Back to Featured Releases

Effects of COVID-19 on Hospital Utilization Trends

With the onset of COVID-19, hospitals statewide saw a sharp drop in inpatient discharges, emergency department utilization, and ambulatory surgeries. Emergency department utilization has been affected more dramatically, with much sharper declines and slower rebound to pre-pandemic utilization trends.

After January 2020, hospitals throughout California began to see utilization drop sharply across all settings. The downward trend continued until April 2020, after which utilization began to rise. This drop is reflected when focusing on patients with common health conditions, such as diabetes, hypertension, and asthma. However, the utilization for people who experience homelessness, cardiac arrest, respiratory arrest/failure, and stroke did not show similar downward trends in utilization compared to other groups.

The release of these visualizations with updated data now includes filters to view trends in utilization at the hospital system and individual facility level.

HCAI’s visualizations are derived from patient administrative data from hospital billing systems and is not disease surveillance data. At the time of publishing, hospital data for 2021 is the most current and validated data available to HCAI. For the latest COVID-19 data from the state’s disease surveillance systems, please visit https://covid19.ca.gov.

Key Findings

  • Individuals with asthma, cancer, chronic obstructive pulmonary disease (COPD), diabetes, hypertension, obesity, sepsis, mental and behavioral disorders, and stroke all reflected a sharp decline in utilization across settings after January 2020 through April 2020. Utilization of the emergency department remained low in late 2020 and throughout 2021 relative to previous years. Inpatient discharges and ambulatory surgeries rebounded nearly to levels typically seen pre-pandemic.
  • The trendlines for patients who experience homelessness do not reflect the same change, aside from a decline in ambulatory surgeries in early 2020. Trendlines for homeless patients may not reflect actuality due to variations in reporting over time. The homeless population is undercounted in 2018 due to the inability to designate homeless persons by ZIP code that year.
  • In the emergency department, in-hospital mortality peaked sharply in December 2020 among nearly all groups and remained relatively low throughout the remainder of 2021. Patients experiencing homelessness were the only group to show an even higher mortality peak in December 2021 when compared to December 2020.
  • Among inpatient discharges, in-hospital mortality peaked sharply in January 2021, with deaths remaining relatively low throughout the remainder of 2021.
  • Patients experiencing obesity and pneumonia demonstrated a greater likelihood of suffering COVID-19 as a primary diagnosis when looking at in-hospital deaths over all settings. In the emergency department, the same was true of patients experiencing pneumonia, cardiac arrest, and respiratory arrest/failure.

The visualization below focuses on utilization trends in inpatient discharges, emergency department treat and release utilization, and ambulatory surgeries beginning in 2018. The sharp downward trend in all three settings begins after January 2020 with the onset of COVID-19 and hits a low point in April 2020 before beginning to rise again. Utilization of the emergency department remained low in late 2020 and throughout 2021 relative to previous years. Inpatient discharges and ambulatory surgeries rebounded nearly to levels typically seen pre-pandemic beginning in March 2021. Utilization from 201-2021 within specific healthcare systems and facilities can be viewed using the dropdown filters.

Note: This product uses suppression of small numbers. Any count of diagnoses below 11 was changed to 11 in the underlying dataset for de-identification purposes.

This visualization displays utilization trends in the same settings as above but focuses on trends in key health-related topics. A sharp decline in utilization is seen starting after January 2020 among nearly all patients. Trends in hospital utilization in patients experiencing homelessness, cardiac arrest, respiratory arrest/failure, sepsis, and stroke were much less affected.

By default, this visualization displays a handful of selected health-related topics and hospital utilization trends among inpatient discharges in 2018-2021. The dropdown filters can be used to change the setting, healthcare system, healthcare facility name, health-related topics, and the time span of focus. All trends reflect encounters and not unique patients.

Note: The number of encounters is all recorded health care encounters for a specific health-related topic. Individual patients may be counted in more than one category if they received multiple diagnoses during a single encounter (e.g., a person who had diabetes and pneumonia would be counted in both diabetes and pneumonia categories). This product uses suppression of small numbers. Any count of diagnoses below 11 was changed to 11 in the underlying dataset for de-identification purposes. The COVID-19 trend lines begin with April 2020 when the COVID-19 specific diagnosis code U07.1 was added to the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM).

This visualization focuses on mortality trends in inpatient discharges, emergency department treat and release utilization, and ambulatory surgeries beginning in 2018. In the emergency department, in-hospital mortality peaked sharply in December 2020 among nearly all groups and remained relatively low throughout the remainder of 2021. Patients experiencing homelessness were the only group to show an even higher mortality peak in December 2021 when compared to December 2020. Among inpatient discharges, in-hospital mortality peaked sharply in January 2021, with deaths remaining relatively low throughout the remainder of 2021.

By default, this visualization displays a handful of selected health-related topics and in-hospital deaths among inpatient discharges in 2018-2021. The dropdown filters can be used to change the setting, select as many health-related topics as desired, and change the time span of focus.

Note: The number of deaths is all recorded in-hospital deaths for a specific health-related topic. Individual patients may be counted in more than one category if they received multiple diagnoses during the encounter in which the death occurred (e.g., a person who had diabetes and pneumonia would be counted in both diabetes and pneumonia categories). The COVID-19 trend lines begin with April 2020 when the COVID-19 specific code U07.1 was added to ICD-10-CM.

This visualization focuses on mortality trends and diagnosis breakdown in inpatient discharges, emergency department treat and release utilization, and ambulatory surgeries in 2020 and 2021. Patients experiencing sepsis and cardiac arrest demonstrated a greater likelihood of COVID-19 as a primary diagnosis. COVID-19 was more often present as a primary diagnosis in deaths among inpatient discharges than in emergency department deaths.

The diagnosis legend is meant to demonstrate a breakdown of how each diagnosis was assigned. For example, among in-hospital deaths including a cancer diagnosis, the breakdown includes a count of cancer as primary diagnosis, cancer as a secondary diagnosis with COVID-19 as the primary diagnosis, and cancer as a secondary diagnosis with any primary diagnosis except COVID-19.

Note: The number of deaths includes all recorded in-hospital deaths for a specific health-related topic. Individual patients may be counted in more than one category if they received multiple diagnoses during the encounter in which the death occurred (e.g., a person who had diabetes and pneumonia would be counted in both diabetes and pneumonia categories). This visualization only includes data for 2020 and 2021 due to the lack of diagnosis data specific to COVID-19 in previous years. Codes specific to COVID-19 were not included in the ICD-10-CM until April 2020. Treat and release emergency department encounters include patients who were seen and treated in an emergency department setting, and then released without admission to an inpatient setting.

This visualization focuses on mortality trends and secondary diagnosis breakdown in inpatient discharges, emergency department treat and release utilization, and ambulatory surgeries in 2020 and 2021. Among all in-hospital deaths, patients with obesity and pneumonia demonstrated greater likelihood of having COVID-19 as a primary diagnosis compared to other patients. When looking at deaths occurring in the emergency department, patients experiencing cardiac arrest and pneumonia demonstrated a higher likelihood of COVID-19 as a primary diagnosis.

Note: The number of deaths includes all recorded in-hospital deaths for a specific health-related topic. Individual patients may be counted in more than one category if they received multiple diagnoses during the encounter in which the death occurred (e.g., a person who had diabetes and pneumonia would be counted in both diabetes and pneumonia categories). This visualization only includes data for 2020 and 2021 due to the lack of diagnosis data specific to COVID-19 in previous years. Codes specific to COVID-19 were not included in the ICD-10-CM until April 2020.

How HCAI Created This Product

  • HCAI identified encounters and in-hospital deaths within inpatient discharges, emergency department treat and release utilization, and ambulatory surgeries from 2018-2021. HCAI’s visualizations are derived from patient administrative data from hospital billing systems and is not disease surveillance data. At the time of publishing, hospital data for 2021 is the most current and validated data available to HCAI. For the latest COVID-19 data from the state’s disease surveillance systems, please visit https://covid19.ca.gov.
  • The COVID-19 trend lines begin with April 2020 when the COVID-19 specific diagnosis code U07.1 was added to the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM).
  • Trendlines for homeless patients may not reflect actuality due to variations in reporting over time. The homeless population is undercounted in 2018 due to the inability to designate homeless persons by ZIP code that year.
  • This product uses suppression of small numbers in the first two visualizations. Any count below 11 was changed to 11 in the underlying dataset for de-identification purposes.
  • In visualizations related to in-hospital mortality, the number of deaths equals all recorded in-hospital deaths for a specific health-related topic. Individual patients may be counted in more than one category if they received multiple diagnoses during the encounter in which the death occurred (e.g., a person who had diabetes and pneumonia would be counted in both diabetes and pneumonia categories).
View Underlying Data

Additional Information

Topic: Healthcare Utilization
Source Link: Healthcare Utilization – Patient-Level Administrative Data
Citation: HCAI – Patient-Level Administrative Data – Effects of COVID-19 on Hospital Utilization Trends, 2018-2021
Temporal Coverage: 2018-2021
Spatial/Geographic Coverage: Statewide
Frequency: Monthly
Additional Resources: COVID-19 Data Reporting Guidance