Cohort characteristics
Over the six-year observation time, a total of 12,435 hospitalizations (6146 female, 6289 male) occurred in the Internal Medicine Unit. Of those, 1481 (or 11.9%) patients presented a diagnosis of HF. Their mean age was 82.3 ± 9.3 years. More specifically, 610 patients (41.2% of those with HF) had HFp and 871 (58.8%) had HFs.
Patients with zero comorbidities included 27.2% of hospitalized patients without HF, followed by 37.5% with one comorbidity, 23.7% with two comorbidities, 9.0% with three comorbidities, and 2.6% with four or more comorbidities. The corresponding picture for HF patients was 13.7% with zero comorbidities, followed by 35.5% with one comorbidity, 30.0% with two comorbidities, 15.4% with three comorbidities, and 5.4% with four or more comorbidities (Fig. 1).
Distribution of overall comorbidities in hospitalized patients with or without HF (left columns), and because of HFp or with HFs (right columns)
A greater prevalence of multiple comorbidities was observed in patients with HFs as compared to those with HFp: for example, 25.8% of HFs patients had three or more comorbidities as compared to 13.7% of HFp patients (Fig. 1).
Occurrence of cardiac and non-cardiac comorbidity
HF hospitalized patients with zero cardiac comorbidities were 55.2%, followed by 31.7% with one cardiac comorbidity, 10.9% with two cardiac comorbidities, and 2.2% with three or more cardiac comorbidities (Fig. 2).
Distribution of the number of cardiac (top panel) and non-cardiac (bottom panel) comorbidities
The corresponding picture for non-cardiac comorbidities was 23.4% with zero non-cardiac comorbidities, followed by 44.1% with one non-cardiac comorbidity, 20.6% with two non-cardiac comorbidities, and 5.9% with three or more non-cardiac comorbidities (Fig. 2).
Patients with HFp had more cardiac comorbidities than those with HFs, who had at least one non-cardiac comorbidity in 84.7% of cases as compared to 50.3% in HFp.
Atrial fibrillation was the most prevalent cardiac comorbidity (25.6%) in HF patients, followed by hypertension (17.0%), diabetes mellitus (9.9%), and prevalent coronary heart disease (7.9%) (Table 1). Respiratory failure and pneumonia were the most prevalent non-cardiac comorbidity within HF population (38.9%), followed by kidney disease (27.9%), anemia (15.5%), sepsis (8.7%), and dementia (4.7%) (Table 1).
Atrial fibrillation was more common in HFp (31.0% versus 21.7%), while respiratory failure or pneumonia (55.0% versus 15.9%) and sepsis (14.0% versus 1.2%) were more common in HFs.
The burden of cardiac and non-cardiac comorbidities on length of stay
The number of comorbidities was associated with the in-hospital length of stay. As illustrated in Fig. 3, the average length of stay progressively increased in association with the number of comorbidities from 7.6 ± 0.5 days in subjects with no comorbidity to 13.9 ± 1.5 days in subjects with 4 + comorbidities (p < 0.0001).
Association between the number of comorbidities and the length of stay in hospital in HF patients
Interestingly, the length of stay was greater in HFs than in HFp patients, at any level of comorbidity.
Multiple regression analyses, controlling for age and sex, confirmed this observation (number of comorbidities: beta coefficient 1.41 ± 0.22, p < 0.0001). It is noteworthy that, after controlling for age and sex, only non-cardiac comorbidities remained significantly associated with an increased duration of hospitalization (number of non-cardiac comorbidity: beta coefficient 2.86 ± 0.27, p < 0.0001).
The burden of cardiac and non-cardiac comorbidities on in-hospital mortality
No significant linear correlation was found between in-hospital mortality and length of stay. Rather, the association between in-hospital mortality and length of stay showed a tendency toward a J-shaped relationship (Supplemental Fig. 1) – possibly reflecting early death upon admission.
In-hospital mortality was associated with the number of comorbidities: it was 4.4% in HF patients with zero comorbidities, increased to 13.7% in HF patients with one comorbidity, 18.9% with two comorbidities, 21.5% with three comorbidities, and 23.8% with four or more comorbidities (Fig. 4). In-hospital mortality was similar in HFs and in HFp patients if they presented zero or one comorbidity; it was dramatically greater in HFs than in HFp patients with two or more comorbidities (Fig. 4).
Association between the number of comorbidities and in-hospital mortality in HF patients
Multivariable logistic regression analyses, controlling for age, sex, and length of stay, revealed that the number of comorbidities is associated with significantly greater odds of in-hospital mortality (adjusted OR 1.42, 95% CI 1.25–1.62; p < 0.0001). The impact of comorbidity on in-hospital mortality appeared attributable to non-cardiac (adjusted OR 1.90, 95% CI 1.60–2.23; p < 0.0001) but not to cardiac comorbidity in HF patients (adjusted OR 0.97, 95% CI 0.80–1.18; p = 0.76).
Having two comorbidities was associated with six-fold higher odds of in-hospital mortality; the odds did not further increase with a greater number of comorbidities. Of note, as illustrated in Table 2, the increase in the number of non-cardiac—but not in the number of cardiac—comorbidities resulted in greater mortality in HF patients.
link