Research Letter
November 2, 2023
Changes in Employment in the US Health Care Workforce, 2016-2022
Thuy Nguyen, Christopher Whaley, Kosali I. Simon, et al
JAMA. 2023;330(20):2018-2019. doi:10.1001/jama.2023.18932
The COVID-19 pandemic has caused a variety of disruptions to the health care workforce.1Data on select health care workforce sectors chronicle patterns in employment through 2021.1-3 Lack of national studies on the broad health care workforce using more recent data in this period of rapid change limits the ability of health care organizations and policy makers to ensure that patient needs are met. This study estimated recovery since the initial decrease in employment after the March 2020 public health emergency across US health care subsectors through the end of 2022.
Methods
We used industry- and national-level employment data from the 2016-2022 Quarterly Census of Employment and Wages, a national labor force census database covering more than 95% of jobs in the US,4 to assess health care employment changes vs changes in non–health care sectors. Full- and part-time employees in private and government sectors were included. In a subgroup analysis, we analyzed data of several key health care subsectors (offices of physicians and hospitals). Given long-term care staffing challenges during the COVID-19 pandemic,1,5 we also described employment changes among skilled nursing facilities (SNFs). See eMethods in Supplement 1 for details on data and sector descriptions.
We used ordinary least-squares regression with seasonality adjustments and data in the prepandemic period (quarter 1 [Q1; January-March] 2016 to Q4 [October-December] 2019) to estimate predicted employment levels in the pandemic period (Q1 2020 to Q4 2022), simulating what employment would have been in the absence of the pandemic, assuming prepandemic trends continued. We calculated percentage differences between actual and predicted employment and used mean comparison tests to compare these values. Analyses were conducted with Stata version 17.0; P < .05 defined statistical significance (2-sided). The study was granted not-regulated status by the University of Michigan Medical School institutional review board and abided by the STROBE guidelines for cross sectional studies.
Results
In the prepandemic period, health care employment increased by 0.6% per quarter (from 20 636 222 to 22 456 615), whereas non–health care employment increased by 0.4% per quarter (from 119 492 688 to 127 548 688). During the pandemic period, health care employment increased by 0.1% per quarter to 22 710 834 vs 0.3% to 129 607 080 for non–health care employment (Figure 1).
Health care employment decreased less rapidly than non–health care employment in 2020 (difference in observed vs predicted from Q1 to Q4, −4.3% [95% CI, −9.0% to 0.3%; P = .06] vs −7.2% [95% CI, −14.1% to −0.3%; P = .04]) but recovered less quickly in 2022 (difference in observed vs predicted Q1 to Q4, −5.0% [95% CI, −5.8% to −4.2%; P < .001] vs −2.4% [95% CI, −3.1% to −1.8%; P = .001]). Thus, percentage differences during the entire pandemic were comparable in these sectors (health care employment, −4.9% [95% CI, −6.0% to −3.9%; P < .001]; non–health care employment, −5.0% [95% CI, −7.1% to −3.0%; P < .001]).
Employment in hospitals increased by 0.4% per quarter to 6 620 127 prepandemic and by 0.03% per quarter to 6 619 070 during the pandemic, whereas employment in physicians’ offices increased by 0.6% per quarter to 2 783 994 and 0.5% per quarter to 2 925 099, respectively (Figure 2). Employment in SNFs declined prepandemic by 0.2% per quarter, from 1 702 025 to 1 649 920, and by 1.1% to 1 414 326 during the pandemic. The percentage differences in observed vs predicted levels were −3.3% (95% CI, −4.1% to −2.5%; P < .001) for hospitals, −1.6% (95% CI, −2.7% to −0.6%; P = .007) for physicians’ offices, and −10.5% (95% CI, −13.0% to −8.1%; P < .001) for SNFs.
Discussion
Health care employment growth declined after the onset of the COVID-19 pandemic and recovery patterns varied by health care subsectors. Measurement of subsector recovery is essential for informing future workforce policy, particularly for SNFs, given increasing efforts to promote adequate employment. Staffing in SNFs had already declined before the pandemic and further declines after the pandemic are concerning. The differential employment trends across health care subsectors may be driven by worker concerns of infectious disease threats,5 modest wage levels, and high turnover rates among many long-term care occupations.6
Limitations include the inability to examine mechanisms for diverging recovery patterns and lack of granular data on occupation compositions and patient capacity within each subsector. Future research seeking to understand causes and consequences of diverging recovery patterns is needed, especially among SNFs.
