Discontinuation of Universal Admission Testing for SARS-CoV-2 and Hospital-Onset COVID-19 Infections in England and Scotland | Global Health | JAMA Internal Medicine

Some hospitals test all patients for SARS-CoV-2 infection on admission to prevent transmission from patients with unrecognized infection to staff and other patients, since two-thirds of SARS-CoV-2 transmissions are from people with asymptomatic or presymptomatic infections, transmission risk between patients sharing rooms is high, and universal medical masks can reduce but not eliminate transmission.¹ However, the utility of universal admission testing has been questioned due to resource constraints, care delays, and sparse data demonstrating it reduces nosocomial infections.² England and Scotland stopped requiring hospitals to test all admitted patients starting August 31, 2022, and September 28, 2022, respectively. We examined whether this discontinuation was associated with increases in hospital-onset SARS-CoV-2 infections.

This time series analysis used public data sets from Public Health Scotland (Hospital Onset COVID-19 Cases in Scotland) and National Health Service England (COVID-19 Hospital Activity) to obtain weekly counts of hospital-onset cases of SARS-CoV-2 infections, defined as newly positive SARS-CoV-2 tests more than 7 days after admission between July 1, 2021, and December 16, 2022. Polymerase chain reaction testing was recommended for admission testing during most of the study period. Hospital-onset SARS-CoV-2 infection rates closely correlate with community prevalence,³ so we calculated weekly rates of new hospital-onset cases per 1000 community infections to assess the relative increases in hospital-onset cases. We used the UK Office for National Statistics (ONS) COVID-19 Infection Survey to estimate community prevalence. The ONS conducts near-weekly testing of randomly selected households, avoiding the potential biases associated with health care–related or patient-initiated testing.⁴ We evaluated level changes in this weekly rate after admission testing ended using interrupted time series with a first-order autoregressive structure accounting for temporal correlation; we used R, version 4.2.1, package nlme (R Foundation for Statistical Computing).

The study was divided into 3 periods: Delta dominance with admission testing (July 1, 2021, to December 13, 2021), Omicron dominance with admission testing (December 14, 2021, when the Omicron variant exceeded 50% of sequenced SARS-CoV-2 infections in both countries, to August 30, 2022 [England], and September 27, 2022 [Scotland]), and Omicron dominance without admission testing ([Scotland] September 28, 2022, to December 16, 2022; [England] August 31, 2022, to December 16, 2022). Sensitivity analysis was performed using community-acquired SARS-CoV-2–related hospitalizations as the denominator. A 2-tailed P < .05 indicated statistical significance. Neither institutional review board approval nor informed consent was required because the study used public, deidentified data. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

During the study period, there were 46 517 COVID-19–related admissions (34 183 community-onset, 12 334 hospital-onset cases) in Scotland, and 518 379 COVID-19–related admissions (398 264 community-onset, 120 115 hospital-onset cases) in England. The mean (SD) weekly rate of new hospital-onset SARS-CoV-2 infections per 1000 estimated community infections in Scotland increased from 0.78 (0.37) during the Delta dominance period, to 0.99 (0.21) during Omicron dominance, to 1.64 (0.37) after universal admission testing ended (Figure, A). The immediate level change was statistically significant after admission testing ended (41% relative increase; 95% CI, 6%-76%) but not after the Delta-to-Omicron transition (Table). Likewise, during analogous periods in England, the same mean (SD) rate increased from 0.64 (0.14) to 1.00 (0.17) to 1.39 (0.34) (Figure, B). The immediate level change was significant after admission testing ended (26% relative increase; 95% CI, 8%-45%) but not after the Delta-to-Omicron transition. Findings were similar with incident community-acquired SARS-CoV-2 hospitalizations as the denominator (Figure, C and D; Table).

Stopping universal admission testing in the national health systems of 2 countries (England and Scotland) was associated with significant increases in hospital-onset SARS-CoV-2 infections relative to community-onset infections. Potential mechanisms include more unrecognized present-on-admission infections causing transmissions to other patients and health care workers, who in turn infected other patients.

Limitations of our analysis include its before-and-after design without concurrent controls and possibly spurious associations if policy compliance was minimal or delayed; however, the associations in 2 countries that stopped admission testing at different times were similar. Some community-onset cases may have been misclassified as hospital-onset cases, but this is unlikely considering the Omicron variant’s 3-day incubation period and sequencing data⁵ that suggest that a 7-day cutoff underestimates rather than overestimates nosocomial infections.

Nosocomial SARS-CoV-2 Omicron infections remain common, with crude mortality estimates ranging from 3% to 13%.^3,6 Hospitals should exercise caution before stopping universal admission testing for SARS-CoV-2 infections.

Accepted for Publication: February 28, 2023.

Published Online: June 5, 2023. doi:10.1001/jamainternmed.2023.1261

Corresponding Author: Theodore R. Pak, MD, PhD, Massachusetts General Hospital, 55 Fruit St, Cox Building, 5th Floor, Boston, MA 02114 ([email protected]).

Author Contributions: Dr Pak had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Pak, Rhee, Klompas.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Pak, Klompas.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Pak, Wang.

Obtained funding: Pak.

Supervision: Rhee, Klompas.

Conflict of Interest Disclosures: Dr Pak reported receiving grants from the National Institute of Allergy and Infectious Diseases (NIAID) during the conduct of the study. Dr Wang reported receiving funding from the National Institutes of Health, the Agency for Healthcare Research and Quality, the Centers for Disease Control and Prevention, and Patient-Centered Outcomes Research Institute outside the submitted work. Dr Klompas and Dr Rhee reported receiving grants from the Centers for Disease Control and Prevention and the Agency for Healthcare Research and Quality and personal fees from UpToDate outside the submitted work. No other disclosures were reported.

Funding/Support: This project was supported by grant number T32AI007061 from the NIAID (Dr Pak).

Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIAID.

Data Sharing Statement: See the Supplement.