Advertisement

The burden of healthcare-associated infection in Australian hospitals: A systematic review of the literature

  • Brett G. Mitchell
    Correspondence
    Corresponding author. Avondale College, Clinical Education Centre, 185 Fox Valley Road, Wahroonga, NSW, 2076, Australia.
    Affiliations
    Avondale College of Higher Education, Faculty of Arts, Nursing and Theology, Australia

    Griffith University, Menzies Health Institute Queensland, Queensland, Australia
    Search for articles by this author
  • Ramon Z. Shaban
    Affiliations
    Griffith University, Menzies Health Institute Queensland, Queensland, Australia

    Gold Coast University Hospital, Gold Coast Health Queensland, Australia
    Search for articles by this author
  • Deborough MacBeth
    Affiliations
    Griffith University, Menzies Health Institute Queensland, Queensland, Australia

    Gold Coast University Hospital, Gold Coast Health Queensland, Australia
    Search for articles by this author
  • Claudia-Jayne Wood
    Affiliations
    Avondale College of Higher Education, Faculty of Arts, Nursing and Theology, Australia
    Search for articles by this author
  • Philip L. Russo
    Affiliations
    Avondale College of Higher Education, Faculty of Arts, Nursing and Theology, Australia

    Deakin University, School of Nursing and Midwifery Victoria, Australia

    Griffith University, Menzies Health Institute Queensland, Queensland, Australia
    Search for articles by this author
Open AccessPublished:August 15, 2017DOI:https://doi.org/10.1016/j.idh.2017.07.001

      Highlights

      • This systematic review reports on the incidence burden of HAIs in Australian hospitals.
      • There are approximately 165,000 HAIs occurring each year in Australia.
      • There is a need for improved surveillance and reporting of HAIs in Australia.

      Abstract

      Introduction

      Central to all efforts to control and prevent healthcare associated infections (HAIs) is the inherent need to measure the burden of infection and disease, classically referred to as surveillance. Australia does not have a national HAI surveillance system making it very difficult to systematically assess and report on the burden of hospital-acquired HAIs. This systematic review reports the incidence burden of HAIs in Australian hospitals as reported in the peer-reviewed literature from 2010 to 2016.

      Methods

      Systematic review of the peer-reviewed literature reporting the incidence of HAIs in Australian hospitals between from 2010 to 2016 was identified using MEDLINE and CINAHL databases. The study protocol is registered with PROSPERO (registration number: CRD42016052997).

      Results

      Of the 844 articles identified in the search, 24 articles were included in this review. Overall, these data suggest 83,096 HAIs per year in Australia, comprising 71,186 urinary tract infections, 4902 Clostridium difficile infections, 3946 surgical site infections, 1962 respiratory infections in acute stroke patients and 1100 hospital-onset Staphylococcus aureus bacteraemia. This is very large underestimate given the lack of or incomplete data on common infections such as pneumonia, gastroenterological and bloodstream infection, thus potentially missing up to 50%–60% of infections. If that is the case, the incidence of HAIs in Australia may be closer to 165,000 per year.

      Conclusion

      There is a dearth of peer-reviewed literature reporting the incidence of HAIs in Australian hospitals, making it very difficult to an accurate burden of infection. On the eve of a global ‘post antibiotic era’, the need for national consensus on definitions, surveillance methodology and reporting is paramount.

      Keywords

      Introduction

      Healthcare-associated infections (HAIs) are a major patient safety issue in hospitals. While research into infection prevention and control has led to improvements in our understanding of effective HAI prevention strategies [
      • Yokoe D.S.
      • Anderson D.J.
      • Berenholtz S.M.
      • Calfee D.P.
      • Dubberke E.R.
      • Ellingson K.D.
      • et al.
      A compendium of strategies to prevent healthcare-associated infections in Acute Care Hospitals: 2014 updates.
      ], HAIs continue to occur and lead to morbidity, mortality and excess healthcare expenditure [
      • Yokoe D.S.
      • Anderson D.J.
      • Berenholtz S.M.
      • Calfee D.P.
      • Dubberke E.R.
      • Ellingson K.D.
      • et al.
      A compendium of strategies to prevent healthcare-associated infections in Acute Care Hospitals: 2014 updates.
      ,
      • Magill S.S.
      • Edwards J.R.
      • Bamberg W.
      • Beldavs Z.G.
      • Dumyati G.
      • Kainer M.A.
      • et al.
      Multistate point-prevalence survey of health care–associated infections.
      ]. Unlike most developed countries, Australia does not have a national system to monitor these infections and cannot provide an estimate of the burden of hospital-acquired HAIs. The last national point prevalence study conducted in Australia occurred in 1984 [
      • McLaws M.
      • Gold J.
      • King K.
      • Irwig L.
      • Berry G.
      The prevalence of nosocomial and community-acquired infections in Australian hospitals.
      ]. Healthcare delivery, technology and infection prevention and control initiatives have advanced considerably since this time.
      In 2008, Cruickshank and Ferguson [
      ] estimated that there are about 200,000 HAIs each year in Australia, which if correct makes them the most common complication affecting patients in hospital. There has not been a subsequent evidence-based estimate of the incidence of HAIs in Australia despite HAIs being widely reported as the most common complication affecting patients in hospital globally. The purpose of this systematic review was to explore the burden of HAIs in Australian hospitals by determining the incidence of HAIs in Australian hospitals, as reported in the peer-reviewed literature from 2010 to 2016.

      Methods

      Protocol and registration

      The protocol for this review is registered with PROSPERO, an international prospective register of systematic reviews (available at http://www.crd.york.ac.uk/prospero/ with registration number: CRD42016052997). Ethics approval was not sought because this review used data from published studies.

      Search strategy

      A systematic search was conducted using the electronic databases MEDLINE (PubMed) and the Cumulative Index to Nursing and Allied Health Literature (CINAHL) to search for articles published between 1 January 2010 and 31 December 2016. The timeframe was chosen to include the most current HAI data. During the 2000s, a number of state and national initiatives were undertaken to reduce HAIs, including the national hand hygiene initiative, development of National Health and Medical Research Council guidelines, revision of hospital accreditation standards, surveillance initiatives and public reporting of some infection data [
      • Graves N.
      • Page K.
      • Martin E.
      • Brain D.
      • Hall L.
      • Campbell M.
      • et al.
      Cost-effectiveness of a national initiative to improve hand hygiene compliance using the outcome of healthcare associated Staphylococcus aureus bacteraemia.
      ,
      • Australian Commission on Safety and Quality in Health Care
      National safety and quality health service standards.
      ,
      • National Health and Medical Research Council
      Australian guidelines for the prevention and control of infection in healthcare (2010).
      ]. Reference lists of eligible articles relating to these initiatives were reviewed to identify any additional articles.
      To identify articles from MEDLINE and CINAHL, the search terms “surveillance”, “incidence”, “prevalence”, “frequency”, “rates or statistics” and “performance indicators” were used in combination with “nosocomial infection”, “hospital acquired infection”, “healthcare associated infection”, “cross infection” and “infection”. To identify and limit the search to articles from Australian hospitals, the search term “hospital” was used with “Australia”, “Australian”, “Queensland”, “New South Wales”, “Australian Capital Territory”, “Victoria”, “Tasmania”, “South Australia”, “Western Australia” and “Northern Territory”. These terms were applied using an all text search. For the MEDLINE search, the MeSH terms “epidemiology”, “cross Infection” and “disease transmission, infectious” were also used.

      Inclusion criteria

      Articles were eligible if they reported the results of cohort studies, case-control studies, cross-sectional studies, randomised controlled trials or case reports (reporting incidence) of HAI. For this study, cohort studies include studies where a population is followed up to determine whether they subsequently acquire an infection. To be eligible, data collection had to occur after 1 January 2010. Articles were limited to studies conducted in Australian hospitals. For international or multi-centre studies, data from Australian hospitals was included if the data was reported at this level. Where Australia-specific data was not available, the study was excluded.

      Exclusion criteria

      The following exclusion criteria applied: all grey literature, non-peer reviewed, conference abstracts, papers written in languages other than English, reviews, editorials, commentaries or policy statements.

      Definitions

      For this review, the definition of HAI used in the articles must have referenced a recognised standard such as a definition agreed or published by professional association or government agency, a definition widely used in the published literature or an ICD10 code that constitutes a HAI (not any infection). Any disputes about whether an article had used an appropriate definition were resolved by discussion among the reviewers.

      Study selection

      The titles and abstracts of the articles identified from MEDLINE and CINAHL were examined and assessed for relevance and appropriateness to the review question by one reviewer (CW). Those not relevant were excluded. For a random selection (10%) of articles, the abstracts were independently reviewed by another reviewer (BM) to identify any discrepancies; none were identified. Of the remaining articles, a full text review was undertaken (CW) to assess eligibility. Where there was uncertainty as to whether an article should be included, two researchers independently reviewed the articles (PR and RS). If there was disagreement, a third reviewer made the final determination (BM).

      Data collection process

      A data extraction form was developed for extracting data for the systematic review (Supplementary 1). All data items extracted were cross-checked. No attempt was made to contact the authors of papers that had missing data or unclear information.

      Risk of bias

      An assessment of study quality and risk of bias in the articles included in the review was conducted using the Newcastle–Ottawa Scale, as recommended by the Cochrane Collaboration [
      • Higgins J.
      • Green S.
      Cochrane handbook for systematic reviews of interventions.
      ,
      • Wells G.A.
      • Shea B.
      • O'connell D.
      • Peterson J.
      • Welch V.
      • Losos M.
      • et al.
      The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses.
      ]. The tool enables a maximum of nine stars to be awarded to a study. The content validity and inter-rater reliability of this tool has been established [
      • Wells G.A.
      • Shea B.
      • O'connell D.
      • Peterson J.
      • Welch V.
      • Losos M.
      • et al.
      The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses.
      ]. Two reviewers undertook this assessment independently (PR and RS). For any discrepancies, a third reviewer (BM) made the final determination. We were unable to determine level of bias associated with not including all HAIs in the review. Publications on the HAI incidence (or prevalence) varies considerably and includes the possibility of not publishing any data, not reporting on all HAIs a lack of sensitivity in HAI surveillance or methodological approaches used and a lack of validation. Such an issue is not limited to this study and is widespread in the area of HAI data, particularly in Australia [
      • Russo P.L.
      • Cheng A.C.
      • Mitchell B.G.
      • Hall L.
      Healthcare-associated infections in Australia: tackling the known unknowns.
      ,
      • Russo P.L.
      • Barnett A.G.
      • Cheng A.C.
      • Richards M.
      • Graves N.
      • Hall L.
      Differences in identifying healthcare associated infections using clinical vignettes and the influence of respondent characteristics: a cross-sectional survey of Australian infection prevention staff.
      ,
      • Mitchell B.G.
      • Gardner A.
      A model for influences on reliable and valid health care-associated infection data.
      ].

      Data analysis

      All extracted data was entered into a comprehensive evidence-base table for descriptive analysis; 95% confidence intervals were calculated using Poisson regression. Meta-analysis was not performed due to the studies' heterogeneous methods, in particular, data collection. Meta-regression was not possible for the same reason.

      Estimating the annual incidence of HAIs in Australia

      We estimated the annual incidence of HAIs in Australia by extrapolating and modelling data from multi-centre studies that observed the same type of infection. Only multi-centre studies were used in the estimation, as they are likely to be more representative than single site studies. To estimate the annual number of cases of hospital onset Staphylococcus aureus bacteraemia and hospital identified Clostridium difficile, the total number of cases of each infection (identified in the included studies) was divided was divided by population data and subsequently multiplied by the Australian population. Population data was aggregate population of the states and territories contributing to the data. Population data for states territories and Australia was obtained from the Australian Bureau of Statistics, with 2016 population estimates used [
      • Australian Bureau of Statistics
      Australian demographic statistics, Jun 2016.
      ].
      To calculate the estimated number of cases of respiratory infections and surgical site infections, the raw data from studies was recalculated (if not reported) to determine the incidence (number of cases/number of procedures surveyed × 100). Confidence intervals were also calculated. Where studies contained incidence data for more than one year, the data from the last reported year in the study was used for modelling to ensure the most recent data was used. If incidence data was not reported by year, the incidence for the entire study period was used for modelling.
      Data from the Australian Institute of Health and Welfare and the Australian Commission on Safety and Quality was used to determine the number of procedures performed each year in Australia [
      • Australian Institute of Health and Welfare
      Australian hospital statistics 2012–13.
      ,
      • Australian Institute of Health and Welfare
      Elective surgery waiting times 2014–15.
      ,
      • Australian Institute of Health and Welfare
      Admitted patient care 2014-15.
      ,
      • Australian Commission on Safety and Quality in Health Care
      Australian atlas of healthcare variation.
      ]. The calculated incidence was multiplied by the number of reported procedures per year, to give an estimated number of cases per year in Australia.
      The number of healthcare associated urinary tract infections was reported as an incidence per 100 patient admissions in the literature. To determine the number of patient admissions per year in Australia, separation data from the Australian Institute of Health and Welfare was used [
      • Australian Institute of Health and Welfare
      Admitted patient care 2014-15.
      ]. Overnight separations are the closest available denominator to patient admissions. The reported incidence in the literature was multiplied by the number of overnight separations, to provide an estimated number of cases per year in Australia.

      Results

      Study selection

      The initial database search identified 843 articles, with one additional article identified through hand searching. After duplicate articles were removed, 751 abstracts were reviewed and 105 articles were deemed suitable for a full text review. Twenty-four studies met the eligibility criteria and were included in this review. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow chart describing the papers identified from the search strategy is presented in Fig. 1.
      Fig. 1
      Figure 1PRIMSA flow diagram for study selection.

      Risk of bias

      When studies were assessed for risk of bias using the Newcastle–Ottawa scale, the number of stars awarded to studies ranged from five to eight, the median being seven.

      Study characteristics

      A summary of the included articles with an overview of their key characteristics is provided in Table 1. Of the 24 included articles, the most common infections surveyed were bloodstream infections (eight articles), surgical site or procedure infections (six articles) and respiratory infections (four articles). C. difficile infection was reported in three articles, with the remainder reporting multiple or other infections. There was variation in the methodological approaches used to identify infections, the definitions used to determine an infection, the patients included in the studies and the number of hospitals involved. Four of the articles involved hospitals in more than one state. Eight articles involved hospitals from Victoria only, making this state the most represented.
      Table 1Overview of study findings.
      Lead authorLocation (no hospitals)Hospital typeData collectionInfection typePopulationProcessKey resultsDefinitions
      Bloodstream infections
      Alcorn
      • Alcorn K.
      • Gerrard J.
      • Macbeth D.
      • Steele M.
      Seasonal variation in health care-associated bloodstream infection: increase in the incidence of gram-negative bacteremia in nonhospitalized patients during summer.
      QLD (1)Principal referral2001–2011Bloodstream infection (BSI)All hospitalRetrospectiveAll: 841/10,000 overnight discharges

      Gram negative: 7.8/10,000

      Gram positive: 12.6/10,000
      AUS
      Anderson
      • Anderson E.
      • Leahy O.
      • Cheng A.C.
      • Grummet J.
      Risk factors for infection following prostate biopsy – a case control study.
      NSW (1)Private2009–2013 (5 years)Bacteraemia following TRUS biopsy.One surgeon's patientsICD, data warehouse & chart review.1.5% (71 patients) bacteraemia over 5 years.EST
      Aubron
      • Aubron C.
      • Suzuki S.
      • Glassford N.J.
      • Garcia-Alvarez M.
      • Howden B.P.
      • Bellomo R.
      The epidemiology of bacteriuria and candiduria in critically ill patients.
      VIC (1)Principal referral2006–2011 (6 year)BSI associated with positive ICU-acquired positive urine cultureICURetrospectiveCandiduria 3.5 episodes/1000 ICU day

      Bacteriuria 2.8 episodes/1000 ICU day
      COM
      Coombs
      • Coombs G.W.
      • Nimmo G.R.
      • Pearson J.C.
      • Collignon P.J.
      • Bell J.M.
      • McLaws M.L.
      • et al.
      Australian group on antimicrobial resistance hospital-onset Staphylococcus aureus surveillance programme annual report, 2011.
      All (29)Various2011S. aureus bacteraemiaAll hospital, only first 100 collectedLaboratory based. Period prevalence.MRSA accounted for 30.3%, 2357 isolates.COM
      Mitchell
      • Mitchell B.G.
      • Collignon P.J.
      • McCann R.
      • Wilkinson I.J.
      • Wells A.
      A major reduction in hospital-onset Staphylococcus aureus bacteremia in Australia-12 years of progress: an observational study.
      WA, SA, TAS, ACT (132)VariousS. aureus bacteraemia1All hospitalProspective0.90/10,000 patient days (95% Cl 0.86–0.93)COM
      Mumford
      • Mumford V.
      • Reeve R.
      • Greenfield D.
      • Forde K.
      • Westbrook J.
      • Braithwaite J.
      Is accreditation linked to hospital infection rates? A 4-year, data linkage study of Staphylococcus aureus rates and accreditation scores in 77 Australian acute hospitals.
      NSW (77)Various2009–12S. aureus bacteraemia2All hospitalRetrospectivePeer group A = 1.52 per 10,000 bed days

      Peer group B = 0.93 per 10,000 bed days

      Peer group C = 1.06 per 10,000 bed days

      Peer group D = 0.62 per 10,000 bed days

      All (2012) = 0.77 per 10,000 bed days
      AUS
      Si
      • Si D.
      • Runnegar N.
      • Marquess J.
      • Rajmokan M.
      • Playford E.G.
      Characterising health care-associated bloodstream infections in public hospitals in Queensland, 2008–2012.
      QLD (23)Various2008–12S. aureus bacteraemia1,2All hospitalProspectiveInpatient BSI 6.0/10,000 patient days

      Inpatient, intravascular associated: 1.9/10,000 patient days

      Inpatient, S. aureus bacteraemia 1.0/10,000 patient days

      HCA S. aureus bacteraemia

      1.3/000 patient days
      Calculated from available data.
      AUS
      Wong
      • Wong S.W.
      • Gantner D.
      • McGloughlin S.
      • Leong T.
      • Worth L.J.
      • Klintworth G.
      • et al.
      The influence of intensive care unit-acquired central line-associated bloodstream infection on in-hospital mortality: a single-center risk-adjusted analysis.
      VIC (1)Principal referral2008–2014 (5 years, 10m)ICU-acquired central line-associated bloodstream infection (CLABSI)ICURetrospectiveICU-acquired CLABSI 1.1 per 1000 ICU CVC-daysINT
      Surgical site/procedure related
      Chandrananth
      • Chandrananth J.
      • Rabinovich A.
      • Karahalios A.
      • Guy S.
      • Tran P.
      Impact of adherence to local antibiotic prophylaxis guidelines on infection outcome after total hip or knee arthroplasty.
      VIC (3)Unknown2011–2014 (3 year)Surgical site infectionHip or knee arthroplasty.Retrospective2.7% infection rateINT
      Phan
      • Phan K.
      • Schultz K.
      • Huang C.
      • Halcrow S.
      • Fuller J.
      • McDowell D.
      • et al.
      External ventricular drain infections at the Canberra Hospital: a retrospective study.
      ACT (1)Principal referral2007–2010External ventricular drains infectionPatient with external ventricular drainsRetrospective13 infections from 114 patients (11.5%)CLIN
      Roth
      • Roth H.
      • Millar J.L.
      • Cheng A.C.
      • Byrne A.
      • Evans S.
      • Grummet J.
      The state of TRUS biopsy sepsis: readmissions to Victorian hospitals with TRUS biopsy-related infection over 5 years.
      VIC (311)Various2007–2012Infective complication following TRUS biopsyPatients post TRUS biopsyVictorian Admitted Episodes Data Set1.73%, 95% (Cl 1.63–1.92%) readmitted with infection (604 of 34,865 patients).CLIN
      Tao
      • Tao P.
      • Marshall C.
      • Bucknill A.
      Surgical site infection in orthopaedic surgery: an audit of peri-operative practice at a tertiary centre.
      VIC (1)Principal referral2012 (1 month)Surgical site infectionOrthopaedic patientsRetrospective7% surgical site infection. 95 procedures (6 superficial. 1 deep)
      Unadjusted rate/not risk adjusted.
      INT
      Worth
      • Worth L.J.
      • Bull A.L.
      • Spelman T.
      • Brett J.
      • Richards M.J.
      Diminishing surgical site infections in Australia: time trends in infection rates, pathogens and antimicrobial resistance using a comprehensive Victorian surveillance program, 2002–2013.
      VIC (81)Various2002–15Surgical site infectionAll hospitalProspective2.8 per 100 procedures surgical site infection. (95% CI 2.7–2.9). 5123 infections from 183,625 procedures.
      • -
        53.5% super-ficial
      • -
        23.6% deep incisional
      • -
        22.9% organ/space infections
      Coronary artery bypass surgery = 4.9/100 procedures (1137/23,108)

      Caesarean section = 2.0/100 procedures (1044/50,833)

      Hip/knee = 2.6/1000 procedures (1453/56,503)

      Colorectal = 8.6/100 procedures (753/8688)

      Other abdominal = 2.6/100 procedures (290/10,993)

      Cardiac non coronary artery bypass surgery = 3.8/100 procedures (133/3458)

      Hysterectomy = 2.0/100 procedures (131/6504)

      Vascular = 11.8/100 procedures (48/407)
      INT
      Ezzatzadegan
      • Ezzatzadegan S.
      • Chen S.
      • Chapman J.R.
      Invasive fungal infections after renal transplantation.
      NSW (1)Principal referral2000–2010 (10 years)Invasive fungal infectionRenal transplant recipientsRetrospective2.1%, 95% (CI 0.7%–3.4%)

      (10/471 patients)
      CLIN
      Respiratory
      Brogan
      • Brogan E.
      • Langdon C.
      • Brookes K.
      • Budgeon C.
      • Blacker D.
      Respiratory infections in acute stroke: nasogastric tubes and immobility are stronger predictors than dysphagia.
      NSW (6)Principal referral2010Respiratory infections in acute stroke.Stroke patientsRetrospective11.3 % (60/(533 patients))NS
      Gautam
      • Gautam A.
      • Ganu S.S.
      • Tegg O.J.
      • Andresen D.N.
      • Wilkins B.H.
      • Schell D.N.
      Ventilator-associated pneumonia in a tertiary paediatric intensive care unit: a 1-year prospective observational study.
      NSW (1)Principal referral(1 year)Ventilator-associated pneumoniapaediatric intensive care unit:Prospective observational7.02/1000 ventilation days

      (18 episodes/269 patients)
      INT
      Macesic
      • Macesic N.
      • Kotsimbos T.C.
      • Kelly P.
      • Cheng A.C.
      Hospital-acquired influenza in an Australian sentinel surveillance system.
      All (15)Various2010–11Hospital-acquired influenzaAll hospitalSentinel4.3% of cases were nosocomial. (26/598 cases)CLIN
      Sanagou
      • Sanagou M.
      • Leder K.
      • Cheng A.C.
      • Pilcher D.
      • Reid C.M.
      • Wolfe R.
      Associations of hospital characteristics with nosocomial pneumonia after cardiac surgery can impact on standardized infection rates.
      NS (16)Various2001–2011HCA pneumonia post cardiac surgery.Retrospective (registry)5.1% (2229 cases/43,691 patients)
      C. difficile
      Foster
      • Foster N.F.
      • Collins D.A.
      • Ditchburn S.L.
      • Duncan C.N.
      • Schalkwyk J.W.
      • Golledge C.L.
      • et al.
      Epidemiology of Clostridium difficile infection in two tertiary-care hospitals in Perth, Western Australia: a cross-sectional study.
      WA (2)Principal referral2011–12 (6 months)C. difficile infectionAll hospitalRetrospectiveHospital 1 = 6.8 cases/10,000 OBDs

      Hospital 2 = 8.0 cases/10,000 OBDs.

      53.8% of cases were hospital-onset.
      Unable to determine combined incidence from available data.
      AUS/INT
      Slimings
      • Slimings C.
      • Armstrong P.
      • Beckingham W.D.
      • Bull A.L.
      • Hall L.
      • Kennedy K.J.
      • et al.
      Increasing incidence of Clostridium difficile infection, Australia, 2011–2012.
      TAS, VIC, NSW, QLD, SA, WA, ACT (450)Various2011–12C. difficile infectionAll hospitalProspectiveHospital identified: 3.65 (95% CI 3.58–3.71) per 10,000 patient days (12,683 cases/34786323 patient day)

      Hospital associated 2.95 (95% CI 2.86–3.04) per 10,000 patient days (WA, SA, ACT,TAS, VIC only. 6759 cases/5,585,533 patient days)
      AUS
      Worth (2016)
      • Worth L.J.
      • Spelman T.
      • Bull A.L.
      • Brett J.A.
      • Richards M.J.
      Epidemiology of Clostridium difficile infections in Australia: enhanced surveillance to evaluate time trends and severity of illness in Victoria, 2010–2014.
      VIC (136)Various2010–2014C. difficile infectionAll hospitalProspectiveHealthcare associated 2.49 cases per 10,000 occupied bed days. (4826 cases)INT
      All other
      Jarratt
      • Jarratt L.S.
      • Miller E.R.
      The relationship between patient characteristics and the development of a multi-resistant healthcare-associated infection in a private South Australian hospital.
      SA (1)Private hospital2003–2011 (9 years)HCA infections (various)All hospital (>18 years)Retrospective1017 HAIs

      ∼0.63% of all patients admitted to the hospital
      CON
      Mathot
      • Mathot F.
      • Duke T.
      • Daley A.J.
      • Butcher T.
      Bacteremia and pneumonia in a tertiary PICU: an 11-year study.
      VIC (1)Children2002–12 (11 years)Bacteraemia and pneumoniaPaediatric ICUProspective8.9% patients had serious bacterial infections (blood culture or bronchial alveolar lavage) (881 of 9947 patients)CLIN
      Mitchell
      • Mitchell B.G.
      • Ferguson J.K.
      • Anderson M.
      • Sear J.
      • Barnett A.
      Length of stay and mortality associated with healthcare-associated urinary tract infections: a multi-state model.
      NSW (8)VariousUrinary tract infection1All hospitalRetrospective1.73% (96% CI 1.67–1.80) of admitted patients. (2821/162,503 patient admissions)EST
      Note: HCA = healthcare associated. BSI = bloodstream infection. ICU = intensive care unit. TRUS = transrectal ultrasound. ICD = International Statistical Classification of Diseases and Related Health Problems. WA = Western Australia. SA = South Australia. ACT = Australian Capital Territory. TAS = Tasmania. NSW = New South Wales. QLD = Queensland. MRSA = Methicillin resistant S. aureus AUS = Consistent with Australian or nationally used/agreed definition. INT = Consistent with national Healthcare Safety Network (NHSN) and or Centres for Disease Control definitions. EST = consistent with an established or previously documented definition (for example, professional association). NS = Not stated. COM = Common approach to defining a healthcare associated infection – consistent with not incubating on admission and or acquired 48 h after admission. CLIN = defined as an infection based on clinical grounds, assessment or review of notes. 1Hospital onset. 2Healthcare associated.
      a Calculated from available data.
      b Unadjusted rate/not risk adjusted.
      c Unable to determine combined incidence from available data.

      Results by type of infection

      Bloodstream infection

      Four of the eight articles reported bloodstream infections at a hospital level; the remainder examined either bloodstream infections within an intensive care unit or patients who were undergoing a procedure. There was further variation in the types of bloodstream infections surveyed, with S. aureus bacteraemia the most common. There was some consistency in the methods used to determine the incidence of S. aureus bacteraemia. The reported incidence of hospital-acquired S. aureus bacteraemia was 0.77 per 10,000 bed days from New South Wales hospitals and 1.33 per 10,000 bed days from Queensland hospitals [
      • Si D.
      • Runnegar N.
      • Marquess J.
      • Rajmokan M.
      • Playford E.G.
      Characterising health care-associated bloodstream infections in public hospitals in Queensland, 2008–2012.
      ]. Mitchell et al. reported an incidence of 0.90 per 10,000 bed days for hospital-onset S. aureus bacteraemia [
      • Mitchell B.G.
      • Collignon P.J.
      • McCann R.
      • Wilkinson I.J.
      • Wells A.
      A major reduction in hospital-onset Staphylococcus aureus bacteremia in Australia-12 years of progress: an observational study.
      ], but did not include hospitals from New South Wales, Queensland or Victoria. In this study by Mitchell et al. [
      • Mitchell B.G.
      • Collignon P.J.
      • McCann R.
      • Wilkinson I.J.
      • Wells A.
      A major reduction in hospital-onset Staphylococcus aureus bacteremia in Australia-12 years of progress: an observational study.
      ], hospital-acquired S. aureus bacteraemia was defined as a S. aureus bacteraemia occurring more than 48 h after admission, whereas the article based on New South Wales and Queensland data used these criteria in addition to other criteria. The combined incidence of hospital-acquired S. aureus bacteraemia was 0.93 cases per 10,000 bed days (95% CI 0.90–0.96) [
      • Si D.
      • Runnegar N.
      • Marquess J.
      • Rajmokan M.
      • Playford E.G.
      Characterising health care-associated bloodstream infections in public hospitals in Queensland, 2008–2012.
      ,
      • Mitchell B.G.
      • Collignon P.J.
      • McCann R.
      • Wilkinson I.J.
      • Wells A.
      A major reduction in hospital-onset Staphylococcus aureus bacteremia in Australia-12 years of progress: an observational study.
      ]. Other articles reporting bloodstream infection rates were limited to one hospital [
      • Alcorn K.
      • Gerrard J.
      • Macbeth D.
      • Steele M.
      Seasonal variation in health care-associated bloodstream infection: increase in the incidence of gram-negative bacteremia in nonhospitalized patients during summer.
      ,
      • Anderson E.
      • Leahy O.
      • Cheng A.C.
      • Grummet J.
      Risk factors for infection following prostate biopsy – a case control study.
      ,
      • Aubron C.
      • Suzuki S.
      • Glassford N.J.
      • Garcia-Alvarez M.
      • Howden B.P.
      • Bellomo R.
      The epidemiology of bacteriuria and candiduria in critically ill patients.
      ,
      • Wong S.W.
      • Gantner D.
      • McGloughlin S.
      • Leong T.
      • Worth L.J.
      • Klintworth G.
      • et al.
      The influence of intensive care unit-acquired central line-associated bloodstream infection on in-hospital mortality: a single-center risk-adjusted analysis.
      ] or did not report the incidence of bloodstream infection [
      • Coombs G.W.
      • Nimmo G.R.
      • Pearson J.C.
      • Collignon P.J.
      • Bell J.M.
      • McLaws M.L.
      • et al.
      Australian group on antimicrobial resistance hospital-onset Staphylococcus aureus surveillance programme annual report, 2011.
      ].

      Surgical site or procedure related infection

      Three studies reported the incidence of surgical site infection in orthopaedic patients [
      • Chandrananth J.
      • Rabinovich A.
      • Karahalios A.
      • Guy S.
      • Tran P.
      Impact of adherence to local antibiotic prophylaxis guidelines on infection outcome after total hip or knee arthroplasty.
      ,
      • Tao P.
      • Marshall C.
      • Bucknill A.
      Surgical site infection in orthopaedic surgery: an audit of peri-operative practice at a tertiary centre.
      ,
      • Worth L.J.
      • Bull A.L.
      • Spelman T.
      • Brett J.
      • Richards M.J.
      Diminishing surgical site infections in Australia: time trends in infection rates, pathogens and antimicrobial resistance using a comprehensive Victorian surveillance program, 2002–2013.
      ], but only two of these reported rates at a procedure level [
      • Chandrananth J.
      • Rabinovich A.
      • Karahalios A.
      • Guy S.
      • Tran P.
      Impact of adherence to local antibiotic prophylaxis guidelines on infection outcome after total hip or knee arthroplasty.
      ,
      • Worth L.J.
      • Bull A.L.
      • Spelman T.
      • Brett J.
      • Richards M.J.
      Diminishing surgical site infections in Australia: time trends in infection rates, pathogens and antimicrobial resistance using a comprehensive Victorian surveillance program, 2002–2013.
      ]. Both of these studies included hospitals from Victoria and had overlapping study periods. In addition to these studies, Roth et al. [
      • Roth H.
      • Millar J.L.
      • Cheng A.C.
      • Byrne A.
      • Evans S.
      • Grummet J.
      The state of TRUS biopsy sepsis: readmissions to Victorian hospitals with TRUS biopsy-related infection over 5 years.
      ] undertook a multi-centre study to determine the incidence of infectious complications following trans-rectal ultrasound. This study examined data from 31,209 patients from public and private hospitals over a five-year period and found that 1.73% of these patients were readmitted with a biopsy-related infection [
      • Roth H.
      • Millar J.L.
      • Cheng A.C.
      • Byrne A.
      • Evans S.
      • Grummet J.
      The state of TRUS biopsy sepsis: readmissions to Victorian hospitals with TRUS biopsy-related infection over 5 years.
      ].

      Respiratory

      The articles that reported respiratory HAI were heterogeneous in terms of the type of infections, the participants monitored and the methodological approaches used. Three were multi-centre studies but reported on different HAI: respiratory infections in acute stroke patients [
      • Brogan E.
      • Langdon C.
      • Brookes K.
      • Budgeon C.
      • Blacker D.
      Respiratory infections in acute stroke: nasogastric tubes and immobility are stronger predictors than dysphagia.
      ], influenza [
      • Macesic N.
      • Kotsimbos T.C.
      • Kelly P.
      • Cheng A.C.
      Hospital-acquired influenza in an Australian sentinel surveillance system.
      ] and pneumonia following cardiac surgery [
      • Sanagou M.
      • Leder K.
      • Cheng A.C.
      • Pilcher D.
      • Reid C.M.
      • Wolfe R.
      Associations of hospital characteristics with nosocomial pneumonia after cardiac surgery can impact on standardized infection rates.
      ].

      C. difficile infection

      Three multi-centre studies examined C. difficile infection with overlapping data collection periods [
      • Foster N.F.
      • Collins D.A.
      • Ditchburn S.L.
      • Duncan C.N.
      • Schalkwyk J.W.
      • Golledge C.L.
      • et al.
      Epidemiology of Clostridium difficile infection in two tertiary-care hospitals in Perth, Western Australia: a cross-sectional study.
      ,
      • Slimings C.
      • Armstrong P.
      • Beckingham W.D.
      • Bull A.L.
      • Hall L.
      • Kennedy K.J.
      • et al.
      Increasing incidence of Clostridium difficile infection, Australia, 2011–2012.
      ,
      • Worth L.J.
      • Spelman T.
      • Bull A.L.
      • Brett J.A.
      • Richards M.J.
      Epidemiology of Clostridium difficile infections in Australia: enhanced surveillance to evaluate time trends and severity of illness in Victoria, 2010–2014.
      ]. The study by Slimings et al. [
      • Slimings C.
      • Armstrong P.
      • Beckingham W.D.
      • Bull A.L.
      • Hall L.
      • Kennedy K.J.
      • et al.
      Increasing incidence of Clostridium difficile infection, Australia, 2011–2012.
      ] was conducted in seven Australian states and territories and involved 450 public hospitals. The incidence of hospital-identified C. difficile infection was 3.65 per 10,000 bed days, while the incidence of hospital-associated C. difficile infection was 2.95 per 10,000 bed days (from five states and territories). This finding is consistent with that of Worth et al. [
      • Worth L.J.
      • Spelman T.
      • Bull A.L.
      • Brett J.A.
      • Richards M.J.
      Epidemiology of Clostridium difficile infections in Australia: enhanced surveillance to evaluate time trends and severity of illness in Victoria, 2010–2014.
      ], who examined C. difficile infection in 136 healthcare facilities over a four-year period and found the incidence of healthcare-associated C. difficile infection to be 2.49 cases per 10,000 bed days. From the published data, we were unable to calculate the incidence of healthcare-associated C. difficile infection in the study undertaken in Western Australia [
      • Foster N.F.
      • Collins D.A.
      • Ditchburn S.L.
      • Duncan C.N.
      • Schalkwyk J.W.
      • Golledge C.L.
      • et al.
      Epidemiology of Clostridium difficile infection in two tertiary-care hospitals in Perth, Western Australia: a cross-sectional study.
      ].

      Other

      Three studies examined other HAI but only one study was multi-centre [
      • Mitchell B.G.
      • Ferguson J.K.
      • Anderson M.
      • Sear J.
      • Barnett A.
      Length of stay and mortality associated with healthcare-associated urinary tract infections: a multi-state model.
      ]. In this study examining the incidence of healthcare-associated urinary tract infections in eight hospitals over a four-year period, the authors found that 1.7% of admitted patients acquired a HAI of the urinary tract.

      Estimates of yearly HAI incidence

      Using data from multi-centre studies, we sought to estimate the number of HAI occurring each year in Australia (Table 2). From the available data, we estimate 83,096 HAIs occur each year in Australia, however this number is a significant underestimation, it was only based on peer-reviewed published multi-centre studies Included in our review. The 83,086 HAIs comprised of 71,186 urinary tract infections, 4902 healthcare associated C. difficile infections, 3946 surgical site infections, 1962 respiratory infections in acute stroke patients and 1100 hospital onset S. aureus bacteraemia infections.
      Table 2Multi-centre studies and estimates of healthcare-associated infections in Australia.
      InfectionLead authorData extracted or recalculated from studies for modelEstimated denominator.-Modelled number of cases in Australia each year
      Bloodstream infection
      Hospital onset S. aureus bacteraemiaMitchell
      • Mitchell B.G.
      • Collignon P.J.
      • McCann R.
      • Wilkinson I.J.
      • Wells A.
      A major reduction in hospital-onset Staphylococcus aureus bacteremia in Australia-12 years of progress: an observational study.
      253 cases in WA, SA, ACT, TAS in 2012.5.24 million population1100 cases per year (95% CI 995–1193)
      Si
      • Si D.
      • Runnegar N.
      • Marquess J.
      • Rajmokan M.
      • Playford E.G.
      Characterising health care-associated bloodstream infections in public hospitals in Queensland, 2008–2012.
      207 cases from QLD in 2012.4.84 million population.
      Healthcare associated S.aureus bacteraemiaMumford
      • Mumford V.
      • Reeve R.
      • Greenfield D.
      • Forde K.
      • Westbrook J.
      • Braithwaite J.
      Is accreditation linked to hospital infection rates? A 4-year, data linkage study of Staphylococcus aureus rates and accreditation scores in 77 Australian acute hospitals.
      0.77 (SD 0.77)Unable to model
      Insufficient data published in the study (numerator and or denominator).
      Subtotal1100
      Surgical site infection
      Hip and kneeWorth
      • Worth L.J.
      • Bull A.L.
      • Spelman T.
      • Brett J.
      • Richards M.J.
      Diminishing surgical site infections in Australia: time trends in infection rates, pathogens and antimicrobial resistance using a comprehensive Victorian surveillance program, 2002–2013.
      2.57/100 procedures (95% CI 2.44–2.71)25,938 procedures/year.Hip/knee = 667 cases
      Data from Worth et al. [27] only used. Data collection from Chandrananth et al. [25] overlaps and may include the same hospitals.
      ,
      Unadjusted number of infections.
      (Public hospitals)
      Chandrananth
      • Chandrananth J.
      • Rabinovich A.
      • Karahalios A.
      • Guy S.
      • Tran P.
      Impact of adherence to local antibiotic prophylaxis guidelines on infection outcome after total hip or knee arthroplasty.
      2.75/100 procedures (95% CI 1.86–3.9)ˆ
      Coronary artery bypass surgeryWorth
      • Worth L.J.
      • Bull A.L.
      • Spelman T.
      • Brett J.
      • Richards M.J.
      Diminishing surgical site infections in Australia: time trends in infection rates, pathogens and antimicrobial resistance using a comprehensive Victorian surveillance program, 2002–2013.
      4.51/100 procedures (95% CI 2.72–7.05)3837 procedures/year.189 cases
      Unadjusted number of infections.
      (Public hospitals)
      Caesarian sectionWorth
      • Worth L.J.
      • Bull A.L.
      • Spelman T.
      • Brett J.
      • Richards M.J.
      Diminishing surgical site infections in Australia: time trends in infection rates, pathogens and antimicrobial resistance using a comprehensive Victorian surveillance program, 2002–2013.
      2.75 (95% CI 2.26–3.32)95,894 procedures/year.1969 cases
      Unadjusted number of infections.
      (All hospitals)
      HysterectomyWorth
      • Worth L.J.
      • Bull A.L.
      • Spelman T.
      • Brett J.
      • Richards M.J.
      Diminishing surgical site infections in Australia: time trends in infection rates, pathogens and antimicrobial resistance using a comprehensive Victorian surveillance program, 2002–2013.
      2.01 (95% CI 1.68–2.39)9989 procedures/year.201 cases
      Unadjusted number of infections.
      (Public hospitals)
      VascularWorth
      • Worth L.J.
      • Bull A.L.
      • Spelman T.
      • Brett J.
      • Richards M.J.
      Diminishing surgical site infections in Australia: time trends in infection rates, pathogens and antimicrobial resistance using a comprehensive Victorian surveillance program, 2002–2013.
      11.79 (8.70–15.64)4002 procedures/year.472 cases
      Unadjusted number of infections.
      .
      Based on a denominator of the number of varicose veins procedures in one year. Only available data.
      (Public hospitals)
      Infective complication following TRUS biopsyRoth
      • Roth H.
      • Millar J.L.
      • Cheng A.C.
      • Byrne A.
      • Evans S.
      • Grummet J.
      The state of TRUS biopsy sepsis: readmissions to Victorian hospitals with TRUS biopsy-related infection over 5 years.
      1.73%, (95% Cl 1.63–1.92%) readmitted with infection25,869 procedures/year.448 cases (All hospitals)
      Subtotal3946
      Respiratory
      Respiratory infections in acute strokeBrogan
      • Brogan E.
      • Langdon C.
      • Brookes K.
      • Budgeon C.
      • Blacker D.
      Respiratory infections in acute stroke: nasogastric tubes and immobility are stronger predictors than dysphagia.
      11.3% (95% CI 8.7–14.4%) of patients15,684 cases on hospital per year1766 cases
      Based on a denominator of cerebral infarction and stroke ischaemic hospital data.
      (All hospitals)
      InfluenzaMacesic
      • Macesic N.
      • Kotsimbos T.C.
      • Kelly P.
      • Cheng A.C.
      Hospital-acquired influenza in an Australian sentinel surveillance system.
      4.5% (95% CI 2.9–6.3%)ˆUnable to model
      Insufficient data published in the study (numerator and or denominator).
      Pneumonia post cardiac surgery.Sanagou
      • Sanagou M.
      • Leder K.
      • Cheng A.C.
      • Pilcher D.
      • Reid C.M.
      • Wolfe R.
      Associations of hospital characteristics with nosocomial pneumonia after cardiac surgery can impact on standardized infection rates.
      5.1% (95% CI 4.9–5.3%)ˆ3837 procedures/year.196 cases (Public hospital)
      Subtotal1962
      Other
      C. difficile infection (healthcare associated)Slimings
      • Slimings C.
      • Armstrong P.
      • Beckingham W.D.
      • Bull A.L.
      • Hall L.
      • Kennedy K.J.
      • et al.
      Increasing incidence of Clostridium difficile infection, Australia, 2011–2012.
      2.95 (95% CI 2.86–3.04) per 10,000 patient days. 2298 cases from WA, SA, AC, VIC and TAS.#11.3m population.4902 cases per year
      Data from Slimings et al. [33] only used. Data collection from Worth et al. [34] overlaps.
      C. difficile infection (healthcare associated)Worth
      • Worth L.J.
      • Spelman T.
      • Bull A.L.
      • Brett J.A.
      • Richards M.J.
      Epidemiology of Clostridium difficile infections in Australia: enhanced surveillance to evaluate time trends and severity of illness in Victoria, 2010–2014.
      2.49 cases per 10,000 OBDs +
      Healthcare associated urinary tract infectionMitchell
      • Mitchell B.G.
      • Ferguson J.K.
      • Anderson M.
      • Sear J.
      • Barnett A.
      Length of stay and mortality associated with healthcare-associated urinary tract infections: a multi-state model.
      1.73% (95% CI 1.67–1.80) of admitted patients4.1m overnight separations per year
      Overnight separations used as the closest denominator to patient admissions.
      71,186 cases (95% CI 68,481–73,812) (All hospitals)
      Subtotal76,088
      Total83,096
      Notes: WA = Western Australia. SA = South Australia. ACT = Australian Capital Territory. TAS = Tasmania. TRUS = transrectal ultrasound biopsy. ˆ 95% CI calculated from data provided in the original study.-Estimated denominators as based on the best available data from the Australian Bureau of Statistics, Australian Institute of Health and Welfare and the Australian Commission on Safety and Quality in Healthcare. + 95% CI could not be calculated from data provided in the original study. # 2298 cases calculated from data a reported incidence of 2.95 per 10,000 bed days (4597 cases over two years) from WA, SA, ACT, VIC and TAS.
      a Insufficient data published in the study (numerator and or denominator).
      b Data from Worth et al.
      • Worth L.J.
      • Bull A.L.
      • Spelman T.
      • Brett J.
      • Richards M.J.
      Diminishing surgical site infections in Australia: time trends in infection rates, pathogens and antimicrobial resistance using a comprehensive Victorian surveillance program, 2002–2013.
      only used. Data collection from Chandrananth et al.
      • Chandrananth J.
      • Rabinovich A.
      • Karahalios A.
      • Guy S.
      • Tran P.
      Impact of adherence to local antibiotic prophylaxis guidelines on infection outcome after total hip or knee arthroplasty.
      overlaps and may include the same hospitals.
      c Unadjusted number of infections.
      d Based on a denominator of the number of varicose veins procedures in one year. Only available data.
      e Based on a denominator of cerebral infarction and stroke ischaemic hospital data.
      f Data from Slimings et al.
      • Slimings C.
      • Armstrong P.
      • Beckingham W.D.
      • Bull A.L.
      • Hall L.
      • Kennedy K.J.
      • et al.
      Increasing incidence of Clostridium difficile infection, Australia, 2011–2012.
      only used. Data collection from Worth et al.
      • Worth L.J.
      • Spelman T.
      • Bull A.L.
      • Brett J.A.
      • Richards M.J.
      Epidemiology of Clostridium difficile infections in Australia: enhanced surveillance to evaluate time trends and severity of illness in Victoria, 2010–2014.
      overlaps.
      g Overnight separations used as the closest denominator to patient admissions.

      Discussion

      This paper presents the first systematic review of Australian HAI rates published in the peer-reviewed literature. We identified 24 studies published between 2010 and 2016. The infections surveyed, their definitions and the methods used varied considerably. To understand the magnitude of HAIs and associated burden of disease in Australia more fully, we attempted to estimate the incidence of HAIs using data from multi-centre studies.
      A previous report estimated the incidence of HAIs in Australia to be 177,392 HAIs per year (headline number of 200,000 HAIs) [
      ]. This estimate was made several years ago and was based on one study conducted in one hospital; the best available data at that time. Unfortunately, the evidence base for estimating the incidence of HAIs in Australia remains lacking, certainly when using data from the peer-reviewed literature. Our review has highlighted a myriad of different approaches to measure and define HAIs. In many instances, the data presented was limited, making comparisons or extrapolation of data impossible. In the absence of a national system for reporting there is a real need for those undertaking HAI surveillance – both at a hospital and a state level – to publish HAI surveillance data in the peer-reviewed literature. Accessing such information from government websites is fraught with challenges, including lack of detail about methods, delays in publishing data or failure to publish data. These issues do not apply to all states and territory health departments, but there are significant gaps across the country. This was also reflected in the results of our literature review, where more articles were published from Victorian data than any other state. Limited data is available from other states or on the MyHospitals website (www.myhospitals.gov.au).
      National point prevalence studies are undertaken periodically in numerous countries including the United States, Canada and all European countries [
      • Magill S.S.
      • Edwards J.R.
      • Bamberg W.
      • Beldavs Z.G.
      • Dumyati G.
      • Kainer M.A.
      • et al.
      Multistate point-prevalence survey of health care–associated infections.
      ,
      • Zarb P.
      • Coignard B.
      • Griskeviciene J.
      • Muller A.
      • Vankerckhoven V.
      • Weist K.
      • et al.
      The European Centre for Disease Prevention and Control (ECDC) pilot point prevalence survey of healthcare-associated infections and antimicrobial use.
      ,
      • Gravel D.
      • Taylor G.
      • Ofner M.
      • Johnston L.
      • Loeb M.
      • Roth V.
      • et al.
      The Canadian Nosocomial Infection Surveillance Program
      Point prevalence survey for healthcare-associated infections within Canadian adult acute-care hospitals.
      ]. Australia is one of the few Organisation for Economic Co-operation and Development countries that does not undertake national HAI point prevalence studies, or have a national surveillance program [
      • Russo P.L.
      • Cheng A.C.
      • Mitchell B.G.
      • Hall L.
      Healthcare-associated infections in Australia: tackling the known unknowns.
      ]. These studies provide valuable insights into the most frequent HAIs: gastroenterological, respiratory, surgical site and urinary tract. Our review has identified few studies that explored the incidence of gastroenterological and respiratory infections, despite these being the most common. Surgical site infection data was well represented from studies from Victoria and urinary tract infection from New South Wales [
      • Tao P.
      • Marshall C.
      • Bucknill A.
      Surgical site infection in orthopaedic surgery: an audit of peri-operative practice at a tertiary centre.
      ,
      • Worth L.J.
      • Bull A.L.
      • Spelman T.
      • Brett J.
      • Richards M.J.
      Diminishing surgical site infections in Australia: time trends in infection rates, pathogens and antimicrobial resistance using a comprehensive Victorian surveillance program, 2002–2013.
      ,
      • Roth H.
      • Millar J.L.
      • Cheng A.C.
      • Byrne A.
      • Evans S.
      • Grummet J.
      The state of TRUS biopsy sepsis: readmissions to Victorian hospitals with TRUS biopsy-related infection over 5 years.
      ,
      • Mitchell B.G.
      • Ferguson J.K.
      • Anderson M.
      • Sear J.
      • Barnett A.
      Length of stay and mortality associated with healthcare-associated urinary tract infections: a multi-state model.
      ]. The data included in our review misses a large proportion of HAIs due to the absence of published incidence data.
      Using data from the multi-centre studies to improve representativeness, we estimated the incidence of some types of HAIs. We estimate that each year in Australia, there are 71,186 urinary tract infections, 4902 C. difficile infections, 3946 surgical site infections, 1962 respiratory infections in acute stroke patients and 1100 S. aureus bacteraemia. These total 83,096 HAIs each year, but would be a very large underestimation given the lack of data on common infections. These estimates also come with considerable caveats due to the different methodologies and definitions. In the only national data available, our estimates for S. aureus bacteraemia are broadly consistent with that reported by the Australian Institute of Health and Welfare, who report 1440 cases of healthcare associated S. aureus bacteraemia in the 2015–16 financial year [
      • Australian Institute of Health and Welfare
      Staphylococcus aureus bacteraemia in Australian public hospitals 2015–16: Australian hospital statistics. Health services series no. 74. Cat. No. HSE 184.
      ]. The difference is likely explained by case definitions (hospital onset Vs healthcare associated).
      Data from the United States and Europe suggests that pneumonia accounts for 21.8%–25.7% of HAIs: 7.8%–17.1% are gastroenterological infections, 9.9%–14.2% are bloodstream infections, and 3.1%–5.6% are ear, nose and throat infections [
      • Magill S.S.
      • Edwards J.R.
      • Bamberg W.
      • Beldavs Z.G.
      • Dumyati G.
      • Kainer M.A.
      • et al.
      Multistate point-prevalence survey of health care–associated infections.
      ,
      • Zarb P.
      • Coignard B.
      • Griskeviciene J.
      • Muller A.
      • Vankerckhoven V.
      • Weist K.
      • et al.
      The European Centre for Disease Prevention and Control (ECDC) pilot point prevalence survey of healthcare-associated infections and antimicrobial use.
      ]. Our review only had data on the incidence of pneumonia from one group of patients (post-stroke) and one bloodstream infection (S. aureus bacteraemia). C. difficile infection was the only gastroenterological infection included in our review and we had no data on ear, nose or throat infections or many other HAIs such as central nervous system, reproductive tract and skin and soft tissue infections. Therefore, our estimate of 83,096 infections does not take into account a large proportion of pneumonia, gastroenterological, bloodstream infection or other infection data, thus potentially missing up to 50%–60% of infections. If that is the case, the incidence of HAIs in Australia may be closer to 165,000 per year.
      One limitation of our review was that we did not source grey literature, specifically HAI data published on state and territory government websites. These data are difficult to locate, are reported differently and often used methodology that is not adequately described in an open forum. We recommend a follow-up review, focussing on collating and analysing this data. This approach was outside the scope of our review and available resources, and should be undertake in a future review.
      Having a reliable estimate for the incidence of HAIs is important for a number of reasons. First, in an era of finite health resources, it helps prioritise infection prevention and control strategies. Second, it provides a benchmark against which future achievements or targets can be measured and evaluated. Third, it enables industry and those involved in healthcare innovation to have more reliable data for investment in products and research. Finally, it helps determine the resources required for HAI prevention and control relative to other health issues. On the eve of a global ‘post-antibiotic era’ [
      • Jones C.A.
      • Davis J.S.
      • Looke D.F.
      Death from an untreatable infection may signal the start of the post-antibiotic era.
      ], the need for coordinated and systematic national surveillance and reporting of HAIs and established contributing factors, namely antimicrobial resistance and antibiotic usage, has never been greater.
      We believe three concurrent strategies are required to address this gap. There needs to be a determination and action by state and national government bodies to achieve consensus on definitions, approaches to surveillance and transparent regular reporting. This should happen in parallel to the establishment of a national HAI surveillance program. There have been calls for a national centre for disease control [
      • Australian Medical Association
      Australian national centre for Disease Control (CDC) – 2017.
      ], potentially a national HAI surveillance program could be incorporated into such a centre. Finally, a national point prevalence study would provide valuable insight in the short term on the burden of HAIs in Australia. Given that the latter does not require a sophisticated study design and provides descriptive results, it has not been favourably supported by funding bodies. Whilst these suggested strategies have to date failed to eventuate, we call on all those involved in undertaking HAI surveillance in Australia to work collaboratively and publish data in the peer-reviewed literature.

      Authorship statement

      BM and PR conceived the study and drafted the research protocol. BM, PR, RS and DM provided critical review of and approved the study design. BM conducted the database searches. CW made the primary selection of eligible papers including data extraction. BM, PR and RS supervised and checked the study selection process and data extraction. BM analysed the data. All authors contributed to interpretation of the analysis. BM and CW wrote the manuscript. All authors provided critical review and approved the final manuscript.

      Conflicts of interest

      BM, PR and DM were authors on some of the included papers in this review. The decision to include/exclude studies which they authored, in addition to a risk of bias assessment was made independently by other members of the research team. All data extracted from these studies were independently checked by other members of the research team. The authors have no other conflicts to declare.

      Funding

      Funding to support this study was provided by Avondale College of Higher Education, via a summer scholarship grant.

      Provenance and peer review

      Not commissioned; externally peer reviewed.

      Appendix A. Supplementary data

      The following is the supplementary data related to this article:

      References

        • Yokoe D.S.
        • Anderson D.J.
        • Berenholtz S.M.
        • Calfee D.P.
        • Dubberke E.R.
        • Ellingson K.D.
        • et al.
        A compendium of strategies to prevent healthcare-associated infections in Acute Care Hospitals: 2014 updates.
        Am J Infect Control. 2014; 42: 820-828
        • Magill S.S.
        • Edwards J.R.
        • Bamberg W.
        • Beldavs Z.G.
        • Dumyati G.
        • Kainer M.A.
        • et al.
        Multistate point-prevalence survey of health care–associated infections.
        N Engl J Med. 2014; 370: 1198-1208
        • McLaws M.
        • Gold J.
        • King K.
        • Irwig L.
        • Berry G.
        The prevalence of nosocomial and community-acquired infections in Australian hospitals.
        Med J Aust. 1987; 149: 582-590
      1. Cruickshank M. Ferguson J. Reducing harm to patients from healthcare associated infections: an Australian infection prevention and control model for acute hospitals. Australian Commission on Safety and Quality in Health Care, Sydney2009
        • Graves N.
        • Page K.
        • Martin E.
        • Brain D.
        • Hall L.
        • Campbell M.
        • et al.
        Cost-effectiveness of a national initiative to improve hand hygiene compliance using the outcome of healthcare associated Staphylococcus aureus bacteraemia.
        PLoS One. 2016; 11: e0148190
        • Australian Commission on Safety and Quality in Health Care
        National safety and quality health service standards.
        Australian Commission on Safety and Quality in Health Care, Sydney2011
        • National Health and Medical Research Council
        Australian guidelines for the prevention and control of infection in healthcare (2010).
        National Health and Medical Research Council, Canberra2010
        • Higgins J.
        • Green S.
        Cochrane handbook for systematic reviews of interventions.
        Cochrane Collaboration, 2011 ([updated March 2011])
        Version: Version 5.1.0
        • Wells G.A.
        • Shea B.
        • O'connell D.
        • Peterson J.
        • Welch V.
        • Losos M.
        • et al.
        The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses.
        Ottawa Hospital Research Institute, 2014
        • Russo P.L.
        • Cheng A.C.
        • Mitchell B.G.
        • Hall L.
        Healthcare-associated infections in Australia: tackling the known unknowns.
        Aust Health Rev. 2017 Mar 7; ([Epub ahead of print])https://doi.org/10.1071/AH16223
        • Russo P.L.
        • Barnett A.G.
        • Cheng A.C.
        • Richards M.
        • Graves N.
        • Hall L.
        Differences in identifying healthcare associated infections using clinical vignettes and the influence of respondent characteristics: a cross-sectional survey of Australian infection prevention staff.
        Antimicrob Resist Infect Control. 2015; 4: 29
        • Mitchell B.G.
        • Gardner A.
        A model for influences on reliable and valid health care-associated infection data.
        Am J Infect Control. 2014; 42: 190-192
        • Australian Bureau of Statistics
        Australian demographic statistics, Jun 2016.
        2016
        • Australian Institute of Health and Welfare
        Australian hospital statistics 2012–13.
        in: Australian Institute of Health and Welfare. Australian Government, Canberra2014
        • Australian Institute of Health and Welfare
        Elective surgery waiting times 2014–15.
        in: Australian Institute of health and Welfare. Australian Government, Canberra2015
        • Australian Institute of Health and Welfare
        Admitted patient care 2014-15.
        in: Australian Institute of health and Welfare. Australian Government, Canberra2016
        • Australian Commission on Safety and Quality in Health Care
        Australian atlas of healthcare variation.
        Australian Commission on Safety and Quality in Health Care, Canberra2015
        • Si D.
        • Runnegar N.
        • Marquess J.
        • Rajmokan M.
        • Playford E.G.
        Characterising health care-associated bloodstream infections in public hospitals in Queensland, 2008–2012.
        Med J Aust. 2016; 204: 276
        • Mitchell B.G.
        • Collignon P.J.
        • McCann R.
        • Wilkinson I.J.
        • Wells A.
        A major reduction in hospital-onset Staphylococcus aureus bacteremia in Australia-12 years of progress: an observational study.
        Clin Infect Dis: Off Publ Infect Dis Soc Am. 2014; 59: 969-975
        • Alcorn K.
        • Gerrard J.
        • Macbeth D.
        • Steele M.
        Seasonal variation in health care-associated bloodstream infection: increase in the incidence of gram-negative bacteremia in nonhospitalized patients during summer.
        Am J Infect Control. 2013; 41: 1205-1208
        • Anderson E.
        • Leahy O.
        • Cheng A.C.
        • Grummet J.
        Risk factors for infection following prostate biopsy – a case control study.
        BMC Infect Dis. 2015; 15: 580
        • Aubron C.
        • Suzuki S.
        • Glassford N.J.
        • Garcia-Alvarez M.
        • Howden B.P.
        • Bellomo R.
        The epidemiology of bacteriuria and candiduria in critically ill patients.
        Epidemiol Infect. 2015; 143: 653-662
        • Wong S.W.
        • Gantner D.
        • McGloughlin S.
        • Leong T.
        • Worth L.J.
        • Klintworth G.
        • et al.
        The influence of intensive care unit-acquired central line-associated bloodstream infection on in-hospital mortality: a single-center risk-adjusted analysis.
        Am J Infect Control. 2016; 44: 587-592
        • Coombs G.W.
        • Nimmo G.R.
        • Pearson J.C.
        • Collignon P.J.
        • Bell J.M.
        • McLaws M.L.
        • et al.
        Australian group on antimicrobial resistance hospital-onset Staphylococcus aureus surveillance programme annual report, 2011.
        Commun Dis Intell Q Rep. 2013; 37: E210-E218
        • Chandrananth J.
        • Rabinovich A.
        • Karahalios A.
        • Guy S.
        • Tran P.
        Impact of adherence to local antibiotic prophylaxis guidelines on infection outcome after total hip or knee arthroplasty.
        J Hosp Infect. 2016; 93: 423-427
        • Tao P.
        • Marshall C.
        • Bucknill A.
        Surgical site infection in orthopaedic surgery: an audit of peri-operative practice at a tertiary centre.
        Healthc Infect. 2015; 20: 39-45
        • Worth L.J.
        • Bull A.L.
        • Spelman T.
        • Brett J.
        • Richards M.J.
        Diminishing surgical site infections in Australia: time trends in infection rates, pathogens and antimicrobial resistance using a comprehensive Victorian surveillance program, 2002–2013.
        Infect Control Hosp Epidemiol. 2015; 36: 409-416
        • Roth H.
        • Millar J.L.
        • Cheng A.C.
        • Byrne A.
        • Evans S.
        • Grummet J.
        The state of TRUS biopsy sepsis: readmissions to Victorian hospitals with TRUS biopsy-related infection over 5 years.
        BJU Int. 2015; 116: 49-53
        • Brogan E.
        • Langdon C.
        • Brookes K.
        • Budgeon C.
        • Blacker D.
        Respiratory infections in acute stroke: nasogastric tubes and immobility are stronger predictors than dysphagia.
        Dysphagia. 2014; 29 (0179051X): 340-345
        • Macesic N.
        • Kotsimbos T.C.
        • Kelly P.
        • Cheng A.C.
        Hospital-acquired influenza in an Australian sentinel surveillance system.
        Med J Aust. 2013; 198: 370-372
        • Sanagou M.
        • Leder K.
        • Cheng A.C.
        • Pilcher D.
        • Reid C.M.
        • Wolfe R.
        Associations of hospital characteristics with nosocomial pneumonia after cardiac surgery can impact on standardized infection rates.
        Epidemiol Infect. 2016; 144: 1065-1074
        • Foster N.F.
        • Collins D.A.
        • Ditchburn S.L.
        • Duncan C.N.
        • Schalkwyk J.W.
        • Golledge C.L.
        • et al.
        Epidemiology of Clostridium difficile infection in two tertiary-care hospitals in Perth, Western Australia: a cross-sectional study.
        New Microbes New Infect. 2014; 2: 64-71
        • Slimings C.
        • Armstrong P.
        • Beckingham W.D.
        • Bull A.L.
        • Hall L.
        • Kennedy K.J.
        • et al.
        Increasing incidence of Clostridium difficile infection, Australia, 2011–2012.
        Med J Aust. 2014; 200: 272-276
        • Worth L.J.
        • Spelman T.
        • Bull A.L.
        • Brett J.A.
        • Richards M.J.
        Epidemiology of Clostridium difficile infections in Australia: enhanced surveillance to evaluate time trends and severity of illness in Victoria, 2010–2014.
        J Hosp Infect. 2016; 93: 280-285
        • Mitchell B.G.
        • Ferguson J.K.
        • Anderson M.
        • Sear J.
        • Barnett A.
        Length of stay and mortality associated with healthcare-associated urinary tract infections: a multi-state model.
        J Hosp Infect. 2016; 93: 92-99
        • Zarb P.
        • Coignard B.
        • Griskeviciene J.
        • Muller A.
        • Vankerckhoven V.
        • Weist K.
        • et al.
        The European Centre for Disease Prevention and Control (ECDC) pilot point prevalence survey of healthcare-associated infections and antimicrobial use.
        Euro Surveill. 2012; 17: 20316
        • Gravel D.
        • Taylor G.
        • Ofner M.
        • Johnston L.
        • Loeb M.
        • Roth V.
        • et al.
        • The Canadian Nosocomial Infection Surveillance Program
        Point prevalence survey for healthcare-associated infections within Canadian adult acute-care hospitals.
        J Hosp Infect. 2007; 66: 243-248
        • Australian Institute of Health and Welfare
        Staphylococcus aureus bacteraemia in Australian public hospitals 2015–16: Australian hospital statistics. Health services series no. 74. Cat. No. HSE 184.
        Australian Institute of Health and Welfare, Canberra2017
        • Jones C.A.
        • Davis J.S.
        • Looke D.F.
        Death from an untreatable infection may signal the start of the post-antibiotic era.
        Med J Aust. 2017; 206: 292-293
        • Australian Medical Association
        Australian national centre for Disease Control (CDC) – 2017.
        Australian Medical Association, 2017
        • Mumford V.
        • Reeve R.
        • Greenfield D.
        • Forde K.
        • Westbrook J.
        • Braithwaite J.
        Is accreditation linked to hospital infection rates? A 4-year, data linkage study of Staphylococcus aureus rates and accreditation scores in 77 Australian acute hospitals.
        Int J Qual Health Care: J Int Soc Qual Health Care. 2015; 27: 479-485
        • Phan K.
        • Schultz K.
        • Huang C.
        • Halcrow S.
        • Fuller J.
        • McDowell D.
        • et al.
        External ventricular drain infections at the Canberra Hospital: a retrospective study.
        J Clin Neurosci: Off J Neurosurg Soc Australas. 2016; 32: 95-98
        • Ezzatzadegan S.
        • Chen S.
        • Chapman J.R.
        Invasive fungal infections after renal transplantation.
        Int J Organ Transplant Med. 2012; 3: 18-25
        • Gautam A.
        • Ganu S.S.
        • Tegg O.J.
        • Andresen D.N.
        • Wilkins B.H.
        • Schell D.N.
        Ventilator-associated pneumonia in a tertiary paediatric intensive care unit: a 1-year prospective observational study.
        Crit Care Resusc: J Australas Acad Crit Care Med. 2012; 14: 283-289
        • Mathot F.
        • Duke T.
        • Daley A.J.
        • Butcher T.
        Bacteremia and pneumonia in a tertiary PICU: an 11-year study.
        Pediatr Crit Care Med: J Soc Crit Care Med World Fed Pediatr Intensive Crit Care Soc. 2015; 16: 104-113
        • Jarratt L.S.
        • Miller E.R.
        The relationship between patient characteristics and the development of a multi-resistant healthcare-associated infection in a private South Australian hospital.
        Infect Dis Health. 2013; 18: 94-101