

a = 0.76open-ended items: ‘What did you like

best about using

program?’; ‘How

could the program

be better?’

free response

Decisional Balance Scale

(Velicer et al., 1985)

20 five-point Likert-type decision making about

smoking: Pros and

Cons subscales

Cronbach’s a Pros = 0.87 (0.84);

Cons = 0.90 (0.83)

POC–SF

(Prochaska et al., 1988)

20 five-point Likert-type cognitive processes associated

with quitting smoking:

Experiential and Behavioral

subscales

Median Cronbach’s

a = 0.88 (0.87)

a Reliability in parentheses indicate this study

Table II. Sample characteristics and relationship to

satisfaction with program usability (N = 100)

Characteristic Mean (SD) a

Test statistic P

Ethnicity t = 2.156 0.034

Caucasian (n = 68) 59.1 (5.92)

minority (n = 32) 62.0 (6.68)

Educational level F = 4.332 0.016

did not complete high

school (n = 35) 57.7 (6.08)

high school graduate

(n = 38)

61.0 (6.47)

attended college (n = 27) 61.8 (5.49)

Stage of change F = 10.15 <0.00 precontemplation (n = 19) 55.7 (5.47)

contemplation (n = 68) 60.4 (6.14)

preparation (n = 13) 64.9 (3.73)

a Range of scores across entire sample was 40–70.

A. M. McDaniel et al.

382

by guest on A pril 19, 2014

http://her.oxfordjournals.org/ D

ow nloaded from

the Experiential subscale of the POC-SF (e.g.

substituting another behavior when tempted to

smoke, consciously avoiding smoking ‘triggers’)

at follow-up (RMANOVA, F = 4.47, P = 0.037) across the three stages of change. Overall, 15% of

participants progressed at least one stage of change

after completing the program (v2 = 88.8, d.f. = 6, P < 0.001).

Discussion

The purpose of this study was to design and test

the usability of an interactive computer-mediated

smoking cessation intervention for inner-city

women. Incorporating user beliefs into content de-

livery is critical to the design of an effective in-

teractive patient education software program. The

findings of this study suggest that information

technology has potential for delivering brief smok-

ing cessation intervention for low-income women in

primary care.

The application of focus group findings as a user-

centered design strategy resulted in important in-

formation for targeting content and delivery of

tailored smoking cessation messages. For example,

findings from the focus group indicated that pref-

erence for source of information was dependent

upon message content. Results of the usability

survey and open-ended feedback suggest that in-

corporating this information into the program was

successful. Eliciting user input in designing in-

teractive consumer health products is essential to

maximize impact and acceptability.

Users found the program easy to use via the

touch-screen interface. Although a research assist-

ant was available to assist participants in using the

program, only one of the users required assistance

with the touch-screen input process. Participants

were highly satisfied with the application as in-

dicated by the usability survey results. Although

most usability studies employ strictly direct obser-

vational methods with a small purposive sample in

a laboratory setting, this study examined evidence

from multiple sources to support usability. We

evaluated time to complete the program, field notes

and reported satisfaction with a relatively large

sample of end-users in a naturalistic environment,

i.e. patients within the clinic settling. We believe

this methodology is critical for testing usability in

the target population of low-income women, who

have less experience with interactive technology.

Another important factor in usability is the utility

of the program for achieving intended goals. In this

case, the purpose of the interactive program was to

motivate cognitive and behavior change in smoking

behaviors. Assessment of the short-term cognitive

and behavioral indicators of change demonstrated

progress in cognitive processes of change and

decisional balance after use of the interactive

computer program. Participants reported engaging

in more behavior-oriented strategies, such as seek-

ing help from others, removing smoking cues from

the environment or avoiding tempting situations.

The findings suggest that this method may be

effective for presenting motivational content.

The results of this study must be interpreted with

caution due to the limitations of the study design,

particularly the lack of a control group. Without

a control group, the impact of the interactive pro-

gram cannot be evaluated. In addition, long-term

effects of the program cannot be determined because

the follow-up period was brief (1 week). The

relatively high refusal rate also dictates cautious

interpretation of the findings. It is likely that women

who were less motivated to quit smoking had higher

rates of refusal, resulting in a biased sample. Never-

theless, this study supports the use of interactive

computer technology as a feasible method for

motivating smoking cessation in low-income

women. Based on our findings, future studies em-

ploying randomized controlled trial methods are

needed to determine both efficacy and long-term

effects of the intervention on smoking cessation.

Implications for practitioners

The women in this study reported a high level of

satisfaction with usability of the program despite

relatively little experience using computers. Usabil-

ity testing is an important dimension of evaluation

An interactive smoking cessation intervention for inner-city women

383

by guest on A pril 19, 2014

http://her.oxfordjournals.org/ D

ow nloaded from

for such technologically based health interventions.

Usability testing incorporating evidence from mul-

tiple sources may enable practitioners to have

greater confidence in the efficacy of such interven-

tions in the clinical arena.

Low-income women, such as the participants in

this study, have high rates of smoking and limited

access to resources to assist with cessation. An

important implication of this study is that a theoret-

ically derived, interactive program can provide

brief smoking cessation intervention in the primary

care setting with minimal burden to users. The

results of this study indicate that computer technol-

ogy can be used to deliver tailored information

intended to change health behaviors in underserved

populations.

Acknowledgements

The authors would like to acknowledge the con-

tributions of Raymond T. Ford for programming

assistance, and Drs William Tierney and Clement J.

McDonald of the Regenstrief Medical Informatics

National Library of Medicine Fellowship Program.

This study was conducted in affiliation with the

Mary Margaret Walther Program for Cancer Care

Research, Walther Cancer Institute.

References

Etter, J.F. and Perneger, T.V. (2001) Effectiveness of a computer- tailored smoking cessation program: a randomized trial. Archives of Internal Medicine, 1611, 2596–2601.

Fiore, M.C., Bailey, S.J., Cohen, S.J., et al. (2000) Treating Tobacco Use and Dependence: Clinical Practice Guideline. USDHHS, Rockville, MD.

Jaen, C.R., Crabtree, B.F., Zyzanski, S.J., Goodwin, M.A. and Strange, K.C. (1998) Making time for tobacco cessation counseling. Journal of Family Practice, 46, 425–428.

Lewis, D. (2003) Computers in patient education. CIN: Com- puters, Informatics, Nursing, 21, 88–96.

McBride, P.E., Plane, M.B., Underbakke, G., Brown, R.L. and Solberg, L.I. (1997) Smoking screening and management in primary care practices. Archives of Family Medicine, 6, 165–172.

McDaniel, A.M., Hutchison, S., Casper, G.R., Ford, R.T., Stratton, R. and Rembusch, M. (2002) Usability testing and outcomes of an interactive computer program to promote smoking cessation in low income women. In Proceedings of the American Medical Informatics Association Annual Sym- posium, San Antonio, TX, pp. 509–513.

National Telecommunications and Information Administration, US Department of Commerce. (2000) Falling through the Net: Toward Digital Inclusion. Available: www.ntia.doc.gov/ ntiahome/digitaldivide/; retrieved 14 June 2004.

Nielsen, J. (1993) Usability Engineering. Academic Press, Boston, MA.

Prochaska, J.O. and DiClemente, C.C. (1983) Stages and processes of self-change of smoking: toward an integrative model of change. Journal of Consulting and Clinical Psych- ology, 51, 390–395.

Prochaska, J.O., Velicer, W.F., DiClemente, C.C. and Fava, J.L. (1988) Measuring the processes of change: Applications to the cessation of smoking. Journal of Consulting and Clinical Psychology, 56, 520–528.

Prochaska, J.O., Velicer, W.F., Fava, J.L., Ruggiero, L., Laforge, R.G., Rossi, J.S., Johnson, S.S. and Lee, P.A. (2001) Counselor and stimulus control enhancements of a stage- matched expert system intervention for smokers in a managed care setting. Preventive Medicine, 32, 23–32.

Prochaska, J.O., Velicer, W.F., Guadagnoli, E., Rossi, J.S. and DiClemente, C.C. (1991) Patterns of change: dynamic typ- ology applied to smoking cessation. Multivariate Behavioral Research, 26, 83–107.

Revere, D. and Dunbar, P.J. (2001) Review of computer- generated outpatient behavioral interventions: clinical encoun- ters ‘in Absentia’. Journal of the American Informatics Association, 8, 62–79.

Shiffman, S., Paty, J.A., Rohay, J.M., DiMarino, M.E. and Gitchell, J. (2000) The efficacy of computer-tailored smoking cessation material as a supplement to nicotine polacrilex gum therapy. Archives of Internal Medicine, 160, 1675–1681.

Stanley, L.D. (2003) Beyond access: psychosocial barriers to computer literacy. The Information Society, 19, 407–416.

Thorndike, A.N., Rigotti, N.A., Stafford, R.S. and Singer, D.E. (1998) National patterns in the treatment of smokers by physicians. Journal of the American Medical Association, 279, 604–608.

US Department of Health and Human Services (2004) Usability Basics. Available: www.usability.gov/; retrieved: 14 June 2004.

Velicer, W.F., DiClemente, C.C., Prochaska, J.O. and Brandenburg, N. (1985) Decisional balance measure for assessing and predicting smoking status. Journal of Person- ality and Social Psychology, 48, 1279–1289.

Wichansky, A.M. (2000) Usability testing in 2000 and beyond. Ergonomics, 43, 998–1006.

Witte, K., Meyer, G. and Martell, D. (2001) Effective Health Risk Messages. Sage, Thousand Oaks, CA.

Received on November 19, 2003; accepted on September 28, 2004

A. M. McDaniel et al.

384

by guest on A pril 19, 2014

http://her.oxfordjournals.org/ D

ow nloaded from