Factors Associated With Use of Telemedicine for Follow-up of SLE in the COVID-19 Outbreak

Ho SO Kwong Wah Hospital https://orcid.org/0000-0001-7113-9390 Evelyn CHOW The Chinese University of Hong Kong Isaac T Cheng The Chinese University of Hong Kong Sze-Lok LAU The Chinese University of Hong Kong Tena K Li The Chinese University of Hong Kong Cheuk-Chun SZETO The Chinese University of Hong Kong Lai-Shan TAM (  lstam@cuhk.edu.hk ) Department of Medicine and Therapeutics, The Prince of Wales Hospital, The Chinese University of Hong Kong https://orcid.org/0000-0001-6410-8852


Background
Since coronavirus disease 2019 (COVID- 19) was declared a pandemic, the rapidly increasing number of cases and deaths overwhelmed the health care system worldwide. Community lockdowns, social distancing and quarantine restrictions have been implemented to curb the spread of the infection.
Systemic lupus erythematosus (SLE) is a chronic remitting-relapsing disease that affects multiple organ systems (1). Patients with SLE are at heightened risk of infection due to the underlying disease and the use of immunosuppressive therapies (2). The increased prevalence of comorbidities, such as hypertension and cardiovascular diseases, have been reported to be poor prognostic factors of COVID-19 (3,4).
During this extraordinary time, vulnerable patients such as those with lupus nephritis who are likely to be with multiple comorbidities and on immunosuppressants (IS) will face the di cult choice between risking iatrogenic COVID-19 exposure during a clinician visit and postponing needed care. Indeed, the attendance of our lupus nephritis clinic dropped by 14.4% after the outbreak compared with the same time last year (5). Lupus patients typically require regular follow-up visits to ensure early detection of ares and to monitor the toxicity of immunosuppressive therapy. The unattended patients are at risk of sub-optimal disease control which will lead to damage accrual and high costs (6, 7). As care could be interrupted for a prolonged period, an alternative option would be to adopt telemedicine (TM) or telehealth, the use of telecommunication technologies to provide medical information and services. In fact, the use of TM to reduce potential exposure to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has been recommended by international rheumatology societies (8, 9).
So far, the use of TM in rheumatology has been very limited with few experiences reported. According to a systematic review in 2017, there is no good evidence in supporting the use of TM for the management of rheumatic diseases (10). In a subsequent randomized controlled trial, it was concluded that in rheumatoid arthritis patients with low disease activity or remission, a TM follow-up could achieve similar disease control as conventional care (11). Two studies done during the COVID-19 outbreak reported moderate acceptance of TM as the mode of care in patients with connective tissue diseases (5,12).
However, there is no data on the clinical factors associated with the use of TM in patients with SLE. We hypothesized the patient's decision of choosing TM as the mode of follow-up could be in uenced by their clinical condition on top of the socio-economic pro le.
In this study, we aimed to examine the demographic, socio-economic, psychological, disease and treatment factors predicting the patient's preference of use of TM for follow-up of SLE.

Study design and patients
This was a single-center prospective case-control study. The study was performed at the lupus nephritis clinic of a regional hospital in Hong Kong. From 1st May to 30th November 2020, all consecutive adult patients with a diagnosis of SLE according to the 2019 EULAR/ACR classi cation criteria were invited to participate in the study (13). Patients (or their carers) needed to possess the technology required to conduct a TM visit. Patients were excluded if they were incapable of answering a questionnaire. All patients who had given written informed consent were asked for their interest in changing the coming scheduled follow-up to TM-based in the form of a videoconference. The rst 140 patients agreed to use TM care were recruited. Another 140 patients who preferred to continue standard follow-up were enrolled as controls. All participants were asked to complete a set of questionnaires including the LupusQoL, Health Assessment Questionnaire Disability Index (HAQ-DI), and Hospital Anxiety and Depression Scale (HADS). They also lled in an online questionnaire regarding their perceptions of TM follow-up (supplementary Fig. 1) (5). The responses were assigned a value of − 2 to 2 (strongly disagree to strongly agree), with a higher score indicating that the respondent was more receptive to TM and a zero indicating a neutral response. The study was conducted according to the principles of the Declaration of Helsinki.

Assessments
The disease variables recorded included: disease duration, comorbidities, nephritis class, ever presence of rash/ joint pain, proteinuria, use of IS, SLE disease activity and Systemic Lupus International Collaborating Clinics/American College of Rheumatology (SLICC/ACR) Damage Index (SDI) (14). SLE disease activity was assessed by the Safety of Estrogens in Lupus Erythematosus National Assessment (SELENA) version of the Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI-2k) and physician global assessment (PGA) (14). SLE disease remission was de ned as absence of clinical activity with no use of systemic glucocorticoids and IS; and lupus low disease activity state (LLDAS) as a SLEDAI 2k ≤ 4, PGA ≤ 1 with GC ≤ 7.5 mg of prednisone and well tolerated IS agents (15). Recently, remission and LLDAS were agreed to be the meaningful targets for managing lupus patients in order to prevent damage accrual and improve quality of life (15). All the investigations and assessments were performed within one month before or after the patient was recruited. The clinical assessment of the control group was done face-to-face while that for the TM group was by a videoconference.

Statistical analysis
The overall demographic and clinical characteristics of the recruited patients were reported as mean values with standard deviations (SD) for continuous variables and as numbers and percentages for categorical variables. The patients in the TM and control groups were compared by chi-square test or sher exact test and student t-test where appropriate. Binary logistic regression was conducted for the multivariate analysis of independent predictors with respect to preference over TM follow-up. Spearman correlation was used to investigate the association between SLE disease activity and patient's perception about TM follow-up. A 2-tailed probability value of p < 0.05 was considered statistically signi cant. Statistical analyses were performed using the Statistics Package for Social Sciences (IBM SPSS V.26.0, IBM Corporation, Armonk, NY, USA).

Results
The data of 280 patients (140 TM and 140 controls) with SLE were analyzed. The demographic and disease characteristics are presented in Table 1. The mean age of the recruited patients was 45.6 ± 11.8 years. There was a female predominance (91.4%). The mean disease duration was 15.8 ± 9.5 years. The majority of them (88.2%) had lupus nephritis class III, IV or V. The mean SLEDAI-2k was 3.39 ± 2.35, mean PGA was 0.46 ± 0.62 and SDI was 0.97 ± 1.23. Almost all of them (90%) were on prednisolone with a mean daily dose of 5.82 ± 6.10 mg. Three quarters of the patients (67.1%) were on IS with the commonest being mycophenolate mofetil followed by calcineurin inhibitors. While 70% of the patients were in LLDAS, only 5 (1.8%) had disease remission. A signi cant proportion of the patients (72.1%) had one or more comorbidities. The mean HAQ-DI of the patients was 0.20 ± 0.40. Regarding the anxiety and depression scale, 32.9% and 29.6% of the patients had HADS anxiety score and depression score equals or larger than 8 respectively. The socio-economic pro le of the patients is presented in Table 2.  Binary logistic regression analysis revealed higher PGA, family monthly income > USD3, 800 and nonfulltime employment status remained independent predictors of TM care (OR 1.05 95% CI 1.01-1.09, OR 1.90 95% CI 1.24-3.79, OR 1.89 95% CI 1.13-3.17 respectively) after adjustment for age. PGA was found to be positively correlated with the perceptions that TM follow-up could reduce (r = 0.13, p = 0.036) and routine visit could increase (r = 0.12, p = 0.04) the risk of infection during the COVID-19 outbreak (supplementary table 1).

Discussion
As we de ne the new normal for ambulatory care in the COVID era, we need a new approach to provide routine follow-up for our SLE patients and TM is an obvious option. Our study found that patients with higher physician-assessed disease activity were more willing to use TM instead of standard in-person follow-up. This could be due to the fear of infection exposure during the clinic visits as we also noted higher PGA was associated with the perception that TM follow-up would reduce the risk of infection while routine care would increase that risk. In fact it has been postulated that the more stringent behavioral measures adopted by patients due to the perceived risk could potentially explain the initial reports of paucity of SLE patients with COVID-19 (16). A survey distributed to 199 patients with lupus nephritis during the outbreak showed that their median fear of COVID-19 was 8 out of a maximum scale of 10 (17). However, in subsequent case series, SLE patients with quiescent disease did not seem to be protected from COVID- 19 (18, 19). Comorbidities and glucocorticoid therapy were noted to be overpresented in these infected SLE patients which could explain the higher rate of severe COVID-19 compared to patients with other rheumatic diseases (18). However, no other disease speci c factors including lupus manifestations, glucocorticoid therapy, comorbidities and IS use were found to be predictive of TM use in our study. Another possible explanation for the higher PGA in the TM group could be the perceived less stable disease when the patients were assessed virtually. The accuracy of disease activity assessment via TM warrants further investigations. Interestingly, in a study done before the COVID-19 outbreak, when offered as an option, video TM was also more likely to be used by rheumatoid arthritis patients with higher disease activity (20).
In this study, we also found that higher monthly family income favored TM use. Cavagna et al reported the results of a survey on the propensity for adopting TM in 175 patients with connective tissue disease of whom 49 had SLE (12). It was found that a college degree and distance from the hospital were independent predictors for the acceptance of TM. It might seem conceivable that patients who are socioeconomically more privileged would be more keen to use TM. The issue needs be addressed before universal integration between TM and standard care in order not to exacerbate health care disparities. On the other hand, we found no association in the distance from hospital with the preference of TM. This could be related to the fact that most of our patients were residing close to the hospital. In another study on the perception of SLE patients with regard to choosing TM for follow-up, the privacy/security issue, the accuracy of assessment, as well as the infection risks were important factors considered by the patients (5). When deciding on the mode of care delivery, the patient's perception is also important.
Another intriguing nding of the study is the association of fulltime employment status with standard inperson visit. Border restrictions, quarantine, and social distancing were the anti-endemic measures adopted in Hong Kong. Complete society lock-down or prohibition of social mobility was not in place which meant patients with fulltime employment still had to go to work. As a result, the increased infection risk associated with attending the scheduled clinic follow-up might seem to be negligible. Another potential explanation could be that sick-leave certi cates were not issued for TM care due to administrative reasons in the institution where the study was conducted.
There are several limitations in this study. First, the results should be interpreted in the context of the local outbreak status and mitigation measures implemented. They may not be generalizable to other parts of the world in different outbreak stages adopting different strategies. Second, SLE patients on intravenous cyclophosphamide were not included in the study as they received regular assessment by rheumatologists as day-patients. Lastly, there were logistic issues that might affect the patient's acceptance to TM. For example, under the current work ow, patient still had to come to the hospital for blood and urine tests. The home or working environment of the patients may also be important.

Conclusions
When offered as an option, we found SLE patients with higher physician-determined disease activity and family income who were not full-time employed, were more willing to use TM for follow-up. With the availability of vaccines and anticipated loosening of containment measures, the results could provide important information on the factors to consider when we choose the mode of care delivery during, and also after the COVID-19 outbreak. The subjective disease activity and socio-economic statues but not treatments or comorbidities appeared to be the important determinants in patients with SLE.