ORIGINAL ARTICLE
Karuna Tiwari1, Samruddhi Patil1, Aparna Naik1, Anjali Shetty1, Kamini Walia2 and Camilla Rodrigues1*
1Department of Microbiology, P. D. Hinduja Hospital and Medical Research Centre, Mumbai, India; 2Indian Council of Medical Research (ICMR), New Delhi, India
Empiric antimicrobial therapy in hospitalized patients is guided by an institution’s cumulative antibiogram, which may not be adequate in giving information on decision-making for optimal treatment in different patient populations. Adding patient risk factors can make it more useful for clinicians in guiding empiric therapy and for antimicrobial stewardship. Cumulative data were obtained for blood culture and urine isolates from the laboratory information system of a tertiary care hospital for 6 months (January to June 2019). Further stratification of organism types and resistance rates on the basis of patient risk factors (Patient Types 1, 2, and 3) was performed and analyzed. Salmonella spp. was seen in community-acquired ward patients (Types 1 and 2). Streptococcus pneumoniae was seen in Type 1 patients, and Acinetobacter spp. was seen in Type 3 patients. Extended-spectrum beta-lactamase-producing gram-negative infection rates were higher in community patients than in hospital patients. Carbapenem-resistant Enterobacteriaceae rates were high in Type 3 hospitalized patients. Cumulative blood methicillin-resistant Staphylococcus aureus rates were 43% but stratification showed it only in Type 2 and Type 3 ICU patients with 0% in ward patients. Stratified antibiograms based on patient risk factors are valuable for antimicrobial stewardship and help to optimize empiric therapy and increase the understanding of antimicrobial resistance trends.
Keywords: antimicrobial stewardship; antibiogram; drug resistance; risk factors; India
Citation: Int J Infect Control 2021, 17: 20089 – http://dx.doi.org/10.3396/ijic.v17.20089
Copyright: © 2021 Karuna Tiwari et al. This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material for any purpose, even commercially, provided the original work is properly cited and states its license.
Received: 10 February 2020; Accepted: 22 June 2020; Published: 21 April 2021
Competing interests and funding: Camilla Rodrigues is a Scientific Advisory Board member and has received speaker honoraria from Pfizer, Sanofi and Biomerieux; and has received speaker honoraria from B Braun, Becton Dickinson , Cipla, Glenmark, Novartis, and Cepheid. Anjali Shetty is an Advisory Board member for Bharat Biotech and has received honoraria from Pfizer.
*Camilla Rodrigues, Department of Microbiology, P. D. Hinduja Hospital and Medical Research Centre, Veer savarkar marg, Mahim, Mumbai – 400016, Maharashtra, India. Email: dr_crodrigues@hindujahospital.com
Antimicrobial resistance is increasing worldwide, so it is crucial to monitor the emerging trends in drug resistance at the local hospital level to support decision-making for clinicians, infection control interventions, and antimicrobial resistance containment strategies. Cumulative susceptibility data and derived resistance patterns are used to guide empiric antimicrobial therapy and to detect changes in antibiotic resistance over time, as surveillance of local antibiotic resistance is an integral part of antimicrobial stewardship. In our healthcare facilities, local antimicrobial resistance trend monitoring is commonly performed using an annual summary of susceptibility rates known as a cumulative antibiogram. To ensure applicability to specific patient populations, the Clinical and Laboratory Standards Institute (CLSI) has published guidelines recommending division of susceptibility data by patient location, clinical service, specimen type, and patient population (1).
Clinicians rely on an institution wide cumulative antibiogram that may be skewed because it does not accurately reflect susceptibility rates in different patient populations. Thus, proper history details and patient risk factors, a history of prior admissions, and co-morbid conditions should be included in order to make it meaningful for the clinicians who have to prescribe drugs on the basis of our data.
Standard antibiogram studies segregate only laboratory data. A stratified antibiogram gives clinicians more information than cumulative antibiogram—provides better understanding of resistance patterns to guide optimum empirical therapy and the formulation of an effective local antibiotic policy in hospitals.
Our objective was to compare the hospital-wide cumulative antibiograms of patients with the results of additional stratification of susceptibility data to help with appropriate empiric antibiotic selection and improved surveillance of antimicrobial resistance trends in our institution.
The antibiogram is assessed 6 monthly in our tertiary care hospital, P. D. Hinduja Hospital and Medical Research Centre, Mumbai, India, by aggregating the susceptibility data of all isolates from different locations (outpatient, wards, ICU) and sites of infection (blood, urine, lower respiratory tract, intra-abdominal infection) of patients of all ages. For this study, antimicrobial susceptibility data and cumulative antibiogram were analyzed only for blood and urine cultures from the laboratory information system for a period of 6 months from January to June 2019.
The same blood and urine sample data were used to further stratify our antibiogram into community-acquired and hospital-acquired infections by including patient history details and risk factors such as hospitalization of the patient during the previous 90 days, invasive procedures performed during hospitalization, antibiotic exposure in the previous 90 days, and any co-morbid conditions such as immunosuppression, diabetes mellitus, and chronic kidney disease (CKD) (2). On the basis of these risk factors, patients were divided into three types, Types 1, 2, and 3, as shown in Table 1. For every positive blood and urine culture isolate, the patient’s contact detail (for outpatients) or admission detail (for inpatients) was noted. Each patient was contacted by phone and asked for history and risk factors. Depending upon these risk factors, patient type (Type 1, 2, or 3) was determined. The patient’s admission date was also noted which helped us to evaluate whether the infection was community acquired (within 48 h of admission) or healthcare associated (after 48 h of admission).
| Patient risk factors (2) | Patient type 1 | Patient type 2 | Patient type 3 |
| Hospitalization in last 90 days | No | Yes | Yes |
| Invasive procedures during hospitalization | Not applicable | No | Yes |
| Antibiotics received in last 90 days | No | ≤2 | ≥3 |
| Co-morbid conditions | No | ≤2 | ≥3* |
| Criteria to be fulfilled | All | Any 1 | Any 1 |
| *Patients with malignancy/CKD (on hemodialysis) are considered patient Type 3. | |||
The study was approved by the Institutional Ethics Committee.
Our cumulative organism data for 6 months (January to June 2019) showed 317 isolates for blood culture in wards and ICU collectively, of which 155 were ward isolates and 162 were ICU isolates. The most common organisms in wards were Escherichia coli (32%), followed by Salmonella spp. (9%), Klebsiella pneumoniae (8%), Staphylococcus aureus (all MSSA) (6%), and Enterococcus spp. (6%); in the ICU, the most common organisms were K. pneumoniae (20%) and E. coli (20%), followed by Enterococcus spp. (6%), Pseudomonas aeruginosa (5%), and S. aureus (4%).
Similarly, the total number of isolates was 432 for urine cultures in wards and ICU collectively, of which 315 were ward isolates and 117 were ICU isolates. Urine isolates from wards were mainly E. coli (44%), K. pneumoniae (19%), P. aeruginosa (8%), Enterococcus spp. (5%), and Proteus mirabilis (3%) and those from ICU were mainly E. coli (32%), K. pneumoniae (16%), Enterococcus spp. (5%), P. aeruginosa (5%), and P. mirabilis (3%).
Further stratification of the most common organisms based on cumulative data was performed on the basis of patient risk factors, and the results were different as shown in Table 2.
| Sample type | Location type | Infection acquired type | Patient types based on risk factors (refer to Table 1) | ||
| Patient type 1 | Patient type 2 | Patient type 3 | |||
| Blood | Wards | Community acquired (within 48 h of admission) | 1. Salmonella spp. 2. E. coli |
1. E. coli 2. Salmonella spp. 3. K. pneumoniae 4. S. aureus 5. P. aeruginosa |
1. E. coli 2. Klebsiella spp. 3. Pseudomonas spp. 4. Enterococcus spp. 5. S. aureus |
| Hospital acquired (after 48 h of admission) | * | 1. E. coli 2. Proteus spp. 3. S. aureus 4. Streptococcus spp. |
1. E. coli 2. Klebsiella spp. 3. Pseudomonas spp. 4. S. aureus 5. Enterococcus spp. |
||
| ICU | Community acquired | 1. Streptococcus pneumoniae 2. E. coli 3. K. pneumoniae |
1. E. coli 2. Salmonella spp. 3. S. aureus 4. Enterococcus spp. |
1. E. coli 2. Klebsiella spp. 3. Acinetobacter spp. 4. S. aureus 5. Burkholderia cepacia |
|
| Hospital acquired | * | 1. E. coli 2. Acinetobacter spp. 3. Enterococcus spp. |
1. E. coli 2. Klebsiella spp. 3. Pseudomonas spp. 4. Candida spp. 5. Enterococcus spp. 6. Enterobacter spp. |
||
| Urine | Wards | Community acquired | 1. E. coli 2. Klebsiella spp. 3. Enterococcus spp. 4. Pseudomonas spp. 5. Salmonella spp. |
1. E. coli 2. Candida spp. 3. Pseudomonas spp. 4. Klebsiella spp. 5. Enterococcus spp. |
1. E. coli 2. Klebsiella spp. 3. Pseudomonas spp. 4. Enterococcus spp. 5. Acinetobacter spp. |
| Hospital acquired | * | 1. E. coli 2. Klebsiella spp. 3. Enterococcus spp. 4. Proteus spp. 5. Candida sp. |
1. E. coli 2. Klebsiella spp. 3. Pseudomonas spp. 4. Providencia spp. 5. Enterobacter spp. |
||
| ICU | Community acquired | 1. K. pneumoniae 2. Candida spp. |
1. E. coli 2. Candida spp. 3. K. pneumoniae 4. Pseudomonas spp. 5. Streptococcus spp. |
1. E. coli 2. Candida spp. 3. K. pneumoniae 4. Enterococcus spp. 5. Morganella morganii 5. S. aureus |
|
| Hospital acquired | 1. E. coli 2. K. pneumoniae 3. Candida spp. 4. Pseudomonas spp. 5. Enterococcus spp. |
1. K. pneumoniae 2. E. coli 3. Candida spp. 4. Pseudomonas spp. 5. Enterococcus spp. |
|||
| *Type 1 patients have not been hospitalized and therefore cannot have hospital-acquired infection. | |||||
Resistance rates in blood and urine isolates are shown in Tables 3 and 4, respectively, with cumulative data in the left column and stratified data in the remaining columns.
Inclusion of different patient types based on risk factors at our hospital was transformative to the hospital antibiogram.
The most common blood culture isolates in wards and in ICU as per cumulative data were E. coli, followed by Salmonella spp. and K. pneumoniae. Further stratification showed that community-acquired Salmonella spp. was the common organism in Type 1 and Type 2 community-acquired infections and not in Type 3 and hospital-acquired infections, which our cumulative data were not able to differentiate. In ICU, Streptococcus pneumoniae was seen in Type 1 patients and Acinetobacter spp. was seen in Type 3 patients. This lent clarity about common organisms in certain patient types.
Our ESBL rates in blood and urine in wards and ICU were almost the same as shown by cumulative data, while risk factor stratification showed that ESBL rates were higher in community patients than in hospital patients as hospital CRE rates could have masked the actual ESBL rates.
CRE rates in ward blood isolates were 20% and 37% in ICU as per the cumulative data. When we stratified it further, CRE rates were higher in Type 3 patients, while they were nil in the blood isolates of Type 1 and Type 2 patients. The same pattern was also seen in urine isolates. CRE coverage is important in Type 3 patients with multiple risk factors and not in other patient types; our cumulative antibiogram did not differentiate patient types because it extrapolates CRE rates for whole wards and ICU populations. On the basis of this observation, we can change our antibiotic policy to add colistin with carbapenem in Type 3 hospitalized patients.
Cumulative VRE rates in blood isolates in wards and ICU were 22% and 10%, respectively. Our stratified data showed that VRE rates were around 20% in Type 3 patients in wards and 50% in Type 3 community-acquired ICU patients, although numbers were lower. Similarly, our cumulative blood MRSA rates were 43%, while stratification showed that this was seen only in Type 2 and Type 3 ICU patients. Our observation of 6 months data hints at how clinical stratification can help us with antimicrobial stewardship by including risk factors and relevant history of patients in our antibiogram. These early data on patient stratification with risk factors and relevant history are vital to the understanding and implementation of antibiogram.
As our study included 6 months antibiogram data, numbers are small and statistical analysis could not be performed. We therefore do not know if the difference between cumulative and stratified antibiograms is significant, but certainly a difference is observable and warrants further investigation.
Clinicians’ reliance on institution wide antibiograms that do not accurately reflect susceptibility rates in certain patient groups might lead to inappropriate empiric antibiotic prescribing. Usual antibiogram studies have only laboratory data, but our study went further and incorporated relevant patient-related clinical history and other factors to stratify susceptibility data.
Stratifying the antibiogram by incorporating patient risk factors is a valuable antibiotic stewardship tool that helps in appropriate empiric antibiotic selection and improved surveillance of antimicrobial resistance trends. Such stratification is key to antimicrobial stewardship program (3).
We acknowledge the Indian Council of Medical Research (ICMR) for supporting us as a Regional Centre for the National Antimicrobial Resistance (AMR) surveillance study.