• Disease management;
  • Effectiveness;
  • Rheumatoid arthritis


  1. Top of page
  2. Abstract


To perform a systematic review of the published literature on disease management of rheumatoid arthritis (RA) and to use metaanalysis to estimate the magnitude of benefit these programs have on functional status in patients with RA.


Computerized databases for English articles from 1966 to September 2001 were searched. Two reviewers evaluated 1,029 published titles, identified 11 studies meeting explicit inclusion criteria, and extracted data about study characteristics, interventions used, and outcomes measured. Pooled effect sizes for functional status were calculated using a random-effects model.


Four out of 8 disease management programs showed significant improvements in functional status; however, the pooled effect size (ES) was small and statistically non-significant (ES 0.27; 95% confidence interval [95% CI] −0.01, 0.54). Studies with longer intervention durations (>5 weeks) had significantly improved patient functional status (ES 0.49; 95% CI 0.12, 0.86), compared with studies with shorter intervention durations (≤5 weeks, ES 0.13; 95% CI −0.25, 0.52).


There were limited data to support or refute the effectiveness of disease management programs in improving functional status in patients with RA. Additional studies are needed to confirm if a more intensive intervention may be of benefit to patients with RA, as suggested by our study.


  1. Top of page
  2. Abstract

Rheumatoid arthritis (RA) is a chronic progressive disease, with more than 2 million cases reported in the US (1). Women are 2.5 times more likely than men to develop RA, with an estimated 1.5 million women and 600,000 men diagnosed with this disease in the US (2, 3). Worldwide prevalence is estimated at 1% (4, 5).

In a majority of cases, the disease is chronic and debilitating in nature, and is one of the most common causes of disability, particularly in persons 65 years of age and older. More than 80% of patients diagnosed with RA are partially disabled and 16% are completely disabled after 12 years of disease (4). Disability and functional status drive the major cost items such as hospitalizations, surgery, loss of employment, and need for long-term care (6).

It has been estimated that over 8.7 billion dollars are spent annually in the US on RA, with the average annual direct cost of medical care for a single patient estimated at $4,798 (7). One study estimated that each 1-point increase in the Health Assessment Questionnaire (HAQ) score was associated with an incremental increase of $3,000 in yearly direct costs (6). These costs are expected to rise as new therapies to treat RA become increasingly available. Thus, improvements in functional status and reducing disability could lead to meaningful savings in the direct and indirect medical costs associated with the management and treatment of RA.

Due to escalating costs in managing and treating chronic conditions such as RA, many health care organizations have developed disease management programs in an attempt to improve quality of care and reduce overall costs. We defined disease management as a systematic and multidisciplinary approach to care for chronic conditions. By implementing disease management programs in the care of patients with RA, it is possible that joint destruction, disability, and premature death could be prevented or minimized (4).

Unlike the Vliet Vlieland et al (8) review, which assessed the effects of a multidisciplinary team care, our study evaluates a combined effect of both a multidisciplinary team care approach and patient education on functional status in patients with RA. Therefore, our objective was to perform a systematic appraisal of the published literature to identify the body of evidence in disease management of RA and to use metaanalysis to estimate the magnitude of benefit these programs have on functional status in patients with RA.


  1. Top of page
  2. Abstract

We performed a systematic review of the published medical literature to evaluate the effectiveness of disease management programs in improving functional status and reducing disability for patients with RA.

Search strategy.

We conducted a search of the Medline, Healthstar, Embase, and Cochrane databases for English language articles published between January 1966 and September 2001. The search used the following medical subject headings: patient care team, patient care planning, primary nursing care, case management, critical pathways, primary health care, continuity of patient care, guidelines, practice guidelines, disease management, comprehensive health care, ambulatory care, patient education, and the specific words “rheumatoid arthritis” and “disease management.” We also performed a manual search of bibliographies from relevant articles and reviews. Finally, we solicited other references from physicians and researchers in the field.

Inclusion criteria.

We defined disease management intervention as follows: A multidisciplinary intervention delivered by a team of health care professionals, providing a systematic approach to care, and including a patient education component. A multidisciplinary team was defined as 2 or more disciplines cooperatively involved in patient care.

In order to be included in our review, the studies had to meet the following criteria 1) have a target population of adult patients with RA, 2) the evaluation of a disease management intervention was as defined above, 3) contained an outcome assessment of patients' functional status, 4) the study design included a control group and was either a randomized controlled clinical trial (RCT) or a quasi-experimental, nonrandomized study with a before-and-after design, as defined by the Cochrane Effective Practice and Organization of Care (EPOC) Group's criteria for acceptable study designs (9). Functional status was defined as the ability to carry out routine activities of daily living, as well as to participate in life situations of recreational, leisure, and/or vocational nature (work, school, homemaking) that are age and sex-specific (10).

Explicit inclusion and exclusion criteria were developed for titles, abstracts, and articles. Two reviewers (EB, AH) trained in health services research and the principles of critical appraisal, independently reviewed samples of titles, abstracts, and reviews. Inter-rater agreement was assessed using the kappa statistic, with kappa values of 0.70–0.75 at each stage of the review. Disagreements were resolved by consensus.

Articles were rejected at the title stage if they did not pertain to RA, included only pediatric cases, or if they were reviews, case reports, editorials, letters, or meeting abstracts. Selected abstracts were rejected if they described disease management programs in general, but did not report any objective measurements of the processes or outcomes of care, did not assess functional status, or if they were clinical trials evaluating single treatment or diagnostic modalities. Full texts of selected articles were reviewed and rejected if they did not implement a multidisciplinary treatment approach, if patient education was not a part of the overall intervention strategy, or if they did not employ acceptable experimental or quasi-experimental study designs as defined by the Cochrane EPOC group's criteria for acceptable study designs (9).

Data collection.

Using a standardized abstraction form, specific elements describing study design, population characteristics, sample size, practice settings, intervention strategies, and the effects of the programs on functional status were collected from unmasked articles that passed the article-screening phase.


The statistical pooling of effects across studies was performed using a random effects model following the empirical Bayes method of Hedges and Olkin (11). Since the measures of function were not the same across studies, we used effect sizes to estimate the disease management effects instead of weighted mean differences.

Effect sizes (ESs), defined as the difference between the means of the treatment and control arms divided by the pooled estimate of the standard deviation (continuous variables) or the log odds ratio multiplied by a constant variance term (binary outcomes) (12), were calculated for eight studies to allow for pooling of the results (11, 13–16). When appropriate, we used changes over baseline values as opposed to followup values in our analysis. Variances for the changes over baseline were used if they were reported. In those instances when variances for the change were absent, we assumed that the variance for the change was equal to the average of the variances of the baseline and followup distributions, if both were given; if not, we assumed that the variance for the change was equal to the variance of the followup distribution. If neither of these was reported, we assumed that the pooled results for the same instrument from other studies were equal to the unreported variances of the study. The ESs were constructed such that positive numbers denote treatment benefit and are reported as an ES with the 95% confidence interval (95% CI).

Sensitivity analyses.

We evaluated the possibility that differences of the intervention attributes employed by disease management programs might have different impacts on the functional status of patients with RA. We used the RAND definitions to assess different attributes of the intervention (17). The duration of intervention was defined as the length of time in days from the start of intervention to the end of intervention, e.g. educational sessions for 5 weeks. The number of units of intervention was defined as the number of times the intervention occurred for each target population, e.g., 1 educational session each week for 5 weeks, resulting in 5 units of intervention.

In order to estimate the effects on these intervention attributes on functional status, we categorized the included studies into 2 broad categories based on duration of intervention and number of units of intervention and then compared their pooled results. All, except the Vliet Vlieland et al study, were done in an outpatient setting. Therefore, we excluded this study from the sensitivity analyses. The following categories were included into metaanalyses: 1) studies with a duration of intervention ≤ 5 weeks (n = 3) versus studies with a duration of intervention for > 5 weeks (n = 4), and 2) studies employing less than or equal to 6 units of intervention (n = 4) versus studies employing more than 6 units of intervention (n = 3).

Publication bias.

We explored the evidence for publication bias using a funnel plot (18) for functional status (19). We plotted the ES on the x-axis by the inverse of its standard error on the y-axis. A plot that is asymmetrical or that has a relative paucity of estimates favoring no treatment with large standard errors (the bottom left region of the graph) suggests publication bias.


  1. Top of page
  2. Abstract

Literature review.

The initial literature search yielded 1,029 relevant references published between January 1966 and September 2001. We accepted 89 titles for further screening and of these, 67 abstracts met our explicit inclusion criteria and were retrieved for more detailed evaluation. In total, 11 articles met our definition of disease management. Table 1 outlines the key characteristics of each study included in the review.

Table 1. Key characteristics of the included studies*
Author (year)CountryStudy designSample size (TX/CTL)Patient populationFollowupFrequency/intensity of interventionTX group interventionCTL group interventionOutcome measure(s)
  • *

    TX/CTL = Treatment group/control group; RCT = randomized controlled trial; RA = rheumatoid arthritis; Nonrandomized = prospectived controlled before and after; ARA = American Rheumatism Association. OA = osteoarthritis; VA = Veterans Administration.

Ahlmen (1988)SwedenRCT31/28Female patients ages 38–73 years old, who had definite RA and who were due for appointments at the outpatient clinic of the university-based rheumatology department12 monthsPatients attended five 2-hour education sessionsFive, 2-hour education sessions led by a multidisciplinary team of providers and given individually to patients at regular appointments.Usual care1. Health status 2. Emotional state 3. Functional status
Anderson (1988)USNonrandomized12/8Patients with definite RA, who had been under regular outpatient treatment and who had a sustained flare up of RA over an 18-month interval24 monthsPatients were admitted to a hospital-based rheumatic unit for approximately 12.4 days for intensive inpatient treatmentPatient received multidisciplinary, comprehensive inpatient management. Rheumatology clinic nurse specialist facilitated patient education and team care coordination.Usual care1. Functional status 2. Cost
Katz (1968)USRCT20/20Patients with definite RA who had 2 or 3 ARA functional criteria and who lived within the area served by Visiting Nurse Association12 monthsIn 12 months nurses made 864 visits to the 20 patients in the treatment group.Patient received comprehensive treatment program supplied by arthritis outpatient clinic and visiting home nurse. Goals and individual treatment plan set for each patient.Usual care1. Functional status 2. Patient compliance with scheduled visits
Lindroth (1995)SwedenNonrandomized53/39Patients with OA and RA were referred to the education program3 monthsSix biweekly 2 1/2 hour sessionsGroup of 7–8 patients with RA or OA received 6 × 150 minutes education sessions held biweekly led by multidisciplinary team.Usual care1. Functional status 2. Patient knowledge 3. Patient compliance
Lindroth (1997)SwedenRCT49/47Patients with established diagnosis of RA12 monthsEight weekly 2.5 hour sessionsPatients enrolled in the RA school. Eight 2.5-hour educational sessions led by team of health professionals.Usual care1. Functional status 2. Patient knowledge 3. Patient compliance
Parker (1984)USRCT18Male patients who were hospitalized at a VA hospital3 monthsIntensive educational program that lasted for 7 hoursPatients received inpatient rheumatology care on a multidisciplinary arthritis unit and intensive 7-hour education program delivered by 2 rheumatology educators.Usual care1. RA knowledge 2. Functional status 3. Emotional status 4. Patient compliance
Schned (1995)USRCT37/42Patients with chronic inflammatory arthritis including RA between 18 and 72 years of age12 monthsHalf day education and management programPatients enrolled in a Comprehensive Care Program. Patients participated in half-day education and management program. Team of rheumatologists and allied professionals reviewed all patients on quarterly basis.Usual care1. Functional status 2. Emotional state 3. Overall health
Scholten (1999)AustriaRCT38/30Patients with RA12 monthsEducational lectures were scheduled biweekly for nine afternoonsPatients participated in training program led by multidisciplinary team of providers. Biweekly lectures on RA management and clinical aspects provided to patients.Usual care1. Functional status 2. Emotional state 3. Patient compliance 4. Patient knowlege
Spiegel (1986)USNonrandomized49/43Patients with definite RA, who had the disease for at least 1 year13 days13 days inpatient treatment plus educational programPatients hospitalized in a multidisciplinary rheumatology rehabilitation unit. Team designed individualized management program. Nurses educated patients about arthritis and encouraged them to participate in management of disease.Usual care1. Functional status 2. Emotional state 3. Knowledge of arthritis
Taal (1993)IrelandRCT27/30RA patients who entered the practices of 3 rheumatologists less than 8 years were selected for this study1.5 monthIntervention consisted of 5 weekly 2-hour group sessionsPatients received 5 weekly 2-hour group sessions led by nurses specialized in rheumatic disease, physiotherapists, or social workerUsual care1. Functional status 2. Health status 3. Emotional status 4. Patient compliance 5. Patient Education
Vlieland (1997)NetherlandsRCT39/41Patients with active RA11 daysIn-patient treatment consisted of a fixed period of 11 days of hospitalization in a rheumatology clinic.Patients hospitalized in multidisciplinary in-patient rheumatology unit. Team members conducted educational sessions and information on principle of joint protection during hospital stay.Usual care1. Functional status 2. Emotional status

Of the 11 studies that met all the inclusion criteria, 8 (20–27) were randomized controlled trials (RCTs) and 3 (28–30) were nonrandomized trials. The total number of patients studied in each report ranged from 18 to 96, and the mean duration of illness was 8.4 years. Approximately half (5/11) of these studies were conducted in the US, 3 studies were carried out in Sweden, and the remaining 3 studies were from Austria, Ireland, and the Netherlands.

Our primary outcome of interest was functional status. All 11 studies evaluated the impact of disease management programs on functional status (20–30). We also listed other outcomes assessed in these programs (Table 1) and provided a detailed overview of the selected studies (see Appendix A). The selected studies all employed patient educational programs led by a multidisciplinary team of providers. In a majority of the studies, a rheumatologist, nurse, physiotherapist, occupational therapist, and social worker made up the core of each multidisciplinary team. In addition, some of the studies included a dietician, physiatrist, chaplain, internist, orthopedist, podiatrist, or psychotherapist.

Functional status.

Eleven studies (20–30) representing a total of 701 patients evaluated the impact of disease management programs on improving functional status and reducing disability for patients with RA (Table 2). Of these, 8 studies (20–24, 26, 27, 29) provided sufficient data for metaanalysis. The remaining 3 studies (25, 28, 30) failed to report required variances for an ES calculation.

Table 2. Assessment of disease management program effects on functional status and disability*
Author (year)Study designOutcomeSample size (TX/CTL)Results TX vs CTL Mean ± SDMean difference (95% CI)Effect size (95% CI)
  • *

    TX/CTL = treatment group/control group; 95% CI = 95% confidence interval; RCT = randomized controlled trial; HAQ = Health Assessment Questionnaire (possible range 0–3, higher score denotes greater dysfunction); mHAQ = modified HAQ (possible range 0–12, higher score denotes greater dysfunction); Nonrandomized = prospective controlled before and after; RA = Rheumatoid arthritis; SIP = Sickness Impact Profile (possible range 0–100, higher score denotes greater dysfunction); FSQ = Functional Status Questionnaire (possible range 0–100, higher score denotes better physical functioning; NA = not assessed; AIMS = Arthritis Impact Measurement Scale (possible range 0–6, higher score denotes better physical functioning). Pooled effect size for first 8 studies = 0.279–0.01, 0.54).

  • Pooled estimates for 6 studies that employed HAQ; Mean difference (95% CI) = 0.12 (−0.01, 0.25); Effect size (95% CI) = 0.16 (−0.13, 0.44)

  • Statistically significant studies at 0.05 level.

Lindroth (1997)RCTMean change in HAQ49/470.1 ± 4.6 vs 0 ± 3.70.1 (−1.6, 1.8)0.02 (−0.37, 0.43)
Lindroth (1995)NonrandomizedMean mHAQ (RA only)31/351.6 ± 2.8 vs −0.70 ± 3.52.3 (0.71, 3.89)0.71 (0.21, 1.2)
Taal (1993)RCTMean change in mHAQ27/300.01 ± 0.3 vs −0.16 ± 0.30.17 (0.02, 0.31)0.63 (0.10, 1.17)
Scholten (1999)RCTMean change in HAQ disability score38/300.4 ± 0.6 vs 0.3 ± 0.70.1 (−0.21, 0.41)0.16 (−0.32, 0.64)
Vliet Vlieland (1997)RCTMean change in HAQ39/390.32 ± 0.7 vs 0.42 ± 0.7−0.1 (−0.42, 0.22)−0.14 (−0.59, 0.30)
Schned (1995)RCTMean score on mHAQ (SD)37/420.5 ± 3.1 vs 1.4 ± 2.9−0.9 (−2.2, 0.45)−0.30 (−0.75, 0.14)
Katz (1968)RCTProportion of patients who had deterioration in activities of daily living20 vs 201/20 vs 7/20OR = 10.2 (0.86, 123.9)1.1 (0.05, 2.1)
Ahlmen (1988)RCTMean change in SIP scores31 vs 283.6 ± 6.2 vs 0.1 ± 5.33.5 (0.43, 6.5)0.60 (0.07, 1.12)
Studies not included in metaanalysis
Anderson (28)NonrandomizedMean values of FSQ18 vs 132.15 vs 1.490.66NA
Parker (25)RCTMean change in AIMS9 vs 91.1 vs −0.81.9NA
Spiegel (30)NonrandomizedMean change in AIMS49 vs 430.5 vs 00.50NA

The 11 selected studies utilized the following instruments to assess functional status and disability in RA patients: 6 studies employed validated scales such as the HAQ or modified (m) HAQ (20, 21, 24, 26, 27, 29), 1 study (22) used the functional status subscale from the Sickness Impact Profile, 1 study (23) used an Activities of Daily Living scale, 2 studies (25, 30) employed the Arthritis Impact Measurement Scale, and 1 study (28) used a self-reported functional status questionnaire. Analyzed individually, 4 (20, 22, 23, 29) of 8 studies (50%) showed statistically significant improvements in functional status and disability. Overall, disease management had a small effect, which did not reach statistical significance, in improving functional status in patients with RA (ES 0.27; 95% CI −0.01, 0.54) (Figure 1).

thumbnail image

Figure 1. Effects of disease management programs on functional status (n = 8).

Download figure to PowerPoint

Assessment of studies using HAQ or mHAQ. When studies that used the HAQ or mHAQ to evaluate functional status were pooled separately, we found that on average, there was a difference of 0.12 in HAQ score observed between intervention and control groups, which corresponded to an ES of 0.16; 95% CI −0.13, 0.44. Although both estimates showed improvements in functional status of patients treated in the disease management group, these improvements were not statistically significant (Figure 2).

thumbnail image

Figure 2. Effects of disease management programs on functional status including only studies that utilized Health Assessment Questionnaire (HAQ) or modified HAQ (n = 6).

Download figure to PowerPoint

Sensitivity analysis.

Seven studies were included in the sensitivity analyses (Table 3). Of these, 4 studies (20–22, 26) reported a duration of intervention ≤ 5 weeks, with the remaining 3 studies (23, 24, 29) having interventions that lasted > 5 weeks. Based on the number of units of intervention, we found 4 studies (20–22, 29) employed ≤ 6 units of intervention, and 3 other studies (23, 24, 26) used more than 6 units of intervention for each target population.

Table 3. Impact of duration of intervention and number of units of intervention on functional status (n = 7)*
Intervention attributesNumber of studies analyzedPooled effect size (95% CI)
  • *

    95% CI = 95% confidence interval.

  • P < 0.05

 Studies with duration of intervention ≤5 weeks40.13 (−0.25, 0.52)
 Studies with duration of intervention >5 weeks30.49 (0.12, 0.86)
Number of units of intervention  
 Studies with number of units of intervention ≤640.39 (−0.30, 0.82)
 Studies with number of units of intervention >630.17 (−0.10, 0.44)

When metaanalyzed, we found that the pooled estimate for studies with interventions lasting > 5 weeks had statistically significant improvements in patient functional status (ES 0.49; 95% CI 0.12, 0.86) as compared with studies with interventions lasting ≤ 5 weeks (ES 0.13; 95% CI −0.25, 0.52). However, when we compared the pooled estimates of the studies based on the number of units of intervention, we found that the studies offering ≤ 6 units of intervention were associated with higher effects (ES 0.39; 95% CI −0.03, 0.82) compared with the studies having more than 6 units of intervention (ES 0.17; 95% CI −0.1, 0.44). Neither estimates reached statistical significance.

Publication bias.

Figure 3 includes a funnel plot exploring the possibility of publication bias. There appears to be asymmetry, with a missing area in the lower left portion of the plot. This suggests that small studies with negative results were less likely to have been published than studies of similar size reporting positive results.

thumbnail image

Figure 3. Funnel plot for publication bias detection (n = 8).

Download figure to PowerPoint


  1. Top of page
  2. Abstract

The results of our study suggest that implementation of disease management programs incorporating a multidisciplinary team and a patient education component were associated with small, but not statistically significant, improvements in functional status and disability in patients with RA (Figure 1). On average, there was a decrease of 0.12 points in the HAQ score observed in the intervention group due to disease management programs, which is equal to an improvement of 13% above baseline values when compared to usual care.

The Outcomes Measures in Rheumatoid Arthritis Clinical Trials set the standards to define clinically important changes in different outcomes of RA patients (31). It has been suggested that improvements of 22–36% (approximately 0.22–0.46 points) in the HAQ scores would be considered as clinically important and meaningful changes to both patients and providers (31, 32). The percentage of improvement observed in our study is considerably below these suggested threshold values.

There could be several reasons for the obtained results. One possible explanation is that disease management programs need standardized implementation strategies that are necessary to achieve improved patient outcomes; for example, was the education based on individual patients' needs, was the education delivered using the cognitive behavioral technique proven to influence patient's behaviors, or was the intervention long enough to yield desirable patient outcomes? Such standardized program development and implementation might ultimately determine the success of the program. However, there are limited data from the published literature on the effectiveness of these strategies that could inform program developers which interventions are proven to work. If such standardized programs are effectively developed and implemented, they should produce significant results. Adequate funding and rigorous evaluation of these programs will be required in order for the rheumatology community to adopt them as a means for evaluating the management of RA patients. Funding could be problematic given the nature of our current health care system.

Another possible explanation for our findings could be that the overall efficacy of disease management programs is a function of the relative efficacies of both the pharmacologic and non-pharmacologic components of the programs. Over the last decade more effective pharmacologic therapies have become available for RA. The studies included in our review may not be reflective of the current pharmacologic treatment of RA patients. Studies using more effective drugs in conjunction with disease management programs might yield different results.

The evaluation of disease management programs is a continuously evolving field. To date, there is no standardized approach to evaluate these programs and no studies have assessed the correlations between different attributes of the interventions, such as intensity or type of interventions, with the processes and outcomes of care.

Attempts to synthesize the evidence regarding the impact of specific disease management interventions on outcomes of RA patients have been reported (33, 34). Superio-Cabuslay et al (33) assessed the effects of patient education on clinical outcomes in patients with RA and concluded that patient education had a favorable effect in improving patient functional status. Mullen et al (34) evaluated the efficacy of psychoeducational interventions and demonstrated that patient education, with a psychobehavioral component, can improve the health status of patients with RA. Our study, while complementing the above reviews, stresses the potential importance of an intervention assessment on the outcomes of disease management programs.

The RAND report, assessing different interventions to increase utilization of preventive services for the Medicare population (17), suggests that any disease management intervention can be described using the following three attributes: 1) intensity, defined as the length of time in minutes for each unit of intervention (e.g., 120-minute educational session), 2) duration, defined as length of time in weeks for the start of the intervention to the end of intervention (e.g., computerized reminders to providers starting December 1 and ending December 30, duration = 4 weeks), and 3) number of units of intervention – defined as number of units of intervention that occurred for each target population (e.g., 2 reminders sent to each patient, units of intervention = 2).

When these criteria were applied to the selected studies, we found that studies varied in terms of the duration of intervention and number of units of the intervention employed for each target population. For example, Taal et al (20) implemented 2 hour weekly educational sessions during a 5-week intervention period, while Schned et al (21) employed a half-day educational session.

Information on the intensity of the intervention was rarely reported in the selected studies; thus, we were unable to assess its impact on functional status of RA patients. However, we assessed the effects of the duration and the number of units of intervention occurring for each target on functional status. Studies with the duration of intervention lasting for more than 5 weeks, compared to those lasting less than or equal to 5 weeks, revealed that those with longer intervention durations were associated with statistically significant improvements in functional status (ES 0.49 versus 0.13).

Our analyses were done using studies that had outpatient programs only. Vliet Vlieland et al (8) compared the efficacy of a multidisciplinary approach to care, provided in the inpatient and outpatient settings, versus usual outpatient care on patient outcomes. The study concluded that a short-term inpatient multidisciplinary approach to care was associated with significant improvements in disease activity when compared to usual care. However, Vliet Vlieland et al noted that the effects of an outpatient multidisciplinary approach to care were less evident. Although our sensitivity analyses did not incorporate inpatient programs, the results suggest that significant improvements in functional status can also be achieved in an outpatient setting, given the duration of the intervention is of sufficient length.

In addition to performing a qualitative analysis, we attempted to quantify the effectiveness of the disease management programs by pooling studies to derive the programs' estimate. According to Kazis et al (35), the effect size represents a standardized measure of change in a group or a difference in changes between two groups and provides a standard for comparison. However, the interpretation of this unitless measure poses some difficulties for clinicians (35). To overcome this limitation, we pooled studies that utilized the same assessment tool (e.g., HAQ) as a mean difference in addition to providing the corresponding effect size.

Another limitation of this study is that disease management is an evolving concept, which lacks a single definition. The present study included only studies that utilized a multidisciplinary team care approach (36) and patient education as a part of the overall intervention strategy. Our results may be dependent upon our operational definition of disease management. Consequently, our definition of disease management may have limited the number of studies that could have been evaluated. The small number of studies limited our ability to draw more meaningful conclusions about the effectiveness of disease management in RA. Our sensitivity analyses were based on 7 of the 8 evaluable studies, and given the small number of studies analyzed, the results should be considered with caution. The final limitation is that there appears to be evidence of publication bias based on our funnel plot analysis (Figure 3). This indicates that smaller studies with negative results may not have been published. Thus, our findings may overestimate the true impact of disease management in RA.

Functional status remains the most important clinical outcome for RA patients. There are limited data to support or refute the effectiveness of disease management programs in improving functional status and reducing disability due to this serious disease. Additional studies are needed to confirm whether a longer lasting intervention may be of benefit to patients with RA, as suggested by this review. Further research should seek patient subsets that are more likely to benefit from disease management programs and evaluate whether such a targeted approach is cost-effective in improving functional status. Additional studies utilizing more effective drugs, in combination with disease management programs, are needed to evaluate the effects of these programs.


  1. Top of page
  2. Abstract
  • 1
    Moder KG. New medications for use in patients with rheumatoid arthritis. Ann Allergy Asthma Immunol 2000; 84: 2804.
  • 2
    Lawrence RC, Helmick CG, Arnett FC, Deyo RA, Felson DT, Giannini EH, et al. Estimates of the prevalence of arthritis and selected musculoskeletal disorders in the United States. Arthritis Rheum 1998; 41: 77899.
  • 3
    Alarcon GS. Epidemiology of rheumatoid arthritis. Rheum Dis Clin North Am 1995; 21: 589604.
  • 4
    Matteson EL. Current treatment strategies for rheumatoid arthritis. Mayo Clin Proc 2000; 75: 6974.
  • 5
    Schuna AA, Megeff C. New drugs for the treatment of rheumatoid arthritis. Am J Health Syst Pharm 2000; 57: 22534.
  • 6
    Fries JF. Safety, cost and effectiveness issues with disease modifying anti-rheumatic drugs in rheumatoid arthritis. Ann Rheum Dis 1999; 58 Suppl 1: I869.
  • 7
    Ramsburg KL. Rheumatoid arthritis. Am J Nurs 2000; 100: 403.
  • 8
    Vliet Vlieland TP, Hazes JM. Efficacy of mltidisciplinary team care programs in rheumatoid arthritis. Semin Arthritis Rheum 1997; 27: 11022.
  • 9
    Cochrane Effective Practice and Organization of Care Group (EPOC) Information page. The Cochrane Database of Systematic Reviews Web site. Accessed September 12, 2001. URL:
  • 10
    Subcommittee on Populations. National Committee on Vital and Health Statistics: Classifying and reporting functional status 2001. URL: Accessed April 19, 2002.
  • 11
    Hedges LV, Olkin I. Statistical methods for meta-analysis. Orlando (FL): Academic Press; 1985.
  • 12
    Cohen J. Statistical power analysis for the behavioral sciences. revised edition. New Jersey: Lawrence Erlbaum Associates; 1988.
  • 13
    Cohen J. Statistical power analysis for the behavior sciences. 2nd ed. New York: Academic Press; 1977.
  • 14
    Rosenthal R, Rubin D. Combining results from independent studies. Psychol Bull 1979; 85: 18593.
  • 15
    Glass G. Summarizing effect sizes. New directions for methodology of social and behavioral science: quantitative assessment of research domains. San Francisco: Jossey-Bass; 1980.
  • 16
    Rosenthal R. Meta-analytic procedures for social research. Beverly Hills (CA): Sage Publications; 1984.
  • 17
    Evidence Report and Evidence-Based Recommendations: interventions that increase the utilization of Medicare-funded preventive services for persons age 65 and older. Washington (DC): US Department of Health and Human Services, Health Care Financing Administration; 1999. Publication No. 02145
  • 18
    Egger M, Davey SG, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997; 315: 62934.
  • 19
    Light RJ, Pillemer DB. Summing up: the science of reviewing research. Cambridge (MA): Harvard University Press; 1984.
  • 20
    Taal E, Riemsma RP, Brus HL, Seydel ER, Rasker JJ, Wiegman O. Group education for patients with rheumatoid arthritis. Patient Educ Couns 1993; 20: 17787.
  • 21
    Schned ES, Doyle MA, Glickstein SL, Schousboe JT, Reinertsen JL, Baglioni AJ, et al. Team managed outpatient care for early onset chronic inflammatory arthritis. J Rheumatol 1995; 22: 11418.
  • 22
    Ahlmen M, Sullivan M, Bjelle A. Team versus non-team outpatient care in rheumatoid arthritis: a comprehensive outcome evaluation including an overall health measure. Arthritis Rheum 1988; 31: 4719.
  • 23
    Katz S, Vignos PJJ, Moskowitz RW, Thompson HM, Svec KH. Comprehensive outpatient care in rheumatoid arthritis: a controlled study. JAMA 1968; 206: 124954.
  • 24
    Lindroth Y, Brattstrom M, Bellman I, Ekestaf G, Olofsson Y, Strombeck B, et al. A problem-based education program for patients with rheumatoid arthritis: evaluation after three and twelve months. Arthritis Care Res 1997; 10: 32532.
  • 25
    Parker JC, Singsen BH, Hewett JE, Walker SE, Hazelwood SE, Hall PJ, et al. Educating patients with rheumatoid arthritis: a prospective analysis. Arch Phys Med Rehabil 1984; 65: 7714.
  • 26
    Scholten C, Brodowicz T, Graninger W, Gardavsky I, Pils K, Pesau B, et al. Persistent functional and social benefit 5 years after a multidisciplinary arthritis training program. Arch Phys Med Rehabil 1999; 80: 12827.
  • 27
    Vliet Vlieland TP, Breedveld FC, Hazes JM. The two-year follow-up of a randomized comparison of in-patient multidisciplinary team care and routine out-patient care for active rheumatoid arthritis. Br J Rheumatol 1997; 36: 825.
  • 28
    Anderson RB, Needleman RD, Gatter RA, Andrews RP, Scarola JA. Patient outcome following inpatient vs outpatient treatment of rheumatoid arthritis. J Rheumatol 1988; 15: 55660.
  • 29
    Lindroth Y, Bauman A, Brooks PM, Priestley D. A 5-year follow-up of a controlled trial of an arthritis education programme. Br J Rheumatol 1995; 34: 64752.
  • 30
    Spiegel JS, Spiegel TM, Ward NB, Paulus HE, Leake B, Kane RL. Rehabilitation for rheumatoid arthritis patients: a controlled trial. Arthritis Rheum 1986; 29: 62837.
  • 31
    Goldsmith CH, Boers M, Bombardier C, Tugwell P, and the OMERACT Committee. Criteria for clinically important changes in outcomes: development, scoring and evaluation of rheumatoid arthritis patient and trial profiles. J Rheumatol 1993; 20: 5615.
  • 32
    Wells GA, Tugwell P, Kraag GR, Baker PR, Groh J, Redelmeier DA. Minimum important difference between patients with rheumatoid arthritis: the patient's perspective. J Rheumatol 1993; 20: 55760.
  • 33
    Superio-Cabuslay E, Ward MM, Lorig KR. Patient education interventions in osteoarthritis and rheumatoid arthritis: a meta-analytic comparison with nonsteroidal antiinflammatory drug treatment. Arthritis Care Res 1996; 9: 292301.
  • 34
    Mullen PD, Laville EA, Biddle AK, Lorig K. Efficacy of psychoeducational interventions on pain, depression, and disability in people with arthritis: a meta-analysis. J Rheumatol 1987; 14 Suppl 15: 339.
  • 35
    Kazis LE, Anderson JJ, Meenan RF. Effect sizes for interpreting changes in health status. Med Care 1989; 27 Suppl 3: S17889.
  • 36
    Ellrodt G, Cook DJ, Lee J, Cho M, Hunt D, Weingarten S. Evidence-based disease management. JAMA 1997; 278: 168792.


  1. Top of page
  2. Abstract


Ahlmen et al. (1988) conducted a controlled trial to assess the effectiveness of a multidisciplinary team approach in treating patients with rheumatoid arthritis (RA). Sixty female patients were sequentially randomized to either a multidisciplinary treatment team or a regular outpatient clinic team. The intervention consisted of 5, 2-hour individualized patient training sessions led by a multidisciplinary team of providers. A rheumatologist, a nurse, a physiotherapist, an occupational therapist, and a medical social worker conducted the sessions. Regular outpatient clinic team patients received only usual care.

Results of this study showed that patients treated in the multidisciplinary team had statistically significant improvement in the Sickness Impact Profile Physical Index score when compared with the regular outpatient team (effect size [ES] 0.60; 95% confidence interval [CI] 0.07, 1.12). Other clinical outcomes measured include the Ritchie Articular Index (RAI), the Keitel Index for specific joint function, the Body Symptom Scale questionnaire for self-assessed physical problems, and the Mood Adjective Check List. No significant differences were observed for these tests.

A controlled before and after trial performed by Anderson (1988) evaluated the effects of short periods of hospitalization to treat active RA on long-term outcomes. Patients were given the option of admission into the rheumatic disease unit (intervention group) or continuation of regular outpatient care (comparison group). Sixteen patients enrolled in the intervention group and 10 patients continued with outpatient care. A multidisciplinary team of physicians, nurses, physical and occupational therapists, social workers, nutritionists, and chaplain treated the hospitalized intervention group patients. Upon discharge, each patient continued to receive care from the same rheumatologist they interacted with while hospitalized. The comparison group maintained their regular outpatient medical program. Clinical outcomes measured were serum rheumatoid factor, erythrocyte sedimentation rate (ESR), morning stiffness, visual analog scale (VAS), grip strength, number of inflamed joints, and a self-reported global assessment. The intervention group had statistically significant improvements in morning stiffness, pain, and joint score at all time points throughout the 2-year followup. The comparison group saw significant improvement only at the end of the 2 years. The results indicated that outcomes improved much quicker in the hospitalized patients and that these improvements were maintained over time. Authors concluded that short hospitalization and multidisciplinary treatment of patients with active RA would lead to a more rapid improvement in clinical outcomes compared to regular outpatient therapy.

Katz (1968) measured the effectiveness of extending the efforts of a multidisciplinary team into the home environment. Forty hospitalized patients were randomized evenly into treatment and control groups. Treatment patients received care from a multidisciplinary team at the arthritis clinic. The multidisciplinary team consisted of rheumatologists, internists, orthopedists, resident physicians, nurses, physical and occupational therapists, a podiatrist, a social worker, and various support staff. In addition, the treatment group received home care service from a public health nurse. The control group only received care from the multidisciplinary team at the arthritis clinic. The Lansbury's Index was used to measure disease activity and functional status. A 5-grade walking scale measured patients' walking ability and the index of independence in activities of daily living assessed functional performance in day-to-day activities. House confinement and socioeconomic function were also recorded. Results showed that the decrease in deterioration of activities of daily living was statistically significant for the treatment group (ES 1.08; 95% CI 0.05, 2.11). The treatment group also had more numerous improvements on the other outcomes, but these did not reach statistical significance. The authors agreed that patients with severe RA would benefit from a multidisciplinary team approach extending into coordinated home care.

In their controlled trial, Taal et al (1993) evaluated the effects of group education on health status, behavior, self-efficacy and outcomes expectations, and disease knowledge of patients with RA. Thirty-eight patients were randomized to the group education program and 37 patients to the comparison group not receiving the education program. The intervention was comprised of a 1-hour educational seminar meeting for 5 consecutive weeks. Nurses specializing in rheumatic diseases, physiotherapists, or social workers led the group sessions. A program book and an audiotape with relaxation exercises were also distributed. A Dutch version of the modified Health Assessment Questionnaire (mHAQ) was used to measure functional disability 6 weeks after the start of the intervention. A statistical improvement was seen for patients in the group education program (ES 0.63; 95% CI 0.10, 1.17). This positive effect was still present at the 14-month followup.

Schned (1995) examined how effective a multidisciplinary team intervention approach would be if provided early in the course of chronic inflammatory arthritis. Employing a pairwise randomization scheme, 57 patients were enrolled into the experimental group and 50 to the control group. The experimental group participated in a Comprehensive Care Program, consisting of a baseline assessment interview, normal primary care services, and a half-day self-management and education seminar within 4 months of enrollment. In addition, members of the multidisciplinary team (a rheumatologist, a mental health specialist, a social worker, a podiatrist, a nurse, a dietitian, a physical therapist, and an occupational therapist) reviewed the status of all experimental group patients on a monthly basis. The control group received traditional care from rheumatologists and primary care physicians. Various clinical outcomes were assessed: mHAQ for functional disability, VAS for overall health, Arthritis Impact Measurement Scale (AIMS) for health status, Acquired Helplessness Index for arthritis control, Beck Depression Inventory (BDI) for psychological depression. No statistically significant differences were observed between the experimental and control groups. However, the control group did show a slight improvement on the mHAQ compared to the experimental group (ES −0.30; 95% CI −0.75, 0.14). Authors concluded that a different team care approach might improve outcomes.

Scholten (1999) conducted a prospective, randomized, controlled trial to study the effects of multidisciplinary training and education. The study enrolled 38 patients in the experimental group and 30 patients in the control group. Experimental group patients attended training programs that met 9 times over a 2-week period. Multidisciplinary team members, consisting of rheumatologists, orthopedists, physiotherapists, psychologists, and social workers led the training programs. Patients also received Lorig's Arthritis Helpbook, which covered the contents of the training programs. Control group patients received no training. The following tests were used: Stanford HAQ to measure disability, the Freiburg Questionnaire of Coping with Illness to evaluate coping with illness, and the BDI to measure depression. Results at one year showed that the experimental group had statistically significant improvement in these clinical tests compared to baseline. No statistical differences were noted in the control group when compared to baseline. In terms of functional disability between the 2 groups, the experimental group displayed a very small improvement compared to the control group. However, this was not significant (ES 0.16, 95% CI −0.32, 0.64).

Vliet Vlieland et al (1997) evaluated whether treatment conducted by a multidisciplinary team produced significant change in disease-modifying antirheumatic drugs. Thirty-nine patients with active RA were randomized to the experimental inpatient group receiving treatment from a multidisciplinary team followed by routine outpatient care. The multidisciplinary team consisted of primary medical and nursing care, a physical therapist, an occupational therapist, and a social worker. The intervention period lasted for 11 days. Forty-one active RA patients were randomized to the control group, receiving only routine outpatient care. A VAS, the modified RAI, ESR, C-reactive protein, and number of swollen joints assessed disease activity. The Dutch version of the HAQ and AIMS assessed functional and emotional status respectively. At the 2-year followup, the experimental group displayed greater improvement in all outcomes, but these differences did not reach statistical significance. However, patients in the control group showed a slight improvement in HAQ score compared to the experimental group patients at follow-up (ES −0.14; 95% CI −0.59, 0.30). The authors concluded that treatment effects from the 11-day intervention period decreased over time with no statistical differences observed at 2 years.

Spiegel (1986) compared the outcomes of RA patients admitted to a rehabilitation unit with those of RA patients treated by rheumatologists on an outpatient basis. Using a controlled before-and-after study design, 49 consecutive non-surgical patients were enrolled into the rehabilitation group, while 43 RA outpatients comprised the comparison group. A questionnaire based on the AIMS and the Rand Corporation's Health Insurance Experiment was used to evaluate outcomes. Outcomes include morning stiffness, grip strength, ESR, swollen joint count, social activity, manual dexterity, activities of daily living, knowledge of arthritis, depression, and anxiety. The rehabilitation group received treatment from a multidisciplinary team, consisting of rheumatologists, physical and occupational therapists, a nurse educator, and a social worker. Orthopedic surgeons were available for consultation. Those in the comparison group received usual care. Results at the 1-year followup showed that the rehabilitation group significantly improved in all questionnaire outcomes, except for physical activity and anxiety, when compared with the comparison group. In addition, more comparison patients required hospitalization for an arthritis problem, but more rehabilitation patients required orthopedic surgery. The authors suggested that this discrepancy is due to the fact that those in the rehabilitation group were clinically worse than those in the comparison group at the beginning of the study. Nonetheless, the authors concluded that the rehab group showed continued and significant improvement in their functional ability, mental health, and disease activity when compared with the comparison group.

Lindroth et al (1995) carried out a controlled before-and-after trial to examine the effects of a patient education program at the 5-year followup. One hundred consecutively referred outpatients were enrolled into the intervention group and 100 consecutive patients at a rheumatology clinic were enrolled into the control group. Those in the intervention group attended six educational sessions, each 2.5 hours in duration. A multidisciplinary team of rheumatologists, nurses, physiotherapists, occupational therapists, and social workers directed the sessions. Those in the control group received basic and usual care from a rheumatology clinic. Clinical outcomes measured include a VAS for pain intensity, a modified HAQ for disability, and a true/false quiz for patient knowledge. Ninety-two patients (53 intervention and 39 control) responded to the questionnaire mailed out at the 5-year followup. The results showed that only knowledge about the pathophysiology of arthritis was statistically better in the intervention group. Interestingly, stratified results for RA patients in the intervention group demonstrated a statistically significant reduction in disability compared with RA patients in the control group (ES 0.71; 95% CI 0.21, 1.21.

Parker et al (1984) published a RCT of 18 men hospitalized at a Veterans Administration hospital. Patients were randomized into an educational group or a control group. Both groups received inpatient care from a multidisciplinary treatment team. But only those in the educational group received the formal education program, a 7-hour educational seminar given by rheumatology educators. Outcomes measured included RA knowledge by the Arthritis Knowledge Inventory, functional status by AIMS, and emotional status by BDI. At the end of the 3-month followup, the additional patient education did not affect the majority of AIMS subscales or the BDI. However, the educational group did report significantly more impaired physical activity and increased pain intensity on the AIMS Physical Activity Scale and AIMS Pain Scale respectively, compared to the control group. The authors suggested that the patient education program did not offer additional benefits and, in some cases, resulted in a worse clinical condition.

In their RCT, Lindroth et al (1997) developed an RA school to increase patients' behaviors in practicing exercise and work simplification. The goal was to examine whether this educational intervention would improve pain and disability outcomes. Forty-nine patients were enrolled into the intervention group to receive the educational sessions. Sessions were run by a multidisiciplinary group to receive the educational sessions. Sessions were run by a multidisciplinary team, which included a physician, nurse, physiotherapist, occupational therapist, social worker, and dietitian. Sessions lasted for 8 weeks and were 2.5 hours in length. Forty-seven patients enrolled into the control group and received only usual care. The Stanford HAQ measured disability, a VAS-recorded pain scores, and a multiple choice quiz assessed patient knowledge. At 1 year of followup, results showed significant improvements in the intervention group when compared to the control group. Significantly fewer intervention patients reported lack of knowledge on disease, diet, and physical therapy. Joint protection and capacity to relieve pain also significantly improved. The authors concluded that this patient education program improves knowledge, health behaviors, and some disease-related problems and should be incorporated into RA management.