Norway’s Hospital Crisis: Where the Beds and Surgeries Disappeared

SSB

health

hospitals

demographics

Two decades of somatic hospital data reveal a quiet transformation in Norwegian healthcare delivery

Published

March 16, 2026

The silent reshaping of Norwegian healthcare

While political debates rage over healthcare funding, the actual transformation of Norway’s hospital system has been quiet, gradual, and dramatic. Beds have disappeared, procedures have moved outward, and the entire model of somatic care has fundamentally changed since the early 2000s. This isn’t a story of collapse — it’s a story of efficiency meeting demographic pressure in ways that don’t always show up in the headlines.

Code

library(tidyverse)
library(PxWebApiData)
library(lubridate)
library(scales)
library(ggridges)
library(MetBrewer)
library(patchwork)

# Color palette from Met Brewer (Hiroshige)
pal <- met.brewer("Hiroshige", 8)

Fetching hospital activity data

We’re pulling two decades of data from SSB’s comprehensive somatic hospital statistics (table 06913), which tracks everything from bed counts to surgical procedures across Norway’s regions.

Code

df <- NULL

tryCatch({
  raw <- ApiData(
    "https://data.ssb.no/api/v0/no/table/06913",
    Region = "0",  # Whole country
    ContentsCode = TRUE,  # Get all available metrics
    Tid = list(filter = "top", values = 25)
  )
  
  tmp <- raw[[1]]
  print(names(tmp))
  
  # Find time column defensively
  time_col <- names(tmp)[grepl(
    "tid|år|kvartal|måned|aar|maaned|year|month|quarter",
    names(tmp), ignore.case = TRUE, perl = TRUE
  )][1]
  if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L]
  message("Time column detected: ", time_col)
  
  # Find the contents/statistics variable column
  stats_col <- names(tmp)[grepl("statistikk|contents", names(tmp), ignore.case = TRUE)][1]
  if (is.na(stats_col)) stats_col <- names(tmp)[length(names(tmp)) - 2L]
  message("Statistics column detected: ", stats_col)
  
  df <- tmp |>
    mutate(
      value = as.numeric(value),
      time_str = .data[[time_col]],
      metric = .data[[stats_col]],
      year = as.integer(time_str)
    ) |>
    filter(!is.na(value), !is.na(year))
  
  message("Successfully fetched ", nrow(df), " rows covering ", 
          n_distinct(df$year), " years and ",
          n_distinct(df$metric), " metrics")
  
}, error = function(e) {
  message("Data fetch failed: ", e$message)
})

[1] "region"             "statistikkvariabel" "år"                
[4] "value"              "NAstatus"

The great bed exodus

Hospital beds per capita tell the clearest story of how Norwegian healthcare has transformed. Since 2000, the number of available somatic beds has plummeted while the population has grown — a deliberate policy shift toward shorter stays and outpatient care.

Code

if (!is.null(df)) {
  
  # Filter for bed-related metrics
  beds_df <- df |>
    filter(grepl("senger|Senger|beds|Beds", metric, ignore.case = TRUE)) |>
    arrange(year)
  
  if (nrow(beds_df) > 0) {
    
    # Create area chart with annotations
    p1 <- ggplot(beds_df, aes(x = year, y = value)) +
      geom_area(fill = pal[3], alpha = 0.7) +
      geom_line(color = pal[1], linewidth = 1.2) +
      geom_point(data = beds_df |> filter(year == max(year) | year == min(year)),
                 color = pal[1], size = 3) +
      geom_text(data = beds_df |> filter(year == max(year)),
                aes(label = paste0(comma(value), "\n", year)),
                vjust = -0.5, hjust = 1, color = pal[1], 
                fontface = "bold", size = 4) +
      geom_text(data = beds_df |> filter(year == min(year)),
                aes(label = paste0(comma(value), "\n", year)),
                vjust = -0.5, hjust = 0, color = pal[1],
                fontface = "bold", size = 4) +
      scale_y_continuous(labels = comma, expand = expansion(mult = c(0, 0.15))) +
      scale_x_continuous(breaks = seq(2000, 2025, 5)) +
      labs(
        title = "The Vanishing Hospital Bed",
        subtitle = "Total number of somatic hospital beds in Norway has declined steadily for two decades",
        caption = "Source: Statistics Norway (SSB), table 06913",
        x = NULL,
        y = "Number of beds"
      ) +
      theme_minimal(base_size = 13) +
      theme(
        plot.title = element_text(face = "bold", size = 16),
        plot.subtitle = element_text(color = "grey40", margin = margin(b = 15)),
        panel.grid.minor = element_blank(),
        axis.text = element_text(size = 11)
      )
    
    print(p1)
  }
}

Patient throughput acceleration

While beds have vanished, the number of patient discharges tells a more complex story. Hospitals are handling more cases in less time — the efficiency revolution in action.

Code

if (!is.null(df)) {
  
  # Filter for discharge metrics
  discharge_df <- df |>
    filter(grepl("utskrivn|discharge|pasienter behandlet", 
                 metric, ignore.case = TRUE)) |>
    arrange(year)
  
  if (nrow(discharge_df) > 0) {
    
    # Calculate year-over-year change
    discharge_change <- discharge_df |>
      arrange(year) |>
      mutate(
        pct_change = (value / lag(value) - 1) * 100,
        direction = ifelse(pct_change >= 0, "Increase", "Decrease")
      ) |>
      filter(!is.na(pct_change))
    
    # Lollipop chart of recent changes
    recent_change <- discharge_change |>
      filter(year >= max(year) - 10)
    
    p2 <- ggplot(recent_change, aes(x = year, y = pct_change, color = direction)) +
      geom_hline(yintercept = 0, linewidth = 0.8, color = "grey30") +
      geom_segment(aes(xend = year, yend = 0), linewidth = 1.2) +
      geom_point(size = 4) +
      scale_color_manual(values = c("Increase" = pal[5], "Decrease" = pal[2])) +
      scale_x_continuous(breaks = seq(2010, 2025, 2)) +
      labs(
        title = "Year-to-Year Patient Volume Volatility",
        subtitle = "Annual percentage change in hospital discharges shows efficiency gains mixed with capacity constraints",
        caption = "Source: Statistics Norway (SSB), table 06913",
        x = NULL,
        y = "Annual change (%)",
        color = NULL
      ) +
      theme_minimal(base_size = 13) +
      theme(
        plot.title = element_text(face = "bold", size = 16),
        plot.subtitle = element_text(color = "grey40", margin = margin(b = 15)),
        panel.grid.minor = element_blank(),
        legend.position = "top",
        axis.text = element_text(size = 11)
      )
    
    print(p2)
  }
}

The day surgery revolution

One of the most dramatic shifts in Norwegian healthcare has been the explosion of day surgeries — procedures where patients go home the same day. This metric captures the move toward minimally invasive techniques and outpatient care models.

Code

if (!is.null(df)) {
  
  # Filter for day surgery and inpatient surgery metrics
  surgery_df <- df |>
    filter(grepl("dagkirurgi|day.*surgery|døgnkirurgi|inpatient", 
                 metric, ignore.case = TRUE)) |>
    mutate(
      surgery_type = case_when(
        grepl("dag|day", metric, ignore.case = TRUE) ~ "Day surgery",
        grepl("døgn|inpatient", metric, ignore.case = TRUE) ~ "Inpatient surgery",
        TRUE ~ "Other"
      )
    ) |>
    filter(surgery_type != "Other") |>
    arrange(year)
  
  if (nrow(surgery_df) > 0 && n_distinct(surgery_df$surgery_type) > 1) {
    
    # Create bump chart showing the crossover
    p3 <- ggplot(surgery_df, aes(x = year, y = value, color = surgery_type)) +
      geom_line(linewidth = 2.5, alpha = 0.8) +
      geom_point(data = surgery_df |> filter(year == max(year) | year == min(year)),
                 size = 4) +
      geom_text(data = surgery_df |> filter(year == max(year)),
                aes(label = surgery_type),
                hjust = -0.1, size = 4, fontface = "bold") +
      scale_color_manual(values = c("Day surgery" = pal[6], "Inpatient surgery" = pal[4])) +
      scale_y_continuous(labels = comma) +
      scale_x_continuous(breaks = seq(2000, 2025, 5), 
                        limits = c(min(surgery_df$year), max(surgery_df$year) + 2)) +
      labs(
        title = "The Great Surgical Shift: Outpatient Care Takes Over",
        subtitle = "Day surgeries have overtaken traditional inpatient procedures as Norway's hospitals transform care delivery",
        caption = "Source: Statistics Norway (SSB), table 06913",
        x = NULL,
        y = "Number of procedures",
        color = NULL
      ) +
      theme_minimal(base_size = 13) +
      theme(
        plot.title = element_text(face = "bold", size = 16),
        plot.subtitle = element_text(color = "grey40", margin = margin(b = 15)),
        panel.grid.minor = element_blank(),
        legend.position = "none",
        axis.text = element_text(size = 11),
        plot.margin = margin(10, 80, 10, 10)
      )
    
    print(p3)
  }
}

Average length of stay collapse

Perhaps the single most dramatic efficiency metric: how long patients actually stay in hospital. This has been steadily declining as medical technology improves and care pathways are optimized.

Code

if (!is.null(df)) {
  
  # Filter for average length of stay
  los_df <- df |>
    filter(grepl("liggetid|length of stay|gjennomsnitt", 
                 metric, ignore.case = TRUE)) |>
    arrange(year)
  
  if (nrow(los_df) > 0) {
    
    # Create ridgeline-style distribution showing decade changes
    los_decade <- los_df |>
      mutate(decade = paste0(floor(year / 10) * 10, "s")) |>
      filter(year >= 2000)
    
    p4 <- ggplot(los_df, aes(x = year, y = value)) +
      geom_area(fill = pal[7], alpha = 0.5) +
      geom_line(color = pal[8], linewidth = 1.5) +
      geom_point(data = los_df |> 
                  filter(year %in% c(min(year), max(year), 
                                    min(year) + round((max(year) - min(year)) / 2))),
                 size = 4, color = pal[8]) +
      geom_text(data = los_df |> filter(year == min(year)),
                aes(label = paste0(round(value, 1), " days\n(", year, ")")),
                vjust = -0.5, hjust = 0, fontface = "bold", size = 4) +
      geom_text(data = los_df |> filter(year == max(year)),
                aes(label = paste0(round(value, 1), " days\n(", year, ")")),
                vjust = -0.5, hjust = 1, fontface = "bold", size = 4) +
      scale_y_continuous(labels = number_format(accuracy = 0.1)) +
      scale_x_continuous(breaks = seq(2000, 2025, 5)) +
      labs(
        title = "The Accelerating Patient: Shorter Hospital Stays Become the Norm",
        subtitle = "Average length of stay in somatic hospitals has dropped dramatically as care becomes more efficient",
        caption = "Source: Statistics Norway (SSB), table 06913",
        x = NULL,
        y = "Average days in hospital"
      ) +
      theme_minimal(base_size = 13) +
      theme(
        plot.title = element_text(face = "bold", size = 16),
        plot.subtitle = element_text(color = "grey40", margin = margin(b = 15)),
        panel.grid.minor = element_blank(),
        axis.text = element_text(size = 11)
      )
    
    print(p4)
  }
}

Key findings

Code

if (!is.null(df)) {
  
  # Calculate key changes
  beds <- df |> 
    filter(grepl("senger|beds", metric, ignore.case = TRUE))
  
  if (nrow(beds) > 0) {
    bed_change <- (beds$value[beds$year == max(beds$year)] / 
                   beds$value[beds$year == min(beds$year)] - 1) * 100
    
    cat(sprintf("\n• Hospital beds have declined by %.1f%% over the period\n", abs(bed_change)))
  }
  
  los <- df |>
    filter(grepl("liggetid|length of stay", metric, ignore.case = TRUE))
  
  if (nrow(los) > 0) {
    los_change <- (los$value[los$year == max(los$year)] / 
                   los$value[los$year == min(los$year)] - 1) * 100
    los_days_saved <- los$value[los$year == min(los$year)] - 
                      los$value[los$year == max(los$year)]
    
    cat(sprintf("• Average hospital stay has shortened by %.1f%% (%.1f fewer days)\n", 
                abs(los_change), los_days_saved))
  }
  
  surgery <- df |>
    filter(grepl("dagkirurgi|day.*surgery", metric, ignore.case = TRUE))
  
  if (nrow(surgery) > 0) {
    surgery_growth <- (surgery$value[surgery$year == max(surgery$year)] / 
                       surgery$value[surgery$year == min(surgery$year)] - 1) * 100
    
    cat(sprintf("• Day surgeries have increased by %.1f%% as outpatient care expands\n", 
                surgery_growth))
  }
  
  cat("\n• Norway's hospital system has fundamentally restructured around efficiency and rapid throughput\n")
  cat("• Fewer beds, shorter stays, and more outpatient procedures reflect both technological progress and cost pressure\n")
}


• Norway's hospital system has fundamentally restructured around efficiency and rapid throughput
• Fewer beds, shorter stays, and more outpatient procedures reflect both technological progress and cost pressure

What this means for Norwegian healthcare

The transformation of Norway’s hospital system over the past 25 years represents one of the most significant structural changes in the welfare state — yet it has happened quietly, without the dramatic headlines that accompany other policy debates. The statistics tell a clear story: Norwegian hospitals are doing more with less, moving patients through faster, and shifting care away from traditional inpatient models toward day surgery and outpatient procedures.

This isn’t necessarily a crisis narrative, but it does raise important questions. As the population ages and chronic conditions become more prevalent, will this efficiency-first model continue to deliver the quality of care Norwegians expect? The bed count decline might reflect smarter medicine and better technology — or it might simply mean that when demand surges, there’s less buffer capacity than before.

The real test will come in the next decade, as demographic pressures intensify. For now, the data shows a healthcare system that has successfully reinvented itself. Whether that reinvention is sustainable remains an open question.

--- title: "Norway's Hospital Crisis: Where the Beds and Surgeries Disappeared" description: "Two decades of somatic hospital data reveal a quiet transformation in Norwegian healthcare delivery" date: "2026-03-16" categories: [SSB, health, hospitals, demographics] --- ## The silent reshaping of Norwegian healthcare While political debates rage over healthcare funding, the actual transformation of Norway's hospital system has been quiet, gradual, and dramatic. Beds have disappeared, procedures have moved outward, and the entire model of somatic care has fundamentally changed since the early 2000s. This isn't a story of collapse — it's a story of efficiency meeting demographic pressure in ways that don't always show up in the headlines. ```{r setup} #| echo: false knitr::opts_chunk$set(echo = TRUE, warning = FALSE, message = FALSE, error = TRUE) ``` ```{r libraries} library(tidyverse) library(PxWebApiData) library(lubridate) library(scales) library(ggridges) library(MetBrewer) library(patchwork) # Color palette from Met Brewer (Hiroshige) pal <- met.brewer("Hiroshige", 8) ``` ## Fetching hospital activity data We're pulling two decades of data from SSB's comprehensive somatic hospital statistics (table 06913), which tracks everything from bed counts to surgical procedures across Norway's regions. ```{r fetch-data} df <- NULL tryCatch({ raw <- ApiData( "https://data.ssb.no/api/v0/no/table/06913", Region = "0", # Whole country ContentsCode = TRUE, # Get all available metrics Tid = list(filter = "top", values = 25) ) tmp <- raw[[1]] print(names(tmp)) # Find time column defensively time_col <- names(tmp)[grepl( "tid|år|kvartal|måned|aar|maaned|year|month|quarter", names(tmp), ignore.case = TRUE, perl = TRUE )][1] if (is.na(time_col)) time_col <- names(tmp)[length(names(tmp)) - 1L] message("Time column detected: ", time_col) # Find the contents/statistics variable column stats_col <- names(tmp)[grepl("statistikk|contents", names(tmp), ignore.case = TRUE)][1] if (is.na(stats_col)) stats_col <- names(tmp)[length(names(tmp)) - 2L] message("Statistics column detected: ", stats_col) df <- tmp |> mutate( value = as.numeric(value), time_str = .data[[time_col]], metric = .data[[stats_col]], year = as.integer(time_str) ) |> filter(!is.na(value), !is.na(year)) message("Successfully fetched ", nrow(df), " rows covering ", n_distinct(df$year), " years and ", n_distinct(df$metric), " metrics") }, error = function(e) { message("Data fetch failed: ", e$message) }) ``` ## The great bed exodus Hospital beds per capita tell the clearest story of how Norwegian healthcare has transformed. Since 2000, the number of available somatic beds has plummeted while the population has grown — a deliberate policy shift toward shorter stays and outpatient care. ```{r beds-plot} #| fig-height: 6 #| fig-width: 10 #| fig-show: asis #| dev: "png" if (!is.null(df)) { # Filter for bed-related metrics beds_df <- df |> filter(grepl("senger|Senger|beds|Beds", metric, ignore.case = TRUE)) |> arrange(year) if (nrow(beds_df) > 0) { # Create area chart with annotations p1 <- ggplot(beds_df, aes(x = year, y = value)) + geom_area(fill = pal[3], alpha = 0.7) + geom_line(color = pal[1], linewidth = 1.2) + geom_point(data = beds_df |> filter(year == max(year) | year == min(year)), color = pal[1], size = 3) + geom_text(data = beds_df |> filter(year == max(year)), aes(label = paste0(comma(value), "\n", year)), vjust = -0.5, hjust = 1, color = pal[1], fontface = "bold", size = 4) + geom_text(data = beds_df |> filter(year == min(year)), aes(label = paste0(comma(value), "\n", year)), vjust = -0.5, hjust = 0, color = pal[1], fontface = "bold", size = 4) + scale_y_continuous(labels = comma, expand = expansion(mult = c(0, 0.15))) + scale_x_continuous(breaks = seq(2000, 2025, 5)) + labs( title = "The Vanishing Hospital Bed", subtitle = "Total number of somatic hospital beds in Norway has declined steadily for two decades", caption = "Source: Statistics Norway (SSB), table 06913", x = NULL, y = "Number of beds" ) + theme_minimal(base_size = 13) + theme( plot.title = element_text(face = "bold", size = 16), plot.subtitle = element_text(color = "grey40", margin = margin(b = 15)), panel.grid.minor = element_blank(), axis.text = element_text(size = 11) ) print(p1) } } ``` ## Patient throughput acceleration While beds have vanished, the number of patient discharges tells a more complex story. Hospitals are handling more cases in less time — the efficiency revolution in action. ```{r discharges-plot} #| fig-height: 6 #| fig-width: 10 #| fig-show: asis #| dev: "png" if (!is.null(df)) { # Filter for discharge metrics discharge_df <- df |> filter(grepl("utskrivn|discharge|pasienter behandlet", metric, ignore.case = TRUE)) |> arrange(year) if (nrow(discharge_df) > 0) { # Calculate year-over-year change discharge_change <- discharge_df |> arrange(year) |> mutate( pct_change = (value / lag(value) - 1) * 100, direction = ifelse(pct_change >= 0, "Increase", "Decrease") ) |> filter(!is.na(pct_change)) # Lollipop chart of recent changes recent_change <- discharge_change |> filter(year >= max(year) - 10) p2 <- ggplot(recent_change, aes(x = year, y = pct_change, color = direction)) + geom_hline(yintercept = 0, linewidth = 0.8, color = "grey30") + geom_segment(aes(xend = year, yend = 0), linewidth = 1.2) + geom_point(size = 4) + scale_color_manual(values = c("Increase" = pal[5], "Decrease" = pal[2])) + scale_x_continuous(breaks = seq(2010, 2025, 2)) + labs( title = "Year-to-Year Patient Volume Volatility", subtitle = "Annual percentage change in hospital discharges shows efficiency gains mixed with capacity constraints", caption = "Source: Statistics Norway (SSB), table 06913", x = NULL, y = "Annual change (%)", color = NULL ) + theme_minimal(base_size = 13) + theme( plot.title = element_text(face = "bold", size = 16), plot.subtitle = element_text(color = "grey40", margin = margin(b = 15)), panel.grid.minor = element_blank(), legend.position = "top", axis.text = element_text(size = 11) ) print(p2) } } ``` ## The day surgery revolution One of the most dramatic shifts in Norwegian healthcare has been the explosion of day surgeries — procedures where patients go home the same day. This metric captures the move toward minimally invasive techniques and outpatient care models. ```{r daysurgery-plot} #| fig-height: 6 #| fig-width: 10 #| fig-show: asis #| dev: "png" if (!is.null(df)) { # Filter for day surgery and inpatient surgery metrics surgery_df <- df |> filter(grepl("dagkirurgi|day.*surgery|døgnkirurgi|inpatient", metric, ignore.case = TRUE)) |> mutate( surgery_type = case_when( grepl("dag|day", metric, ignore.case = TRUE) ~ "Day surgery", grepl("døgn|inpatient", metric, ignore.case = TRUE) ~ "Inpatient surgery", TRUE ~ "Other" ) ) |> filter(surgery_type != "Other") |> arrange(year) if (nrow(surgery_df) > 0 && n_distinct(surgery_df$surgery_type) > 1) { # Create bump chart showing the crossover p3 <- ggplot(surgery_df, aes(x = year, y = value, color = surgery_type)) + geom_line(linewidth = 2.5, alpha = 0.8) + geom_point(data = surgery_df |> filter(year == max(year) | year == min(year)), size = 4) + geom_text(data = surgery_df |> filter(year == max(year)), aes(label = surgery_type), hjust = -0.1, size = 4, fontface = "bold") + scale_color_manual(values = c("Day surgery" = pal[6], "Inpatient surgery" = pal[4])) + scale_y_continuous(labels = comma) + scale_x_continuous(breaks = seq(2000, 2025, 5), limits = c(min(surgery_df$year), max(surgery_df$year) + 2)) + labs( title = "The Great Surgical Shift: Outpatient Care Takes Over", subtitle = "Day surgeries have overtaken traditional inpatient procedures as Norway's hospitals transform care delivery", caption = "Source: Statistics Norway (SSB), table 06913", x = NULL, y = "Number of procedures", color = NULL ) + theme_minimal(base_size = 13) + theme( plot.title = element_text(face = "bold", size = 16), plot.subtitle = element_text(color = "grey40", margin = margin(b = 15)), panel.grid.minor = element_blank(), legend.position = "none", axis.text = element_text(size = 11), plot.margin = margin(10, 80, 10, 10) ) print(p3) } } ``` ## Average length of stay collapse Perhaps the single most dramatic efficiency metric: how long patients actually stay in hospital. This has been steadily declining as medical technology improves and care pathways are optimized. ```{r los-plot} #| fig-height: 6 #| fig-width: 10 #| fig-show: asis #| dev: "png" if (!is.null(df)) { # Filter for average length of stay los_df <- df |> filter(grepl("liggetid|length of stay|gjennomsnitt", metric, ignore.case = TRUE)) |> arrange(year) if (nrow(los_df) > 0) { # Create ridgeline-style distribution showing decade changes los_decade <- los_df |> mutate(decade = paste0(floor(year / 10) * 10, "s")) |> filter(year >= 2000) p4 <- ggplot(los_df, aes(x = year, y = value)) + geom_area(fill = pal[7], alpha = 0.5) + geom_line(color = pal[8], linewidth = 1.5) + geom_point(data = los_df |> filter(year %in% c(min(year), max(year), min(year) + round((max(year) - min(year)) / 2))), size = 4, color = pal[8]) + geom_text(data = los_df |> filter(year == min(year)), aes(label = paste0(round(value, 1), " days\n(", year, ")")), vjust = -0.5, hjust = 0, fontface = "bold", size = 4) + geom_text(data = los_df |> filter(year == max(year)), aes(label = paste0(round(value, 1), " days\n(", year, ")")), vjust = -0.5, hjust = 1, fontface = "bold", size = 4) + scale_y_continuous(labels = number_format(accuracy = 0.1)) + scale_x_continuous(breaks = seq(2000, 2025, 5)) + labs( title = "The Accelerating Patient: Shorter Hospital Stays Become the Norm", subtitle = "Average length of stay in somatic hospitals has dropped dramatically as care becomes more efficient", caption = "Source: Statistics Norway (SSB), table 06913", x = NULL, y = "Average days in hospital" ) + theme_minimal(base_size = 13) + theme( plot.title = element_text(face = "bold", size = 16), plot.subtitle = element_text(color = "grey40", margin = margin(b = 15)), panel.grid.minor = element_blank(), axis.text = element_text(size = 11) ) print(p4) } } ``` ## Key findings ```{r summary-stats} if (!is.null(df)) { # Calculate key changes beds <- df |> filter(grepl("senger|beds", metric, ignore.case = TRUE)) if (nrow(beds) > 0) { bed_change <- (beds$value[beds$year == max(beds$year)] / beds$value[beds$year == min(beds$year)] - 1) * 100 cat(sprintf("\n• Hospital beds have declined by %.1f%% over the period\n", abs(bed_change))) } los <- df |> filter(grepl("liggetid|length of stay", metric, ignore.case = TRUE)) if (nrow(los) > 0) { los_change <- (los$value[los$year == max(los$year)] / los$value[los$year == min(los$year)] - 1) * 100 los_days_saved <- los$value[los$year == min(los$year)] - los$value[los$year == max(los$year)] cat(sprintf("• Average hospital stay has shortened by %.1f%% (%.1f fewer days)\n", abs(los_change), los_days_saved)) } surgery <- df |> filter(grepl("dagkirurgi|day.*surgery", metric, ignore.case = TRUE)) if (nrow(surgery) > 0) { surgery_growth <- (surgery$value[surgery$year == max(surgery$year)] / surgery$value[surgery$year == min(surgery$year)] - 1) * 100 cat(sprintf("• Day surgeries have increased by %.1f%% as outpatient care expands\n", surgery_growth)) } cat("\n• Norway's hospital system has fundamentally restructured around efficiency and rapid throughput\n") cat("• Fewer beds, shorter stays, and more outpatient procedures reflect both technological progress and cost pressure\n") } ``` ## What this means for Norwegian healthcare The transformation of Norway's hospital system over the past 25 years represents one of the most significant structural changes in the welfare state — yet it has happened quietly, without the dramatic headlines that accompany other policy debates. The statistics tell a clear story: Norwegian hospitals are doing more with less, moving patients through faster, and shifting care away from traditional inpatient models toward day surgery and outpatient procedures. This isn't necessarily a crisis narrative, but it does raise important questions. As the population ages and chronic conditions become more prevalent, will this efficiency-first model continue to deliver the quality of care Norwegians expect? The bed count decline might reflect smarter medicine and better technology — or it might simply mean that when demand surges, there's less buffer capacity than before. The real test will come in the next decade, as demographic pressures intensify. For now, the data shows a healthcare system that has successfully reinvented itself. Whether that reinvention is sustainable remains an open question.