A retrospective, quasi-experimental study found that patients admitted to the intensive care unit (ICU) for acute respiratory failure (ARF) saw better outcomes when an inter-ICU transfer came within the first 2 days of admittance.
Patients with acute respiratory failure (ARF) have a lower risk of mortality and shorter length of stay when they are transferred to another intensive care unit (ICU) early after admission, according to a recent study published in Critical Care Explorations. The aim of the study was to describe inter-ICU transfer patterns and evaluate the impact of timing of transfer on patient-centered outcomes.
For the retrospective study, the researchers used the Agency for Healthcare Research and Quality’s Healthcare Cost and Utilization Project (HCUP) State Inpatient Databases for 5 states. Inpatient databases from 2015 to 2017 in Washington, New York, Florida, Maryland, and Mississippi were used for this study.
Adult patients who had International Classification of Diseases, Ninth Revision and Tenth Revision diagnosis codes corresponding to ARF and procedure codes corresponding to ventilation were identified for this study. Patients with a length of stay greater than 25 days were excluded from this study, as they were likely to be in long-term acute care facilities. Patients in psychiatric hospitals were also excluded.
All patients with ARF were then organized into transfer and nontransfer groups. All patients with ARF with a transfer code of interfacility transfer to a second acute care hospital were included in the transfer group, as were all patients with ARF who were discharged from a hospital and then subsequently admitted to a different acute care hospital within the same calendar day.
The researchers found that approximately 70% of all inter-ICU transfers were done within 2 days. Using this information, they defined early transfers as those that occurred within 2 days of admission to the first hospital’s ICU and later transfers as those that occurred at least 3 days after admission.
Patients were categorized by age into groups younger than 65 years, 65 to 79 years, and 80 years and older. Race was classified into 4 groups: White, Black (non-Hispanic), Hispanic, and other/missing. Patients were also split into groups based on insurance, with groups of Medicare, Medicaid, commercial, and other. Hospitals were categorized as for profit and nonprofit. Small hospitals had less than 100 beds, medium hospitals had between 100 and 300 beds, and large hospitals had more than 300 beds. Emergency department utilization was also calculated using inpatient records.
The study authors examined a total of 245,626 acute care hospital records with ARF diagnoses from the aforementioned states. Of those records, 9576 had undergone a transfer. A total of 2858 patients were excluded for incomplete data, which left 6718 patients to be evaluated in the study. Of note, those in Maryland were younger and more often had commercial insurance, and Mississippi and Washington had lower rates of transfer and lower ARF volume in hospitals.
The investigators used propensity score matching to obtain unbiased estimates of the effect of early vs later transfer. The primary outcome of the study was the mortality rate at hospital 2, which was defined as the discharge destination “deceased,” and the secondary outcomes were cost and length of stay.
Unadjusted outcomes were each lower for the early transfer group: in-hospital mortality (24.4% vs 36.1%), length of stay (8 vs 22 days), and cumulative charges ($118,686 vs $308,977). In the final adjusted model, the risk of in-hospital mortality for the early transfer group was lower than that for the later transfer group (odds ratio [OR], 0.44; 95% CI, 0.40-0.49). This represented a 20.5% absolute risk reduction of in-hospital mortality (16.2% vs 36.7%) for early vs later transfers.
Adjusted models also indicated that patients who were transferred early had a length of stay 20.7 days shorter than those who were not (13.0 vs 33.7 days; P < .0001). Lastly, the early transfer group had adjusted charges that were $66,201 lower than those of the later transfer group ($192,182 vs $258,383; P < .0001).
To test their findings, the researchers recalculated their findings with different parameters for early vs late transfers: Early transfers were defined as those occurring within 1 day of admittance, and later transfers were defined as occurring 2 days or more after admittance. The reanalysis yielded a similar risk of the primary outcome (OR, 0.4251; 95% CI, 0.34-0.53). The authors wrote that this demonstrates that their research is robust to changes in the definitions of early and late transfer.
There were some limitations to this study. It involved only 5 states, none of which were in the Midwest or Southwest. The constraints of HCUP data include variability in coding practices and limitations of diagnosis codes. The study also lacked information on clinically derived illness severity scores, stakeholder reports on reason for transfer, and staffing models, among others. Because the study focused on patients from 2015 to 2017, it did not account for the scarcity of ICU beds due to the COVID-19 pandemic. Further work also needs to be done to estimate the inherent variation in implementation of best practices in ARF care among transferring hospitals.
Summarizing their findings, the researchers wrote that early transfers affect patient-centered outcomes, like mortality, and health system outcomes, like length of stay and cost. Although the study was retrospective and the results cannot be taken as causal, the fact that all outcomes demonstrated a favorable association is quite compelling, they wrote.
The authors recommended that a multicenter study prospectively evaluating early and late transfers could be the next step to better understand the structural and process measures that affect these outcomes.
Reference
Nadig NR, Brinton DL, Simpson KN, et al. The impact of timing on clinical and economic outcomes during inter-ICU transfer of acute respiratory failure patients: time and tide wait for no one. Crit Care Med. 2022;4(3):e0642. doi:10.1097/CCE.0000000000000642