Article Text
PDF
Abstract
Objectives The objective of this study was to evaluate epidemiological characteristics, clinical course and outcome of mechanically ventilated non-surgical intensive care unit (ICU) patients, with the aim of improving the strategic planning of ICU capacities.
Design We conducted a retrospective observational cohort analysis. Data from mechanically ventilated intensive care patients were obtained by investigating electronic health records. The association between clinical parameters and ordinal scale data of clinical course was evaluated using Spearman correlation and Mann-Whitney U test. Relations between clinical parameters and in-hospital mortality rates were examined using binary logistic regression analysis.
Setting A single-centre study at the non-surgical ICU of the University Hospital of Frankfurt, Germany (tertiary care-level centre).
Participants All cases of critically ill adult patients in need of mechanical ventilation during the years 2013–2015 were included. In total, 932 cases were analysed.
Results From a total of 932 cases, 260 patients (27.9%) were transferred from peripheral ward, 224 patients (24.1%) were hospitalised via emergency rescue services, 211 patients (22.7%) were admitted via emergency room and 236 patients (25.3%) via various transfers. In 266 cases (28.5%), respiratory failure was the reason for ICU admission. The length of stay was higher in non-geriatric patients, patients with immunosuppression and haemato-oncological disease or those in need of renal replacement therapy. 431 patients died, which corresponds to an all-cause in-hospital mortality rate of 46.2%. 92 of 172 patients with presence of immunosuppression (53.5%), 111 of 186 patients (59.7%) with pre-existing haemato-oncological disease, 27 of 36 patients (75.0%) under extracorporeal membrane oxygenation (ECMO) therapy, and 182 of 246 patients (74.0%) undergoing renal replacement therapy died. In logistic regression analysis, these subgroups and older age were significantly associated with higher mortality rates.
Conclusions Respiratory failure was the main reason for ventilatory support at this non-surgical ICU. Immunosuppression, haemato-oncological diseases, the need for ECMO or renal replacement therapy and older age were associated with higher mortality.
- RESPIRATORY MEDICINE (see Thoracic Medicine)
- Adult intensive & critical care
- EPIDEMIOLOGY
- Risk management
- Adult intensive & critical care
- Respiratory infections
Data availability statement
Data are available upon reasonable request. The datasets used and analysed during the current study are available from the corresponding author on reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
- RESPIRATORY MEDICINE (see Thoracic Medicine)
- Adult intensive & critical care
- EPIDEMIOLOGY
- Risk management
- Adult intensive & critical care
- Respiratory infections
STRENGTHS AND LIMITATIONS OF THIS STUDY
We gathered detailed data about the clinical course of intensive care unit patients of a large academic hospital via an electronic patient file assessment.
The cohort is characterised by a large spectrum of diseases; all possible organ replacement therapies were available during the period of evaluation.
This study is subject to the usual limitation of a single-centre, retrospective study.
Data do not include validated risk scores and the evaluation of neurological outcome and course, including readmission to the hospital.
Introduction
The demand for intensive care capacity is increasing.1 2 In addition, the current COVID-19 pandemic was the biggest challenge for intensive and respiratory care medicine in decades. The pandemic has influenced intensive care unit (ICU) occupancy and clinical ICU course (length of stay and mortality).3 To create more capacities for critically ill patients with COVID-19, in many hospitals, elective procedures were cancelled, responsibilities for ICU patients were changed, and the differentiation between surgical and non-surgical wards was abrogated.4
The demand for ventilatory support for treatment of COVID-19-induced respiratory failure revealed the capacity of intensive care beds as the most critical resource for patient management.5 6 This was especially concerning in the beginning of the pandemic. During the first waves, prolonged ICU stay and extensive duration of ventilator therapy contributed to the scarcity of ICU capacity.7 8 An additional severe point of concern was the restriction of human resources, which limited the available ICU capacities and demanded new strategies for recruiting more staff.9
As a result of this increase in patients with critical disturbance of pulmonary gas exchange, our knowledge about the clinical course and management of acute and acute-on-chronic respiratory failure has increased rapidly.10 Above all, experience in acute respiratory distress syndrome (ARDS) treatment and differentiated ventilatory support, including extracorporeal treatment, advanced.11
Irrespective of the pandemic situation, in non-surgical ICUs, treatment of respiratory failure is the most important reason for admission to ICUs. Different pathological processes lead to disturbance of oxygenation or ventilation and manifest as respiratory failure in clinical settings. Independent of the COVID-19 outbreak, the main causes of impaired oxygenation are inflammatory processes of the lung parenchyma-like infectious pneumonia with its worst manifestation of disturbance of oxygenation, and the pneumonia-induced ARDS.12 Accordingly, adequate intensive care capacity is needed for effective management of respiratory failure. Therefore, planning of intensive care medicine demand must incorporate demographic factors as well as future pandemic and disaster preparation.
During the last two decades, several mathematical models with statistical or arithmetical methods were evaluated with the aim to improve prediction of length of stay, complication or mortality rates at ICUs.13 There is an extensive literature of studies which focuses on the association between scores, clinical parameters, length of hospital stay and mortality, respectively. A recent example might be the study of Aronsson Dannewitz et al who demonstrated an association between comorbidity of ICU patients and mortality.14 However, every form of statistical modelling needs detailed information about the specific patient characteristics. Prognosis and duration of ventilatory support are quite variable. This is of particular concern in patients with chronic pulmonary diseases or other severe comorbidities who need prolonged ventilatory support. Occasionally, this cohort must be transferred to specialised units for successful weaning from ventilatory support. Nevertheless, few patients with prolonged weaning failure rely on ventilatory support in the outpatient setting.15 The limited capacities in specialised weaning units aggravate the overload in intensive care settings.
The aim of this study was to evaluate the typical epidemiological characteristics, the clinical course and the outcome of mechanically ventilated non-surgical ICU patients, not confounded by the COVID-19 outbreak. The primary reason for respiratory failure was to be recorded and the underlying diagnosis to be documented. We employed a single-centre approach deliberately as it has been shown that a disadvantage of multicentre studies involving advanced treatment modalities is that physicians have different educational backgrounds, decide differently according to local protocols and beliefs, and judge patients differently, even though they may have been trained under comparable study protocols.16
Methods
Study design and setting
We conducted a retrospective single-centre cohort study at the non-surgical ICU and intermediate care unit of the University Hospital, Frankfurt/Main, Germany.
The University Hospital of the Goethe University Frankfurt/Main is a large academic hospital and as a tertiary care-level centre, offers the whole spectrum of intensive care therapy including all forms of organ replacement therapy (renal replacement therapy, extracorporeal liver assist devices, cardiac assist systems, extracorporeal membrane oxygenation (ECMO)). It is a centre for kidney and liver transplantation. Physicians working in the emergency room, intermediate and intensive care wards are trained following a structured approach, and clinical decision-making follows established protocols.
Procedures
We collected data from all invasively and non-invasively ventilated patients aged >18 years during the period 2013–2015. The data were obtained by investigating electronic health records of the hospital data system ‘AGFA-Orbis’.
The focus was to gather information about the way and cause of admission and the typical clinical course of non-surgical ICU patients. The parameters recorded are summarised in online supplemental table 1.
Supplemental material
Patients with structural lung diseases (interstitial lung disease, chronic obstructive pulmonary disease (COPD), cystic fibrosis) and geriatric patients (≥70 years) were analysed separately.
Data analysis
Data were documented by using the software Excel (Microsoft Home and Business V.2016). Statistical analysis was performed using the statistical software ‘Bias’ (V.11.12-05/2020, epsilon) and ‘SPSS’ (IBM SPSS Statistics V.27). Kolmogorov-Smirnov test was used for testing for normal distribution. Continuous data are presented as median and range (minimum and maximum) and categorical data as percentages.
All results are presented using common descriptive statistics. The association between several clinical parameters and clinical course was statistically evaluated by using Spearman correlation and Mann-Whitney U test. The association with mortality as a dependent variable was analysed by using binary logistic regression analysis.
Patient and public involvement
None.
Results
The patient population included 932 consecutive ICU cases with acute respiratory failure needing mechanical ventilation. The median age of the patients was 66 years (17–98 years). Three hundred fifty-six (38.2%) of all cases were patients aged >70 years. One hundred eighty-six patients (20.0%) had diagnosed haemato-oncological diseases. In 223 cases (23.9%), structural lung diseases like COPD, interstitial lung disease or cystic fibrosis pre-existed (see table 1).
Baseline characteristics: sociodemographic data, main diagnosis and comorbidities
A total of 27.9% of patients were transferred from peripheral ward, 24.1% of all subjects were hospitalised directly via emergency rescue services and 22.7% were admitted via the emergency room. Furthermore, patients were transferred from other hospitals, directly from the cardiac intervention suite, from operating theatres or from other ICUs of hospitals.
Acute/acute-on-chronic respiratory failure and sepsis were the reasons for initiating intermediate care unit and/or ICU admission. Two hundred sixty-six patients (28.5%) were admitted because of non-septic respiratory diseases including descending order acute exacerbation of COPD, pneumonia, aspiration, exacerbated interstitial lung disease and exacerbation of cystic fibrosis pulmonary disease (table 2 summarises cause and way of ICU admission).
Cause and way of ICU admission
Of the 932 individuals studied, 773 patients (82.9%) had to be intubated and received invasive ventilation, whereas 159 (17.0%) patients were ventilated non-invasively. One hundred twenty-five (13.4%) of the invasively ventilated patients were intubated after failure of non-invasive ventilation. Most patients who had non-invasive ventilation failure were admitted with respiratory failure caused by acute exacerbation of COPD (16.8%), pneumonia (15.2%) or sepsis of respiratory origin (16.2%).
We found a median time of ventilation of 59 hours (range <1–2348 hours) and detected a median length of stay in hospital of 14 days (range <1–214 days).
Altogether, 94 patients (10.0%) underwent tracheotomy, primarily in the form of percutaneous dilatational tracheotomy (38 patients; 40.9% of all tracheostoma patients).
Thirty-one (3.3%) of all ventilated patients were directly transferred to specialised weaning units in external hospitals.
Seventy-three patients were discharged with home ventilatory support (7.8%). Of these, 16 patients needed invasive ventilation in an outpatient setting (20.5%). Table 3 shows the most important characteristics of mechanical ventilation and reflects important parameters of clinical stay.
Characteristics of mechanical ventilation and outcomes (all cases)
The proportion of patients who needed non-invasive ventilation was 64.6%.
The in-hospital mortality in this unselected series was 46.2%. Main causes of admission in deceased patients were out-of-hospital resuscitation (18.9%), respiratory failure (16.9%), and sepsis of pulmonary and non-pulmonary origin (15.3% and 14.6%, respectively). Online supplemental table 2 gives the summary of causes of admission in deceased patients.
Taking a closer look at the subgroup of 223 patients with structural lung disease, 35.4% were non-invasively and 64.6% were invasively ventilated. Seven patients of this subgroup received ECMO therapy. In-hospital mortality in this cohort was 27.4%. Online supplemental table 3 gives an analysis of this subgroup of patients with known structural lung disease and reflects the course of mechanical ventilation in these cases.
Of the total of 356 geriatric patients (aged >70 years), 22.8% were non-invasively and 77.2% were invasively ventilated. The in-hospital mortality in geriatric patients was 47.5%. The course of mechanical ventilation and outcome of geriatric patients is presented in table 4.
Characteristics of mechanical ventilation and outcomes in geriatric patients
The results of Spearman correlation analysis are presented in online supplemental table 4. A statistically significant association between immunosuppression, haemato-oncological disease, renal replacement therapy and length of hospital stay, respectively, with in-hospital mortality was observed.
The comparison of the duration of ventilatory support and the length of hospital stay between these subgroups was further analysed by using the Mann-Whitney U test.
The results revealed increased duration of ventilatory support in patients who received a renal replacement therapy (p<0.01), and patients undergoing ECMO therapy (p=0.013). The median of duration of ventilatory support was lower in geriatric patients compared with non-geriatric patients (p<0.01).
The length hospital of stay was longer in immunosuppressed patients (p=0.01), patients with haemato-oncological disease (p=0.01) and patients who received renal replacement therapy (p<0.01) compared with patients without these features. The median of hospital stay was lower in geriatric patients (p=0.016) compared with non-geriatric patients (see online supplemental table 5).
The binary logistic regression analysis using in-hospital mortality as the dependent variable revealed a significant association between the presence of immunosuppression (p=0.048), haemato-oncological disease (p=0.006), use of renal replacement therapy (p<0.001) or ECMO therapy (p=0.005) and age (p=0.016) (see online supplemental table 6). Mortality rate was 53.5% in patients with presence of immunosuppression, 59.7% in patients with pre-existing haemato-oncological disease, 75.0% in case of ECMO and 74.0% in case of renal replacement therapy.
Discussion
This study illustrates some important aspects of medical intensive care demand, which might support planning of intensive and respiratory care resources.
In contrast to surgical intensive care,17 at non-surgical ICUs, the number of elective admissions is low. We observed that most patients were transferred to the ICU with an emergency admission via peripheral ward, via emergency rescue service or via the emergency room.
To improve planning of non-surgical intensive care demand and to secure adequate ICU capacity for this cohort, detailed knowledge about patient characteristics and typical clinical course of each disease is warranted.18
Our results revealed acute and acute-on-chronic respiratory failure in nearly half of cases as the most important cause leading to ICU admission, with the necessity of non-invasive and invasive ventilation.
First, pneumonia including pneumogenic sepsis causes most cases of acute respiratory failure. This result is in keeping with previous studies, which reported the high risk of serious disturbance of respiratory homeostasis induced by community-acquired pneumonia.19 This result provides support for considering community-acquired pneumonia as a serious pulmonary emergency.12 20 Another relevant cause of ICU admission and necessity for mandatory ventilatory support is acute exacerbation of COPD. In our cohort, more than 60% of these patients received non-invasive ventilation, which is in accordance with current recommendations for treatment of acute respiratory failure in this subgroup.21
Because of epidemiological developments, geriatric patients represented a relevant group of all patients evaluated. The sociodemographic distribution in our cohort was like in former findings. We could demonstrate an association between age and increased risk of in-hospital mortality. Referring to this, earlier studies reported diverging results. Wang et al, who examined predictive factors of in-hospital mortality in ventilated ICU patients, reported about a cohort with a mean age of 69.8 years. The authors could not show a statistical significant association between age and in-hospital mortality.22 Data from a multicentric, prospective cohort study of Flaatten et al indicate that frailty had a small impact on ICU survival.23 According to our results, a more recent study of Guidet et al24 revealed age as the most important predictor of 1-month mortality in ICU patients.
Furthermore, the mentioned study of Guidet et al24 showed that ICU admission diagnosis, Sequential Organ Failure Assessment (SOFA) score and Clinical Frailty Scale are further relevant predictors of 1-month mortality. This is consistent with our data showing a higher mortality in patients under ECMO or renal replacement therapy.
We also showed that comorbidity might contribute to the relatively high mortality observed in our cohort. The concurrence of haemato-oncological disease and immunosuppression was especially associated with high mortality, length of stay and duration of ventilatory support in our cohort.
This is in accordance with recent data of de Vries et al, who analysed the outcome of patients with haemato-oncological disease in five university hospitals in the Netherlands.25 This study demonstrated an increased risk of mortality in the case of multiorgan failure.
Moreover, we found an increased length of stay in this cohort. Hospitals with emphasis on haemato-oncological therapy units should consider this relevant demand for ICU capacity.
Similar data exist for immunosuppressed patients. Raymond et al demonstrated in a recent observational study a threefold increased mortality in immunosuppressed patients with influenza at the ICU.26
In our cohort, the pre-existence of structural lung disease was not correlated with length of clinical stay or mortality. This observation is in accordance with Gadre et al, who reported an ICU mortality of 25% in invasively ventilated patients with COPD with acute respiratory failure.27
It is remarkable that in spite of enormous progress in intensive care medicine, findings of mortality are comparable with data of Seneff et al, who observed in 1995 in a prospective, multicentre, inception cohort study including 362 ICU admissions, an in-hospital mortality of 24% in patients with COPD, rapidly increasing in patients older than 65 years up to 30% and higher, depending on length of hospital stay.28
What conclusions can be drawn from these results?
The relevant group of geriatric patients, which will further increase in the future, leads us to the conclusion that the sociodemographic development will fundamentally impact future evaluation of ICU performance and statistics.
Moreover, our data let us assume that the quality of comorbidity might influence the clinical course and mortality risk more than comorbidity or frailty indices. Therefore, the knowledge about the therapeutic spectrum of the hospital will help to estimate the ICU demand. This might be of vital consequence for hospitals with a high proportion of haemato-oncological or immunocompromised patients.
What are the consequences of the clinical course following intensive care therapy?
Only 352 (37.8%) patients were discharged home, and 149 patients (33.5%) were transferred to different medical institutions (other hospitals, weaning units or rehabilitation). These results indicate the strong need for prolonged healthcare support after ICU discharge. A relevant number of patients were discharged with home ventilatory support. Few patients were discharged with invasive mandatory ventilation support. The recent results of the WEAN SAFE Study confirmed the considerable proportion of patients with difficultly weaning. However, regional differences must be considered to identify the local demand for specialised weaning units. Therefore, a structured assessment of the characteristics of the local patient cohorts might be helpful.
A further challenge in public health is the increasing number of patients with invasive home ventilation support. It is a special feature of German outpatient therapy that patients undergo invasive ventilatory support in an outpatient setting. In situations of scarce intensive care capacity like in pandemic situations, this type of care would reduce the load of intensive care. However, capacity in nursing homes with expertise in caring for ventilated patients is low and the cost for outpatient treatment of this population is high. Nonetheless, the tendency for home ventilation support is increasing. At the present time, in Germany, an estimated 20 000 patients are ventilated invasively in a non-clinical setting.29 In contrast to this rising number, there is a lack of systematic registration of out-of-hospital invasive ventilation capacities. Especially in the current SARS-CoV-2 pandemic, knowledge about these capacities would have been extremely helpful.
A first step for systematic evaluation of out-of-hospital intensive care capacities is the current implementation of the registry of the German Society for Mechanical Home Ventilation (‘Deutsche Gesellschaft für außerklinische Beatmung’).
Finally, our results showed a further important aspect of management of emergency and critical care patients. One-third of patients were transferred via peripheral ward. Most of these subjects were transferred to hospital via the emergency room or the outpatient clinic before reaching the peripheral ward. Obviously, these patients developed serious deterioration after initial stabilisation of the ventilatory and circulatory situation. Similar observations were made by Wardi et al in patients with sepsis: 23.5% of patients with severe sepsis who were admitted from the emergency room to a non-ICU ward had to be transferred to the ICU within 48 hours because of serious deterioration.30 This illustrates the high importance of vigilance of medical staff in non-ICU or non-immediate care units. This observation is in accordance with the recommendations of current pneumonia guidelines, which advise on close clinical observation of hospitalised patients during the first 72 hours.20 31
Accordingly, first, results indicate that concepts of integrated respiratory failure or lung attack units could improve assessment and management of patients with acute/acute-on-chronic respiratory failure.32 These specialised units might prevent emergency transfers from peripheral wards to ICUs.
Limitations
This study is subject to the inherent limitations of any retrospective-based data. Accordingly, there was a lack of structured registration of validated risk scores (like Acute Physiology And Chronic Health Evaluation (APACHE) II score, SOFA score), which should be included in further studies with focus on ICU course and outcome.
Another drawback is the single-centre setting. There may be a selection bias due to the level of care (tertiary centre), which does not reflect the other average intensive care units. Presumably, the proportion of patients with acute liver failure or acute-on-chronic liver failure, and particularly of patients with exacerbation of cystic fibrosis, must be attributed to the specific focus on these diseases at our hospital.
Furthermore, the data gathered were not suitable to obtain information about patient outcome (including quality of life or neurological outcome) and we were not able to gather data on follow-up mortality.
Even though we know that the patient’s quality of life is occasionally disturbed for months after recovery, there is still a lack of knowledge on which factors determine the long-term outcome during ICU stay. The results of the international DACAPO Study impressively demonstrated this issue.33
Szakmany et al demonstrated in a recent retrospective observational study that one-fifth of ICU survivors died within 1 year.34 Consequently, further studies which focus on ICU course and outcome should collect these follow-up data.
Conclusion
In non-surgical ICUs, most transfers were emergency admissions with respiratory failure as the main reason leading to mandatory ventilation in a non-surgical ICU.
Most of these patients were admitted directly or indirectly via the emergency room to the ICU. Hence, the establishment of respiratory failure or lung attack units should be further advocated.
The high frequency of comorbidities like immunosuppression or haemato-oncological diseases explains the considerable high mortality in our cohort, long in-hospital stay and extended time of ventilation. Having in mind the progress in haematological and oncological therapy in the last decade, these facts underscore the increasing demand for intensive care capacity in the future.
Our retrospective data could be underscored by an integrative prospective registry of ventilated patients. Registries such as the registry of the German Interdisciplinary Association for Intensive and Emergency Medicine focusing on intensive care capacities and the recently launched German registry of outpatient respiratory care capacities are encouraging first steps into this direction.35
Data availability statement
Data are available upon reasonable request. The datasets used and analysed during the current study are available from the corresponding author on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
The study protocol was approved by the Local Ethics Committee of the University Hospital of Goethe University Frankfurt (study number 465/15).
Acknowledgments
We thank Matthias Angersbach for his support in data capturing. Furthermore, we thank Professor Dr Eva Herrmann for her support in statistical analysis.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors Conceptualisation—AG. Formal analysis—AG. Writing (original draft preparation)—AG. Guarantor_AG. Writing (review and editing)—JB, K-HP and GGUR. All authors read and approved the final manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Pre-registration Internal Reference No: NUTH R&D: 8822; Ethics Ref: 19/NE/0027; IRAS project ID: 24743718-23; Trial registration: International Standard Randomised Controlled Trial Number: ISRCTN47889316 https://doi-org.ezproxy.u-pec.fr/10.1186/ISRCTN47889316
Competing interests AG reports personal fees from GSK and Novartis for lectures. GGUR reports personal fees from AstraZeneca, Berlin Chemie, BMS, Boehringer Ingelheim, Chiesi, Essex Pharma, Grifols, GSK, Insmed, MSD, Roche, Solvay, Takeda, Novartis, Pfizer and Vertex for consultancy during advisory board meetings, as well as personal fees from AstraZeneca, Berlin Chemie, BMS, Boehringer Ingelheim, Chiesi, Essex Pharma, Grifols, GSK, Insmed, MSD, Roche, Solvay, Takeda, Novartis, Pfizer and Vertex for lectures including service on speakers’ bureaus. JB and K-HP have no potential conflict of interest.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.