Hospital occupancy in the Omicron wave

Hospital occupancy during the Omicron wave in Lithuania

The Omicron wave is different

As has been seen in many other countries, the Omicron wave of COVID-19 has shown a weaker link between case incidence and hospitalisation data. This may be attributed to the variant itself, to prior exposure to other variants, or to increased vaccination rates.

Lithuania is not much different and using data from the Official Statistics Portal we can compare time-averaged values for case incidence, hospital admissions with COVID-19, admissions requiring oxygen, admissions requiring mechanical ventilation and deaths due directly to COVID.1

The Omicron wave compared to others

As seen elsewhere, hospitalisation and deaths are much less correlated with incidence in the latest wave than in previous waves. This graph (similar to graphs from the Financial Times by John Burn-Murdoch) adjusts all curves, setting 100 as the maximum of the December 2020 wave (Lithuania’s pre-Omicron maximum). This also shows the time delay seen between the curves in previous waves.

Show code 
lt_hosp_cv19 <- lt_hosp_data %>%
  filter(str_detect(event, "^cv19.*admissions")
         |str_detect(event, "^cv19.*in_use")
         |str_detect(event, "^cv19.*delta")
         |str_detect(event, "^cv19.*occupied")) %>%
  mutate( event = gsub("cv19_","", event)) %>%
  mutate( event = gsub("_r7m.*","", event)) %>%
  mutate( event = str_to_title(gsub("_"," ", event))) %>%
    mutate(event = case_when(
    event == "Admissions Oxygen" ~ "Admissions with Oxygen",
    event == "Admissions Mv"     ~ "Admissions on Ventilator",
    event == "Deaths Cov1"       ~ "Deaths", 
    #event == "Hospital Beds Occupied" ~ "Admissions with Oxygen",
    
    #event == "Admissions Mv"     ~ "Admissions on Ventilator",
    TRUE ~ event
  ))
events_to_plot <- c(
               "Hospital Beds Occupied",
               "Oxygen Beds Occupied",
              # "Mv Beds Occupied",
               "Icu Beds Occupied"
              # "Icu Mv Beds Occupied", 
              # "Icu Ecmo In Use",
              #"Admissions", 
              #"with Oxygen",
              #"on Ventilator"
              )

lt_case_deaths <- lt_aggregate %>%
  group_by(date) %>%
  summarise(across(where(is.numeric), ~ sum(.x, na.rm=TRUE)), .groups = "drop") %>%
  mutate(date = as_date(date)) %>%
  pivot_longer(
    cols = !c("date"),
    values_to = "count",
    names_to = "event") %>%
    mutate( event = str_to_title(gsub("_"," ", event))) %>%
  group_by(event) %>%
  mutate(count = zoo::rollmean(count,k=7, fill=NA, align="right") ) %>%
  ungroup()

lt_hosp_cv19_max <- lt_hosp_cv19 %>%
  bind_rows(lt_case_deaths) %>%  
  filter(date > ymd("2020-11-01") & date < ymd("2021-7-01") ) %>%
  group_by(event) %>%
  summarise(max = max(count)) %>%
  # fudge things slightly: admissions must be > admissions O2 so we'll set it to 250
  mutate(max = if_else(event == "Admissions", 250, max))
  
lt_hosp_nrml <- left_join(bind_rows(lt_hosp_cv19, lt_case_deaths), lt_hosp_cv19_max, by=c("event")) %>%
  # mutate( count = case_when(
  #   event=="Admissions" && count == 0  ~ NA_real_,
  #   TRUE ~ count ) ) %>%
  # filter(!(event == "Admissions" & count == 0)) %>%
  mutate(normalised= count / max * 100) %>%
    filter(event %in% c(
      #"Icu Ecmo In Use",
                      #  "Admissions",
                        # "with Oxygen",
                        # "on Ventilator",
                        "Hospital Beds Occupied",
                        "Oxygen Beds Occupied",
                        # "Mv Beds Occupied",
                        "Icu Beds Occupied",
                      "New Cases",
                      "Deaths Cov1"
                      #"Icu Ecmo In Use"
                      )) %>%
  mutate(event = case_when(
    event == "Deaths Cov1"            ~ "Deaths", 
    event == "Hospital Beds Occupied" ~ "COVID-19 Beds",
    event == "Oxygen Beds Occupied"   ~ "Beds with O2",
    event == "Icu Beds Occupied"      ~ "ICU",
    TRUE ~ event
  )) %>%
  #filter(!(event=="Admissions" && normalised == 0)) %>%
  arrange(date)

lt_hosp_nrml %>%
  ggplot(aes(x=date,y=normalised, colour=event)) +
  geom_line(size=1) +
    geom_text_repel(aes(x=date,y=normalised,label=event,colour=event),
                  nudge_x=100,
                  direction="y",hjust="right",
                  data = arrange(lt_hosp_nrml,event,date)%>%
                    group_by(event)%>%
                    summarise(across(everything(), last)) ) +
                  #data=group_by(lt_hosp_nrml%>%filter(date == ymd("2022-01-13")), event)%>%tail(6)) +
  geom_hline(yintercept = 100) +
  ft_theme() +
  scale_color_ft("origami primary") +
  theme(legend.position = "none") +
    labs(title="Lithuania - Case, Hospitalisation and Deaths data",
       subtitle="7 day averages - December 2020 wave max = 100",
       y="",
       x="Date",
       caption=paste0("", caption_text)) +
  scale_x_date(limits = c(ymd("2020-12-01",NA)),expand = expansion(add = c(0,100))) +
  scale_y_continuous(limits = c(0,NA))

Hospital occupancy since late 2020

This plot shows COVID-19 related hospital bed occupancy since late 2020.

Show code 
lt_hosp_occupancy_labelled <- lt_hosp_cv19 %>%
  filter( event %in% events_to_plot) %>%
    mutate(event = case_when(
    event == "Deaths Cov1"            ~ "Deaths", 
    event == "Hospital Beds Occupied" ~ "COVID-19 Beds",
    event == "Oxygen Beds Occupied"   ~ "Beds with O2",
    event == "Icu Beds Occupied"      ~ "ICU",
    TRUE ~ event
  ))

lt_hosp_occupancy_labelled %>%
  ggplot(aes(x=date,y=count, colour=event)) +
  geom_line(size=1) +
  geom_text_repel(aes(x=date,y=count,label=event,colour=event),
                  nudge_x=75,
                  direction="y",hjust="left",
                  data = arrange(lt_hosp_occupancy_labelled,event,date)%>%
                    group_by(event)%>%
                    summarise(across(everything(), last))) +
  ft_theme() +
  scale_color_ft("teal") +
    #scale_colour_viridis_d() +
  theme(legend.position = "none") +
    labs(title="Lithuania - Hospital bed occupancy",
       subtitle="7 day means",
       y="Beds",
       x="Date",
       caption=caption_text
         ) +
   scale_x_date(limits = c(ymd("2020-12-01",NA)),expand = expansion(add = c(0,100)))

Data source

These charts are drawn using data published by the Official Statistics Portal (OSP) on their COVID-19 open data site.

The R markdown source is available as a github repo.

Dataset description (google translation)

google translation of original Lithuanian2

7-day averages for available and occupied inpatient beds, oxygen-supplied beds, beds in intensive care and resuscitation wards, artificial lung ventilation equipment and its use, ECMO systems and their occupancy, patient flow, and amounts of Remdesivir.

Data is updated in the morning at 10:00.

Description of information provided:

field description
date reporting day (end date of the reporting period 16: 00-16: 00)
hospital_name the name of the hospital
healthcare_region region (five territorial health insurance funds)
hospital_beds_total_r7mean total number of inpatient hospital beds (without RITS!)
hospital_beds_occupied_r7mean number of inpatient hospital beds (without RITS!)
oxygen_beds_total_r7mean all beds with oxygen supply (part of hospital_beds_total)
oxygen_beds_occupied_r7mean occupied beds with oxygen supply
mv_beds_total_r7mean all beds with artificial lung ventilation (DPV) (part of hospital_beds_total)
mv_beds_occupied_r7mean busy beds with DPV options
icu_beds_total_r7mean resuscitation and intensive care (RITS) beds (not part of hospital_beds_total!)
icu_beds_occupied_r7mean busy RITS beds (not part of hospital_beds_occupied!)
icu_ecmo_systems_total_r7mean total number of RITS beds with ECMO systems
icu_ecmo_systems_in_use_r7mean number of occupied RITS beds with ECMO systems
cv19_hospital_beds_assigned_r7mean Number of beds for COVID patients
cv19_hospital_beds_occupied_r7mean number of occupied COVID beds
cv19_oxygen_beds_assigned_r7mean COVID oxygen beds for patients
cv19_oxygen_beds_occupied_r7mean occupied oxygen beds for COVID patients
cv19_mv_beds_assigned_r7mean COVID Patient Beds with Artificial Lung Ventilation (DPV)
cv19_mv_beds_occupied_r7mean occupied COVID patient beds with DPV capabilities
cv19_icu_beds_assigned_r7mean COVID Patient Resuscitation and Intensive Care (RITS) Beds
cv19_icu_beds_occupied_r7mean busy RITS beds for COVID patients
cv19_icu_mv_beds_assigned_r7mean COVID Patient RITS Beds with Artificial Lung Ventilation (DPV)
cv19_icu_mv_beds_occupied_r7mean occupied RITS beds for COVID patients with DPV capability
cv19_icu_highflow_systems_total_r7mean COITS for RV beds with high-flow system
cv19_icu_highflow_systems_in_use_r7mean busy COITS RITS beds with high-flow system
cv19_icu_ecmo_assigned_r7mean RITS beds for COVID patients with EKMO system
cv19_icu_ecmo_in_use_r7mean busy RITS beds for COVID patients with EKMO system
cv19_admissions_r7mean new COVID patients (daily)
cv19_admissions_oxygen_r7mean new COVID patients receiving oxygen
cv19_admissions_mv_r7mean new COVID patients with artificial lung ventilation
cv19_discharges_r7mean COVID patient checkouts / departures
remdesivir_vials_used_r7mean used Remdesivir vials
remdesivir_patients_total_r7mean Number of patients receiving Remdesivir
remdesivir_patients_new_r7mean Number of new patients receiving Remdesivir

  1. Lithuania publishes mortality data according to three different levels of attribution to COVID. The most restrictive is used here.↩︎

  2. https://open-data-ls-osp-sdg.hub.arcgis.com/datasets/f8f5c84b20e24ff99be4dd472d377640_0/about↩︎

Corrections

If you see mistakes or want to suggest changes, please create an issue on the source repository.

Reuse

Text and figures are licensed under Creative Commons Attribution CC BY 4.0. Source code is available at https://github.com/richardmn/lt_covid_calcs, unless otherwise noted. The figures that have been reused from other sources don't fall under this license and can be recognized by a note in their caption: "Figure from ...".