0
views
0
recommends
+1 Recommend
2 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      Estimating the Impact of COVID-19 on Invasive Mechanical Ventilation Trends in the United States

      letter
      * , , , ,
      Annals of the American Thoracic Society
      American Thoracic Society

      Read this article at

      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          To the Editor: The coronavirus disease (COVID-19) pandemic has strained intensive care unit (ICU) resources (1), leading to concerns regarding the capacity to deliver mechanical ventilation (2, 3). However, increases in the burden of invasive mechanical ventilation (IMV) because of the COVID-19 pandemic have not been quantified. Estimates of the increased need for IMV during the pandemic can inform expectations and preparations for future pandemics. Therefore, we sought to quantify the change in the overall burden of IMV attributable to COVID-19 during the year 2020. Methods We used the Premier Healthcare Database (4), an enhanced, claims-based database from approximately 20% of U.S. hospitalizations, to identify adult ICU or step-down unit (SDU) patients (age 18 years or more) hospitalized between January 1, 2016, and December 31, 2020, who received IMV (using billing codes [5]). We limited our data to hospitals that reported data to the Premier database during each month of the study period. For each patient, we determined the number of hospital days during which IMV was billed (IMV-days). We then calculated the total number of IMV-days per month (primary outcome) by summing the IMV-days of all patients. As a secondary outcome, we calculated the monthly rate of IMV per 1,000 ICU + SDU patients (monthly number of ICU and SDU patients with IMV billing codes divided by the total number of ICU and SDU patients admitted that month, multiplied by 1,000). We used interrupted time series (ITS) with segmented regression (6–9) to evaluate changes in IMV trends as a result of the COVID-19 pandemic during the year 2020. Using ITS, we determined the trends of IMV-days and rates of IMV before the pandemic (January 2016 to February 2020), as well as trend changes that occurred near the onset of the COVID-19 pandemic, and during the pandemic period (April 2020 to December 2020). We used the typical ITS method of linear modeling for all trends before the pandemic. Conversely, we modeled trends during the pandemic with quadratic and cubic terms to account for nonlinear trend fluctuations resulting from spikes in COVID-19 cases. We chose cubic modeling as the primary analysis on the basis of superior model fit (Akaike information criterion [10]). To estimate the expected values of IMV-days and rates of IMV during the pandemic period had the COVID-19 pandemic never occurred, we used trends from before the pandemic to model counterfactual outcomes. Finally, we quantified the cumulative impact of COVID-19 on IMV during the pandemic period, by summing the differences between fitted model values from the ITS analysis and the theoretical counterfactual values. We adjusted all ITS models for season, accounted for first-degree autocorrelation, and excluded March 2020 from analysis (because evolving IMV use during that month was unlikely to align with trends before or during the pandemic). We conducted several sensitivity analyses, including March 2020 as a month during the pandemic, as well as using linear and quadratic terms to model IMV trends during the pandemic. To investigate possible geographic variation in the impact of COVID-19 on IMV trends, we performed the primary ITS analysis within distinct U.S. regions (Midwest, Northeast, South, and West). The study was deemed not human subjects research by Boston University Institutional Review Board. All analyses were performed in R (version 4.1.2). Results We included 1,199,986 ICU or SDU admissions (n = 1,010,837 before the pandemic and n = 189,149 during the pandemic) across 423 participating U.S. hospitals between 2016 and 2020. Hospitals came from one of four U.S. regions, including the Midwest (107 [25%]), Northeast (49 [12%]), South (202 [48%]), and West (65 [15%]). The median monthly number of IMV-days before the pandemic was 90,439 (interquartile range [IQR] 85, 731–95,755), compared with 121,296 IMV-days (IQR 117, 162–121,996) during the pandemic. The ITS model showed IMV-days were stable before the pandemic (+38 IMV-days per month, 95% confidence interval [CI], −73 to 149) with a nonstatistically significant increase near the start of the pandemic in the United States (+12,471 IMV-days [95% CI, −10,210 to 35,151]) (Figure 1A). From April 2020 to December 2020, there was a cumulative increase of 305,610 IMV-days (38% increase) above the expected values on the basis of trends before the pandemic. There was monthly variation in the increase above expected values, with the months of April 2020 to May 2020 and October 2020 to December 2020 contributing 190,615 (62%) of the excess IMV-days. Figure 1. Mechanical ventilation trends in the United States before and after the start of the coronavirus disease (COVID-19) pandemic. (A) Changes in invasive mechanical ventilation days (IMV-days). Each dot represents the monthly number of invasive mechanical ventilation days. The blue regression lines demonstrate observed trends in IMV-days before and during the COVID-19 pandemic, with shaded gray areas representing 95% confidence intervals. The orange dashed line represents the counterfactual trend of IMV-days on the basis of trends before the pandemic, estimating IMV-days had the COVID-19 pandemic not occurred. (B) Changes in the rate of invasive mechanical ventilation. Each dot represents the monthly rate of IMV per 1,000 ICU + SDU admissions. The blue segmented regression lines demonstrate the trends in rates of IMV before and during the COVID-19 pandemic, with shaded gray areas representing 95% confidence intervals. The orange dashed line represents the counterfactual trend of IMV rate per 1,000 ICU + SDU admits on the basis of trends before the pandemic, estimating the IMV rate had the COVID-19 pandemic not occurred. ICU + SDU = intensive care unit or step-down unit. The median monthly rate of IMV use before the pandemic was 152 per 1,000 ICU + SDU patients (IQR, 148–157 per 1,000), compared with 159 per 1,000 (IQR, 152–161 per 1,000) during the pandemic. The ITS model demonstrated a decrease in the monthly rate of ICU + SDU patients receiving IMV before the pandemic (−0.3 per 1,000; 95% CI, −0.4 to −0.2 per 1,000), with an immediate increase to 79 per 1,000 (95% CI, 64–94 per 1,000) around the onset of the pandemic (Figure 1B). From April 2020 to December 2020, an additional 183 ICU + SDU patients per 1,000 received IMV, representing a 15% increase in the median monthly rate of IMV among ICU admissions compared with the expected values. Monthly variation was again present, with the months of April 2020 to May 2020 and October 2020 to December 2020 accounting for 143 (78%) of the additional 183 ICU + SDU patients per 1,000 who received IMV. Sensitivity analyses were similar to primary findings (Table 1). Subgroup analyses demonstrated regional differences in which months contributed most to differences in trends above expected values. However, the aggregate estimates during the first 9 months of the pandemic in the United States were comparable across regions and to the primary results (Table 2). Table 1. Difference in outcomes in the United States during the first 9 months of the COVID-19 pandemic compared with expected values on the basis of trends before the pandemic using different strategies of modeling data during the pandemic   Total increase in invasive mechanical ventilation days before vs. during the COVID-19 pandemic (relative change), n (%) Total increase in the rate of intensive care unit and step-down unit patients/1,000 receiving invasive mechanical ventilation before vs. during the COVID-19 pandemic (relative change in median monthly rate), n (%) Primary analysis      Modeling COVID-19 period with cubic term 305,610 (38) 182 (15) Sensitivity analyses      Modeling COVID-19 period with linear term 304,302 (38) 194 (16)  Modeling COVID-19 period with quadratic term 305,444 (38) 183 (15)  Including March 2022 as a COVID-19 month 312,700 (34) 200 (15) Definition of abbreviation: COVID-19 = coronavirus disease. Table 2. Regional differences in outcomes in the United States during the first 9 months of the COVID-19 pandemic compared with expected values on the basis of trends before the pandemic   Total increase in invasive mechanical ventilation days before vs. during the COVID-19 pandemic (relative change), n (%) Total increase in the rate of intensive care unit and step-down unit patients/1,000 receiving invasive mechanical ventilation before vs. during the COVID-19 pandemic (relative change in median monthly rate), n (%) Region        Midwest 62,616 (37) 191 (15)    Northeast 40,518 (33) 280 (18)  South 154,405 (38) 179 (14)  West 48,128 (43) 135 (14) Definition of abbreviation: COVID-19 = coronavirus disease. Discussion In this study of the impact of COVID-19 on the use of IMV in the United States, there was a 38% increase in total IMV-days and a 15% increase in the median monthly rate of ICU + SDU patients that received IMV during the first 9 months of the COVID-19 pandemic in the United States, as compared with expected values on the basis of trends before the pandemic. Understanding the epidemiology of IMV among critically ill patients with COVID-19 is important for pandemic planning. Although studies have evaluated changes in ICU burden and rates of IMV throughout the various stages of the COVID-19 pandemic (11, 12), there remains a lack of data estimating changes in overall IMV burden compared with trends before the pandemic. This study expands current understanding by quantifying the immediate and cumulative impact of COVID-19 on IMV during the first 9 months of the pandemic in the United States. The relative increases in IMV-days and rates of IMV that we quantified can inform estimations of ventilator and staffing resources that may be necessary to care for patients during future pandemics. Despite expected regional variation in monthly changes in IMV trends (because of variation in the timing of local surges), the cumulative effect of the COVID-19 pandemic on the IMV burden was similar nationwide from April 2020 to December 2020. The comparable overall impact of the pandemic across regions suggests that early, local observations of IMV increases are likely generalizable and predictive of demand elsewhere. In addition, the national pandemic response may be enhanced by leveraging regional strategies that have shown success in expanding IMV delivery. Separately, we demonstrated that the number of IMV-days increased disproportionately to the rate of ICU + SDU patients who received IMV, suggesting that ventilator availability during a pandemic may be further strained by the increased duration of IMV. This increase in the duration of IMV highlights that pandemic planning should expand beyond just increasing ventilator capacity. To optimize IMV delivery, it may be equally important to promote sedation strategies that enhance ventilator liberation, determine the optimal timing of tracheostomy, regularly involve palliative care services, and leverage support of long-term acute care facilities. Further study of these care elements is warranted to inform future pandemic responses. Strengths and Limitations This study has important strengths. It uses a large, multicenter database, employs a quasiexperimental design that accounts for temporal trends, and yields comparable findings across a range of sensitivity analyses. Limitations include the assumption that IMV trends in the United States before the pandemic would have continued had COVID-19 not occurred and that the 20% of ICU + SDU admits in the United States included in the data source is not necessarily representative of the total U.S. patient population. In addition, future respiratory pandemics may not have similar effects on IMV use. Conclusion From April 2020 to December 2020, the COVID-19 pandemic resulted in a 38% increase in IMV-days and a 15% increase in median monthly rates of IMV per 1,000 ICU + SDU patients across U.S. hospitals. This information can be applied to local data to anticipate ventilator needs during future pandemics.

          Related collections

          Most cited references11

          • Record: found
          • Abstract: found
          • Article: not found

          Critical Supply Shortages — The Need for Ventilators and Personal Protective Equipment during the Covid-19 Pandemic

          New England Journal of Medicine
            Bookmark
            • Record: found
            • Abstract: found
            • Article: found

            A Framework for Rationing Ventilators and Critical Care Beds During the COVID-19 Pandemic

              Bookmark
              • Record: found
              • Abstract: found
              • Article: not found

              Segmented regression analysis of interrupted time series studies in medication use research.

              Interrupted time series design is the strongest, quasi-experimental approach for evaluating longitudinal effects of interventions. Segmented regression analysis is a powerful statistical method for estimating intervention effects in interrupted time series studies. In this paper, we show how segmented regression analysis can be used to evaluate policy and educational interventions intended to improve the quality of medication use and/or contain costs.
                Bookmark

                Author and article information

                Journal
                Ann Am Thorac Soc
                Ann Am Thorac Soc
                AnnalsATS
                Annals of the American Thoracic Society
                American Thoracic Society
                2329-6933
                2325-6621
                1 December 2022
                1 December 2022
                1 December 2023
                : 19
                : 12
                : 2105-2108
                Affiliations
                Boston University School of Medicine

                Boston, Massachusetts
                Author notes
                [* ]Corresponding author (e-mail: justin.rucci@ 123456bmc.org ).
                Author information
                https://orcid.org/0000-0002-6400-6060
                Article
                202205-467RL
                10.1513/AnnalsATS.202205-467RL
                9743481
                35857409
                41fec156-756f-4b76-b231-56e9fa31c0e7
                Copyright © 2022 by the American Thoracic Society

                This article is open access and distributed under the terms of the Creative Commons Attribution Non-Commercial No Derivatives License 4.0. For commercial usage and reprints, please e-mail dgern@ 123456thoracic.org .

                History
                Page count
                Figures: 1, Tables: 2, References: 12, Pages: 4
                Funding
                Funded by: National Institutes of Health (NIH)/National Heart, Lung, and Blood Institute (NHLBI)
                Award ID: R01HL151607
                Award ID: R01HL139751
                Award ID: R01HL136660
                Funded by: Agency of Healthcare Research and Quality
                Award ID: R01HL151607
                Funded by: Boston Biomedical Innovation Center/NIH/NHLBI
                Award ID: 5U54HL119145-07
                Categories
                Letters

                Comments

                Comment on this article