A Personalized Decision Aid to Help Women with Lupus Nephritis … · 2020. 5. 14. · 3 Abstract Background/Objective: Systemic lupus erythematosus (SLE) is a rare and sometimes

1

A Personalized Decision Aid to Help Women with Lupus Nephritis from Racially and Ethnically Diverse Backgrounds Make Decisions about Taking Immune-Blocking Medicines

Jasvinder Singh, MD,1 Jinoos Yazdany, MD, MPH2 Winn Chatham, MD1 Tara Rizvi, MD3 Liana Fraenkel, MD, MPH4 Graciela Alarcon, 1Kenneth Saag, MD 1 Robert Kimberly, MD1 Maria Suarez-Almazor, 5 Amye Leong, MBA,6 Leslie Hanrahan,7 Sandra Raymond,7 Elyse Reyes, Charity Morgan, PhD1 Nipam Shah, 1 Candace Green, 1 Alexa Meara, MD8 Rick Street, PhD 9 Laura Marrow,10 Brennda Caro,1 Jennifer Barton, MD,11 Jeffrey Sloan, PhD,12

1University of Alabama School of Medicine, Birmingham, AL 2University of California San Francisco Medical Center at Parnassus, San Francisco, CA 3Baylor College of Medicine, Houston, TX 4Yale University School of Medicine, Boston, MA 5University of Texas MD Anderson Cancer Center, Houston, TX 6Healthy Motivations Inc. Santa Barbara, CA 7Lupus Foundation of America, Washington, DC 8Ohio State University, Columbus, OH 9Texas A&M University System, College Station, TX 10Arthritis Foundation, Atlanta, GA 11Oregon State University, Portland, OR 12Mayo Clinic Minnesota, Rochester, MN

Original Project Title: Individualized Patient Decision Making for Treatment Choices among Minorities with Lupus PCORI ID: CE-1304-6631 HSRProj ID: 20143522 ClinicalTrials.gov ID: NCT02319525 _______________________________

To cite this document, please use: Singh J, Yazdany J, Chatham W, et al. (2019). A Personalized Decision Aid to Help Women with Lupus Nephritis from Racially and Ethnically Diverse Backgrounds Make Decisions about Taking Immune-Blocking Medicines. Washington, DC: Patient-Centered Outcomes Research Institute (PCORI). https://doi.org/10.25302/10.2019.CE.13046631

2

Table of Contents

ABSTRACT ............................................................................................................................. 3 BACKGROUND ....................................................................................................................... 5 PARTICIPATION OF PATIENTS AND OTHER STAKEHOLDERS IN RESEARCH DESIGN, CONDUCT, AND DISSEMINATION OF FINDINGS ....................................................................................... 9

Study Intervention Development ............................................................................................................................ 12 Study Design, Population, and Setting: Randomized Trial ...................................................................................... 26 Randomized Trial Study Intervention ...................................................................................................................... 27 Baseline and Follow-up Study Visits ........................................................................................................................ 27 Outcome Measures ................................................................................................................................................. 28 Statistical Analyses .................................................................................................................................................. 31 Conduct of the Study ............................................................................................................................................... 32

RESULTS .............................................................................................................................. 33 1. Primary Outcome ................................................................................................................................................ 40 2. Primary Outcome ................................................................................................................................................ 42 3. Secondary Outcome ............................................................................................................................................ 43 4. Secondary Outcome ............................................................................................................................................ 44 5. Secondary Outcome ............................................................................................................................................ 44

DISCUSSION ........................................................................................................................ 51 Context for Study Result.......................................................................................................................................... 51 Generalizability of the Findings ............................................................................................................................... 55 Implementation of Study Results ............................................................................................................................ 55 Subpopulation Considerations ................................................................................................................................ 57 Study Strengths and Limitations.............................................................................................................................. 57 Future Research ...................................................................................................................................................... 59

CONCLUSION ....................................................................................................................... 61 ACKNOWLEDGMENTS .......................................................................................................... 62 REFERENCES ........................................................................................................................ 63

3

Abstract

Background/Objective: Systemic lupus erythematosus (SLE) is a rare and sometimes fatal disease. Lupus nephritis inflammation of the kidney is a devastating complication of SLE and is often more common in racial/ethnic minorities. Immunosuppressive drugs can effectively treat lupus nephritis, but patients may be reluctant to use them due to concerns about side effects and lack of understanding of their potential benefits. We assessed whether a web-based, individualized, culturally tailored, computerized patient decision aid can improve decision making regarding using immunosuppressive drugs in women with lupus nephritis.

Methods: We developed a patient decision aid for immunosuppressive medication decision making based on formative work with 52 lupus nephritis patients (predominantly racial/ethnic minorities with low socioeconomic status) and systematic review, meta-analyses, and network meta-analyses. In a 6-month randomized controlled trial at 4 US centers, we recruited adult women, largely racial/ethnic minorities with low socioeconomic status, who were making decisions about starting or maintaining treatment of lupus nephritis flares, or who had a history of lupus nephritis flares and were at risk of future flares. Patients were randomized during a clinic visit to receive a patient decision aid on a tablet computer or a standard American College of Rheumatology (ACR) pamphlet that provided information about lupus and its treatment, including the use of immunosuppressive drugs. The study was conducted during clinic visits and outcome assessments occurred immediately after the intervention was administered. Coprimary outcomes were a change in decisional conflict assessed with a low literacy–version Decisional Conflict Scale (0-100; the higher the score, the more conflict was present) and the proportion with an informed choice regarding immunosuppressive drugs (concordance of the patients’ values and their choice for or against immunosuppressives based on an adequate knowledge of the drugs). Secondary outcomes included (1) the concordance between a patient’s desired and actual role in immunosuppressive drugs decision making using the control preference scale, (2) the patient perception of patient–physician communication and care processes using the Interpersonal Process of Care–Short Form (IPC-SF), and (3) the assessment of patient–physician communication by assessing the audiotaped physician–patient conversation about therapy options.

Results: Of the 301 women with lupus nephritis randomized, 3 patients withdrew consent and 298 received an individualized decision aid (n = 151) or the ACR pamphlet (n = 147, control arm). The mean age was 37 years, 35% had an annual income < $20 000, 36% had a high school education or less, the average health literacy score on the Short Assessment of Health Literacy was 16.8 (a score of 0-14 denotes low health literacy), 47% were African American, and 26% were Hispanic. Compared with those who received the pamphlet, patients who received the decision aid had a significantly larger reduction in decisional conflict (21.8 [SD, 30.9] versus

4

12.7 [SD, 24.4]; p = 0.005). The group receiving the decision aid made informed choices regarding immunosuppressive drugs more frequently (41% versus 31%), although this did not meet statistical significance (p = 0.08). Sensitivity analysis that used an alternate definition of informed choice (positive versus negative values rather than the median score for values) showed that significantly more women in the decision aid group made an informed choice compared with those in the pamphlet group (50% versus 35%; p = 0.006). We noted no statistically significant differences in the secondary outcomes of concordance in the desired versus the actual role in decision making (94% versus 85%; p = 0.25) or the IPC-SF scores (83.6 [SD, 7.7] versus 83.1 [SD, 7.3]; p = 0.50). Using an audiotaped patient–physician conversation, the patient-centered communication by doctor showed a statistical trend toward significance in the decision aid versus the pamphlet group (5.1 versus 3.7; p = 0.06).

Conclusions: An electronic, individualized, culturally appropriate patient decision aid was effective in reducing the decisional conflict regarding choosing immunosuppressive drugs in an ethnically and socioeconomically diverse sample of women with lupus nephritis. Future studies should investigate whether this decision aid can be further enhanced to improve its efficacy, modified for other manifestations of lupus, or provided on a mobile platform, so that patients have even easier access to it. The Patient-Centered Outcomes Research Institute (PCORI) lupus nephritis decision aid will be available in the public domain.

5

Background

Systemic lupus erythematosus (SLE), or lupus, is a rare chronic disease that primarily

affects young women, particularly during peak reproductive years.1,2 In the United States, 161

000 individuals have SLE, making it a rare disease, and if not treated appropriately, SLE has

devastating consequences.3 Approximately 35% of SLE patients present initially with nephritis

and 50% to 60% develop nephritis during the first 10 years.4,5 Among racial/ethnic minorities

lupus nephritis accounts for 2% of all end-stage renal disease in the United States.6 Lupus

nephritis is significantly more prevalent and has worse outcomes (e.g., > 3 times higher

mortality) in African Americans and Hispanics than in whites.7,84,4,5,9-15 Thus, racial disparities in

outcomes and high mortality make lupus an optimal disease for patient-centered outcomes

research. While the 2012 American College of Rheumatology (ACR) treatment guidelines for

lupus nephritis incorporated comparative effectiveness research (CER) data,16 a detailed

analysis of comparative toxicity was not completed. Strong evidence exists that the use of

immunosuppressive medications improves lupus outcomes when combined with

glucocorticoids,16 and may reduce the cumulative glucocorticoid dose and associated side

effects.17-19 However, immunosuppressive drugs used to treat lupus nephritis (mycophenolate

mofetil, cyclophosphamide, azathioprine, etc.)16 differ from each other significantly in their

toxicity (e.g., cancer risk, infections), effects on fertility, safety during pregnancy, administration

route, frequency of dosing, and cost. For example, mycophenolate is contraindicated for use in

pregnancy, whereas clinicians consider azathioprine (a less expensive medication) a relatively

safe option.20 Given the differences in toxicity and cost across treatment options, we believed

that it was critical to generate both patient perspectives21-23 (Aim 1) and CER data24-26 (Aim 2)

related to immunosuppressive drugs to enable informed decision making. In the current study,

patients were asked to weigh risks versus benefits differently depending on their values,

preferences, and knowledge of and aversion to risk, to enable development of a patient-

centered decision aid.

Patients are often faced with difficult decisions about how to treat newly diagnosed and

chronic lupus nephritis. Medication choice(s) depend not only on efficacy and drug-specific risk

6

profiles, but also on cost. In addition, efficacy may differ by race and ethnicity. For example, a

subgroup analysis of an open-label randomized controlled trial (RCT)—the Aspreva Lupus

Management Study (ALMS)—found that, compared with whites and Asians, patients of other

races and ethnicities with lupus nephritis responded less favorably to cyclophosphamide as

induction therapy27; and while not differing by race, the response to mycophenolate was

superior to azathioprine as maintenance therapy for lupus nephritis.28 If the existing CER data

can be expanded (Aim 1) and simplified using a decision aid, this may help patients with lupus

nephritis through the complicated decision-making process regarding immunosuppressive

drugs (Aim 3). It should be noted that the example study above, ALMS, was limited by its being

an open-label design and inclusion of relatively few African Americans.27

Many minority patients do not receive quality health care, due in part to lower health

literacy and numeracy,29 poor physician–patient communication,30-33 high medication

nonadherence,34 low socioeconomic status,35,36 more barriers to health care access,37-39 and

more risk aversion to therapies.40-45 Since health literacy is the ability to understand, engage in,

and actively apply health information to improve health,46 poor health literacy can contribute

to poor health outcomes. In the United States, 41% of Hispanics, 24% of African Americans, and

9% of whites have inadequate health literacy skills, highlighting the disproportionate impact of

low health literacy on racial minorities.46 Numeracy is the ability to understand and use

numbers in daily life and, when inadequate, is associated with poor health outcomes.47,48

Compared with whites, studies have shown that African Americans have lower numeracy,

which may explain poor health outcomes in diseases such as diabetes,49 and poor management

of medications, as demonstrated in a study using a simulated HIV medication regimen.50

Previous studies have shown poor medication adherence is common in patients with lupus

nephritis, with racial minorities indicating less willingness to receive treatment for worsening

lupus.51-53 In qualitative studies, we described both facilitators22 and barriers23 to medication

decision making by patients with lupus nephritis, and that a decision aid could overcome these

challenges and present information tailored to health literacy and numeracy. A decision aid is

one potential solution for patients with low literacy and numeracy but is currently lacking for

lupus. A Cochrane systematic review54 that included studies in patients with low literacy

7

showed that a decision aid helped patients comprehend accurate expectations of possible

benefits and harms, make choices that are more consistent with their informed values, and

participate more in decision making.54 Patients who were involved in making medical decisions

had better outcomes55,56 and were more satisfied than patients who were not involved.57-59

Many treatment decisions (for example, mycophenolate mofetil versus cyclophosphamide for

induction) in lupus nephritis have no single “best” choice and are preference sensitive due to

insufficient evidence about outcomes and/or the need to trade off known benefits and harms.

Similarly, there are preference-sensitive decisions for maintenance therapy for lupus nephritis,

including the choice between mycophenolate mofetil versus cyclophosphamide (following

azathioprine failure), calcineurin inhibitors such as cyclosporine/tacrolimus versus

cyclophosphamide (following azathioprine and mycophenolate mofetil failure), or

cyclophosphamide versus azathioprine (following mycophenolate mofetil failure). For others,

the benefits of a drug far outweigh its risks. For example, the use of glucocorticoids combined

with immunosuppressives is superior to glucocorticoids alone for induction—i.e., it is the

dominant choice (not preference sensitive), yet the rate of immunosuppressive use by patients

with lupus nephritis is low in the United States.60 A patient decision aid could facilitate

preference-sensitive decisions, such as the choice between immunosuppressives for induction

or maintenance therapy for lupus nephritis. Decision aids have been used successfully for

diabetes and osteoporosis,61,62 rheumatoid arthritis, and hepatitis C.63-66 Decision aids have

succeeded in improving outcomes in other inflammatory arthritis conditions67 when developed

for the target population.68 To our knowledge, no decision aid exists to assist patients with

lupus nephritis in treatment decision making. As part of Aims 3 and 4, we tested the hypothesis

that a decision aid that can help patients overcome literacy/numeracy challenges can improve

patient decision making and can lead to more informed choices by patients.

We conducted network meta-analyses (NMAs)69-71 to assess the efficacy and

harms/toxicity of lupus nephritis treatments to address the PCOR question “What are my

options and what are the potential benefits and harms of those options?” Second, we aimed to

develop an individualized decision aid for African American and Hispanic women with lupus

nephritis, for whom the disease is often more severe and outcomes are worse, by focusing on

8

culture-specific barriers. Our central research question was whether the use of an

individualized, culturally tailored, patient-centered decision aid would reduce the conflict in

patient decision making and lead to more informed patient choice regarding

immunosuppressives for the treatment of lupus nephritis. Our study aims were the following:

Aim 1: To perform an evidence synthesis by systematic review and NMA. To assess

comparative effectiveness of various immunosuppressives compared with each other and with

glucocorticoids, corresponding to main induction and maintenance treatment decision points

(published24-26), using rigorous systematic review and NMA methods based on Agency for

Healthcare Research and Quality (AHRQ) recommendations72 and the Cochrane handbook,73

and building on the systematic review performed for the 2012 ACR lupus nephritis treatment

recommendations.16 After assessing heterogeneity across trials74 in patient characteristics, trial

methodologies, and treatment protocols, we conducted a Bayesian NMA75-77 for prespecified

outcomes.

Aim 2: To elicit patients’ perceptions of barriers to effective decision making for lupus

nephritis treatments, and their concerns regarding the risks and benefits of various

immunosuppressives compared with each other and with glucocorticoids through the nominal

group technique (NGT; results published21-23).

Aim 3: To develop and pilot test an interactive computerized decision aid based on CER

data and formative work with patients—largely racial/ethnic minorities of low socioeconomic

status—using an iterative process (a detailed protocol has been published78).

Aim 4: To conduct a multicenter, parallel arm, randomized trial comparing the usual

care for decision making against an individualized, culturally tailored patient decision aid in

women— largely racial/ethnic minorities with low socioeconomic status—making lupus

nephritis treatment decisions regarding immunosuppressives.

9

Participation of Patients and Other Stakeholders in Research Design, Conduct, and Dissemination of Findings

We identified our key patients, patient advocacy organization, physicians, researcher,

and institutional stakeholders from various racial and socioeconomic backgrounds to assist with

the development and execution of this project. Our patient stakeholders included Ms. A

(anonymized as per instructions; CEO of Company A; master’s degree in business management;

trained in psychometrics and outcomes); Ms. B (anonymized; Master of Public Administration;

past director of patient programs and community programs at a patient advocacy organization;

bilingual). Our patient advocacy organization leaders included Sandra Raymond (president and

CEO of research, Lupus Foundation of America [LFA]), Leslie Hanrahan (vice president of

education and research, LFA), and Laura Marrow (director, partnerships liaison at the Arthritis

Foundation [AF]). Physician and researcher stakeholders included experts in lupus treatment

(Drs. Chatham, Yazdany, Alarcón) and those with expertise in decision aid development and/or

performing CER including NMA (Drs. Fraenkel, Winthrop, Wells, Suarez-Almazor, Barton,

Grossman, Saag, Singh, and Street). Institutions and organizations that served as key

stakeholders and assisted with the study development and conduct included the University of

Alabama (UAB) Department of Communication, University of California at San Francisco (UCSF),

AHRQ-supported UAB CERTs, and The Eisenberg Center.

We recruited our 2 patient stakeholders (Ms. A, Ms. B) based on their expertise, disease

experience, and experience serving as partners in patient-centered research. Ms. A and Ms. B

represent a key group of people for whom the study results are particularly relevant. Ms. A and

Ms. B played key roles in developing our study questions before submission, and throughout

the funding period. Patient and other key stakeholder engagement had a significant impact on

all aspects of the study quality and contributed heavily to the development of the decision aid.

We held stakeholder and investigator teleconference meetings monthly on Thursdays to discuss

key aspects of the project, from project inception through its conclusion. The stakeholder

committee was charged with finalizing the risks and benefits that were incorporated into the

decision aid, and all patient and clinician stakeholders engaged in a variety of activities,

including formulation of the study design, the NMA, pilot testing of the intervention (i.e.,

10

individualized decision aids in English and Spanish), and patient recruitment and retention in

our trial. Previous experience of key stakeholders helped us meet new challenges during the

study conduct. The monthly meetings allowed interactions of investigators with other key

stakeholders. For example, stakeholders and investigators collaboratively reviewed and

provided feedback about plans for focus group techniques and discussed the results of patient

focus groups. We did not face any significant logistic or budgetary challenges in engaging

patients and stakeholders. Specific examples of key stakeholder involvement are as follows:

• Both researchers and patient partners were involved in creating the manuscript related

to facilitators of medication adherence and cognitive mapping (published21).

Researchers presented a preliminary document for review at one of the recent

stakeholder/investigator calls. All partners provided their feedback on the study design

as a group and later independently grouped concepts/statements as part of the

cognitive mapping. This iterative process was highly effective, allowing

researchers/clinicians and patient representatives to hear the differing opinions of their

peers. The resulting manuscript of cognitive mapping of medication decision-making

facilitators addressed the different views of all 3 groups.21

• Brennda Caro led and Drs. Suarez-Almazor, Alarcón, and Barton helped with the Spanish

translations that led to the development of the Spanish version of the decision aid. Dr.

Fraenkel helped the researchersdevelop the content of our lupus nephritis decision aid

for minorities based on the NGT results, using the rheumatoid arthritis decision aid as a

template. We eliminated from the decision aid several side effects important only to

physicians, but not patients (eg, low white cell counts and bone density values related to

glucocorticoids).

• Patient research partners (Ms. A and Ms. B) helped us design the study and participated

heavily in the development of the decision aid. Both participated in all of the activities of

our study by regular teleconferences and email exchanges.

Additionally, we were able to convene a group of key stakeholders and investigators at

the 2016 annual ACR meeting, at which panel discussions were held about further ideas on

11

patient-centered research for lupus. The theme of the discussion was how to improve lupus

care. Discussion involved key Patient-Centered Outcomes Research Institute (PCORI)

investigators, stakeholders, and attending rheumatologists from sites such as Ohio State

University (OSU), and stakeholders shared ideas on initiatives started by their respective

organizations for better lupus care. The panel concluded that a multicomponent intervention

for the management of lupus was important from the patient perspective. Investigators have

actively engaged and enrolled their patients in this research project. It has helped study team

investigators understand how patient-centered research helps and empowers patients to make

informed decisions about their care for lupus nephritis. Investigators were motivated to

become a part of future outcomes research involving implementation of multifaceted

interventions for management of this condition. We discussed the methods to disseminate this

decision aid with the help of our stakeholder partners, LFA and AF, using one of the PCORI

dissemination grants.

Overall, patients enrolled in this trial liked the decision aid tool and found it very

informative. Most patients did not mind spending the extra time during their visits to

participate in the study, noting it helped them understand the treatment options as well as the

risks and benefits of those options. Patients shared positive experiences about the study

participation, the research study process and conduct, and the technology used (iPad self-

administered).

12

Methods

Study Intervention Development

The study intervention in this 2-arm trial was a decision aid in English or Spanish

targeting patient knowledge and opinions about immunosuppressive drugs. The decision aid

content was based on information generated through qualitative work with our target

population,21,22,79 and further pilot tested in the patient population of interest. The decision aid

included information about benefits and harms that are relevant to patients with lupus

nephritis. A published study protocol provides further details about the decision aid.78 We

designed the decision aid educational tool to improve understanding of a large amount of

information and decision quality.

We achieved the lupus nephritis decision aid development in 3 steps: (1) nominal group

technique in the target population to understand barriers and facilitators to treatment decision

making related to lupus; (2) performance of systematic review and network meta-analysis to

derive the estimates of comparative effectiveness and harms of each treatment option for

preference-sensitive decisions; and (3) pilot testing, iterative modification, and finalization of

the decision aid in English and Spanish for the target population.

The following sections briefly describe the methods for the NGT21-23 and the NMA and

systematic reviews,24-26 which we have reproduced from our peer reviewed publications with

permission from the respective journals.22,24 Detailed methods of the NGT and the NMA and

systematic reviews for other outcomes are available in listed publications.

NGT to Define the Barriers and Facilitators of Medication Decision Making in Lupus

Nephritis

We used the NGT with patients with lupus to assess the barriers and facilitators to

medication decision making for the treatment of lupus.22,79 The NGT is a structured process

to elicit ideas from participants for a formative assessment. We conducted 8 NGT

meetings in English at 2 medical centers, UAB and UCSF, moderated by an expert NGT

researcher. Participants responded separately to 2 questions: (1) Barriers: “What sorts

13

of things make it hard for people to decide to take the medicines that doctors prescribe

for treating their lupus kidney disease?”79 (2) Facilitators: “What sorts of things make it

easier for people to decide to take the medicines that doctors prescribe for treating their

lupus kidney disease?”22 In response to each of the 2 questions, patients nominated,

discussed, and prioritized (1) barriers and (2) facilitators to medication decisional

processes for lupus. In the section below, methods from our previous publication22 have

been replicated (with permission from the publisher).

We recruited patients from the lupus clinics at the University of Alabama at Birmingham

and the University of California at San Francisco. All patients met American College of

Rheumatology classification criteria for systemic lupus erythematosus and had a clinical

diagnosis of lupus nephritis (based on renal biopsy and/or laboratory tests).

We convened 8 NGT meetings including lupus nephritis patients who had received

treatment at UAB or UCSF lupus clinics. An expert NGT researcher (R.S.) conducted and

moderated all NGT meetings in English between February and April 2014. The Institutional

Review Boards at UAB and UCSF approved this study. All patients provided written, informed

consent.

The NGT meeting is a facilitated data collection activity, structured to promote even and

equal subject participation by minimizing the loss of information. Evidence shows that the NGT,

when used correctly, elicits a greater volume of novel and higher-quality responses to a

carefully articulated question than less structured group data collection approaches such as

focus groups and brainstorming.80,81 Moreover, by using the verbatim responses that are

concisely documented on a flipchart as participants present them to the group, the NGT

eliminates a potential source of investigator-induced interpretive bias resulting from

transcribing and coding audio or video recordings.

The purpose of the NGT meetings was to tap into patients’ unique insights, knowledge,

and lived experiences to identify different factors that facilitated their decision-making process

regarding prescribed lupus medications. The NGT leader (R.S.), along with a team member

(H.Q.), started the sessions with a brief explanation of the purpose and the NGT process.

Patients then worked independently for about 5 minutes to develop their own lists of brief

14

statements/phrases in response to the NGT question.

Patients were encouraged to think broadly about the types of factors that enhanced the

likelihood of deciding to take the medications prescribed for their condition. This ensured that

each panel generated a wide array of responses. After 5 minutes of working on their own,

patients were invited to present their responses to the group. To promote open disclosure,

increase response volume, and ensure all patients had an equal opportunity to present

responses, we used a “round-robin” participation format. This format involved having each

patient, in turn, articulate a single response without providing any rationale, justification, or

explanation for his or her response and without discussion or debate from other members in

the group. All responses were immediately recorded verbatim on a flipchart to help participants

recollect previously nominated responses. We continued until no further responses could be

generated. All responses were then discussed in a nonevaluative fashion to ensure that they

were understood from a common perspective and, potentially, to obtain additional insights.82

Patients were asked to silently review the full list of responses generated during the

meeting and to independently select 3 facilitators that they perceived as the most influential to

their decision making regarding lupus nephritis medication. Patients recorded their selected

responses on index cards and prioritized the influence of each of their selections from 1 (least

influential) to 3 (most influential). The votes reflecting these priorities were tabulated across

patients in each NGT panel to determine the perceived relative influence of medication

decision-making facilitators and the level of agreement among patients regarding these

perceptions.

A brief questionnaire was administered to the patients at the conclusion of each NGT

meeting to obtain basic demographic data, education level, and disease duration, and whether

the patient needed assistance in reading materials. Data from this questionnaire were analyzed

at the group level and not linked with individual responses generated during the NGT meetings.

NMA and Systematic Review of Medications Used for the Treatment of Lupus Nephritis

We performed the NMA and systematic review for several benefit and harm outcomes,

which we identified from NGT with patients with lupus nephritis. We published the results in 3

15

peer reviewed manuscripts focused on the following outcomes: (1) serious infections25; (2)

malignancy, herpes zoster, gastrointestinal side effects (gastrointestinal upset, diarrhea, etc),

nausea, alopecia, mycobacterial infections, hyperglycemia/diabetes, avascular

necrosis/osteonecrosis, mortality, amenorrhea, cytopenia, and urinary bladder toxicity

(including hemorrhagic cystitis and hematuria)26; and (3) renal remission/response, renal

relapse/flare, amenorrhea/ovarian failure, and cytopenia.24

In the section below, we describe the methods from our previous publication,24 which

have been replicated from the publication (with permission from the publisher). We chose to

describe methods for only a select few outcomes, rather than all outcomes, since (1) methods

were similar (or in some cases the same) for other outcomes with minor exceptions, as noted

below; and (2) most estimates for the outcomes described in the paper24 were used in the

patient decision aid. In the other 2 papers describing outcomes from the systematic

reviews,25,26 rankograms were not presented.

We used rigorous methods for the systematic review and NMA based on the Agency for

Healthcare Research and Quality (AHRQ) recommendations,72 the Cochrane handbook,83 and

the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

The Institutional Review Board at the University of Alabama at Birmingham approved the study.

The need for informed consent was waived for this systematic review, since no human subjects

were involved. The study protocol was registered in PROSPERO, CRD42016032965

(http://www.crd.york.ac.uk/PROSPERO/).

This systematic review included randomized controlled trials or controlled clinical trials

of immunosuppressive drugs or corticosteroids for lupus nephritis, published in English, that

reported any safety or efficacy outcome. Included medications were corticosteroids ([PRED],

cyclophosphamide [CYC], mycophenolate mofetil [MMF], azathioprine [AZA], cyclosporine,

tacrolimus, or rituximab). Belimumab studies could not be included in this systematic review

since these studies included patients with lupus, and only a small proportion had active lupus

nephritis. A Cochrane systematic review of belimumab for lupus is underway.84 There were no

restrictions regarding medication dose or the duration of medication use.

http://www.crd.york.ac.uk/PROSPERO/

16

Experienced librarians (J.J. and T.R.) updated 2 systematic reviews16,85 from their search

end dates (August 2010 and April 2012, respectively) to September 2013 using the PubMed

database. The search used the following terms:

(Lupus[text word] OR “Lupus Vulgaris”[MeSH] OR “Lupus Erythematosus,

Cutaneous”[MeSH] OR “Lupus Erythematosus, Systemic”[Mesh]) AND (“Kidney

Diseases”[MeSH] OR nephropath*[text word] OR Transplants[MeSH] OR Transplantation[MesH]

OR transplantation[subheading] OR transplant*[text word] OR “Kidney”[Mesh] OR Kidney*[text

word] OR Renal*[text word] OR “End Stage Renal Disease”[text word] OR ESRD[text word] OR

Glomerulonephr*[text word] OR “GN”[text word] OR “crescentic GN”[text word]) NOT

(“animals”[MeSH] NOT “humans”[MeSH]).

Raw data abstracted for the ACR lupus nephritis guidelines systematic review16 were

obtained (courtesy Dr. Jennifer Grossman [J.G.], see acknowledgment section), or were

abstracted from the Revman tables of the Cochrane Systematic Review.85 A librarian (C.H.) also

performed a search for all lupus trials for harms (for conditions other than lupus nephritis) in

PubMed and SCOPUS from inception to February 2014, based on an a priori assumption that

treatment-related harms may not depend on whether kidney is involved. Examination of data

from this search revealed little additive data for harms, but added clinical heterogeneity related

to differences in patient population. Therefore, after careful consideration of pros and cons, we

decided not to use these data in analyses.

We defined the PICO (patient, intervention, comparator, outcome) for our systematic

review and NMA as follows:

P: Adults 18 years or older, meeting the 1987 American College of Rheumatology

Classification criteria for systemic lupus erythematosus,86 who have lupus nephritis.

I: Immunosuppressant drug alone or in combination with other immunosuppressant

drugs or biologics (such as rituximab) or corticosteroids. We categorized medication

doses as low dose (LD), standard dose (SD), or high dose (HD).

C: Placebo or another immunosuppressive with or without biologic.

O: Benefit and harm outcomes (renal remission/response, renal relapse/flare, fertility,

bone marrow suppression), defined as follows.

17

We assessed benefits based on 2 composite outcomes: (1) renal remission/response

(indicating success of therapy), including complete renal remission,27 partial renal

remission,87,88 and renal response; and (2) renal relapse/flare (indicating failure of therapy),

including renal relapse89 and renal flare. We assessed harms based on 2 composite outcomes:

(1) ovarian failure/amenorrhea, including ovarian failure and amenorrhea; and (2) bone

marrow toxicity (cytopenia), including leucopenia.

Two trained abstractors (A.O, A.B) independently reviewed abstracts and titles,

abstracted data in duplicate directly into Microsoft Excel sheets, and assessed the risk of bias

according to the Cochrane risk of bias tool.90 We examined the following domains as low or

high risk of bias or unclear risk (lack of information or uncertainty about potential for bias):

randomization sequence generation, allocation sequence concealment, blinding of participants,

personnel and outcome assessors, incomplete outcome data (primary outcome data reporting,

dropout rates and reasons for withdrawal, appropriate imputation of missing data, an overall

completion rate ≥ 80%), and selective outcome reporting and other potential threats to validity

(considering external validity, e.g., relevant use of cointerventions, bias due to funding source).

An adjudicator (J.S.) resolved any disagreements not resolved by consensus. An expert

rheumatologist (J.S.) and an expert in lupus (J.G.) examined for similarity of studies prior to

performing evidence synthesis by the examination of similarity of study population and

interventions.

We designated doses as follows: (1) CYC: SD: 0.5 through 1.0 gm/m2 intravenously (IV) q

month for 6 to 12 months or 2 through 2.5 mg/kg orally (PO) daily over 3 to 6 months; HD: dose

higher or duration longer than SD; LD: dose lower or duration shorter than SD, including the

EURO-lupus dose, 500 mg IV q 14 days for 6 doses (mean 3 g); (2) AZA: SD: 1 to 3 mg/kg PO

daily; HD: > 3 mg/kg PO daily; (3) LEF: 1 mg/kg PO qd for 3 d then 30 mg PO qd for 6 months;

and (4) PRED: SD: prednisone/methylprednisolone 1 gm/m2 IV q month for 6 months or

prednisone 60 mg PO qd 1 to 3 months then tapered over 3 to 12 months as tolerated; HD:

prednisone/methylprednisolone 1gm/m2 qd IV 3 times, then 1 dose q month for 1 year or

18

prednisone 1mg/kg daily for 4 to 8 weeks (or unspecified period). When dose is not specified,

medication dose is the standard dose.

We used Bayesian mixed treatment comparison (MTC) meta-analyses75-77 to assess the

comparative effectiveness of one immunosuppressive drug versus another and

immunosuppressive drugs versus corticosteroids. We conducted a Bayesian MTC meta-analysis

using a binomial likelihood model using WinBUGS software (Cambridge, UK: MRC Biostatistics

Unit), which allows inclusion of data from multiarm trials.91,92 We conducted random-effects

NMA and assessed model fit and the choice of model (random versus fixed effects) based on

the assessment of the deviance information criterion and the comparison of residual deviance

to the number of unconstrained data points.91,93

We assigned vague priors, such as N (0, 1002) for basic parameters throughout91 and

informative priors for the variance parameter based on Turner et al.94 We evaluated the model

diagnostics including trace plots and the Brooks-Gelman-Rubin statistic to ensure model

convergence.91,95 We fit 3 chains in WinBUGS for each analysis, with 40 000 iterations, and a

burn-in of 40 000 iterations.95,96 Both MTC and traditional meta-analysis require studies to be

sufficiently similar in order to pool their results. We investigated heterogeneity, where

warranted, with subgroup analyses and meta-regressions.92,97 We examined consistency–

inconsistency plots for evidence of inconsistency, and chose the appropriate model for our

analyses. We obtained point estimates using odds ratios (ORs) and 95% credible intervals (CrI)

using Markov Chain Monte Carlo methods. We conducted transformation of the OR to relative

risk and risk difference to allow ease of interpretation for clinicians and patients. We assessed

the quality of evidence as recommended in a recent study.98

We performed sensitivity analysis by limiting analyses to partial/complete remission

rather than combining this with renal response for the composite renal remission/response. We

constructed staircase diagrams, another pictorial way to visualize comparisons of various

treatments against each other. We constructed rankograms to model the probabilities of the

treatment rankings, representing the best to the worst ranks.

Study Intervention Content Finalization and Pilot Testing

19

We performed iterative testing of the decision aid content. Examples from various

sections of the decision aid are provided in Figures 1 through 6. The decision aid for each

scenario provided information about lupus in general and how lupus affects kidneys, general

information about the medication (medication formulation, route of administration, costs of

medication, dosage), information about comparative risks and benefits about the 2

immunosuppressive drugs being compared, a final summary of the information provided, and

information on patient resources and patient support groups. The decision aid also provided 3

optional modules: benefits and side effects of using glucocorticoids, lupus and pregnancy, and

lupus and breast-feeding. We ensured the cultural sensitivity of the tool by developing the

content and the structure of the tool based on our qualitative work with predominantly African

American and Hispanic women (English- and Spanish-speaking) with lupus nephritis (but also

white and Asian women)21,22,79; pilot-testing both English- and Spanish-language versions in this

target group; making the decision aid understandable to people with low literacy, numeracy,

and graphical literacy; and keeping the decision aid at a sixth- to eight-grade reading level. The

control intervention consisted of the American College of Rheumatology pamphlet with

information about lupus, lupus nephritis, and its treatment.

20

Figure 1. Screen Shot of Lupus Nephritis Decision Aid: Introduction Section

21

Figure 2. Screen Shot of Lupus Nephritis Decision Aid: Example of Comparison of a Benefit (Prevention of Kidney Failure) of Treatment Choices

22

Figure 3. Screen Shot of Lupus Nephritis Decision Aid: Example of Comparison of a Harm/Toxicity (Shingles) Related to the Treatment Choices for Lupus Nephritis

23

Figure 4. Screen Shot of Lupus Nephritis Decision Aid: Example of Potential Concerns Related to the Use of Immunosuppressive Drugs During Pregnancy

24

Figure 5. Screen Shot of Lupus Nephritis Decision Aid: Example of Other Concerns Related to Medications Based on Qualitative Work With Patients With Lupus Nephritis

25

Figure 6. Screen Shot of Lupus Nephritis Decision Aid: An Example of a Summary Slide

26

Study Design, Population, and Setting: Randomized Trial

This was a multicenter, parallel 2-arm, prospective randomized trial comparing an

individualized computerized decision aid tool to an educational pamphlet (usual care). A

published study protocol provides details.78 Patients with lupus nephritis attending

rheumatology or nephrology clinics at the 4 sites (UAB, UCSF, OSU, and Baylor College of

Medicine) were recruited over 2 years starting in January 2015. Each site has a weekly lupus

clinic and has been actively engaged in lupus research. A multisite study conducted with

geographically diverse centers ensured the representativeness of the study population and the

generalizability of our study findings. The PIs of each lupus clinic were our study

coinvestigators, who helped us develop the recruitment and retention plan, with a focus on

racial and ethnic minorities. We developed our intervention and all assessment forms in both

English and Spanish and included experienced translators as coinvestigators. Study coordinators

at each site were bilingual and/or nonwhite or had extensive experience with recruiting

minorities in studies. A list with hospital clinic appointments of lupus patients, their gender, and

their race/ethnicity was generated each month using the International Classification of

Diseases, code 710.0. We screened this list on a weekly basis to identify potentially eligible

individuals, and then discussed potential study eligibility of the individuals with their lupus care

provider before their visit. Study inclusion criteria were (1) adult women (≥ 18 years) of any

race/ethnicity with lupus nephritis and (2) currently having a flare of lupus nephritis and

considering a change or initiation of an immunosuppressive medication (current flare; including

new incident lupus cases) or patients with a history of lupus nephritis and at risk for a future

lupus nephritis flare (future flare). Exclusion criteria were (1) men, (2) lupus but no nephritis, (3)

current lupus nephritis flare but medication change not considered, (4) end-stage renal disease

on dialysis, and (5) renal transplant or candidate for a renal transplant. Patients were recruited

at the time of their regular clinic visits. All primary and secondary outcome assessments were

patient reported and completed during the baseline visit, to eliminate the burden of additional

study visits. Patients were provided a check for $70 for completing the baseline study

procedures, which usually took 45 to 90 minutes.

27

Randomized Trial Study Intervention

The randomized trial compared the decision aid in English or Spanish against an ACR

pamphlet in this 2-arm trial.

Arm 1, Study Arm: Individualized computerized decision aid for immunosuppressive

drugs for lupus nephritis (for the 4 most common scenarios, i.e., immunosuppressive drug

choice decision points: mycophenolate mofetil versus cyclophosphamide for induction therapy

or the 3 maintenance therapy scenarios; see background section), developed specifically for

this study, with a focus on racial/ethnic minorities from a diverse socioeconomic background.

Patients randomized to the intervention group viewed the tool on a tablet computer. Patients

were prompted to write down questions for their physicians about their treatment choices. The

computerized decision aid was programmed once a scenario was chosen by the coordinator per

the guidance of the referring physician based on the most likely 2 choices (or a provider in the

future); the decision aid shows only that comparison. Therefore, individualization of this web-

based touchscreen tool was achieved at multiple levels by (1) allowing subjects to stop, go back,

and review the presentation, depending on their understanding of the decision aid content; (2)

providing optional links to the management of the adverse effects of medications; (3) providing

3 optional sections, 1 each on glucocorticoids, pregnancy, and lactation, to be viewed based on

its relevance; and (4) making selection of 1 of the 4 treatment scenarios by the physician, based

on individual patient treatment choices for lupus nephritis, each of which provided a different

sets of comparisons.

Arm 2, Control Arm: The control intervention was an ACR pamphlet with information

about lupus, lupus nephritis, and its treatment.99

Baseline and Follow-up Study Visits

After obtaining written informed consent and the physician’s designation of the

treatment scenario, patients were randomized to either the decision aid or the pamphlet group

in a 1:1 ratio (stratified by study site and language) using the Research Electronic Data Capture

tool.100 Baseline assessments were then administered including the Rapid Estimate of Adult

28

Literacy in Medicine–Short Form,101 the Short Assessment of Health Literacy,102 measures of

graphical and numeric literacy, and the measures of primary outcomes including the Decisional

Conflict Scale103 and informed choice (details in the section on outcome measures below;

knowledge, patient values, and choice of the immunosuppressive drug for lupus

nephritis).104,105 Questionnaires were modified slightly for patients with future lupus nephritis

flare, who were asked to answer questions thinking about a future flare and a decision at that

point in time. After the completion of the preintervention assessment, the intervention was

administered. This was followed by an audiotaped conversation between patients and

physicians (secondary outcome) for patients with current lupus nephritis flares only, who

agreed to recording of the conversation. The postintervention questionnaire measured

coprimary outcomes (decisional conflict and informed choice) and 2 secondary outcomes (i.e.,

the control preference scale106-108 and Interpersonal Process of Care–Short Form) for all

patients.109

We kept follow-up assessments to a minimum due to the nature of the study and to

minimize missing data. At 3 months, study subjects responded again to the IPC-SF

questionnaire either during a routine clinic visit or via phone (if no clinic visit), or via mail (if not

reachable via phone and not seen in the clinic). This was the only study assessment after the

initial visit. The study coordinator extracted laboratory, medication, and other clinical data on

exploratory outcomes (cumulative glucocorticoid dose, serum creatinine, spot

protein/creatinine ratios, renal remission, and missed appointments) at 6 months using

Electronic Health Record (EHR) clinical care data at each study site for patients currently

experiencing lupus nephritis flare.

Outcome Measures

Coprimary Outcome Measures

Decisional Conflict Scale (DCS), low literacy version: The DCS is a patient self-

administered, validated measure of decisional conflict, most commonly used as the primary

outcome in RCTs of decision aids.110,111 The low literacy version consists of 10 items with 3

29

response categories (yes, unsure, no) with overall scores ranging from 0 (no decisional conflict)

to 100 (extreme decisional conflict), available in English and Spanish.112 We scored responses as

yes = 0; unsure = 2; no = 4. We summed 10 items and multiplied them by 2.5 to provide a score

ranging from 0 to 100. Decisional conflict represents a state of uncertainty about a choice or

course of action and is more likely in situations involving high-stakes choices with important

potential gains and losses, value tradeoffs in choosing one selection or one course of action

(versus the alternative), or uncertain outcomes. We chose the DCS scale based on our

intervention and the formative work showing significant doubts and decisional conflict

regarding immunosuppressive drugs in lupus patients.

Informed choice: We assessed informed choice by using a validated multidimensional

model of informed choice104,105 that individually assesses and then combines 3 constructs:

values regarding immunosuppressive drugs, knowledge about immunosuppressive drugs, and

treatment choices. We assessed informed choice after each patient had viewed the decision aid

or the ACR pamphlet at the baseline study visit before any treatment decision making. We

assessed values with a list consisting of patients’ views regarding immunosuppressive drugs as a

treatment option and their side effects generated by our study team based on patient concerns

regarding immunosuppressive drugs. The values statements consisted of both positive and

negative values about immunosuppressive drugs, mixed in a random order, with responses

ranging from strongly disagree to strongly agree. We scored positive and negative value

statements with appropriate signs (+ or –) and aggregated this into a total score. A higher total

score indicated more positive values regarding using immunosuppressive drugs, and we used a

median score to classify values as positive versus negative regarding using immunosuppressive

drugs. We assessed knowledge related to immunosuppressive drugs for lupus nephritis using 20

questions. We considered patients to have adequate knowledge if they answered at least 75%

of questions correctly. We assessed choice based on response to a single item on a nominal

scale with anchors of “start vs. don’t start immunosuppressives” and “uncertain” in the middle

and 15 circles, asking each patient’s choice in response to the question, “If your doctor asked

you right now to make a choice about immunosuppressives, please show where you would be

on the scale below by choosing a circle below.” Informed choice referred to a choice that is

30

based on accurate knowledge and is concordant with one’s values. A higher proportion of

patients with an informed choice is optimal. Table 1 provides details for this outcome. We

performed a sensitivity analysis for informed choice by reclassifying subjects according to net

score (positive or negative) on value statements, comparing value statements favorable toward

immunosuppressives with those not favorable, rather than the median score; we classified

those with a net positive score as favoring immunosuppressives and those with a net negative

score as against immunosuppressives.

Table 1. Classification Criteria for Informed Choice Willing to Take Immunosuppressive Drugs Not Willing to Take Immunosuppressive

Drugs Values Values Knowledge Favor

immunosuppressives Against immunosuppressives

Favor immunosuppressives

Against immunosuppressives

Adequate Informed choice - - Informed choice Not adequate

- - - -

- Represents a choice that was not informed; subjects who did not make a decision were considered to be in a the “not informed choice” group

Secondary Outcome Measures

Control preferences scale: This validated measure assessed patient participation in

decision making for only those patients with a current flare. The scale assessed how much

decision-making control a patient would like to have versus control actually experienced by

each patient. It distinguishes between those who feel involved in the decision versus those who

do not.106-108 The measure included 5 responses corresponding to 5 control options: active,

active shared, collaborative, passive shared, and passive. We categorized these roles as active

(combining active and active shared), collaborative, or passive (combining passive and passive

shared).113-115 We examined the concordance between the desired and the actual role played

by each patient with a current lupus nephritis flare.

Patient–physician communication and care processes: We used the IPC-SF, an 18-item

validated patient-reported measure of patient–physician communication and care processes,

31

available in English and Spanish.109,116-119 The IPC-SF score ranges from 18 (worst) to 90 (best);

higher scores represent better patient–physician communication and care processes.

Analysis of an audiotaped physician–patient interaction: For the current flare patients

only, we recorded the patient–physician discussion about immunosuppressives and used the

Active Patient Participation Coding Scheme, a validated instrument, to assess indicators and

facilitators of patient participation.120 We coded 3 types of speech acts (question asking,

assertive responses, and expressions of concern) as active patient participation, because they

may influence a doctor’s behavior as well as the content and structure of the consultation.121-124

We coded physician communication using speech acts such as supportive talk and partnership

building. We summed these units (range 0 to infinity) for each interaction to create a frequency

for the degree of active participation.

Acceptability and feasibility of the decision aid: We assessed information quality and

quantity, presentation style, and usefulness of the decision aid using a validated acceptability

survey62 on a 4-point scale ranging from excellent to poor. We assessed feasibility of the

decision aid versus the pamphlet and the study procedures with a self-administered

questionnaire.125 Patients rated the feasibility of the decision aid versus the pamphlet by

responding to the statement “the education guide was easy to use” on a 5-point Likert scale

(strongly agree to strongly disagree).

Clinically meaningful difference: We considered a 10% difference between study arms in

the proportion of patients achieving a favorable or unfavorable outcome as clinically

meaningful. We designated a 5% difference in proportions as “possibly” clinically meaningful.

Statistical Analyses

Sample size calculation: Our study had an 80% power to assess the treatment effect for

the coprimary outcomes, with an estimated enrollment of 200 patients, after allowing for a

10% loss to follow-up. We anticipated that 100 African American and 100 Hispanic/Caucasian

women with lupus nephritis should be enrolled. We anticipated the ability to detect a medium

effect size difference (effect size = difference between 2 means/standard deviation of the data;

according to Cohen small, medium, and large effect sizes correspond to 0.2, 0.5, and 0.8,

32

respectively126) between group means on decisional conflict (range 0-100) using a 2-sample t

test and 2-tailed type 1 error rate of 0.05 (hypothesis 1)112,127 and a 15% absolute difference in

the proportion of patients with informed choice using a 1-sided type 1 error rate of 0.05

(hypothesis 2).128

Analysis of outcomes measures: We compared baseline characteristics including

demographic variables between the decision aid and pamphlet groups. We compared primary

and secondary outcome measures using student t test or analysis of variance or comparison of

proportions. For continuous measures, we checked the normality assumption using normal

probability plots; for any measures that showed possible departures from normality, we used

the Wilcoxon rank sum test (a nonparametric test) to verify results. No conclusions differed by

method (parametric versus nonparametric test). Therefore, we report the results of the

parametric tests for easy interpretation of the study results. We performed sensitivity analysis

for informed choice by reclassifying subjects according to net score (positive or negative) on

value statements, instead of the median, comparing value statements favorable toward

immunosuppressives with those not favorable. We considered a 2-sided P value < 0.05

significant. We stratified variables based on language (English or Spanish) and study site and

compared them using a chi-square test or Fisher’s exact test. We conducted all statistical

analyses conducted using SAS, Version 9.4 ([computer program] Cary, NC).

Conduct of the Study

We made the following major protocol amendments due to a recruitment shortfall: (1)

We added 2 additional sites, Baylor University and Ohio State University, to allow us to reach

our target goal; (2) we enrolled patients with current flare or at risk of future flare, since the

total number of patients making treatment decisions for a current flare only was fewer than

estimated; and (3) we enrolled Caucasian and Asian women to increase the generalizability of

the study findings and potentially make this decision aid relevant to all women with lupus

nephritis.

33

Results

We conducted a systematic review, meta-analysis, and NMA to assess the comparative

efficacy and harms of immunosuppressive drugs (Aim 1); performed a nominal group technique

with primarily minority women with lupus nephritis to prioritize barriers and facilitators to

immunosuppressive drug decision making (Aim 2); and iteratively modified and finalized our

decision aid by testing it in women with lupus nephritis (Aim 3).

Detailed results of the NMA and systematic review have been published.24-26 This study

provided comparative estimates for benefits and harms of immunosuppressive drugs in lupus

nephritis for our decision aid. We provide a brief summary below. We included 65 RCTs that

met the inclusion and exclusion criteria. Significantly lower risk of end-stage renal disease

(ESRD; 17 studies) was seen with cyclophosphamide (CYC; odds ratio 0.49, 95% credible interval

0.25-0.92) or CYC + azathioprine (AZA; OR 0.18, 95% CrI 0.05-0.57) compared with standard-

dose glucocorticoids, and with high-dose (HD) CYC (OR 0.16, 95% CrI 0.03-0.61) or CYC + AZA

(OR 0.10, 95% CrI 0.03-0.34) compared with high-dose glucocorticoids. High-dose

glucocorticoids were associated with higher risk of ESRD compared with CYC (OR 3.59, 95% CrI

1.30-9.86), AZA (OR 2.93, 95% CrI 1.08-8.10), or mycophenolate mofetil (MMF; OR 7.05, 95% CrI

1.66-31.91).26 No differences were noted between medications for the risk of malignancy (15

studies). The risk of herpes zoster versus glucocorticoids (17 studies) was as follows, OR (95%

CrI): MMF, 4.38 (1.02-23.87); CYC, 6.64 (1.97-25.71); tacrolimus [TAC], 9.11 (1.13-70.99); and

CYC + AZA, 8.46 (1.99-43.61). We concluded that renal benefits and the risk of herpes zoster

were higher for immunosuppressive drugs versus glucocorticoids. In another systematic review,

we compared the risk of serious infections with various drugs in lupus nephritis (32 RCTs with

2611 patients provided data). Tacrolimus was associated with significantly lower risk of serious

infections compared with glucocorticoids, CYC, MMF, and AZA, with odds ratios (95% CrI) of

0.33 (0.12-0.88), 0.37 (0.15-0.87), 0.34 (0.18-0.81), and 0.32 (0.12-0.81), respectively.25 We also

found that MMF treatment followed by AZA (MMF-AZA sequential treatment) was associated

with significantly lower risk of serious infections compared with CYC LD, CYC HD, CYC-AZA, or

HD glucocorticoids, with odds ratios (95% CrI) of 0.09 (0.01-0.76), 0.07 (0.01-0.54), 0.14 (0.02-

0.71), and 0.03 (0.00-0.56), respectively. Sensitivity analyses confirmed these findings. We

34

concluded that tacrolimus and MMF-AZA combination each were associated with a lower risk of

serious infections compared with other immunosuppressive drugs or glucocorticoids for lupus

nephritis. In another analysis, we assessed the rates of renal remission/response (37 trials; 2697

patients), renal relapse/flare (13 studies; 1108 patients), amenorrhea/ovarian failure (8 trials;

839 patients), and cytopenia with immunosuppressive drugs (16 trials; 2257 patients). CYC LD

and CYC HD were less likely than MMF-AZA, CYC LD, CYC HD, and plasmapheresis less likely

than cyclosporine to achieve renal remission/response.24 TAC was more likely than CYC LD to

achieve renal remission/response. MMF and CYC were associated with lower odds of renal

relapse/flare compared with glucocorticoids and MMF was associated with a lower rate of renal

relapse/flare than AZA. CYC was more likely than MMF and glucocorticoids to be associated

with amenorrhea/ovarian failure. Compared with MMF, CYC, AZA, CYC LD, and CYC HD were

associated with a higher risk of cytopenia. We used these estimates in the decision aid tool to

depict the comparative benefits and harms of various immunosuppressive medications, as

applicable to a given comparison/scenario for induction or maintenance therapy.

We incorporated themes generated from NGTs into our decision aid. Since we recruited

a similar target patient population for the NGT as we aimed for in the trial (racially/ethnically

diverse with diverse socioeconomic status and literacy level), themes and content generated

were culturally appropriate. An NGT expert moderated 8 patient group meetings at

Birmingham and San Francisco, in which 52 women with lupus nephritis participated (27 African

American, 13 Hispanic, and 12 Caucasian). The average age was 40.6 years (standard deviation

(SD) = 13.3), and the mean disease duration was 11.8 years (SD = 8.3); 36.5% had at least some

college education, and 55.8% had difficulty in reading health materials (Aim 1).

Detailed results of the NGT have been published.21-23 Briefly, participants (n = 52)

responded to the question “What sorts of things make it easier for people to decide to take the

medicines that doctors prescribe for treating their lupus kidney disease?” and generated 280

decision-making facilitators.22 Of these, 102 (36%) facilitators were perceived by patients as

having relatively more influence in decision-making processes than others. Prioritized

facilitators included effective patient–physician communication regarding benefits and harms,

patient desire to live a normal life and improve quality of life, patient concern for their

35

dependents, experiencing benefits and few/infrequent/no harms with lupus medications, and

medication affordability (Aim 2). In a similar NGT (n = 51), participants with lupus nephritis

responded to the question “What sorts of things make it hard for people to decide to take the

medicines that doctors prescribe for treating their lupus kidney disease?”23 The most salient

perceived barriers were known/anticipated side effects (15.6%), medication expense/ability to

afford medications (8.2%), and the fear that the medication could cause other diseases (7.8%).

We discussed these barriers in detail and incorporated them into the development of the

decision aid. These barriers and facilitators guided the content of the decision aid. For example,

we included slides on medication cost in the decision aid and focused the content on the key

side effects of these medications identified by patients during the NGT to be most relevant to

women with lupus nephritis. We also focused benefits on relevant aspects (dialysis, patient-

relevant benefits) identified by patients in the NGT, rather than benefits and harms that are

relevant to providers only, such as decreased white cell counts and improvement in laboratory

measures of proteinuria. We also included 3 optional sections, 1 each on glucocorticoids,

pregnancy, and breast-feeding, based on the additional concerns identified from the NGT. After

finalization of our decision aid content, we tested it iteratively in 19 subjects with lupus (mean

age 39 years). We made edits, word substitutions, and corrections as suggested by patients.

Three patients reviewed a pre–pilot test version of the decision aid website for color and the

ease of use. They suggested a change in the background color and the exclusion of a navigation

bar; we made both changes before the finalization of the electronic version of the decision aid

(Aim 3).

We performed a randomized trial of the decision aid versus the pamphlet (Aim 4). In the

section below, we provide results using tables adapted from clinicaltrials.gov as per instructions

including the CONSORT flow chart (Figure 7) and a study flow diagram showing the study

procedures (Figure 8).

36

37

38

Participant Flow Recruitment details Preassignment details Arm/group title Decision aid Pamphlet Total

(not public)

Arm/group description Participants received decision aid tool

Participants received the ACR lupus pamphlet

Period Title: Overall Study Started 153 148 301 Primary research completion 151 147 298 Completed 151 147 298 Not completed 2 1 3 Reason not completed Withdrew consent before

receiving intervention Withdrew consent before receiving intervention

Withdrawal by subject 2 1 3 (Not public) Not completed = 2

Total from all reasons =2 Not completed = 1 Total from all reasons = 1

Baseline Characteristics Arm/Group Title Decision Aid Pamphlet Total Overall Number of Baseline Participants 151 147 298 Age, categorical Measure type: count of participants Unit of measure: participants

Number analyzed

151 participants

147 participants

298 participants

≤ 18 years 0 0%

0 0%

0 0%

Between 18 and 65 years

148 98.01%

146 99.32%

294 98.66%

≥ 65 years 3 1.99%

1 0.68%

4 1.34%

Age, continuous Mean (full range) Unit of measure: years

Number analyzed

151 participants

147 participants

298 participants

37.1 (19 to 69)

37.6 (19 to 66)

37.3 (19 to 69)

Sex: female, male Measure type: count of participants Unit of measure: participants

Number analyzed

151 participants

147 participants

298 participants

Female 151 100%

147 100%

298 100%

Male 0 0%

0 0%

0 0%

Race/ethnicity, customized Measure type: number Unit of measure: participants

Number analyzed

151 participants

147 participants

298 participants

Not answered

2 0 2 Asian

11 9 20

39

Hispanic/Latino

41 37 78 Non-Hispanic black

70 71 141

Non-Hispanic white

20 24 44 Other

7 6 13

Region of enrollment Measure type: number Unit of measure: participants

Number analyzed

151 participants

147 participants

298 participants

United States

151 147 298 Decisional conflict[1] Mean (standard deviation) Unit of measure: units on a scale

Number analyzed

151 participants

147 participants

298 participants

33.37 (29.55)

37.48 (29.85)

35.4 (29.72)

[1] Measure description: The Decisional Conflict Scale is a patient self-administered, validated measure of decisional conflict—a state of uncertainty about a course of action. Scores range from 0 (no decisional conflict) to 100 (extremely high decisional conflict).

Knowledge about immunosuppressives[1] Measure type: number Unit of measure: participants

Number analyzed

151 participants

147 participants

298 participants

Adequate knowledge

90 89 179 Inadequate knowledge

61 58 119

[1] Measure description: Patients were asked 20 true/false questions regarding lupus nephritis and immunosuppressive drugs. Patients answering at least 75% of questions correctly were considered to have adequate knowledge about immunosuppressives. Those answering fewer than 75% of questions correctly were considered to have inadequate knowledge.

Preintervention unresolved clinically significant decisional conflict[1] on Decisional Conflict Scale (score ≥ 25) Measure type: count of participants Unit of measure: participants

Number analyzed

151 participants

147 participants

298 participants

Unresolved conflict 85 93 178 Not unresolved conflict 66 54 120 [1]

Measure description: Unresolved conflict on the Decisional Conflict Scale is defined as a score of 25 or more. The Decisional Conflict Scale is a patient self-administered, validated measure of decisional conflict—a state of uncertainty about a course of action. Scores range from 0 (no decisional conflict) to 100 (extremely high decisional conflict).

Of patients, 68 (35 decision aid, 33 pamphlet) had current lupus flare; 107 (52 decision

aid, 55 pamphlet) were newly diagnosed.

40

1. Primary Outcome

Title: Change From Baseline in Decisional Conflict Scale Scores (Reduction)

Description Patient self-administered, validated measure of decisional conflict, most commonly used as the primary outcome in RCTs of decision aids (change score). The score ranges from 0 (no decisional conflict) to 100 (extreme decisional conflict). Decisional conflict represents a state of uncertainty about a choice or course of action and is more likely in situations involving high-stakes choices with important potential gains and losses, value tradeoffs in selecting a choice or a course of action (versus the alternative), or uncertain outcomes.

Time Frame Baseline and after viewing the decision aid or the standard pamphlet on the same visit as the intervention (preferred) but before treatment decision making (usually within 1 week)

Outcome Measure Data

Analysis Population Description All participants who received either the decision aid or pamphlet

Arm/Group Title Decision Aid Pamphlet


Participants received the standard ACR pamphlet

Overall number of participants analyzed

151 147

Mean reduction (standard deviation) Unit of measure: units on a scale

21.80 (30.89) 12.69 (24.41)

Statistical Analysis 1 Statistical analysis overview

Comparison group selection Decision aid, pamphlet comments [Not specified] Type of statistical test Superiority or other (legacy) Comments [Not specified]

Statistical test of hypothesis

P value 0.005 Comments [Not specified] Method t test, 2 sided Comments [Not specified]

The effect size (Cohen’s d) for change in decisional conflict was 0.33. The proportion of

patients with unresolved clinically significant decisional conflict (score ≥ 25) postintervention

was greater in the pamphlet than in the decision aid group—44.2% versus 22.5% (p < 0.001).

Median (IQR) was 10 (45) for the decision aid group and 10 (25) for the pamphlet group.

Sensitivity Analysis: Change in Decisional Conflict Scale scores was almost normally

distributed (see Figure 9). For sensitivity analysis, we used the Wilcoxon rank sum test, a

41

nonparametric test, to compare change in DCS (in case of deviation from a normal distribution).

We found that the decision aid group had a significantly larger change in DCS compared with

the pamphlet group using the Wilcoxon rank sum test (p = 0.04).

Figure 9. Distribution of Change in DCS Scores

42

2. Primary Outcome

Title Informed Choice (Validated Instruments for Values Regarding Immunosuppressives, Knowledge About Immunosuppressives, and Treatment Decision Making)

Description

Concordance between values related for or against starting immunosuppressive drugs with patients’ decision on (starting or not starting) immunosuppressive drugs, in those with adequate knowledge about benefits/harms of immunosuppressive drugs, assessed using validated instruments for values regarding immunosuppressive drugs, knowledge about immunosuppressive drugs, and treatment decision making (patient’s decision to start immunosuppressive drug)

Time Frame

After viewing the guide or standard pamphlet on the same visit as the intervention (preferred) but before treatment decision making (usually within 1 week)


Analysis Population Description All participants who received either the decision aid or the pamphlet




Overall number of participants analyzed 151 147

Measure type: number Unit of measure: participants

Row Title

Informed choice made 62 46 No informed choice 89 101

Statistical Analysis 1

Statistical analysis overview

Comparison group selection

Decision aid, pamphlet

Comments [Not specified] Type of statistical test Superiority or other (legacy) Comments [Not specified]


P value 0.08 Comments We compared informed choice versus no informed choice

made between the decision aid and the pamphlet groups. Method Chi-square Comments [Not specified]

Sensitivity Analysis: Using an alternate definition for patient values regarding

immunosuppressives (sensitivity analysis), more women in the decision aid group made an

informed choice compared with those in the pamphlet group (50.3% versus 34.7%; p = 0.006).

43

3. Secondary Outcome

Title Control Preferences Scale: Patient Participation in Decision Making Description This scale assessed how much decision-making control patients would like to have

versus what they actually experienced. There are 5 responses for 5 control options: active, active shared, collaborative, passive shared, and passive, which we collapsed into active (active, active shared), collaborative, and passive (passive shared, passive), as previously indicated (and prespecified). We assessed concordance between desired and actual role played by each patient. We present these data for patients with current flare only, since only they were making a decision about the immunosuppressive drugs; patients with past lupus flare were not included in the denominator.

Time Frame After viewing the guide or standard pamphlet on the same visit as the intervention (preferred) but before treatment decision making (usually within 1 week)


Analysis Population Description Only participants having current lupus nephritis and requiring immunosuppressive

medication change/initiation or participants with newly diagnosed lupus nephritis starting an immunosuppressive medication




Overall number of participants analyzed 35 33 Measure type: number Unit of measure: participants

Row Title

Concordance between roles 33 28 No concordance between roles 2 5



Comparison group selection Decision aid, pamphlet Comments We compared concordance between preferred and

actual roles versus no concordance between roles between decision aid and pamphlet.

Type of statistical test Superiority or other (legacy) Comments [Not specified]


P value 0.252 Comments [Not specified] Method Chi-square Comments [Not specified]

44


Title Patient Physician Communication (Interpersonal Processes of Care [IPC-SF]) Description

This was assessed using the IPC-SF, an 18-item validated patient-reported measure of patient–physician communication and care processes. The score ranges from 18 (worst) to 90 (best) and the scale is a patient-reported measure of patient–physician communication and care processes.

Time Frame

After viewing the guide or standard pamphlet on the same visit as the intervention (preferred) (usually within 1 week)


Analysis Population Description All participants who received either the decision aid or the pamphlet




Overall number of participants analyzed 149 147 Mean (standard deviation) Unit of measure: units on a scale

83.64 (7.69) 83.06 (7.28)

Statistical analysis overview Comparison group selection Decision aid, pamphlet


Statistical test of hypothesis P value 0.504

Comments [Not specified] Method t test, 2-sided Comments [Not specified]


Title Analysis of Audiotaped Physician–Patient Interaction (Using the Active Patient Participation Coding Scheme [APPC]): Doctor Patient-centered Communication

Description

We performed this by analyzing the audiorecorded patient–physician discussion in patients with current lupus nephritis flare. The APCC is a validated instrument to measure active patient participation. APCC assesses indicators and facilitators of patient participation. The unit of coding is the utterance, the oral analogue of a sentence. The range is 0 to unlimited. Patient participation is measured by the number of questions, number of concerns expressed, and act of assertiveness (e.g., preferences, introducing topics, making requests). These are active forms of participation because of their influence on clinician behavior and the structure and content of the consultation. The APPC also assesses clinician behaviors that facilitate and support patient participation,

45

partnership building, and supportive talk (e.g., reassurance, empathy). We present patient-centered communication by doctor/health care provider. Higher scores indicate better patient participation and communication.

Time Frame

After viewing the guide or standard pamphlet on the same visit as the intervention (preferred) (usually within 1 week)


Analysis Population Description Only participants having current lupus nephritis and requiring immunosuppressive

medication change/initiation or participants with newly diagnosed lupus nephritis starting an immunosuppressive medication, who also agreed to an audiorecorded conversation




Overall number of participants analyzed 16 17 Mean (standard deviation) Unit of measure: units on a scale

5.1 (2.1) 3.7 (1.9)

Statistical analysis overview Comparison group selection Decision aid, pamphlet


Statistical test of hypothesis P value 0.06

Comments [Not specified] Method t test, 2-sided Comments [Not specified]

Acceptability

Title Acceptability (Number of Participants Rating Each Statement as “Excellent”)

Description We assessed acceptability of the decision aid (information quality and quantity, presentation style, and usefulness) using a validated acceptability survey on a 4-point scale ranging from excellent to poor (response options were excellent, good, fair, and poor). We compared the number of patients rating each of the five statements as excellent (versus other ratings) between the 2 treatment arms.

Time Frame After viewing the guide or standard handout on the same visit as the intervention (preferred) (usually within 1 week)


Analysis Population Description Title Decision Aid Pamphlet



46


151 147

Measure type: count of participants Unit of measure: participants

Row Title

Impact of lupus nephritis 74 49 Risk factors 64 40 Medication options 76 49 Evidence about medications 71 35 Studies about other patients 64 32



Comparison group selection Decision aid, pamphlet Comments Statistical analysis comparing the decision aid versus

pamphlet for patient rating of acceptability of information and presentation related to the impact of lupus nephritis







pamphlet for patient rating of acceptability of information and presentation related to the risk factors







pamphlet for patient rating of acceptability of

47

information and presentation related to the medication options







pamphlet for patient rating of acceptability of information and presentation related to the evidence about medications



P value < 0.001 Comments [Not specified] Method Chi-square Comments [Not specified]




pamphlet for patient rating of acceptability of information and presentation related to the evidence about other patients



P value < 0.001 Comments [Not specified] Method Chi-square Comments [Not specified]

Feasibility

Title Feasibility (Number of Participants Rating the Feasibility of Using Decision Aid or Pamphlet—Referred to as Education Guide in This Statement)

48

Description We assessed feasibility of the decision aid versus pamphlet using a single statement: “The education guide was easy to use.” Patients rated this on a 5-point ordinal scale ranging from strongly agree to strongly disagree (response options were strongly agree, agree, neither agree nor disagree, disagree, strongly disagree). We compared the number of patients between the 2 treatment arms.

Time frame After viewing the guide or standard handout on the same visit as the intervention (preferred) (usually within 1 week)


Analysis Population Description One patient from the pamphlet group did not respond to this question; therefore, valid

responses from the pamphlet were 146, not 147. Title Decision Aid Pamphlet




151 147

Measure type: count of participants Unit of measure: participants

Row Title

Strongly disagree 1 3 Disagree 1 13 Neither disagree nor agree 73 74 Agree 75 55 Strongly agree 1 1 Missing 0 1



Comparison group selection Decision aid, pamphlet Comments The test of significance compared all the rows—i.e., all

response options for the statement. Type of statistical test Superiority or other (legacy) Comments [Not specified]



49

Adverse Events Time frame 3 months Adverse event reporting description Source vocabulary name for table [Not specified] Collection approach for table default Nonsystematic assessment Arm/group title Decision aid Pamphlet



All-cause Mortality Decision Aid Pamphlet Affected/at risk

(%)

Affected/at risk (%)

Total 1/151 (0.66%) 1/147 (0.68%)

Serious Adverse Events Decision Aid Pamphlet Affected/at risk

(%) # events

Affected/at risk (%)

# events

Total 1/151 (0.66%) 1/147 (0.68%) Vascular Disorders

Right ventricular failure[1]∗ 1/151 (0.66%) 1 0/147 (0%) 0 Subarachnoid hemorrhage[2]∗ 0/151 (0%) 0 1/147 (0.68%) 1 ∗ Indicates events were collected by nonsystematic methods. [1]

Decision aid: Patient with mitral regurgitation died due to right ventricular failure after cardiovascular operation (Day 53).

[2] Pamphlet: Subarachnoid hemorrhage from posterior circulation aneurysm. Patient died due to central herniation (Day 22).

Other (Not Including Serious) Adverse Events

Frequency threshold for reporting other adverse events

0%

Decision Aid Pamphlet Affected/at risk (%) #

events Affected/at risk (%)

# events

Total 0/151 (0%) 0/147 (0%)

6-month Outcomes (Exploratory)

We were unable to analyze any exploratory outcome in a meaningful way, due to the

sporadic availability and heterogeneity of these outcomes (laboratory values, medication use

data, etc.). This was related to the differences in clinical protocols, follow-up visit times, and the

50

EHR systems, a passive data collection across each site as per clinical follow-up, and the

limitation of most exploratory outcomes to patients with current flares only.

Subgroup Analyses

Table. Subgroup Analyses of Outcomes by Race/Ethnicity Postintervention

Decision Aid Mean (SD) or n (%)

Postintervention Pamphlet Mean (SD) or n (%) P Value

Change in DCS Score

Non-Hispanic black (n = 141) 25.52 (3.71) 16.99 (3.18) < 0.001

Hispanic/Latino (n = 78) 13.5 (4.29) 6.76 (3.99) 0.07 Non-Hispanic white (n = 44) 30.25 (8.6) 12.29 (3.85) 0.002

Asian/other (n = 33) 19.17 (6.34) 7.44 (4.61) 0.05

Informed Choice

Non-Hispanic black 28 (40%) 20 (28.2%) 0.22

Hispanic/Latino 14 (35%) 12 (32.4%) 0.61

Non-Hispanic white 14 (70%) 5 (20.8%) 0.003

Asian/other 6 (33.3%) 9 (60%) 0.62

t test or chi-square test, as appropriate.

51

Discussion

Context for Study Result

We found that an individualized, culturally tailored, computerized patient decision aid

improved decision making related to using immunosuppressive drugs in women with lupus

nephritis. Specifically, compared with viewing a standard ACR lupus information pamphlet,

women with lupus nephritis who viewed our computerized patient decision aid had a clinically

meaningful and statistically significant reduction in decisional conflict. Women with lupus

nephritis also had clinically meaningfully higher informed choice (41% versus 31%, a 10%

difference in proportions between arms) that was statistically nonsignificant (p = 0.08) in the

main analysis, and statistically significant and clinically meaningful in sensitivity analyses (50%

versus 35%; p = 0.006). There were no significant differences in the 2 secondary outcomes of

control preferences scale (concordance between desired role and actual role in decision

making) and IPC-SF, a patient-reported measure of patient–physician communication and care

processes. Patient-centered communication by the doctor/health care provider had a statistical

trend toward significance in the decision aid versus the pamphlet group in audiotaped

conversations with patients with current flare of lupus nephritis (p = 0.06). We also found that

the proportion of women with unresolved clinically significant conflict postintervention (score ≥

25) was statistically significantly lower in the decision aid group compared with the pamphlet

group—22.5% versus 44.2% (p < 0.001). When we examined coprimary outcomes by

race/ethnicity (subgroup analyses), we found that improvements in decisional conflict were

clinically meaningfully and statistically significantly higher in African Americans and Caucasian

women, and clinically meaningful with a nonsignificant statistical trend in Hispanic and

Asian/other women (small sample sizes; p = 0.075 and p = 0.053, respectively). Informed choice

did not show differences by race/ethnicity and was statistically not significant in this subgroup

analysis, likely related to small sample size.

52

Comparison With Other Published Studies

A literature search with terms “decision aid” and “lupus or SLE” on January 28, 2017,

revealed 103 titles in PubMed. We found no studies that tested a decision aid in patients with

lupus nephritis, but found one related study.129 The authors of this related study tested the

validity and reliability of a decision board for lupus nephritis in 172 Brazilian lupus patients, 75%

of whom had taken or were taking some immunosuppressive drug.129 The decision board

consisted of 5 parts, each presented during different times in the patient clinic visit, including

general information related to lupus and lupus nephritis, information on 2 therapies

(cyclophosphamide and mycophenolate mofetil), most common side effects, patient

prioritization of the 3 worst side effects, and probability of each of the 3 side effects with

treatment options. The decision board was valid and reliable. Most patients easily understood

the content of the decision board. Patients with a higher education level showed a better

understanding. Patients favored oral medications and were most worried about cancer, hair

loss, and infections.129 The authors set a goal to “develop a decision-aid that will help clinicians

communicate.” We were unable to find any subsequently published studies of the development

or testing of a decision aid.

In the absence of published studies of decision aids in patients with lupus nephritis, we

examined relevant indirect evidence regarding other interventions used for medication

adherence (a related concept to medication decision making that we studied) in patients with

lupus and similar decision aid studies in other rheumatic conditions. For example, in a

randomized study of 50 lupus patients in India that excluded patients with “advanced” lupus

nephritis (“advanced” not defined in the article), patients were randomized to 3 counseling

sessions by the pharmacist and given a handout developed to improve patient knowledge

versus a single physician-led session at the second follow-up visit (control group).130 Medical

knowledge (p < 0.001) and the adherence to medication score (only 18% on

immunosuppressive drugs; P value not provided) improved significantly more in the group that

received the pharmacist counseling sessions with a handout versus the control group, although

only means were provided without standard errors or standard deviations.130 Study limitations

were that the results were described inadequately, it was a single-center study, a low

53

proportion of patients were on immunosuppressives, and patients with advanced lupus

nephritis were excluded. A direct comparison to our study results is difficult, given the nature of

the intervention, study outcomes, country setting, and the type of patients. However, this study

provided some evidence that patient-centered education and counseling may improve lupus

outcomes.130

In another study, 41 patients with childhood-onset lupus were randomized in a 1:1 ratio

to receive text messages to improve their adherence to hydroxychloroquine. This study used

daily text messages before each patient’s scheduled time of medication intake to remind him or

her to take the medication (n = 19) or daily text messages to provide standard of care education

about hydroxychloroquine (n = 22).131 No significant difference in hydroxychloroquine

adherence was noted between groups (80% in each group). Our study was focused on

immunosuppressives, not hydroxychloroquine, and on medication decision making rather than

on medication adherence.

A recent systematic review of medication adherence in rheumatic conditions concluded

that interventions that were tailored to patients, delivered by the health care provider, and

directed at adherence were most likely to improve medication adherence outcomes.132 This

principle is consistent with the International Patient Decision Aid Standard (IPDAS) for the

development of effective decision aids,133 which we followed for the development of our lupus

nephritis decision aid. Tailoring to patients and using content that is easily understood by the

target population likely explains the higher efficacy of our decision aid compared with an

information pamphlet.

Implications of These Findings

We developed our decision aid intervention as a patient-focused intervention based on

IPDAS principles.133 We included benefit and harm data from a state-of-the-art NMA,69-71 and

based the content and messages of our lupus nephritis decision aid on the qualitative work with

patients with lupus nephritis similar to our target population.21-23 This information was

iteratively modified and finalized by constructive feedback from patients with lupus nephritis

who were similar to our target population. This study provides the proof that our individualized,

54

culturally tailored, computerized patient decision aid for medication decision making improved

decision making for immunosuppressive drugs in patients with lupus nephritis by reducing

decisional conflict. Patients were also more likely to make an informed choice using the

decision aid—41% versus 31%, a clinically meaningful difference that was statistically

nonsignificant (p = 0.08) for the main analysis. We used a 2-tailed analysis as a conservative

approach, but when viewed through the lens of a superiority perspective (as stated in our

original protocol78; also see the statistical analysis section above), a P value of 0.04 would be

the result. Thus, the effect for informed choice was alternatively statistically significant or

nonsignificant depending on whether a 1-sided or 2-sided test was used for the main analysis.

More important, the size of the effect (i.e., a 10% difference clinically in the proportion of

patients with favorable outcome is important simply because 10% more people are getting

help, and it is likely to impact clinical practice. Ultimately, the decision regarding the veracity of

the results will come with further review and perhaps more data. A statistically significantly

higher proportion of patients had informed choice in the sensitivity analysis, which was also

clinically meaningful (50% versus 35%; p = 0.006). Compared with the usual care group that

received the ACR lupus pamphlet, the decision aid group had higher knowledge scores

postintervention, which may be a mediator of the reduced decisional conflict.

To our knowledge, this is the first RCT of a patient-centered intervention for medication

decision making in lupus nephritis. Our finding that an individualized, computerized lupus

nephritis decision aid reduced decisional conflict related to immunosuppressive drugs and

possibly improved informed choice advances this field. We will make this decision aid available

in the public domain, as per instructions from the funder, PCORI. We plan to achieve this in

collaboration with our stakeholders, the Lupus Foundation of America and the Arthritis

Foundation.

The development of a lupus nephritis decision aid should help patients and providers

alike, since lupus nephritis is the most common manifestation of lupus,4,5 and has associated

morbidity of end-stage renal disease, in the United States.6 In addition, the rate of

immunosuppressive drug use by patients with lupus nephritis is low in the United States, at

34% in the Medicaid patient population.60 While we did not measure whether the use of a

55

decision aid can improve adherence to immunosuppressive drugs, this can be addressed with

future research and tested in clinical settings. The decision aid will need careful modification to

focus it on this related, but separate, domain of inquiry; medication adherence was not the

focus of our investigation.

Generalizability of the Findings

We conducted this study at 4 major academic centers (West Coast, Deep South,

Southwest, and Northeast United States) that provided health care to patients with lupus in

urban and suburban settings. Therefore, our findings may not be generalizable to patients

receiving care in rural settings. While we oversampled for minority women (as lupus is more

common in racial/ethnic minorities, who also have worse outcomes related to the disease), to

improve the generalizability of study results we included all women with lupus nephritis seeking

health care in urban and suburban settings. Additionally, to improve the generalizability of our

decision aid, as well as all study materials, we developed and tested it in both English and

Spanish. The low rate of refusal to participate in this study by eligible patients (i.e., 19 of the

320 patients), inclusion of patients from both outpatient and inpatient settings, inclusion of

geographically diverse sites, and conduct of the research study during a regular scheduled clinic

visit improve the generalizability of our study findings. We did not include in this study patients

with renal transplant, dialysis, or anticipated renal transplant; therefore, our findings are not

generalizable to these patients.

Implementation of Study Results

The decision aid can guide the choice of immunosuppressive medications by women with

lupus nephritis for shared decision making about immunosuppressive drugs for either induction

or maintenance therapy in consultation with their health care provider, in outpatient and

inpatient settings. Our decision aid addresses 4 scenarios when the following drug choices are

considered for the treatment of lupus nephritis:

1. Induction therapy: Mycophenolate mofetil versus cyclophosphamide

2. Maintenance therapy: Mycophenolate mofetil versus cyclophosphamide

56

3. Maintenance therapy: Calcineurin inhibitors such as cyclosporine/tacrolimus versus

cyclophosphamide

4. Maintenance therapy: Cyclophosphamide versus azathioprine

The ideal time to provide this information might be after the lupus diagnosis has been

briefly discussed with the health care provider with the patient during an outpatient visit or

inpatient consultation using a touch-pad computer. Once the patient has reviewed the

information, the patient can ask questions of the health care provider about the information

and make a treatment decision concordant with her values and preferences. Another potential

use of this decision aid is to have patients view the module on corticosteroid benefits and risks

once they start receiving corticosteroids as a concomitant therapy. Young premenopausal

women with lupus nephritis can view the sections on pregnancy and breast-feeding to

understand the risks and benefits of these medications and be guided properly regarding future

planning related to conception.

In order to keep patient burden low and based on the main focus of the study to

develop a decision aid and to assess its efficacy in a randomized trial, we did not include any

qualitative or quantitative evaluation of barriers or facilitators for future implementation of the

decision aid. However, patients frequently shared their positive and negative experiences about

the study, which are summarized below and provide information about future potential

facilitators and barriers to its implementation.

Positive experiences (ie, potential facilitators): Patients shared their positive experiences

about study participation. Most patients told the research team members that the entire

research study process (informed consent, randomization, responding to surveys, follow-ups)

was very easy. They found that the technology used in this research was user-friendly.

Participants also reported that they were motivated to ask questions of their physicians about

treatment options for lupus nephritis and in general about lupus care after viewing this

information, which they were able to do before seeing their provider. Following the conclusion

of the study, we have received requests from patients for copies of the decision aid tool, so that

they can refer to it when they have questions before or after their visits with their attending

57

physicians. Some quotes from patients who were either involved in the development of the

decision aid content and/or pilot testing are as follows:

• “I wish I had this information when I had to make a decision about my lupus. It was

really difficult.”

• “This tool shows me what I want to know.”

• “I can actually understand what this says; sometimes the doctors are trying to tell you

something and you lose them, and get scared.”

• “Every lupus patient should get this tool.”

Negative experiences (ie, potential barriers): A few patients had difficulty navigating the

decision aid, both with using an iPad independently and with the decision aid graphical

representation of harms and benefits. Some patients noticed that it took a long time for them

during the clinic visit to navigate the decision aid. Many patients wondered if having a paper

copy of the decision aid in addition to the iPad version would help them make the best use of

this information.

Subpopulation Considerations

The improvement/reduction in decisional conflict was smaller in the Hispanic/Latino

subgroup and was not statistically significant. The reduction in decisional conflict was

statistically significant and clinically meaningful in all other racial/ethnic subgroups, including

African American, Caucasian, and Asian women with lupus nephritis. This indicated that the

Hispanic/Latino subgroup may not benefit as much from the use of the decision aid as the other

racial/ethnic groups. The magnitude of reduction in decisional conflict was slightly higher in

Caucasian women.

Study Strengths and Limitations

Our study had several strengths. We developed our decision aid based on IPDAS

principles.133 We based our estimates of benefits and risks of immunosuppressive drugs on an

58

updated systematic review, meta-analysis, and NMA.69-71 The design and the content of our

decision aid were based on qualitative work with patients with lupus nephritis similar to our

target population.21-23 We oversampled African Americans and Hispanic women, since

compared with Caucasian women, minority women have a higher prevalence of lupus, higher

lupus severity, and worse outcomes related to lupus, including higher mortality.7,8,14,134 Our

study was a multicenter randomized trial that recruited patients from 4 geographically diverse

medical centers, making our study population representative of female patients with lupus

nephritis in the United States. We developed the decision aid in English and Spanish. Our

decision aid takes just a short time to be viewed and it will be available in the public domain.

This will make the decision aid a practical tool available for use by any adult woman with lupus.

The decision aid can also be further modified to be contextually relevant.

Our study has several limitations. We assessed 2 coprimary outcomes, since they

captured 2 complementary aspects of decision making, related to our individualized decision

aid intervention. We do not know which component of the decision aid was responsible for

efficacy—i.e., the reduction in decisional conflict. Since lupus is a female-predominant disease,

we excluded males; therefore, these study findings cannot be generalized to men. Another

limitation is that, while the decision aid and outcome instruments are available in English and

Spanish, we lack the translation of materials into other languages. Translation can be done in

the future and will make this decision aid even more accessible for lupus patients who speak

and read languages other than English and Spanish. In order to reduce patient burden, we did

not assess patient satisfaction and quality of life in this study. These effects need to be

examined in future studies. It will also be important to see if a similar decision aid will improve

medication adherence. Our study was not designed to assess coprimary outcomes after

patient–physician interaction, or to examine long-term medication adherence, both of which

could have provided additional insights into shared patient decision making and/or medication

adherence. Future studies should consider examining these outcomes. We examined the effect

of the decision aid on outcomes immediately after the visit during which each patient used the

decision aid or control pamphlet. Our study was not designed to assess whether the

information retention was short lived and whether the desired improvement in patients’

59

decisional conflict or informed choice about immunosuppressives remained stable over weeks

or months.

Future Research

In an experimental setting, we found a computerized, patient self-administered lupus

nephritis decision aid was more effective than a standard lupus pamphlet in reducing decisional

conflict and improving informed choice about immunosuppressive drugs. Women with lupus

nephritis who are making treatment decisions related to immunosuppressive drugs currently

and patients who may face such a decision in the future can use this decision aid. Given the

active participation by patients and stakeholders in the research process, including the

development and the testing of this decision aid, we are confident that this tool can now be

disseminated widely to patients with lupus nephritis. While the effectiveness of the decision aid

has been established, a wider dissemination will need further work.

More work is also needed to figure out ways to make this tool accessible to all low-

income, rural, undereducated, and minority patients and caregivers who may have limited or

no access and/or knowledge related to computers. However, 57% of our trial participants had

an annual income < $40 000, 85% were racial/ethnic minorities, and 36% had a high school

education or less, indicating that we succeeded in enrolling underrepresented, disadvantaged,

minority female patients in our trial. However, more work needs to be done to make this tool

accessible to every female lupus patient. Therefore, we plan to target future modifications to

accommodate the elderly and additional underprivileged patient populations. More qualitative

work with study participants, responders, and nonresponders could provide insights for further

improvements to the tool. Such work includes continuing and expanding collaborative efforts

with our partners, the Lupus Foundation of America and the Arthritis Foundation, to

disseminate the decision aid, along with the following: (1) performing implementation research

to understand how to incorporate use of the decision aid into a busy clinical practice workflow;

(2) figuring out how to use social media and patient networks to further disseminate the

decision aid; and (3) developing alternate designs of the decision aid including, but not limited

60

to, smartphone applications for iPhone and Windows platforms. These steps can help ensure

that our decision aid is widely used and proves effective for patients with lupus nephritis. The

goal is to implement it in many more clinics across the United States, including more variations

in the type of practice (private practice versus academic centers), specialty (rheumatology

versus combined rheumatology–nephrology clinics), and location type (urban versus suburban

versus rural). Research is needed to assess barriers to implementation and how to overcome

them in unique settings and to make the use of a decision aid part of standard care of lupus

patients. Research and work are also needed to translate the decision aid into many more

languages. In addition, more work is needed to assess the reasons for lower efficacy of the

decision aid for Hispanic/Latino populations, and ways to improve its efficacy in this subgroup

and possibly other underserved populations.

61

Conclusion

This multicenter randomized study tested an individualized, culturally tailored,

computerized patient decision aid that addressed treatment induction and the failure of

maintenance therapy in patients with a flare of lupus nephritis. The decision aid was superior to

the standard ACR lupus pamphlet for the primary outcome: reducing decisional conflict about

immunosuppressive drugs in women from diverse racial/ethnic and socioeconomic

backgrounds. The lupus nephritis decision aid had higher acceptability and was easier to

understand than the standard ACR lupus pamphlet for patients.

The major strengths of our study were a randomized design as well as the negligible

number of patients without outcome data. The shortcoming was the lack of evidence about

how long the effect of the decision aid persisted. Based on these findings, we conclude that the

decision aid is ready for testing in additional settings and with further follow-up to measure the

persistence of the knowledge gained. This tool, available in English and Spanish, can facilitate

and improve shared decision making for lupus nephritis treatments in clinical practice and

should lead to higher patient satisfaction and engagement. Whether it might improve disease

outcomes remains to be seen.

62

Acknowledgments

We thank Jeffrey Foster, Mazin Khalil, Candace Green, and Diana Florence for proofreading this

report.

63

References

1. Weckerle CE, Niewold TB. The unexplained female predominance of systemic lupus erythematosus: clues from genetic and cytokine studies. Clin Rev Allergy Immunol. 2011;40(1):42-49. doi:10.1007/s12016-009-8192-4.

2. Lopez P, Mozo L, Gutierrez C, Suarez A. Epidemiology of systemic lupus erythematosus in a northern Spanish population: gender and age influence on immunological features. Lupus. 2003;12(11):860-865. doi:10.1191/0961203303lu469xx.

3. Helmick CG, Felson DT, Lawrence RC, et al. Estimates of the prevalence of arthritis and other rheumatic conditions in the United States. Part I. Arthritis Rheum. 2008;58(1):15-25. doi:10.1002/art.23177.

4. Alarcon GS, McGwin G Jr, Petri M, et al. Baseline characteristics of a multiethnic lupus cohort: PROFILE. Lupus. 2002;11(2):95-101. doi:10.1191/0961203302lu155oa.

5. Kasitanon N, Magder LS, Petri M. Predictors of survival in systemic lupus erythematosus. Medicine. 2006;85(3):147-156. doi:10.1097/01.md.0000224709.70133.f7.

6. Maisonneuve P, Agodoa L, Gellert R, et al. Distribution of primary renal diseases leading to end-stage renal failure in the United States, Europe, and Australia/New Zealand: results from an international comparative study. Am J Kidney Dis. 2000;35(1):157-165. doi:10.1016/S0272-6386(00)70316-7.

7. Odutola J, Ward MM. Ethnic and socioeconomic disparities in health among patients with rheumatic disease. Curr Opin Rheumatol. 2005;17(2):147-152. PubMed

8. Alarcon GS, Friedman AW, Straaton KV, et al. Systemic lupus erythematosus in three ethnic groups: III. a comparison of characteristics early in the natural history of the LUMINA cohort. LUpus in MInority populations: NAture vs. Nurture. Lupus. 1999;8(3):197-209.

9. Dooley MA, Aranow C, Ginzler EM. Review of ACR renal criteria in systemic lupus erythematosus. Lupus. 2004;13(11):857-860. PMID:15580982.

10. Ward MM, Pyun E, Studenski S. Mortality risks associated with specific clinical manifestations of systemic lupus erythematosus. Arch Intern Med. 1996;156(12):1337-1344.

11. Costenbader KH, Desai A, Alarcon GS, et al. Trends in the incidence, demographics, and outcomes of end-stage renal disease due to lupus nephritis in the US from 1995 to 2006. Arthritis Rheum. 2011;63(6):1681-1688. doi:10.1002/art.30293. PubMed PMID: 21445962;

64

12. Bernatsky S, Boivin JF, Joseph L, et al. Race/ethnicity and cancer occurrence in systemic lupus erythematosus. Arthritis Rheum. 2005;53(5):781-784. doi:10.1002/art.21458.

13. Cervera R, Khamashta MA, Font J, et al. Morbidity and mortality in systemic lupus erythematosus during a 10-year period: a comparison of early and late manifestations in a cohort of 1,000 patients. Medicine (Baltimore). 2003;82(5):299-308. doi:10.1097/01.md.0000091181.93122.55.

14. Krishnan E, Hubert HB. Ethnicity and mortality from systemic lupus erythematosus in the US. Ann Rheum Dis. 2006;65(11):1500-1505. doi:10.1136/ard.2005.040907.

15. Barr RG, Seliger S, Appel GB, et al. Prognosis in proliferative lupus nephritis: the role of socio-economic status and race/ethnicity. Nephrol Dial Transplant. 2003;18(10):2039-2046. doi:10.1093/ndt/gfg345.

16. Hahn BH, McMahon MA, Wilkinson A, et al. American College of Rheumatology guidelines for screening, treatment, and management of lupus nephritis. Arthritis Care Res (Hoboken). 2012;64(6):797-808. doi:10.1002/acr.21664. PMID:22556106.

17. van Staa TP, Leufkens HG, Abenhaim L, Zhang B, Cooper C. Oral corticosteroids and fracture risk: relationship to daily and cumulative doses. Rheumatology (Oxford). 2000;39(12):1383-1389.

18. Walsh LJ, Wong CA, Oborne J, et al. Adverse effects of oral corticosteroids in relation to dose in patients with lung disease. Thorax. 2001;56(4):279-284.

19. Zonana-Nacach A, Barr SG, Magder LS, Petri M. Damage in systemic lupus erythematosus and its association with corticosteroids. Arthritis Rheum. 2000;43(8):1801-1808. doi:10.1002/1529-0131(200008)43:8<1801::AID-ANR16>3.0.CO;2-O.

20. Houssiau FA, D'Cruz D, Sangle S, et al. Azathioprine versus mycophenolate mofetil for long-term immunosuppression in lupus nephritis: results from the MAINTAIN Nephritis Trial. Annals Rheumatic Dis. 2010;69(12):2083-2089. doi:10.1136/ard.2010.131995.

21. Qu H, Shewchuk RM, Alarcon G, et al. Mapping perceptions of lupus medication decision-making facilitators: the importance of patient context. Arthritis Care Res (Hoboken). 2016;68(12):1787-1794. doi:10.1002/acr.22904.

22. Singh JA, Qu H, Yazdany J, Chatham W, Shewchuk R. Minorities with lupus nephritis and medications: a study of facilitators to medication decision-making. Arthritis Res Ther. 2015;(17):367. doi:10.1186/s13075-015-0883-z.

23. Singh JA, Qu H, Yazdany J, Chatham W, Dall'era M, Shewchuk RM. Barriers to medication decision making in women with lupus nephritis: a formative study using nominal group technique. J Rheumatol. 2015. doi:10.3899/jrheum.150168.

65

24. Singh JA, Hossain A, Kotb A, Wells GA. Comparative effectiveness of immunosuppressive drugs and corticosteroids for lupus nephritis: a systematic review and network meta-analysis. Syst Rev. 2016;5(1):155. doi:10.1186/s13643-016-0328-z.

25. Singh JA, Hossain A, Kotb A, Wells G. Risk of serious infections with immunosuppressive drugs and glucocorticoids for lupus nephritis: a systematic review and network meta-analysis. BMC Med. 2016;14(1):137. doi:10.1186/s12916-016-0673-8.

26. Singh JA, Hossain A, Kotb A, et al. Treatments for lupus nephritis: a systematic review and network metaanalysis. J Rheumatol. 2016;43(10):1801-1815. doi:10.3899/jrheum.160041.

27. Appel GB, Contreras G, Dooley MA, et al. Mycophenolate mofetil versus cyclophosphamide for induction treatment of lupus nephritis. J Am Soc Nephrol. 2009;20(5):1103-1112. doi:10.1681/ASN.2008101028. PMID:19369404.

28. Dooley MA, Jayne D, Ginzler EM, et al. Mycophenolate versus azathioprine as maintenance therapy for lupus nephritis. N Engl Journal Med. 2011;365(20):1886-1895. doi:10.1056/NEJMoa1014460.

29. Langford AT, Resnicow K, Roberts JS, Zikmund-Fisher BJ. Racial and ethnic differences in direct-to-consumer genetic tests awareness in HINTS 2007: sociodemographic and numeracy correlates. J Genet Couns. 2012;21(3):440-447. doi:10.1007/s10897-011-9478-2.

30. Cene CW, Roter D, Carson KA, Miller ER III, Cooper LA. The effect of patient race and blood pressure control on patient-physician communication. J Gen Intern Med. 2009;24(9):1057-1064. doi:10.1007/s11606-009-1051-4.

31. Gordon HS, Street RL Jr, Kelly PA, Souchek J, Wray NP. Physician-patient communication following invasive procedures: an analysis of post-angiogram consultations. Soc Sci Med. 2005;61(5):1015-1025. doi:10.1016/j.socscimed.2004.12.021.

32. Johnson RL, Roter D, Powe NR, Cooper LA. Patient race/ethnicity and quality of patient-physician communication during medical visits. Am J Public Health. 2004;94(12):2084-2090.

33. Manfredi C, Kaiser K, Matthews AK, Johnson TP. Are racial differences in patient-physician cancer communication and information explained by background, predisposing, and enabling factors? J Health Commun. 2010;15(3):272-292. doi:10.1080/10810731003686598.

34. Jeon-Slaughter H. Economic factors in of patients' nonadherence to antidepressant treatment. Soc Psychiatry Psychiatr Epidemiol. 2012. doi:10.1007/s00127-012-0497-6.

66

35. Cox ED, Nackers KA, Young HN, Moreno MA, Levy JF, Mangione-Smith RM. Influence of race and socioeconomic status on engagement in pediatric primary care. Patient Educ Couns. 2012;87(3):319-326. doi:10.1016/j.pec.2011.09.012.

36. Yao L, Robert SA. Examining the racial crossover in mortality between African American and white older adults: a multilevel survival analysis of race, individual socioeconomic status, and neighborhood socioeconomic context. J Aging Res. 2011;2011:132073 doi:10.4061/2011/132073.

37. Chu DI, Moreira DM, Gerber L, et al. Effect of race and socioeconomic status on surgical margins and biochemical outcomes in an equal-access health care setting: results from the Shared Equal Access Regional Cancer Hospital (SEARCH) database. Cancer. 2012. doi:10.1002/cncr.27456.

38. Trinh QD, Sun M, Sammon J, et al. Disparities in access to care at high-volume institutions for uro-oncologic procedures. Cancer. 2012. doi:10.1002/cncr.27440.

39. FitzGerald JD, Soohoo NF, Losina E, Katz JN. Potential impact on patient residence to hospital travel distance and access to care under a policy of preferential referral to high-volume knee replacement hospitals. Arthritis Care Res (Hoboken). 2012;64(6):890-897. doi:10.1002/acr.21611.

40. Byrne MM, Souchek J, Richardson M, Suarez-Almazor M. Racial/ethnic differences in preferences for total knee replacement surgery. J Clin Epidemiol. 2006;59(10):1078-1086. doi:S0895-4356(06)00117-X [pii] 10.1016/j.jclinepi.2006.01.010.

41. Rosen AB, Tsai JS, Downs SM. Variations in risk attitude across race, gender, and

education. Med Decis Making. 2003;23(6):511-517. doi:10.1177/0272989X03258431. PMID:14672111.

42. Ibrahim SA, Siminoff LA, Burant CJ, Kwoh CK. Differences in expectations of outcome mediate African American/white patient differences in "willingness" to consider joint replacement. Arthritis Rheum. 2002;46(9):2429-2435. doi:10.1002/art.10494.

43. Hofmann JC, Wenger NS, Davis RB, et al. Patient preferences for communication with physicians about end-of-life decisions. SUPPORT investigators. study to understand prognoses and preference for outcomes and risks of treatment. Ann Intern Med. 1997;127(1):1-12.

44. Whittle J, Conigliaro J, Good CB, Lofgren RP. Racial differences in the use of invasive cardiovascular procedures in the Department of Veterans Affairs medical system. N Engl J Med. 1993;329(9):621-627. doi:10.1056/NEJM199308263290907.

67

45. Constantinescu F, Goucher S, Weinstein A, Fraenkel L. Racial disparities in treatment preferences for rheumatoid arthritis. Med Care. 2009;47(3):350-355. doi:10.1097/MLR.0b013e31818af829.

46. Institute of Medicine. Health Literacy: A Prescription to End Confusion. Washington, DC: National Academies Press; 2004.

47. Rothman RL, Montori VM, Cherrington A, Pignone MP. Perspective: the role of numeracy in health care. J Health Commun. 2008;13(6):583-595. doi:10.1080/10810730802281791.

48. Peters E, Hibbard J, Slovic P, Dieckmann N. Numeracy skill and the communication, comprehension, and use of risk-benefit information. Health Aff (Millwood). 2007;26(3):741-748. doi:10.1377/hlthaff.26.3.741.

49. Osborn CY, Cavanaugh K, Wallston KA, White RO, Rothman RL. Diabetes numeracy: an overlooked factor in understanding racial disparities in glycemic control. Diabetes Care. 2009;32(9):1614-1619. doi:10.2337/dc09-0425.

50. Waldrop-Valverde D, Osborn CY, Rodriguez A, Rothman RL, Kumar M, Jones DL. Numeracy skills explain racial differences in HIV medication management. AIDS Behav. 2010;14(4):799-806. doi:10.1007/s10461-009-9604-4. PMCID:2891293.

51. Julian LJ, Yelin E, Yazdany J, et al. Depression, medication adherence, and service utilization in systemic lupus erythematosus. Arthritis Rheum. 2009;61(2):240-246. doi:10.1002/art.24236.

52. 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(3):329-340. doi:10.1016/j.berh.2013.07.001.

53. Vina ER, Masi CM, Green SL, Utset TO. A study of racial/ethnic differences in treatment preferences among lupus patients. Rheumatology (Oxford). 2012;51(9):1697-1706. doi:10.1093/rheumatology/kes128.

54. Stacey D, Bennett CL, Barry MJ, et al. Decision aids for people facing health treatment or screening decisions. Cochrane Database Syst Rev. 2011(10):CD001431. doi:10.1002/14651858.CD001431.pub3.

55. Petrella RJ, Petrella M. A prospective, randomized, double-blind, placebo controlled study to evaluate the efficacy of intraarticular hyaluronic acid for osteoarthritis of the knee. J Rheumatol. 2006;33(5):951-956. doi:0315162X-33-951 [pii].

56. Greenfield S, Kaplan S, Ware JE Jr. Expanding patient involvement in care. effects on patient outcomes. Ann Intern Med. 1985;102(4):520-528.

68

57. Gattellari M, Butow PN, Tattersall MH. Sharing decisions in cancer care. Soc Sci Med. 2001;52(12):1865-1878. doi:S0277953600003038 [pii].

58. Keating NL, Guadagnoli E, Landrum MB, Borbas C, Weeks JC. Treatment decision making in early-stage breast cancer: should surgeons match patients' desired level of involvement? J Clin Oncol. 2002;20(6):1473-1479. doi:10.1200/jco.2002.20.6.1473.

59. Davidson JR, Brundage MD, Feldman-Stewart D. Lung cancer treatment decisions: patients' desires for participation and information. Psychooncology. 1999;8(6):511-520. doi:10.1002/(SICI)1099-1611(199911/12)8:6<511::AID-PON415>3.0.CO;2-T [pii].

60. Yazdany J, Feldman CH, Liu J, Ward MM, Fischer MA, Costenbader KH. Quality of care for incident lupus nephritis among medicaid beneficiaries in the United States. Arthritis Care Res (Hoboken). 2014;66(4):617-624. doi:10.1002/acr.22182.

61. Montori VM, Shah ND, Pencille LJ, et al. Use of a decision aid to improve treatment decisions in osteoporosis: the osteoporosis choice randomized trial. Am J Med. 2011;124(6):549-556. doi:10.1016/j.amjmed.2011.01.013.

62. Mullan RJ, Montori VM, Shah ND, et al. The diabetes mellitus medication choice decision aid: a randomized trial. Arch Intern Med. 2009;169(17):1560-1568. doi:10.1001/archinternmed.2009.293.

63. de Achaval S, Fraenkel L, Volk RJ, Cox V, Suarez-Almazor ME. Impact of educational and patient decision aids on decisional conflict associated with total knee arthroplasty. Arthritis Care Res (Hoboken). 2012;64(2):229-237. doi:10.1002/acr.20646.

64. Fraenkel L, Street RL Jr, Fried TR. Development of a tool to improve the quality of decision making in atrial fibrillation. BMC Med Inform Decis Mak. 2011;(11):59. doi:10.1186/1472-6947-11-59.

65. Berman L, Curry L, Goldberg C, Gusberg R, Fraenkel L. Pilot testing of a decision support tool for patients with abdominal aortic aneurysms. J Vasc Surg. 2011;53(2):285-292 doi:10.1016/j.jvs.2010.08.075.

66. Fraenkel L, Chodkowski D, Lim J, Garcia-Tsao G. Patients' preferences for treatment of hepatitis C. Med Decis Making. 2010;30(1):45-57. doi:10.1177/0272989X09341588. PMID:19636065.

67. Nota I, Drossaert CHC, Melissant HC, et al. Development of a web-based patient decision aid for initiating disease modifying anti-rheumatic drugs using user-centred design methods. BMC Med Inform Decis Mak. 2017;17(1):51. doi:10.1186/s12911-017-0433-5.

68. Meade T, Dowswell E, Manolios N, Sharpe L. The motherhood choices decision aid for women with rheumatoid arthritis increases knowledge and reduces decisional conflict: a

69

randomized controlled trial. BMC Musculoskelet Disord. 2015;(16):260. doi:10.1186/s12891-015-0713-0.

69. Singh JA, Christensen R, Wells GA, et al. A network meta-analysis of randomized controlled trials of biologics for rheumatoid arthritis: a Cochrane overview. CMAJ. 2009;181(11):787-796. doi:10.1503/cmaj.091391.

70. Singh JA, Christensen R, Wells GA, et al. Biologics for rheumatoid arthritis: an overview of Cochrane reviews. Cochrane Database Syst Rev. 2009(4):CD007848. doi:10.1002/14651858.CD007848.pub2.

71. Singh JA, Wells GA, Christensen R, et al. Adverse effects of biologics: a network meta-analysis and Cochrane overview. Cochrane Database Syst Rev. 2011;(2):CD008794. doi:10.1002/14651858.CD008794.pub2.

72. Methods guide for effectiveness and comparative effectiveness reviews. Agency for Healthcare Research and Quality website. http://effectivehealthcare.ahrq.gov/index.cfm/search-for-guides-reviews-and-reports/?pageaction=displayProduct&productID=318. . Published 2012. Accessed July 25, 2012.

73. The Cochrane Collaboration. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0. The Cochrane Collaboration; 2011.www.cochrane-handbook.org. Accessed July 25, 2012.

74. Deeks J, Higgins J, Altman D. Chapter 9: Analysing data and undertaking meta-analyses. In: Higgins J, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.0.0. The Cochrane Collaboration; 2008. www.cochrane-handbook.org. Accessed July 25, 2012.

75. Lumley T. Network meta-analysis for indirect treatment comparisons. Stat Med. 2002;21(16):2313-2324. doi:10.1002/sim.1201. PMID:12210616.

76. Psaty BM, Lumley T, Furberg CD, et al. Health outcomes associated with various antihypertensive therapies used as first-line agents: a network meta-analysis. JAMA. 2003;289(19):2534-2544. doi:10.1001/jama.289.19.2534. PMID:12759325.

77. Lu G, Ades AE. Combination of direct and indirect evidence in mixed treatment comparisons. Stat Med. 2004;23(20):3105-3124. doi:10.1002/sim.1875.

78. Singh JA, Shah N, Green C. Individualized patient decision-aid for immunosuppressive drugs in women with lupus nephritis: study protocol of a randomized, controlled trial. BMC Musculoskelet Disord. 2017;18(1):53. doi:10.1186/s12891-017-1408-5.

70

79. Singh JA, Qu H, Yazdany J, Chatham W, Dall'era M, Shewchuk RM. Barriers to medication decision making in women with lupus mephritis: a formative study using nominal group technique. J Rheumatol. 2015;42(9):1616-1623. doi:10.3899/jrheum.150168.

80. Asmus CL, James K. Nominal group technique, social loafing, and group creative project quality. Creat Res J. 2005;17(4):349-354.

81. Ruyter KD. Focus versus nominal group interviews: a comparative analysis. Marketing Intelligence & Planning. 1996;14(6):44-50.

82. Delbecq AL, Van de Ven AH, Gustafson DH. Group Techniques for Program Planning. Glenview, IL: Scott, Foresman, and Co.; 1975.

83. In: Higgins JP, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0. The Cochrane Collaboration; 2011 . http://handbook.cochrane.org..

84. Singh JA, Noorbaloochi S, Tucker MD. Belimumab for systemic lupus erythematosus (protocol). Cochrane Database Syst Rev. 2013. doi:10.1002/14651858.CD010668.

85. Henderson L, Masson P, Craig JC, et al. Treatment for lupus nephritis. Cochrane Database Syst Rev. 2012;(12):CD002922. doi:10.1002/14651858.CD002922.pub3. PMID:23235592.

86. Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 1997;40(9):1725. doi:10.1002/1529-0131(199709)40:9<1725::AID-ART29>3.0.CO;2-Y.

87. El-Shafey EM, Abdou SH, Shareef MM. Is mycophenolate mofetil superior to pulse intravenous cyclophosphamide for induction therapy of proliferative lupus nephritis in Egyptian patients? Clin Exp Nephrol. 2010;14(3):214-221.

88. Mitwalli AH, Al Wakeel JS, Hurraib S, et al. Comparison of high and low dose of cyclophosphamide in lupus nephritis patients: a long-term randomized controlled trial. Saudi J Kidney Dis Transpl. 2011;22(5):935-940.

89. Houssiau FA, D'Cruz D, Sangle S, et al. Azathioprine versus mycophenolate mofetil for long-term immunosuppression in lupus nephritis: results from the MAINTAIN Nephritis Trial. Ann Rheum Dis. 2010;69(12):2083-2089.

90. Higgins JP, Altman DG, Sterne JA. Chapter 8: Assessing risk of bias in included studies. In: Higgins JP, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0. The Cochrane Collaboration; 2011. www.cochrane-handbook.org.

91. Dias S, Sutton A, Welton N, et al. NICE DSU Technical Support Document 2: Generalised Linear Modelling Framework for Pairwise and Network Meta-Analysis of Randomised

http://handbook.cochrane.org/

71

Controlled Trials. May 2011:1-96. http://www.nicedsu.org.uk/TSD6 Software.final.08.05.12.pdf.

92. Dias S, Sutton A, Welton N, et al. NICE DSU Technical Support Document 3: Heterogeneity: Subgroups, meta-regression, bias and bias-adjustment. May 2011:1-24. http://www.nicedsu.org.uk/TSD6 Software.final.08.05.12.pdf.

93. Cooper NJ, Sutton AJ, Morris D, Ades AE, Welton NJ. Addressing between-study heterogeneity and inconsistency in mixed treatment comparisons: application to stroke prevention treatments in individuals with non-rheumatic atrial fibrillation. Stat Med. 2009;28(14):1861-1881.

94. Turner RM, Thompson SG, Spiegelhalter DJ. Prior distributions for the intracluster correlation coefficient, based on multiple previous estimates, and their application in cluster randomized trials. Clin Trials. 2005;2(2):108-118. PMID:16279132.

95. Spiegelhalter D, Thomas A, Best N, Lunn D. WinBUGS User Manual. Version 1.4. 2003. http://www.mrcbsu.cam.ac.uk/bugs/winbugs/manual114.pdf.

96. Ades AE, Welton NJ, Caldwell D, Price M, Goubar A, Lu G. Multiparameter evidence synthesis in epidemiology and medical decision-making. J Health Serv Res Policy. 2008;13(suppl 3):12-22. doi:10.1258/jhsrp.2008.008020.

97. Dias S, Sutton A, Welton N, et al. NICE DSU Technical Support Document 4: Inconsistency in Networks of Evidence Based on Randomised Controlled Trials. May 2011:1-39. http://www.nicedsu.org.uk/TSD6 Software.final.08.05.12.pdf.

98. Puhan MA, Schunemann HJ, Murad MH, et al. A GRADE working group approach for rating the quality of treatment effect estimates from network meta-analysis. BMJ. 2014;(349):g5630. doi:10.1136/bmj.g5630.

99. Rheumatology ACo website. http://www.rheumatology.org/I-Am-A/Patient-Caregiver/Diseases-Conditions/Lupus. Published 2016.

100. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377-381. doi:10.1016/j.jbi.2008.08.010.

101. Arozullah AM, Yarnold PR, Bennett CL, et al. Development and validation of a short-form, rapid estimate of adult literacy in medicine. Med Care. 2007;45(11):1026-1033. doi:10.1097/MLR.0b013e3180616c1b.

102. Lee SY, Stucky BD, Lee JY, Rozier RG, Bender DE. Short assessment of health literacy-Spanish and English: a comparable test of health literacy for Spanish and English

http://www.nicedsu.org.uk/TSD6%20Software.final.08.05.12.pdf




http://www.rheumatology.org/I-Am-A/Patient-Caregiver/Diseases-Conditions/Lupus

http://www.rheumatology.org/I-Am-A/Patient-Caregiver/Diseases-Conditions/Lupus

72

speakers. Health Serv Res. 2010;45(4):1105-1120. doi:10.1111/j.1475-6773.2010.01119.x.

103. O'Connor AM. Validation of a decisional conflict scale. Med Decis Making. 1995;15(1):25-30.

104. Marteau TM, Dormandy E, Michie S. A measure of informed choice. Health Expect. 2001;4(2):99-108. PMID:11359540.

105. Michie S, Dormandy E, Marteau TM. The multi-dimensional measure of informed choice: a validation study. Patient Educ Couns. 2002;48(1):87-91. PMID:12220754.

106. Degner LF, Russell CA. Preferences for treatment control among adults with cancer. Res Nurs Health. 1988;11(6):367-374. PMID:3231738.

107. Degner LF, Kristjanson LJ, Bowman D, et al. Information needs and decisional preferences in women with breast cancer. JAMA. 1997;277(18):1485-1492. PMID:9145723.

108. Adams JR, Drake RE, Wolford GL. Shared decision-making preferences of people with severe mental illness. Psychiatr Serv. 2007;58(9):1219-1221. doi:58/9/1219 [pii] 10.1176/appi.ps.58.9.1219. PMID:17766569.

109. Stewart AL, Napoles-Springer AM, Gregorich SE, Santoyo-Olsson J. Interpersonal

processes of care survey: patient-reported measures for diverse groups. Health Serv Res. 2007;42(3, pt 1):1235-1256. doi:10.1111/j.1475-6773.2006.00637.x. PMID:17489912. PMCID:1955252.

110. O'Connor AM. User manual – Decisional Conflict Scale (10 item question format) . http://decisionaid.ohri.ca/docs/develop/User_Manuals/UM_Decisional_Conflict.pdf. Updated 2010. Accessed March 17, 2013.

111. Siminoff LA, Ravdin P, Colabianchi N, Sturm CM. Doctor-patient communication patterns in breast cancer adjuvant therapy discussions. Health Expect. 2000;3(1):26-36. PMID:11281909.

112. O'Connor AM. User manual – Decisional Conflict Scale. http://decisionaid.ohri.ca/docs/develop/User_Manuals/UM_Decisional_Conflict.pdf 2003. Updated 2010. Accessed December 25, 2016, 2016.

113. Degner LF, Sloan JA, Venkatesh P. The Control Preferences Scale. Can J Nurs Res. 1997;29(3):21-43.

114. Florin J, Ehrenberg A, Ehnfors M. Clinical decision-making: predictors of patient participation in nursing care. J Clin Nurs. 2008;17(21):2935-2944.

http://decisionaid.ohri.ca/docs/develop/User_Manuals/UM_Decisional_Conflict.pdf

73

115. Singh JA, Sloan JA, Atherton PJ, et al. Preferred roles in treatment decision making among patients with cancer: a pooled analysis of studies using the Control Preferences Scale. Am J Manag Care. 2010;16(9):688-696. doi:12718 [pii].

116. Stewart AL, Napoles-Springer A, Perez-Stable EJ. Interpersonal processes of care in diverse populations. Milbank Q. 1999;77(3):305-339, 274.

117. Schillinger D, Handley M, Wang F, Hammer H. Effects of self-management support on structure, process, and outcomes among vulnerable patients with diabetes: a three-arm practical clinical trial. Diabetes Care. 2009;32(4):559-566. doi:10.2337/dc08-0787.

118. Napoles AM, Gregorich SE, Santoyo-Olsson J, O'Brien H, Stewart AL. Interpersonal processes of care and patient satisfaction: do associations differ by race, ethnicity, and language? Health Serv Res. 2009;44(4):1326-1344. doi:10.1111/j.1475-6773.2009.00965.x.

119. Schenker Y, Stewart A, Na B, Whooley MA. Depressive symptoms and perceived doctor-patient communication in the Heart and Soul study. J Gen Intern Med. 2009;24(5):550-556. doi:10.1007/s11606-009-0937-5.

120. Street RL Jr, Millay B. Analyzing patient participation in medical encounters. Health Commun. 2001;13(1):61-73. doi:10.1207/S15327027HC1301_06. PMID:11370924.

121. Cegala DJ, Street RL Jr, Clinch CR. The impact of patient participation on physicians' information provision during a primary care medical interview. Health Commun. 2007;21(2):177-185. doi:10.1080/10410230701307824.

122. Kravitz RL, Epstein RM, Feldman MD, et al. Influence of patients' requests for direct-to-consumer advertised antidepressants: a randomized controlled trial. JAMA. 2005;293(16):1995-2002. doi:10.1001/jama.293.16.1995.

123. Street RL Jr, Gordon H, Haidet P. Physicians' communication and perceptions of patients: is it how they look, how they talk, or is it just the doctor? Soc Sci Med. 2007;65(3):586-598. doi:10.1016/j.socscimed.2007.03.036. PMID:17462801. PMCID:2811428.

124. Street RL Jr, Richardson MN, Cox V, Suarez-Almazor ME. (Mis)understanding in patient-health care provider communication about total knee replacement. Arthritis Rheum. 2009;61(1):100-107. doi:10.1002/art.24371.

125. Mudano AS, Gary LC, Oliveira AL, et al. Using tablet computers compared to interactive voice response to improve subject recruitment in osteoporosis pragmatic clinical trials: feasibility, satisfaction, and sample size. Patient Prefer Adherence. 2013;(7):517-523. doi:10.2147/PPA.S44551.

126. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Erlbaum; 1988.

74

127. Cohan D, Gomez E, Greenberg M, Washington S, Charlebois ED. Patient perspectives with abbreviated versus standard pre-test HIV counseling in the prenatal setting: a randomized-controlled, non-inferiority trial. PLoS One. 2009;4(4):e5166. doi:10.1371/journal.pone.0005166.

128. Smith SK, Trevena L, Simpson JM, Barratt A, Nutbeam D, McCaffery KJ. A decision aid to support informed choices about bowel cancer screening among adults with low education: randomised controlled trial. BMJ. 2010;341:c5370. doi:10.1136/bmj.c5370.

129. De Abreu MM, Gafni A, Ferraz MB. Development and testing of a decision board to help clinicians present treatment options to lupus nephritis patients in Brazil. Arthritis Rheum. 2009;61(1):37-45. doi:10.1002/art.24368.

130. 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):7-12.

131. 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(1):174-179. doi:10.3899/jrheum.110771. PMID:22089460.

132. Galo JS, Mehat P, Rai SK, Avina-Zubieta A, De Vera MA. What are the effects of medication adherence interventions in rheumatic diseases: a systematic review. Ann Rheum Dis. 2016;75(4):667-673. doi:10.1136/annrheumdis-2014-206593.

133. International Patient Decision Aid Standard (IPDAS) Collaboration website. http://www.ipdas.ohri.ca/. Published 2013. Accessed January 28, 2017.

134. Centers for Disease Control and Prevention. Trends in deaths from systemic lupus erythematosus--United States, 1979-1998. MMWR Morb Mortal Wkly Rep. 2002;51(17):371-374.

http://www.ipdas.ohri.ca/

75

Copyright© 2019. University of Alabama at Birmingham. All Rights Reserved.

Disclaimer:

The [views, statements, opinions] presented in this report are solely the responsibility of the author(s) and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute® (PCORI®), its Board of Governors or Methodology Committee.

Acknowledgement:

Research reported in this report was [partially] funded through a Patient-Centered Outcomes Research Institute® (PCORI®) Award (#CE-1304-6631) Further information available at: https://www.pcori.org/research-results/2013/personalized-decision-aid-help-women-lupus-nephritis-racially-and-ethnically

Documents

A Personalized Decision Aid to Help Women with Lupus Nephritis … · 2020. 5. 14. · 3 Abstract Background/Objective: Systemic lupus erythematosus (SLE) is a rare and sometimes