Clinical integration of patient-reported outcome measures to enhance the care of patients with SLE: a multi-centre prospective cohort study

Abstract

Objective

To assess the feasibility and impact of integrating electronic patient-reported outcome measures (PROMs) into the routine outpatient care of patients with SLE.

Methods

We conducted a prospective cohort study, utilizing a mixed-methods sequential explanatory design, of SLE outpatients receiving rheumatology care at two academic medical centres. Participants completed electronic PROMs at enrolment and then prior to their next two routine rheumatology visits. PROM score reports were shared with patients and rheumatologists before visits. Patients and rheumatologists completed post-visit surveys evaluating the utility of PROMs in the clinical encounters. Focus groups of patients and interviews with treating rheumatologists were conducted to further explore their experience utilizing PROMs.

Results

A total of 105 SLE patients and 17 rheumatologists participated in the study. Patients completed PROMs in 159 of 184 encounters (86%), with 93% of surveys completed remotely. Patients reported that PROMs were ‘quite a bit’ or ‘very’ useful (55% of encounters) and beneficial to communication (55% of encounters). In contrast, physicians found PROMs useful (20%) and beneficial to communication (17%) less frequently. There was no significant change in visit length, health-related quality of life or disease activity after implementation of PROMs; however, patient satisfaction improved slightly. Qualitative analyses revealed that patients felt PROMs provided utility primarily by facilitating communication, particularly when physicians discussed the surveys.

Conclusion

The remote capture and integration of electronic PROMs into clinical care was feasible in a diverse cohort of SLE outpatients. PROMs were useful to patients and enhanced their clinical experience primarily by facilitating communication.

Rheumatology key messages.

• The clinical integration of electronic PROMs was feasible and well-received by SLE patients.

• Patients found PROMs facilitated communication by highlighting their perspective and bolstering therapeutic relationships.

• Patients valued PROMs more than their rheumatologists, highlighting differing perspectives on what constitutes patient-centred communication.

Introduction

Advances in therapy have substantially reduced mortality from SLE, but morbidity from this chronic illness remains high. Individuals with SLE suffer from impaired function and diminished health-related quality of life (HRQOL), even compared with those with other chronic diseases [1, 2]. HRQOL and subjective symptoms, including fatigue, anxiety and cognitive impairment, are central to patients’ experience of SLE and assessment of disease activity, but are often not prioritized in clinical visits [3, 4]. Physicians instead emphasize objective measures more directly correlated with disease activity, such as joint or skin manifestations or laboratory abnormalities. This disconnect can leave patients feeling that their concerns are unheard and their needs are unmet [5, 6]. The discordance between the perceptions and priorities of SLE patients and physicians is well described and can undermine the therapeutic alliance, compromise treatment adherence, reduce patient satisfaction, and ultimately affect patient outcomes [7–9].

Patient-reported outcomes measures (PROMs) have the potential to bridge this disconnect. PROMs are tools measuring the status of a patient’s health condition directly from the patient, without interpretation of the patient’s response by a clinician [10]. PROMs can elucidate the patient perspective by providing standardized assessments of symptoms, function and HRQOL, domains that complement the objective measures used by physicians to evaluate SLE activity. PROMs are well-established endpoints in clinical trials and are used as quality and performance metrics for rheumatoid arthritis, but are not routinely integrated into patient care [11–13]. While SLE patients are enthusiastic about completing PROMs as part of their clinical care, rheumatologists cite numerous barriers to utilizing the surveys, including the perceived administrative burden and lack of proven benefit [14–16]. Despite growing evidence in other fields that PROMs can improve patient communication, satisfaction with care, and health outcomes, including mortality, few studies have evaluated the use of PROMs in clinical rheumatology [17–21].

The objective of this study was to assess the feasibility and impact of integrating electronic PROMs into the routine outpatient care of patients with SLE. We hypothesized that it would be feasible to complete, review and discuss PROMs as part of clinical encounters and that doing so would enhance communication and patient satisfaction.

Methods

Study design: This was a prospective cohort study of patients with SLE receiving outpatient rheumatology care at two academic medical centres. A mixed-methods sequential explanatory design was employed in which quantitative (survey) data was first collected and analysed [22]. Results of quantitative analyses were further explored through the subsequent collection and analysis of qualitative data (focus groups and interviews).

Participants: English-speaking patients receiving ongoing care (≥2 visits in the previous year with the same provider) meeting 1997 ACR or 2012 SLICC SLE classification criteria were eligible to participate [23, 24]. Patients were approached at routine outpatient appointments and provided written informed consent; treating rheumatologist provided verbal consent. The study was approved by the Tufts Health Sciences Institutional Review Board (IRB# 13159).

Procedures

PROMs: Patients completed electronic PROMs in the context of usual care. PROMs were administered through the ArthritisPower research registry mobile or web-based application (app) at the time of study enrolment and prior to two consecutive routinely scheduled outpatient rheumatology appointments. Patients received an email reminder and/or app notification to fill out the surveys one week prior to their scheduled appointments. Participants who had not completed PROMs prior to arriving for their visit were assisted by research staff to fill out the surveys onsite. Patients who had not completed surveys by the time of their clinical encounter were asked to complete surveys prior to their next scheduled visit. Patients could access PROM score reports through the ArthritisPower website or app.

PROMs consisted of eight PROMIS^® computerized adaptive tests (CATs) (Physical Function, Pain Interference, Fatigue, Participation in Social Roles and Activities, Applied Cognition- Abilities, Sleep Disturbance, Anxiety, and Depression) and four domains from the LupusQoL-US (Planning, Burden to Others, Body Image, Intimate Relationship). These instruments measure domains identified by SLE patients as important to their HRQOL and have been validated in this population [15, 25, 26]. Patients also completed the Systemic Lupus Activity Questionnaire, a validated patient-reported measure of SLE disease activity [27].

Rheumatologists participated in a 30 min in-person training on study procedures and PROM score interpretation. They were given sample score reports and information on general and lupus-specific population norms for PROMIS^® domains and estimates of minimally important changes in scores. Each participating physician also received a set of PROMIS^® T-score maps (https://www.healthmeasures.net/score-and-interpret/interpret-scores/promis/t-score-maps). Physicians were encouraged to review scores prior to clinical encounters and discuss results with patients as they felt appropriate. They were given electronic or printed score reports according to their preference.

In addition to HRQOL measured in PROM surveys, several other outcomes were assessed. Visit length, specifically face-to-face time between provider and patient, was recorded at baseline and at subsequent visits by research staff. The impact of PROMs on the clinical encounter was evaluated through a brief survey immediately following visits administered using Research Electronic Data Capture (REDCap) [28]. Patients and rheumatologists answered Likert-scale and open-ended questions about the use of PROMs in the encounter. Patient activation, patient satisfaction, and disease activity and damage were measured at baseline and at subsequent visits using the Patient Activation Measure (PAM), the provider rating question from the Consumer Assessment of Healthcare Providers and Systems Clinician & Group Survey (CG-CAHPS^®) Version 3.0, Physician’s Global Assessment and SELENA-SLEDAI, and SLICC/ACR Damage Index, respectively [29–31].

Focus groups and semi-structured interviews: Focus groups of patients and semi-structured interviews of rheumatologists were conducted in order to further understand survey data. Focus group and interview guides explored themes identified from free-text responses and probed the impact of using web-based PROMs in clinical care, including potential benefits and challenges. Patients completing the study were invited to participate in the groups, which were led by an experienced moderator (A.L.) and conducted in-person at each site. Physicians were purposefully sampled for interviews to include attending rheumatologists and rheumatology fellows from both sites who had >2 patients in the study. Interviews took place via web-based teleconference. All focus groups and interviews were audio-recorded, professionally transcribed, and de-identified. Transcripts were uploaded into Dedoose™ software for coding and analysis.

Data analysis

Statistical analysis: Baseline descriptive statistics were calculated. The primary outcome was the feasibility of implementing PROMs in clinical care, which was assessed by calculating the proportion of eligible subjects completing surveys at scheduled visits. A completion rate of ≥80% was pre-specified as indicative of feasibility. Remote survey completion rates were also calculated as part of evaluating feasibility. Paired t-tests were performed to assess whether visit length changed between baseline and subsequent visits.

Secondary outcomes explored the impact of integrating PROMs into clinical care. Impact on clinical encounters was evaluated through patient and physician post-visit surveys and the proportion of responses were calculated. The impact of implementing PROMs on patient outcomes was investigated by comparing patient activation, patient satisfaction, HRQOL and disease activity between baseline and subsequent visits in exploratory analyses using McNemar tests for categorical variables and paired t-tests or Wilcoxon signed-rank tests as appropriate for continuous variables.

Qualitative analysis: Responses to open-ended questions in post-visit surveys were analysed for themes. Transcripts from focus groups and interviews were analysed utilizing an integrated approach in which inductive codes emerging from line-by-line reading of the data were employed alongside a priori codes generated from formative work and analyses of the survey data [16, 32]. Each transcript was coded independently by two research staff (S.S., N.P.) and codes were evaluated and revised after each coding session using a comparison and consensus approach. All coding was reviewed by two investigators (S.K., A.L.) and discrepancies were discussed and resolved by consensus. Thematic analysis focussed on the feasibility and utility of integrating PROMs into clinical care, specifically the impact of PROMs on communication and treatment planning. Illustrative quotes were identified for each theme and the perspectives of patients and providers were compared and contrasted.

Results

Patients were recruited from February 2019 through August 2019. Of 120 eligible patients approached at the two sites, 105 were enrolled (88%) (Fig. 1). The most common reason for non-participation was time constraints. Patients were predominantly female and nearly half identified as non-white, with >1/3 insured by Medicaid/Medicare (Table 1). Mean SELENA-SLEDAI was consistent with mildly active disease. Nine attending rheumatologists and eight rheumatology fellows participated, contributing a median of four patients.

Fig. 1.

Study flowchart

Table 1.

Participant characteristics

Patient characteristics (n = 105)
Age, median (25–75%iles) years	39 (32–51)
Female, n (%)	102 (97.1)
Race, n (%)
White	47 (44.8)
Black or African-American	25 (23.8)
Asian	16 (15.2)
More than one	9 (8.6)
Not reported	8 (7.6)
Ethnicity, n (%)
Hispanic or Latino	20 (19)
Insurance, n (%)
Medicaid	25 (23.8)
Medicare	14 (13.3)
Commercial/private	64 (61.0)
None	2 (1.9)
Education, n (%)
High school or below	25 (23.8)
College	53 (50.5)
Post-college	26 (24.8)
Disabled (self-report), n (%)	16 (15.2)
Use of technology, n (%)
Smartphone	98 (93.3)
Tablet	32 (30.5)
Computer	63 (60.0)
None	1 (1.0)
Medications, n (%)
Steroid	64 (61.0)
Prednisone equivalent dose, mean (s.d.) mg	7.5 (3.5)
HCQ	82 (78.1)
DMARD	75 (71.4)
Disease duration, median (25–75%iles) years	8 [4–15]
Baseline SELENA-SLEDAI, mean (s.d.)	3.8 [3.8]
Baseline SLICC-ACR Damage Index, mean (s.d.)	1.1 [1.8]
Comorbid medical diagnoses (self-report), n (%)
Anxiety	22 (21.0)
Depression	24 (22.9)
Fibromyalgia	18 (17.1)
Physician characteristics (n = 17)
Attending rheumatologists	9 (52.9)
Rheumatology fellows	8 (47.1)
Years in practice, median (25–75%iles)	6 (2–18)
SLE patients seen/week, median (25–75%iles)	4 (3–5)
Female, n (%)	10 (58.8)
Race, n (%)
White	14 (82.3)
Asian	2 (11.8)
More than one	1 (5.9)
Ethnicity, n (%)
Hispanic or Latino	1 (5.9)

DMARD: AZA, MMF, mycophenolic acid, MTX, belimumab, CYC, rituximab, or tacrolimus; SELENA-SLEDAI: Safety of Estrogens in Lupus Erythematosus National Assessment-SLEDAI;

SLICC-ACR Damage Index: Systemic Lupus International Collaborating Clinics-ACR Damage Index.

Feasibility: During the study period, 98/105 patients had one or more scheduled follow-up visits, with 95/98 (97%, 95% CI 91, 99) completing at least one follow-up PROM survey and 86/95 (91%) doing so for their first follow-up visit. Seventy-six patients were able to attend two or more follow-up visits before the study was prematurely terminated in March 2020 due to the global COVID-19 pandemic. Sixty-four of these 76 (84%, 95% CI 74, 92) completed two follow-up PROM surveys, exceeding the pre-specified feasibility threshold of 80%, and 55/64 (86%) did so at consecutive follow-up visits.

Overall, patients completed PROMs in 159 of 184 follow-up encounters (86%, 95% CI 81, 91). Nearly all follow-up surveys (148/159 = 93%) were completed remotely. Mean (s.d.) time between baseline and first follow-up PROM was 15.4 (7.1) weeks and 13.8 (6.7) weeks between first and second follow-up PROMs. Median visit length was 17 min (Q1 12–Q3 25.5, range 5–50) at baseline. There was no statistically significant change in visit length after implementation of PROMs (mean change at visit 1 = 0.7 min, P = 0.24).

Impact on clinical encounters: Post-visit surveys were completed by patients and physicians in 157/159 (99%) and 159/159 (100%) of encounters, respectively. Patients reported reviewing and discussing PROM results in 115/157 (73%) and 94/157 (60%) of encounters, respectively. Physicians reported reviewing and discussing the surveys in 146/159 (92%) and 103/159 (65%) of encounters.

Patients and rheumatologists differed in their perceptions of the utility of PROMs (Fig. 2). Patients found PROMs ‘quite a bit’ or ‘very much’ useful in 87/157 (55%, 95% CI 47, 63) of encounters, compared with physicians in 31/159 (20%, 95% CI 14, 27) of encounters. Similarly, patients reported the PROM surveys facilitated communication ‘quite a bit’ or ‘very much’ in 73/157 (47%, 95% CI 39, 55) of encounters vs physicians in 27/159 (17%, 95% CI 12, 24) of encounters. Patients and physicians both found PROMs impacted patient management in some way in about half of encounters. PROMs impacted the treatment plan primarily by guiding conversations (patients = 52, providers = 68), but also by prompting referrals (patients = 13, providers = 8) and influencing medication changes (patients = 5, providers = 12).

Fig. 2.

Patient and rheumatologist views on impact of Patient-Reported Outcome Measures (PROMs)

Among physicians, there was heterogeneity in the reported use of PROMs: the median rate of discussing PROMs was 73% (Q1 59.5–Q3 90.6, range 8.3–100). Rates of discussing PROMs during visits strongly correlated with physician ratings of PROMs utility (Spearman’s ρ = 0.79, P <0.01), impact on communication (ρ = 0.80, P < 0.01) and impact on treatment planning (ρ = 0.78, P < 0.01). Fellows rated PROMs as slightly more useful and beneficial for communication and treatment planning than attendings (∼0.5-point difference in mean Likert scale rating, Wilcoxon rank sum tests P < 0.01).

In responses to open-ended questions regarding the utility of PROMs, patients emphasized the role of the surveys in promoting self-awareness, reflection or validation (n = 47) facilitating tracking of symptoms (n = 38) and helping them prepare for and communicate during visits (n = 23). Many found PROMs ‘informative’ (n = 16) with six patients specifically citing the usefulness of population norms. In contrast, providers described using PROMs to track domains (n = 22), facilitate communication (n = 18) and contribute to their clinical impression (n = 13). Several physicians (n = 12) noted that PROMs identified new topics for discussion, including fatigue, sleep, social stressors and psychiatric concerns. A few (n = 5) commented on the utility of the PROMs in preparing them for the clinical encounter.

Impact on patient outcomes: Exploratory analyses of the impact of PROMs on outcomes are described in Table 2. There were no significant changes in HRQOL or SLE disease activity between assessments at baseline and follow-up visits. Mean T-scores for PROMIS pain, fatigue and physical function domains were about a half standard deviation worse than the general population mean of 50 at both timepoints, while other PROMIS domains remained stable around the general population mean. Patient satisfaction, as measured by patients’ rating of their rheumatologists, was high at baseline and slightly increased at follow-up visits (P < 0.01). A majority of participants exhibited moderate-to-high patient activation on enrolment, and nearly 30% improved from baseline at follow-up visits.

Table 2.

Changes in patient outcomes

Domain	Baseline mean [s.d.] n = 104	Change: Baseline to Visit 1 median [25–75%iles] n = 95	P-valuea	Change: Baseline to Visit 2 median [25–75%iles] n = 64	P-valuea
Health-related quality of life
PROMIS Physical function	46.8 [10.3]	0.0 [–3.3–3.3]	0.66	–0.1 [–3.9–2.8]	0.38
PROMIS Pain interference	54.7 [10.7]	0.0 [–3.6–3.1]	0.93	0.0 [–4.5–5.6]	0.94
PROMIS Fatigue	55.0 [12.4]	–0.1 [-5.6–3.3]	0.42	–0.1 [–5.6–7.6]	0.83
PROMIS Social participation	48.8 [10.8]	0.0 [–4.4–2.4]	0.35	0.0 [–5.2–5.7]	0.78
PROMIS Cognitive abilities	48.8 [9.5]	–1.2 [–4.3–1.4]	0.01	0.0 [–3.4–3.0]	0.83
PROMIS Sleep disturbance	53.7 [11.6]	–0.0 [–6.1–4.1]	0.50	0.0 [–4.6–5.0]	0.70
PROMIS Anxiety	52.8 [9.3]	0.0 [–4.6- 5.4]	0.56	0.4 [–4.3–6.1]	0.19
PROMIS Depression	48.9 [9.5]	0.0 [–5.2–5.3]	0.53	0.0 [–2.7–7.5]	0.18
LupusQoL Planning	74.3 [28.5]	0.0 [–8.3–8.3]	0.66	0.0 [–12.5–0.0]	0.12
LupusQoL Burden to others	68.8 [32.2]	0.0 [–8.4–8.4]	0.94	0.0 [–8.3–16.7]	0.18
LupusQoL Body image	84.2 [23.5]	0.0 [–5.0–10.0]	0.23	0.0 [–5.0–7.5]	0.25
LupusQoL Intimate relationship	66.8 [35.3]	0.0 [–12.5–12.5]	0.64	0.0 [0.0–25.0]	0.15
Disease activity
SLAQ	12.3 [7.3]	0.0 [–2.0–4.0]	0.47	–0.5 [–3.0–3.0]	0.72
SELENA-SLEDAI	3.8 [3.8]	0.0 [0.0–2.0]	0.08	0.0 [0.0–1.3]	0.56
PGA	0.85 [0.6]	0.0 [–0.3–0.3]	0.44	0.0 [–0.5–0.3]	0.68

Patient satisfaction	Baseline n = 104	Visit 1 n = 92b	P-value	Visit 2 n = 64	P-value
CG-CAHPS physician rating (mean [s.d.])	9.6 [0.9]	9.8 [0.5]	<0.01c	9.8 [0.6]	<0.01c
Physician rating ‘10/10’ (n (%))	79 (76.0)	77 (83.7)	0.06d	55 (86.0)	0.26d

Patient activation (PAM)	Baseline n (%) n = 104	Visit 1 n (%) n = 92b	Visit 2 n (%) n = 64
Level 1	13 (12.5)	5(5.4)	9 (14.1)
Level 2	10 (9.6)	9 (9.8)	4 (6.3)
Level 3	39 (37.5)	30 (33.0)	24 (37.5)
Level 4	42 (40.4)	46 (50.0)	27 (42.5)
Improved from baseline	n/a	27 (29.3)	18 (28.1)
Unchanged from baseline	n/a	55 (59.8)	34 (53.1)
Worsened from baseline	n/a	10 (10.9)	12 (18.8)

^aWilcoxon signed rank tests comparing change from baseline to Visit 1 or 2.

^b3 post-visit surveys were incomplete.

^cT-tests comparing means at baseline vs Visit 1 and 2.

^dMcNemar tests comparing proportions at baseline vs Visit 1 and 2.

PROMIS T-scores are normalized around a general population mean of 50 and range from 0 to 100 where higher indicates more of the measured domain.

LupusQoL scores range from 0 to 100 where higher indicates better health-related quality of life.

SLAQ: Systemic Lupus Activity Questionnaire; scores range from 0 to 47, where higher indicates more disease activity.

SELENA-SLEDAI: Safety of Estrogens in Lupus Erythematosus National Assessment-SLEDAI; scores range from 0 to 105, where higher indicates more disease activity and >4 signifies active disease.

PGA: Physician Global Assessment; scores range from 0 to 3, where higher indicates more disease activity.

CG-CAHPS physician rating ranges from 0 to 10, where 10 is the best physician possible.

PAM: Patient Activation Measure, where higher levels indicate greater activation.

Focus groups and interviews: On study completion, 12 patients participated in two focus groups and eight physicians (six attending rheumatologists, two rheumatology fellows) participated in semi-structured interviews. Several themes regarding the impact of PROMs on SLE clinical care emerged, with considerable overlap between patients and physicians (Table 3). Recurring themes were organized into a conceptual model (Fig. 3).

Table 3.

Representative quotes regarding impact of patient-reported outcome measures (PROMs) on systemic lupus erythematosus clinical care

Category/Theme	Quote
Communication
Patient Perspective	‘Even though I knew these patients pretty well, I think [PROMs] made me aware of more aspects of their life and how they see themselves. What is their body image, their relationships, and those kinds of factors that I never really touched on before I knew these patients pretty well, but I still felt that it was useful. It helped me know them better.’ [Rheumatology Fellow]
	‘There are many different aspects of a patient’s sense of well-being and we as physicians have traditionally focussed on the very most medical aspects … I think for some patients that is what their big issue is. I think for some other patients it's just hard because they really don't feel well and yet they're told that each visit, ‘okay, you're doing fine, stay on the same medicine’. I think the impact of [PROMs] is that they sort of bring forward some of their complaints and their feelings about how they're doing without them having to say to the doctor, ‘you know, by the way, I am really tired’. Some patients will say that, but others may feel like, ‘they don't want to know that, they just want to know if my joints are swollen’. I think it definitely adds another dimension in terms of important information that your patient can give you.’ [Rheumatologist, 28 years in practice]
Patient Perspective Therapeutic Relationship	‘The surveys definitely helped because my provider did bring up things that I would never talk about. Things like my feeling like a burden on my spouse. I can be feeling really good in a visit and I won’t talk about anything like that. And I don’t usually bring that stuff up. I think that the survey made him talk to me about that and so I think it kind of strengthened our connection a little bit.’ [Patient, 37F, White, SLE for 8 years]
	‘Did this change my prescription for the patient? Not really, not much anyway. But, did it change our encounter? Yes, sometimes it did, because sometimes it prompted more of a discussion about their feelings and some of these other issues that they're dealing with that we tend not to focus on as much on painful joints. And oftentimes those things would be stable and yet the patient would have other stuff going on. So yeah, I do think it was useful. I think it impacted my interaction with the patient during the course of the visit, which I like to think is, in part, my care of the patient.’ [Rheumatologist, 28 years in practice]
Agenda Setting Therapeutic Relationship	‘It was very helpful because when I went into the room, he knew exactly how I felt … Instead of me telling him and spending that time, we could actually spend time on what happened to me before I got there. So, it was very helpful in that we were able to connect.’ [Patient, 57F, Black, SLE for 17 years]
Agenda Setting Therapeutic Relationship	‘[The PROMs were] nice because it gave the patient the idea that there was something about their well-being that I had looked into before I came into the room and that I had a concern about them that wasn't just that very limited amount of time in the room. That I was checking up on them over time, and thinking about them, and then coming into the room with that information. I think it was helpful to the both of us to focus the problems, but also as bonding, if you will.’ [Rheumatologist, 18 years in practice]
Agenda Setting	‘[PROMs] guide the conversation in a way that's meaningful to the patient clinically. Whether that's lupus or not, it’s still clinically meaningful to the patient. It's almost like the thing where the patient comes in and they save the most important thing for last. That’s a big problem for doctors. This kind of tool actually puts that important stuff up front so that doesn't happen and so that it gets addressed.’ [Rheumatologist, 18 years in practice]
	‘It was helpful sometimes to talk to the doctor about things that you forget in the moment.’ [Patient, 32F, Hispanic, SLE for 3 years]
Patient Satisfaction	‘I guess I probably forgot I did the survey and I came in. I’m like “what are you psychic?” [My rheumatologist] got it. It just makes for a better visit when your doctor’s on the same page as you are.’ [Patient, 57F, Black, SLE for 17 years]
Time Saving	‘It’s very time efficient where they can lead that discussion for you.’ [Patient, 32F, Black, SLE for 1 year]
	‘We just got right to the point. He didn’t have to ask me a hundred questions.’ [Patient, 57F, Black, SLE for 17 years]
	‘It actually is sort of time saving because they think about an answer and then you can focus on what seems to be most of an issue at that point in time.’ [Rheumatologist, 28 years in practice]
Monitoring
Monitoring	‘I use it to track some of my symptoms … especially during the summer months, the exposure to the sun affects me a lot and gives more fatigue, more joint pain. So that was helpful, to see the change in the weather, and how my symptoms change.’ [Patient, 32F, Hispanic, SLE for 3 years]
Monitoring Treatment Plan	‘I think it would strike me mostly when there were psychological things that were worse, and I wouldn't realize otherwise. That made me pay more attention and try to discuss how do we address that.’ [Rheumatologist, 30 years in practice]
Monitoring Patient Activation	‘[PROM surveys] enlighten me as to what to look for, or if I see any changes or notice any kind of change.’ [Patient, 61F, Black, SLE for 32 years]
Awareness of Impact of SLE	‘It was interesting to see that the patients had insight into the changes in their well-being and mood. It was more over long periods of time and not just the short period of time they were being asked. I think that was helpful for them because they could reflect more on their disease in a sort of geologic time.’ [Rheumatologist, 18 years in practice]
Self-Management	‘It’s like an accountability check … what was I doing here where my numbers were up? And, what have I not been doing? And you’re trying to get back there. In this very nonjudgmental way, it just kind of presents you data … This is all information I’ve provided that I can be like, okay, this is where I need to focus.’ [Patient, 32F, Black, SLE for 1 year]
Self-Management Agenda Setting	‘It’s like a self-reminder more than anything else. That’s the most important part for me, I need to remind myself to eat healthy, try to avoid any food that would cause inflammation. And then try to manage my sleep hours and keep myself active. But it also helped at the moment right before the appointment – to try to feel what kind of symptoms I had been experiencing so I can talk to my doctor more about that.’ [Patient, 38F, Asian, SLE for 13 years]
Patient Validation	‘Many patients may find these types of surveys validating even if they don't have a direct impact right at that second on treatment plans. Having people's own experiences reflected in some of the questions and realizing that other people experience these things … I think people find that very validating so I can see that as being a benefit.’ [Rheumatology Fellow]
Moderators/Useful Populations
Longstanding Disease	‘I think I found it more useful in patients who have long-standing disease because I think over time, they tend to under-report how they're feeling because they expect certain things not to be right.’ [Rheumatologist, 18 years in practice]
Inactive Disease	‘Those that don't have as active lupus or don't have as much of a change, you want to bring out the more subtle things that maybe you don't tend to cover. I think that that's where I see [PROMs] maybe doing a bigger service. When they're really flaring it's going to be more often more obvious.’ [Rheumatologist, 27 years in practice]
Inactive Disease Patient–Doctor Relationship	‘The surveys kind of caught me at a time where I was doing pretty good … I felt that I wasn’t really experiencing as many symptoms. And so I didn’t have as much feedback to give at that time. Plus, because I’ve had lupus for so long, and I pretty much had the same doctor for my whole experience, we have a pretty solid rapport, in and out. So I already know what questions he’s going to ask me. I already know what we’re going to talk about or things that I want to talk about. I don’t think the surveys really—for me—added to our appointments.’ [Patient, 29F, Asian/White, SLE for 7 years]
Patient–Doctor Relationship	‘I know a lot of my patients well, and for a long time, and I think they're pretty open with telling me how they're doing. It’s not that [the PROMs] didn't have information that would have been useful, it’s just that I felt like it didn't add that much more to what I could get just from the regular way that I would do things with the patient.’ [Rheumatologist, 27 years in practice]
Personality: Reserved Patient	‘On one occasion, a patient who is usually psychologically well balanced and would be in the middle [on PROMs] for many of these things was definitely different. It led me to ask her what was going on and she shared with me that her father was very sick, and, obviously, this was pre-COVID so she was flying out to see him. This patient is a somewhat reserved kind of person, and although I had taken care of her at least 15 years, she did not tell me that until I asked what was going on. Other patients, it might have been the first thing they said when they came into the room. I think in some ways, it was most helpful for that type of patient; the patient who is a little more reserved, not as prone to sharing personal issues that might be impacting their health.’ [Rheumatologist, 28 years in practice]
Coping Style	‘Doing the survey helps me to know what to expect to go on, but I’m already experiencing a lot of things that’s on it. And it gets real depressing. It can, especially when I’ve been fighting with this since 1985.’ [Patient, 61F, Black, SLE for 32 years]

Fig. 3.

Conceptual model of the impact of patient-reported outcome measures on SLE clinical care

Both patients and physicians emphasized the primary utility of PROMs in facilitating communication by highlighting the patient perspective, enhancing therapeutic relationships and setting the agenda for encounters. PROMs effectively broadened the agenda by giving patients permission to raise concerns important to their well-being. In prioritizing the discussion of these issues, PROMs strengthened the connection between patients and doctors and saved time. Patients suggested that by promoting communication, PROMs contributed to a better clinical experience. Participants also highlighted the role of PROMs in enabling monitoring, which could facilitate self-reflection and awareness of the impact of SLE, in turn stimulating patient activation and self-management. Rheumatologists explained that using PROMs for monitoring could validate patients’ experience of illness and identify emotional and psychological issues affecting health, sometimes leading to changes in the treatment plan. Patients and physicians noted that SLE duration and activity, the strength of the existing patient–doctor relationship, and the personality or coping style of patients could moderate the utility of PROMs. Some providers suggested PROMs could potentially be more useful in the care of patients with less active disease or less established therapeutic relationships. Others emphasized the value of PROMs even in encounters with patients they had known for over a decade.

Participants also commented on challenges of implementing PROMs in clinical care (Supplementary Table S1, available at Rheumatology online), including time constraints, the length of surveys and score reports, and the logistics of workflow integration outside the context of a research study. Patients and physicians suggested that ideal implementation requires detailed guidance on score interpretation and population norms, survey customization options, and integration of PROMs with the electronic health record. Importantly, some patients emphasized the necessity of doctors reviewing and discussing PROMs: ‘It’s only as valuable and useful as you choose to use it in conjunction with your patient … It’s useless if you don’t use it proactively’ (32-year-old Black woman, SLE for 1 year); ‘I’m happy to take the survey, but I’d like the doctor to mention it to me. Am I getting better? Is my condition stable? Am I getting worse? That’s important.’ (63-year-old White woman, SLE for 6 years).

Discussion

In this study we evaluated the implementation of PROMs in the routine clinical care of patients with SLE from the perspective of both patients and physicians. Using mixed methods, we demonstrated the feasibility of integrating electronic PROMs into SLE patient care and posit that PROMs enhanced clinical encounters by promoting communication. Over 80% of participants completed PROMs at all clinical visits and 93% of PROMs were completed remotely. Physicians reviewed and discussed PROMs in over 90% and 65% of encounters, respectively, and there was no increase in visit length. Notably, patients found PROMs useful and beneficial to communication in a majority of encounters and more than twice as often as physicians. They emphasized the importance of their rheumatologists reviewing and discussing PROMs, but many found the surveys valuable even when not directly discussed during visits.

Our findings suggest that PROMs, when integrated into clinical encounters, can enhance SLE patient care in ways that are not always obvious to clinicians. PROMs provided benefit to patients with SLE primarily by facilitating communication, consistent with what has been reported in other populations [20, 33, 34]. As patients noted, the completion of PROMs was not a passive act of transferring information, but an active process that promoted self-reflection. The inclusion of PROMs in routine clinical care also signalled to patients that their physicians cared about aspects of their well-being beyond the objective signs and symptoms of SLE, strengthening the therapeutic bond and giving patients permission to broaden the agenda for the encounter. In addition, PROMs validated the invisible manifestations of SLE, including fatigue, anxiety and cognitive dysfunction, which can be a source of disconnect between patients and providers [35]. Thus, for many patients, the act of completing PROMs in and of itself positively affected clinical encounters, regardless of whether surveys were directly discussed by physicians. For other patients, however, reviewing results with their rheumatologists was an important part of the process, as others have also shown [36–38].

We did not find that that the clinical integration of PROMs improved health outcomes such as HRQOL or SLE activity. Given the small sample size of our pilot study and the short duration of follow-up, this was not unexpected and is similar to what others have reported [17, 33, 39]. We did, however, find that PROMs significantly impacted treatment plans in about 10% of encounters, which may have implications on long-term outcomes. In addition, we hypothesized that by improving patient-centred communication, PROMs may increase patient activation, a metric that has been associated with improved treatment adherence and health outcomes in patients with chronic disease [40]. We did note a group of patients whose activation improved, and larger studies may be helpful in characterizing these patients who may most benefit from utilizing PROMs in their care.

We found a small but statistically significant increase in patient satisfaction, an important quality metric and independent patient-oriented outcome. This is particularly notable given the high baseline satisfaction level in our cohort of patients who had established relationships with their providers. Improvement in patient satisfaction has been demonstrated with the use of PROMs in other clinical settings and may offer an immediate tangible benefit of the clinical integration of PROMs [41].

Our study indicates that PROMs were beneficial to a majority of SLE patients; however, it also strikingly demonstrates that the experience among physicians was much more heterogeneous and at times discordant with the experience of patients. That physicians found PROMs ‘little’ to ‘not at all’ useful in half of the encounters is consistent with data suggesting that many rheumatologists feel that PROMs either provide information that is irrelevant or redundant to what they are already capturing through the clinical encounter [16]. Our finding, that patients derived benefit not only from reviewing PROM scores, but also from the simple process of completing PROMs as part of their clinical care, is an important reframing of value for physicians who aspire to provide patient-centred care. The positive experience and improved communication patients associated with PROMs are themselves important goals for health care delivery. Increasing physician awareness of patients’ desire to include PROMs as part of their clinical care and the benefits of doing so with data driven approaches is critical to overcoming this barrier to implementation.

Other common challenges to the broader clinical implementation of PROMs such as the lack of resources or interventions to address PROMs, time constraints and challenges interpreting scores [42, 43] were infrequently raised by physicians in our study. Physician training at the outset of the study may have mitigated these concerns. Logistical challenges such as survey customization and integration with the clinical workflow linger but may be less daunting with the widespread use of EHRs with built-in PROMs.

Our study has many strengths. The use of mixed methods and inclusion of patient and physician perspectives provides a comprehensive assessment of the feasibility and utility of these tools in clinical care. The inclusion of two clinical sites with diverse patient populations and the high survey response and retention rates are further strengths. Our study also has several limitations. Duration of follow-up was relatively short and the study was prematurely terminated in the setting of the COVID-19 pandemic prior to all follow-up data being collected. Patients were English-speaking, recruited from two tertiary care academic medical centres and treated by academic rheumatologists, so our findings may not be generalizable to patients in other settings. In addition, our cohort had established SLE and high baseline provider satisfaction, limiting our ability to comment on the utility of PROMs in patients with new diagnoses or less amicable relationships with their treating physicians. It is possible, as patients and providers suggested, that the surveys may be of additional benefit in these contexts. Finally, our findings may not be generalizable to PROMs that were not used in this study.

In summary, our study provides valuable insight into the feasibility of integrating PROMs into the clinical care of patients with SLE and the benefits of these surveys in augmenting the patient voice and experience. Broader implementation of PROMs in clinical rheumatology requires building physician confidence in the value of these tools. Clarifying the impact of PROMs on patient activation and health outcomes in longer controlled pragmatic studies may provide further evidence supporting their routine clinical utilization. Regardless, enhanced communication and patient satisfaction may be adequate reasons alone to warrant active integration of PROMs into clinical visits for individuals with SLE.

Data availability statement

The data underlying this article will be shared on reasonable request to the corresponding author.

Supplementary data

Supplementary data are available at Rheumatology online.