Season, weather and predictors of healthcare-associated Gram-negative bloodstream infections: a case-only study

https://doi.org/10.1016/j.jhin.2018.06.015Get rights and content

Summary

Background

Recent studies reported seasonality in healthcare-associated infections (HCAI). The association of this phenomenon with other risk factors for HCAI is not clear.

Aim

To analyse the interplay of season, weather and usual predictors of healthcare-associated bloodstream infections caused by Gram-negative bacilli (GNB-BSI).

Methods

A case-only study was conducted in a teaching hospital in Brazil. The study enrolled 446 subjects with GNB-BSI diagnosed from July 2012 to June 2016. Demographic data, comorbidities, invasive procedures and use of antimicrobials were reviewed in medical charts. The season in which GNB-BSI occurred, and weather parameters on the day of diagnosis were recorded. Factors associated with occurrence of GNB-BSI in different seasons (reference category: winter) and caused by different GNB (reference category: Escherichia coli) were analysed. Uni- and multi-variable models of multi-nomial logistic regression were used for analysis.

Findings

GNB-BSI diagnosed in summer was more likely to be caused by Klebsiella spp. [odds ratio (OR) 5.33; 95% confidence interval (CI) 2.04–13.96] or Acinetobacter baumannii (OR 2.69; 95% CI 1.04–6.96), and there was an association between Klebsiella spp. and spring (OR 2.86; 95% CI 1.14–7.18). Average temperature on the day of diagnosis was associated with Klebsiella spp. (OR 1.19; 95% CI 1.07–1.33) and A. baumannii (OR 1.20; 95% CI 1.07–1.34).

Conclusion

Warm seasons and daily temperature impact on the aetiology of GNB-BSI, even in models adjusted for usual risk factors. One possible explanation for these findings is that seasonality of healthcare-associated pathogens is intrinsic to micro-organisms, and not associated with comorbidities, procedures or use of antimicrobials.

Introduction

Seasonality of healthcare-associated infections (HCAI) caused by Gram-negative bacilli (GNB) has been reported recently [1]. It is usually characterized by ‘summer peaks’ in incidence [2], [3], [4], [5] or association with environmental temperature [6], [7]. Coincidentally, a multi-centre study reported increased incidence of GNB bloodstream infections (GNB-BSI) in hospitals located closer to the Equator [8].

Factors underlying this phenomenon have not been elucidated. Increased environmental reservoirs (inside or outside healthcare settings), understaffing associated with summer holidays, and expression of virulence factors in GNB have been suggested as contributing causes [1], [9]. Ecological studies [3], [6], [7], although valuable for identifying seasonality, have not contributed to reinforcing either of these hypotheses.

Seasonality in the incidence of healthcare-associated pathogens could reflect seasonal variations in their epidemiological determinants. For some years, traditional risk factors for healthcare-associated GNB (especially those that are multi-drug-resistant) have been described repeatedly [10]. They include severity of illness, comorbidities, time of exposure to health care (‘time at risk’), invasive procedures and use of antimicrobials [11], [12]. It is therefore worth investigating if those factors vary between seasons, thus contributing to seasonality.

This was the rationale of this study. Briefly, an observational, individual-based study was conducted to identify the interplay of season, weather and usual predictors of healthcare-associated GNB-BSI.

Section snippets

Setting

This study was conducted in a 450-bed teaching hospital in Botucatu, São Paulo State, Brazil. The hospital is a centre for tertiary care for several cities in an area comprising one million inhabitants [7].

Study design

A case-only study was conducted, enrolling subjects with healthcare-associated GNB-BSI diagnosed from July 2012 to June 2016.

Inclusion criteria

All adult patients (aged ≥18 years) admitted to the hospital who acquired healthcare-associated GNB-BSI caused by Escherichia coli, Enterobacter spp., Klebsiella spp.,

Ethical issues

  • This study was approved by the local committee for ethics in research.

Results

This study enrolled 446 subjects with GNB-BSI caused by the following micro-organisms: E. coli, 75; Enterobacter spp., 65; Klebsiella spp., 126; P. aeruginosa, 64; and A. baumannii, 116. Central-line-associated BSI (CLABSI) accounted for 68.8% of cases. Overall, cases were more frequent in summer (27.4%) and spring (28.9%) than in winter (19.5%) and autumn (24.2%) (P = 0.01). Table I presents a description of meteorological data from the days of collection of positive blood cultures, with

Discussion

Seasonality of infectious diseases is a complex phenomenon, the determinants of which are not fully understood [18]. The impact of season and climate on parasite–host systems is relatively easy to understand for vector-borne diseases, but not as clear for human-to-human transmission, or for acquisition of pathogens from endogenous or environmental sources (such as occurs in HCAI) [19]. Even so, understanding the mechanisms and pattern of seasonality may provide clues regarding how and when

Conflict of interest statement

None declared.

Funding source

FACL was a medical student for part of this study, and received a student grant from the National Council of Scientific and Technological Development of Brazil (CNPq).

References (24)

  • F. Schwab et al.

    The warmer the weather, the more Gram-negative bacteria – impact of temperature on clinical isolates in intensive care units

    PLoS One

    (2014)
  • C.M. Fortaleza et al.

    Tropical healthcare epidemiology: weather determinants of the etiology of bloodstream infections in a Brazilian hospital

    Infect Control Hosp Epidemiol

    (2014)
  • Cited by (13)

    • A systematic review of the global seasonality of infections caused by Acinetobacter species in hospitalized patients

      2020, Clinical Microbiology and Infection
      Citation Excerpt :

      To this end, most studies (7/10; 70%) relied on multivariable generalized linear models (Poisson [8,13,19,20], negative binomial [22] or logistic [12,21] regression), while the remaining three studies used bivariate correlation coefficients [7,15,18]. All nine studies that examined ambient temperature found a positive association with Acinetobacter incidence [7,8,12,13,15,19–22]. Of note, one study was conducted exclusively in ICUs [22] and another found that temperature-associated variation in Acinetobacter infections persisted in the subgroup of patients admitted to ICUs [21], which were units with climate control.

    • Seasonal variation in antimicrobial resistance rates of community-acquired Escherichia coli bloodstream isolates

      2019, International Journal of Antimicrobial Agents
      Citation Excerpt :

      Second, AMR is often multifactorial involving various interactions between the host and the environment. Transmission of AMR in hospitals, intensive care units and other healthcare settings includes much more complex dynamics than the community [36,37]. Third, the study examined E. coli bloodstream isolates from one large healthcare system.

    View all citing articles on Scopus
    View full text