Surge capacity of Australian intensive care units associated with COVID-19 admissions

This is a preprint only. The final version of this article is available at:
https://www.mja.com.au/journal/2020/212/10/surge-capacity-intensive-care-units-case-acute-increase-demand-caused-covid-19

Suggested citation: Litton E, Bucci T, Chavan S, et al. Surge capacity of Australian intensive care units associated with COVID-19 admissions. Med J Aust 2020; https://www.mja.com.au/journal/2020/212/10/surge-capacity-australian-intensive-care-units-associated-covid-19-admissions [Preprint, 30 March 2020]

UPDATED 6 APRIL 2020

The burden on health care resources associated with the global coronavirus disease 2019 (COVID-19) pandemic is unparalleled in modern times. Early experience in severely affected countries suggests that between 5 and 16% of laboratory-confirmed COVID-19 patients require admission to an Intensive Care Unit (ICU) (1, 2).

While public health measures aimed at reducing virus transmission are the primary means to reduce the overall disease burden and ICU requirement, guidelines recommend that ICUs have a coordinated local and regional surge plan that is staged to respond to increased demand (3). 

The primary aim of this study was to describe reported intensive care bed and invasive ventilation surge capacity in Australia. A secondary aim was to describe variation in these capabilities according to ICU type (tertiary, metropolitan, regional/rural, or private) and jurisdiction as well as other ICU equipment and workforce.

Methods

Data was aggregated from a number of sources, primarily using existing data from the Australian and New Zealand Intensive Care Society (ANZICS) Critical Care Resources (CCR) registry, supplemented by a survey of ICU surge capacity distributed to each ICU in Australia.

Data extracted from the ANZICS CCR 2018/19 survey, or where not available 2017/18 data were used. This included baseline available beds for all ICUs in Australia, workforce data, including number of senior and junior medical staff plus nursing staff, and elective surgical, mechanical ventilation and total ICU bed days.

The surge survey was designed by the study investigators with feedback from other critical care clinicians and health policy makers with the aim of informing local, jurisdictional and national critical care capacity. The survey contained questions on the incremental capacity to increase ICU beds [baseline available beds, additional physical ICU beds not staffed or operational during baseline activity, surge beds already configured to ICU standards, and other beds that could rapidly be converted to meet the minimal standards required for the provision of intensive care (4)]. Equipment questions included those relevant to providing invasive mechanical ventilation (standard ICU ventilators, additional ICU ventilators such as transport ventilators, anaesthetic machine ventilators, non-invasive ventilators capable of providing invasive ventilator and others), dialysis machines, extra-corporeal membrane oxygenation (ECMO) machines, intravenous infusion (IV) pumps, and workforce (senior doctors defined as consultants working in ICU, and ICU registered nursing staff). The survey is provided in the supplementary appendix (available in the PDF version).

The survey was distributed via email to the ICU director of all Australian ICUs through the ANZICS registries mailing list. ICU directors, or when unavailable the nurse unit manager, were asked to complete and return the survey as a word document via email or link to an online survey. Surge intensive care beds and equipment were identified by each ICU within their hospital. Responses were collated into a spreadsheet by ANZICS Centre for Outcome and Resource Evaluation (CORE). Non-responders were contacted twice more by email for follow up. With the assistance of the Australian Veterinary Boards Council, the Veterinary Board in each state also surveyed superintendents of every veterinary hospital across Australia about model, number and location of all machines capable of invasive ventilation.

Findings were collated and analysed in Microsoft Excel and reported using descriptive statistics. No range was provided for absolute values (e.g. number of beds). Parametric data were described using mean and standard deviation (SD), and  non-parametric data using median and interquartile range (IQR). Incremental, increasing capacity to surge based on the increase in ICU beds were compared to available ventilators and provided as both mean (SD) and median (IQR). Workforce surge were based on ANZICS CCR from respondent ICUs that was then extrapolated to all 2378 available Australian ICU beds to provide an estimated baseline senior medical staff (n=1671) and baseline registered nursing staff (n=15857) . The estimate for the total surge beds for all Australian ICUs were extrapolated from proportionate increase in beds in the 92% of respondent ICUs (i.e. maximal surge was 191% of baseline available beds). For senior medical staff, the “Best Case” estimate assumed no increase in absenteeism and that the requirement for additional staffing could be reduced by 20% in the short term,  by increasing the clinical full time equivalent (FTE) of available staff through reduced leave, part-time employees increasing FTE and by temporarily delegating components of roles outside of ICU to ward staff  (e.g. follow up clinics, tracheostomy and total parenteral nutrition rounds).  Conversely, “Worst Case” estimates made no allowance for available staff to increase FTE and assumed an additional 30% would be required to be added to the workforce to cover absenteeism. ‘Worst Case’ nursing staff estimate also included a 10% increase to account for an increase in 1:1 ratio for all patients receiving mechanical ventilation from a baseline of 50% to 70% under pandemic conditions. The study was approved by the Central Australian Human Research Ethics Committee (CAHREC-20-3687).

Results

ICU beds, ventilators and other equipment

There are 191 ICUs in Australia (119 public and 72 private) recorded in the ANZICS CCR database. At these sites there are 2378 ICU beds that are available during baseline activity. This equates to 9.4 ICU beds per 100 000 population in Australia. The surge survey was distributed between 13-24 March 2020. Responses were received from 175 (92%) of ICUs, accounting for 2228 (95%) of available beds. The results represent surge capacity for respondent ICUs (listed in the supplementary appendix, available in the PDF version).

The characteristics of respondent ICUs are presented in Table 1 (available in the PDF version). At these 175 ICUs, there are a total of 4261 additional physical intensive care beds, surge beds outside of the ICU already configured to intensive care standards, and other beds that could rapidly be converted to meet the minimal requirements of providing intensive care (Table 2, Figure 1 and 2; available in the PDF version).

At the 175 respondent sites who reported equipment data, there are 4815 machines capable of delivering invasive mechanical ventilation including 2184 standard ICU invasive ventilators, 532 additional ventilators (e.g. for transport), 1476 anaesthetic machine ventilators and 471 non-invasive ventilators capable of invasive ventilation. The proportion of ventilators to ICU beds during incremental surge capacity is provided in Table 3 (see supplementary appendix eTable 4 for ventilator to bed ratio by jurisdiction and ICU category; available in the PDF version). The availability of other ICU equipment including to deliver renal replacement therapy, ECMO and IV infusion pumps are provided in the supplementary appendix eTables 5-7 (available in the PDF version).

Workforce

Based on current workforce data from the 175 respondent ICUs, maximal surge would require an additional 4092 senior medical staff (345% above baseline staffing), and an additional 42 720 registered nursing staff would be required (365% above baseline staffing) (Table 4; available in the PDF version).

Ventilators in veterinary facilities

There are 188 invasive ventilators in 120 Australian veterinary facilities (147 in metropolitan areas and 41 in rural/regional areas). Of these, 179 (95%) were human model ventilators (supplementary appendix eTable 10 Ventilators in veterinary hospitals; available in the PDF version).

Discussion

There are 191 ICUs in Australia with 2378 available intensive care beds during baseline activity, equating to 9.4 ICU beds per 100 000 population. The 175 (92%) ICUs that responded to our surge survey reported capacity to increase intensive care bed capacity by 4258 beds, a 191% increase over baseline capacity of 2228 at these sites, with 4815 machines identified at these sites capable of delivering invasive mechanical ventilation.

Compared to baseline available Australian ICU beds of 9.4 per 100 000 population, China and the UK are substantially lower (3.6 and 4.4 per 100,000 population respectively), while the US is substantially higher (25.8 per 100 000) (5, 6). However, our survey suggests substantial capacity for Australian ICUs to surge beds (up to the equivalent of 26.5 per 100 000 with the addition of only beds in the respondent sites). This demonstrates the value of additional time provided for planning compared to the remarkable achievement, but lower immediate surge capacity associated with unexpected and sudden presentations in Northern Italy (2). Furthermore, international comparisons are complex and require consideration of additional factors including the availability of suitably trained specialists. Intensive Care Specialists are present in nearly all Australian ICUs compared with only 48% of acute care hospitals in the US (7, 8).

There are other important findings of our study. First, even under a ‘Best Case’ scenario, workforce requirements to meet maximal intensive care bed surge capacity are extremely high. Nursing workforce surge may be susceptible to strain due to a requirement for a bedside nurse to care for each additional surge patient receiving mechanical ventilation.  Strategies and resources to rapidly upskill registered nurses to manage mechanically ventilated patients are required urgently. Similarly, senior ICU doctors may be required to adopt a supervisory role over multiple 'pods' of patients, assisted by other senior clinicians with overlapping skill sets such as anaesthetists in order to maintain safe patient to physician ratios. The optimal alterations to standard ICU workforce models during surge are likely to vary by site and according to local capabilities. The association with resource use and patient outcomes is uncertain and will require further investigation.

Secondly, cross-disciplinary collaboration allowed identification of potential additional veterinary ventilators. The majority were models already in use in ICUs in Australia. These ventilators with appropriate preparation may be lower risk to use than non-ICU capable ventilators or equipment that is unfamiliar to critical care staff. In addition, the veterinary sector may also contain additional physical capacity, pharmaceutical resources and staff who could assist with provision of intensive care. Finally, our survey demonstrates substantial differences in capacity to surge between ICU category and between jurisdictions. Paediatric ICUs appear to have relatively little capacity to surge. Rural/regional ICUs have more capacity for surge than metropolitan but much less than private and tertiary ICUs.  Rural/regional ICUs have a high proportion of vulnerable patients and are reliant on interhospital transfer systems that may become overwhelmed (9). While it is unlikely that transfer of patients can occur between jurisdictions, our study demonstrates that equipment and workforce may be a greater limiter. National real-time monitoring of ICU bed capacity could provide a mechanism to rapidly redeploy resources to areas of need.

There are limitations of this study. Most importantly, surge capacity represents those beds identified at the time of reporting. Further beds could potentially be created over time and additional facilities explored both to take over urgent elective surgical ICUs admissions or other critical care admissions.  Equally, invasive ventilator capacity did not include state, commonwealth, Australian Defence Force stores or health care facilities which do not have an ICU. Reported surge capacity may also be greater or less than actual capacity. Finally, response rate to our surge survey was incomplete. However, a response rate of 92% including all tertiary ICUs is likely to be nationally representative. 

Conclusion

Australian ICUs report potential to nearly triple intensive care bed capacity in response to predicted increased demand associated with pandemic COVID-19. Maximal surge could result in an invasive ventilator shortfall and would require a large increase in workforce. There is variation between jurisdictions and greater capacity in tertiary ICUs.

References

1. Guan WJ, Ni ZY, Hu Y, et al. Clinical Characteristics of Coronavirus Disease 2019 in China.  N Engl J Med. 2020. DOI: 10.1056/NEJMoa2002032

2. Grasselli G, Pesenti A, Cecconi M. Critical Care Utilization for the COVID-19 Outbreak in Lombardy, Italy: Early Experience and Forecast During an Emergency Response. JAMA. 2020. doi:10.1001/jama.2020.4031

3. Australian and New Zealand Intensive Care Society (2020) ANZICS COVID-19 Guidelines Version 1. ANZICS 277 Camberwell Road, Melbourne 3124. https://www.anzics.com.au/coronavirus-guidelines/ (accessed 27 March 2020)

4. College of Intensive Care Medicine, 168 Greville Street, Prahan, Australia 3181 https://www.cicm.org.au/CICM_Media/CICMSite/CICM Minimum Standards for Intensive Care Units (accessed 27 Mar 2020)

5. Halpern NA. U.S. ICU Resource Availability for COVID-19. Society of Critical Care Medicine. 2020. https://sccm.org/Blog/March-2020/United-States-Resource-Availability-for-COVID-19 (accessed 27 March 2020)

6. Emanuel EJ, Persad G, Upshur R, et al. Fair Allocation of Scarce Medical Resources in the Time of Covid-19. N Engl J Med. 2020. DOI: 10.1056/NEJMsb2005114

7. Halpern NA, Tan KS, DeWitt M, Pastores SM. Intensivists in U.S. Acute Care Hospitals. Critical care medicine. 2019;47(4):517-25.

8. Litton E. Bucci T Intensive Care Resources and Activity in Australia and New Zealand - Activity Report 2017/18, ANZICS Melbourne. 2020. https://www.anzics.com.au/wp-content/uploads/2020/02/2018-CCR-Report.pdf (accessed 27 March 2020)

9. Secombe P, Brown A, McAnulty G, Pilcher D. Aboriginal and Torres Strait Islander patients requiring critical care: characteristics, resource use, and outcomes. Critical care and resuscitation. 2019;21(3):200-11.