Abstract
Background
In trauma patients, unplanned intensive care unit (ICU) readmission (UIR) is associated with poor clinical outcomes. In this study, we aimed to analyze associated factors for UIR in trauma patients.
Methods
This retrospective study was conducted on trauma patients admitted to the ICU at a trauma center from January 2016 to December 2022. Clinical information at admission, the first ICU hospitalization, first discharge from the ICU, and reasons for readmission were collected. Patients who were successfully discharge from the ICU were compared to UIR patients. Logistic regression was performed to determine the factors with a significant impact on ICU readmission.
Results
Here, 5,529 patients were admitted to the ICU over 7 years, and 212 patients (3.8%) experienced UIR. Among patients who experienced UIR, 9 (4.2%) died. In the UIR patients, hospital stay (20 days [interquartile range, 13–35] vs. 45 days [28–67], P<0.001), total ICU stay (5 days [3–11] vs. 17 days [9–35], P<0.001), and complications during the first ICU hospitalization were significantly higher. The most common reason for UIR was respiratory problem (53.8%). In multivariable analysis, cervical spine operation during the first ICU hospitalization (odds ratio, 6.56; 95% CI, 3.62–11.91; P<0.001), renal replacement therapy (RRT; 3.52, 2.06–5.99, P<0.001), and massive blood transfusion protocol (MTP; 1.74, 1.08–2.81, P=0.023) were most highly related with UIR.
Conclusions
Because UIR patients had poor outcomes, trauma patients who underwent cervical spine operation, RRT, or MTP require monitoring in the general ward, especially for respiratory problems.
Keywords: intensive care units, patient readmission, trauma centers, risk factors
INTRODUCTION
Severe trauma patients are usually treated in the intensive care unit (ICU) for multiple reasons including shock and ventilator use. There are many things to consider when determining whether to discharge a patient from the ICU including the patient's medical condition and resources on the general ward. However, unplanned ICU readmissions (UIRs) occur, causing negative patient outcomes [1-4]. In previous studies, the mortality rate for UIR was approximately five times higher and hospital length of stay was two times higher compared with patients without UIR [2,5-8]. Therefore, prevention of UIR and knowledge of the risk factors causing UIR in trauma patients are important. UIR is considered a quality management indicator in trauma centers [3,9-11]. Nevertheless, various causes of UIR were identified in each study, and there was a lack of consensus on the most critical factor or factors.
To improve the preventable trauma death rate in Korea, regional trauma centers have been constructed for centralization of trauma patients. Our trauma center is a regional center and recently joined a trauma quality improvement program. Hence, the UIR data were collected prospectively and could be utilized for quality control. In this study, we aimed to find associated factors and outcomes of UIR in trauma patients. To our knowledge, this is the first study of UIR in a trauma center in Korea.
MATERIALS AND METHODS
This study was approved by the Institutional Review Board of Ajou University Hospital (No. AJOUIRB-DB-2024-56), and all methods were conducted in accordance with the Declaration of Helsinki. According to our Institutional Review Board policy informed consent of the research was waived.
Study Protocol
We retrospectively analyzed trauma patients admitted to the ICU at our regional trauma center from January 2016 to December 2022. The UIR group included all patients who experienced UIR within the hospital stay. For excluding non-severe trauma patients who had a very low likelihood of UIR, patients with Injury Severity Score (ISS) less than 15 points and ICU stay less than 2 days were excluded. Patients who died or were transferred to another hospital at the time of first admission to the ICU were also excluded. An additional exclusion was made for one patient who was readmitted to the ICU not for medical reasons but because no caregiver was available on general ward. Patients were divided into the UIR group and the successfully discharged group (Figure 1). General characteristics and clinical outcome were compared. For finding factors associated with UIR, logistic regression was used. The reason for UIR was also reviewed, and only the most impactful cause was selected.
Figure 1.
Patient inclusion/exclusion flowchart. ICU: intensive care unit; ISS: injury Severity Score.
Data Collection
The collected data included patient age, sex, occurrence rate of UIR, Glasgow Coma Scale (GCS) score upon arrival, whether the patient was in shock (systolic blood pressure <90 mm Hg), ISS, Revised Trauma Score, Trauma Score and Injury Severity Score (probability of survival), Abbreviated Injury Scale (AIS) for each body region, procedures, and surgeries during the first ICU hospitalization. Underlying diseases, hospital stay, ICU stay, days on ventilator, mortality, and complications were also investigated. In addition, vital signs, length of stay, use of vasopressor agents, massive blood transfusion protocol (MTP), and renal replacement therapy (RRT) were recorded at the time of first discharge from the ICU.
In our center, the MTP follows the ABC protocol (penetrating mechanism, systolic blood pressure ≤90 mm Hg, heart rate ≥120/min, positive Focused Assessment With Sonography in Trauma) and is employed if the patient meets at least 2 of these). UIR was defined as “patients admitted to the ICU after initial transfer to the floor and/or patients with an unplanned return to the ICU after initial discharge, excluding planned postoperative ICU stay” by the Nation Trauma Data dictionary and confirmed by the responsible physician in our center. Other hospital events such as acute kidney injury (AKI) or adult respiratory distress syndrome (ARDS) were followed using definitions from the same dictionary [12]. The reason for UIR was determined by panel review.
The classification criteria for cause analysis of UIR are as follows. (1) Respiratory: cases with hypoxemia and hypercapnia due to airway or lung problems (e.g., pneumonia, excessive secretion of airway, pneumothorax, effusion). (2) Septic: cases with septic shock (septic shock due to pneumonia is classified under respiratory cause). (3) Neurogenic: cases with seizure or brain lesion objectively confirmed by image studies. (4) Hypovolemic: cases with bleeding or dehydration. (5) Cardiogenic: cases with heart problems such as arrhythmia or heart failure, confirmed by electrocardiogram or echocardiography. (6) Continuous RRT (CRRT): cases where CRRT was required for severe electrolyte imbalance and fluid volume control. (7) Psychiatric: cases involving the use of medication that requires intensive monitoring or absence of a caregiver. (8) Arrest: cases found in cardiac arrest. (9) Anaphylaxis: cases with anaphylactic shock.
Statistical Analysis
All data were stored in a spreadsheet using Microsoft Excel and analyzed using R package MoonBook 0.3.1 (https://cran.r-project.org/web/packages/moonBook/index.html). Data are presented as mean (standard deviation) or median (interquartile range) after normality test. Continuous variables were analyzed using Student t-test, and categorical variables were analyzed using the chi-square or Fisher’s exact tests. Results were considered significant at P<0.05. Multivariate logistic regression was performed with potential associated factors that had significant differences in univariable analysis.
RESULTS
A total of 11,607 patients was admitted to the ICU of our trauma center between January 1, 2016 and December 31, 2022 (Figure 1). Of these, 6,078 patients were excluded: 4,974 had an ISS less than 15, 558 had an ICU stay less than 2 days, 545 were not transferred to a general ward, and 1 had a non-medical cause. As a result, 5,529 patients were finally included, of which 212 (3.8%) were in the UIR group and 5,317 (96.2%) were in the successfully discharged group. Among the UIR group, nine patients (4.2%) died, and 203 (95.8%) were successfully discharged following the second ICU admission.
The UIR group was significantly older (61.5 years [50.5–73.5] vs. 53.0 years [36.0–63.0], P<0.001) and had relatively more male patients (175 [82.5%] vs. 4,087 [76.9%], P=0.065). There was no difference in GCS score at the time of arrival. However, the UIR group had more severe injury with a significantly larger proportion of cases of shock (systolic blood pressure <90 mm Hg), MTP, and higher ISS than the successfully discharged group. Underlying diseases were significantly higher in the UIR group and included hypertension (79 [37.3%] vs. 1,239 [23.3%], P<0.001), diabetes mellitus (52 [23.2%] vs. 717 [13.5%], P<0.001), and cerebrovascular disease (18 [8.0%] vs. 198 [3.7%], P=0.002). In the UIR group, the AIS score of the overall head and neck was significantly higher (4.0 [3.0–5.0] vs. 3.0 [3.0–4.0], P=0.026) (Table 1).
Table 1.
Baseline characteristics according to UIR
| Variable | Non-UIR (n=5,317) | UIR (n=212) | Total (n=5,529) | P-value |
|---|---|---|---|---|
| Age (yr) | 53.0 (36.0–63.0) | 61.5 (50.5–73.5) | 53.0 (37.0–64.0) | <0.001 |
| Sex (male:female, %) | 76.9:23.1 | 82.5:17.5 | 77.1: 22.9 | 0.065 |
| On arrival | ||||
| GCS | 14 (11.0–15.0) | 14 (11.0–15.0) | 14 (11.0–15.0) | 0.744 |
| SBP <90 mm Hg | 1,204 (22.7) | 77 (34.4) | 1,281 (23.2) | <0.001 |
| MTP | 279 (5.3) | 25 (11.2) | 304 (5.5) | <0.001 |
| Mechanism of injury | 0.898 | |||
| Blunt | 5,155 (97.0) | 207 (97.6) | 5,362 (97.0) | |
| Penetrating | 140 (2.6) | 5 (2.2) | 145 (2.6) | |
| Others | 22 (0.5) | 0 (0.0) | 22 (0.4) | |
| Underlying disease | ||||
| Hypertension | 1,239 (23.3) | 79 (37.3) | 1,318 (23.8) | <0.001 |
| Diabetes mellitus | 717 (13.5) | 52 (23.2) | 769 (13.9) | <0.001 |
| CVD | 198 (3.7) | 18 (8.0) | 216 (3.9) | 0.002 |
| Angina | 56 (1.1) | 6 (2.7) | 62 (1.1) | 0.053 |
| Trauma score | ||||
| ISS | 24.9±8.7 | 27.0±9.4 | 25.0±8.7 | <0.001 |
| RTS | 7.34±0.95 | 7.18±1.17 | 7.3±0.96 | 0.247 |
| TRISS | 0.89±0.16 | 0.86±0.16 | 0.89±0.16 | <0.001 |
| AIS | ||||
| Head–neck | 3.0 (3.0–4.0) | 4.0 (3.0–5.0) | 3.0 (3.0–4.0) | 0.026 |
| Face | 2.0 (0.0–2.0) | 1.0 (0.0–2.0) | 2.0 (0.0–2.0) | 0.388 |
| Chest | 3.0 (3.0–3.0) | 3.0 (3.0–3.0) | 3.0 (3.0–3.0) | 0.095 |
| Abdomen–pelvis | 2.0 (2.0–3.0) | 2.0 (2.0–3.0) | 2.0 (2.0–3.0) | 0.850 |
| Extremity | 2.0 (2.0–3.0) | 2.0 (2.0–3.0) | 2.0 (2.0–3.0) | 0.786 |
Values are presented as median (interquartile range), number (%), or mean±standard deviation.
UIR: unplanned intensive care unit readmission; GCS: Glasgow Coma Scale; SBP: systolic blood pressure; MTP: massive blood transfusion protocol; CVD: cerebrovascular disease; ISS: Injury Severity Score; RTS: revised trauma score; TRISS: trauma score and Injury Severity Score; AIS: Abbreviated Injury Scale.
The UIR group also had more frequent complications and longer overall hospital stay (Table 2) as well as total ICU length of stay (17 days [9–35] vs. 5 days [3–11], P<0.001). Complications during the first ICU hospitalization—including pneumonia, unplanned intubation, AKI, and ARDS—were significantly higher in the UIR group. Although the vital signs and laboratory findings at ICU discharge showed significant differences in some variables, all values were within the clinically normal range and no intervention was required (Table 3). The proportion of patients who used vasopressors (180 [80.4%] vs. 2,449 [46.2%], P<0.001) or RRT (20 [8.9%] vs. 104 [2.0%], P<0.001) during the first ICU stay was significantly higher in the UIR group. The frequency of operation for cervical spine during the first ICU hospitalization in the UIR group was also significantly higher (17 [15.7%] vs. 85 [3.7%], P<0.001). In the UIR group, there was no statistical difference in surgeries for other areas.
Table 2.
Clinical results according to UIR
| Variable | Non-UIR (n=5,317) | UIR (n=212) | Total (n=5,529) | P-value |
|---|---|---|---|---|
| Outcome (total hospitalization) | ||||
| Hospital day (day) | 20 (13–35) | 45.0 (28–67) | 21.0 (13–37) | <0.001 |
| ICU LOS (day) | 5 (3–11) | 17.0 (9–35) | 5.0 (3–12) | <0.001 |
| Days of ventilator use (day) | 0 (0–5) | 6.5 (0–17) | 0 (0–5) | <0.001 |
| Mortality | 0 | 9 (4.2) | 9 (0.2) | <0.001 |
| Complication (during the first ICU hospitalization) | ||||
| Pneumonia | 539 (10.1) | 36 (17.0) | 575 (10.4) | 0.003 |
| Unplanned intubation | 144 (2.7) | 18 (8.5) | 162 (2.9) | <0.001 |
| AKI | 106 (2.0) | 15 (7.1) | 121 (2.2) | <0.001 |
| ARDS | 83 (1.6) | 13 (6.1) | 96 (1.7) | <0.001 |
Values are presented as median (interquartile range) or number (%).
UIR: unplanned intensive care unit readmission; ICU: intensive care unit; LOS: length of stay; AKI: acute kidney injury; ARDS: acute respiratory distress syndrome.
Table 3.
Characteristics at the initial ICU discharge according to UIR
| Variable | Non-UIR (n=5,317) | UIR (n=212) | Total (n=5,529) | P-value |
|---|---|---|---|---|
| Last vital signs in ICU | ||||
| SBP (mm Hg) | 133 (119.0 to 148.0) | 137 (118.5 to 154.0) | 133 (119.0 to 148.0) | 0.040 |
| Pulse rate (/min) | 86 (74.0 to 98.0) | 86 (75.0 to 99.5) | 86.0 (74.0 to 98.0) | 0.390 |
| Saturation (%) | 99 (98.0 to 100.0) | 100 (98.0 to 100.0) | 99.0 (98.0 to 100.0) | 0.585 |
| Hospital day (during the first ICU hospitalization) | ||||
| ICU LOS | 4.0 (2.0 to 9.0) | 5.0 (2.0 to 13.0) | 4.0 (2.0 to 9.0) | 0.002 |
| Last laboratory tests in ICU | ||||
| Hemoglobin (g/dl) | 10.0 (8.9 to 11.6) | 9.1 (8.3 to 10.6) | 10.0 (8.8 to 11.5) | <0.001 |
| White blood cell (×103/mm3) | 8.6 (6.8 to 10.9) | 9.1 (7.1 to 11.2) | 8.6 (6.8 to 11.0) | 0.035 |
| Platelet (×103/mm3) | 179.0 (131.0 to 264.5) | 180.5 (126.5 to 261.5) | 179.0 (130.0 to 264.0) | 0.846 |
| Lactate (/mmol) | 1.1 (0.9 to 1.6) | 1.3 (1.0 to 1.6) | 1.1 (0.9 to 1.6) | 0.634 |
| Base excess (/mmol) | 0.9 (–0.9 to 2.7) | 0.7 (–1.0 to 2.6) | 0.9 (–0.9 to 2.7) | 0.582 |
| INR | 1.12 (1.06 to 1.19) | 1.13 (1.07 to 1.25) | 1.12 (1.1 to 1.2) | 0.003 |
| Intervention in ICU | ||||
| Vasopressor use | 2,449 (46.2) | 180 (80.4) | 2,629 (47.5) | <0.001 |
| RRT | 104 (2.0) | 20 (8.9) | 124 (2.2) | <0.001 |
| Operation (during the first ICU hospitalization) | ||||
| Head | 206 (3.9) | 7 (3.3) | 213 (3.9) | 0.808 |
| Cervical spine | 54 (1.0) | 16 (7.5) | 70 (1.3) | <0.001 |
| Thoracic spine | 24 (0.5) | 0 | 24 (0.4) | 0.654 |
| Lumbar spine | 52 (1.0) | 1 (0.5) | 53 (1.0) | 0.702 |
| Abdomen | 189 (3.6) | 11 (5.2) | 200 (3.6) | 0.288 |
| Chest | 35 (0.7) | 2 (0.9) | 37 (0.7) | 0.944 |
| Face | 68 (1.3) | 4 (1.9) | 72 (1.3) | 0.648 |
| Pelvis | 92 (1.7) | 4 (1.9) | 96 (1.7) | >0.999 |
Values are presented as median (interquartile range) or number (%).
ICU: intensive care unit; UIR: unplanned intensive care unit readmission; SBP: systolic blood pressure; LOS: length of stay; INR: international normalized ratio; RRT: renal replacement therapy.
The most common reason for UIR was respiratory problem requiring treatment such as endotracheal intubation and high flow nasal cannula (114 cases, 53.8%). The second most common reason was septic shock (32 cases, 15.1%), followed by a neurologic problem (27 cases, 12.7%) and hypovolemic shock due to hemorrhage (19 cases, 9.0%) (Figure 2). Multivariate logistic regression showed cervical spine operation during the first ICU hospitalization (odds ratio [OR], 6.564; 95% CI, 3.618–11.908; P<0.001), RRT (OR, 3.515; 95% CI, 2.063–5.990; P<0.001), MTP (OR, 1.739; 95% CI, 1.078–2.805; P=0.023), cerebrovascular disease (OR, 1.738; 95% CI, 1.024–2.947; P=0.04), use of vasopressors (OR, 1.644; 95% CI, 1.216–2.222; P=0.001), male (OR, 1.584; 95% CI, 1.094–2.294; P=0.015), and old age (OR, 1.028; 95% CI, 1.019–1.037; P<0.001) were significantly correlated with UIR (Table 4).
Figure 2.
Reasons for unplanned intensive care unit readmission. CRRT: continuous renal replacement therapy.
Table 4.
Associated factors for UIR
| Variable | Unadjusted OR | 95% CI | P-value | Adjusted OR | 95% CI | P-value |
|---|---|---|---|---|---|---|
| Age (yr) | 1.026 | 1.016–1.036 | <0.001 | 1.028 | 1.019–1.037 | <0.001 |
| Sex (male) | 1.617 | 1.115–2.343 | 0.011 | 1.584 | 1.094–2.294 | 0.015 |
| ISS | 1.003 | 0.986–1.020 | 0.719 | - | - | - |
| On arrival | ||||||
| SBP <90 mm Hg | 1.286 | 0.902–1.833 | 0.165 | |||
| MTP | 1.543 | 0.905–2.632 | 0.111 | 1.739 | 1.078–2.805 | 0.023 |
| Underlying disease | ||||||
| Hypertension | 1.084 | 0.776–1.516 | 0.636 | |||
| Diabetes mellitus | 1.196 | 0.830–1.723 | 0.336 | |||
| Cerebrovascular disease | 1.662 | 0.975–2.833 | 0.062 | 1.738 | 1.024–2.947 | 0.040 |
| AIS | ||||||
| Head–neck | 1.039 | 0.954–1.131 | 1.131 | - | - | - |
| Operation (during the first ICU hospitalization) | ||||||
| Cervical spine | 6.315 | 3.410–11.696 | <0.001 | 6.564 | 3.618–11.908 | <0.001 |
| Hospital day (during the first ICU hospitalization) | ||||||
| LOS in ICU | 0.992 | 0.982–1.003 | 0.150 | - | - | - |
| Intervention in ICU | ||||||
| Vasopressor use | 1.612 | 1.160–2.241 | 0.004 | 1.644 | 1.216–2.222 | 0.001 |
| RRT | 3.927 | 2.210–6.978 | <0.001 | 3.515 | 2.063–5.990 | <0.001 |
UIR: unplanned intensive care unit readmission; OR: odds ratio; ISS: Injury Severity Score; SBP: systolic blood pressure; MTP: massive blood transfusion protocol; AIS: Abbreviated Injury Scale; ICU: intensive care unit; LOS: length of stay; RRT: renal replacement therapy.
Of the 70 patients who underwent cervical spine surgery during their initial ICU stay, 16 were readmitted to the ICU, all due to respiratory complications. Among these, 15 cases were directly related to respiratory issues, while 1 patient was found in cardiac arrest, presumably secondary to respiratory failure (Table 5). Pneumonia was the most common diagnosis, with other conditions such as atelectasis, pleural effusion, and vocal cord paralysis also identified. Type 2 respiratory failure, characterized by carbon dioxide hypercapnia, was confirmed in six cases. Of the 16 readmitted patients, three had already undergone a tracheostomy.
Table 5.
Analysis of cases with cervical spine surgery and unplanned returned ICU
| Cause | Ventilator | Diagnosis | Failure typea) | Tracheostomy | |
|---|---|---|---|---|---|
| 1 | Respiratory | Y | Atelectasis | 2 | N |
| 2 | Respiratory | Y | Pneumonia | 1 | Y |
| 3 | Respiratory | Y | Atelectasis, pneumonia | 2 | N |
| 4 | Respiratory | Y | Pneumonia | 1 | N |
| 5 | Respiratory | Y | Pneumonia | 1 | N |
| 6 | Respiratory | N | Atelectasis, pneumonia, pleural effusion | 1 | N |
| 7 | Respiratory | N | Pneumonia | 1 | Y |
| 8 | Respiratory | Y | Pneumonia | 1 | N |
| 9 | Respiratory | N | Atelectasis, pneumonia | 2 | N |
| 10 | Respiratory | N | Pneumonia | 1 | N |
| 11 | Respiratory | Y | Pneumonia, pleural effusion | 1 | N |
| 12 | Arrest | Y | Pneumothorax, pleural effusion | - | N |
| 13 | Respiratory | Y | Pneumonia | 2 | Y |
| 14 | Respiratory | Y | Atelectasis, pneumonia, pleural effusion, vocal cord palsy | 2 | N |
| 15 | Respiratory | N | Pneumonia | 1 | N |
| 16 | Respiratory | Y | Pneumonia | 1 | N |
ICU: intensive care unit.
Type 1: hypoxemic; type 2: hypercapnic.
DISCUSSION
In this study, patients in the UIR group had significantly worse clinical outcomes, including hospital stay, ICU length of stay, and complications. Among many associated factors, cervical spine surgery during the first ICU hospitalization, RRT, and MTP were significantly associated with UIR in trauma patients. Respiratory problems were the leading cause of UIR.
In our study, unplanned readmission rates and mortality rates are relatively lower compared with previous studies, which reported readmission rates of 4%–10% and mortality of 20%–40% [3,6]. Some studies interpreted lower UIR as a good indicator of ICU liberation, but it could also be the result of delayed transfer to a general ward [6,13]. However, the ICU length of stay at our center is equivalent to those of other trauma centers [14]. Differences in UIR rate may also be due to the method of calculating readmission rates, which differed by center [5,2,15,16]; some studies defined UIR within 7 days of ICU discharge [2], while others considered it within 48 hours of ICU discharge [3,5]. However, in our center, admission to the ICU after regular surgery is recorded as a planned admission, which may falsely decrease the UIR. Severe trauma patients may require multiple major surgeries that require preparation in the ICU, and many of these patients might be readmitted after planned surgery. In addition, in our analysis, patients in the UIR group had a longer initial ICU stay than those in the successfully discharged group. This may be interpreted as a longer initial ICU stay affecting UIR, but it may also be due to the higher disease severity of the UIR group.
Various studies have been implemented to reduce UIR. Forster et al. [17] suggested that delayed transfer to the general ward reduced the likelihood of mortality and ICU readmission. In addition, Berardino et al. [18] reported that insufficient verification of patient condition increased the likelihood of complications when transferred from the ICU. However, long ICU stays can cause psychiatric problems including delirium, as well as cost issues and delayed rehabilitation [19]. Therefore, it is important to determine which patients should be delayed from discharge from the ICU and which patients should be discharged quickly. While more research is needed on the timing of transfer to a general ward, our results can aid decision-making by identifying factors associated with UIR.
The associated factors, complications, and prolonged length of stay identified in our study are largely consistent with those identified in previous studies [1-4]. We confirmed that cervical spine surgery during the first ICU hospitalization was the strongest associated factor for readmission to the ICU, with respiratory problems being the most common reason. In cervical surgery, the longer is the surgery time, the greater is the level of damage, or the more common are concomitant diseases, the greater is the level of care needed [20,21]. It is well known that severe cervical spine injury is a significant risk factor for respiratory failure [20,22-24]. Such surgery is typically performed as soon as possible for an unstable or high-level injury accompanied by neurological deterioration [25]. Patients who undergo emergency cervical spine surgery often experience cervical cord swelling and diaphragm fatigue, leading to weakened respiratory muscles, difficulty in clearing sputum, and restricted movements. Therefore, caution is required when transferring such patients to general wards. More intensive monitoring and treatments, including bronchial washes, early tracheostomy, and respiratory rehabilitation in the ICU, may be necessary to prevent respiratory complications [26]. Additionally, considering that there have been cases of readmission to the ICU even after tracheostomy, careful management of excessive airway exudate following the procedure is also crucial.
In this study, as in many others [27-29], RRT the second strongest factor associated with UIR. Patients who received RRT for AKI in the ICU are more likely to experience multi-organ dysfunction and require continued dialysis. RRT would have more complications due to anticoagulants during dialysis and large vascular access [30-32]. Our study could have been improved by analyzing the cause of RRT use. However, detailed reasons for RRT utilization were not available in all medical records. MTP is usually performed on patients with massive bleeding and could reflect injury severity, and patients with severe injury have a higher probability of UIR. Not surprisingly, in this study, variables related with severe injury were associated with UIR in univariable analysis.
Previous studies have pointed out that GCS, unstable vital signs at the time of ICU discharge, and laboratory findings are correlated with UIR [6,33,34]. In our study, some vital signs and laboratory findings showed significant differences, but no clinical difference was noted. While hemoglobin levels were different in the two groups, they were greater than 9.0 g/dl in both, indicating that transfusion was usually not required. GCS at the time of ICU discharge would have been a more informative variable than GCS at admission, but it could not be analyzed due to lack of data. Medical charts containing GCS at ICU discharge would be required for further study.
Nine patients died after UIR, and three cases were classified as potentially preventable trauma deaths by our multi-disciplinary panel review. These three patients died from multi-organ failure or sepsis due to lack of early monitoring or intervention. Additionally, respiratory problems, which were the leading cause of UIR, require careful monitoring. This is particularly crucial for patients needing high-flow oxygen therapy, even if they do not require intubation. While continuous monitoring in the ICU is not necessary for all patients, on the borderline of recovery patients in the general ward may require more intensive monitoring. In this situation, the sub-ICU which provides more intensive care than general ward, could be a good alternative.
This study has several limitations. First, because it is a single-center, retrospective study, it may not represent all trauma patients and may involve selection and information biases. Second, complications or procedures during the general ward stay could have caused UIRs that were not counted. Third, patients typically had poly-trauma, making it difficult to identify a single cause for ICU readmission. Despite this complexity, we made efforts to identify the major causes of UIR. Finally, patients with planned ICU readmission were excluded from this study. Since some of these patients are severely injured and may need more than one surgery, the complication or mortality rate may appear to be relatively low.
In conclusion, UIR was confirmed to have poor clinical outcomes and should be prevented. Many associated factors for UIR have been identified; among them, cervical spine surgery during the first ICU hospitalization, RRT, and MTP seem to have the greatest impact. As respiratory problems account for the most frequent UIR, care is needed when transferring patients to general wards including monitoring of respiratory problems. Furthermore, for patients who undergo cervical spine operation, respiratory rehabilitation might be required to prevent UIR.
KEY MESSAGES
▪ Unplanned intensive care unit (ICU) readmission (UIR) among trauma patients causes poor clinical outcomes in patients, and the most common reason for UIR was respiratory-related problems.
▪ Cervical spine surgery during the first ICU hospitalization, renal replacement therapy (RRT), massive blood transfusion protocol (MTP), cerebrovascular disease, use of vasopressor, male, and old age, in that order, were the strongest factors influencing UIR.
▪ To prevent UIR, patients with cervical spine surgery during the first ICU hospitalization, RRT, or MTP should be carefully evaluated for transfer to a general ward and closely monitored for respiratory problems.
Footnotes
CONFLICT OF INTEREST
No potential conflict of interest relevant to this article was reported.
FUNDING
None.
ACKNOWLEDGMENTS
We would like to express our great appreciation to Sora Kim, Head of Trauma Quality Improvement Department, Ajou Trauma Center, for her invaluable support regarding data collection.
AUTHOR CONTRIBUTIONS
Conceptualization: BHK. Methodology: BHK. Formal analysis: YL. Data curation: YL. Writing–original draft: all authors. Writing–review & editing: all authors. All authors read and agreed to the published version of the manuscript.
REFERENCES
- 1.Christmas AB, Freeman E, Chisolm A, Fischer PE, Sachdev G, Jacobs DG, et al. Trauma intensive care unit 'bouncebacks': identifying risk factors for unexpected return admission to the intensive care unit. Am Surg. 2014;80:778–82. doi: 10.1177/000313481408000827. [DOI] [PubMed] [Google Scholar]
- 2.Fakhry SM, Leon S, Derderian C, Al-Harakeh H, Ferguson PL. Intensive care unit bounce back in trauma patients: an analysis of unplanned returns to the intensive care unit. J Trauma Acute Care Surg. 2013;74:1528–33. doi: 10.1097/TA.0b013e31829247e7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Morgan M, Vernon T, Bradburn EH, Miller JA, Jammula S, Rogers FB. A comprehensive review of the outcome for patients readmitted to the ICU following trauma and strategies to decrease readmission rates. J Intensive Care Med. 2020;35:936–42. doi: 10.1177/0885066619899639. [DOI] [PubMed] [Google Scholar]
- 4.Rubano JA, Vosswinkel JA, McCormack JE, Huang EC, Shapiro MJ, Jawa RS. Unplanned intensive care unit admission following trauma. J Crit Care. 2016;33:174–9. doi: 10.1016/j.jcrc.2016.02.012. [DOI] [PubMed] [Google Scholar]
- 5.Rosenberg AL, Hofer TP, Hayward RA, Strachan C, Watts CM. Who bounces back? Physiologic and other predictors of intensive care unit readmission. Crit Care Med. 2001;29:511–8. doi: 10.1097/00003246-200103000-00008. [DOI] [PubMed] [Google Scholar]
- 6.Rosenberg AL, Watts C. Patients readmitted to ICUs: a systematic review of risk factors and outcomes. Chest. 2000;118:492–502. doi: 10.1378/chest.118.2.492. [DOI] [PubMed] [Google Scholar]
- 7.Nishi GK, Suh RH, Wilson MT, Cunneen SA, Margulies DR, Shabot MM. Analysis of causes and prevention of early readmission to surgical intensive care. Am Surg. 2003;69:913–7. [PubMed] [Google Scholar]
- 8.Chrusch CA, Olafson KP, McMillan PM, Roberts DE, Gray PR. High occupancy increases the risk of early death or readmission after transfer from intensive care. Crit Care Med. 2009;37:2753–8. doi: 10.1097/CCM.0b013e3181a57b0c. [DOI] [PubMed] [Google Scholar]
- 9.Laytin AD, Sims CA. Risk factors for unplanned ICU readmission among trauma patients: age matters. Crit Care Explor. 2022;4:e0778. doi: 10.1097/CCE.0000000000000778. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Johns TJ. Characteristics and risk factors of trauma patients readmitted to the ICU within the same hospitalization. J Trauma Nurs. 2014;21:14–21. doi: 10.1097/JTN.0000000000000023. [DOI] [PubMed] [Google Scholar]
- 11.Bradburn EH, Jammula S, Horst MA, Morgan M, Vernon TM, Gross BW, et al. An analysis of outcomes and predictors of intensive care unit bouncebacks in a mature trauma system. J Trauma Acute Care Surg. 2020;88:486–90. doi: 10.1097/TA.0000000000002550. [DOI] [PubMed] [Google Scholar]
- 12. American College of Surgeons. National trauma data standard data dictionary 2022 [Internet]. American College of Surgeons; 2024 [cited 2024 Aug 4]. Available from: https://www.facs.org/quality-programs/trauma/quality/national-trauma-data-bank/national-trauma-data-standard/data-dictionary/
- 13.Zimmerman JE. Intensive care unit readmission: the issue is safety not frequency. Crit Care Med. 2008;36:984–5. doi: 10.1097/CCM.0B013E318165FC15. [DOI] [PubMed] [Google Scholar]
- 14.Wong EG, Parker AM, Leung DG, Brigham EP, Arbaje AI. Association of severity of illness and intensive care unit readmission: a systematic review. Heart Lung. 2016;45:3–9. doi: 10.1016/j.hrtlng.2015.10.040. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Ho KM, Dobb GJ, Lee KY, Finn J, Knuiman M, Webb SA. The effect of comorbidities on risk of intensive care readmission during the same hospitalization: a linked data cohort study. J Crit Care. 2009;24:101–7. doi: 10.1016/j.jcrc.2007.11.015. [DOI] [PubMed] [Google Scholar]
- 16.Utzolino S, Kaffarnik M, Keck T, Berlet M, Hopt UT. Unplanned discharges from a surgical intensive care unit: readmissions and mortality. J Crit Care. 2010;25:375–81. doi: 10.1016/j.jcrc.2009.09.009. [DOI] [PubMed] [Google Scholar]
- 17.Forster GM, Bihari S, Tiruvoipati R, Bailey M, Pilcher D. The association between discharge delay from intensive care and patient outcomes. Am J Respir Crit Care Med. 2020;202:1399–406. doi: 10.1164/rccm.201912-2418OC. [DOI] [PubMed] [Google Scholar]
- 18.Berardino M, Morrone O, Sciacca PF, Rosato R, Ciccone G, Massaro F. Discharge criteria from intensive care unit in brain injured patients. Acta Neurochir (Wien) 2004;146:453–6. doi: 10.1007/s00701-003-0176-1. [DOI] [PubMed] [Google Scholar]
- 19.Kafantaris S, Kadda O. Advantages and disadvantages of patients’ hospitalization in intensive care units. Health Res J. 2021;7:155–9. [Google Scholar]
- 20.Debkowska MP, Butterworth JF, Moore JE, Kang S, Appelbaum EN, Zuelzer WA. Acute post-operative airway complications following anterior cervical spine surgery and the role for cricothyrotomy. J Spine Surg. 2019;5:142–54. doi: 10.21037/jss.2019.03.01. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sagi HC, Beutler W, Carroll E, Connolly PJ. Airway complications associated with surgery on the anterior cervical spine. Spine (Phila Pa 1976) 2002;27:949–53. doi: 10.1097/00007632-200205010-00013. [DOI] [PubMed] [Google Scholar]
- 22.Boddapati V, Lee NJ, Mathew J, Held MB, Peterson JR, Vulapalli MM, et al. Respiratory compromise after anterior cervical spine surgery: incidence, subsequent complications, and independent predictors. Global Spine J. 2022;12:1647–54. doi: 10.1177/2192568220984469. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Nandyala SV, Marquez-Lara A, Park DK, Hassanzadeh H, Sankaranarayanan S, Noureldin M, et al. Incidence, risk factors, and outcomes of postoperative airway management after cervical spine surgery. Spine (Phila Pa 1976) 2014;39:E557–63. doi: 10.1097/BRS.0000000000000227. [DOI] [PubMed] [Google Scholar]
- 24.Kim M, Choi I, Park JH, Jeon SR, Rhim SC, Roh SW. Airway management protocol after anterior cervical spine surgery: analysis of the results of risk factors associated with airway complication. Spine (Phila Pa 1976) 2017;42:E1058–66. doi: 10.1097/BRS.0000000000002236. [DOI] [PubMed] [Google Scholar]
- 25.Okereke I, Mmerem K, Balasubramanian D. The management of cervical spine injuries: a literature review. Orthop Res Rev. 2021;13:151–62. doi: 10.2147/ORR.S324622. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Colón LF, Barber L, Soffin E, Albert TJ, Katsuura Y. Airway complications after anterior cervical spine surgery: etiology and risk factors. Global Spine J. 2023;13:2526–40. doi: 10.1177/21925682231160072. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ponzoni CR, Corrêa TD, Filho RR, Serpa Neto A, Assunção MS, Pardini A, et al. Readmission to the intensive care unit: incidence, risk factors, resource use, and outcomes. a retrospective cohort study. Ann Am Thorac Soc. 2017;14:1312–9. doi: 10.1513/AnnalsATS.201611-851OC. [DOI] [PubMed] [Google Scholar]
- 28.Jo YS, Lee YJ, Park JS, Yoon HI, Lee JH, Lee CT, et al. Readmission to medical intensive care units: risk factors and prediction. Yonsei Med J. 2015;56:543–9. doi: 10.3349/ymj.2015.56.2.543. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Japiassú AM, Cukier MS, Queiroz AG, Gondim CR, Penna GL, Almeida GF, et al. Early predictive factors for intensive care unit readmission. Rev Bras Ter Intensiva. 2009;21:353–8. [PubMed] [Google Scholar]
- 30.Siqueira Santos MM, Sganzerla D, Pereira IJ, Rosa RG, Granja C, Teixeira C, et al. Long-term mortality and health-related quality of life after continuous versus intermittent renal replacement therapy in ICU survivors: a secondary analysis of the quality of life after ICU study. J Intensive Care Med. 2024;39:636–45. doi: 10.1177/08850666231224392. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Oeyen S, De Corte W, Benoit D, Annemans L, Dhondt A, Vanholder R, et al. Long-term quality of life in critically ill patients with acute kidney injury treated with renal replacement therapy: a matched cohort study. Crit Care. 2015;19:289. doi: 10.1186/s13054-015-1004-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Shingarev R, Wille K, Tolwani A. Management of complications in renal replacement therapy. Semin Dial. 2011;24:164–8. doi: 10.1111/j.1525-139X.2011.00828.x. [DOI] [PubMed] [Google Scholar]
- 33.Makris N, Dulhunty JM, Paratz JD, Bandeshe H, Gowardman JR. Unplanned early readmission to the intensive care unit: a case-control study of patient, intensive care and ward-related factors. Anaesth Intensive Care. 2010;38:723–31. doi: 10.1177/0310057X1003800338. [DOI] [PubMed] [Google Scholar]
- 34.Lee HF, Lin SC, Lu CL, Chen CF, Yen M. Revised Acute Physiology and Chronic Health Evaluation score as a predictor of neurosurgery intensive care unit readmission: a case-controlled study. J Crit Care. 2010;25:294–9. doi: 10.1016/j.jcrc.2009.12.007. [DOI] [PubMed] [Google Scholar]