The impact of COVID-19 pandemic on rheumatology practice and rheumatologists: a cross-sectional multinational study


 ObjectiveTo evaluate the impact of the COVID-19 pandemic on rheumatology practice and on the rheumatologists themselves, and to develop suggestions to improve the practice.MethodsA cross-sectional web survey was designed by members of the Arab League of Associations for Rheumatology (ArLAR), validated by its scientific committee and disseminated through e-mail and social media. It included close-ended questions about the impact of the pandemic on the activities (in percentage, where 100% corresponds to complete suspension), and open-ended questions about unmet needs. Univariate and multivariable logistic regression analyses were used to evaluate the predictors of impact. Suggestions were developed to improve practice.ResultsA total of 858 rheumatologists were included in the analysis (27.3% of registered), 37% were 35-44 years-old, 60% were females and 48% worked in the private sector. The impact of COVID-19 was a decrease of 69% in hospitalizations, 65% in outpatient clinic, 56% in infusion centers and 43% in income. It was associated with the region (highest in the Gulf), the use of telemedicine, the impact on income and the practice sector (lowest in private). Telemedicine was mostly based on traditional telephone contacts and e-mails and reimbursed in 12%. Fifteen rheumatologists (1.8%) were infected and 156 cases of COVID-19 among patients were reported, of whom 22% died. The top-cited unmet needs were: access to drugs and a telemedicine platform.Conclusion The negative COVID-19 pandemic on rheumatology practice may compromise rheumatic diseases control. Better access to drugs and providing telemedicine platforms are recommended to improve the practice.


Introduction
The Coronavirus Disease 2019 (COVID-19) has emerged in December 2019 in Wuhan, China, to quickly become a global outbreak and a signi cant public health issue [1,2]. On January 30, 2020, the World Health Organization (WHO) declared COVID-19 a public health emergency of international concern [3], and, on March 20, 2020, due to the devastating number of new cases reported globally, WHO declared it as a pandemic [4]. At the time of drafting this manuscript (June 28, 2020), the WHO reported more than 9.5 million COVID-19 cases and 495 760 deaths [5].
During the pandemic, health care professionals (HCP) were faced with many challenges in the ght against an unprecedented and rapidly spreading disease [4]. This has prompted the WHO and the Centers for Disease Control and Prevention (CDC) to publish recommendations for the prevention and control of COVID-19 for HCP by the end of January 2020 [6,7]. Similar to all medical and non-medical elds, rheumatology practice was deeply disrupted by the pandemic, partially due to the lockdown and the social distancing constraints that were imposed by the governments [8,9]. Moreover, rheumatologists had to face additional challenges.
Many rheumatology patients are immunocompromised and vulnerable to infection. Whether these patients had higher risk of infection with the novel severe acute respiratory syndrome coronavirus (SARS-CoV-2) was uncertain [10][11][12][13][14][15][16]. Several recommendations were developed worldwide to guide the rheumatologists in their therapeutic choices [17][18][19][20][21][22]. However, the decision to stop, pursue or initiate therapies had to be determined on a case-by-case basis, according to clinical judgment. mostly a hands-on discipline. In some cases, patients even decided to modify their treatment on their own without notifying their physician [23].
Moreover, some symptoms and signs may be overlapping between COVID-19 and immune-mediated disease ares, such as fever, myalgias, arthralgias, elevation of acute phase reactants, leucopenia, thrombopenia, acute interstitial pneumonia, and myocarditis [13,15]. This may cause diagnostic confusion in some cases, resulting in perplexity about therapeutic decisions [24].
Furthermore, several rheumatology drugs are thought to have potential activity against the SARS-CoV-2, such as chloroquine, hydroxychloroquine (HCQ), anti-Interleukin (IL) 6 agents, anti-IL1 agents, and Janus Kinase inhibitors, thus leaving the rheumatology patients in drug shortage sometimes [25][26][27]. In these cases, the rheumatologists were forced to revise the management plan and to adapt it to the available therapies.
Finally, rheumatologists are regularly solicited by chronic patients for advice and by colleagues for assistance in rheumatology drugs prescription for COVID-19 cytokine storm management [28,29], which could be mentally draining.
Therefore, COVID-19 pandemic is a truly challenging time for rheumatologists, particularly in the Arab countries. Unlike USA or Western Europe [30,31], the concept of telemedicine is not well established in the Arab world. The main obstacles include the absence of a legal telemedicine framework and the reluctance of patients to renounce their privileged physician-patient direct communication. For this purpose, a group from the Arab League of Associations for Rheumatology (ArLAR) designed a survey to evaluate the di culties faced by the rheumatologists in the Arab countries.

Objectives
The primary objective was to evaluate the impact of COVID-19 pandemic on rheumatology practice in the Arab countries.
The secondary objectives were to evaluate the impact of COVID-19 on the rheumatologists themselves and to develop suggestions to improve rheumatology practice in the region.

Methods
A 21-items web-based, cross-sectional survey was developed by ArLAR members, with a prospective data collection between May 9 and May 24, 2020. The survey was designed in English by the steering committee, based on the available literature and following the Checklist for Reporting Results of Internet E-Surveys (CHERRIES) guidelines [32]. It was validated by the ArLAR scienti c committee, which consists of rheumatologists representing the 15 ArLAR countries.
The survey instrument was also translated and validated in French, and both versions were made accessible through a Google form link. A pilot test was run in both languages with six bilingual rheumatologists to evaluate the timing, readability, relevance, and acceptability. The questionnaire was considered easy to read, relevant and acceptable, and required around 3 minutes to complete.
The survey comprised 19 close-ended and simple numeric questions about the demographic characteristics, the impact of the pandemic on the rheumatology activity and practice income (expressed in percentages, where 0% corresponds to the absence of impact, and 100% corresponds to a complete suspension of activities), the direct impact on the personal life of the rheumatologists including personal infection with SARS-CoV-2, the impact on mental health, the attitude towards telemedicine, as well as two open-ended questions about the unmet needs and the ways to improve the current practice (Supplementary data 1).
The survey was disseminated to all the rheumatologists working in the 15 ArLAR countries through the ArLAR mass mailing system. It was also advertised on ArLARs' and Arab rheumatology societies' social media (Facebook, Twitter, and Instagram) for 16 days. The study was named HARMONIC (How are the Arab RheuMatologists dealiNg with the COVID-19 pandemIC) on social media. Participation through Google forms was anonymous.
Based on the survey results, the authors developed a set of suggestions to help improve the rheumatology practice across the Arab countries.

Consent to participate and ethical considerations
The rheumatologists received the invitation to participate by mass e-mail through the ArLAR. Clicking on the button " ll out the form" was considered the equivalent of consent to participate in the survey. Con dentiality of personal information was maintained throughout the study by making participants' information anonymous. The study was approved by the Saint-Joseph University Ethics Committee (number CEHDF 1654).

Statistical analysis
The number of rheumatologists in the participating countries is estimated to be around 3137 rheumatologists. We aimed at including around 25% of this total (600). Moreover, for the analysis, the countries of practice were grouped into three

Continuous variables were expressed by mean and standard deviation and categorical variables as numbers and percentages.
Comparison of the rheumatologists' characteristics and responses between the three regions was performed using the Pearson chi-square or Fisher test for the categorical variables and the T-test or ANOVA for the continuous variables. Furthermore, the dependent variable "impact on clinical activity" was categorized into a binary variable: higher impact and lower impact, using the median as a cut-off. Univariate and multivariable binary logistic regression analyses were conducted to identify factors associated with this dependent variable. All independent variables with a p-value 0.1 in the univariate analysis were taken into account in the multivariable logistic regression analysis; p-values <0.05 were accepted as statistically signi cant. All statistical analyses were performed using SPSS v23 (IBM).
The results of the open-ended items were analyzed and assembled into themes. Discordances were resolved through discussion.

Descriptive analysis of the HARMONIC study participants
Out of the 3137 rheumatologists registered in the 15 ArLAR societies, 1214 clicked on the survey link, and 865 responded to the survey (Figure 1: Flowchart of the participants). Seven responses reported a non-Arab current country of practice and were eliminated. Therefore, a total of 858 rheumatologists were included in the analysis (27.3% of total registered rheumatologists).
Over one third of the participants were in the 35-44 years age group. The rheumatologists had been 13.4 years in rheumatology practice (SD 9.7). Sixty percent were females. Around half of the participants worked in private sectors, and 45% worked in university hospitals; notably, 267 rheumatologists (31%) worked in more than one sector. Thirty-nine percent worked in institutions that were implicated in frontline management of COVID-19, and 22% were personally involved in the frontline management of the pandemic. There were signi cant differences between Levant, Gulf, and North Africa. The participants' characteristics by region are presented in Table 1 and their participation rate by country in Supplementary data 2.
Impact of the COVID-19 pandemic on the rheumatology practice The impact of the COVID-19 pandemic on the outpatient activity was estimated to be 65%, highest in the Gulf (76%), and lowest in the Levant (53%) (p<0.001). The impact was 56% on day hospital (infusion centers) and 69% on regular hospitalization, both lowest in the Gulf and highest in North Africa. The impact of the pandemic on the practice income was 43%, highest in the Levant (50%), and lowest in the Gulf (27%) (Figure 2).
The impact on the outpatient clinic activity was categorized into higher versus lower impact, using the median, 70%, as a cutoff. In univariate analysis, a higher impact was associated with age (p<0.001), years of practice (p = 0.03), region (p<0.001), country (p<0.001), number of sectors of activity (p<0.001), private sector of activity (p<0.001), using telemedicine (p<0.001), agreeing using telemedicine (p = 0.001), institution implicated in COVID-19 (p<0.001), frontline management of COVID-19 (p = 0.009), patient with COVID-19 (p = 0.001), HCQ shortage (p = 0.067), personal infection with SARS-CoV-2 (p = 0.063) and impact on income (p<0.001). In multivariable analysis, the region (higher impact in North Africa, followed by the Gulf then the Levant), using telemedicine and higher impact on income remained associated with a higher impact on the clinic activity, whereas working in the private sector was associated with a lower impact on the clinical activity (compared to working in the public sector and in university teaching hospital) ( Table 2). Impact of the COVID-19 pandemic on the logistics of the practice During the visit to the outpatient clinic, 98.3% of the rheumatologists used additional precautions: 93.4% used masks for themselves, 60.7% used masks for the patients, 66.6% used gloves for themselves, 15.4% used gloves for the patients and 78.2% used more antiseptics than usual.
Because of the pandemic, there was a shortage of HCQ in 47% of cases (highest in North Africa (61.2%) and lowest in the Levant (26.1%) (p<0.001). Also, there was a di culty in accessing HCQ in an additional 24.2% of cases. Regarding the practice of telemedicine, the rheumatologists reported to use in 70% of the cases: as a full facility in 10%, as a partial facility in 16%, and more traditional ways (telephone, e-mails) in 51% of cases, with a higher telemedicine practice in the Gulf (p<0.001).
Telemedicine was reimbursed in 22% of cases in the Gulf, 10% in the Levant, and 8% in North Africa (12% in total). As for the agreement to use telemedicine, 54% fully agree to use it, 24% would agree if it's reimbursed, and 22% do not agree.

Impact of the COVID-19 pandemic on the rheumatologists
The mental impact related to the stress caused by COVID-19 was reported in 77% of the respondents. It was minor in 60.4% and major in 16.7% of cases. It was numerically higher in North Africa (80.2% of cases), but the difference between the regions was not statistically signi cant (p = 0.158).
Twenty-ve percent of the rheumatologists felt they were totally prepared in case they were asked to work in the frontline management of COVID-19, 45% felt partially prepared, whereas 30% were not prepared at all. Preparedness was highest in the Gulf (86%) and lowest in North Africa (58%), p<0.001.

Infection with SARS-CoV-2
Fifteen rheumatologists (1.8%) were personally infected with SARS-CoV-2, ve in each of the 3 regions. One third of these rheumatologists were involved as frontline physicians in the management of COVID-19. Also, 34 rheumatologists (3.96%) said they were in quarantine because of COVID-19, 22 of which were from Levant.

Unmet needs and ways to improve the practice
In the open-ended questions, 349 rheumatologists provided 509 comments about the unmet needs in the rheumatology practice in the Arab countries, and 294 rheumatologists provided 565 comments about the way to improve this practice. After the classi cation of the answers into themes, the top-cited needs and ways of improvement were: access to drugs (biologics and HCQ), a telemedicine platform, an organized rheumatology unit, personal protective equipment (PPE), patient education, continuous medical education for physicians and advocacy for rheumatology (the complete ranking of the comments is presented in Table 3).

Discussion
The current study showed a signi cant negative impact of the COVID-19 pandemic on the rheumatology practice. The impact was higher on the regular hospitalization, followed by outpatient clinic and the daily hospitalization for infusion. The differences in impact between the Levant, the Gulf, and North Africa are partly due to differences in the healthcare system (more private-based in the Levant, more public-based in the Gulf), as well as differences in the pandemic response pattern (earlier and more strict quarantine measures in the Gulf and North Africa, compared to the Levant).
This signi cant negative impact on the continuity of rheumatology care, coupled with the shortage in some cornerstone drugs, such as HCQ (47% in the current study), may have a substantial impact on the control of chronic rheumatic diseases, putting the patients at a high risk of disease are and compromising the disease prognosis on the short and long term [33].
The impact on the practice also translated into a 43% impact on the income -so far-, re ecting the serious economic repercussions of the pandemic. This economic impact of the pandemic is global, as announced by a United Nations Conference on Trade and Development report [34]. Nevertheless, it is different from one country to another (lowest in the Gulf countries), re ecting the baseline socio-economic disparities between the countries.
A higher impact on the outpatient clinic was associated with higher use of telemedicine, probably to compensate compromised access of the patient to rheumatology care. However, telemedicine was mostly used in a "traditional way", such as telephone or e-mail contact, which cannot replace a regular visit in rheumatology. Also, this traditional way may not be sustainable on the long term. In particular, the rheumatologists stated that telemedicine was reimbursed in only 12% of cases. This lack of appropriate compensation contributes probably to the reluctance of about half of the rheumatologists if implementing it in their practice.
The mental impact of the pandemic was also considerable, as 77% of the respondents had some sort of impact, even minor in 60%. This is expected, as healthcare professionals, in general, are faced with additional challenges during the pandemic, such as mental stress, physical exhaustion, separation from families, stigma, risk of personal infection, and the pain of losing patients and colleagues. In the current study, 22% of the rheumatologists were implicated in the frontline management of COVID-19, and 39% worked in frontline institutions, meaning that, even in a non-frontline specialty, physicians were highly implicated in the pandemic management.
The rate of rheumatologists' infection with SARS-CoV2 was 1.8%, which may increase with time, as the pandemic evolves.
Although this number may be underestimated, since the seriously infected rheumatologists may not be able to respond to this survey, it gives a preliminary estimate about the direct rate of infection among the rheumatologists, as this information is still lacking worldwide.
As for the reported patients' infection, a high mortality rate (22% of the 156 reported cases) was noted in comparison to the gures reported by the COVID-19 Global Rheumatology Alliance physician-reported registry (9%) [13], and in cohort of patients with chronic rheumatic diseases from Spain and Italy (6 to 10%) [16,35]. This result mostly re ects a bias towards more severe cases that "come to the attention" of rheumatologists, relative to an increased mortality. Also, this bias may be due to the fact that only individuals with severe symptoms are being tested for COVID-19 in most Arab countries.
Based on the open-ended responses, the authors developed suggestions to improve the rheumatology practice during the pandemic and in the decon nement period ( Table 4). The authors highlighted the need for local guidance for rheumatologists and patients, for working with health authorities to guarantee the availability of drugs for patients and PPE for HCP, and for establishing a reliable telemedicine platform that will help to rationalize human resources, reduce infection risk and ensure the proper continuity of rheumatology care.
The study has some limitations. The questionnaire was developed de novo, based on the available literature. However, it was validated by an independent scienti c committee and pilot-tested successfully for readability, acceptability, and timing.
Also, the study was cross-sectional, covering a period where the pandemic had already reached its peak in some countries, whereas it was still in the ascending part of the curves in others. Therefore, the responses re ect the status in each country in a particular time frame of the pandemic. Also, the young age of the participants re ect the users of social media in the Arab region [36], and may jeopardize the external validity of the study.
Besides, the data presented in this study were self-reported and partly dependent on recall ability; thus, they may carry subjectivity and recall bias. Moreover, duplicate data, although very unlikely, cannot be ruled out. However, not collecting personal identi cation data was a choice that the authors had to make it to favor anonymity.
Nevertheless, the current study provides valuable information about the impact of COVID-19 on the rheumatology practice in the Arab regions. Although many studies have addressed the patients' condition during the pandemic, data on physicians and practice remain very limited. The current study has gathered signi cant information from 27% of all rheumatologists across 15 Arab countries within 16 days. It was disseminated in two languages, catching the broadest audience in the region, as most Arab countries study and practice medicine in English and French, except for Syria. Also, the study presented the rst estimates about the rate of rheumatologists' infection and the outcomes of patients' with COVID-19 in the Arab regions. Finally, the results allowed the development of local suggestions to improve the rheumatology practice during the COVID-19 pandemic and for the decon nement period.

Conclusion
The current study highlights the deleterious consequences of the COVID-19 pandemic on the rheumatology practice and on rheumatologists themselves. The compromised access to clinical care and to fundamental drugs is expected to affect the rheumatic disease control signi cantly. Suggestions were developed to improve practice and include developing uni ed local guidelines for rheumatologists and patients. The health authorities are asked to guarantee access of patients to their medications and to establish reliable telemedicine platforms.

Declarations
Funding : None

Con icts of interest
The authors declare no con icts of interest related to this study

Ethics approval
The study complies with the Declaration of Helsinki  Table 3. Summary of the comments of the rheumatologists about the unmet needs in the rheumatology practice (509 comments) and the ways to improve the practice (563 comments), ranked by the number of the mentions by the rheumatologists Figures   Figure 1 Flowchart of the study participants