Professional Documents
Culture Documents
STAGES OF CHANGE
443
J. O. Prochaska & J. C. Prochaska
changing the problem behavior in the next 6 to work to prevent relapse. Termination is defined
months are classified as contemplators. On the as total confidence or self-efficacy across all high-
continuous measure these individuals would be risk situations and zero temptation to relapse.
endorsing such items as "I have a problem and I
really think I should work on it" and "I've been Research Review
thinking that I might want to change something Empirical research on the stages of change has
about myself." taken a number of tacks over the past 20 years
Preparation is a stage that combines intention (for reviews, see DiClemente, 1991; Prochaska,
and behavioral criteria. Individuals in this stage DiClemente, & Norcross, 1992; Prochaska, Redd-
are intending to take action in the next month and ing, & Evers, in press; Norcross, 2002a, 2002b).
have unsuccessfully taken action in the past year. Here, we review only those published research
As a group, individuals who are prepared for ac- studies that have directly examined the stages and
tion report some small behavioral changes— processes of change as they relate to treatment
"baby steps," so to speak. While they have made outcome, broadly defined.
some reductions in their problem, individuals in
the preparation stage have not yet reached a crite- Stages X Processes
rion for effective action, such as abstinence from The transtheoretical model posits that different
smoking or absence of clinical depression. They processes of change are differentially effective in
are intending, however, to take such action in the certain stages of change. In general terms, change
very near future. On the continuous measure they processes traditionally associated with the experi-
score high on both the contemplation and action ential, cognitive, and psychoanalytic orientations
scales. are most useful during the earlier precontempla-
Action is the stage in which individuals modify tion and contemplation stages. Change processes
their behavior, experiences, and environment in traditionally associated with the existential and
order to overcome their problems. Action involves behavioral traditions, by contrast, are most useful
the most overt behavioral changes and requires during action and maintenance.
considerable commitment of time and energy. Twenty years of research in behavioral medi-
Modifications of the problem behavior made in cine and psychotherapy converge in showing that
the action stage tend to be most visible and receive different processes of change are indeed differen-
the greatest external recognition. Individuals are tially effective in certain stages of change. Rosen
classified in the action stage if they have success- (2000) published a meta-analysis of 47 cross-
fully altered the dysfunctional behavior for a pe- sectional studies examining the relationships of
riod from 1 day to 6 months. On the continuous the stages and the processes of change. The stud-
measure individuals in the action stage endorse ies involved smoking, substance abuse, exercise,
statements like, "I am really working hard to diet, and psychotherapy. The mean effect size (d)
change" and "Anyone can talk about changing; I was .70 for variation in cognitive-affect processes
am actually doing something about it." They by stage and .80 for variation in behavioral pro-
score high on the action scale and lower on the cesses by stage, both moderate to large effects.
other scales. At the same time, the sequencing of change pro-
Maintenance is the stage in which people work cesses by stage varied somewhat by disorder or
to prevent relapse and consolidate the gains at- sample. Of particular interest was the finding that
tained during action. Being able to remain free "use of helping relationships was strongly related
of the problem behavior and to consistently en- to stages in studies of psychotherapy" (Rosen,
gage in a new incompatible behavior for more 2000, p. 601).
than 6 months are the criteria for considering The therapist's stance at different stages can
someone to be in the maintenance stage. On the be characterized as follows. With patients in pre-
continuous measure, representative maintenance contemplation, often the role is like that of a nur-
items are, "I may need a boost right now to help turing parent joining with a resistant and defen-
me maintain the changes I've already made" and sive youngster who is both drawn to and repelled
"I'm here to prevent myself from having a relapse by the prospects of becoming more independent.
of my problem." With clients in contemplation, the role is akin
Termination is the stage in which people have to a Socratic teacher who encourages clients to
completed the change process and no longer have achieve their own insights into their condition.
444
Stages of Change
With clients who are in the preparation stage, the by 6 months. For the precontemplators who pro-
stance is more like that of an experienced coach gressed to contemplation at 1 month, 7% took
who has been through many crucial matches and action by 6 months. Similarly, of the contempla-
can provide a fine game plan or can review the tors who remained in contemplation at 1 month,
participant's own plan. With clients who are pro- only 20% took action by 6 months. At 1 month,
gressing into action and maintenance, the psycho- 41% of the contemplators who progressed to the
therapist becomes more of a consultant who is preparation stage attempted to quit by 6 months
available to provide expert advice and support (Prochaska, Norcross, & DiClemente, 1995).
when action is not progressing as smoothly as These data indicate that treatment programs de-
expected. signed to help people progress just one stage in
a month may be able to double the chances of
Predicting Dropout participants taking action on their own in the
When people present for psychotherapy, the near future.
initial challenge is to help them continue. A meta-
analysis across 126 studies found that the dropout Proactive Treatment
rate was about 50% (Wierzbicki & Pekarik, Psychotherapy has traditionally taken a passive
1993). and narrow perspective on its relationship to pa-
Research has identified stage-of-change-related tient populations. Like most health care provid-
variables as the best predictors of dropout across ers, psychotherapists initially relate to patient
a growing number of problems, such as heroin populations in a reactive pattern. Therapists wait
addiction, cocaine abuse, alcoholism, domestic for patients to seek their services. Passive reactive
violence, obesity, chronic mental illness, and relating is appropriate when practicing acute
mental health diagnoses. In one of our studies we care—when patients are acutely ill, in pain, or
were able to predict psychotherapy dropout with distressed. But the major cripplers and cost-
90% accuracy among clients with a variety of drivers of the 20th century are chronic conditions.
mental health problems (Brogan, Prochaska, & One of our recent studies investigated the re-
Prochaska, 1999). The 40% of the patients who sults of reaching out to patient populations. With
terminated quickly (fewer than three sessions) and a representative sample of 5,000 smokers we pro-
prematurely, as judged by their therapists, had a actively offered therapeutic services. Because we
group profile representing the precontemplation knew less than 20% of this population would be
stage. The 20% of patients who terminated ready to take action on their smoking, we let them
quickly but appropriately had a group profile rep- know the services were designed for smokers at
resenting action. The 40% who continued in ther- every stage of change: the 20% or less in the
apy had a mixed profile, with the majority being preparation stage who were ready to act in the
in the contemplation stage. next month; the 40% in the contemplation stage
who were getting ready to quit in the next 6
Stage as an Outcome Predictor months; and the 40% in the precontemplation
The amount of progress clients make during stage who were not ready to quit.
treatment tends to be a function of their pretreat- By proactively reaching out to these patients
ment stage of change. For example, an intensive and customizing our clinical communications to
action- and maintenance-oriented smoking- their stage of change we were able to have 80%
cessation program for cardiac patients achieved participate in our clinical services (Prochaska ,
success for 22% of precontemplators and 43% Velicer et al., 2001). That results in a quantum
of the contemplators; 76% of those in action or increase in our ability to care for this addiction.
prepared for action at the start of the study were We replicated these results with a health mainte-
not smoking 6 months later (Ockene, Kristellar, nance organizations (HMO) population of about
Ockene, & Goldberg, 1992). 4,000 smokers (Prochaska & Velicer et al., 2000)
If clients progress from one stage to the next and a teenage population of about 4,000 teenagers
during the first month of treatment, they can dou- with multiple behavior risks and their parents
ble their chances of taking action in the following (Prochaska, Redding et al., 2002).
6 months. Among smokers, for example, of the Two transformations in therapeutic relating can
precontemplators who were still in precontempla- increase the percentage of high-risk and suffering
tion at 1-month follow-up, only 3% took action people receiving clinical services. The first is to
445
J. O. Prochaska & J. C. Prochaska
reach out proactively and offer them therapeutic The computer alone and computer plus coun-
services. The second is to match the service to selor conditions paralleled each other for 12
each individual's stage of change. months. Then, the effects of the counselor con-
dition flattened out (18%) while the computer
Stage-Matched Treatments condition effects continued to increase (25%
A series of clinical trials applying stage- abstinent).
matched interventions have been conducted. In The next test was to demonstrate the efficacy
our first large-scale clinical trial, we compared of the expert system when applied to an entire
four treatments: a home-based action-oriented population recruited proactively. With over 80%
cessation program (standardized), stage-matched of 5,170 smokers participating and fewer than
manuals (individualized), expert system com- 20% in the preparation stage, we demonstrated
puter reports plus manuals (interactive), and significant benefit of the expert system at each 6-
counselors plus computers and manuals (person- month follow-up (Prochaska, Velicer et al.,
alized). We randomly assigned by stage 739 2001). The point-prevalent abstinence rates for
smokers to one of the four treatments (Prochaska, expert systems versus assessment alone were
DiClemente, Velicer, & Rossi, 1993). 9.7% versus 7.4%; 18.0% versus 14.5%; 21.7%
In the computer condition, participants com- versus 16.6%, and 25.6% versus 19.7% at 6, 12,
pleted by mail or telephone 40 questions that were 18, and 24 months, respectively. The advantages
entered into our central computers that generated over proactive assessment alone increased at each
feedback reports. These reports informed partici- follow-up for the full 2 years assessed. The impli-
pants about their stage of change, their pros and cations here are that stage-matched treatments in
cons of changing, and their use of change pro- a population can continue to demonstrate benefits
cesses appropriate to their stages. At baseline, long after the intervention has ended.
participants were given positive feedback on what The expert system's efficacy was replicated in
they were doing correctly and guidance on which an HMO population of 4,000 smokers with 85%
principles and processes they needed to apply participation (Prochaska et al., 2000). In the first
more in order to progress. In two progress reports population-based study, the expert system was
delivered over the next 6 months, participants 34% more effective than assessment alone; in the
also received positive feedback on any improve- second it was 31% more effective (23.2% absti-
ment they made on any of the variables relevant nent vs. 17.5%). These replicated differences
to progressing. were clinically significant as well. While working
In the personalized condition, smokers re- on a population basis, we were able to produce
ceived four proactive counselor calls over the the level of success normally found only in intense
6-month intervention. Three of the calls were clinic-based programs with low participation rates
based on the computer reports. Counselors re- of much more selected samples of smokers, namely
ported much more difficulty in interacting with about 25% abstinence at long-term follow-up.
participants without any progress data. Without
scientific assessments, it was harder for both Limitations of the Research
clients and counselors to tell whether any sig- Although at least 100 empirical studies have
nificant progress had occurred since their last been conducted on the core transtheoretical con-
interaction. struct of the stages of change, none have directly
Point-prevalence abstinence rates were com- and prospectively matched and mismatched the
pared for each of the four treatments over 18 therapist's relational style in psychotherapy out-
months with treatment ending at 6 months. The come studies. Rather, the available research
two self-help manual conditions paralleled each concerns the predictive utility of the stages of
other for 12 months. At 18 months, the stage- change in terms of outcomes and dropouts, the
matched manuals moved ahead (18% vs. 11% differential use of the processes of change at
abstinent). This is an example of a delayed action various stages of change, and the relative effi-
effect, which we often observe with stage- cacy of diverse forms of treatment. Further, the
matched programs specifically and others have majority of published research concerns self-
observed with self-help programs generally. It help interventions for addictive behaviors, as
takes time for participants in early stages to prog- contrasted to psychotherapy for a wide range of
ress all the way to action. neurotic disorders.
446
Stages of Change
447
J. O. Prochaska & J. C. Prochaska
PROCHASKA, J. O., DICLEMENTE, C. C., VEUCER, W. F., & stage-matched expert systems for multiple behaviors in a
Rossi, J. S. (1993). Standardized, individualized, inter- population of parents. Annals of Behavioral Medicine, 24,
active, and personalized self-help programs for smoking Sxxx (Abstract).
cessation. Health Psychology, 13, 39-46. PROCHASKA, J. O., VELICER, W. F., FAVA, J. L., Rossi,
PROCHASKA, J. O., & NORCROSS, J. C. (2002a). Systems of J. S., & TSOH, J. Y. (2001). Evaluating a population-
psychotherapy: A transtheoretical analysis (5th ed.). Pa- based recruitment approach and a stage-based expert system
cific Grove, CA: Brooks/Cole. intervention for smoking cessation. Addictive Behaviors,
PROCHASKA, J. O., & NORCROSS, J. C. (2002b). Stages of 26. 00-00.
change. In J. C. Norcross (Ed.), Psychotherapy relation- PROCHASKA, J. O., VELICER, W. F., FAVA, J. L., RUGGIERO,
ships that work. New York: Oxford University Press. L., LAFORGE, R., Rossi, J. S., JOHNSON, S. S., & LEE,
PROCHASKA, J. O., NORCROSS, J. C., & DICLEMENTE, C. C. P. A. (2000). Counselor and stimulus control enhancements
(1995). Changing for good. New York: Avon. of a stage-matched expert system intervention for smokers
PROCHASKA, J. O., REDOING, C. A., & EVERS, K. (in press). in a managed care setting. Preventive Medicine, 32,23-32.
The transtheoretical model and stages of change. In K. ROSEN, C. S. (2000). Is the sequencing of change processes by
Glanz, F. M. Lewis, & B. K. Rimer (Eds.), Health behav- stage consistent across health problems? A meta-analysis.
ior and health education (3rd ed.). San Francisco: Health Psychology, 19, 593-604.
Jossey-Bass. WERZBICKI, M., & PEKARIK, G. (1993). A meta-analysis of
PROCHASKA, J. O., VEUCER, W. F., Rossi, J. S., REDOING, psychotherapy dropout. Professional Psychology: Research
C. A., GREENE, G. W., Rossi, S. R., SUN, X., FAVA, and Practice, 29, 190-195.
J. L., & PLUMMER, B. A. (2002). Impact of simultaneous
448