Abstract
Background
Little is known about diagnostic and antibiotic use practices in low and middle-income countries (LMICs) before and during COVID-19 pandemic. This information is crucial for monitoring and evaluation of diagnostic and antimicrobial stewardships in healthcare facilities.
Methods
We linked and analyzed routine databases of hospital admission, microbiology laboratory and drug dispensing of Indonesian National Referral Hospital from 2019 to 2020. Patients were classified as COVID-19 cases if their SARS-CoV-2 RT-PCR result were positive. Blood culture (BC) practices and time to discontinuation of parenteral antibiotics among inpatients who received a parenteral antibiotic for at least four consecutive days were used to assess diagnostic and antibiotic use practices, respectively. Fine and Grey subdistribution hazard model was used.
Results
Of 1,311 COVID-19 and 58,917 non-COVID-19 inpatients, 333 (25.4%) and 18,837 (32.0%) received a parenteral antibiotic for at least four consecutive days. Proportion of patients having BC taken within ±1 calendar day of parenteral antibiotics being started was higher in COVID-19 than in non-COVID-19 patients (21.0% [70/333] vs. 18.7% [3,529/18,837]; p<0.001). Cumulative incidence of having a BC taken within 28 days was higher in COVID-19 than in non-COVID-19 patients (44.7% [149/333] vs. 33.2% [6,254/18,837]; adjusted subdistribution-hazard ratio [aSHR] 1.71, 95% confidence interval [CI] 1.47–1.99, p<0.001). The median time to discontinuation of parenteral antibiotics was longer in COVID-19 than in non-COVID-19 patients (13 days vs. 8 days; aSHR 0.73, 95%Cl 0.65–0.83, p<0.001).
Conclusions
Routine electronic data could be used to inform diagnostic and antibiotic use practices in LMICs. In Indonesia, the proportion of timely blood culture is low in both COVID-19 and non-COVID-19 patients, and duration of parenteral antibiotics is longer in COVID-19 patients. Improving diagnostic and antimicrobial stewardship is critically needed.
Introduction
Antimicrobial-resistant (AMR) bacterial infection is a global threat to public health [1]. Low and middle-income countries (LMICs), including Indonesia, are considered hotspots of AMR, driven by the lack of laboratory support for infectious disease diagnosis and high levels of inappropriate antibiotic use in hospitals [2, 3].
Diagnostic and antimicrobial stewardship (AMS) are increasingly implemented in LMICs [4–6]. Although processes of the stewardship programmes (e.g., presence of standard operating procedures [SOPs] and how many patient medical charts are reviewed) are frequently monitored and reported, outcomes of those stewardship programmes (e.g. increase in specimens submitted for bacterial culture according to SOPs and defined daily dose [DDDs] per 1,000 patient-days) or patient outcomes (e.g. in-hospital mortality) are rarely monitored and reported. This is because measuring and analyzing diagnostic and antibiotic use practices overtime are labor-intensive and prohibitively costly.
In our recent report we found that the etiology of bloodstream infections (BSI) and proportion of AMR-BSI were similar between COVID-19 and non-COVID-19 patients hospitalized at the Indonesian national referral hospital, Jakarta, Indonesia, from 1 January 2019 to 31 December 2020, and that reported incidence rates of hospital-origin AMR-BSI increased in 2020, which was likely attributable to increased blood culture utilization [7]. Here, we aimed to analyze drug dispensing data, and evaluate diagnostic and antibiotic use practices in the Indonesian national referral hospital before and during the COVID-19 pandemic.
Methods
Study design, setting and population
We conducted a retrospective hospital-wide longitudinal surveillance study using routine electronic data on all patients hospitalized at Cipto Mangunkusumo National Hospital (CMNH), the national referral hospital in Jakarta, Indonesia, from 1 January 2019 to 31 December 2020.
Data collection
We used the routine hospital databases containing hospital admission, laboratory microbiology and drug dispensing data. The hospital admission data collected included medical record number (MRN), sex, age, admission and discharge date. The microbiology laboratory data collected included MRN, admission date, specimen type, specimen collection date, culture result using conventional bacterial identification methods and Vitek®2 (bioMerieux, Inc. Durham, USA), antibiotic susceptibility profile according to Clinical and Laboratory Standards Institute guidelines, and result report date. The drug dispensing data collected included MRN, admission date, drug name, route of drug administration, the dosage regimen, drug start date and stop date. The MRN and admission date were used to link the three databases together. Data were accessed since 1 March 2021.
Definitions
Severe infection
Patients with severe infection were defined using the criteria modified from the United States Centers for Disease Control and Prevention [8]. Patients who received a parenteral antibiotic for at least four consecutive days was used as a surrogate for severe infection, with the first calendar day equal to the start date of parenteral antibiotics. Patients who died, were discharged to a hospice or transferred to other hospital before completing four consecutive days of parenteral antibiotics and had parenteral antibiotics continuously until the day prior to death, hospice discharge or transfer were also included as patients with severe infection [9].
Diagnostic practices
To measure diagnostic practices, we estimated (a) the proportion of patients with blood culture (BC) taken within ±1 calendar day of parenteral antibiotics being started, (b) the proportion of patients with BC taken within 28 days of parenteral antibiotics being started, and (c) the median BC turnaround-time [4, 5, 9]. The proportion of patients having BC taken within ±1 calendar day of parenteral antibiotics being started was used to measure the compliance with the recommendations that blood specimens should be taken before starting antimicrobial therapy in patients with suspected sepsis and septic shock if no substantial delay in the start of antimicrobials occurs (i.e. <45 min) and that if the antibiotic is administered first, the blood specimens should be collected for culture within 24 hours [10]. The proportion of patients with BCs taken within 28 days was defined as the ratio of the number of patients with severe infection where BCs were taken from one calendar day before to 28 calendar days after starting parenteral antibiotics, and the total number of patients with severe infection. BC turn-around time was defined as the time interval between BC taken from patients and BC results positive for organisms reported by laboratory [4].
Antibiotic use practices
To measure antibiotic use practices, we estimated (a) the proportion of antibiotic use based on the AWaRe system (Access, Watch and Reserve) [5], (b) median time to discontinuation of parenteral antibiotics [5], (c) the proportion of patients with narrow-spectrum parenteral antibiotics (antibiotics within the Access and Watch category without anti-methicillin-resistant Staphylococcus aureus [MRSA] and antipseudomonal activity) within 2 calendar days following BC report among those who had a BC positive for 3rd-generation cephalosporin-sensitive Escherichia coli (3GCSEC) or Klebsiella pneumoniae (3GCSKP) and remained on a parenteral antibiotic on the day that the BC results were reported [11], and (d) the proportion of patients with parenteral antibiotic discontinuation within 2 calendar days following BC report among those who had BC results negative for organisms and remained on a parenteral antibiotic on the day that the BC results were reported [12].
Antibiotic use among inpatients was measured in terms of DDD per 1,000 inpatient-days. DDD was defined as the assumed average maintenance dose per day for a drug used for its main indication in adults as specified by World Health Organization (WHO) [13]. Pharmacological subgroups of antibacterials for systemic use was defined according to group based on the Anatomical Therapeutic Chemical classification system, including tetracyclines (J01A), amphenicols (J01B), beta-lactam antibacterials, penicillins (J01C), other beta-lactam antibacterials (J01D), sulfonamides and trimethoprim (J01E), macrolides, lincosamides and streptogramins (J01F), aminoglycoside antibacterials (J01G), quinolone antibacterials (J01M), combinations of antibacterials (J01R), other antibacterials (J01X), agents against amoebiasis and other protozoal diseases (P01A).
We also assessed the change of parenteral antibiotic among patients with severe infection within 28 days. Parenteral antibiotic use on each calendar day were categorized as escalation, de-escalation, change to other antibiotics with similar spectrum, no change from the previous antibiotic or stop. Antibiotic escalation was defined as (a) adding a new parenteral antibiotic, (b) changing parenteral antibiotic with an increase in AWaRe categories, or (c) changing parenteral antibiotic within the Watch category but from those without anti-MRSA and antipseudomonal activity to those with anti-MRSA activity (e.g. vancomycin) or antipseudomonal activity (e.g. antipseudomonal cephalosporin, antipseudomonal penicillin and carbapenems) compared to the previous day. De-escalation was defined as the inverse of these. Change to other antibiotics with similar spectrum was defined as changing parenteral antibiotics to other parenteral antibiotics within similar AWaRe categories, and with similar anti-MRSA activity and antipseudomonal activity for parenteral antibiotics within the Watch category compared to the previous day. No change was defined as continuation of current parenteral antibiotics, while stop was defined as discontinuation of parenteral antibiotics [12, 14–17]. We also evaluated initial parenteral antibiotics. Parenteral antibiotics being prescribed within the first calendar day that a parenteral antibiotic was started were regarded as initial parenteral antibiotics [18, 19].
Ethics
The study was approved by the Faculty of Medicine Universitas Indonesia Ethics Committee (KET-115/UN2.F1/ETIK/PPM.00.02/2021) and Oxford Tropical Research Ethics Committee (503–22). The requirement for individual patient consent was waived. Additional permission was obtained from the Innovation and Intellectual Property Directorate CMNH, to use the routine hospital databases for research.
Data analysis
Pearson’s chi-squared test or Fisher’s Exact test was used to compare categorical variables between groups when appropriate. Mann-Whitney test was used to compare continuous variables between groups. We compared time to mortality, having BC taken and discontinuation of parenteral antibiotics within 28 days between groups by using Fine and Grey models to account for competing risks [20, 21]. Mortality included in-hospital mortality and discharge to a hospice. Discharge to home was considered a competing risk for mortality. Mortality, discharge to home and discontinuation of parenteral antibiotics were considered a competing risk for having BC taken. For patients who had parenteral antibiotics up to the calendar date that they were discharged home, all parenteral antibiotics were considered discontinued on the discharge date. Mortality was considered a competing risk for discontinuation of parenteral antibiotics. All data analyses were performed using the STATA version 15.1 (StataCorp, College Station, TX, USA). We visualized data using the GraphPad Prism version 8.3.0 (La Jolla, California, USA) and SankeyMATIC (https://sankeymatic.com/).
Results
Baseline characteristics
Of 91,960 admissions (from 60,228 patients) admitted from 1 Jan 2019 to 31 Dec 2020, 35,766 (38.9%) admissions received at least one antibiotic (Fig 1 and S1 Table). Total consumption of antibiotics among inpatients was 624.1 DDD per 1,000 inpatient-days, of which 77.6% (484.3 DDD per 1,000 inpatient-days) and 22.3% (139.8 DDD per 1,000 inpatient-days) were parenteral and oral antibiotics, respectively. Overall, 29.3% (182.4 DDD per 1,000 inpatient-days), 68.7% (429.3 DDD per 1,000 inpatient-days) and 2.0% (12.4 DDD per 1,000 inpatient-days) were classified as Access, Watch and Reserve categories, respectively (S1 Fig).
Fig 1. Flow diagram.
Of all parenteral antibiotic consumption among inpatients (484.3 DDD per 1,000 inpatient-days), 46.9% (227.1 DDD per 1,000 inpatient-days) were other beta-lactam antibacterials (J01D), 20.2% (97.8 DDD per 1,000 inpatient-days) were quinolone antibacterials (J01M), and 15.4% (74.6 DDD per 1,000 inpatient-days) were penicillins (J01C) (S2 Fig). Of all other beta-lactam antibacterials (J01D) consumption (227.1 DDD per 1,000 inpatient-days), 65.0% (147.6 DDD per 1,000 inpatient-days) were 3rd-generation cephalosporins (J01DD), 21.7% (49.3 DDD per 1,000 inpatient-days) were carbapenems (J01DH), and 8.0% (18.2 DDD per 1,000 inpatient-days) were 4th-generation cephalosporins (J01DE).
Of 91,960 admissions (60,228 patients), 1,373 (from 1,311 patients) were COVID-19 cases. Total antibiotic consumption was 1,502.8 and 606.3 DDD per 1,000 inpatient-days in COVID-19 and non-COVID-19 patients, respectively (S1 Table).
Of 60,228 patients, 19,170 (31.8%) received at least four consecutive days of parenteral antibiotics, were classified as patients with severe infection, and the first admission per patient which fulfilled the criteria of severe infection were included for further analysis.
Patients with severe infection
Among 19,170 patients with severe infection, 333 (1.7%) were COVID-19 patients and 18,837 (98.3%) were non-COVID-19 patients (Table 1). The median age was higher in COVID-19 patients than non-COVID-19 patients (47 vs. 39 years, p<0.001). The sex distribution was not different between the two groups (p = 0.23).
Table 1. Characteristics of 19,170 patients with severe infection.
| Parameters | COVID-19 patients with severe infection (n = 333) | Non-COVID-19 patients with severe infection (n = 18,837) | P values |
|---|---|---|---|
| Median age (years old, IQR) | 47 (IQR 25–62, range 0–93) | 39 (IQR 15–56, range 0–95) | <0.001 |
| Sex | |||
| Female | 156 (46.8%) | 9,445 (50.1%) | 0.23 |
| Male | 177 (53.2%) | 9,392 (49.9%) | |
| Median duration of hospital stay (days) | 8 (IQR 4–16, range 1–90) | 8 (IQR 4–15, range 1–119) | 0.08 |
| Time when a parenteral antibiotic was started a | |||
| 1–2 calendar days after admission | 207 (62.2%) | 13,137 (69.7%) | <0.001 |
| 3–7 calendar days after admission | 97 (29.1%) | 3,840 (20.4%) | |
| 7 calendar days after admission | 29 (8.7%) | 1,860 (9.9%) | |
| Number of parenteral antibiotics received on the day that a parenteral antibiotic was started (%) | |||
| 1 | 255 (76.6%) | 14,091 (74.8%) | 0.66 |
| 2 | 72 (21.6%) | 4,451 (23.6%) | |
| ≥3 | 6 (1.8%) | 295 (1.6%) | |
| Total antibiotic consumption (DDD per 1,000 patient-days) b | 1656.4 | 807.6 | |
| Access category | 153.9 (9.3%) | 263.8 (32.7%) | <0.001 |
| Watch category | 1427.3 (86.2%) | 520.3 (64.4%) | |
| Reserve category | 75.2 (4.5%) | 23.5 (2.9%) | |
| 28-day mortality c | 95 (28.5%) | 4,485 (23.8%) | 0.02 |
Only the first admission per patient which fulfilled the criteria of severe infection were included. Patients with severe infection were defined as those who received a parenteral antibiotic for at least four consecutive days. Patients who died, were discharged to a hospice or transferred to other hospital before completing four consecutive days of parenteral antibiotics and had parenteral antibiotics continuously until the day prior to death, hospice discharge or transfer were also included as patients with severe infection
a Time was measured when a parenteral antibiotic was started and continued for at least four consecutive days. Staring parenteral antibiotics on the day of admission was equal to 1 calendar day after admission.
b Total antibiotic consumption was estimated from the day of parenteral antibiotics being started to the discharge date.
c 28-day mortality included in-hospital mortality and discharged to a hospice within 28 days of parenteral antibiotics being started.
The proportion of parenteral antibiotics consumption under Watch and Reserve category was significantly higher in COVID-19 patients than non-COVID-19 patients (86.2% vs. 64.4% and 4.5% vs. 2.9%, respectively; p<0.001; Table 1). The proportion of patients being prescribed one, two or at least three initial parenteral antibiotics within the first calendar days of parenteral antibiotics being started was not different between two groups (p = 0.66) (Table 1). Ceftriaxone (36.9% [123/333]), levofloxacin (18.9% [63/333]) and meropenem (9.3% [31/333]) were commonly used as the initial antibiotics in COVID-19 patients, while ceftriaxone (30.9% [582/18,837]), ampicillin/sulbactam (8.5% [160/18,837]) and metronidazole (7.8% [147/18,837]) were commonly prescribed as the initial antibiotics in non-COVID-19 patients (S3 Fig).
The 28-day mortality among 19,170 patients with severe infection was 23.9% (Table 1 and S2 Table). Using Fine and Grey model, we found that the risk of mortality was higher in COVID-19 patients than non-COVID-19 patients (28.5% vs. 23.8%, adjusted subdistribution-hazard ratios [aSHR] 1.29; 95%CI 1.04–1.58, p = 0.02; Table 2 and S4A Fig).
Table 2. Multivariable analyses for time to mortality, having blood culture sampled and discontinuation of parenteral antibiotics among 19,170 patients with severe infection.
| Variables | Mortality a | Having blood culture sampled b | Discontinuation of parenteral antibiotics c | |||
|---|---|---|---|---|---|---|
| Adjusted SHR | P values | Adjusted SHR | P values | Adjusted SHR | P values | |
| COVID-19 status | ||||||
| Non-COVID-19 | Reference | Reference | Reference | |||
| COVID-19 | 1.29 (1.04–1.58) | 0.02 | 1.71 (1.47–1.99) | <0.001 | 0.73 (0.65–0.83) | <0.001 |
| Sex | ||||||
| Female | Reference | Reference | Reference | |||
| Male | 1.02 (0.97–1.09) | 0.38 | 1.04 (0.99–1.09) | <0.001 | 0.92 (0.89–0.94) | <0.001 |
| Age (years old) | ||||||
| <1 | 1.09 (0.93–1.28) | <0.001 | 7.54 (6.73–8.45) | <0.001 | 0.84 (0.79–0.89) | 0.72 |
| 1–4 | 1.01 (0.84–1.20) | 4.54 (4.01–5.14) | 0.88 (0.82–0.95) | |||
| 5–14 | 1.07 (0.92–1.25) | 3.98 (3.52–4.51) | 0.94 (0.87–1.00) | |||
| 15–24 | 1.26 (1.07–1.48) | 1.90 (1.65–2.17) | 0.88 (0.82–0.94) | |||
| 25–34 | Reference | Reference | Reference | |||
| 35–44 | 1.24 (1.07–1.41) | 1.25 (1.09–1.43) | 0.99 (0.93–1.05) | |||
| 45–54 | 1.23 (1.07–1.41) | 1.51 (1.33–1.71) | 0.94 (0.89–1.00) | |||
| 55–64 | 1.22 (1.06–1.40) | 1.51 (1.33–1.72) | 0.90 (0.85–0.95) | |||
| ≥65 | 1.30 (1.13–1.49) | 1.44 (1.26–1.64) | 0.87 (0.82–0.92) | |||
| Admission year | ||||||
| 2019 | Reference | Reference | Reference | |||
| 2020 | 0.96 (0.91–1.02) | 0.18 | 1.06 (1.01–1.11) | 0.02 | 0.94 (0.92–0.97) | <0.001 |
SHR = sub-distribution hazard ratio. Fine and Grey sub-distribution hazard models were used.
a Mortality included in-hospital mortality and discharged to a hospice. Discharged home from the hospital was considered as a competing risk for mortality.
b In-hospital mortality, discharged to a hospice and discontinuation of parenteral antibiotics were considered as a competing risk for having blood culture sampled. Blood culture sampled on -1 calendar day when a parenteral antibiotic was started was included in the model as sampled on the calendar day when a parenteral antibiotic was started.
c In-hospital mortality and discharged to a hospice were considered as a competing risk for discontinuation of parenteral antibiotics.
Blood culture practices
Of 19,170 patients with severe infection, 3,599 (18.8%) had BC taken within ±1 calendar day of parenteral antibiotics being started (Table 3 and S5 Fig). The proportion was higher in COVID-19 than in non-COVID-19 patients (21.0% [70/333] vs. 18.7% [3,529/18,837]; p<0.001).
Table 3. Parameters for clinical diagnostic practices and antibiotic use practices among 19,170 patients with severe infection.
| Parameters | COVID-19 patients with severe infection (n = 333) | Non-COVID-19 patients with severe infection (n = 18,837) | P values |
|---|---|---|---|
| Parameters for clinical diagnostic practices | |||
| Proportion of patients with BC taken within ±1 calendar day of parenteral antibiotic being started | 21.0% (70/333) | 18.7% (3,529/18,837) | <0.001 |
| Proportion of patients with BC taken from calendar day 3 to day 28 of parenteral antibiotic being started | 31.8% (106/333) | 21.3% (4,012/18,837) | <0.001 |
| Proportion of patients with at least one BC taken within 28 calendar days of parenteral antibiotic being started | 44.7% (149/333) | 33.2% (6,254/18,837) | <0.001 |
| Median BC turn-around time (days, IQR, n) | 4 (4–5, n = 34) | 4 (4–5, n = 1,593) | 0.69 |
| Parameters for antibiotic use practices | |||
| Proportion of patients with narrow-spectrum parenteral antibiotic within 2 calendar days following BC report among those who had a BC positive for 3GCSEC or 3GCSKP and remained on a parenteral antibiotic on the day that BC results were reported | 0% (0/1) | 38.4% (20/52) | 0.43 |
| Proportion of patient with parenteral antibiotic discontinuation within 2 calendar following BC report among those who had a negative BC and remained on a parenteral antibiotic on the day that BC results were reported | 19.4% (7/36) | 26.8% (616/2,298) | 0.32 |
BC = blood culture, 3GCSEC = 3rd generation cephalosporin sensitive E. coli, 3GCSKP = 3rd generation cephalosporin sensitive K. pneumoniae, Narrow-spectrum parenteral antibiotics were defined as antibiotics within the Access and Watch category without anti-MRSA and antipseudomonal activity
The proportion of patients having BC taken within 28 days after parenteral antibiotics being started was 33.4%. Using Fine and Grey model, we found that having BC taken within 28 days was higher in COVID-19 than non-COVID-19 patients (44.7% vs. 33.2%; aSHR 1.71; 95%CI 1.47–1.99, p<0.001; Table 2) and in those admitted in 2020 than those admitted in 2019 (aSHR 1.06; 95%CI 1.01–1.11, p = 0.02). S4B Fig shows that, of COVID-19 and non-COVID-19 patients, 21.0% and 18.7% had the first BC taken on ±1 calendar day of parenteral antibiotics being started and the other 23.7% and 14.5% gradually had the first BC taken between calendar day 2 and 14 of parenteral antibiotics being started, respectively.
The median BC turn-around time was not different between the two groups (4 vs. 4 days; p = 0.69).
Antibiotic use practices
Fig 2 shows escalation, change, de-escalation and discontinuation of parenteral antibiotics compared to the parenteral antibiotics the patients received the day before. Of 19,170 patients with severe infection, 2,466 (12.9%) had parenteral antibiotics escalated, 3,267 (17.0%) changed to other antibiotics with similar spectrum, and 1,001 (5.2%) de-escalated at least once within 28 days after parenteral antibiotics being started. S6 Fig shows the proportion of the dispensed parenteral antibiotics categorized by AWaRe classification over 28 calendar days.
Fig 2.
Antibiotic stewardship among (A) 333 COVID-19 patients and (B) 18,837 non-COVID-19 patients with severe infection. “Antibiotic escalation” was defined as (a) adding a new antibiotic, (b) changing antibiotics to other antibiotics with an increase in Access, Watch and Reserve categories, or (c) changing antibiotics in the Watch category without anti-MRSA and antipseudomonal activity to other antibiotics within the Watch category with anti-MRSA activity (e.g. vancomycin) or antipseudomonal activity (e.g. antipseudomonal cephalosporin, antipseudomonal penicillin and carbapenems) compared to the parenteral antibiotics received on the day before. “Antibiotic de-escalation” was defined as the inverse of these. “Change to other antibiotics with similar spectrum” was defined as changing antibiotics to other antibiotics within similar Access, Watch and Reserve categories, and with similar anti-MRSA activity and antipseudomonal activity for antibiotics within the Watch category. Only parenteral antibiotics were included in the analysis.
Of 19,170 patients with severe infection, 72.7% had parenteral antibiotics discontinued within 28 days (S4C Fig). The median time to discontinuation of parenteral antibiotics was longer in COVID-19 than non-COVID-19 patients (13 days vs. 8 days; aSHR 0.64; 95%CI 0.55–0.75, p<0.001; Table 2 and S4C Fig), and in those admitted in 2020 than those admitted in 2019 (aSHR 0.94; 95%CI 0.92–0.97, p<0.001).
Of 53 patients with severe infection who had a BC positive for 3GCSEC or 3GCSKP and remained on a parenteral antibiotic on the day that BC results were reported, 20 (37.7%) had narrow-spectrum parenteral antibiotic within 2 calendar days following BC report. The proportion was not different between the two groups (p = 0.43). Of 2,344 patients with severe infection who had BC results negative for any organisms and remained on a parenteral antibiotic on the day that BC results were reported, 623 (26.6%) had parenteral antibiotics discontinued within 2 calendar days following BC report. This proportion was not different between the two groups (p = 0.32).
Discussion
This study demonstrates that the routine electronic hospital databases could be used to monitor and report outcomes of diagnostic and antimicrobial stewardship programmes in hospitals in LMICs. In the study hospital, we found that the proportion of timely BC was low in inpatients, and duration of parenteral antibiotics was longer in COVID-19 patients. Most of the parenteral antibiotics used for inpatients were from the Watch Category. After adjusting for COVID-19 patients, we found that the BC practice minimally improved in 2020 compared to 2019; however, the time to discontinuation of parenteral antibiotics was longer in 2020 compared to 2019. We shared these findings with the AMS committee of the hospital, highlighting the urgency of improving both diagnostic and antimicrobial stewardship programmes in the study hospital.
The proportion of patient with severe infection having BC sampled within ±1 calendar day of parenteral antibiotics being started was lower than those reported in other high-income countries [22, 23] and in Thailand [9]. Furthermore, we observed a high proportion of patients with severe infection having the first BC taken from calendar day 2 to 14 after the start of parenteral antibiotic administration. This shows that delayed BC is not uncommon in the study hospital, although there is an increasing trend of blood culture utilization during the study period [7]. This is of concern as underutilization and delayed BC can have a negative impact on individual patient management [4] and can overestimate proportions and underestimate incidence rates of AMR infections in the cumulative antibiograms [9, 24].
The better BC practice among COVID-19 patients compared to non-COVID-19 patients is encouraging. This could be because our hospital, which is the national referral hospital, manages COVID-19 patients with comorbidities and severe clinical presentations. Additionally, in COVID-19 patients, clinicians tend to consult infectious diseases physicians and sampled BC prior to parenteral antibiotic administration more frequently [7]. This improvement should be strengthened further until all patients have BC sampled within ±1 calendar day of parenteral antimicrobial therapy being started [25].
We observed a longer median BC turn-around time compared to those reported in high-income countries, i.e. 2–3 days [26–28]. The BC turn-around time is critical for gaining the maximum benefits of BC as a guidance for clinicians in deciding rational antibiotic prescriptions, reduce length of hospitalization and increase patient survival [4, 29–33]. Efforts to shorten BC turn-around time by improvement of internal laboratory workflows (e.g., work shifts and laboratory staffing) should be commenced [4, 28, 32, 34].
The longer median time to discontinuation of parenteral antibiotics was observed in inpatients compared to those recommended in the guidelines (i.e. 5–8 days) [10]. The longer time to discontinuation parenteral antibiotics and higher use of Watch and Reserve antibiotics in COVID-19 patients could be due to the fear of high mortality in the early COVID-19 pandemic, in addition to the clinical profile of patients managed in our hospital [35]. Previous studies reported an overuse of antibiotics amongst COVID-19 hospitalized patients. There might be a false belief of antibiotic use as medical prophylaxis to reduce secondary bacterial infections [36–39]. These highlight the need to improve clinician adherence to the well-developed evidence-based antibiotic guidelines [38, 39].
We observed that a de-escalation practice is uncommon in our study hospital using multiple parameters, similar to reports from other LMICs [40]. The common barrier to de-escalation following BC positive for monomicrobial narrow-spectrum antibiotic susceptible Enterobacterales could be the fear that patients may deteriorate if the broad-spectrum antibiotic is de-escalated [11]. The low proportion of antibiotic discontinuation following negative BC could reflect the physician lack of trust in negative results [14]. These data highlight the necessity to emphasize consideration of de-escalation following positive BC with drug sensitive pathogens and discontinuation following negative BC in patients with positive treatment response as an opportunity to reduce the overuse of broad-spectrum antibiotics [12, 14, 41].
Our study has several limitations. First, we could not exclude patients who had parenteral antibiotics for surgical prophylaxis for at least four consecutive days and analyze data for each clinical syndrome due to data limitation. Second, parameters used for evaluating practices are not free from bias, although they have been frequently used in previous studies. Third, use of routine data could not determine whether the practice in every single patient was appropriate or inappropriate. However, the analysis of routine data can represent the overall practice at the study hospital. Future studies could compare the routine data among hospitals of similar bed count and over time, or use the routine data to select specific sets of patients for further retrospective chart reviews. Lastly, the findings may not be generalizable to all hospitals in LMICs.
Conclusions
In the Indonesian referral hospital, the proportion of timely BC is low, and duration of parenteral antibiotics is long in both COVID-19 and non-COVID-19 patients. Improving diagnostic and antimicrobial stewardship is critically needed. We recommend hospitals in LMICs to perform routine monitoring and improvement of diagnostic and antimicrobial stewardship.
Supporting information
For this figure, antibiotics in the Watch category was divided to Watch and Watch+. Watch+ category comprises antibiotics in the Watch category with anti-MRSA activity (e.g. vancomycin) or antipseudomonal activity (e.g. antipseudomonal cephalosporin, antipseudomonal penicillin and carbapenems).
(DOCX)
J01A: tetracyclines; J01B: amphenicols; J01C: beta-lactam antibacterials, penicillins; J01D: other beta-lactam antibacterials; J01E: sulfonamides and trimethoprim; J01F: macrolides, lincosamides and streptogramins; J01G: aminoglycoside antibacterials; J01M: quinolone antibacterials; J01R: combinations of antibacterials; J01X: other antibacterials, P01A: agents against amoebiasis and other protozoal diseases.
(DOCX)
Parenteral antibiotics being prescribed within the first calendar day that a parenteral antibiotic was started were regarded as initial parenteral antibiotics. Patients who received a parenteral antibiotic for at least four consecutive days was used as a surrogate for severe infection, with the first calendar equal to the start date of parenteral antibiotics. Patients who died, were discharged to a hospice or transferred to other hospital before completing four consecutive days of parenteral antibiotics and had parenteral antibiotics continuously until the day prior to death, hospice discharge or transfer were also included as patients with severe infection.
(DOCX)
Cumulative incidence of (A) mortality, (B) having blood culture sampled, and (C) discontinuation of parenteral antibiotics among 19,170 patients with severe infection.
(DOCX)
(DOCX)
For this figure, antibiotics in the Watch category was divided to Watch and Watch+. Watch+ category comprises antibiotics in Watch category with anti-MRSA activity (e.g. vancomycin) or antipseudomonal activity (e.g. antipseudomonal cephalosporin, antipseudomonal penicillin and carbapenems.
(DOCX)
(DOCX)
(DOCX)
Acknowledgments
We thank the Hospital Information System Management Unit, Cipto Mangunkusumo National Referral Hospital for the support for this project. We thank Samuel Susanto for technical assistance.
Data Availability
All relevant data are available from within the paper and Supporting information files. We permit the use of third party data not contained within the paper, available from Hospital Information System Management Cipto Mangunkusumo National Referral Hospital: hospital admission, microbiology, and drug dispensing data set that could be accessed through umsi@rscm.co.id.
Funding Statement
This study is funded by Research Fund of Division of Tropical and Infectious Diseases, Department of Internal Medicine, Cipto Mangunkusumo National Hospital and in part by the Wellcome Trust [220211]. RS is funded by Indonesian Education Scholarship (Beasiswa Pendidikan Indonesia [BPI]) [202101182688] from the Ministry of Education, Culture, Research, and Technology Republic of Indonesia: Directorate General of Higher Education, Research, and Technology (Direktorat Jenderal Pendidikan Tinggi, Riset dan Teknologi Kementerian Pendidikan, Kebudayaan, Riset dan Teknologi Republik Indonesia) and Indonesian Endowment Fund for Education (Lembaga Pengelola Dana Pendidikan [LPDP]). DL is supported by the Wellcome Trust. For the purpose of Open Access, the author has applied a CC BY public copyright license to any Author Accepted Manuscript version arising from this submission. The funders of the investigators and study had no role in the study design, data collection, data analysis, data interpretation, or writing of the manuscript. The corresponding author had the final responsibility for the decision to submit for publication.
References
- 1.Antimicrobial Resistance Collaborators. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet. 2022;399(10325):629–55. doi: 10.1016/S0140-6736(21)02724-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Laxminarayan R BT, Frost I, Kariuki S, Khan EA, Limmathurotsakul D, et al. The Lancet Infectious Diseases Commission on antimicrobial resistance: 6 years later. Lancet Infect Dis. 2020;20(4):e51–e60. doi: 10.1016/S1473-3099(20)30003-7 [DOI] [PubMed] [Google Scholar]
- 3.Walsh TR, Gales AC, Laxminarayan R, Dodd PC. Antimicrobial Resistance: Addressing a Global Threat to Humanity. PLoS Med. 2023;20(7):e1004264. doi: 10.1371/journal.pmed.1004264 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.WHO. Diagnostic stewardship: a guide to implementation in antimicrobial resistance surveillance sites. Geneva: World Health Organization; 2016. [Google Scholar]
- 5.WHO. Antimicrobial stewardship programmes in health-care facilities in low- and middle-income countries. A WHO practical toolkit. Geneva: World Health Organization; 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Cox JA, Vlieghe E, Mendelson M, Wertheim H, Ndegwa L, Villegas MV, et al. Antibiotic stewardship in low- and middle-income countries: the same but different? Clin Microbiol Infect. 2017;23(11):812–8. doi: 10.1016/j.cmi.2017.07.010 [DOI] [PubMed] [Google Scholar]
- 7.Sinto R, Lie KC, Setiati S, Suwarto S, Nelwan EJ, Djumaryo DH, et al. Blood culture utilization and epidemiology of antimicrobial-resistant bloodstream infections before and during the COVID-19 pandemic in the Indonesian national referral hospital. Antimicrob Resist Infect Control. 2022;11(1):73. doi: 10.1186/s13756-022-01114-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.CDC. Hospital Toolkit for Adult Sepsis Surveillance. Atlanta, GA: US Department of Health and Human Services, CDC; 2018. [Google Scholar]
- 9.Lim C, Hantrakun V, Teerawattanasook N, Srisamang P, Teparrukkul P, Sumpradit N, et al. Impact of low blood culture usage on rates of antimicrobial resistance. J Infect. 2021;82(3):355–62. doi: 10.1016/j.jinf.2020.10.040 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Evans L, Rhodes A, Alhazzani W, Antonelli M, Coopersmith CM, French C, et al. Surviving sepsis campaign: international guidelines for management of sepsis and septic shock 2021. Crit Care Med. 2021;49(11):e1063–e143. doi: 10.1097/CCM.0000000000005337 [DOI] [PubMed] [Google Scholar]
- 11.Palacios-Baena ZR, Delgado-Valverde M, Valiente Mendez A, Almirante B, Gomez-Zorrilla S, Borrell N, et al. Impact of De-escalation on Prognosis of Patients With Bacteremia due to Enterobacteriaceae: A Post Hoc Analysis From a Multicenter Prospective Cohort. Clin Infect Dis. 2019;69(6):956–62. doi: 10.1093/cid/ciy1032 [DOI] [PubMed] [Google Scholar]
- 12.Deshpande A, Richter SS, Haessler S, Lindenauer PK, Yu PC, Zilberberg MD, et al. De-escalation of Empiric Antibiotics Following Negative Cultures in Hospitalized Patients With Pneumonia: Rates and Outcomes. Clin Infect Dis. 2021;72(8):1314–22. doi: 10.1093/cid/ciaa212 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.WHO. WHO report on surveillance of antibiotic consumption: 2016–2018 early implementation. Geneva: World Health Organization; 2018. [Google Scholar]
- 14.Hamilton WL, Pires SM, Lippett S, Gudka V, Cross ELA, Llewelyn MJ. The impact of diagnostic microbiology on de-escalation of antimicrobial therapy in hospitalised adults. BMC Infect Dis. 2020;20(1):102. doi: 10.1186/s12879-020-4823-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Huang LJ, Chen SJ, Hu YW, Liu CY, Wu PF, Sun SM, et al. The impact of antimicrobial stewardship program designed to shorten antibiotics use on the incidence of resistant bacterial infections and mortality. Sci Rep. 2022;12(1):913. doi: 10.1038/s41598-022-04819-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.ECDC. European Centre for Disease Prevention and Control. Point prevalence survey of healthcareassociated infections and antimicrobial use in European acute care hospitals–protocol version 5.3.. Stockholm: ECDC; 2016. [Google Scholar]
- 17.Liu P, Ohl C, Johnson J, Williamson J, Beardsley J, Luther V. Frequency of empiric antibiotic de-escalation in an acute care hospital with an established Antimicrobial Stewardship Program. BMC Infect Dis. 2016;16(1):751. doi: 10.1186/s12879-016-2080-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Lambregts MMC, Wijnakker R, Bernards AT, Visser LG, Cessie SL, Boer MGJ. Mortality after Delay of Adequate Empiric Antimicrobial Treatment of Bloodstream Infection. J Clin Med. 2020;9(5). doi: 10.3390/jcm9051378 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Savage RD, Fowler RA, Rishu AH, Bagshaw SM, Cook D, Dodek P, et al. The Effect of Inadequate Initial Empiric Antimicrobial Treatment on Mortality in Critically Ill Patients with Bloodstream Infections: A Multi-Centre Retrospective Cohort Study. PLoS One. 2016;11(5):e0154944. doi: 10.1371/journal.pone.0154944 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Austin PC, Steyerberg EW, Putter H. Fine-Gray subdistribution hazard models to simultaneously estimate the absolute risk of different event types: Cumulative total failure probability may exceed 1. Stat Med. 2021;40(19):4200–12. doi: 10.1002/sim.9023 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lau B, Cole SR, Gange SJ. Competing risk regression models for epidemiologic data. Am J Epidemiol. 2009;170(2):244–56. doi: 10.1093/aje/kwp107 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Cheng MP, Stenstrom R, Paquette K, Stabler SN, Akhter M, Davidson AC, et al. Blood culture results before and after antimicrobial administration in patients with severe manifestations of sepsis: a diagnostic study. Ann Intern Med. 2019. doi: 10.7326/M19-1696 [DOI] [PubMed] [Google Scholar]
- 23.Scheer CS, Fuchs C, Grundling M, Vollmer M, Bast J, Bohnert JA, et al. Impact of antibiotic administration on blood culture positivity at the beginning of sepsis: a prospective clinical cohort study. Clin Microbiol Infect. 2019;25(3):326–31. doi: 10.1016/j.cmi.2018.05.016 [DOI] [PubMed] [Google Scholar]
- 24.Tacconelli T, Angelis GD, Cataldo MA, Pozzi E, Cauda R. Does antibiotic exposure increase the risk of methicillin-resistant Staphylococcus aureus (MRSA) isolation? A systematic review and meta-analysis. J Antimicrob Chemother. 2008;61(1):26–38. doi: 10.1093/jac/dkm416 [DOI] [PubMed] [Google Scholar]
- 25.Yam ELY, Hsu LY, Yap EP, Yeo TW, Lee V, Schlundt J, et al. Antimicrobial resistance in the Asia Pacific region: a meeting report. Antimicrob Resist Infect Control. 2019;8:202. doi: 10.1186/s13756-019-0654-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Arena F, Argentieri M, Bernaschi P, Fortina G, Kroumova V, Manso E, et al. Real life turnaround time of blood cultures in the clinical microbiology laboratory: results of the first Italian survey, May 2015. Microbiologia Medica, 31(3). 2016;31(3). [Google Scholar]
- 27.Grinberg S, Schubert S, Hochauf-Stange K, Dalpke AH, Narvaez Encalada M. Saving Time in Blood Culture Diagnostics: a Prospective Evaluation of the Qvella FAST-PBC Prep Application on the Fast System. J Clin Microbiol. 2022;60(5):e0253321. doi: 10.1128/jcm.02533-21 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Tabak YP, Vankeepuram L, Ye G, Jeffers K, Gupta V, Murray PR. Blood Culture Turnaround Time in U.S. Acute Care Hospitals and Implications for Laboratory Process Optimization. J Clin Microbiol. 2018;56(12). doi: 10.1128/JCM.00500-18 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Barenfanger J, Graham DR, Kolluri L, Sangwan G, Lawhorn J, Drake CA, et al. Decreased mortality associated with prompt Gram staining of blood cultures. Am J Clin Pathol. 2008;130(6):870–6. doi: 10.1309/AJCPVMDQU2ZJDPBL [DOI] [PubMed] [Google Scholar]
- 30.Beekmann SE, Diekema DJ, Chapin KC, Doern GV. Effects of rapid detection of bloodstream infections on length of hospitalization and hospital charges. J Clin Microbiol. 2003;41(7):3119–25. doi: 10.1128/JCM.41.7.3119-3125.2003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Fidelman P, Rygalski K, Riley J, Hernandez D, Wolk D. Characterization of Clinical Improvements After Implementation of New Blood Culture Strategy for Patients With Severe Sepsis. Open Forum Infectious Diseases. 2016;3(suppl_1). [Google Scholar]
- 32.Ombelet S, Barbe B, Affolabi D, Ronat JB, Lompo P, Lunguya O, et al. Best Practices of Blood Cultures in Low- and Middle-Income Countries. Front Med (Lausanne). 2019;6:131. doi: 10.3389/fmed.2019.00131 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Perez KK, Olsen RJ, Musick WL, Cernoch PL, Davis JR, Peterson LE, et al. Integrating rapid diagnostics and antimicrobial stewardship improves outcomes in patients with antibiotic-resistant Gram-negative bacteremia. J Infect. 2014;69(3):216–25. doi: 10.1016/j.jinf.2014.05.005 [DOI] [PubMed] [Google Scholar]
- 34.Savinelli T, Parenteau S, Mermel LA. What happens when automated blood culture instrument detect growth but there are no technologists in the microbiology laboratory? Diagn Microbiol Infect Dis. 2004;48:173–4. doi: 10.1016/j.diagmicrobio.2003.10.001 [DOI] [PubMed] [Google Scholar]
- 35.Sinto R, Suwarto S, Lie KC, Harimurti K, Widodo D, Pohan HT. Prognostic accuracy of the quick Sequential Organ Failure Assessment (qSOFA)-lactate criteria for mortality in adults with suspected bacterial infection in the emergency department of a hospital with limited resources. Emerg Med J. 2020;0:1–7. doi: 10.1136/emermed-2018-208361 [DOI] [PubMed] [Google Scholar]
- 36.Chedid M, Waked R, Haddad E, Chetata N, Saliba G, Choucair J. Antibiotics in treatment of COVID-19 complications: a review of frequency, indications, and efficacy. J Infect Public Health. 2021;14(5):570–6. doi: 10.1016/j.jiph.2021.02.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Rashid MM, Akhtar Z, Chowdhury S, Islam MA, Parveen S, Ghosh PK, et al. Pattern of antibiotic use among hospitalized patients according to WHO Access, Watch, Reserve (AWaRe) classification: findings from a point prevalence survey in Bangladesh. Antibiotics 2022;11(810). doi: 10.3390/antibiotics11060810 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Adebisi YA, Jimoh ND, Ogunkola IO, Uwizeyimana T, Olayemi AH, Ukor NA, et al. The use of antibiotics in COVID-19 management: a rapid review of national treatment guidelines in 10 African countries. Trop Med Health. 2021;49(1):51. doi: 10.1186/s41182-021-00344-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Hamada S, Tokuda Y, Honda H, Watari T, Suzuki T, Moromizato T, et al. Prevalence and characteristics of antibiotic prescription for acute COVID-19 patients in Japan. Sci Rep. 2022;12(1):22340. doi: 10.1038/s41598-022-26780-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Russell NJ, Stohr W, Plakkal N, Cook A, Berkley JA, Adhisivam B, et al. Patterns of antibiotic use, pathogens, and prediction of mortality in hospitalized neonates and young infants with sepsis: A global neonatal sepsis observational cohort study (NeoOBS). PLoS Med. 2023;20(6):e1004179. doi: 10.1371/journal.pmed.1004179 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.De Waele JJ, Schouten J, Beovic B, Tabah A, Leone M. Antimicrobial de-escalation as part of antimicrobial stewardship in intensive care: no simple answers to simple questions—a viewpoint of experts. Intensive Care Med. 2020;46: 236–44. doi: 10.1007/s00134-019-05871-z [DOI] [PMC free article] [PubMed] [Google Scholar]