OBJECTIVES:
The COVID-19 pandemic led to rapid changes in care delivery for critically ill patients, due to factors including increased numbers of ICU patients, shifting staff roles, and changed care locations. As these changes may have impacted the care of patients without COVID-19, we assessed changes in common ICU practices for mechanically ventilated patients with non-COVID acute respiratory failure at the onset of and during the COVID-19 pandemic.
DESIGN:
Interrupted time series analysis, adjusted for seasonality and autocorrelation where present, evaluating trends in common ICU practices prior to the pandemic (March 2016 to February 2020), at the onset of the pandemic (April 2020) and intra-pandemic (April 2020 to December 2020).
SETTING:
Premier Healthcare Database, containing data from 25% of U.S. discharges from January 1, 2016, to December 31, 2020.
PATIENTS:
Patients without COVID-19 receiving mechanical ventilation for acute respiratory failure.
INTERVENTIONS:
We assessed monthly rates of chest radiograph (CXR), chest CT scans, lower extremity noninvasive vascular testing (LENI), bronchoscopy, arterial catheters, and central venous catheters.
MEASUREMENTS AND MAIN RESULTS:
We identified 742,096 mechanically ventilated patients without COVID-19 at 545 hospitals. At the onset of the pandemic, CXR (–0.5% [–0.9% to –0.2%; p = 0.001]), LENI (LENI: –2.1% [–3.3% to –0.9%; p = 0.001]), and bronchoscopy rates (–1.0% [–1.5% to –0.6%; p < 0.001]) decreased; use of chest CT increased (1.5% [0.5–2.5%; p = 0.006]). Use of arterial lines and central venous catheters did not change significantly. Intra-pandemic, LENI (0.5% [0.3–0.7%; p < 0.001]/mo) and bronchoscopy (0.1% [0.05–0.2%; p < 0.001]/mo) trends increased relative to pre-pandemic trends, while the remainder of practices did not change significantly.
CONCLUSIONS:
We observed several statistically significant changes to practice patterns among patients without COVID-19 early during the pandemic. However, most of the changes were small or temporary, suggesting that routine practices in the care of mechanically ventilated patients in the ICU was not drastically affected by the pandemic.
Keywords: COVID, health services, intensive care unit, mechanical ventilation, practice patterns
KEY POINTS
Question: How did common ICU practices change at the start of and during the pandemic for patients receiving mechanical ventilation for acute respiratory failure not attributable to COVID-19?
Findings: Several practice patterns changed among patients without COVID-19 early during the pandemic, although changes were generally small or temporary.
Meaning: Understanding practices changes is the first step in preparing for future situations of increased ICU demand.
The onset of the COVID-19 pandemic increased demand for critical care worldwide and necessitated changes to resource allocation within ICUs (1). Hospitals responded in multiple ways, including: 1) increasing the number of critical care beds within ICUs and in units not typically designated as ICUs, 2) changing staffing ratios and redeploying healthcare personnel, and 3) redistributing or rationing ICU resources (2). Around March 2020 to April 2020, it was speculated that ICUs would need to make adjustments in routine care of patients without COVID-19 (3, 4); however, ICU practice changes for patients without COVID-19 across the United States have not been well characterized during this time period. Thus, we sought to evaluate changes to common diagnostic and therapeutic interventions among a group requiring high levels of ICU resources—patients with acute respiratory failure receiving invasive mechanical ventilation (MV) (5), but without COVID-19.
METHODS
We used the Premier Healthcare Database (6) from January 2016 to December 2020, containing enhanced multicenter claims data from ~25% of U.S. hospitalizations, to identify adult patients (≥ 18 yr) admitted to an ICU or step-down unit receiving MV (identified by billing codes) for acute respiratory failure (identified by International Classification of Diseases, 10th Edition, [ICD-10] codes [Supplemental Table 1, http://links.lww.com/CCX/B162] [7, 8]) during a hospitalization. Patients were excluded if they had a diagnosis of COVID-19 (ICD-10 U07.1) (Supplemental Table 2, http://links.lww.com/CCX/B162) (9). For patients with multiple hospitalizations meeting inclusion criteria, we randomly selected one hospitalization for inclusion in the study. Only hospitals with data for all 5 years were included.
We used billing codes to examine rates of common ICU imaging studies (chest radiograph [CXR], chest CT scan, and lower extremity noninvasive vascular testing [LENI]) and bedside procedures (bronchoscopy, arterial catheter placement, and central venous catheter [CVC] placement) (Supplemental Table 3, http://links.lww.com/CCX/B162). Monthly rates were calculated by dividing the number of patients receiving the imaging study or procedure of interest by the total number of patients discharged each month. As patients often receive multiple CXRs per admission, we also assessed changes in frequency of CXRs by assessing monthly trends in average number of CXRs/d.
To quantify the impact of the early pandemic, we conducted interrupted time series analyses with segmented regression to quantify: 1) the immediate level change in rate of practices between February 2020 and April 2020 and 2) the change in trends in practices, comparing April 2020 to December 2020 trends with 2016 to February 2020 trends, for each imaging study or procedure of interest and mortality. As the pandemic surge began throughout March 2020 without a singular start date, this month was censored from interrupted time series analysis (10). Models were adjusted for seasonality and first order autocorrelation where present (per Durbin-Watson test). In sensitivity analysis, quadratic models were explored for best fit by comparing Akaike information criterion. Pandemic-onset level change in rate was reassessed if quadratic models were found to be better fit. To assess the effect of clinical strain during a surge, we also performed an exploratory analysis restricting the cohort to the Northeast alone, which experienced a surge in COVID-19 beginning in March 2020 (11). Statistical testing was two-tailed with α = 0.05 using R Version 4.2.2. This study was designated not Human Subjects Research by Boston University’s Institutional Review Board on September 15, 2021 (No. H-41991, “Practice Patterns and Outcomes of Critically Ill Patients”). Procedures followed by the study were in accordance with the ethical standards of the Boston University Clinical and Translational Research Institute and with the Helsinki Declaration of 1975, as recently amended.
RESULTS
We identified 727,538 mechanically ventilated patients with non-COVID acute respiratory failure at 545 unique hospitals across the United States (Supplemental Fig. 1, http://links.lww.com/CCX/B162). Patient characteristics across all years of admission are shown in Table 1. Adjusted estimates of pre-pandemic trends, pandemic-onset level change, and intra-pandemic trend change for all practices and monthly proportions of patients receiving each imaging study or procedure are depicted in Figure 1. The proportion of patients receiving a CXR was decreasing slightly pre-pandemic; this dropped more at pandemic onset (–0.5% [–0.9% to –0.2%]; p = 0.001), and intra-pandemic trends did not change significantly. Chest CTs were increasing pre-pandemic and continued to increase at pandemic onset (1.5% [0.5–2.5%]; p = 0.006); intra-pandemic there was no significant trend change in chest CT. At pandemic onset, use of LENIs, which was not changing significantly pre-pandemic, dropped (–2.1% [–3.3% to –0.9%]; p = 0.001) followed by an increase in trend intra-pandemic. Bronchoscopy, which was decreasing pre-pandemic, dropped more at pandemic onset (–1.0% [–1.5% to –0.6%]; p < 0.001), but then saw a subsequent increase in trend intra-pandemic. Pre-pandemic, use of arterial catheters decreased, with no significant in level or trend at pandemic onset or intra-pandemic. CVC catheter rates did not change significantly in trend or level pre-, at onset, or post-pandemic.
TABLE 1.
Patient Characteristics by Date of Discharge
| Characteristic | Pre-Pandemic (January 2019 to February 2020) (n = 628,914) | After Pandemic Onset (April 2020 to December 2020) (n = 98,624) | Total (n = 727,538) |
|---|---|---|---|
| Age, mean ± sd | 62.3 (16.8) | 61.6 (16.8) | 62.2 (16.8) |
| Male (%) | 351,890 (56.0) | 57,029 (57.8) | 408,919 (56.2) |
| Race (%) | |||
| White | 454,661 (72.3) | 69,997 (71.0) | 524,658 (72.1) |
| Black | 96,278 (15.3) | 15,801 (16.0) | 112,079 (15.4) |
| Insurance (%) | |||
| Medicare | 358,294 (57.0) | 53,962 (54.7) | 412,256 (56.7) |
| Medicaid | 101,954 (16.2) | 18,252 (18.5) | 120,206 (16.5) |
| Private | 17,132 (17.4%) | 129,266 (17.8) | 112,134 (17.8) |
| Urban hospital (%) | 559,369 (88.9) | 87,491 (88.7) | 646,860 (88.9) |
| Teaching hospital (%) | 345,677 (55.0) | 55,031 (55.8) | 400,708 (55.1) |
| Hospital bed size (%) | |||
| 0–199 | 85,419 (13.5) | 13,334 (13.5) | 98,753 (13.6) |
| 200–500 | 265,240 (41.1) | 39,146 (39.6) | 304,386 (41.8) |
| 500+ | 278,255 (44.2) | 46,144 (46.8) | 324,399 (44.6) |
| Combined Comorbidity Scorea, median (minimum–maximum) | 4.00 (0–13.0) | 4.00 (0–13.0) | 4.00 (0–13.0) |
| Acute organ dysfunction (%)b(7, 8) | |||
| Cardiovascular | 328,743 (52.3) | 56,064 (56.8) | 384,807 (52.9) |
| Neurologic | 140,601 (22.4) | 20,264 (20.5) | 160,865 (22.1) |
| Hematologic | 134,860 (21.4) | 22,737 (23.1) | 157,597 (21.7) |
| Hepatic | 50,864 (8.1) | 9,263 (9.4) | 60,127 (8.3) |
| Renal | 307,865 (49.0) | 51,846 (52.6) | 359,711 (49.4) |
The Combined Comorbidity Score (12) is a summative score based on 20 comorbid conditions identified by International Classification of Diseases, 10th Edition (ICD-10) codes, adapted from the Charlson and Elixhauser scores. Score range (0–20).
Figure 1.
Unadjusted trends (lines) are plotted before and after April 2020 for each practice. Pre-pandemic, pandemic onset, and intra-pandemic trends, adjusted for seasonality and autocorrelation where present, were calculated for each practice studied nationwide. LENI = lower extremity noninvasive vascular testing.
Pre-pandemic, there was a decrease in mortality of –0.03%/mo (–0.04% to –0.01%/mo; p < 0.001). At pandemic onset there was an increase in mortality of +1.8% (0.5–3.2%; p = 0.01). Intra-pandemic, there was no significant change in mortality trend at +0.04%/mo (–0.2% to 0.3%/mo; p = 0.68). Monthly rates of mortality are seen in Supplemental Figure 2 (http://links.lww.com/CCX/B162).
Pre-pandemic, the average number of CXR/d decreased with no level change at pandemic onset. Intra-pandemic, CXRs/day declined more steeply compared with pre-pandemic trends (–0.003%/mo [–0.005% to –0.002%/mo]; p < 0.001) (Supplemental Table 4, http://links.lww.com/CCX/B162 and Supplemental Fig. 3, http://links.lww.com/CCX/B162). In sensitivity analysis, models with quadratic terms were better fit (Supplemental Table 5, http://links.lww.com/CCX/B162); however, level changes at pandemic onset were similar for all practices except for CVC placement, where a statistically significant increase was detected at pandemic onset (Supplemental Table 6, http://links.lww.com/CCX/B162).
The subgroup hospitalized in the Northeast, strongly affected by the March 2020 COVID surge, similarly showed a decrease in CXRs and LENIs; there were no statistically significant changes in the other interventions (chest CT, bronchoscopy, arterial catheters, or CVCs) at pandemic onset (Supplemental Table 7, http://links.lww.com/CCX/B162).
DISCUSSION
In this analysis of practice changes for patients with non-COVID-19 respiratory failure receiving MV before and during the COVID-19 pandemic, we found a statistically significant drop in CXR, LENI, and bronchoscopy at the pandemic onset (April 2020), followed by subsequent increases in LENI and bronchoscopy trend during the pandemic. Chest CT was the only practice that increased at pandemic onset. We did not detect changes in arterial catheter or CVC placement in primary analysis. The frequency of CXRs decreased prior to the pandemic and decreased further during the pandemic. While many changes were statistically significant, the magnitude of the changes were generally small and/or temporary, suggesting that care for mechanically ventilated patients in the ICU was largely stable at the start of the COVID-19 pandemic.
The observed trend increase in chest CT scans prior to COVID-19 (with a concomitant decrease in rates of CXRs) is consistent with literature recommending CT scan over CXR for detecting underlying causes of acute respiratory failure (13). Our observed increase of about 10% in chest CT scan use from 2016 to 2020 may reflect evolving ICU physician practice in favor of more advanced diagnostics despite no clear evidence for improved outcomes (14). The further increase in CT chest at the start of the pandemic potentially reflects clinician desire to assess for COVID-19 among other causes of respiratory failure, especially in the initial stages of the pandemic when COVID-19 testing was less available (15). In the Northeast subgroup, chest CT use did not significantly change at pandemic onset, potentially due to higher levels of strain experienced by hospitals in the Northeast around April 2020 (11) as transporting a ventilated ICU patients to the CT scanner requires staff and time (16).
The 2018 American Society of Hematology guidelines on the diagnosis of venous thromboembolism recommended use of lower extremity over d-dimer testing in those with a high suspicion of venous thromboembolism (17); this recommendation is very similar to the 2012 American College of Chest Physicians recommendations on diagnosis of Deep Vein Thrombosis (18) and likely accounts for the steady rate of LENIs seen in the pre-pandemic period. Although LENIs decreased at pandemic onset, LENI use subsequently increased during the pandemic, rising to rates higher than pre-pandemic rates. Given the known increased risk of venous thromboembolism in COVID-19 patients (19, 20), it is possible that there was increased clinical momentum to order LENIs, a practice which carried over to all patients with respiratory failure, even those without COVID.
Pre-pandemic, bronchoscopy rates were known to vary widely between centers, likely in part due to a lack of clear evidence of guidelines delineating indications for bronchoscopy in acute respiratory failure (21). Similar to LENIs, we observed a drop in bronchoscopies at the start of the pandemic, followed by gradual increase in rates back to pre-pandemic levels as the pandemic progressed. Physicians and health systems may have initially decreased bronchoscopy use due to concern for transmission of undetected infection to the bronchoscopist (22), only to gradually rebound in bronchoscopy use once evidence showed low transmission of COVID-19 with proper personal protective equipment (23). It is surprising, however, that the subgroup in the Northeast did not have a clinically significant level decrease at pandemic onset in bronchoscopy; it is possible that early lack of consensus on optimal diagnostic testing for COVID played a role in this regional difference (24).
In primary analysis, there was no significant change in rates of central and arterial catheter placements for patients without COVID at the start of the pandemic or as the pandemic progressed. Many hospitals adopted specialized procedure services—typically consisting of surgeons, anesthesiologists, and interventional radiologists impacted by cancelations in elective procedures—to place arterial and central catheters to address the increased access needs of overburdened intensivists (25, 26). It is possible that support from additional redeployed physician staff allowed for consistency in line placement for patients with and without COVID-19 (perhaps even a slight increase in CVC placement, as detected in our sensitivity analysis). Additionally, prior data has demonstrated wide practice variation between individual ICUs in central and arterial catheter use prior to the pandemic in light of Choosing Wisely campaign recommendations to only use these catheters if indicated (27). The wide variability in ICU policies likely continued through the pandemic leading to no significant trend changes in central and arterial catheter use.
We noted a statistically significant decrease in mortality at pandemic onset.
Almost all patients received at least one CXR during admission, which is not unexpected given the common use of CXR after endotracheal intubation. However, the frequency of CXRs was decreasing pre-pandemic and decreased further as the pandemic progressed. Studies have suggested that daily CXRs in the ICU do not impact clinical outcomes (28). Whether the decreased use of CXRs during the pandemic impacted patient outcomes remains to be seen.
There are limitations to our study. As with most studies using administrative data, there is a risk of misclassification of patients requiring MV for acute respiratory failure and of practices measured as outcomes. However, we relied on billing codes for MV and practices (which are directly tied to reimbursement) and used validated ICD-10 codes for acute respiratory failure, COVID-19, comorbidities, and acute organ dysfunction (7–9, 12); thus, it is unlikely that there was differential misclassification over time (29). Additionally, if indications to initiate MV fluctuated during the pandemic, cohort characteristics and indications for procedures/studies may also have changed; however, patient characteristics appeared to be generally stable throughout the study period. Last, as the statistical power of interrupted time series analyses depends on having sufficient data points in both pre- and post-periods (10), we were underpowered to perform exploratory analyses of the South, Midwest, and West which experienced pandemic surges later in 2020.
CONCLUSIONS
We observed many changes in common ICU practices at the onset of the COVID-19 pandemic among patients without COVID-19 who had acute respiratory failure requiring MV. Many of the significant changes were small, however, suggesting that dramatic shifts in routine practices for mechanically ventilated patients in the ICU did not occur at the start of the COVID-19 pandemic.
Supplementary Material
Footnotes
Dr. Shankar (0000-0002-2433-4895), Dr. Bosch (0000-0001-7161-5254), Dr. Walkey (0000-0003-4685-6894), and Dr. Law (0000-0003-3616-0316).
Dr. Shankar received funding from National Institutes of Health (NIH) T32 HL007035. Dr. Walkey received funding from NIH R01HL139751, NIH R01HL151607, NIH R01HL136660, and NIH OT2HL156812-01. Dr. Bosch received funding from NIH National Center for Advancing Translational Sciences 1KL2TR001411. Dr. Law received funding from NIH K23HL 153482, Boston University School of Medicine Department of Medicine Career Investment Award, and Doris Duke Charitable Foundation Fund to retain clinician scientists.
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website (http://journals.lww.com/ccejournal).
REFERENCES
- 1.Supady A, Curtis JR, Abrams D, et al. : Allocating scarce intensive care resources during the COVID-19 pandemic: Practical challenges to theoretical frameworks. Lancet Respir Med 2021; 9:430–434 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Wilcox ME, Rowan KM, Harrison DA, et al. : Does unprecedented ICU capacity strain, as experienced during the COVID-19 pandemic, impact patient outcome?. Crit Care Med 2022; 50:e548–e556 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Song H, Ezaz G, Greysen SR, et al. : How Hospitals Can Meet the Needs of Non-Covid Patients During the Pandemic. Harvard Business Review. 2020. Available at: https://hbr.org/2020/07/how-hospitals-can-meet-the-needs-of-non-covid-patients-during-the-pandemic. Accessed August 25, 2022
- 4.Grimm CA: Hospital Experiences Responding to the Covid-19 Pandemic: Results of a National Pulse Survey March 23–27, 2020. 2020. Available at: https://oig.hhs.gov/oei/reports/oei-06-20-00300.pdf. Accessed September 11, 2022
- 5.Stefan MS, Shieh M-S, Pekow PS, et al. : Epidemiology and outcomes of acute respiratory failure in the United States, 2001 to 2009: A national survey. J Hosp Med 2013; 8:76–82 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.White Paper: Premier Hospital Database: Premier Healthcare Database: Data That Informs and Performs. 2022. Available at: https://products.premierinc.com/downloads/PremierHealthcareDatabaseWhitepaper.pdf. Accessed February 28, 2022
- 7.Angus DC, Linde-Zwirble WT, Lidicker J, et al. : Epidemiology of severe sepsis in the United States: Analysis of incidence, outcome, and associated costs of care. Crit Care Med 2001; 29:1303–1310 [DOI] [PubMed] [Google Scholar]
- 8.Bosch NA, Law AC, Rucci JM, et al. : Predictive validity of the Sequential Organ Failure Assessment score versus claims-based scores among critically ill patients. Ann Am Thorac Soc 2022; 19:1072–1076 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bosch NA, Law AC, Peterson D, et al. : Validation of the International Classification of Diseases Code for coronavirus disease 2019 among critically ill patients. Ann Am Thorac Soc 2022; 19:861–863 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Penfold RB, Zhang F: Use of interrupted time series analysis in evaluating health care quality improvements. Acad Pediatr 2013; 13:S38–S44 [DOI] [PubMed] [Google Scholar]
- 11.Centers for Disease Control and Prevention: COVID Data Tracker. 2020. Available at: https://covid.cdc.gov/covid-data-tracker. Accessed November 9, 2022
- 12.Sun JW, Rogers JR, Her Q, et al. : Adaptation and validation of the combined comorbidity score for ICD-10-CM. Med Care 2017; 55:1046–1051 [DOI] [PubMed] [Google Scholar]
- 13.Zompatori M, Ciccarese F, Fasano L: Overview of current lung imaging in acute respiratory distress syndrome. Eur Respir Rev 2014; 23:519–530 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.van den Berk IAH, Kanglie MMNP, van Engelen TSR, et al. ; OPTIMACT study group: Ultra-low-dose CT versus chest X-ray for patients suspected of pulmonary disease at the emergency department: A multicentre randomised clinical trial. Thorax 2022. Jun 10. [online ahead of print] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Laino ME, Ammirabile A, Lofino L, et al. : Prognostic findings for ICU admission in patients with COVID-19 pneumonia: Baseline and follow-up chest CT and the added value of artificial intelligence. Emerg Radiol 2022; 29:243–262 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Aliaga M, Forel J-M, De Bourmont S, et al. : Diagnostic yield and safety of CT scans in ICU. Intensive Care Med 2015; 41:436–443 [DOI] [PubMed] [Google Scholar]
- 17.Lim W, Le Gal G, Bates SM, et al. : American Society of Hematology 2018 guidelines for management of venous thromboembolism: Diagnosis of venous thromboembolism. Blood Adv 2018; 2:3226–3256 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Bates SM, Jaeschke R, Stevens SM, et al. : Diagnosis of DVT: Antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 2012; 141:e351S–e418S [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kollias A, Kyriakoulis KG, Lagou S, et al. : Venous thromboembolism in COVID-19: A systematic review and meta-analysis. Vasc Med Lond Engl 2021; 26:415–425 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Alfageme M, González Plaza J, Méndez S, et al. : Venous Doppler ultrasound in critically ill COVID-19 patients: Game changer in anticoagulation therapy. Ultrasound J 2020; 12:54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Wayne MT, Seelye S, Molling D, et al. : Variation in US hospital practices for bronchoscopy in the intensive care unit. Ann Am Thorac Soc 2022; 19:1061–1065 [DOI] [PubMed] [Google Scholar]
- 22.Lentz RJ, Colt H: Summarizing societal guidelines regarding bronchoscopy during the COVID‐19 pandemic. Respirol Carlton Vic 2020; 25:574–577 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Chang SH, Jiang J, Kon ZN, et al. : Safety and efficacy of bronchoscopy in critically ill patients with coronavirus disease 2019. Chest 2021; 159:870–872 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Baron A, Hachem M, Tran Van Nhieu J, et al. : Bronchoalveolar lavage in patients with COVID-19 with invasive mechanical ventilation for acute respiratory distress syndrome. Ann Am Thorac Soc 2021; 18:723–726 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ilonzo N, Rao A, Soundararajan K, et al. : The importance of a centralized line service during the COVID-19 pandemic. J Vasc Surg 2020; 72:403–404 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Woo K, Rigberg D, Lawrence PF: Safe central venous access in an overburdened health system. JAMA 2021; 325:299–300 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Gershengorn HB, Garland A, Kramer A, et al. : Variation of arterial and central venous catheter use in United States intensive care units. Anesthesiology 2014; 120:650–664 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Choukalas CG, Vu T-G: The problem of daily imaging in the intensive care unit: When you care so much it hurts. JAMA Intern Med 2020; 180:1369–1370 [DOI] [PubMed] [Google Scholar]
- 29.Funk MJ, Landi SN: Misclassification in administrative claims data: Quantifying the impact on treatment effect estimates. Curr Epidemiol Rep 2014; 1:175–185 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.