Temporal changes in co-morbidities and mortality in patients hospitalized for COVID-19 in Italy.

To the Editor.

In response to the growing pandemic of Coronavirus Disease 2019 (COVID-19) in Italy, the Italian government imposed a strict nationwide quarantine which began on March 9, 2020. The restrictive measures were associated with a remarkable reduction in the spread of clinical manifestations of coronavirus. Positive effects in terms of fewer hospitalizations and fatalities occurred within a few weeks [1].

In this setting, a study by Patti and co-workers [2], recently published in this Journal, investigated the temporal changes in the clinical severity of COVID-19 patients at the time of their hospital presentation. The Authors retrospectively reviewed the clinical records of patients admitted for Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2) infection between March 1 and April 13, and between April 14 and May 12, 2020 in three hospitals of the Piedmont region [2]. At hospital admission, the severity of respiratory tract involvement, as assessed by the rapid development of severe respiratory distress syndrome requiring mechanical ventilation, and the inflammatory status were more pronounced during the earlier time window than during the later one [2]. The Authors speculated that the exposition to higher viral loads and the preferential hospitalization of more severe cases during the earlier phase of the pandemic were reasonable mechanisms supporting their findings. However, the impact of disease on mortality was not assessed in that study [2]. It is well established that advanced age and underlying medical comorbidities are powerful determinants of clinical severity of SARS-CoV-2 infection and mortality [3,4].

To elucidate these findings, in the setting of an ongoing prospective registry of patients affected by COVID-19 [5], we evaluated whether and to what extent the clinical characteristics of patients and the rate of all-cause mortality changed over time. We analyzed data from consecutive patients hospitalized from February 23 to May 22 2020, in 5 hospitals of the Lombardy region and belonging to the Maugeri Care and Research Institutes Network. Notably, our study was not a retrospective collection of clinical notes of patients hospitalized for COVID-19, but rather a pre-designed protocol with subsequent prospective collection of data. The protocol was approved by the Ethical Committee of our Institution and patients gave their written informed consent to participate. Diagnosis of viral infection was confirmed in all patients by nasopharyngeal swab. The presence of comorbidities was defined according to documented medical history, as collected by investigators at study site-level. All clinical evaluations were performed by attending physicians during the clinical interview and through interrogation of medical records. The study outcome was all-cause mortality during hospitalization.

Logistic and time-series regression analyses were used to weight the prognostic impact of comorbidities (age, hypertension, coronary artery disease [CAD], heart failure, type 2 diabetes mellitus [T2DM], chronic kidney disease [CKD], and obstructive pulmonary diseases [COPD]) and to compute their changes in prevalence among hospitalized patients over time. We used STATA 15 (StataCorp, USA) and R software version 3 (R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org) for data analysis and in 2-tailed tests, pvalues < 0.05 were considered statistically significant.

Overall, we studied 856 patients with complete medical history. During hospitalization, 67 patients died. Mean age was 72 years and 488 patients were males. The most prevalent comorbidity was hypertension (60%), followed by T2DM (24%), CAD (16%), COPD (13%), heart failure (8%), and CKD (7%). According to distribution of the time of hospital admission, the study population was categorized in 6 different subgroups (from A to F). Each subgroup included patients admitted to hospital over a period of two weeks (Fig. 1 ). We also computed the prevalence of comorbidities and the rate of in-hospital mortality (%) in each subgroup (Fig. 1).

Fig. 1

Changes over time of in-hospital mortality and prevalence of medical comorbidities (see text for details).

Legend: CHF = congestive heart failure; CAD = coronary artery disease; COPD = chronic obstructive pulmonary disease; T2DM = type 2 Diabetes Mellitus.

Among comorbidities, older age (≥75 years; odds ratio [OR]: 6.7, 95% confidence interval [CI]: 3.4–13.3; p < 0.0001), heart failure (OR: 3.3, 95% CI: 1.7–6.3; p < 0.0001), CAD (OR: 2.2, 95% CI: 1.2–3.8; p = 0.007), hypertension (OR: 1.7, 95% CI: 1.1–3.0; p = 0.048), T2DM (OR: 1.7, 95% CI: 1.1–2.9; p = 0.049) and COPD (OR: 2.1, 95% CI: 1.1–3.8; p = 0.020) were significant predictors of mortality in the univariate analysis. When forcing these variables in the same multivariate model, only age and heart failure achieved the highest goodness-of-fit score and retained statistical significances.

Time series analyses (Fig. 1) documented a marked and progressive fall in the rate of in-hospital mortality over time (from 24% to 1%, p < 0.05). Of note, the positive trend of in-hospital mortality was associated with a significant reduction in the number of patients with comorbidities (all p < 0.05, Fig. 1). Notably, time of hospitalization significantly affected the risk of mortality. When compared with patients hospitalized before March 9 (group A), the risk of in-hospital mortality progressively decreased by 40%, 80%, 87%, 87%, and 95% in groups B, C, D, E, and F, respectively (Fig. 2 , left panel). Such effect remained significant after adjustment for the impact of age (for each increment of 5 years: OR 1.63, 95% CI: 1.39–1.92, p < 0.0001) and heart failure (OR 2.52, 95% CI: 1.24–5.11; p = 0.011) as medical comorbidity (Fig. 2, right panel).

Fig. 2

Risk of all-cause mortality according to the time period in which the hospital admission occurred. Group A was the reference and included patients hospitalized before March 9, 2020. The left panel shows the results of univariate analysis. The right panel shows the results of multivariable analysis, after adjustment for the significant effect of age and heart failure.

Legend: OR = odds ratio; Ref = reference; CI = confidence interval.

In summary, we prospectively evaluated the association between calendar date of hospitalization for COVID-19 and the clinical characteristics of patients, along with the occurrence of all-cause mortality. In order to better define the time course of changes, we sub-divided the observation period into six time-windows. Our findings are consistent with the study by Patti et al. [2], who noted a reduction in the clinical severity of patients hospitalized in the period April 14-May 12 compared to the period between March 1 and April 13, 2020. Conversely, a study by Flacco et al. [6] provided discordant findings by showing that COVID-19 patients hospitalized from April 1 to May 3 had a more advanced age and a significant increase in the prevalence of comorbidities (i.e., diabetes, major cardiovascular diseases, cancer, renal diseases) when compared with those hospitalized from March 3 to March 31, 2020. Such increase was paralleled by an increase in mortality from 9.5% in March to 12.1% afterwards [6]. In a recent meta-analysis of data gathered from 36 European countries [7], the case-fatality rate of patients with COVID-19 progressively rose from 1.5% in the week subsequent to the first death to 6.5% at the sixth week, with an average case-fatality of 4.5%. However, the meta-analysis could not adjust for the changes in the clinical characteristics of subjects across the different Countries over the examined time period [7].

The main novelty of the present study is the temporal association between the clinical characteristics of patients admitted to hospital for COVID-19 and the case-mortality rate over a period of strict lock-down imposed by the Government. The severity of COVID-19 at hospital admission and the case-mortality rate decreased significantly and substantially over the explored time period, although the adverse prognostic impact of age and heart failure remained independent of earlier or later date of hospitalization. These data suggest that older age and comorbidities retain an important adverse impact on case-fatality of COVID-19 even taking into account the favorable temporal trend.

The basic mechanisms underlying our results are elusive. It has been speculated that the measures of social constraint during lock-down and the large use of facial masks may have progressively reduced the viral load directed to individuals destined to become infected [8,9]. Such a view is supported by a study from China in which the mean viral load in nasopharyngeal swabs was about 60 times higher in more severe than in less severe cases. [10] Further studies on larger population are needed to verify the temporal trend of COVID-19 case-fatality and the impact of the less stringent measures implemented after the more rigid lock-down phase.

Declaration of Competing Interest

The Authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.