Introduction and objectives

People with inflammatory rheumatic diseases (IRDs) being treated with immunomodulatory medications have higher prevalence of COVID-19 and are at an increased risk of poor outcomes [1]. We aim to characterize affected Filipino patients at the outset of the pandemic and to determine the relationships between patient variables and their outcomes, intended as a historical perspective to guide rheumatologists as vaccinations and variants emerge.

Methods

Upon protocol approval by institutional review boards, Philippine rheumatologists registered patients with IRDs and confirmed COVID-19 infection into the Global Rheumatology Alliance (GRA) registry between March 2020 and August 2021. De-identified data were entered into REDCap. Descriptive statistics and regression analyses were performed.

Results

Patient characteristics

There were 164 Filipinos (70% female) with IRDs diagnosed with COVID-19 registered. Median age was 44 years; 85.4% were 65 years old and younger (Table 1). The most common IRDs were SLE (41%), RA (15%), and gout (15%). Most were on csDMARDs (59%), most commonly hydroxychloroquine and methotrexate, and/or GC (51%). Of those on DMARDs, 49% were on a stable dose while 36% had a recent dose increase. It was either stopped or tapered during their COVID-19 illness in 14%. Among patients on GC therapy, 26% were receiving ≥ 10 mg of prednisone equivalent per day.

Table 1 Characteristics of included patients (n = 164)
Full size table

COVID-19 and its outcomes

Table 2 shows the most common symptoms. Most infections occurred between July and October 2020 (37%) and between March and May 2021 (31%), and were community acquired (51%) or through close contact with a confirmed case (40%). Very few (8.5%) acquired it from a health facility.

Table 2 Most common symptoms of COVID-19 infection (n = 164)
Full size table

Among those who had a blood count (n = 78), 42%, 26%, 31%, and 13% had anemia, leukopenia, lymphopenia, and thrombocytopenia, respectively. Other abnormalities were elevated liver transaminases (44%), elevated D-dimer (61%), ferritin > 2000 ng/mL (34%), elevated triglyceride (22%), fibrinogen < 250 mg/dL in 4/9, and elevated IL-6 in 12/16 patients.

Half (59%) were hospitalized which increased to 83% among the those aged > 65 years. Multivariate analysis showed increased odds of hospitalization among COVID-19 patients with active IRD (OR 3.7), heart disease (OR 8.52), and hypertension (OR 8.73), and reduced odds for those with SLE (OR 0.15) compared to those with gout (Table 3).

Table 3 Association between patient factors and outcomes of hospitalization, need for oxygen supplementation or mechanical ventilation, or death
Full size table

Among hospitalized patients, 76% (52% of all patients) needed supplemental oxygen, with 14% needing mechanical ventilation. Multivariate analysis showed that heart disease (OR 6.28), hypertension (OR 7.6), and moderate to high rheumatic disease activity (OR 3.37) were associated with need for oxygen supplementation, while hypertension increased the odds (OR 8.23) for mechanical ventilation.

Two-thirds had no serious complication. Sepsis or secondary infection was the most common (15%), followed by ARDS (10%), cytokine storm (7%), and myocarditis or new heart failure (6%). Twenty-four (15%) died after a mean hospital stay of 17 days. Among those who survived, recovery was after a median of 13 days from symptom onset. Odds of dying were lower among those receiving prednisone ≥ 10 mg/day [OR 0.17 (0.04 to 0.68)] and with other autoimmune diseases besides SLE and RA, including systemic sclerosis (SS), mixed connective tissue disease (MCTD), and inflammatory myopathy (IIM) (OR 0.05).

Discussion

On patient characteristics

Most of our patients were females, similar to reports of Gianfrancesco et al., which described the first 600 patients included into the GRA from March 24 to April 20, 2020, and the meta-analysis of Xu et al. that included 1138 patients from North America, Europe, and Asia, excluding data from the GRA [2, 3]. Our patients were younger (median age of 44 years) than those reported by Gianfrancesco (56 years) [2].

Gout is the most prevalent inflammatory arthritis in the Philippines (0.5 to 1.6% of the population) [4,5,6]. RA and SLE are less common, affecting 0.6% and 0.03–0.05% of the population, respectively [6, 7]. The high proportion of SLE in our study may reflect reporting bias since these patients were more likely referred to rheumatologists when hospitalized. Similar to the GRA data, hypertension was the most common comorbidity.

More than half (62%) of our patients were in remission or low disease activity when they got infected, but this is lower than in the overall GRA data (80%), reflecting limited access to, and cost of, second-line therapies in the Philippines, worsened by pandemic restrictions, and decreased availability of hydroxychloroquine and methotrexate in March to July 2020 [7].

On COVID-19 and its outcomes

Very few patients acquired the virus from a healthcare facility, a point of emphasis since decline in consults due to fear of acquiring infection from health facilities may worsen primary rheumatologic conditions [8]. The most common symptoms in our patients with rheumatic disease were similar to the general population in the Philippines and Wuhan, China [9]. High baseline levels of inflammatory biomarkers and decreased blood counts are associated with severe disease and mortality [10, 11].

The hospitalization rate in our study (59%) was higher than that in GRA (46%) and in the meta-analysis of Akiyama (35%). Using the gout subset as the reference group in our cohort, patients with SLE had lower odds (OR 0.15) of hospitalization, likely because while gout by itself is not life-threatening, hyperuricemia is predictive of comorbidities that increase severity and poor outcome with COVID-19 [12, 13].

One-third of our patients had serious complications signifying critical illness, compared to the general population where critical illness occurs in about 5% and in 22% of hospitalized COVID-19 patients [14]. Higher occurrence of serious complications in our patients may be due to delayed health-seeking behavior and out-of-pocket expenses.

The death rate in our series (15%) is higher than GRA (9%) and Akiyama’s (7%), but lower than the case fatality rate Xu reported for Asia (27%). We found a lower risk of dying among people other autoimmune conditions (SS, MCTD, IIM, and others) compared with gout, in contrast to a UK population-based study which found an increased risk of death by COVID-19 for those with RA independent of co-morbidities, but not for those with gout [15]. It is likely our cohort with gout had serious comorbidities that led to their poorer prognosis.

We found reduced risk of dying among those receiving prednisone ≥ 10 mg/day [OR 0.2 (0.05 to 0.68)], contrary to GRA’s result, but this finding supports a possible protective role of steroids due to its anti-inflammatory effects. In a study by Ye et al., patients on low- to medium-dose glucocorticoids (5 mg to 15 mg/day prednisone) had less severe/critical covid conditions [16]. Andersen’s retrospective cohort also found no increase in the risk of in-hospital death and risk for mechanical ventilation among people on long-term (at least 14 days) immunosuppressive medications, except rituximab [17].

Limitations

There is underreporting due to a lack of rheumatologists in the country, reduced consults, lack of healthcare facilities, and psychological well-being effects [18].

Conclusion

Majority of Filipinos with inflammatory rheumatic diseases infected with COVID-19 prior to vaccinations and variants reported in this registry were females with a mean age of 44 years. The most frequently reported IRDs were SLE, RA, and gout. Majority were on cDMARDs and/or GC. Hospitalization rate is about half. Mortality rate (15%) was higher than reported in other populations but relatively lower for Asia. Poor rheumatic disease control was associated with increased odds of hospitalization and need for supplemental oxygen. The decreased odds of death in patients on prednisone ≥ 10 mg/day is similar to some studies but contrary to most. Reporting bias is a limitation. Continued registry reporting of COVID-19 cases in people with rheumatic diseases worldwide as well as historical documentation are crucial in order to have evidence-based recommendations for the management of our patients.