Re: More questions raised about efficacy of Arthritis Self-ManagementProgram

[Guest guest]

a, I think they should keep the course online for the patients, give

all patients copies of the book, and spend the rest of the time and

money sending the physicians to classes or developing online instruction

for them.

You know what I'm thinking: " It's the Pain, Stupid! 101. "

[ ] More questions raised about efficacy of Arthritis

Self-ManagementProgram

Jun 17, 2003

More questions raised about efficacy of Arthritis Self-Management

Program

Boston, MA, and Stanford, CA < Questions raised by Drs H

and

H Lee (Brigham and Women's Hospital, Boston, MA) about the widely

recommended Arthritis Self-Management Program (ASMP) triggered a sharp

response in the June 2003 issue of the Journal of Rheumatology from the

program's developers and may put the ASMP back under the microscope of

outcomes researchers [1-3]. The ASMP has been endorsed by the Arthritis

Foundation, the US Centers for Disease Control, and the American College

of

Rheumatology.

Dr H

tested the ASMP in patients recruited from primary physicians'

practices and concluded that the 6-week program (handbook and group

interactive sessions) was no more effective than simply providing

patients

with the handbook [1]. Drs Fries, Kate Lorig, and Halsted R Holman

(Stanford University School of Medicine, CA) argue that 's study

is

deeply flawed and that the main problem with the ASMP is that not enough

patients have access to it [2].

" But even if this study were not methodologically flawed, we would urge

caution in basing conclusions on it when the preponderance of the

evidence

is that the program has helped a lot of people over an extended period, "

Lorig tells rheumawire.

" If we can get the same effect with a $20 book as with this

time-consuming

program, why not just use that? "

That may be true, counters, " but if we can get the same effect

with

a $20 book as with this time-consuming program, why not just use that? I

am

a strong supporter of patient education, but it seems clear that we need

more evaluation of the ASMP and why it seemed to work less well in this

setting than in the hands of Dr Lorig and colleagues. "

No benefit found for patients from primary-care settings

The controversy highlights some of the challenges in applying outcomes

measurement to behavioral interventions. Like drugs, such programs often

work less well in the " real world " than in developmental trials.

et al recruited 12 primary physician practices from a network

linked

by common insurance contracts. Sites were randomized as a block into

either

the intervention or control arm of the study. Eligible patients were

identified using claims data as having either osteoarthritis (OA) or

rheumatoid arthritis (RA).

The researchers sent letters from intervention sites inviting them to

participate in the ASMP course. Those who did not respond were called to

determine their interest. Patients from the control sites were also

contacted by letter and similarly enrolled, dropped from further contact

at

their request, or contacted by telephone. One intervention site dropped

out

of the network several months into the trial, at which point the

investigators switched a large control site into the intervention arm.

Intervention patients were given copies of the Arthritis Helpbook and

asked

to attend 6 weekly ASMP interactive discussions led by a trained

facilitator

following a structured syllabus. Control patients were sent copies of

the

book but not offered the interactive sessions. This use of an " active

control " is an interesting departure from previous trials, in which

control

groups generally received " usual treatment " plus a questionnaire or very

basic information such as Arthritis Foundation pamphlets. Patients were

surveyed at baseline and by a questionnaire at 4-month follow-up.

Primary

end points included pain, disability, self-efficacy for pain control,

mental

health, and vitality.

The final report included data for 184 patients: 104 in 6 intervention

practice sites and 74 in 5 control practice sites (Table 1).

Table 1. Primary diagnoses in study

Diagnosis

Intervention

(n=104)

Control

(n=74)

OA

61%

49%

RA

23%

42%

Other

17%

16%

et al found " no significant improvement from baseline in any end

point and no difference between patients in the intervention and control

groups. " There were also no differences in patient satisfaction or in

use of

healthcare resources such as follow-up physician visits, visits for

arthritis, hospitalizations or emergency department visits for

arthritis,

and medication.

This lack of effect was a surprise to the investigators. " Our hypothesis

was

that the ASMP would be effective. We wanted to test that and also to get

the

program out to a large number of people, which is why we moved patients

from

1 site originally randomized to be a control group into the intervention

category. This change did not greatly alter the power of the study, and

the

fact remains that in this 'real-world' test, we saw absolutely no effect

from the program, " says.

Study design criticized

Fries et al attribute this result to recruitment bias, lack of true

randomization, possibly ineffective program delivery, high dropout

rates,

and confounding by the active control. They charge that the

study

was not a randomized trial, since the investigators randomized treatment

sites as blocks rather than individual patients and switched one block

from

the control to the intervention category after the trial had begun.

" It is of course permissible to use groups as the unit of randomization,

but

use of only an n of 6 is inadequate to ensure balanced groups, " Fries et

al

write. " Thus, this is not a randomized trial, and the unbalanced

treatment

groups further document this. The intervention group had about half the

number of patients with rheumatoid arthritis, was 7 years older, [and]

was

more numerous, less educated, and less affluent. " Fries et al suggested

that

there might be " other, unmeasured, major differences between groups as

well

as those reported. " The differences between the 2 groups as reported by

are reproduced in table 2.

Table 2. Baseline differences among treatment groups (p=0.05)

Characteristic

Intervention

(n=104)

Control

(n=74)

Annual household income

>US$60 000

7%

16%

Employment status

Retired

52%

34%

Disabled or unemployed

4%

11%

Paid employment

20%

37%

" The patient groups were quite small, " Lorig points out. " If we have

fewer

than 8 patients, we don't give the course, since it depends highly on

peer

interactions. "

" While it is true that the intervention group had a smaller proportion

of RA

patients, it is hard to see how that could have obscured efficacy of the

intervention unless you assume that self management is easier in RA than

in

OA, " tells rheumawire.

Fries et al also doubt that the ASMP as delivered in the study

was

enough like the " classic " program to be comparable to previous studies

in

which that program was validated. " It is not clear that the intervention

was

delivered as it was meant to be. There was 1 leader rather than 2, and

there

was not a peer leader, and it is not clear what quality control was

done, "

Lorig said. " The ASMP course is tightly scripted, and quality control is

important. "

" It is not clear that the intervention was delivered as it was meant to

be. "

and Lee responded, " While we have no objective data to support

the

integrity of the classes taught in the trial, they were all facilitated

by

the statewide trainer for the arthritis self-help class from the

Arthritis

Foundation. She has trained dozens of other teachers, taught over 50

different self-help classes, and attended numerous workshops on the

arthritis self-help class. If we are to question her ability to teach

the

class as outlined by Lorig and Fries [4], the potential to successfully

transfer the class needs to be assessed " [3].

" I want to emphasize that I am not dismissive of this program, "

adds. " I send patients to it if they want to go, and I think they learn

a

lot. I have even been medical director for the program at our hospital.

Patients like the program, but if pain, functional status, and similar

measures are the end points of interest, there clearly is not a

consensus on

the efficacy of this program. "

How do we know it works?

Although Fries et al state that all studies of the ASMP except the

paper have been positive [2], earlier work by British researchers also

raised questions about the program's effectiveness [5].

Cohen et al compared an intervention modeled on the ASMP with a program

with

similar content but using health professionals rather than laypersons as

instructors. They reported, " Both interventions resulted in an increase

in

patients' knowledge of arthritis and in their use of exercise compared

with

a control group that received no intervention. However, neither

intervention

was any more effective than nonintervention in lessening patients' pain,

improving their functioning, enhancing social support systems, lessening

their depression, or improving their health behaviors beyond that of

exercise. " [5]

" Neither intervention was any more effective than nonintervention in

lessening patients' pain, improving their functioning, enhancing social

support systems, lessening their depression, or improving their health

behaviors beyond that of exercise. "

Also, and colleagues have previously carried out a meta-analysis

of

17 studies of arthritis self-management education. In data presented at

the

2001 American College of Rheumatology annual meeting, they reported that

ASMP programs produced no significant improvement in either pain or

disability [6]. The summary effect size was 0.12 for pain (95% CI

0.00-024)

and 0.07 for disability (95% CI 0.00-0.15). Since the confidence

intervals

overlap zero, neither is statistically significant. In addition, there

was a

significant publication bias toward studies reporting reductions in

disability.

" If we assume that statistical significance would be achieved with

greater

numbers of studies (and patients), these effect sizes suggest that

arthritis

self-management education programs have a very small benefit, "

writes [3].

Patients must be ready for change

Dr Kate Lorig

Lorig says that the study has raised points that should be

considered, such as how to recruit patients from a closed practice

network.

More work is needed to determine whether outcomes were affected by

differences in patients actively recruited by et al vs the

perhaps

more motivated volunteers passively recruited through print and radio

public

announcements by her own group.

In addition, she points out that research on how behavior change happens

shows that the efficacy of intervention programs depends in part on the

patient's stage of psychological " readiness " for change. For instance,

in a

study reported by Dr FJ Keefe (Ohio University, Athens, OH) [7],

arthritis

patients fell into 1 of 5 subgroups. About 44% were in the stage of

" precontemplation " < not even thinking about changing. Eleven percent

were

in " contemplation " < considering change. Another 22% were in

" preparation "

for change. Six percent were undertaking some type of " unprepared

action, "

and 17% were in the " prepared maintenance " stage of managing their

disease.

" One may be able to enhance the outcomes of self-management

interventions

for arthritis by tailoring treatment to the patient's particular stage. "

" It is possible that the arthritis subgroups identified may predict

arthritis patients' participation in and responsiveness to pain-coping

skills training, exercise interventions, or other formal self-management

training programs. Also, one may be able to enhance the outcomes of

self-management interventions for arthritis by tailoring treatment to

the

patient's particular stage, " Keefe said [7].

ASMP to be tested online

From the viewpoint of the Stanford researchers who developed the ASMP,

the

program's main weakness is that too few arthritis patients have access

to

it. Lorig says a large part of the problem is that US insurers are

generally

unwilling to pay the costs.

" There is almost no payment for self-management education in the

fee-for-service setting, " she tells rheumawire. " If a useful new

medication

is developed, it becomes part of the formulary and costs are reimbursed,

even if it is given in an unusual way. For example, infusions were not

often

used in treating arthritis until the new generation of disease-modifying

agents appeared, but physicians quickly learned how to manage infusions

of

these drugs. " Managed-care organizations may be somewhat more open to

patient-education issues, she comments. Lorig says that a similar

self-management program for patients with chronic diseases is now a

routine

benefit for most participants in the Kaiser programs in California.

" There is almost no payment for self-management education in the

fee-for-service setting. "

Lorig et al are addressing the access issue by moving self-management

programs from the classroom to the Web. They will soon begin recruiting

500

to 600 US patients with OA or RA for an NIH-funded randomized trial

comparing a new online version of the ASMP with usual care.

The online version has 3 parts: a 6-week interactive didactic section,

threaded message bulletin boards with discussions, and an individual

" med

log " information storage area. Lorig tells rheumawire the investigators

will

assemble test groups of 25 patients plus 2 moderators, each with a

control

group of 25 patients. An initial version of this program tested for

patients

with various types of chronic disease showed that patients tend to log

on 2

or 3 times per week for a total of about 2 hours per week. The patients

randomized to usual care will receive a $10.00 credit at amazon.com

every 6

months for filling out the study questionnaires. Study end points will

be

health behaviors, health status, and healthcare utilization.

And if it works, how?

Although behavior change has been the goal of many patient-education

programs, Lorig's own work raises doubts about its role in the ASMP. The

benefits from the ASMP had been expected to correlate with behavior

changes

such as more exercise and better use of nonpharmacologic pain-management

techniques. When Lorig examined outcomes more closely, she found only a

weak

association between behavior changes and changes in health status [8].

The

efficacy was associated most strongly with an increase in " perceived

self-efficacy " < confidence in being able to manage the consequence of

arthritis. In essence, patients who thought they had more control over

their

disease and were emotionally upbeat also had better health status.

Whether

perceived self-efficacy is the cause or the result of improved health

status

is unknown.

" It's almost as if I've stopped aging and started to get younger. "

While the long-term benefits for most arthritis patients seen in

ordinary

clinical practice may be somewhat uncertain, 1 group clearly benefits

from

the ASMP: patients trained as lay leaders reported significantly less

pain,

less depression, better communication with physicians, " more confidence,

happiness, and a changed outlook on life in general " [9]. One of them

commented: " It's almost as if I've stopped aging and started to get

younger. "

Janis

[Guest guest]

Amen !!!!

a

> a, I think they should keep the course online for the patients, give

> all patients copies of the book, and spend the rest of the time and

> money sending the physicians to classes or developing online instruction

> for them.

>

> You know what I'm thinking: " It's the Pain, Stupid! 101. "

>

>

>