Clinical and epidemiological research
EXTENDED REPORT
Handling editor Tore K Kvien
Additional material is
published online only. To view
please visit the journal online
(http://dx.doi.org/10.1136/
annrheumdis-2014-206593).
1
University of British Columbia
Faculty of Pharmaceutical
Sciences, Vancouver, British
Columbia, Canada
2
Arthritis Research Canada,
Richmond, British Columbia,
Canada
3 databases to identify studies meeting inclusion criteria
Department of Experimental
Medicine, University of British
Columbia Faculty of Medicine,
Vancouver, British Columbia,
Canada
4
Division of Rheumatology,
University of British Columbia
Faculty of Medicine,
Vancouver, British Columbia,
Canada
Correspondence to
Dr Mary De Vera, University of
British Columbia Faculty of
Pharmaceutical Sciences, 2405
Wesbrook Mall, Vancouver,
British Columbia, Canada V6T
1Z3; mdevera@mail.ubc.ca
Received 8 September 2014
Revised 13 January 2015
Accepted 15 January 2015
Published Online First
9 February 2015
To cite: Galo JS, Mehat P,
Rai SK, et al. Ann Rheum
Dis 2016;75:667673.
What are the effects of medication adherence
interventions in rheumatic diseases: a systematic
review
Jessica S Galo,1 Pavandeep Mehat,1,2 Sharan K Rai,2,3 Antonio Avina-Zubieta,2,3,4
Mary A De Vera1,2
ABSTRACT
Objectives Consistent reports of suboptimal treatment
adherence among patients with inammatory arthritis
underscore the importance of understanding how
adherence can be promoted and supported. Our
objectives were to identify and classify adherence
interventions; and assess the evidence on the effects of
adherence interventions on outcomes of patients with
rheumatic diseases.
Methods We conducted a mapped search of Medline,
Embase and International Pharmaceutical Abstract
of: (1) patient population with inammatory arthritis;
(2) evaluation of an intervention or programme targeting
medication adherence directly or indirectly; (3) reporting
of one or more measures of medication adherence and
disease outcome; (4) publication in English, French or
Spanish. For our rst objective, we applied a structured
framework to classify interventions according target
( patient vs provider), focus (educational vs behavioural
vs affective), implementation (generalised vs tailored),
complexity (single vs multifaceted) and provider. For the
second objective, we appraised the evidence of effects of
interventions on adherence and disease outcomes.
Results We identied 23 studies reporting adherence
interventions that directly or indirectly addressed
treatment adherence in rheumatic diseases and further
appraised included RCTs. Interventions that were shown
to impact adherence outcomes were generally
interventions directed at adherence, tailored to patients
and delivered by a healthcare provider. For interventions
that were not shown to have impacts, reasons may be
those related to the intervention itself, patient
characteristics or study methodology.
Conclusions Our systematic review shows limited
research on adherence interventions in rheumatic
diseases with inconsistent impacts on adherence or
disease outcome.
INTRODUCTION
Adherence to long-term pharmacotherapy is para-
mount to the management of inammatory arthritis
including rheumatoid arthritis (RA), systemic
lupus erythematosus (SLE), gout and juvenile idio-
pathic arthritis ( JIA)as left untreated, these condi-
tions can result in disability, complications,
morbidity and mortality.111 Yet systematic reviews
have synthesised adherence rates as low as 10% in
gout,12 30% in RA13 and 49% in SLE.14
Galo JS, et al. Ann Rheum Dis 2016;75:667673. doi:10.1136/annrheumdis-2014-206593
While the problem of medication non-adherence
is well described in rheumatology, solutions are
not. Systematic reviews on adherence interventions
in chronic disease including elderly15 patients with
hypertension16 and diabetes17 are available, yet to
our knowledge, there are none specically among
patients with rheumatic diseases. A 2014 update of
a 2008 Cochrane review of adherence interventions
increased included trials in RA from 219 20 to 6.18
However, an extension of this information is
needed along with comprehensive knowledge of
published evidence in patients with rheumatology.
Thus, our objectives were to: (1) identify and clas-
sify adherence interventions; and (2) assess the evi-
dence on the effects of adherence interventions on
outcomes of patients with rheumatic diseases.
METHODS
Literature search strategy
We searched Medline (1946June 2014), Embase
(1974June 2014) and International Pharmaceutical
Abstracts (1970June 2014). We used Medical
Subject Headings for concepts underlying our search,
medication adherence, intervention and inam-
matory arthritis and applied keywords for concepts
that did not map (see online supplementary table S1).
We also conducted a hand search of bibliographies as
well as a Google Scholar search.
Study selection
Two authors reviewed all titles and abstracts to
identify studies meeting the following inclusion cri-
teria: (1) patient population with inammatory
arthritis (eg, RA, SLE, JIA); (2) evaluation of an
intervention or programme targeting medication
adherence directly or indirectly; (3) reporting of
one or more measures of medication adherence
and disease outcome; (4) publication in English,
French or Spanish. As one of our objectives was to
comprehensively identify instances where treatment
adherence has been addressed in rheumatic disease,
at this point, we did not exclude according to study
design. Disagreement between reviewers was
resolved by consensus.
Identication of adherence interventions in
rheumatic diseases
To address our rst objective, we classied adher-
ence interventions according to: (1) target (patient
vs provider);21 (2) focus (educational vs behavioural
vs affective);22 (3) complexity (single vs
667
 Clinical and epidemiological research
multifaceted);15(4) implementation (generalised vs tailored);15
and (5) provider (eg, physician, nurse, pharmacist) of the inter-
vention. The target refers to the receiver of the intervention.21
The focus of the interventions were categorised as educational
(designed to deliver information with a knowledge-based
emphasis), behavioural (designed to change adherence by target-
ing behavioural patterns) or affective (designed to inuence
adherence through appeals to feelings or social relationships).22
The complexity of the intervention distinguishes ones that have
a single focus from those with multiple foci.15 22
Implementation of the intervention may be categorised as tai-
lored to patient or generalised.15 Finally, we specied the pro-
vider of the intervention.
Quality assessment of controlled trials of adherence
interventions in rheumatic diseases
To address our second objective of assessing the evidence on
impacts of adherence interventions, we focused on included ran-
domised controlled trials (RCTs). We developed a standardised
quality assessment using the Consolidated Standards of Reporting
Trials checklist.23 Studies were assessed for quality of the design,
conduct and reporting, using the scoring criteria: 0=not done or
reported, 0.5=partially done or reported and 1=well done and
clearly reported (see online supplementary table S2).
Data extraction and outcome measurement
We extracted information including study design, patient popu-
lation, disease duration, sample size and medications. Our
primary outcome was the effect of the intervention on adher-
ence outcomes and we extracted information on whether this
was reported or analysed, adherence measure, time point of
measurement, value of adherence measure for the intervention
and control group, and effect size. We calculated effect size
where one was not reported. Specically, for studies reporting
mean values of the adherence measure, we calculated the stan-
dardised mean difference (SMD)the difference of the group
means divided by the standard deviation (SD) for the study
group24 and for studies reporting adherence in terms of propor-
tion of patients, we calculated the relative risk (RR)the pro-
portion of adherent patients in the intervention group divided
by the proportion of adherent patients in the control group.
Our secondary outcome was the effect of the intervention on
disease outcomes and where possible, we similarly extracted
information as with adherence outcomes. Due to heterogeneity
and limited data availability, we could not perform similar calcu-
lations to estimate effect sizes for disease outcome measures.
RESULTS
Search strategy
We identied 2726 studies, forwarding 43 for full manuscript
review after screening (gure 1). With exclusion of 22 manu-
scripts and inclusion of 2 after hand search of bibliographies,
we identied 23 studies to satisfy our systematic review
objectives.
Identication and classication of adherence interventions
in rheumatic diseases
We classied adherence interventions described in 23 studies
16 in RA, 3 in SLE, 1 in gout, 2 in JIA and 1 in unspecied
arthritis. Study details along with intervention classications are
in online supplementary table S3. The description of interven-
tions ranged from one-on-one instruction,25 group educa-
tion,26 telephone-based pharmacy advisory,27 to text message
reminders.28 Fourteen of 23 interventions were designed to
668
directly target medication adherence, with all but one interven-
tion29 directed at patients with rheumatology. This exception
was an instrument providing rheumatologists with structured
information about their patients non-adherence, which was not
shown to have an impact on adherence or disease-related
outcomes.29
With respect to focus, 21 interventions incorporated an edu-
cational component, one was a behavioural intervention involv-
ing text message reminders to patients with SLE,28 and the
aforementioned rheumatologist-targeted intervention was classi-
ed as affective.29 Fourteen interventions were multifaceted and
implementation was tailored to patients in 18 interventions.
Interventions were delivered by rheumatologists,20 3033
nurses,19 25 3336 pharmacists,25 27 29 3638 and other profes-
sionals including therapists39 and educators.4044
Effects of adherence interventions in rheumatic diseases
We critically appraised 13 RCTs10 in RA, 2 in SLE and 1 in
JIA. We did not include here a pilot RCT of group versus indi-
vidual counselling for RA34 and a pilot RCT of internet-based
self-management with telephone support for JIA.44 Quality
scores varied from 9 to 28 out of a maximum possible score of
32 (mean=18.6; median=18.5). We ranked 719 20 27 3840 43 as
having good or high quality (above median) and 6 as having fair
or poor quality. Key results of our synthesis are tabulated in
table 1. Detailed information on adherence and disease out-
comes are provided in online supplementary tables S4 and S5.
Rheumatoid arthritis
In McEvoy DeVellis et als42 study, all patients rst participated
in a psychosocial interview to assess problems related to their
RA. The intervention group then received a problem-solving
intervention to address problems identied in the interview.42
Authors cited no signicant difference between groups, though
did not report data.42 Authors also stated that all patients who
participated in study, regardless of group assignment, improved
on physical and psychological function measures, which may be
attributed to unintended positive effects of the psychosocial
interview received by both groups.42
Brus et al20 used group patient education meetings delivered
by a rheumatologist four times over 1 month and subsequently
at 4 months and 8 months for patients with recent-onset active
RA. From reported mean adherence rates of 8916 and 8421
in the intervention and control groups, respectively, we esti-
mated a SMD of 0.26 (95% CI 0.27 to 0.80). Changes from
baseline did not differ signicantly between groups on disease
activity score (DAS) and C reactive protein.20 Authors contex-
tualised that patients with active, recent-onset RA have high
levels of adherence, in line with previous ndings that patients
with high disease activity tend to be more adherent than those
with low disease activity,45 which may explain why their inter-
vention did not show an effect.
Hill et als19 intervention among patients with RA involved
seven 30-min individual monthly consults covering information
about RA medication and disease processes, pain control and
coping strategies. Authors reported signicant differences in
adherence rates in the intervention and control groups (89% vs
55%; RR, 1.55; 95% CI 1.14 to 2.10).19 Among disease out-
comes evaluated, authors reported an impact of the intervention
on plasma viscosity.19
Evers et al evaluated the effect of individualised cognitive-
behaviour therapy directly targeting frequently experienced pro-
blems in RA as well as indirectly, adherence. Authors reported
that adherence in the intervention group increased at the
Galo JS, et al. Ann Rheum Dis 2016;75:667673. doi:10.1136/annrheumdis-2014-206593
 Clinical and epidemiological research
multifaceted);15(4) implementation (generalised vs tailored);15
and (5) provider (eg, physician, nurse, pharmacist) of the inter-
vention. The target refers to the receiver of the intervention.21
The focus of the interventions were categorised as educational
(designed to deliver information with a knowledge-based
emphasis), behavioural (designed to change adherence by target-
ing behavioural patterns) or affective (designed to inuence
adherence through appeals to feelings or social relationships).22
The complexity of the intervention distinguishes ones that have
a single focus from those with multiple foci.15 22
Implementation of the intervention may be categorised as tai-
lored to patient or generalised.15 Finally, we specied the pro-
vider of the intervention.
Quality assessment of controlled trials of adherence
interventions in rheumatic diseases
To address our second objective of assessing the evidence on
impacts of adherence interventions, we focused on included ran-
domised controlled trials (RCTs). We developed a standardised
quality assessment using the Consolidated Standards of Reporting
Trials checklist.23 Studies were assessed for quality of the design,
conduct and reporting, using the scoring criteria: 0=not done or
reported, 0.5=partially done or reported and 1=well done and
clearly reported (see online supplementary table S2).
Data extraction and outcome measurement
We extracted information including study design, patient popu-
lation, disease duration, sample size and medications. Our
primary outcome was the effect of the intervention on adher-
ence outcomes and we extracted information on whether this
was reported or analysed, adherence measure, time point of
measurement, value of adherence measure for the intervention
and control group, and effect size. We calculated effect size
where one was not reported. Specically, for studies reporting
mean values of the adherence measure, we calculated the stan-
dardised mean difference (SMD)the difference of the group
means divided by the standard deviation (SD) for the study
group24 and for studies reporting adherence in terms of propor-
tion of patients, we calculated the relative risk (RR)the pro-
portion of adherent patients in the intervention group divided
by the proportion of adherent patients in the control group.
Our secondary outcome was the effect of the intervention on
disease outcomes and where possible, we similarly extracted
information as with adherence outcomes. Due to heterogeneity
and limited data availability, we could not perform similar calcu-
lations to estimate effect sizes for disease outcome measures.
RESULTS
Search strategy
We identied 2726 studies, forwarding 43 for full manuscript
review after screening (gure 1). With exclusion of 22 manu-
scripts and inclusion of 2 after hand search of bibliographies,
we identied 23 studies to satisfy our systematic review
objectives.
Identication and classication of adherence interventions
in rheumatic diseases
We classied adherence interventions described in 23 studies
16 in RA, 3 in SLE, 1 in gout, 2 in JIA and 1 in unspecied
arthritis. Study details along with intervention classications are
in online supplementary table S3. The description of interven-
tions ranged from one-on-one instruction,25 group educa-
tion,26 telephone-based pharmacy advisory,27 to text message
reminders.28 Fourteen of 23 interventions were designed to
668
directly target medication adherence, with all but one interven-
tion29 directed at patients with rheumatology. This exception
was an instrument providing rheumatologists with structured
information about their patients non-adherence, which was not
shown to have an impact on adherence or disease-related
outcomes.29
With respect to focus, 21 interventions incorporated an edu-
cational component, one was a behavioural intervention involv-
ing text message reminders to patients with SLE,28 and the
aforementioned rheumatologist-targeted intervention was classi-
ed as affective.29 Fourteen interventions were multifaceted and
implementation was tailored to patients in 18 interventions.
Interventions were delivered by rheumatologists,20 3033
nurses,19 25 3336 pharmacists,25 27 29 3638 and other profes-
sionals including therapists39 and educators.4044
Effects of adherence interventions in rheumatic diseases
We critically appraised 13 RCTs10 in RA, 2 in SLE and 1 in
JIA. We did not include here a pilot RCT of group versus indi-
vidual counselling for RA34 and a pilot RCT of internet-based
self-management with telephone support for JIA.44 Quality
scores varied from 9 to 28 out of a maximum possible score of
32 (mean=18.6; median=18.5). We ranked 719 20 27 3840 43 as
having good or high quality (above median) and 6 as having fair
or poor quality. Key results of our synthesis are tabulated in
table 1. Detailed information on adherence and disease out-
comes are provided in online supplementary tables S4 and S5.
Rheumatoid arthritis
In McEvoy DeVellis et als42 study, all patients rst participated
in a psychosocial interview to assess problems related to their
RA. The intervention group then received a problem-solving
intervention to address problems identied in the interview.42
Authors cited no signicant difference between groups, though
did not report data.42 Authors also stated that all patients who
participated in study, regardless of group assignment, improved
on physical and psychological function measures, which may be
attributed to unintended positive effects of the psychosocial
interview received by both groups.42
Brus et al20 used group patient education meetings delivered
by a rheumatologist four times over 1 month and subsequently
at 4 months and 8 months for patients with recent-onset active
RA. From reported mean adherence rates of 8916 and 8421
in the intervention and control groups, respectively, we esti-
mated a SMD of 0.26 (95% CI 0.27 to 0.80). Changes from
baseline did not differ signicantly between groups on disease
activity score (DAS) and C reactive protein.20 Authors contex-
tualised that patients with active, recent-onset RA have high
levels of adherence, in line with previous ndings that patients
with high disease activity tend to be more adherent than those
with low disease activity,45 which may explain why their inter-
vention did not show an effect.
Hill et als19 intervention among patients with RA involved
seven 30-min individual monthly consults covering information
about RA medication and disease processes, pain control and
coping strategies. Authors reported signicant differences in
adherence rates in the intervention and control groups (89% vs
55%; RR, 1.55; 95% CI 1.14 to 2.10).19 Among disease out-
comes evaluated, authors reported an impact of the intervention
on plasma viscosity.19
Evers et al evaluated the effect of individualised cognitive-
behaviour therapy directly targeting frequently experienced pro-
blems in RA as well as indirectly, adherence. Authors reported
that adherence in the intervention group increased at the
Galo JS, et al. Ann Rheum Dis 2016;75:667673. doi:10.1136/annrheumdis-2014-206593

Figure 1 Systematic review study ow.
12 month follow-up assessment (t=2.08, p<0.05) while it
decreased in the control group (t=1.80, p=0.08). However, we
calculated a SMD of 0.43 (95% CI 0.07 to 0.93) based on
reported adherence values at follow-up assessment.39 Authors
reported favourable effects of the intervention on their primary
outcomes of fatigue and depression, however, they reported no
differences between groups in disease activity outcome.39
Clifford et als27 intervention was a telephone-based, patient-
tailored pharmacy advisory service delivered by community
pharmacists to elderly patients which included RA. Authors
reported 91% and 84% adherent patients in intervention and
control groups, respectively (RR, 1.08; 95% CI 1.00 to 1.17),
however this was based on the overall sample and no disease-
specic estimates were provided.27 No disease outcomes were
analysed.
Rudd et als43 RCT intervention indirectly addressed medica-
tion adherence through targeting literacy, with the intervention
group receiving two individualised counselling sessions. Authors
reported no difference in primary adherence between study
groups (SMD, 0.04; 95% CI 0.38 to 0.31). Findings were
inconsistent with respect to disease outcomes, with authors
showing improvement in mental health for the intervention
group, but not other outcomes evaluated. Authors surmised that
Galo JS, et al. Ann Rheum Dis 2016;75:667673. doi:10.1136/annrheumdis-2014-206593
Clinical and epidemiological research
patient characteristics, namely the fact that both groups had
high health literacy and long disease duration may have contrib-
uted to their intervention not altering medication taking
behaviour.43
El Miedany et al31 evaluated a tailored, multifaceted interven-
tion incorporating evaluation and review of patient reported
outcomes measures, education on skills for self-care and deci-
sion making, and a joint tness programme. Follow-up was
every 3 months for 18 months.31 Altogether, 89% and 64% of
intervention and control patients, respectively, were adherent
(RR, 1.39; 95% CI 1.15 to 1.68). Authors also reported that
the intervention signicantly improved patients DAS, pain,
ares, functional disability and quality of life.31
Conn et al40 evaluated the effect of the instructor-led, group-
based Arthritis Self-Management Program (ASMP), which we clas-
sied as an indirect intervention given that ASMP focuses on a
wide array of issues in RA. However, as appropriate use of medica-
tions was also taught in ASMP and adherence (to oral steroids and
methotrexate) was evaluated,40 we included the study. Authors did
not report adherence measures in their results but reported no dif-
ference in adherence between groups. Furthermore, they reported
similar percentages of patients achieving American College of
Rheumatology (ACR)20 in the ASMP and usual care groups.40
669
670

Clinical and epidemiological research

Table 1 Effect of adherence interventions in rheumatic diseases
Adherence outcome Classification of adherence intervention
Improved
Standardised mean Reported/ disease Direct/ Quality
Study difference (95% CI) Relative risk (95% CI) calculated outcome indirect Focus Complexity Implement Provider score

Rheumatoid arthritis (RA)

McEvoy DeVellis et al42 Not reported* Not reported D Educational affective Multifaceted Tailored patient educator 16.5
Brus et al20 0.26 (0.27 to 0.80) Calculated Not reported D Educational affective Multifaceted Generalised rheumatologist 18.5
Hill et al19 1.55 (1.14 to 2.10) Calculated Inconsistent D Educational affective Multifaceted Tailored rheumatology 26.5
nurse
Evers et al39 0.43 (0.07 to 0.93) Calculated Inconsistent I Educational affective Multifaceted Tailored trained therapists 21
behavioural
Clifford et al27 1.08 (1.00 to 1.17) Calculated Not analysed D Educational Single Tailored community 22
pharmacist
Rudd et al43 0.04 (0.38 to 0.31) Calculated Inconsistent I Educational behavioural Multifaceted Tailored educator 24
affective
Galo JS, et al. Ann Rheum Dis 2016;75:667673. doi:10.1136/annrheumdis-2014-206593

El Miedany et al31 1.39 (1.15 to 1.68) Calculated Y D Educational behavioural Multifaceted Tailored rheumatologist 13.5
affective
Conn et al40 Not reported N I Educational Single Generalised instructor 20
Ravindran and Jadhav32 1.18 (1.04 to 1.33) Calculated N D Educational Single Tailored rheumatologist 10.5
Zwikker et al38 0.96 (0.92 to 0.98)** Calculated N D Educational affective Multifaceted Generalised pharmacist 28
Systemic lupus erythematosus (SLE)
Ting et al28 <0.25 Reported Not reported D Behavioural Single Tailored n/a 16.5
Ganachari and Almas37 Not reported Not analysed I Educational Single Tailored pharmacist 9
Juvenile idiopathic arthritis (JIA)
Rapoff et al35 0.77 (0.07 to 1.47) Calculated N D Educational behavioural Multifaceted tailored nurse 16.5
*Not enough participants considered adherence as a problem in psychosocial interviews.
Did not report disease outcomes separately for intervention and control groups but stated in text that all patients improved on a number of global measures of physical and psychological functioning.
Calculated based on reported adherence values reported at follow-up assessment.
Study sample included patients with chronic diseases including rheumatoid arthritis (RA) but no RA-specific estimates reported.
Authors did not report adherence outcomes but stated in text that there was no difference in medication compliance between the groups.
**Calculated based on reported proportion of patients and not reported adjusted effect size.
Not enough data provided to allow calculation.

As patients in this study were African-Americans served by a public
hospital, ndings may be explained by sociodemographic factors
as only a half of patients randomised to ASMP attended four or
more sessions (due to reasons including lack of transportation,
inability to pay for bus).40
Ravindran and Jadhav32 described a rheumatologist-delivered,
education-focused intervention, leading to an agreed treatment,
and ongoing assessment of this treatment. Authors reported
98% of patients in the intervention group as adherent compared
with 83% in the control group (RR, 1.18; 95% CI 1.04 to
1.33).32 They reported lower DAS in 28 joints scores in the
intervention group compared with the control group, though
not statistically signicant (2.931.23 vs 3.261.47,
p=0.02).32
Zwikker et al38 evaluated pharmacist-delivered motivational
interviewing-guided group sessions, designed to improve
patients balance between necessity and concern beliefs about
medication to resolve practical barriers to medication taking.
No signicant difference in adherence outcomes was reported
for the control and intervention groups. Reasons discussed for
these ndings included potential researcher bias and selection
bias by including patients with a long disease duration (>14
years).38 Indeed, similar to the RCT by Rudd et al,43 modifying
adherence in patients with a long disease duration may be
harder to establish than forming new behaviour in recently diag-
nosed patients.43
Systemic lupus erythematosus
In Ganachari and Almass37 RCT, the intervention group
received three individualised counselling sessions from a clinical
pharmacist while the control group received routine counselling
from their physician. There was improved adherence in the
intervention group compared with the control group (mean 5.8
vs 4.6); however no SDs or p values were reported.37
Ting et al28 evaluated daily text message reminders to patients
with SLE in a two-part study. First, 70 patients received text
message reminders for clinic visits and authors reported a
decline in non-adherence to visits from 19% to 10% ( p=0.01).
In the adherence intervention substudy where 41 patients were
randomised to either text message reminders or usual care,
there were no signicant differences in adherence outcomes,
which authors attributed to the novelty of the intervention
wearing off due to its frequency.28
Juvenile idiopathic arthritis
Rapoff et al35 evaluated a nurse-administered intervention
involving one 30-min clinic visit involving a review and
rehearsal of adherence enhancement strategies followed by a
12-month problem solving educational session. Authors
reported signicant differences in adherence as measured by
Medication Event Monitoring Systems between the intervention
and control groups (77.721.5 vs 56.933.0, p=0.02).35
DISCUSSION
The objectives of this systematic review were to identify and
classify adherence interventions and assess evidence of their
effects on outcomes of patients with rheumatic diseases. We
identied 23 studies reporting adherence interventions that dir-
ectly or indirectly addressed treatment adherence in rheumatic
diseases. On further quality appraisal and synthesis of 13 RCTs
of adherence interventions, 4 studies had an effect on adherence
outcomes in RA,19 27 31 32 1 study35 had an effect on adherence
outcome in JIA, and no studies had an effect on adherence
outcome in SLE. Synthesising effects on disease outcomes was
Galo JS, et al. Ann Rheum Dis 2016;75:667673. doi:10.1136/annrheumdis-2014-206593
Clinical and epidemiological research
of greater challenge as only seven RCTs reported or analysed
disease outcomes19 31 32 35 39 40 43 and within these, there were
inconsistent ndings. Altogether, there are a limited number of
studies of medication adherence interventions in rheumatic dis-
eases, as compared with other chronic diseases, for example,
hypertension where a systematic review included 97 articles.16
As with other chronic diseases,18 effects of adherence interven-
tions in rheumatic diseases are disappointing with less than half
of appraised RCTs showing impacts on adherence outcome and
with inconsistent impacts on disease outcome.
In synthesising the interventions that were shown to impact
outcomes, relevant observations were made with respect to our
classication framework. First, in distinguishing whether the
intervention was specically designed to target medication
adherence (direct) or not (indirect), more direct interventions
had positive effects on adherence outcomes19 27 31 32 35 as com-
pared with indirect interventions. Second, with respect to focus,
12 of 13 RCT interventions incorporated at least an educational
component and all the interventions that were shown to have an
impact on adherence included an educational compo-
nent,19 20 27 31 32 35 highlighting the importance of educating
patients about their therapies, including proper administration,
and risks and benets. Third, of the 13 appraised RCTs, 8
involved multifaceted interventions, 3 of which showed positive
impact on adherence outcome.19 31 35 With no clear trend, this
adds to the debate on whether multifaceted approaches are
better as they address various adherence barriers15 22 35 or
single approaches are better as they are simpler and easily imple-
mented.46 Fourth, we did observe a clear trend with respect to
implementation of the interventions as all interventions shown
to have positive impacts were those that were tailored to
patients, which is in line with van Eijken et als15 suggestion
that improved adherence can be attained through tailored inter-
ventions. Finally, we also observed that effective interventions
were those delivered by a healthcare provider, whether a
nurse,19 35 pharmacist39 or rheumatologist.31 32
Lessons can also be drawn from studies that did not show
impacts of interventions, and we categorise reasons as those
related to either the intervention or characteristics of patients
studied. In McEvoys study, intervention and control groups
received a psychosocial interview, with the intervention group
receiving an additional problem-solving intervention.42 Thus, the
control group did not necessarily comprise an intervention-nave
group and unintended positive effects of the psychosocial inter-
view received by both groups may have contributed to lack of dif-
ferences in outcomes.42 With respect to characteristics of patients
included in the studies, an important one may be disease dur-
ation. Specically, Rudd et al43 and Zwikker et al38 had popula-
tions with RA with long disease durations (>14 years) and both
cited that modifying adherence in patients with a long disease
duration may be harder to establish than forming new behaviour
in recently diagnosed patients. Other patient characteristics to be
considered are sociodemographic factors, which may impact
access to40 or acceptance of37 the intervention.
Comments are also warranted with respect to the design of
studies themselves, particularly with outcome measurement.
Adherence measures varied across studies from question-
naires,26 2830 32 3739 4144 47 pill counts20 34 and rell pat-
terns.25 28 3436 38 There was further variation within measures,
for example questionnaires included validated instruments
(Morisky Scale,32 37 compliance questionnaire rheumatology
(CQR)29 38) and researcher developed questions.26 27 30 39 4143 47
A recommendation for future studies evaluating adherence inter-
ventions is standardisation of adherence measures to facilitate
671
 Clinical and epidemiological research
comparison. Another recommendation is a requirement for
studies to also incorporate disease outcomes, which was a limita-
tion of many of the included studies in our systematic review as
well as other reviews in other chronic diseases.21 Related to this is
that follow-up times in included studies were relatively short
(RCTs, mean 10 months; all studies mean 10.4 months). This may
be particularly relevant to disease outcomes which may manifest
over a longer term.
Strengths and limitations of our systematic review deserve dis-
cussion. Applying a structured framework to classify interventions
is a novelty of our systematic review that facilitated synthesis and
allows the opportunity to inform the design of future interven-
tions. Nonetheless, our systematic review may be vulnerable to
publication bias. Consideration of a study that targeted an inter-
vention to patients with arthritis without explicit information on
specic rheumatic diseases may also be a limitation,47 however we
did not appraise this study as it was a RCT. Finally, due to the het-
erogeneity across interventions, adherence and disease outcomes,
a quantitative meta-analysis was not performed.
Overall, our systematic review shows limited research on
adherence interventions in rheumatic diseases with inconsistent
impacts on adherence or disease outcome. Given the substantial
burden of treatment non-adherence across inammatory arth-
ritis, there is need for further work in designing and evaluating
interventions that promote and support treatment adherence.
Correction notice This article has been corrected since it was published Online
First. Rheumatology arthritis has been corrected to Rheumatoid arthritis (RA)
in table 1.
Contributors JSGExecuted searches, extracted data, interpreted ndings, drafted
and revised manuscript; PMinterpreted ndings, revised manuscript; SKR
interpreted ndings, revised manuscript; AA-Zinterpreted ndings, revised
manuscript; MADVExecuted searches, extracted data, interpreted ndings, drafted
and revised manuscript.
Funding MADV is a recipient of a Network Scholar Award from The Arthritis
Society/Canadian Arthritis Network and a Scholar Award from the Michael Smith
Foundation for Health Research. AA-Z is a recipient of a Scholar Award from the
Michael Smith Foundation for Health Research and the BC Lupus Society.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
REFERENCES
1 Bykerk V, Emery P. Delay in receiving rheumatology care leads to long-term harm.
Arthritis Rheumatol 2010;62:351921.
2 Goekoop-Ruiterman YP, de Vries-Bouwstra JK, Allaart CF, et al. Clinical and
radiographic outcomes of four different treatment strategies in patients with early
rheumatoid arthritis (the BeSt study): a randomized, controlled trial. Arthritis
Rheumatol 2005;52:338190.
3 Klareskog L, Catrina AI, Paget S. Rheumatoid arthritis. Lancet 2009;373:65972.
4 Lard LR, Visser H, Speyer I, et al. Early versus delayed treatment in patients with
recent-onset rheumatoid arthritis: comparison of two cohorts who received different
treatment strategies. Am J Med 2001;111:44651.
5 Mottonen T, Hannonen P, Korpela M, et al. Delay to institution of therapy and
induction of remission using single-drug or combination-disease-modifying
antirheumatic drug therapy in early rheumatoid arthritis. Arthritis Rheumatol
2002;46:8948.
6 Quinn MA, Conaghan PG, OConnor PJ, et al. Very early treatment with iniximab
in addition to methotrexate in early, poor-prognosis rheumatoid arthritis reduces
magnetic resonance imaging evidence of synovitis and damage, with sustained
benet after iniximab withdrawal: results from a twelve-month randomized,
double-blind, placebo-controlled trial. Arthritis Rheumatol 2005;52:2735.
7 Costedoat-Chalumeau N, Pouchot J, Guettrot-Imbert G, et al. Adherence to
treatment in systemic lupus erythematosus patients. Best Pract res Clin Rheumatol
2013;27:32940.
8 Li-Yu J, Clayburne G, Sieck M, et al. Treatment of chronic gout. Can we determine
when urate stores are depleted enough to prevent attacks of gout? J Rheumatol
2001;28:57780.
9 Neogi T, Hunter DJ, Chaisson CE, et al. Frequency and predictors of inappropriate
management of recurrent gout attacks in a longitudinal study. J Rheumatol
2006;33:1049.
672
10 Wu EQ, Patel PA, Mody RR, et al. Frequency, risk, and cost of gout-related
episodes among the elderly: does serum uric acid level matter? J Rheumatol
2009;36:103240.
11 Pelajo CF, Sgarlat CM, Lopez-Benitez JM, et al. Adherence to methotrexate in
juvenile idiopathic arthritis. Rheumatol Int 2012;32:497500.
12 De Vera MA, Marcotte G, Rai S, et al. Medication adherence in gout: a systematic
review. Arthritis Care Res 2014;66:15519.
13 Salt E, Frazier SK. Adherence to disease-modifying antirheumatic drugs in patients
with rheumatoid arthritis: a narrative review of the literature. Orthop Nurs
2010;29:26075.
14 de Achaval S, Suarez-Almazor ME. Treatment adherence to disease-modifying
antirheumatic drugs in patients with rheumatoid arthritis and systemic lupus
erythematosus. Int J Clin Rheumatol 2010;5:31326.
15 van Eijken M, Tsang S, Wensing M, et al. Interventions to improve medication
compliance in older patients living in the community: a systematic review of the
literature. Drugs Aging 2003;20:22940.
16 Gwadry-Sridhar FH, Manias E, Lal L, et al. Impact of interventions on medication
adherence and blood pressure control in patients with essential hypertension:
a systematic review by the ISPOR medication adherence and persistence special
interest group. Value Health 2013;16:86371.
17 Williams JL, Walker RJ, Smalls BL, et al. Effective interventions to improve
medication adherence in Type 2 diabetes: a systematic review. Diabetes Manag
2014;4:2948.
18 Nieuwlaat R, Wilczynski N, Navarro T, et al. Interventions for enhancing medication
adherence (Review). Cochrane Database Syst Rev 2014;11:CD000011.
19 Hill J, Bird H, Johnson S. Effect of patient education on adherence to drug
treatment for rheumatoid arthritis: a randomised controlled trial. Ann Rheum Dis
2001;60:86975.
20 Brus HL, van de Laar MA, Taal E, et al. Effects of patient education on compliance
with basic treatment regimens and health in recent onset active rheumatoid
arthritis. Ann Rheum Dis 1998;57:14651.
21 Viswanathan M, Golin CE, Jones CD, et al. 4. Medication Adherence Interventions:
Comparative Effectiveness Closing the Quality Gap: Revisiting the State of the
Science. 2012. Report No.: 208.
22 Roter DL, Hall JA, Merisca R, et al. Effectiveness of interventions to improve patient
compliancea meta-analysis. Med Care 1998;36:113861.
23 The Consort Group. Consort 2010. http://www.consort-statement.org/consort-2010.
24 Cohen J. Statistical power analysis for the behavioral sciences. 2nd edn. USA:
Lawrence Erlbaum Associates, 1988. 567p.
25 Bond CA, Monson R. Sustained improvement in drug documentation, compliance,
and disease control. A four-year analysis of an ambulatory care model. Arch Intern
Med 1984;144:115962.
26 Lindroth Y, Bauman A, Barnes C, et al. A controlled evaluation of arthritis
education. Br J Rheumatol 1989;28:712.
27 Clifford S, Barber N, Elliott R, et al. Patient-centred advice is effective in improving
adherence to medicines. Pharm World Sci 2006;28:16570.
28 Ting TV, Kudalkar D, Nelson S, et al. Usefulness of cellular text messaging for
improving adherence among adolescents and young adults with systemic lupus
erythematosus. J Rheumatol 2012;39:1749.
29 van den Bemt BJ, den Broeder AA, van den Hoogen FH, et al. Making the
rheumatologist aware of patients non-adherence does not improve medication
adherence in patients with rheumatoid arthritis. Scand J Rheumatol
2011;40:1926.
30 Berry D, Bradlow A, Bersellini E. Perceptions of the risks and benets of medicines
in patients with rheumatoid arthritis and other painful musculoskeletal conditions.
Rheumatology 2004;43:9015.
31 El Miedany Y, El Gaafary M, El Arousy N, et al. Arthritis education: the integration
of patient-reported outcome measures and patient self-management. Clin Exp
Rheumatol 2012;30:899904.
32 Ravindran V, Jadhav R. The effect of rheumatoid arthritis disease education on adherence
to medications and followup in Kerala, India: to the editor. J Rheumatol 2013;40:14601.
33 Rees F, Jenkins W, Doherty M. Patients with gout adhere to curative treatment if
informed appropriately: proof-of-concept observational study. Ann Rheum Dis
2013;72:82630.
34 Homer D, Nightingale P, Jobanputra P. Providing patients with information about
disease-modifying anti-rheumatic drugs: Individually or in groups? A pilot randomized
controlled trial comparing adherence and satisfaction. Musculoskelet 2009;7:7892.
35 Rapoff MA, Belmont J, Lindsley C, et al. Prevention of nonadherence to
nonsteroidal anti-inammatory medications for newly diagnosed patients with
juvenile rheumatoid arthritis. Health Psychol 2002;21:6203.
36 Stockl KM, Shin JS, Lew HC, et al. Outcomes of a rheumatoid arthritis disease
therapy management program focusing on medication adherence. J Manag Care
Pharm 2010;16:593604.
37 Ganachari MS, Almas SA. Evaluation of clinical pharmacist mediated education and
counselling of systemic lupus erythematosus patients in tertiary care hospital. Ind J
Rheumatol 2012;7:712.
38 Zwikker HE, van den Ende CH, van Lankveld WG, et al. Effectiveness of a
group-based intervention to change medication beliefs and improve medication
Galo JS, et al. Ann Rheum Dis 2016;75:667673. doi:10.1136/annrheumdis-2014-206593