Fever Correlation with Erythrocyte Sedimentation Rate (ESR) and C-Reactive Protein (CRP) Concentrations in Patients with Isolated Polymyalgia Rheumatica (PMR): A Retrospective Comparison Study between Hospital and Out-of-Hospital Local Registries

Abstract

Background: Polymyalgia rheumatica (PMR) is the most common systemic inflammatory rheumatic disease affecting the elderly. Giant cell arteritis (GCA) is a granulomatous vasculitis affecting the aorta and its branches associated with PMR in up to 20% of cases. In recent studies based on university hospital registries, fever correlated with the erythrocyte sedimentation rate (ESR) but not with C-reactive protein (CRP) concentrations at the time of diagnosis in patients with isolated PMR. A long delay to a PMR diagnosis was suggested to explain this discrepancy, possibly caused by laboratory alterations (for instance, anemia of chronic disease type) that can influence only ESR. We performed a retrospective comparison study between the university hospital and two out-of-hospital public ambulatory databases, searching for any differences in fever/low-grade fever correlation with ESR and CRP. Methods: We identified all patients with newly diagnosed PMR between 2013 and 2020, only including patients who had a body temperature (BT) measurement at the time of diagnosis and a follow-up of at least two years. We considered BT as normal at <37.2 °C. Routine diagnostic tests for differential diagnostics were performed at the time of diagnosis and during follow-ups, indicating the need for more in-depth investigations if required. The GCA was excluded based on the presence of suggestive signs or symptoms and routine ultrasound examination of temporal, axillary, subclavian, and carotid arteries by experienced ultrasonographers. Patients with malignancies, chronic renal disease, bacterial infections, and body mass index (BMI) > 30 kg/m2 were excluded, as these conditions can increase CRP and/or ESR. Finally, we used the Cumulative Illness Rating Scale (CIRS) for quantifying the burden of comorbidities and excluded patients with a CIRS index > 4 as an additional interfering factor. Results: We evaluated data from 169 (73 from hospital and 96 from territorial registries) patients with newly diagnosed isolated PMR. Among these, 77.7% were female, and 61.5% of patients had normal BT at the time of diagnosis. We divided the 169 patients into two cohorts (hospital and territorial) according to the first diagnostic referral. Age at diagnosis, ESR, CRP, median hemoglobin (HB), and diagnostic delay (days from first manifestations to final diagnosis) were statistically significantly different between the two cohorts. However, when we assessed these data according to BT in the territorial cohort, we found a statistical difference only between ESR and BT (46.39 ± 19.31 vs. 57.50 ± 28.16; p = 0.026). Conclusions: ESR but not CRP correlates with fever/low-grade fever at the time of diagnosis in PMR patients with a short diagnosis delay regardless of HB levels. ESR was the only variable having a statistically significant correlation with BT in a multilevel regression analysis adjusted for cohorts (β = 0.312; p = 0.014).