Study population

94,108,335 hospital cases between 2015 and 2019 were screened of which in 35,542,031 cases surgery was performed. Cases with age under 18 years (n = 1,618,823), prior lung transplantation (n = 5,054) or prior heart-lung transplantation (n = 187) were excluded. Of the remaining 33,917,967 surgical cases, 1,339,079 (3.9%) cases were diagnosed with COPD and 1,642,377 cases (4.8%) suffered from any perioperative organ injury according to our definition and were analysed in this study (Fig. 1). Most perioperative organ injuries were observed in operations on the digestive (32.5%) and on the musculoskeletal system (25.3%) (Additional File 3).

Fig. 1

Flow chart of patient inclusion. Flow chart of patient inclusion of a population-based retrospective analysis investigating the impact of COPD on in-hospital mortality, hospital length of stay (HLOS) and ventilation time (VT) in 1,642,377 surgical cases with perioperative organ failure

Table 1 summarises the characteristics of the study population suffering from perioperative organ injury. The median age of all patients was 76 years (IQR, 66–82), and most of the patients were of male gender (59.0%). The patients suffered most frequently from congestive heart failure (n = 569,922; 34.7%), followed by renal disease (n = 549,690; 33.5%). The median CCI was 3 (IQR, 2–5).

Table 1 Characteristics and outcome of 1,642,377 hospitalised surgical patients with chronic obstructive pulmonary disease (COPD) and without COPD suffering from any perioperative organ injury

The in-hospital mortality was 16.3% for patients with any perioperative organ injury, the median HLOS was 17 days (IQR, 10–29) and the median VT was 135 h (IQR, 33–404).

COPD and any perioperative organ injury

10.8% (n = 177,070) of the study population (n = 1,642,377) suffered from COPD (Table 1).

The median age in COPD patients was 75 years (IQR, 67–81) and most of the patients were of male gender (64.1%). Patients with COPD were of poorer health condition represented by a higher CCI compared to patients without COPD (4 (IQR, 3–6) vs. 3 (IQR, 1–5), p < 0.001). Patients with and without COPD suffered most frequently from congestive heart failure (50.4% vs. 32.9%, p < 0.001), followed by renal disease (41.3% vs. 32.5%, p < 0.001) (Table 1).

Impact of COPD on in-hospital mortality, HLOS and VT

COPD patients demonstrated a higher in-hospital mortality compared to patients without COPD (20.6% vs. 15.8%, p < 0.001), when suffering from any perioperative organ injury. In addition, median HLOS (21 days (IQR, 12–34) vs. 16 days (IQR, 10–28, p < 0.001)) and median VT (199 h (IQR, 43–547) vs. 125 h (IQR, 32–379, p < 0.001)) were also higher in patients with COPD compared to patients without COPD (Table 1).

COPD and different types of perioperative organ injury

In COPD patients with perioperative organ injury, the analysed perioperative organ injuries were observed in the following descending order: AKI (57.8%), delirium (36.5%), AMI (12.8%), stroke (8.5%), ALI (4.3%) and ARDS (3.1%). Within the perioperative organ injuries analysed, patients diagnosed with COPD suffer more often from AKI (57.8% vs. 50.4%, p < 0.001), delirium (36.5% vs. 34.6%, p < 0.001) and ARDS (3.1% vs. 2.3%, p < 0.001) compared to patients without COPD. In contrast, patients without COPD suffered more often from stroke (13.3% vs. 8.5%, p < 0.001) and ALI (5.4% vs. 4.3%, p < 0.001) (Table 2). AMI was not significantly different between groups (12.8% vs. 13.0%, p = 0.088) (Table 2).

Table 2 Outcomes of hospitalised surgical patients with chronic obstructive pulmonary disease (COPD) (n = 177,070) and without COPD (n = 1,465,307) suffering from each type of perioperative organ injuryImpact of COPD on in-hospital mortality, HLOS and VT in patients with different types of perioperative organ injury

Patients diagnosed with COPD demonstrated a higher in-hospital mortality compared to patients without COPD in all examined different types of perioperative organ injuries (Table 2; Fig. 2), except in ARDS (40.9% vs. 45.4%, p < 0.001). The highest in-hospital mortality was observed in perioperative ALI (54.2%), the lowest in perioperative delirium (15.5%).

Fig. 2

In-hospital mortality of patients with and without COPD. In-hospital mortality of patients with (grey) and without (green) COPD in different types of perioperative organ injuries, namely: delirium, stroke, acute myocardial infarction (AMI), acute respiratory distress syndrome (ARDS), acute liver injury (ALI) and acute kidney injury (AKI)

Median HLOS was longer in all examined different perioperative organ injuries when suffering from COPD compared to patients without COPD (Table 2). The longest HLOS was observed in ARDS (30 days (IQR, 18–48)), followed by delirium (24 days (IQR, 14–39)), the lowest in perioperative acute myocardial infarction (18 days (IQR, 11–30)) and perioperative acute liver injury (18 days (IQR, 9–34)) (Table 2).

In addition, median VT was also prolonged in COPD patients in all examined different perioperative organ injuries compared to patients without COPD (Table 2). The most prolonged median VT was observed in ARDS (478 h (IQR, 248–764)), followed by perioperative stroke 291 h (IQR, 75–590)). The lowest in perioperative acute myocardial infarction (121 h, (IQR, 25–438) (Table 2).

Regression model analyses

Table 3 summarises the key findings of regression model analyses. The complete results are available as supplementary material (Additional File 4–24).

Table 3 Associations of COPD with mortality, hospital length of stay and perioperative ventilation time in hospitalised surgical patients with perioperative organ injury from different multivariable regression modelsIn-hospital mortality

COPD was associated with a higher odds of in-hospital mortality (odds ratio (OR), 1.19; 95% confidence interval (CI): 1.18–1.21, p < 0.001), when suffering from any perioperative organ injury.

For all different perioperative organ injuries examined, COPD was associated with higher odds of in-hospital mortality, except for ARDS (OR, 0.82; 95% CI: 0.77–0.88, p < 0.001).

Within the perioperative organ injuries analysed, COPD was associated with the highest risk for in-hospital mortality when suffering from perioperative delirium (OR, 1.30, 95% CI: 1.27–1.34, p < 0.001) (Table 3).

HLOS

COPD was associated with a longer HLOS (beta, 2.62 days; 95% CI: 2.51–2.73, p < 0.001), when suffering from any perioperative organ injury.

Further regression model analyses showed a longer HLOS for all different perioperative organ injuries examined (Table 3). The longest HLOS was observed for COPD, when suffering from perioperative delirium (beta, 2.80 days; 95% CI: 2.60–2.99; p < 0.001) (Table 3).

VT

In addition, COPD was also associated with a longer VT (beta, 107.47 h, 95% CI: 104.01-110.93, p < 0.001), when suffering from any perioperative organ injury.

For all different perioperative organ injuries analysed, COPD was associated with a longer VT, with the longest VT in perioperative delirium (beta, 143.45 h, 95% CI: 137.67-149.22, p < 0.001).

Confounders

We addressed different confounders in the regression model analysis. The regression model analysis identified SIRS/sepsis (OR, 4.52; 95% CI: 4.47–4.56, p < 0.001) and moderate to severe liver disease (OR, 3.80; 95% CI: 3.72–3.89, p < 0.001) as the leading risk factors for in-hospital mortality (Additional File 4) when suffering from any perioperative organ injury. SIRS/Sepsis increases the odds for in-hospital mortality by almost 4.5 folds and moderate to severe liver disease by almost 4.0 folds. When considering different types of perioperative organ injury, perioperative liver injury is one of the leading risk factors for in-hospital mortality (Additional File 5–10). Emergency hospital admission was associated with a higher odds of in-hospital mortality (OR 1.23; 95% CI: 1.22–1.24, p < 0.001), when suffering from any perioperative organ injury. Within the different perioperative organ injuries examined, emergency hospital admission was associated with a higher odds of in-hospital mortality when suffering from perioperative delirium (OR, 1.31, 95% CI: 1.28–1.33, p < 0.001), stroke (OR, 1.20, 95% CI: 1.17–1.23, p < 0.001), AMI (OR, 1.22, 95% CI: 1.19–1.25, p < 0.001) and AKI (OR, 1.15, 95% CI: 1.14–1.16, p < 0.001).

In addition, we found that neither age (OR: 1.02; 95% CI: 1.02–1.02, p < 0.001) nor chronic renal disease (OR: 0.94; 95% CI: 0.93–0.95, p < 0.001) or diabetes with (OR: 0.85; 95% CI: 0.84–0.86, p < 0.001) or without complications (OR: 0.93; 95% CI: 0.93–0.95, p < 0.001) is associated with a considerably risk of in-hospital mortality in any perioperative organ injury (Additional File 4).