Nocebo-Prone Behaviour in Patients with Autoimmune Rheumatic Diseases during the COVID-19 Pandemic

Background: The COVID-19 pandemic is associated with emotional distress and significant disruptions in health-care services. These are key players in the development of nocebo phenomena. We aimed to investigate nocebo-prone behaviour in patients with autoimmune rheumatic diseases (ARD) amid the COVID-19 pandemic-associated lockdown. Methods: Consecutive patients were telephone-interviewed during the COVID-19 pandemic in Greece. Clinical and socioeconomic characteristics (eg, level of education) were recorded. For nocebo behaviour, a four-item validated questionnaire (Q-No, cut-off score>15), was used. Results were compared with pre-COVID-19 Q-No scores collected from patients followed-up in our department. Results: Nocebo behaviour was detected in 51/500 (10.2%) individuals. In patients with nocebo behaviour, use of anti-hypertensives was less common (17.6% vs 31.8%, p=0.04), but a higher level of education was more common (58.8% vs 35.9%, p=0.002), compared with patients with Q-No score ≤15; the latter retained statistical significance in multivariate regression analysis (p=0.009, OR [95%CI]: 2.29, [1.23–4.25]). Total Q-No scores were higher in the COVID-19-period compared to the pre-COVID-19 era [median (range); 12 (4–20) vs 11 (4–20), p=0.02]. Among 78 patients with available Q-No questionnaires in the pre-COVID-19 era, 11 (14.1%) displayed nocebo behaviour, which increased to 16 (20.5%) amid the COVID-19 pandemic. Interim development of nocebo behaviour was also associated with higher educational level (p=0.049, OR: 3.65, 95%CI: 1.005–13.268). Conclusion: A considerable proportion of ARD patients manifested nocebo-prone behaviour during the COVID-19 pandemic, which was more common among those with high educational level.


INTRODUCTION
The coronavirus disease 2019 (COVID-19) pandemic has significantly affected our everyday life in many aspects. Several public health restrictive measures, such as social distancing and home isolation, were taken to impede virus spreading. Apparently, outpatient clinics, including appointments with rheumatologists, were heavily affected and therefore patient-doctor interaction has been severely disrupted. Besides, the COVID-19 pandemic is being linked with emotional distress similar to that seen in other crises. 1 One could speculate that the psychological effects of COVID-19 pandemic could be even more pronounced in patients with autoimmune rheumatic diseases (ARD), since many of the latter are on treatment with immunosuppressives. The aforementioned factors have been characterised as key players in the nocebo effect. 2 The latter has been described as the opposite of the placebo effect, and denotes the non-intentional negative physical symptoms in response to a medical treatment, which are not a result of pharmacologic action, but occur solely due to negative expectations on behalf of the patient. Neuro-behavioural mechanisms seem to associate with its occurrence, but are so far poorly understood. 3 In recent years, nocebo effects have been recognized as a cause of treatment non-adherence and worse disease outcomes, 2 which is especially relevant in patients with ARD switching to treatment with biosimilars; 4-6 several studies in this field have incorporated interviews and questionnaires aiming to quantify and/or reduce nocebo-prone behaviour. 7,8 Herein, we aimed to investigate nocebo-prone behaviour in ARD patients during the COVID-19 pandemic and whether this has changed compared to the pre-COVID-19 era.

Study population and recorded parameters
Five hundred patients with various ARDs were interviewed via telephone from rheumatologists of our department during the COVID-19 pandemic period in Greece. In particular, interviews were conducted from 14/04/2020 to 22/04/2020. Patients included in this study had been followed-up for at least 6 months in the outpatient rheumatology clinics of our academic center. They were consecutively enrolled, following an alphabetical order of their surname as recorded in our files. Patients with osteoarthritis, crystal arthropathies, undifferentiated rheumatic disease, metabolic bone disease, and fibromyalgia as primary diagnosis, were excluded. The following demographic and clinical characteristics were recorded: age, sex, cohabitation (defined as living or not with a partner), region of residence (urban, semi-urban, rural), level of education (primary, secondary, higher), employment status, and disease duration. Current treatment for ARDs and for the presense of co-morbidities (hypertension, hyperlipidemia, coronary heart disease, diabetes mellitus, depression, anxiety) were also recorded. For detection of nocebo behaviour, a validated questionnaire (Q-No), 7 composed of four questions was performed (Supplementary Table 1). These questions referred to the period of the COVID-19 pandemic in Greece, starting on 26 February 2020 (first reported case). Nocebo behaviour was cosidered to be present when score of this questionaire was above 15. 7 In the pre-COVID-19 era, data for nocebo behaviour were collected prospectively from patients who were routinely followed-up in our outpatients' rheumatology clinics. Also, as part of another recently published study by our group, 9 specific questions, with predefined answer-options were asked: discontinuation or tapering of the medication being received for ARDs, possible reasons that led to drug discontinuation (including but not confined to: fear of immunosuppresion and lack of reourses/drug shortage), whether advice was received from a clinician or other sources, possible symptomatology compatible with COVID-19 infection, testing for COVID-19, and subjective assessment of the activity of their disease (Supplementary Table 2). This study was conducted according to the Declaration of Helsinki and was approved by the Scientific Council of our hospital.

Statistical analysis
Patients who manifested or developed nocebo behaviour were compared to those who did not. Two-sided Fisher's exact and Mann-Whitney tests were used to compare categorical and continuous characteristics, respectively. Wilcoxon matched paired test was used for paired samples. Binary logistic regression analysis was performed using "Positive Nocebo behaviour"as dependent variable, and sex, age, living alone, urban area of residence, unemployment, level of education, disease duration, presence of arterial hypertension, coronary heart disease, hyperlipidemia, depression, anxiety, diabetes mellitus and chronic obstructive pulmonary 2020 290 SUPPLEMENT 2 MEDITERRANEAN JOURNAL OF RHEUMATOLOGY 31 3 2020 SUPPLEMENT 2 disease (COPD) as independent variables. Statistical analysis was performed using SPSS 21.0 (Armonk, NY: IBM Corp, USA).

Associations of Nocebo behaviour and demographic, clinical and COVID-19-pandemic-related characteristics
Comparing patients who had nocebo behaviour with those who did not, we showed that the former were less frequently on anti-hypertensives (17.6% Vs 31.8%, p=0.04) and had more frequently high level of education (58.8% Vs 35.9%, p=0.002). No other differences were detected regarding demographic and clinical characteristics. COVID-19-pandemic related features, including, but not confined to, discontinuation due to fear and/ or due to lack of resources, were comparable between the two groups (

Change in Nocebo behaviour during Covid-19-pandemic
Regarding the pre-COVID-19 era, data about nocebo effect were available for 78 of these 500 patients, 11 of which (14.1%) had displayed nocebo behaviour. This proportion did not differ significantly from the respective during the pandemic (p=0.324).

DISCUSSION
The COVID-19 pandemic, along with the measures that came with that, affected significantly our every-day life and activities. As observed in other disasters, mental health can be also affected. 1,10 Although hard evidence is lacking, it is not irrational to say that this is expected to be more pronounced in patients with chronic diseases, such as ARDs. 1,2 Indeed, qualitative data from a recent study showed that patients with ARDs identified emotional disturbance during the pandemic as a key issue. 11 We hypothesised that alterations in the mental health status along with disturbances in the health-care setting would lead to nocebo-phenomena. We showed that nocebo behaviour was observed in 10% of our cohort which was representative of the patients who are followed-up in our university rheumatology clinic. Interestingly, patients with a high educational level were more prone to nocebo behaviour, while patients  13 Moreover, nocebo effect is more likely to be triggered during oral drug provocation tests in subjects with high education. 14 It is possible that individuals with higher education levels might be more suspicious to drug-related adverse events. Interestingly, a recent study from China showed that people with higher education level also tended to have more psychological distress during the COVID-19 pandemic. 15 On the other hand, there is a lack of data regarding the nocebo effect in patients under anti-hypertensives. As these drugs are effective and generally well-tolerated, one could hypothesise that they can strengthen patient's optimism about drugs. This could indirectly decrease the nocebo effect, as optimists are less likely to follow the nocebo expectation. 16,17 Comparing the nocebo susceptibility classification before and during the COVID-19 outbreak, we found that the proportion of patients with "positive nocebo behaviour" did not differ significantly. However, an increase in the absolute number of nocebo score was noticed. This in-