Attributable Mortality and Excess Length of Stay associated with Third-Generation Cephalosporin Resistant Enterobacterales Bloodstream Infections - a prospective cohort study in Suva, Fiji

OBJECTIVES

There are scant primary clinical data on antimicrobial resistance (AMR) burden from low- and middle-income countries (LMICs). We adapted recent World Health Organization methodology to measure the impact of third-generation cephalosporin resistance (3GC-R) on mortality and excess length of hospital stay in Fiji.

METHODS

We conducted a prospective cohort study of inpatients with Enterobacterales bloodstream infections (BSIs) at Colonial War Memorial Hospital, Suva. We used cause-specific Cox proportional hazards models to estimate the effect of 3GC-R on the daily risk (hazard) of in-hospital mortality and being discharged alive (competing risks), and multistate modelling to estimate the excess length of hospital stay.

RESULTS

From July 2020 to February 2021 we identified 162 consecutive Enterobacterales BSIs, 3GC-R was present in 66 (40.7%). Crude mortality for patients with 3GC-susceptible and 3GC-R BSIs was 16.7% (16/96) and 30.3% (20/66), respectively. 3GC-R was not associated with the in-hospital mortality hazard rate (adjusted hazard ratio (aHR) 1.13, 95% CI 0.51-2.53) or being discharged alive (aHR 0.99, 95% CI 0.65-1.50), whereas Charlson comorbidity index score (aHR 1.62, 95% CI 1.36-1.93) and Pitt bacteraemia score (aHR 3.57, 95% CI 1.31-9.71) were both associated with an increased hazard rate of in-hospital mortality. 3GC-R was associated with an increased length of stay of 2.6 days (95% CI 2.5-2.8). 3GC-R was more common among hospital-associated infections, but genomics did not identify clonal transmission.

CONCLUSION

Patients with Enterobacterales BSIs in Fiji had high mortality. There were high rates of 3GC-R, which was associated with increased hospital length of stay but not with in-hospital mortality.

The impact of 10-valent pneumococcal vaccine introduction on invasive disease in Fiji

BACKGROUND

In 2012, Fiji introduced the 10-valent pneumococcal conjugate vaccine (PCV10). We assessed the impact of PCV10 on invasive pneumococcal disease (IPD), probable bacterial or pneumococcal meningitis (PBPM), meningitis and sepsis 3-5 years post-introduction.

METHODS

Laboratory-confirmed IPD and PBPM cases were extracted from national laboratory records. ICD-10-AM coded all-cause meningitis and sepsis cases were extracted from national hospitalisation records. Incidence rate ratios were used to compare outcomes pre/post-PCV10, stratified by age groups: 1-23m, 2-4y, 5-9y, 10-19y, 20-54y, ≥55y. To account for different detection and serotyping methods in the pre-and post-PCV10 period, a Bayesian inference model estimated serotype-specific changes in IPD, using pneumococcal carriage and surveillance data.

FINDINGS

There were 423 IPD, 1,029 PBPM, 1,391 all-cause meningitis and 7,611 all-cause sepsis cases. Five years post-PCV10 introduction, IPD declined by 60% (95%CI: 37%, 76%) in children 1-23m months old, and in age groups 2-4y, 5-9y, 10-19y although confidence intervals spanned zero. PBPM declined by 36% (95%CI: 21%, 48%) among children 1-23 months old, and in all other age groups, although some confidence intervals spanned zero. Among children <5y of age, PCV10-type IPD declined by 83% (95%CI; 70%, 90%) and with no evidence of change in non-PCV10-type IPD (9%, 95%CI; -69, 43%). There was no change in all-cause meningitis or sepsis. Post-PCV10, the most common serotypes in vaccine age-eligible and non-age eligible people were serotypes 8 and 23B, and 3 and 7F, respectively.

INTERPRETATIONS

Our study demonstrates the effectiveness of PCV10 against IPD in a country in the Asia-Pacific of which there is a paucity of data.

FUNDING

This study was support by the Department of Foreign Affairs and Trade of the Australian Government and Fiji Health Sector Support Program (FHSSP). FHSSP is implemented by Abt JTA on behalf of the Australian Government.

Prevalence of healthcare-associated infections and antimicrobial use among inpatients in a tertiary hospital in Fiji: a point prevalence survey

Background

Healthcare-associated infections (HAIs) and antimicrobial use (AMU) are important drivers of antimicrobial resistance, yet there is minimal data from the Pacific region. We sought to determine the point prevalence of HAIs and AMU at Fiji’s largest hospital, the Colonial War Memorial Hospital (CWMH) in Suva. A secondary aim was to evaluate the performance of European Centre for Diseases Prevention and Control (ECDC) HAI criteria in a resource-limited setting.

Methods

We conducted a point prevalence survey of HAIs and AMU at CWMH in October 2019. Survey methodology was adapted from the ECDC protocol. To evaluate the suitability of ECDC HAI criteria in our setting, we augmented the survey to identify patients with a clinician diagnosis of a HAI where diagnostic testing criteria were not met. We also assessed infection prevention and control (IPC) infrastructure on each ward.

Results

We surveyed 343 patients, with median (interquartile range) age 30 years (16–53), predominantly admitted under obstetrics/gynaecology (94, 27.4%) or paediatrics (83, 24.2%). Thirty patients had one or more HAIs, a point prevalence of 8.7% (95% CI 6.0% to 12.3%). The most common HAIs were surgical site infections (n = 13), skin and soft tissue infections (7) and neonatal clinical sepsis (6). Two additional patients were identified with physician-diagnosed HAIs that failed to meet ECDC criteria due to insufficient investigations. 206 (60.1%) patients were receiving at least one antimicrobial. Of the 325 antimicrobial prescriptions, the most common agents were ampicillin (58/325, 17.8%), cloxacillin (55/325, 16.9%) and metronidazole (53/325, 16.3%). Use of broad-spectrum agents such as piperacillin/tazobactam (n = 6) and meropenem (1) was low. The majority of prescriptions for surgical prophylaxis were for more than 1 day (45/76, 59.2%). Although the number of handwashing basins throughout the hospital exceeded World Health Organization recommendations, availability of alcohol-based handrub was limited and most concentrated within high-risk wards.

Conclusions

The prevalence of HAIs in Fiji was similar to neighbouring high-income countries, but may have been reduced by the high proportion of paediatric and obstetrics patients, or by lower rates of inpatient investigations. AMU was very high, with duration of surgical prophylaxis an important target for future antimicrobial stewardship initiatives.

Short communication: Epidemiological assessment of Strongyloides stercoralis in Fijian children

As a part of the lymphatic filariasis (LF) transmission assessment survey (TAS)/soil-transmitted helminths (STH) prevalence survey in Western Division of Fiji, a pilot screen for Strongyloides stercoralis (SS) in school children was undertaken using a combination of the Baermann concentration (BC) method and real-time PCR assays.

Using BC, faecal samples collected from 111 children of 7 schools were examined. A single child was positive for larvae of SS and underwent a clinical examination finding an asymptomatic infection. Other members of this child's household were screened with BC, finding none infected. Aliquots of 173 faecal samples preserved in ethanol originating from all schools were examined by real-time PCR, and the prevalence of SS infection was 3.5%.

Our study confirms the existence of SS infection on Fiji and showed that assessing SS prevalence alongside TAS/STH survey is a convenient access platform, allowing introduction of other surveillance techniques such as BC and real-time PCR.