Assumptions of US PPE Needs
Updated and revised based on Kaiser data from California and Washington
https://www.medrxiv.org/content/10.1101/2020.04.12.20062943v1
April 18, 2020
Epidemic modeling
We estimate 252,000 deaths, based on the confirmed deaths in the JHU dashboard on 4/18/20 (31,500), assuming the wave is halfway over and that only half of ultimate deaths among current cases have yet occurred and that only 50% of actual deaths are confirmed (tested).
We assume a 50% ICU in-hospital fatality rate, consistent with published case series data from Seattle, Wuhan, and Kaiser, and that 70% of deaths occur in ICU (Kaiser).
We estimate 25% of US hospitalizations need ICU care (Kaiser).
We assume hospital care is needed for 15% of clinical cases. We do not consider asymptomatic cases.
Length of stay in hospitals (Kaiser data)
Total hospital: 11 days
ICU: 8 days
PPE use in hospitals
These assumptions are based on a combination of expert judgment of ICU clinicians and actual counts of use. The fact that healthcare workers may treat multiple patients at the same time is taken into account, and in a cohorted setting, masks, respirators, and gowns may not be changed between patients within the cohort.
ICU:
Gloves: 2 gloves for each of 170 changes per patient per day. This assumes a change with each patient encounter, as per normal practice by all healthcare workers.
Gowns: Assumes that COVID patients are cohorted and that a single gown is worn for 4 hours by each healthcare worker assigned to the COVID cohort, unless it becomes visibly soiled. Accounting for all the various healthcare workers involved in the care of an ICU patient, 20 changes per patient per day.
Simple masks: Assumes that COVID patients are cohorted and that a single mask is worn for 4 hours by each healthcare worker assigned to the COVID cohort, unless it becomes visibly soiled. 10 changes per patient per day.
N95 respirators: Assumes that COVID patients are cohorted and that a single mask is worn for 4 hours by each healthcare worker assigned to the COVID cohort, unless it becomes visibly soiled. N95s are used only by healthcare workers in proximity (3 feet) to COVID patients. 6 changes per patient per day.
Non-ICU:
Gloves: 2 gloves for each of 80 changes per patient per day. This assumes a change with each patient encounter, as per normal practice by all healthcare workers.
Gowns: Assumes that COVID patients are cohorted and that a single gown is worn for 4 hours by each healthcare worker assigned to the COVID cohort, unless it becomes visibly soiled. 20 changes per patient per day.
Simple masks: Assumes that COVID patients are cohorted and that a single mask is worn for 4 hours by each healthcare worker assigned to the COVID cohort, unless it becomes visibly soiled. 10 changes per patient per day.
N95 respirators: Respirators worn only for nebulizer treatments, intubations, and other aerosol-generating procedures. An average of 2.6 changes per patient per day.
ED:
Gloves: 2 gloves for each of 25 changes per visit. This assumes a change with each patient encounter, as per normal practice by all healthcare workers.
Gowns: Assumes that COVID patients are cohorted and that a single gown is worn for 4 hours by each healthcare worker assigned to the COVID cohort, unless it becomes visibly soiled. An average of 1 change per visit.
Simple masks: Assumes that COVID patients are cohorted and that a single mask is worn for 4 hours by each healthcare worker assigned to the COVID cohort, unless it becomes visibly soiled. An average of 1 change per visit.
N95 respirators: Respirators worn only for nebulizer treatments, intubations, and other aerosol-generating procedures. An average of 1 change per visit.
PPE use in outpatient settings
Number of visits: We assume 25% of clinical cases will seek an in-person outpatient visit (considering many cases may be handled by telemedicine).
Gloves: 2 gloves for each of 25 changes per visit. This assumes a change with each patient encounter, as per normal practice by all healthcare workers.
Gowns: Assumes that COVID patients are cohorted and that a single gown is worn for 4 hours by each healthcare worker assigned to the COVID cohort, unless it becomes visibly soiled. An average of 1 change per visit.
Simple masks: Assumes that COVID patients are cohorted and that a single mask is worn for 4 hours by each healthcare worker assigned to the COVID cohort, unless it becomes visibly soiled. An average of 1 change per visit.
N95 respirators: Respirators worn only for nebulizer treatments, intubations, and other aerosol-generating procedures. An average of 1 change per visit.
PPE use in nursing homes
We assume a 10% attack rate across 1.5 million nursing home residents in the United States.
We assume COVID patients are cohorted.
Gloves: 2 gloves for each of 12 changes per patient per day. This assumes a change with each patient encounter, as per normal practice by all healthcare workers.
Gowns: Assumes that COVID patients are cohorted and that a single gown is worn for 4 hours by each healthcare worker assigned to the COVID cohort, unless it becomes visibly soiled. An average of 3 changes per visit.
Simple masks: Assumes that COVID patients are cohorted and that a single mask is worn for 4 hours by each healthcare worker assigned to the COVID cohort, unless it becomes visibly soiled. An average of 1.5 changes per visit.
N95 respirators: Not anticipated to be used in this setting.
PPE use by EMS
We assume 10% of COVID hospital admissions arrive by ambulance, in addition to the normal EMS volume of 49,315 runs per day.
Gloves: 2 gloves for each of 3 crew members for all runs.
Gowns: 1 gown for each of 3 crew members per COVID suspected run.
Simple masks: 1 mask for each of 3 crew members for all runs.
N95 respirators: 1 respirator for an average of 1 crew member for each COVID-like run.
Mask use by the general public
We assume 10% of the US population of 330 million are essentially house-bound and will not need masks. Of the remaining 300 million, we assume 75% will adhere to guidance and wear 1 mask per 5 days on average.
PPE Calculations
Download the spreadsheet used for the calculations.