COVID-19 epidemiology in rheumatic diseases in Tuscany: A case-control study

SARS-CoV-2 infection for patients with inflammatory rheumatic diseases has raised two main issues. Firstly, given the moderate increase of infection risk associated with conventional synthetic DMARDs biologic DMARDs, do we expect a higher number of COVID-19 cases in patients receiving these drugs? Secondly, considering the anti-cytokine effect of most therapies, may bDMARDs and csDMARDs be potentially protective against the COVID-19? We performed a case control study, in which cases were represented by symptomatic COVID-19 infections occurred from February to July 2020, with a positive PCR test result compared to the general population, and rheumatic diseases and related therapies were considered as exposures. Patients affected by rheumatic diseases were identified through previous records of hospital discharge (SDO) or payment exemption [1] and were aggregated into two groups: connective tissue disease or chronic rheumatic arthritis. Therapies administered in 2019 (two prescriptions spaced at least six months apart) with at least one of the following drug groups:

• •hydroxychloroquine;
• •corticosteroids;
• •immunosuppressants or csDMARD;
• •tsDMARD;
• •bDMARDs were also considered.

Odds ratios (with 95% confidence interval) of COVID-19 were performed for each single covariate (type of rheumatic diseases, age band, gender and therapies), crude and adjusted for all the other covariates, by multivariate logistic regression model. All the analyses were performed by using STATA 15. Over a 6-month period, 4,581 COVID-19 symptomatic cases in 2,589,374 Tuscany people older than 30 years were recorded. Age and gender distribution of the general population and of the COVID-19 cases are shown in Table 1 . As expected, the majority of the COVID-19 cases were observed in older people, with a cumulative 49.8% of cases occurring in the subjects older than 70. Overall, chronic rheumatic arthritis were more frequent among cases than in the general population (OR 1.87), while connective tissue diseases did not. Also, previous use of corticosteroids (OR 2.13) and, to a lesser extent, of csDMARDs (OR 1.52), was more frequent among cases than in the general population, while previous use of hydroxychloroquine and of bDMARDs did not. After the adjustment for age and gender, chronic rheumatic arthritis and previous use of corticosteroids resulted significantly associated with the development of symptomatic COVID-19 (Table 2 ). The epidemiological impact of COVID-19 in patients with rheumatic diseases being treated with bDMARD has been the subject of several reports by Italian groups. The methodology for capturing incident cases could have determined different values, with a low incidence using e-mails, telephone calls, survey [2], [3], [4]. Conversely, a prevalence study or evaluation with serological tests has shown an increased risk in patients with rheumatic diseases [5], [6]. The methodology of our SDO-based study for disease and drug prescription confirms that patients with chronic rheumatic diseases (RA, PsA, AS) have an increased incidence, as observed by a recent Italian registry [7]. The present study confirms the risk association with corticosteroids and the indifferent role of bDMARD and hydroxychloroquine as reported by the Global Rheumatology Alliance Physician [8]. Regarding sex and age, we have not observed a gender correlation, while the progression of attendance by age groups 40–49, 50–59, 60–69 years has already been demonstrated in other researches, with a progressive increase in hospitalisations and the demand for intensive care [9]. Furthermore, the epidemiological methodology [10] can determine different results even in a population coming from the same geographic area. This study confirms an increase risk of symptomatic COVID-19 disease associated with chronic rheumatic arthritis and use of corticosteroids, but not with connective tissue diseases and use of hydroxychloroquine, csDMARDs and bDMARDs.

Table 1.

Age class, gender, rheumatic diseases and previous therapies among general population and cases. Prevalence per 100 and odds ratios of COVID-19.

	General population 30+ (n = 2,589,374)		COVID-19 cases 30+ (n = 4581)		Odds ratios
	n	%	n	%
Age groups
30–39	361,733	14.0	214	4.7	1.00 (ref.)
40–49	519,718	20.1	485	10.6	1.58 (1.34–1.85)***
50–59	558,404	21.6	834	18.2	2.52 (2.17–2.93)***
60–69	445,748	17.2	766	16.7	2.90 (2.50–3.38)***
70–79	396,999	15.3	918	20.0	3.91 (3.37–4.54)***
80+	306,772	11.8	1,364	29.8	7.52 (6.51–8.68)***
Gender
F	1,378,141	53.2	2,269	49.5	1.00 (ref.)
M	1,211,233	46.8	2,312	50.5	1.16 (1.09–1.23)***
Rheumatic diseases
Chronic rheumatic arthritis	28,366	1.1	93	2.0	1.87 (1.52–2.30)***
Connective tissue diseases	11,879	0.5	23	0.5	1.11 (0.73–1.67)
Therapies
Corticosteroids	35,088	1.4	130	2.8	2.13 (1.79–2.53)***
Hydroxychloroquine	8,570	0.3	15	0.3	0.99 (0.60–1.64)
cs-DMARDs	9,695	0.4	26	0.6	1.52 (1.03–2.23)*
b-DMARDs	6,128	0.2	9	0.2	0.83 (0.43–1.60)

*P-value < 0.05; **P-value < 0.01; ***P-value < 0.001.

Table 2.

Age and gender and fully adjusted odds ratios.

	Age and gender adjusted odds ratios	Fully adjusted odds ratios
Rheumatic diseases
Chronic rheumatic arthritis	1.62 (1.32–1.99)***	1.64 (1.32–2.05)***
Connective tissue diseases	1.06 (0.70–160)	1.09 (0.72–1.66)
Therapies
Corticosteroids	1.46 (1.22–1.74)***	1.41 (1.18–1.69)***
Hydroxychloroquine	0.79 (0.48–1.31)	0.62 (0.37–1.04)
cs-DMARDs	1.28 (0.87–1.88)	0.96 (0.64–1.44)
b-DMARDs	0.91 (0.47–1.75)	0.63 (0.32–1.23)

*P-value < 0.05; **P-value < 0.01; ***P-value < 0.001.

Disclosure of interest

The authors declare that they have no competing interest.