Twumasi et al.

Arthritis Research & Therapy (2019) 21:155

https://doi.org/10.1186/s13075-019-1942-7

RESEARCH ARTICLE Open Access

Health service utilization among African

American women living with systemic
lupus erythematosus: perceived impacts of
a self-management intervention
Abena A. Twumasi1, Anna Shao1, Charmayne Dunlop-Thomas2, Cristina Drenkard2* and Hannah L. F. Cooper1

Abstract
Background: Healthcare access, utilization, and quality play critical roles in shaping mortality and morbidity among
patients diagnosed with systemic lupus erythematosus (SLE), and yet healthcare access, utilization, and quality can
be suboptimal for many people living with SLE. The aim of this qualitative study was to explore the perceived impact
of a peer-led, group-based educational intervention (the Chronic Disease Self-Management Program [CDSMP]) on
healthcare engagement behaviors among African American women with SLE.
Methods: Participants were recruited from the WELL (Women Empowered to Live with Lupus) study, a behavioral trial
of the effectiveness of the CDSMP on African American women diagnosed with SLE. We conducted two waves of
qualitative, one-on-one, semi-structured interviews with 24 purposively sampled WELL participants; one interview was
conducted before CDSMP participation and one after. Wave 1 interviews explored health service use behaviors at
baseline; Wave 2 interviews focused on changes in these behaviors post-intervention and women’s perceptions of
whether and how the CDSMP shaped these changes. Transcripts were analyzed using thematic analysis methods.
Results: Study participants perceived the CDSMP to be a valuable resource for supporting two distinct health service
use behaviors: communicating with doctors (N = 16 [88.9%]) and managing medication side effects (N = 17 [41.2%]).
Women perceived that the CDSMP had the most potent and widespread effects on patients’ communication with
doctors. Strategies that women believed generated improvements in patient-doctor communication included
enhancing preparation for appointments and boosting patient participation during doctor’s visits. Women’s
reported post-CDSMP improvements in health service use behaviors varied by disease severity and depression.
Insurance coverage, while not probed directly during baseline interviews, emerged organically as a key factor
affecting health service use behaviors; the CDSMP did not seem to improve participants’ ability to circumvent
insurance-related barriers to accessing care.
Conclusions: Our findings suggest that the CDSMP may help enhance healthcare service utilization among
African American women with SLE by improving doctor/patient communication and medication side effect
management. If future research confirms this conclusion, African American women living with SLE should be
encouraged to participate in CDSMP workshops to enhance health service use behaviors.
Trial registration: NCT02988661. Registered 12/07/2016
Keywords: Systemic lupus erythematosus, Health service utilization, Qualitative methods

* Correspondence: cdrenka@emory.edu
2
Department of Medicine, Emory University School of Medicine, Atlanta, GA,
USA
Full list of author information is available at the end of the article

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Twumasi et al. Arthritis Research & Therapy (2019) 21:155 Page 2 of 11

Introduction healthcare utilization in people with chronic conditions

Systemic lupus erythematosus (SLE) is a chronic auto- [27]. For instance, findings from a longitudinal nation-
immune condition characterized by a broad spectrum of wide study reported 5% reduction in ER visits at both 6
clinical manifestations, including, pain, skin rashes, arth- and 12 months, and 3% reduction in hospitalizations at
ritis, and damage to vital internal organ systems [1]. Al- 6 months [28, 29]. While these and other studies suggest
though therapeutic advances in recent years have led to that generic self-management interventions may contrib-
improved health outcomes among SLE patients [2], racial/ ute to improve healthcare service use, their findings may
ethnic disparities persist [3]. Compared to White adults, not generalize to people living with lupus because these
African American adults have a threefold higher incidence studies often sampled older and White participants. In
and prevalence of SLE [4–7], experience more severe dis- contrast, lupus patients tend to be younger and African
ease activity, and have a worse prognosis [8, 9]. African American, and may thus face unique psychosocial bar-
American women are especially vulnerable to SLE morbid- riers to healthcare service use. For instance, socioeco-
ity and mortality: SLE prevalence is 10- to 15-fold higher nomically disadvantaged patients with lupus and other
in women compared to that in men [4–7]. rheumatic diseases are likely to experience treatment
Healthcare access, healthcare use, and quality of non-adherence, financial problems, fear of side effects,
healthcare play a critical role in disease outcomes and difficulty navigating the health system, and perceived
health disparities among patients with SLE [3, 10–13]. treatment inefficacy [30].
For example, patients with SLE who reported poor Currently, there are no widely available, evidence-
patient-provider communication and poor care coordin- based, SLE-specific self-management resources. The SLE
ation showed significantly greater accrual of disease Self-Help Course (SLESH) was the only SLE-specific
damage over time [14]; conversely, SLE patients who ac- self-management intervention in the USA in the 1990s
tively engaged clinicians during consultations showed [31]. Despite promising results and valiant effort by the
less organ damage over a nearly 5-year median Arthritis Foundation to disseminate SLESH across the
follow-up period [15]. Unfortunately, significant ra- USA, the program has not been available since the late
cial/ethnic inequities exist in healthcare use among 1990s (Personal communication with J Galloway, on July
SLE patients: an analysis of hospital and emergency 3, 2012). One challenge of SLESH in particular is the
room (ER) admissions over a 13-year period for inci- relatively significant effort needed to implement and dis-
dent SLE patients from the Georgia Lupus Registry seminate the program in light of the relatively low
(GLR) revealed that African American adults had a prevalence of SLE. Thus, effective self-management in
significantly higher rate of hospital and ER admissions SLE comes from what patients acquire on their own or
compared to White adults [10]. from their providers.
Educational self-management programs are integral to The present longitudinal, qualitative study is designed
the chronic disease standard of care [16–19] and may to explore women’s perceptions of the ways that one
help patients better navigate healthcare services. These self-management program, the Chronic Disease Self-
programs aim to empower individuals in the utilization Management Program (CDSMP), affects healthcare ser-
of social support resources to optimize health and re- vice engagement among African American women living
duce the impact of disease on daily functioning [20]. with SLE. The CDSMP, a globally recognized, evidence-
Self-management education is a process helping patients based self-management intervention, is a group-based
to bridge the gap between disease knowledge and workshop designed to foster self-efficacy and encourage
health-related behaviors [21]. Evidence-based self- patients to actively partner with their healthcare pro-
management education programs have been designed to viders to optimize their care [24]. Because the CDSMP
increase patients’ skills and confidence in managing their is a cost-effective, nationally available intervention, it
symptoms and medications, as well as to teach patients could potentially facilitate reduction in use of healthcare
how to effectively communicate with healthcare pro- services and help mitigate high medical care costs
viders and navigate the healthcare system [22–25]. through improved self-management among African
While the self-management needs, goals, and strategies American women living with SLE. The underpinning
of each patient may vary, Lorig has identified five core mechanism of the behavioral change driven by the
skills that enhance self-management: (a) problem solv- CDSMP is based on Bandura’s self-efficacy theory, which
ing, (b) decision making, (c) resource utilization, (d) for- in turn is derived from the social cognitive theory. Peo-
mation of patient-clinician partnerships, and (e) taking ple’s perceived efficacy in their ability to produce desired
action [25, 26]. effects by their action may influence their behavior pat-
A growing body of evidence suggests that in addition terns [32]. Moreover, beliefs about individual self-
to improving health-related behaviors and disease out- efficacy develop from cognitive appraisal of information
comes, self-management programs can also impact that arises from performance, vicarious experience,
Twumasi et al. Arthritis Research & Therapy (2019) 21:155 Page 3 of 11

verbal persuasion, and emotional arousal or physio- second interview (“Wave 2”) was conducted 2–6 weeks
logical feedback [32]. after the CDSMP had ended. This pre-/post-interven-
A pilot study conducted by our group found that the tion design allowed us to explore African American
CDSMP was accepted by African American women with women’s perceptions of their self-management behav-
lupus [33]. Moreover, pre-post intervention quantitative iors “in real time” before and after participating in the
data showed significant improvements in physical health, CDSMP, and learn their assessment of the processes
cognitive symptom management, self-efficacy, communi- through which the CDSMP had affected these behav-
cation with physicians, and medication adherence [33]. iors. The semi-structured interview guides are available
In this study, we examined whether African American as Additional file 1.
women with SLE perceived that participation in the The Wave 1 guide covered women’s perceptions of (1)
CDSMP affected their health care engagement behaviors, the SLE diagnosis experience, (2) SLE symptoms and
and the perceived processes through which the program how they affected women’s health-related quality of life,
might have enhanced these behaviors. and (3) health service engagement and doctor/patient
communication. This guide was reviewed and piloted
Materials and methods with two African American women living with SLE who
We used a pre-/post-intervention qualitative design to were on the WELL advisory board; these advisors shared
explore our research question with a sample of African insights into how to strengthen the domains and items.
American women living with SLE. We chose to use The final Wave 2 guide covered participant perceptions
qualitative methods because they excel at capturing per- of the (1) workshop and (2) changes in health service
ceptions and process [34]. use behaviors; when participants did not attend all six
sessions, we asked about reasons for missing sessions.
Participant recruitment When Wave 2 participants only attended one class, in-
Participants were recruited from the WELL (Women terviewers did not probe perceptions of CDSMP impact.
Empowered to Live with Lupus) study, a behavioral trial Two trained interviewers (both young adult women),
of the effectiveness of the CDSMP on African American AS and AT, conducted all interviews in a private space
women diagnosed with SLE (Trial registration number accessible to participants. Wave 1 interviews lasted ap-
NCT02988661). For this study, we conducted a longitu- proximately 85 min; Wave 2 interviews lasted approxi-
dinal quantitative assessment of the CDSMP’s impact on mately 65 min. Interviews were audio-recorded and
SLE-related outcomes among African American women. transcribed verbatim.
WELL randomly sampled 150 African American women
from the GOAL (Georgians Organized Against Lupus) Analysis
cohort; GOAL is representative of the full disease and Transcripts were analyzed using thematic analysis
sociodemographic spectra of those living with SLE in the methods [43], and we searched for variations by partici-
southeastern USA [35]. We purposively sampled 24 Afri- pant age, educational attainment, patient-reported de-
can American women from the WELL cohort to take pressive symptoms (referred since now as depression),
part in this qualitative substudy. A priori, we set a target and severity of SLE activity; given that some people did
sample size of 24 women because we believed we would not attend all CDSMP sessions, we also examine varia-
reach “theoretical saturation” for key themes by the 24th tions by the number of sessions attended. Consistent
interview (i.e., the point at which no new themes with Lorig’s definition, attendance of four or more ses-
emerge) based on past studies with SLE women about sions was deemed as CDSMP completion [24, 25]. To
their perceptions of healthcare [36, 37]. Consistent with create a preliminary Wave 1 codebook, the first 3 tran-
purposive sampling, the qualitative study sought vari- scripts were read repeatedly by AS, AT, and HC to de-
ation in the sample by factors that might affect re- velop a list of codes and their definitions. Two analysts
sponses to the CDSMP, including baseline depressive (AS and AT) then independently applied these codes to
symptoms, SLE disease activity, age, and education [38– these three transcripts. We then compared coded tran-
42]. Data on these characteristics were drawn from scripts. Where inter-coder differences in coding were
WELL surveys, which also provided information on identified, we discussed the causes of these differences
sample sociodemographic characteristics. (e.g., unclear definitions, coding errors) and resolved the
differences (e.g., clarifying definitions, correcting codes).
Data collection Thereafter, AT and AS each coded half of all Wave 1
We conducted two waves of one-on-one, semi- transcripts; to enhance inter-coder agreement, every
structured interviews with each participant. The first third transcript was “double coded” by AS and AT, and
interview (“Wave 1” or “baseline”) was conducted in the coding differences were identified and resolved, as de-
month before the participant started the CDSMP; the scribed above. The team developed themes through
Twumasi et al. Arthritis Research & Therapy (2019) 21:155 Page 4 of 11

extensive memoing (i.e., writing analytical notes analyt- Table 1 Demographic information for African American women
ical notes capturing thoughts, questions, and/or synthe- in the WELL sample (n = 24)
sis of the data). Characteristic Value
A Wave 2 codebook was developed and applied Current age, years, mean (range) 48.6 ± 13.5
through a similar process. Analyses were conducted Age at diagnosis, years, mean (range) 34.5 ± 8.7
using MAXQDA version 12.3.2 [44]. As noted, some Years since diagnosis, mean ± SD 14.1 ± 8.1
participants (N = 6) only attended one CDSMP session, Educational attainment, n (%)
and interviewers did not probe perceived CDSMP im- Less than high school 2 (8.3)
pact. We excluded these individuals from the Wave 2 High school/some college 10 (41.7)
analysis, unless they volunteered information on per-
Completed college 12 (50.0)
ceived CDSMP impact on a particular health service use
Depressed, n (%)
behavior.
No (PROMIS Depression T-score < 56) 13 (54.2)
Yes (PROMIS Depression T-score > 56) 11 (45.8)
Ethics
The study was approved by the Emory Institutional Re- Severity of disease activity score (SLAQ), n (%)

view Board and Grady Health System Research Over- Mild disease activity (score 0–10) 6 (25.0)
sight Committee. All study participants signed informed Moderate disease activity (score 11–16) 7 (29.2)
consent. Severe disease activity (score ≥ 17) 11 (45.8)
Severity of organ damage score (SA-BILD), n (%)
Results No organ damage (score 0) 8 (33.3)
Participant characteristics Mild organ damage (score 1–2) 8 (33.3)
Twenty-four women participated in the baseline qualita- Moderate to severe organ damage (score ≥ 3) 8 (33.3)
tive interviews; 23 women also completed Wave 2 inter-
Current household income, n (%)
views. As depicted in Table 1, on average, participants
Less than $20,000 11 (45.8)
were 48.6 years old (standard deviation [SD] = 13.5) and
$20, 000–$49,000 9 (37.5)
had been diagnosed with SLE 14 years ago (SD = 8.1). By
$50,000+ 3 (16.7)
design, the sample was balanced with regard to educa-
tion, depression, and SLE severity (disease activity and Refuse to answer 1 (4.2)

organ damage). Most participants (91.7%) were currently Employment, n (%)

employed; one third had an annual income of less than Working full-time or part-time 22 (91.7)
$20,000; and about half were currently uninsured or Unemployed/retired/student/homemaker/disabled 2 (8.3)
underinsured. Over half (62.5%) of participants were Insurance status , n (%)
a

currently in a relationship and about half (54.2%) re- No insurance or underinsured 11 (45.8)
ported living in the household with a child. Possibly, Insured 13 (54.2)
CDSMP attendance was slightly lower for the 15 women Below 100% povertya, n (%)
who were currently in a relationship (7/15 attended <
No 14 (58.3)
3 sessions vs. 3/9 women who were not in a relation-
Yes 10 (41.7)
ship). Attendance was similar among women who re-
Number of CDSMP sessions attendedb, n (%)
ported living with a child in the household (5/12
1–3 sessions 9 (37.5)
attended < 3 sessions) and those who did not (5/13
attended < 3 sessions). 4–6 sessions 14 (58.3)
Currently in a relationship, n (%)
Qualitative findings Yes 15 (62.5)
The qualitative analysis identified three distinct domains No 9 (37.5)
of perceived CDSMP impact: communication with doc- Number of children in household, n (%)
tors; medication side effect management; and insurance’s 0 children 13 (54.2)
effects on healthcare engagement (the latter domain 1–2 children 8 (33.5)
emerged from the transcripts and was not a part of the 3–4 children 2 (8.3)
original guide). We describe findings domain by domain,
5–6 children 1 (4.2)
presenting baseline findings and then post-CDSMP
Abbreviations: PROMIS Patient-Reported Outcomes Measure System, SLAQ
changes within each domain. Six participants only Systemic Lupus Activity Questionnaire, SA-BILD Self-administered Brief Index of
attended one workshop. We did not ask these partici- Lupus Damage. aData on insurance and poverty were obtained from the GOAL
survey, and all other demographics from the WELL Baseline survey. bOne
pants about perceived CDSMP impact because of their participant completed the pre-intervention interview but did not attend
slight exposure. Depending on the domain, however, any sessions
Twumasi et al. Arthritis Research & Therapy (2019) 21:155 Page 5 of 11

some of these six participants volunteered information Some participants were dismayed that providers auto-
about CDSMP impact on health service engagement. matically ascribed any new health problems to SLE.
When such data were available, we analyzed them and Others were concerned that physicians, regardless of
include them in our findings. For the sake of transpar- specialty, trivialized non-SLE health problems. One
ency, we note the number of Wave 2 participants who woman disclosed that her rheumatologist had ignored
provided Wave 2 data for each domain (see Table 2. her complaints about symptoms in her finger, attributing
them to an ordinary “nail fungus;” because of delayed
Communication with doctors care, she had to have a partial finger amputation.
Pre-CDSMP At baseline, 10 out of 24 participants re- Conversely, women (N = 14) who were satisfied with
ported dissatisfaction with their communication with their communications with healthcare providers at base-
their healthcare providers; 14 women were satisfied with line praised them for sitting with them and taking the
their provider communication (Table 2). Women re- time to explain medical jargon, test results, and treat-
ceived care from several providers (e.g., nurses, physi- ment options, and helping women feel that “no question
cian’s assistants, rheumatologists), and in the following is too stupid.” These patients reported that they could
sections, women’s accounts may refer to any one, or sev- be transparent about all health concerns and had an
eral, of these healthcare providers. Our analysis indicated agency to discuss their treatment options, refuse new
that women who expressed dissatisfaction with provider medication, and negotiate changes in dosages. As one
communication tended to be less educated; no differ- participant expressed,
ences were observed by age, depression status, or SLE
severity. [The rheumatologist] takes the time out to sit down
There were multiple reasons why participants were and talk with me for a minute and just try to figure
dissatisfied with their communications with providers: out, like, what’s going on. So he just seems – he
women critiqued providers for having a discourteous seems more sincere to me, because I’ve had it – I’ve
“bedside manner;” providing too little information (e.g., been to plenty of doctors where they’re not as great
about comorbidities, medication); being unapproachable; and they don’t care. They spend five minutes in there
listening poorly; lacking empathy towards patient’s de- with you, but that’s why I like … . He listens to me …
scription of pain; and rushing them. This participant
expressed concerns about her provider’s ability to listen (Age 28, College graduate, Mild Moderate Lupus,
to her: Depressed)

Um, I never really hear them say too often, like, “Do
you have any questions or concerns?” I don’t really Post-CDSMP Sixteen out of 18 women reported that
hear them say that too much. They just kind of just – the CDSMP helped enhance their communication with
they just kind of review what you’re on and then they doctors; eight of these women had been dissatisfied with
just, you know, start talking about your medicines. communication at baseline and eight had been satisfied
Sometimes I wish they’ll listen, like just really listen to (Table 2). Five participants were excluded from the
you … You know, it’s a short period of time that Wave 2 analysis because they only attended one CDSMP
you’re in and out, I just feel like it’s like, “Okay, you’ve session and volunteered no information on communica-
got 10 minutes …” tion with providers at Wave 2. Women who reported
post-CDSMP improvements in their communication
(Age 46, College graduate, Mild Moderate Lupus, Not with healthcare providers were more likely to have mild
Depressed) SLE (N = 9), or be non-depressive (N = 9). Participants
who attended 4 or more workshop sessions (N = 13)
Women also worried that their SLE medical history af- were more likely to report improvements in communica-
fected their communication with healthcare providers. tion with providers.
Table 2 Summary of patient’s perceived CDSMP impact on health service utilization by domain
Domain Perceived as satisfactory Perceived as unsatisfactory
Pre-CDSMP Post-CDSMP* Pre-CDSMP Post-CDSMP*
Communication with doctors 14/24 16/18 10/24 2/18
Medication side effect management 3/24 7/17 21/24 10/17
Insured 12/24 14/23 12/24 9/23
*Six participants only attended one session. We did not ask these participants about perceived CDSMP impact because of their minimal exposure. Some of these
participants, however, volunteered information about perceived CDSMP impacts on communication, medication side effect management, or insurance
Twumasi et al. Arthritis Research & Therapy (2019) 21:155 Page 6 of 11

Eight out of 10 participants who expressed dissatis- Medication side effect management
faction with healthcare provider communication at Pre-CDSMP At baseline, 21 out of 24 women reported
baseline noted improvement in this area after the experiencing at least one side effect as a result of SLE
workshop. Several of these women reported that the medications (Table 2). Three participants did not experi-
CDSMP had taught them to prepare for doctor’s visits ence any adverse side effects, one because she was not
using several strategies. Strategies included making currently taking medications for her SLE symptoms. In
lists of their current medications, questions, new the following sections, we only describe experiences of
symptoms, medication side effects, or other concerns the 21 women who reported at least one medication side
they would like to discuss during the visit. At base- effect at baseline. Typically, women suffered from mul-
line, only one woman reported making such lists tiple SLE-medication-related side effects at a time; we
prior to seeing the doctor. List making helped women focus on those side effects that were most commonly
to remember to ask key questions about their health mentioned. Our analysis did not reveal patterns regard-
and care. Discussions at the CDSMP also fostered a ing women’s experiences by demographic characteristics
desire among women for more open communication or health status.
between them and healthcare providers regarding Several women (N = 16) reported experiencing weight
their SLE. Women reported that facilitated discus- fluctuations and facial bloating linked to SLE medica-
sions during the CDSMP about being “rushed” during tions. Weight gain was most commonly reported by par-
doctor’s visits empowered them to insist on “get [ting] ticipants (N = 12), although a few participants (N = 2)
what paid for” by demanding that providers spend had experienced both weight gain and loss. Four women
sufficient time with them. Two women remained dis- described facial bloating, commonly referred to as
satisfied with their communication with healthcare “moon face” or “Prednisone face.” For many, weight fluc-
providers (Table 2). tuations and facial bloating were the most challenging of
Eight out of 13 women who had reported satisfactory all medication side effects reported. Such fluctuations
communication with doctors at baseline experienced en- harmed women’s body image, particularly when they
hanced communication after the CDSMP. Similar to the gained weight:
improvements reported by participants above, women
noted that the CDSMP workshop encouraged them to I never been a thick person... I never been this size in
be more open about their SLE pain and symptoms when my life. Before kids I’ve never weighed over 170
communicating with providers and reminded them that pounds full term so I was really skinny … when I
they had an agency to ask questions, particularly about started taking the Prednisone, I just blew up like crazy
new medications. One participant, for example, de- and um I didn’t like that.
scribed how the class had helped her:
(Age 58, High School graduate, Severe Lupus,
Uh, I’ve seen a couple [of doctors], rheumatology, Depressed)
neurology [since completing the CDSMP]. Um, they
went fine. I think I was able to speak a little bit more Many women reported that managing weight fluctu-
to them. Um, cause I think before [the CDSMP] I ations was outside of their control: “I couldn’t control
was, I didn’t complain as much about the medication it [weight gain] at the time.” Other SLE symptoms
not helping. Um, but [now] I have kind of let them (e.g., fatigue) and medication side effects often pre-
know that this [medication], it’s not working for me. vented women from engaging in behaviors to manage
None of this stuff [medication] has helped at all. And their weight. One woman mentioned exercising to
I’m still in a lot of pain every day. So that’s why they combat weight gain; however, recent health complica-
gave me the additional [medication]. tions and surgery had prevented her from exercising
regularly.
(Age 50, College graduate, Severe Lupus, Depressed) Five women stated that they experienced deteriorating
or blurred vision as a result of SLE medications. At least
Many women described that they had known about two participants attributed the deterioration of their vi-
these strategies previously, but that the workshop en- sion to hydroxychloroquine. Three participants de-
couraged them to act on this knowledge. scribed that they often saw “little spots,” commonly
Two participants reported no change in doctor-patient referred to as “floaters,” that interfered with their vision.
communication: one did not recall the session in which Women feared that eventually they might lose their vi-
this topic was discussed; the other reported that she did sion completely. A participant shared that she avoided
not learn anything new about communication from the driving at night as a result of her poor eyesight and
workshop. double vision:
Twumasi et al. Arthritis Research & Therapy (2019) 21:155 Page 7 of 11

… what has changed, and I believe it’s because the or intensity of women’s medication side effects, it
medications have changed, is my vision. [long pause] seemed to help them refocus their energies:
At night, um … I have decreased my driving at night,
um. My boyfriend drives more at night. Um, because I think [with] the lower dosage I don’t see [my facial
the best way to explain it, um, I see multiple images bloating]. And maybe it’s there. I just don’t see it. Um,
of things... So if I’m goin’ down the highway I see I don’t know. It doesn’t bother me. Um, I think if I
trailings of the street signs, or um, cars... start to see it would bother me, but I don’t see it.
Maybe just fooling myself, but I do have the round
(Age 50, College graduate, Severe Lupus, Depressed) Prednisone face, but I just don’t see it, so, again,
putting negative thoughts out of my head. If I get a
A few (N = 4) participants mentioned that their SLE round face I’m healthy, I’m alive, I’m here, I’m not in
medications, particularly Prednisone, caused their skin pain, so, I’m dealing with it … Being around, again,
to become hypersensitive and bruise easily. Women just like a support group. You hear that often enough
had also experienced rashes, typically on the face, you really start to believe it and every time I went in
arms, and legs. Some reported that these rashes could there [CDSMP workshop] they’d reinforce what we
be “itchy.” Rashes harmed women’s body image and learned last week or just like you – you’re more than
made them feel self-conscious, especially when rashes your disease; that you stay up, don’t stay down –
occurred on their face. To cope, women wore pro- don’t - down. Your disease will spiral if you go down.
tective clothing, including long-sleeved shirts and You got to stay up.
long dresses, both to hide rashes from public view
and protect their skin from overexposure to sunlight, (Age 54, College graduate, Mild Moderate Lupus, Not
which could intensify irritations. Depressed)

Among the remaining 10 participants (Table 2) who

Post-CDSMP Seven of the 17 women who experienced did not report any changes in their management of side
side effects at baseline reported that the CDSMP helped effects, 9 reported that they did not learn any useful
them to improve the management of these side effects skills that helped them manage side effects; 1 had
(Table 2). Four participants were excluded from the stopped taking the medications that caused her side
Wave 2 analysis because they only attended one CDSMP effects.
class and volunteered no information on side effect
management at Wave 2. With one exception, women Insurance
who reported having improved side effect management Women reported that insurance coverage was a signifi-
were older (49.6 years and above) and had attended at cant factor affecting how they utilized healthcare ser-
least 4 CDSMP classes. vices and managed their disease; while not probed
Following the CDSMP, women felt equipped to [1] en- directly during interviews, this theme emerged organic-
gage their healthcare providers to either reduce the dos- ally from women’s discussions of healthcare utilization.
age of medications with overly detrimental side effects,
or to explore alternate types of medications; or [2] better Pre-CDSMP Based on participants’ views, we identified
accommodate medication side effects. Women discussed two distinct groups of women at baseline: participants
three key CDSMP-inspired strategies that aided them to who report ever being uninsured (including those who
improve medication side effect management. One strat- were currently uninsured; N = 12) and participants who
egy was to discuss medications with doctors, to review were currently insured and did not report ever being un-
medications they were taking with the doctors regularly insured in the past (N = 12) (Table 2). Current insurance
and explore alternate medications with less severe side coverage was strongly patterned by age, education, and
effects, as needed. Another strategy learned from the disease severity. Women who were older, had fewer
CDSMP was to adhere to prescription labels and in- years of education, or had mild/moderate SLE were
structions. One woman, for example, learned to identify more likely to be uninsured than women who were
which of her medications were required to be taken with younger, college educated, or had severe SLE. Many cur-
meals, which in turn had helped reduce her previously rently insured participants (N = 8) were enrolled in ei-
persistent stomach irritations. A third strategy women ther Medicaid or Medicare as part of their social
learned from the CDSMP was practicing positivity (i.e., security disability benefits. Notably, five women reported
shifting focus away from difficulties with medication side that their SLE diagnosis or symptoms had caused them
effects to more positive aspects of their lives). While to lose their jobs and that this job loss in turn caused
practicing positivity may not have changed the frequency them to lose their health insurance. Some women
Twumasi et al. Arthritis Research & Therapy (2019) 21:155 Page 8 of 11

worried about their inability to receive approval for dis- group) reported any improvement in health insurance
ability benefits despite their diagnosis, noting that “… it’s coverage or navigation after participating in the CDSMP:
so hard for lupus people to get disability.” two reported that they had started to maximize their use
While uninsured, these women used multiple strat- of covered services since Wave 1, and one reported that
egies to manage their SLE, including purchasing over- she had started negotiating with healthcare and insur-
the-counter pain medications that were affordable but ance providers to receive additional benefits since Wave
often less potent; going into debt to pay for emergency 1. Two out of these three participants attributed the
room visits; paying out-of-pocket; enrolling in research changes directly to the CDSMP. Our analysis did not re-
studies to obtain blood tests; and seeking care from hos- veal patterns in women’s responses by health status,
pitals that had special policies for the uninsured. Several demographic characteristics, or workshop attendance.
reported forfeiting needed medications:
Discussion
… when I first had to leave work … I did not have The present qualitative study explored whether and how
insurance so I didn’t have um any of my medications participating in the CDSMP affected healthcare use,
from the perspectives of African American women living
(Age 50, College graduate, Severe Lupus, Depressed) with SLE. We found evidence that many women in this
sample perceived the CDSMP as a valuable resource to
Three participants reported that lack of insurance pre- improve healthcare use behaviors, including communi-
vented them from [1] seeking needed medical care from cating with doctors and managing medication side ef-
preferred hospitals [2], seeking care from specialists when fects. Many women also volunteered that insurance was
referred, and [3] obtaining timely care (e.g., some reported one of the key barriers to healthcare use, though the
wait times up to 3 months). One woman reported, CDSMP did not seem to affect this barrier.
Our analysis suggests that the CDSMP had the most
…with so many appointments that [my primary care potent and widespread effects on patients’ communica-
physician, and my rheumatologist] want me to go to tion with doctors, according to these participants. Strat-
… I can’t afford to pay all those co-pays. And to go to egies that generated improvements in participants’
the rheumatologist or specialty doctors, which is what ability to communicate with doctors included preparing
I mostly have to see lately [coughs] it’s like fifty dollars for appointments (e.g., making a list of questions and
every time I go to the doctor and I can’t afford that, medications) and boosting patient participation during
so it’s kinda tricky. doctor’s visits (e.g., increased assertiveness, taking notes,
more openness about perceived ineffectiveness of medica-
(Age 45, High School, Severe Lupus, Depressed) tions or fear of comorbidities). However, improvements
were not experienced uniformly across the sample:
women with mild/moderate disease severity and non-
Post-CDSMP The CDSMP did not seem to help women depressed women were more likely to report improve-
improve their health insurance or better navigate sys- ments in communication with doctors after the CDSMP.
tems for uninsured people. Insurance remained a major These qualitative findings are consistent with prior quanti-
barrier to accessing health services: tative data obtained among 698 African American patients
with SLE from our GOAL cohort, indicating that subopti-
I wish I learned more about, ah, like say for the mal interactions with providers may be explained by the
people who don’t have the insurance right now, like mental and physical symptoms of the patient [45]. Thus,
say I don’t even have the disability and can’t afford to our data suggest that in addition to standard of care treat-
pay for... like one of these medicines is very high, two ment and self-management education, African American
of them. [I wish] There was more ways that I could patients with more severe SLE activity and depression
find out how to get doctors I need to see for myself … might need provider-based interventions focused on im-
they said you can go to all the free clinics but the free proving physician-patient communication.
clinics don’t have specialists at their clinic … .they’re Our analysis also revealed that older women were
not a rheumatologist more likely to report post-CDSMP improvements in
medication side effect management. Because older age
(Age 64, High School graduate, Severe Lupus, Not can potentially increase polypharmacy associated with
Depressed) greater cumulative organ damage [46] and contribute to
cognitive impairments to perform instrumental self-
Only three of 23 participants (two belonging to the management activities in the SLE population [47], our
uninsured group at baseline and one in the insured findings support the relevance of the CDSMP program
Twumasi et al. Arthritis Research & Therapy (2019) 21:155 Page 9 of 11

to improve drug side effect management among those at including through memos and team discussions. To test
higher risk. the team’s research questions, guides, and interpreta-
Women who attended 4 or more CDSMP classes were tions, we partnered with African American women who
more likely to report improvements in both communica- were living with SLE as we developed the interview
tion with physicians and management of medication’s guide, and to review and challenge emerging findings.
side effects, supporting the need of weekly CDSMP Our study has several strengths. Each interview was
workshop encounters with peers for successful self- audio-recorded and transcribed verbatim to strengthen
management skill-building [24, 25]. descriptive validity (i.e., the data’s accuracy and com-
Although the CDSMP was not designed to improve in- pleteness). To improve interpretive validity (i.e., hewing
surance coverage, our study participants repeatedly to participant perspectives), we conducted member
noted that insurance was an important barrier to effect- checks with two African American women living with
ive utilization of health services. A major cause of loss of SLE, both of whom corroborated our findings. The
insurance is unemployment, which we previously de- study’s longitudinal design further enhanced its inter-
scribed to occur in nearly 50% of our SLE cohort after pretive validity. We included negative cases to enhance
13 years since diagnosis [48]. We also reported that Afri- theoretical validity (i.e., considerations of plausible alter-
can American patients and those with more severe dis- nate explanations). The study’s longitudinal design, with
ease were more likely to be unemployed [48]. Thus, a high retention rate (95.8%), supported valid compari-
those vulnerable SLE groups are often at a much greater sons of self-management behaviors reported by partici-
risk of entering in the vicious circle of poverty, lack of pants, pre- and post-intervention.
healthcare access, and poor outcomes. Our study partici- In conclusion, findings from this novel longitudinal,
pants lived in Georgia, a state that has resisted Medicaid qualitative study suggest that African American women
expansion, and so under- and lack of insurance remain living with SLE, a population vulnerable to high rates of
critical barriers to healthcare services in the state. How- SLE morbidity and mortality, perceived benefits from
ever, there is a lack of data on the proportion of high- CDSMP participation on healthcare engagement. Specif-
risk individuals with SLE who receive disability benefits ically, our analysis suggests that CDSMP engagement
and Medicaid coverage in Georgia. improved their communication with doctors and their
medication side effect management. Should additional
Strengths and limitations quantitative studies reach similar conclusions, African
These findings should be considered in light of the American women living with SLE should be encouraged
study’s limitations. As with qualitative studies generally, to participate in CDSMP workshops to enhance health
our findings may not be generalizable to the broader service use behaviors, behaviors that have been
population of African American women diagnosed with proven to improve health outcomes among SLE pa-
SLE. However, purposive sampling enabled us to explore tients. Because the CDSMP is free and widely dissem-
variations in themes by participant age, depression, SLE inated across the USA through community centers, it
severity, and CDSMP participation. Most GOAL partici- may be a suitable and accessible option for African
pants have graduated from high school, however, and so American women with SLE, even though it is not
we could not achieve our aim of exploring differences in specifically tailored to SLE.
CDSMP experiences by high school graduation status;
instead, the sample varies by college graduation, which Additional file
may be less salient to health care use behaviors and
CDSMP experiences than high school graduation status. Additional file 1: WELL Qualitative Study Interview Guide: Wave 1 and
Further, the study was not able to explore whether or Wave 2. (DOCX 41 kb)
how provider characteristics and responsiveness might
shape doctor-patient relationships. Future studies of Abbreviations
CDSMP: Chronic Disease Self-Management Program; GLR: Georgia Lupus
doctor/patient dyads can explore the effects of physician Registry; GOAL: Georgians Organized Against Lupus; SLE: Systemic lupus
characteristics on doctor-patient communication. The erythematosus; WELL: Women Living Well with Lupus
relationship between the interviewer and the participant
can shape multiple dimensions of the research process, Acknowledgements
The authors would like to acknowledge members of the WELL cohort who
including interview guides, participant disclosure, and participated in this study for sharing their personal experiences with lupus.
analyses. In this case, both interviewers were women, The authors also thank Gaobin Bao for his assistance with statistical analysis
neither of whom was living with SLE; one interviewer for the manuscript.
was Asian American and the other was African Ameri-
Authors’ contributions
can. The team engaged in multiple forms of critical CD and HC conceived the study concept and design. AT, AS, and HC
reflexivity about team relationships with participants, collected, analyzed, and interpreted the data regarding African American
Twumasi et al. Arthritis Research & Therapy (2019) 21:155 Page 10 of 11

women SLE patients’ perceptions of the CDSMP. AT, AS, CDT, CD, and HC 2018;70(suppl 10). Abstract 265. https://acrabstracts.org/abstract/disparities-in-
were major contributors in writing the manuscript. All authors read and utilization-and-direct-costs-of-hospitalizations-and-emergency-room-visits-in-
approved the final manuscript. sle-the-georgia-lupus-registry/. Accessed 21 June 2019.
11. Demas KL, Costenbader KH. Disparities in lupus care and outcomes. Curr
Funding Opin Rheumatol. 2009;21(2):102–9.
Research reported in this publication was supported by the National Institute 12. Lim SS, Helmick CG, Bao G, Gordon C, Hootman JM, Drenkard C. The
On Minority Health And Health Disparities, housed within the National Distribution of Insurance in a Population-Based Cohort of SLE: Georgians
Institutes of Health, under Award Number R01MD010455 (PI: Drenkard). The Organized Against Lupus Cohort, 2012-2016 [abstract]. Arthritis Rheumatol.
content is solely the responsibility of the authors and does not necessarily 2018;70(suppl 10). Abstract 1183. https://acrabstracts.org/abstract/the-
represent the official views of the National Institutes of Health. distribution-of-insurance-in-a-population-based-cohort-of-sle-georgians-
organized-against-lupus-cohort-2012-2016/. Accessed 21 June 2019.
Availability of data and materials 13. Yazdany J, Feldman CH, Liu J, Ward MM, Fischer MA, Costenbader KH.
The datasets generated and analyzed during the current study are not Quality of care for incident lupus nephritis among Medicaid beneficiaries in
publicly available because they are identifiable. Methods to de-identify quali- the United States. Arthritis Care Res (Hoboken). 2014;66(4):617–24.
tative data are still under development; stories disclosed during interviews, 14. Yelin E, Yazdany J, Trupin L. Relationship between process of care and a
for example, could easily reveal an individual’s identity when paired with subsequent increase in damage in systemic lupus erythematosus. Arthritis
state of residence, SLE status, race/ethnicity, gender, and age. Data are avail- Care Res (Hoboken). 2017;69(6):927–32.
able from the corresponding author on reasonable request. 15. Chambers SA, Rahman A, Isenberg DA. Treatment adherence and clinical
outcome in systemic lupus erythematosus. Rheumatology (Oxford). 2007;
Ethics approval and consent to participate 46(6):895–8.
All study participants signed informed consent. The study was approved by 16. Brady TJ, Jernick SL, Hootman JM, Sniezek JE. Public health interventions for
the Emory Institutional Review Board and Grady Health System Research arthritis: expanding the toolbox of evidence-based interventions. J Women’s
Oversight Committee. Health. 2009;18(12):1905–17.
17. Sapag JC, Lange I, Campos S, Piette JD. Innovative care and self-care
Consent for publication strategies for people with chronic diseases in Latin America. Rev Panam
Not applicable. Salud Publica. 2010;27(1):1–9.
18. van Eijk-Hustings Y, van Tubergen A, Boström C, Braychenko E, Buss B,
Competing interests Felix J, et al. EULAR recommendations for the role of the nurse in the
The authors declare that they have no competing interests. management of chronic inflammatory arthritis. Ann Rheum Dis. 2012;
71(1):13–9.
Author details 19. Department of Health. Self care–a real choice. Self care support–a practical
1
Department of Behavioral Sciences and Health Education, Emory University option. London: Department of Health; 2005.
Rollins School of Public Health, Atlanta, GA, USA. 2Department of Medicine, 20. Clark NM, Janz NK, Dodge JA, Sharpe PA. Self-regulation of health behavior:
Emory University School of Medicine, Atlanta, GA, USA. the “take PRIDE” program. Health Educ Q. 1992;19(3):341–54.
21. Glasgow RE, Davis CL, Funnell MM, Beck A. Implementing practical
Received: 26 December 2018 Accepted: 14 June 2019 interventions to support chronic illness self-management. Jt Comm J Qual
Saf. 2003;29(11):563–74.
22. Lorig KR, Mazonson PD, Holman HR. Evidence suggesting that health
References education for self-management in patients with chronic arthritis has
1. Petri M. Clinical features of systemic lupus erythematosus. Curr Opin sustained health benefits while reducing health care costs. Arthritis Rheum.
Rheumatol. 1995;7(5):395–401. 1993;36(4):439–46.
2. Urowitz MB, Gladman DD, Abu-Shakra M, Farewell VT. Mortality studies in 23. Lorig KR, Ritter P, Stewart AL, Sobel DS, Brown BW Jr, Bandura A, et al.
systemic lupus erythematosus. Results from a single center. III. Improved Chronic disease self-management program: 2-year health status and health
survival over 24 years. J Rheumatol. 1997;24(6):1061–5. care utilization outcomes. Med Care. 2001;39(11):1217–23.
3. Alarcon GS. Lessons from LUMINA: a multiethnic US cohort. Lupus. 2008; 24. Lorig KR, Sobel DS, Ritter PL, Laurent D, Hobbs M. Effect of a self-
17(11):971–6. management program on patients with chronic disease. Eff Clin Pract. 2001;
4. Lim SS, Bayakly AR, Helmick CG, Gordon C, Easley KA, Drenkard C. The 4(6):256–62.
incidence and prevalence of systemic lupus erythematosus, 2002-2004: The 25. Lorig K, Holman H, Sobel D, Laurent D, González V, Minor M. Living a
Georgia Lupus Registry. Arthritis Rheumatol. 2014;66(2):357–68. healthy life with chronic conditions. Self-management of heartd disease,
5. Somers EC, Marder W, Cagnoli P, Lewis EE, DeGuire P, Gordon C, et al. arthritis, diabetes, asthma, bronchitis, emphysema and others. 4th Edition.
Population-based incidence and prevalence of systemic lupus Boulder: Bull Publishing Company; 2012.
erythematosus: the Michigan lupus epidemiology and surveillance program. 26. Lorig K. Self-management education: more than a nice extra. Med Care.
Arthritis Rheumatol. 2014;66(2):369–78. 2003;41(6):699–701.
6. Izmirly PM, Wan I, Sahl S, Buyon JP, Belmont HM, Salmon JE, et al. The 27. Ahn S, Basu R, Smith ML, Jiang L, Lorig K, Whitelaw N, et al. The impact of
incidence and prevalence of systemic lupus erythematosus in New York chronic disease self-management programs: healthcare savings through a
County (Manhattan), New York: the Manhattan lupus surveillance program. community-based intervention. BMC Public Health. 2013;13:1141.
Arthritis Rheumatol. 2017;69(10):2006–17. 28. Ory MG, Ahn S, Jiang L, Smith ML, Ritter PL, Whitelaw N, et al. Successes of
7. Dall'Era M, Cisternas MG, Snipes K, Herrinton LJ, Gordon C, Helmick CG. The a national study of the chronic disease self-management program: meeting
incidence and prevalence of systemic lupus erythematosus in San Francisco the triple aim of health care reform. Med Care. 2013;51(11):992–8.
County, California: the California Lupus Surveillance Project. Arthritis 29. Ory MG, Ahn S, Jiang L, Lorig K, Ritter P, Laurent DD, et al. National study of
Rheumatol. 2017;69(10):1996–2005. chronic disease self-management: six-month outcome findings. J Aging
8. Bernatsky S, Boivin JF, Joseph L, Manzi S, Ginzler E, Urowitz M, et al. Race/ Health. 2013;25(7):1258–74.
ethnicity and cancer occurrence in systemic lupus erythematosus. Arthritis 30. Garcia Popa-Lisseanu MG, Greisinger A, Richardson M, O'Malley KJ, Janssen
Care Res. 2005;53(5):781–4. NM, Marcus DM, et al. Determinants of treatment adherence in ethnically
9. Webb R, Kelly JA, Somers EC, Hughes T, Kaufman KM, Sanchez E, et al. Early diverse, economically disadvantaged patients with rheumatic disease. J
disease onset is predicted by a higher genetic risk for lupus and is Rheumatol. 2005;32(5):913–9.
associated with a more severe phenotype in lupus patients. Ann Rheum 31. Braden CJ, McGlone K, Pennington F. Specific psychosocial and behavioral
Dis. 2011;70(1):151–6. outcomes from the systemic lupus erythematosus self-help course. Health
10. Lim SS, Drenkard C, Bao G, Mukherjee J, Connolly SE, Napoli A, Paul D, Alemao Educ Q. 1993;20(1):29–41.
E. Disparities in utilization and direct costs of hospitalizations and emergency 32. Bandura A. Self-efficacy: toward a unifying theory of behavioral change.
room visits in SLE: The Georgia Lupus Registry [abstract]. Arthritis Rheumatol. Psychol Rev. 1977;84(2):191.
Twumasi et al. Arthritis Research & Therapy (2019) 21:155 Page 11 of 11

33. Drenkard C, Dunlop-Thomas C, Easley K, Bao G, Brady T, Lim SS. Benefits of

a self-management program in low-income African-American women with
systemic lupus erythematosus: results of a pilot test. Lupus. 2012;21(14):
1586–93.
34. Hennink M, Hutter I, Bailey A. Qualitative research methods. SAGE, London,
California, New Dehli, Singapure; 2010.
35. Drenkard C, Rask KJ, Easley KA, Bao G, Lim SS. Primary preventive services in
patients with systemic lupus erythematosus: study from a population-based
sample in southeast U.S. Semin Arthritis Rheum. 2013;43(2):209–16.
36. Hale ED, Treharne GJ, Lyons AC, Norton Y, Mole S, Mitton DL, et al. “Joining
the dots” for patients with systemic lupus erythematosus: personal
perspectives of health care from a qualitative study. Ann Rheum Dis. 2006;
65(5):585–9.
37. Gallop K, Nixon A, Swinburn P, Sterling KL, Naegeli AN, Silk ME.
Development of a conceptual model of health-related quality of life for
systemic lupus erythematosus from the patient’s perspective. Lupus. 2012;
21(9):934–43.
38. Harrison M, Reeves D, Harkness E, Valderas J, Kennedy A, Rogers A, et al. A
secondary analysis of the moderating effects of depression and
multimorbidity on the effectiveness of a chronic disease self-management
programme. Patient Educ Couns. 2012;87(1):67–73.
39. Reeves D, Kennedy A, Fullwood C, Bower P, Gardner C, Gately C, et al.
Predicting who will benefit from an expert patients programme self-
management course. Br J Gen Pract. 2008;58(548):198–203.
40. Ward MM. Education level and mortality in systemic lupus erythematosus
(SLE): evidence of underascertainment of deaths due to SLE in ethnic
minorities with low education levels. Arthritis Rheum. 2004;51(4):616–24.
41. Rojas-Serrano J, Cardiel MH. Lupus patients in an emergency unit. Causes of
consultation, hospitalization and outcome. A cohort study. Lupus. 2000;9(8):
601–6.
42. Petri M, Genovese M. Incidence of and risk factors for hospitalizations in
systemic lupus erythematosus: a prospective study of the Hopkins lupus
cohort. J Rheumatol. 1992;19(10):1559–65.
43. Braun V, Clarke V. Using thematic analysis in psychology. Qual Res Psychol.
2006;3(2):77–101.
44. VERBI Software. MAXQDA version 12.3.2 [computer software]. Berlin: VERBI
Software; 2017.
45. Drenkard C, G GB, Lewis TT, Pobiner B, J JP, Lim SS. Physician-patient
interactions in African American patients with systemic lupus
erythematosus: demographic characteristics and relationship with disease
activity and depression. Semin Arthritis Rheum. 2019;48(4):669-77.
46. Alarcon GS, McGwin G Jr, Bartolucci AA, Roseman J, Lisse J, Fessler BJ, et al.
Systemic lupus erythematosus in three ethnic groups. IX. Differences in
damage accrual. Arthritis Rheum. 2001;44(12):2797–806.
47. Plantinga L, Tift BD, Dunlop-Thomas C, Sam Lim S, Barrett Bowling C,
Drenkard C. Geriatric assessment of physical and cognitive functioning in a
diverse cohort of systemic lupus erythematosus patients: a pilot study.
Arthritis Care Res (Hoboken). 2018;70(10):1469–77.
48. Drenkard C, Bao G, Dennis G, Kan HJ, Jhingran PM, Molta CT, et al. Burden
of systemic lupus erythematosus on employment and work productivity:
data from a large cohort in the southeastern United States. Arthritis Care
Res. 2014;66(6):878–87.

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.