Identifying multimorbidity profiles associated with COVID-19 severity in chronic patients using network analysis in the PRECOVID Study

Abstract

A major risk factor of COVID-19 severity is the patient's health status at the time of the infection. Numerous studies focused on specific chronic diseases and identified conditions, mainly cardiovascular ones, associated with poor prognosis. However, chronic diseases tend to cluster into patterns, each with its particular repercussions on the clinical outcome of infected patients. Network analysis in our population revealed that not all cardiovascular patterns have the same risk of COVID-19 hospitalization or mortality and that this risk depends on the pattern of multimorbidity, besides age and sex. We evidenced that negative outcomes were strongly related to patterns in which diabetes and obesity stood out in older women and men, respectively. In younger adults, anxiety was another disease that increased the risk of severity, most notably when combined with menstrual disorders in women or atopic dermatitis in men. These results have relevant implications for organizational, preventive, and clinical actions to help meet the needs of COVID-19 patients.

Multimorbidity Profiles and Infection Severity in COVID-19 Population Using Network Analysis in the Andalusian Health Population Database

“…We used the Jaccard index (JI) to measure the similarity between patients due to the binary nature of the diagnostic variables (i.e., absence/presence), as performed in previous studies on multimorbidity patterns [ 19 , 23 , 29 ]. This index measures the similarity between patients based on individual and shared chronic diseases and ignores clinical diagnostics that none of them has.…”

“…A link between two given patients was created if the JI between them was ≥0.33 to include patients who share half or more of their chronic diseases with another patient. Thus, each node represents a different patient, and a link means a JI ≥ 0.33 between patients, as already applied in previous work [ 23 ], combining clinical and statistical criteria for building the patient networks. This cutoff allowed the inclusion of almost all the patients with multimorbidity (144,990 out of 145,070 patients).…”

“…We used the network’s modularity to search for communities of patients within each network, as previously used in comorbidity and multimorbidity pattern studies [ 20 , 22 , 23 , 30 ]. We used the Leiden algorithm to guarantee well-connected communities [ 31 ].…”

“…Network analysis is a powerful tool to detect networks of individuals and chronic conditions through correlations amongst them, [ 20 , 22 , 23 ] with high applicability in epidemiology and public health to unlock the potential of real-world data for health research, primarily when used in large-scale epidemiologic studies [ 24 ]. This methodology was previously used in a study analyzing COVID-19 patient networks and their infection severity predictors, but considering only patients’ diagnoses recorded in the primary care setting [ 23 ].…”

“…Network analysis is a powerful tool to detect networks of individuals and chronic conditions through correlations amongst them, [ 20 , 22 , 23 ] with high applicability in epidemiology and public health to unlock the potential of real-world data for health research, primarily when used in large-scale epidemiologic studies [ 24 ]. This methodology was previously used in a study analyzing COVID-19 patient networks and their infection severity predictors, but considering only patients’ diagnoses recorded in the primary care setting [ 23 ]. By applying network analysis in real-world data, this study explores multimorbidity profiles in the whole population with laboratory-confirmed SARS-CoV-2 infection in the Spanish region of Andalusia and estimates each patient network’s likelihood of hospitalization and mortality considering age, sex, and the whole spectrum of patients’ chronic diseases from both primary and hospital care.…”