Factors Associated with Transport Instability and Outcome Deterioration During Interhospital Intensive Care Transfers of Patients with COVID-19 in Germany

1. Introduction

The COVID-19 pandemic posed an unprecedented challenge to healthcare systems worldwide. The rapid spread of SARS-CoV-2 led to a surge in critically ill patients requiring advanced respiratory support and intensive care. Early reports from Italy and France highlighted the overwhelming burden on regional hospitals, with capacity shortages forcing patient transfers across regions and even national borders [1]. These developments underscored the importance of safe interhospital transport for critically ill patients during the pandemic.

In Germany, healthcare provision is highly decentralized, with intensive care transport services regulated at the state level and delivered by a mix of public and private providers, including fire brigades, aid organizations, and aeromedical services such as ADAC and DRF Luftrettung. Despite Germany’s high intensive care capacity, uneven regional resource distribution necessitated large-scale COVID-19 patient transfers during infection peaks.

Before COVID-19, literature on intensive care transfers focused on intrahospital transport complications [2], air transport safety [3], or logistics for acute respiratory distress syndrome (ARDS) patients [4]. Complication rates in critically ill patient transfers varied widely, with some studies reporting up to 38% deterioration in outcomes [5] and others suggesting that, under structured conditions, interhospital transfers were safe and did not increase mortality [6,7]. However, comprehensive data on large-scale, pandemic-related interhospital transports were lacking at the outset of COVID-19. International studies showed that large-scale interhospital transfers were feasible but logistically challenging [8,9,10].

For Germany, however, systematic analyses of COVID-19 patient transfers were absent at the beginning of the pandemic. The “cloverleaf” system introduced in late 2020 established regional coordination hubs to facilitate cross-state redistribution of patients, but its effectiveness and clinical impact had not been empirically assessed. Particularly vulnerable were patients requiring invasive ventilation, extracorporeal membrane oxygenation (ECMO), or vasopressor support — interventions associated with high risk during transport. Moreover, there was no unified national documentation system for intensive care transfers, despite earlier recommendations by the German Interdisciplinary Association for Intensive and Emergency Medicine (DIVI) [11]. The absence of standardized data limited opportunities to evaluate outcomes, quality, and risk factors.

Against this backdrop, the present study was designed as the first nationwide investigation of interhospital transfers of critically ill COVID-19 patients in Germany. The study aimed to describe transport characteristics, patient status at takeover and handover, complication rates, and patient- and system-related factors associated with outcomes. In addition, the study sought to identify organizational determinants of transport duration and quality, such as takeover and handover times, as well as clinical predictors of instability.

The study aimed to identify factors associated with transport instability, deterioration, and transfer logistics (Figure 1).

The study improves understanding of intensive care transport under pandemic conditions and provides evidence for quality assurance and future preparedness. The findings are of direct public health relevance, as they highlight systemic challenges and risks in the interhospital transfer of critically ill patients, informing preparedness, resource allocation, and coordination strategies during large-scale health emergencies such as pandemics.

2. Materials and Methods

2.1. Study Design

This investigation was conducted as a multicentred, retrospective, cross-sectional study within the field of health services research. The study analysed the characteristics and outcomes of interhospital intensive care transports of COVID-19 patients in Germany. The study combined retrospective analysis of transport reports with prospective collection of supplementary organizational data from participating providers.

2.2. Patient Population

The study included adult patients with confirmed COVID-19 who underwent intensive care transport between January 2020 and June 2021. A total of 414 transport reports were obtained and analysed. Cases were contributed by six ground-based intensive care transport vehicle (ITW) providers (Augsburg, Regensburg, Warendorf, Kiel, Hamburg, Cambs) and one aeromedical provider (Cologne). Data on helicopter transfers were not available. Participation was voluntary, and providers were contacted repeatedly to maximize response rates. Only anonymized or pseudonymized data were used.

2.3. Data Collection

The primary data source was the DIVI intensive care transport report, where available (see supplementary material, translated). In 46% of cases the DIVI form was used, while the remainder were documented using alternative formats. The DIVI report captures clinical status, interventions, monitoring, and transport outcomes. In addition, a short supplementary questionnaire was distributed to providers, addressing reasons for transfer, distances, coordination structures, and special resource requirements. All data were digitized, standardized, and coded for analysis. Free-text entries were categorized using predefined groups.

2.4. Definitions and Variables

Three composite variables were defined:

• Severe illness at transport start, defined by the presence of one or more of the following: catecholamine therapy, FiO₂ ≥ 0.5, P/F ratio <100 mmHg, mean arterial pressure (MAP) ≤60 mmHg, or PEEP >12 cmH₂O.
• Transport instability, defined as two or more critical deviations in heart rate <50 bpm / >100 bpm, blood pressure <90 mmHg / >160 mmHg, oxygen saturation <90%, or end-tidal CO₂ <33 mmHg / >43 mmHg during transfer.
• Outcome deterioration, defined as MAP decrease by ≥20%, Glasgow Coma Scale decline by ≥2 points, oxygen desaturation by ≥4%, or initiation of catecholamines during transport.

The definitions of transport instability and deterioration were based on clinically relevant thresholds derived from prior literature on critical care transport and ARDS severity and were adapted to reflect the specific context of COVID-19-related respiratory and circulatory failure. Demographic variables (age, sex), clinical factors (ventilation, comorbidities, ECMO), and operational parameters (priority, transfer times) were also included. In addition, a new variable “Severe Impairment” was created because of the markedly reduced number of eligible cases, to identify patients with physiological impairment defined as MAP < 60 mmHg and/or a P/F ratio < 100 mmHg.

2.5. Outcome

The primary outcome was patient instability and deterioration during transfer. Secondary outcomes included takeover time at the sending hospital, duration of transportation, handover time at the receiving hospital, frequency of necessary interventions during transport, monitoring adequacy, and factors influencing delays.

2.6. Statistical Analysis

Data were entered into SPSS statistical software (version 29, IBM Inc., Armonk, NY, USA). Continuous variables were checked for plausibility, corrected where necessary, and summarized as mean with standard deviation, or median with quartiles, and range (minimum to maximum), depending on the distribution. Categorical data were reported as frequencies and percentages. Logistic regression models were used to identify predictors of transport instability and outcome deterioration, while multiple linear regression was applied to evaluate factors affecting transfer times. Variables were selected for multivariable models based on clinical relevance and univariate associations, to balance statistical robustness and avoidance of overfitting given the sample size. Given the retrospective and heterogeneous data sources, missing data were expected and handled using complete case analysis for regression models. The proportion of missing values varied across variables and was reported transparently.

2.7. Ethical Considerations

The study was approved by the Ethics Committee of the Medical Faculty of Christian-Albrechts-University of Kiel (D 573/20). As only anonymized retrospective data were analyzed, no patient consent was required. Data handling complied with relevant data protection and confidentiality regulations.

3. Results

3.1. Operational and Tactical Data

Most transports (93%) were conducted with intensive care transport vehicles, while 3% (n=10) involved fixed-wing aircraft. Despite frequent use, no data on helicopter transports could be provided within the scope of this study. The largest contribution came from Augsburg (43%) and Regensburg (38%), with smaller numbers from Warendorf, Kiel, Hamburg, Cambs, and Cologne.

Physician staffing was dominated by anaesthesiologists (98% of reported cases). Where qualification was specified, 80% were specialists and 20% were residents in training.

Transfer timing varied widely. Patient takeover averaged 53 minutes (range 4 - 279), pure transport time averaged 59 minutes (range 6 - 415), and handover at the receiving facility averaged 64 minutes (range 0 - 241) (Figure 2).

Linear regression analysis indicated that high priority (as stated in the patient care report) shortened takeover time by ~10 minutes, while unconsciousness/sedation (+7 minutes), catecholamine therapy (+9 minutes), obesity (+14 minutes), and ECMO support (+21 minutes) prolonged the process (Table 1).

For handover, catecholamine therapy (+11 minutes) and ECMO (+20 minutes) were significant prolonging factors. No significant associations were found for hypertension (as comorbidity), age, sex or physician seniority (Table 2).

Most transfers (94%) occurred during pandemic “wave” phases as defined by the Robert Koch Institute, with the majority in the second wave (51%) and third wave (28%).

3.2. Priority for Transfer

Priority as per patient care report was documented in 67% of cases (n = 277): 48% were deemed transferable within 24 hours, 40% required urgent transfer within two hours, and 11% required immediate transfer within 30 minutes.

3.3. Patient Demographics

Age and sex were inconsistently reported. Of patients with available data, 67% were male and 33% female. The largest age group was 60–69 years (28%), followed by 50–59 (23%) and 70–79 (22%). Only 16% were under 50, and 11% were 80 years or older.

3.4. Patient Status at Takeover

Neurological assessment showed that three-quarters of patients were sedated, with 19% alert and oriented. The Glasgow Coma Scale (GCS) was reported in 71% of cases: 60% scored 3 (deep coma) and 25% scored 15 (fully awake).

Cardiovascular data were more complete: median systolic blood pressure was 123 mmHg, median MAP 81 mmHg, and median heart rate 82 bpm. However, 64% of patients required catecholamines, and only 34% were hemodynamically stable. Respiratory support was common. Three-quarters of patients were documented as mechanically ventilated (n=310), most frequently with BiPAP (48%), CPAP (17%), or PCV (10%). Median FiO₂ was 0.6 (range 0.2-1.0), and median PEEP 12 cmH₂O (range 4-22 cmH₂O), consistent with ARDS. Median oxygen saturation at takeover was 94% (range 39–100%). The median P/F ratio was 155 mmHg (range 39-440 mmHg).

3.5. Measures and Monitoring

Patients frequently required invasive lines: peripheral intravenous catheters, central venous catheters, and arterial lines were common. ECMO was initiated in a small but critical subgroup, significantly prolonging takeover and handover times.

Medication use reflected high disease severity. Monitoring typically included continuous hemodynamic and respiratory surveillance, but missing entries limited analysis of completeness.

3.6. Instability During Transport

Transport instability, defined as two or more critical deviations in vital parameters, was observed in 44% of patients (n=182). Of these, 16% experienced multiple unstable events. Instability was most frequent among severely ill patients—over half of those classified as severe had at least one episode, compared to markedly fewer among less severe cases.

Logistic regression identified preexisting catecholamine therapy as the strongest predictor of instability (odds ratio 2.04; 95% CI 1.30–3.19; p=0.002). Other contributing factors included ECMO and high ventilatory requirements, though these did not always reach statistical significance.

3.7. Outcome Deterioration

Outcome deterioration occurred in 16.4% of patients (n=68), most commonly as MAP drops, reduced GCS, oxygen desaturation, or new catecholamine therapy. Obesity (p=0.020) and severe hypoxemia (P/F ratio <100 mmHg; p=0.010) were significant predictors of deterioration. Analgesics were associated with a twofold increased risk; however, this finding did not reach statistical significance (Table 3). Ventilation, sedation, catecholamines, and according to statistical analysis that seniority of the attending physician showed trends toward increased risk but were not statistically significant.

3.8. Summary of Key Results

• Takeover and handover times were prolonged by patient complexity, particularly ECMO and catecholamines.
• The majority of patients were critically ill, requiring ventilation and vasopressors.
• Instability occurred in nearly half of transports; deterioration in one in six.
• Key predictors of adverse events included catecholamines, obesity, and severe hypoxemia.

4. Discussion

4.1. Immediate Post-Transport Deterioration

The key finding of this study is that interhospital transfer of critically ill COVID-19 patients is feasible but associated with a substantial rate of physiological instability, with identifiable clinical risk factors that can inform future transfer strategies. A total of 16.4% of patients (n=68) showed deterioration immediately after transport. Previous studies generally found no increase in mortality associated with interhospital transfer, although transient physiological deterioration has been reported [6,7,10,12,13,14].

Other investigations documented substantially higher rates of deterioration. Differences in reported deterioration rates across studies are likely explained by heterogeneity in definitions, transport modalities (e.g., ground vs. air), patient selection, and the extraordinary operational conditions during the COVID-19 pandemic. In one study, 38% of patients worsened after transfer [5], while evacuations from Mayotte to Réunion showed deterioration of the P/F ratio in 34% of cases, particularly in patients with hypertension or diabetes [12]. In France, 82% of critically ill patients transported by air had worsening oxygenation, correlated with transport duration [12]. It should be noted that P/F ratio values were sometimes estimated from SpO₂/FiO₂ rather than PaO₂/FiO₂ [12,13,14]. Additional research linked P/F ratio <100 mmHg and hemodynamic instability to higher three-month mortality after transfer [12].

Signs of deterioration were reported immediately after transport in 16.4% of patients, despite previous reports describing intensive care transport as generally safe. The comparative studies, however, indicate that adverse outcomes like transport instability or outcome deterioration do occur, particularly in COVID-19 patients who may rapidly develop ARDS with high morbidity and mortality. The findings may also reflect selective documentation of unstable patients. Overall, the results point to the combined impact of disease severity, novelty of COVID-19, and heterogeneous standards in intensive care transport and documentation.

4.2. Instability During Transport

Unstable events in this study were frequent during transfer: 51% of patients with high illness severity experienced up to four unstable episodes, representing 89% of all instabilities. Transport instability was strongly associated with disease severity, consistent with previous studies identifying vasopressor use and severe hypoxemia as key risk factors [15,16,17].

4.3. Transfer Durations

Transfer durations increased with clinical complexity, consistent with previous reports [18,19,20].

4.4. Strategic Transfers

Strategic transfers were an important component of the pandemic response. Previous studies and national frameworks suggest that careful patient selection and coordination enable safe strategic transfers, although criteria vary across settings [6,21,22,23,24]. However, a consistent differentiation between capacity-driven and clinically indicated transfers was not available in the dataset, limiting further subgroup analysis.

4.5. Study Limitations

Several limitations must be considered. First, participation was voluntary and limited to selected providers, introducing potential selection bias, as described in observational transport research [25,26].

Second, documentation was heterogeneous and incomplete, introducing potential reporting bias. Inconsistent documentation is a known limitation in interhospital transport studies [25,26,27].

Third, the use of different documentation formats (DIVI vs. alternative records) has affected comparability, reflecting the lack of standardized national reporting despite prior recommendations [11,25].

Fourth, no data on helicopter transports were available, and the number of participating centers was limited, restricting generalizability [3,17].

Finally, outcome assessment was limited to immediate post-transport changes, and the retrospective design precludes causal inference [6,7].

4.6. Practical Recommendations

Based on these results, several practical recommendations can be drawn. First, interhospital transfers of critically ill patients should be regarded as high-risk interventions requiring specialized training, redundant monitoring, and structured preparation. Second, standardized, preferably digital, documentation patient care reports must be implemented nationwide to ensure complete data capture, facilitate quality assurance, and enable ongoing research. Third, risk stratification should be systematically applied: patients with catecholamine dependence, severe hypoxemia, or obesity require heightened preparation and may benefit from enhanced staffing or additional safeguards during transfer. Fourth, supra-regional coordination mechanisms, such as the German “cloverleaf” model, should be further developed to integrate local and national resources, ensuring equitable care distribution during future crises.

5. Conclusions

This study provides the first nationwide analysis of interhospital intensive care transport of COVID-19 patients in Germany based on 414 transport records. The findings highlight measurable risks of deterioration, frequent instability, extended transfer times, and challenges in documentation. They align with international evidence and underscore the need for further refinement of transport reports and strategic transfer criteria.

The findings show that transfers were feasible but associated with considerable risks. Nearly half of the patients experienced instability during transport, and about one in six deteriorated by objective measures. Predictors of adverse events included catecholamine therapy, obesity, and severe hypoxemia. Operationally, takeover and handover phases were prolonged by clinical complexity, especially when patients were sedated, receiving vasopressors, or supported with ECMO.

Despite these challenges, most patients were safely transferred. Documentation remained inconsistent, limiting systematic evaluation.

Improving documentation, coordination, and preparation for high-risk patients may further enhance the safety of future large-scale transfers. Although conducted in Germany, the findings are internationally relevant and contribute to the evidence base for intensive care transport.