Report: 1 in 5 nursing homes face serious lack of PPE, staff shortages

A comprehensive review of data submitted by more than 15,000 US nursing homes revealed that about 20% faced a staff shortage or severe shortage of personal protective equipment (PPE) in early July, months into the COVID-19 pandemic, according to a study in Health Affairs yesterday.

US researchers analyzed staff and PPE shortage data submitted by 15,035 nursing homes (98% of the total) into the Medicare and Medicaid Services (CMS) COVID-19 Nursing Home Database. They looked at data submitted from May 18 to Jun 14 and compared it to information submitted from Jun 24 to Jul 19. PPE included N95 respirators, masks, gowns, and gloves.

At the end of the first study period, 20.7% of nursing homes reported a severe PPE shortage with 1 week or less of available supply, with shortages of N95s and gowns being the most common (13.4% and 12.6% of all nursing homes, respectively). Similarly, 20.8% of facilities reported a staff shortage, with 15.1%, 17.2%, and 9.2% indicating a shortage of nurses, nursing aides, and other staff, respectively.

The rate of both staff and PPE shortages did not meaningfully improve from May to July, the authors noted. The facilities most likely to report shortages were those with COVID-19 cases among residents and staff, those serving a high proportion of Medicaid recipients, and those scoring lower on CMS's five-star quality rating system.

The authors of the study write, "As in most crises, the most vulnerable nursing homes are at the highest risk for shortages that put the health of their residents and staff at risk…. Unless these shortages are prioritized by policymakers, long-term care residents will continue to be at a great disadvantage in the pandemic."

"Twenty percent is a lot, given where we are in the course of this pandemic," said lead author Brian E. McGarry, PhD, of the University of Rochester Medical Center (URMC), in a URMC news release. "While there has been some shifting in which nursing homes have been reporting these problems, from a national level, we're still not on the right trajectory."
Aug 20 Health Aff study
Aug 20 URMC news release

COVID-19 hospital bed occupancy tied to death rates in US states

Using hospital data from 23 states, researchers at the University of Minnesota (UMN) and University of Washington found a correlation between the number of hospital beds occupied by COVID-19 patients in a state and reported death rates in that state, according to a study yesterday in the Journal of General Internal Medicine.

The investigators used the UMN's COVID-19 Hospitalization Tracking Project to examine data from 23 states that reported daily percentages of intensive care unit (ICU) and non-ICU bed use by COVID patients.

They found that COVID-19 patients occupied nearly 20% of all ICU bed use in all states examined and about 5% of non-ICU beds. A 1% increase in ICU bed use (17 beds on average) was associated with 2.8 more COVID-19 deaths over the next 7 days, and a 1% increase in non-ICU bed use (130 beds on average) was associated with 17.8 more COVID-19 deaths.

"These estimates provide a better understanding of the projections of the COVID-19 pandemic in the U.S., especially when states are monitoring economic activities, and provide important practice insights for hospitals in terms of assessment of hospital bed and ICU bed capacity and preparedness," said study lead Pinar Karaca-Mandic, PhD, with the UMN's Carlson School of Management, in a UMN news release.

Co-author Soumya Sen, PhD, also with the Carlson School, said the findings will be helpful with mathematical modeling estimates in determining how not to overwhelm hospitals during the pandemic.
Aug 20 J Gen Intern Med study
Aug 20 UMN news release

Our weekly wrap-up of antimicrobial stewardship & antimicrobial resistance scans

Study finds antibiotic time-outs not tied to lower overall antibiotic use

Originally published by CIDRAP News Aug 20

Implementation of a pharmacist-led antibiotic time-out (ATO) at an academic medical center was feasible and well-accepted, but did not change overall antibiotic use, researchers from the University of Nebraska Medical Center reported today in Infection Control & Hospital Epidemiology.

In the two-phased cluster-randomized study, three academic inpatient medical teams were randomly selected in the first phase (ATO-A) to implement the pharmacist-led time-out, in which initial antibiotic therapy in a patient is reassessed, and three teams maintained usual care (UC-A). In phase B, the usual-care teams implemented the ATO process (UC ATO-B), while ATO use continued in the other group (ATO-B).

The study, conducted from November 2014 through February 2015, included all patients who were treated by the teams and received antibiotics. The researchers targeted two ATO points: early (less than 3 days after antibiotics were initiated) and late (from 3 to 5 days after antibiotics were started).

In total, 290 ATOs were documented (181 early, 87 late, and 22 subsequent) among 538 admissions. The most common ATO recommendations were narrow therapy (148 of 290), no change (124 of 290), and change to oral (30 of 290).

Measured in days of therapy (DOT) per 1,000 patient-days (PD), overall antibiotic use was not different between any of the groups (P = .51), although intravenous (IV) levofloxacin use decreased in the UC group after ATO implementation (49 DOT/1,000 PD vs 20 DOT/1,000 PD; P = .022). The ratio of oral to IV DOT was lower in the UC group than in any of the ATO groups (P = .032). The researchers detected no differences in mortality, length of stay, readmission, Clostridioides difficile infection, or antibiotic adverse events.

While email surveys indicated the ATO process was well-accepted, the authors of the study say the findings mirror those of other studies that have found that ATOs have no impact on overall antibiotic use. But they also said that not that all studies of ATOs have been conducted in hospital with active antibiotic stewardship programs (ASPs).

"An ATO may be beneficial when an active ASP is absent or the program lacks [infectious disease] expertise, although ATOs have not been studied in these situations," they wrote.
Aug 20 Infect Control Hosp Epidemiol abstract

Use of common antibiotics linked to resistant bacteria colonization

Originally published by CIDRAP News Aug 20

A nationwide case-control study in Denmark found that the risk of colonization with multidrug-resistant bacteria (MRB) was associated with consumption of commonly used antibiotics for at least 2 years after treatment, Danish researchers reported today in the Journal of Antimicrobial Chemotherapy.

Using data from the AB-RED (Antibiotic Resistance in Emergency Departments in Denmark) study and from a national register of antibiotic consumption, researchers from the University of Southern Denmark and Odense University Hospital compared 256 patients colonized with MRB—including extended-spectrum beta-lactamase–producing Enterobacteriaceae (ESBL-E), carbapenemase-producing Enterobacteriaceae, methicillin-resistant Staphylococcus aureus, and vancomycin-resistant Enterococcus—based on throat, nose, and rectum swabs and 4,763 control patients. They performed multivariate analyses to examine the association between antibiotic consumption and MRB colonization, with a subgroup analysis of ESBL-E-colonized patients.

In the 2 years prior to study inclusion, 77% of patients colonized with MRB and 68% of control patients had at least one antibiotic prescription (P = 0.002). The multivariate analysis showed a significant increase in risk of colonization with ESBL-E if penicillins (odds ratio [OR], 1.58 to 1.65) or fluoroquinolones (OR, 2.25 to 6.15) were prescribed. Analysis of all MRB-colonized patients showed similar results. An assessment of the timeline showed a significant increase in risk of colonization up to 2 years after exposure to penicillins, fluoroquinolones, and macrolides, with no association observed for other antibiotics.

"The findings of our study draw a clear link between more commonly used antibiotics such as macrolides and penicillins, especially amoxicillin, and an increased risk of MRB colonization, particularly ESBL-E colonization," the authors of the study wrote. "This indicates that more restricted antibiotic stewardship might be needed to reduce the prevalence of ESBL-E, and maybe all MRB, in the future."
Aug 20 J Antimicrob Chemother abstract

Clinical support tool linked to better antibiotic prescribing for cystitis

Originally published by CIDRAP News Aug 19

The inclusion of a clinical decision support system (CDSS) and order set within the electronic medical record, in combination with the use of local urine antibiograms, was associated with improved antibiotic prescribing for acute cystitis at a veterans' health system in North Carolina, researchers reported today in Infection Control & Hospital Epidemiology.

In the quasi-experimental, interrupted time-series analysis, researchers from Duke University School of Medicine and the Durham Veterans' Affairs Health Care System analyzed treatment of outpatient urinary tract infections from April 2016 through October 2019.

The study period consisted of the pre-intervention phase, an intervention period when the CDSS highlighting nitrofurantoin and beta-lactams as preferred agents over fluoroquinolones for uncomplicated acute cystitis was integrated within the electronic medical record, and a post-intervention phase. The primary outcomes measured were changes in monthly proportions of antibiotic classes prescribed for cystitis.

Prior to the intervention, monthly fluoroquinolone prescriptions accounted for 45% of all outpatient prescriptions of cystitis. After the intervention, fluoroquinolone prescriptions accounted for a median of 32% of antibiotics prescribed monthly for cystitis. Conversely, prescriptions for beta-lactams increased from a monthly median of 14% in the pre-intervention phase to 24.5%. Nitrofurantoin and trimethoprim/sulfamethoxazole prescribing was unaltered by the intervention.

In the interrupted time series analysis, CDSS implementation resulted in a −20.7% level change (95% CI, −33.8% to −7.5%; P = .002) and −1.4% change in slope (95% CI, −3.0% to 0.2%; P = .09) in fluoroquinolone prescribing for cystitis, and a 28.5% level change (95% CI, 15.5% to 41.7%; P < .001) and 1.2% change in slope (95% CI, −0.3% to 2.8%; P = .13) in beta-lactam prescriptions.

"In summary, CDSS combined with local urine antibiograms, even without prescriber education or audit and feedback, can be an effective tool for antimicrobial stewardship," the authors concluded.
Aug 19 Infect Control Hosp Epidemiol abstract

Information card boosts patient knowledge of antibiotics, UK study finds

Originally published by CIDRAP News Aug 17

Introduction of an "antibiotic information card" (AIC) for patients being discharged from the acute medical unit (ACU) of an English hospital significantly increased patient knowledge about antibiotic prescriptions but did not affect the readmission rate, UK researchers reported late last week in the American Journal of Infection Control.

The card was introduced in the ACU at William Harvey Hospital from November 2019 to January 2020 as part of an effort to improve patient compliance with antibiotic regimens. The ultimate aim was to reduce the 30-day readmission and reattendance rate due to the same infection by 75% in patients discharged from the AMU with antibiotics, and to increase patient understanding regarding their antibiotic prescription and treatment to 75% or higher, as assessed by a questionnaire completed 30 days after discharge.

Recent research has shown that 30-day emergency readmissions in English hospitals increased by 1.3% from 2013-14 to 2017-18, with infection-related conditions accounting for 59% of readmissions.

The researchers designed the AIC after collecting baseline data from AMU patients discharged with antibiotics, then provided it to 22 of 23 eligible patients in the first cycle of the intervention and conducted follow-up telephone surveys with those patients. There was no significant improvement in 30-day reattendance and readmission rates, but patient knowledge of their antibiotics increased from the baseline range of 14% to 71% to over 75%, with a range of 86% to 100%. In the second cycle of the intervention, which did not include on-ward support from medical students, only 1 of 23 eligible patients received the card, and patient knowledge of antibiotics ranged from 43% to 100%.

The authors of the study say the AIC is an easily sustainable intervention with minimal financial costs that could be replicated on other wards. But they acknowledge that more staff engagement will be needed to make sure every patient receives the card on discharge.

"Change is more likely to be successful and sustainable as a team effort, and so staff should be encouraged to continue to complete the AICs prior to a patient's discharge," the authors wrote.
Aug 15 Am J Infect Control abstract