Infection preventionist staffing levels and rates of 10 types of healthcare-associated infections: A 9-year ambidirectional observation

Department

Quality and Patient Safety

Document Type

Article

Publication Title

Infection Control and Hospital Epidemiology

Abstract

OBJECTIVE: To quantitatively evaluate relationships between infection preventionists (IPs) staffing levels, nursing hours, and rates of 10 types of healthcare-associated infections (HAIs).

DESIGN AND SETTING: An ambidirectional observation in a 528-bed teaching hospital.

PATIENTS: All inpatients from July 1, 2012, to February 1, 2021.

METHODS: Standardized US National Health Safety Network (NHSN) definitions were used for HAIs. Staffing levels were measured in full-time equivalents (FTE) for IPs and total monthly hours worked for nurses. A time-trend analysis using control charts, t tests, Poisson tests, and regression analysis was performed using Minitab and R computing programs on rates and standardized infection ratios (SIRs) of 10 types of HAIs. An additional analysis was performed on 3 stratifications: critically low (2-3 FTE), below recommended IP levels (4-6 FTE), and at recommended IP levels (7-8 FTE).

RESULTS: The observation covered 1.6 million patient days of surveillance. IP staffing levels fluctuated from ≤ 2 IP FTE (critically low) to 7-8 IP FTE (recommended levels). Periods of highest catheter-associated urinary tract infection SIRs, hospital-onset Clostridioides difficile and carbapenem-resistant Enterobacteriaceae infection rates, along with 4 of 5 types of surgical site SIRs coincided with the periods of lowest IP staffing levels and the absence of certified IPs and a healthcare epidemiologist. Central-line-associated bloodstream infections increased amid lower nursing levels despite the increased presence of an IP and a hospital epidemiologist.

CONCLUSIONS: Of 10 HAIs, 8 had highest incidences during periods of lowest IP staffing and experience. Some HAI rates varied inversely with levels of IP staffing and experience and others appeared to be more influenced by nursing levels or other confounders.

First Page

1641

Last Page

1646

DOI

10.1017/ice.2021.507

Volume

43

Issue

11

Publication Date

11-2022

Comments

Record updated with published article citation 2022-12-01 LB.

Published online ahead of print 2022-01-17.

PubMed ID

35034676

Link to Full Text

Share

COinS