Hospital-acquired influenza (HAI) caused by influenza A (subtype H3N2) was associated with an increased risk of hospital dissemination, according to a study in the Journal of Hospital Infection.
Researchers conducted an observational study to identify and quantify HAI and HAI-associated mortality for influenza seasons 2013 to 2014 through 2018 to 2019 in Skåne County in southern Sweden. They also investigated the influence of varying influenza subtypes.
The analysis included hospitalized adults with influenza infection verified with real-time polymerase chain reaction (RT-PCR). Influenza A and B virus was detected with RT-PCR on nasopharyngeal swab samples. Positive findings were confirmed, and influenza A was further subtyped.
The medical records of individuals with suspected HAI were assessed to identify HAI-associated 30-day mortality, in-hospital mortality, and time from diagnosis to death.
H1N1, H3N2, and influenza B most frequently co-circulated during the study period, with the dominating subtype varying depending on the season. A total of 4110 hospitalized patients had a positive influenza test, of whom 950 (23.1%) were infected with H1N1, 1858 (45.2%) with H3N2, and 1277 (31.1%) with influenza B.
Our findings bear relevance for future seasonal influenza preparedness and show that subtyping of influenza A HAI may prove to be useful for defining relevant infection control measures necessary to prevent hospital outbreaks.
Among the 3680 patients with community-acquired influenza (CAI), 50.3% were female, and their median age was 75 years. Of this group, 890 were infected with H1N1, 1577 with H3N2, and 1190 with influenza B (24.2%, 42.9%, and 32.3%, respectively). Patients with HAI were older than those with CAI, with a median age of 78 years.
The criteria for suspected HAI were met by 716 individuals and were further assessed in a review of medical records. After exclusion of 286 cases, 430 cases of confirmed HAI remained. These 430 cases of HAI represented 10.5% of influenza infections in the admitted patients, of which 60 were caused by H1N1 (14%); 281 by H3N2 (65.3%); 87 by influenza B (20.2%); 1 by influenza A that was not subtyped; and 1 by co-infection with H1N1 and H3N2.
The finding that H3N2 accounted for a significantly higher proportion of influenza infections (15.1%), vs H1N1 (6.3%) and influenza B (6.8%) was also was observed in a sensitivity analysis (P <.001). HAI H3N2 represented about two-thirds of total HAI cases, which was 50% greater than what would be expected from the proportion of H3N2 in hospitalized cases of CAI (43%, P <.001).
Among the solitary cases of HAI, 36 (21.4 %) were caused by H1N1, 75 (44.6 %) by H3N2, and 55 (32.7 %) by influenza B. Notably, most cases of HAI H3N2 were part of clusters (73.3%, P <.001), and HAI H1N1 and HAI B were more frequently solitary cases (60% and 63.2%, respectively, P <.001).
Within 30 days from positive influenza virus PCR, 40 patients with HAI died (9.3%), with a median time from PCR diagnosis to death of 10 days (interquartile range, 4-16). The rate of in-hospital mortality was 8.4%. HAI was not predicted by advanced age (≥80 years; odds ratio [OR], 1.80; P =.09), sex (OR, 1.49 for men; P = 0.22), or influenza subtype (H3N2 vs H1N1: OR, 0.83; P =0.70; influenza B vs H1N1: OR, 0.72; P =0.58) in a logistic multivariate model.
The investigators noted that only symptomatic patients had a PCR test for influenza, and some diagnoses may have been missed in patients with mild or asymptomatic infection, leading to an underestimation of HAI. In addition, the researchers did not perform whole genome sequencing and did not assess the presence of comorbidities.
“Our findings bear relevance for future seasonal influenza preparedness and show that subtyping of influenza A HAI may prove to be useful for defining relevant infection control measures necessary to prevent hospital outbreaks,” the study authors concluded.