Highlights
- •A little over half of all patients with suspected UTI had urine cultures obtained.
- •The most common empiric regimens were cephalosporin-based.
- •Most isolates were not susceptible to agents in the empiric regimen.
- •Most antibiotics were not changed to ensure the coverage of the urinary isolate.
Abstract
Background
The inappropriate use of antibiotics is a global threat. Clinical microbiology laboratory testing can support optimal antibiotic prescribing for many conditions. The purpose of this study was to characterize antibiotic usage in the context of treatment of uncomplicated urinary tract infection (UTI) received through a Ghanaian Accident & Emergency, which found high rates of improper usage.
Methods
A prospective cohort study recruited patients >18 years of age who were admitted to a large Ghanaian teaching hospital with suspected UTI. Eligible patients were identified through a daily review of admission logs. Data were collected through a review of medical records and microbiology laboratory urine data.
Results
A total of 90 patients were enrolled, but urine cultures were obtained from only 50 (56%) patients. Most positive urine cultures grew typical urinary pathogens (21 of 25, 84%). The most common empiric regimens were cephalosporin-based and did not vary if additional infections were suspected or a urine culture was not sent. The majority of patients with confirmed UTI had isolates that were not susceptible to antibiotics selected for empiric treatment (18 of 21, 86%). Although more than half had their empiric regimen changed between admission and study follow up, only 42% (5 of 12) were switched to a regimen that included agent(s) to which their urinary isolate was known to be susceptible.
Conclusions
Establishing hospital-wide guidelines regarding the evaluation and treatment of patients with suspected UTI may help improve antibiotic utilization and patient outcomes by increasing the use of urine cultures and tailoring therapy in response to culture results.
Keywords
Introduction
Antimicrobial resistance to drugs used for treating common infectious agents has markedly increased. Common antibiotics, such as fluoroquinolone, are now ineffective against a large proportion of Escherichia coli infections, the most common cause of community-onset urinary tract infections (UTIs) in many parts of the US and globally. Even carbapenem antimicrobials that are considered as “last-line” drugs to treat highly resistant gram-negative bacteria have become ineffective in more than half of the patients treated (
World Health Organization, 2020
).Despite the fact that antimicrobial resistance can occur naturally over time, numerous experts and international groups assert that both misuse and overuse of antimicrobials have dramatically accelerated this process. To slow the rapidly increasing prevalence of infection due to drug-resistant organisms, the Center for Global Development’s Drug Resistance Working Group recommends to broadly adopt antimicrobial stewardship practices to ensure that drugs are properly prescribed, dispensed, and used (
Nugent et al., 2010
).UTIs are common and a frequent reason for in-hospital use of antibiotics. These infections are classically diagnosed using a combination of clinical signs and symptoms, coupled with laboratory data. Often, the empiric treatment of UTI is initiated based on the patient history and examination at the time of presentation. Microbiological laboratory data can play a major role to improve how antibiotics are used in both the empiric and definitive treatment of UTI.
An initial study done by Owusu-Ofori et al., which characterized the usage of antimicrobials at Komfo Anoyke Teaching Hospital in Kumasi, Ghana, found the evidence of high rates of inappropriate usage of antibiotics in the Accident and Emergency (A&E) Department (
Owusu-Ofori et al., 2017
). In this study, we sought to further characterize the usage of antibiotics for a cohort of patients hospitalized with presumed UTIs after they presented for care at A&E Department and identify opportunities for the improved use of clinical microbiology laboratory data.Materials & methods
Study design and site
We conducted a prospective cohort study of adult patients who were hospitalized with a suspected UTI with or without sepsis. This study was conducted at the second largest hospital in Ghana, a 1,200-bed teaching hospital in Kumasi. Two departments, the A&E Department and the Clinical Microbiology Laboratory, were involved in this study. Patients were enrolled from the A&E Department, which receives about 20,000 cases each year. The hospital is developing an Antimicrobial Stewardship Program but there are currently no specific guidelines for the treatment of UTIs or formulary restrictions. The hospital pharmacy did not experience any antibiotic shortages during the study period.
The study protocol was reviewed and approved by the Committee on Human Research, Publications and Ethics at Kwame Nkrumah University of Science and Technology and Komfo Anokye Teaching Hospital in Kumasi, Ghana, and the Institutional Review Board at Children’s Hospital of Philadelphia.
Study population
Patients >18 years of age were considered eligible if they were admitted for suspected UTI (based on admitting diagnosis or because a urine culture was sent). Eligible patients were identified through a daily review of the admissions log and also urine culture data from the study hospital’s Clinical Microbiology Laboratory.
Data collection
Patients’ medical records were reviewed at the time of enrollment and at discharge or hospital day >4 for patients with longer lengths of stay. Results of all urine cultures sent were captured by a review of the Clinical Microbiology Laboratory records or patients’ medical records. Chart review was conducted to identify comorbid conditions present at the time of hospital admission, which were subsequently categorized by the organ system.
Data definitions
A confirmed UTI was defined as urine culture with growth >100,000 cfu/ml of a single organism. A presumed extended-spectrum beta-lactamase-producing (ESBL) organism was defined as an organism with resistance to two or more third or fourth generation cephalosporins.
For patients with confirmed UTIs, we assessed both empiric and definitive treatment regimens. Empiric regimens were those started prior to urine culture results being available. Definitive regimens were those prescribed after urine cultures became available; if the date on which the urine culture results were finalized was unknown, we assigned it to be hospital day (
Schmiemann et al., 2010
).Data management, analysis, and statistical methods
Prospective data were collected and entered into a REDCap database. Baseline patient characteristics, hospitalization characteristics, and outcomes were summarized as medians with interquartile range for continuous variables and as frequencies and percentages for categorical variables.
Results
Characteristics of participants
A total of 90 adults were enrolled, of whom 51 (57%) were female participants (Table 1). The most common admitting diagnoses were suspected UTI or urosepsis, although 6 (7%) participants were suspected of having a non-UTI as their major reason for hospitalization. Cardiac and endocrine conditions were the most common comorbid conditions, although 4% of enrolled patients were previously healthy. Most had unknown HIV status.
Table 1Characteristics of patients hospitalized and treated for suspected urinary tract infection (n = 90).
| Characteristic | Number (%) |
|---|---|
| Female | 51 (57) |
| Median age in yrs (IQR) | 65 (50–77) |
| Reason for admission | |
| UTI | 59 (66) |
| Urosepsis | 25 (28) |
| Other infection | 6 (7) |
| Comorbid conditions | |
| None | 4 (4) |
| Cardiac | 35 (39) |
| Endocrine | 27 (30) |
| Nervous system | 18 (20) |
| Hematology/oncology | 14 (16) |
a other infections included skin and soft tissue infection (n = 3), sepsis without apparent source (n = 2), and pneumonia (n = 1).
b total > 100% because each enrolled patient could have up to 3 comorbid conditions.
Diagnosis and empiric treatment of urinary tract infections
Half (25 of 50, 50%) of urine cultures were positive, although only 21 cultures grew typical urinary pathogens (Table 2). Although 84 (93%) enrolled patients were hospitalized with a primary diagnosis of suspected UTI or urosepsis, many (40 of 90, 44%) did not have a urine culture sent. Patients who did not have a urine culture sent were majority female (21 of 40, 52%) with an average age of 60 years. Most patients who did not have a urine culture sent were admitted with a diagnosis of UTI (31 of 40, 76%).
Table 2Microbiology and antimicrobial susceptibilities of urinary isolates.
| Laboratory result | Number (%) |
|---|---|
| Urine culture results (n = 50) | |
| No culture sent | 40 (44) |
| <100,000 cfu | 25 (28) |
| >100,000 cfu | 25 (28) |
| Organism isolated (n = 25) | |
| E.coli | 9 (36) |
| Klebsiella spp. | 9 (36) |
| Other gram-negative organism | 3 (12) |
| Candida spp. | 3 (12) |
| Staph aureus | 1 (4) |
| Antimicrobial resistance rates | |
| Presumed ESBL | 15/21 (71) |
| Ciprofloxacin | 12/16 (75) |
| Cotrimoxazole | 20/20 (100) |
| Gentamicin | 11/18 (61) |
| Amikacin | 1/21 (5) |
a antimicrobial susceptibility test results presented for only gram-negative organisms (n = 21); not all isolates were tested for susceptibility to the same panel of antimicrobial agents.
b presumed extended-spectrum beta-lactamase-producing organism based on resistance to 2 or more third or fourth generation cephalosporins.
The most common empiric treatment regimens used for all patients were cephalosporin-based and did not vary based on whether a urine culture was sent or if non-UTI was also being treated.
Microbiology and definitive treatment of laboratory-confirmed urinary tract infections
Among the 21 patients with laboratory-confirmed, gram-negative UTI, Klebsiella pneumoniae and E. coli were the most commonly isolated pathogens (both n = 9, 33%). Most isolates (n = 15, 71%) were presumed ESBL-positive organisms, including 8 E. coli and 5 K. pneumoniae (Table 3, Table 4) .
Table 3Empiric treatment of suspected urinary tract infections (n = 90).
| Suspected UTI (n = 65) | Lab-confirmed UTI (n = 21) | |
|---|---|---|
| Cephalosporin only | 37 (57) | 9 (43) |
| Cephalosporin + second agent | 9 (14) | 4 (19) |
| Quinolone based regimen | 6 (9) | 4 (19) |
| None | 8 (12) | 3 (14) |
| Carbapenem | 0 (0) | 1 (5) |
a second agents included metronidazole (n = 2), ciprofloxacin (n = 1), and vancomycin (n = 1).
# includes one patient treated with ciprofloxacin and metronidazole.
b excludes patients with urine cultures positive for Candida spp. (n = 3) and Staph aureus (n = 1).
Table 4Duration of antimicrobial therapy for patients hospitalized with suspected urinary tract infection.
| Urine Culture | |||
|---|---|---|---|
| Not Sent | Negative (<100,000 cfu) | Positive (>100,000 cfu) | |
| Median duration of antibiotics, in days (IQR) | 2 (2–5) | 4 (3–5) | 3 (2–5) |
a No data on planned duration of treatment for 8 patients.
b No data planned duration of treatment for 2 patients.
c gram-negative organisms only. No data on planned duration of 2 patients.
The majority (n = 18, 86%) of patients with laboratory-confirmed gram-negative UTI had isolates that were not susceptible to the antibiotics selected for empiric coverage. Although two-thirds of these patients (n = 12, 67%) had their antibiotic regimen changed between admission and study follow up, only 5 patients were changed to an antibiotic regimen to which their urinary isolate was known to be susceptible.
Outcomes
A total of 18 (20%) of enrolled patients died before hospital discharge. The rate of in-hospital death was the same for patients who did and did not have a positive urine culture (7 of 25, 28%). We observed a lower mortality among patients who did not have a urine culture sent (4 of 40, 10%; p = 0.044).
Of those patients with culture-confirmed gram-negative urinary pathogens, 28% (5 of 18) died during hospitalization. The majority of patients who died (3 of 5, 60%) were presumed to have died as a result of their UTI, while most (4 of 5, 80%) had appropriate coverage of their isolate with their empiric regimen. The remaining patient was placed on appropriate coverage of his isolate between study enrollment and death.
Discussion
In this study of adults hospitalized with suspected UTI in a resource-limited setting, almost half did not have a urine sample sent for culture. Additionally, many patients with documented UTI, did not have their antibiotic therapy adjusted in response to culture results. With the rapid emergence of multidrug-resistant bacteria, global efforts to improve the use of antimicrobial agents are needed. These findings identify possible targets for interventions to optimize the care of patients with suspected UTI.
Almost half of the patients enrolled in this study with presumed UTI did not have urine cultures sent to confirm this diagnosis. However, nearly all patients receive a urine dipstick upon admission to A&E. According to the IDSA guidelines for diagnosis and treatment of acute uncomplicated UTI, clinical decisions regarding therapy should be based upon several factors, including local resistance patterns, presentation, and more specific factors such as allergies. Urine culture and susceptibility testing are a strong recommendation deemed necessary to tailor empiric therapy (
Gupta et al., 2011
). A larger study, which sought to characterize antibiotic usage in the treatment of UTI across 6 US hospitals found that culture collection at the time of empiric therapy initiation strongly affected later decisions to modify therapy. Yet even when cultures were negative, less than half of patients in the study had their regimen narrowed or stopped (- Gupta K.
- Hooton T.
- Naber K.
- Wullt B.
- Colgan R.
- Miller L.G.
- et al.
International clinical practice guidelines for the treatment of acute uncomplicated cystitis and pyelonephritis in women: a 2010 update by the infectious diseases society of America and the European society for microbiology and infectious diseases.
Clin Infect Dis. 2011; 52: e103-e120
Forson et al., 2016
).Additionally, less than half of patients with confirmed UTI had their antibiotic regimen changed to better target the isolated pathogen. These findings are similar to a study conducted at Central Emek Hospital in Israel (
Braykov et al., 2014
). While many patients admitted with suspected UTI did not have a sample sent for culture and susceptibility testing, failure to use available data represents a missed opportunity to ensure patients receive the narrowest effective therapy.We are unable to comment on the rationale underlying the clinical treatment decisions made for each enrolled patient. However, the results of this study demonstrate opportunities for the greater use of KATH’s clinical microbiology lab to promote good antibiotic stewardship in the treatment of UTI. We recommend that cultures should be sent for all patients with suspected UTI based on clinical symptoms and history. Furthermore, antibiotic susceptibility data should be retrieved and used to tailor therapy when possible, including stopping antibiotics when there is no laboratoryevidence of a UTI. Lastly, the usage of urine-focused antibiogram could help clinicians select an empiric regimen more likely to cover infecting pathogens.
Limitations of this study included a lack of urinalysis data, which might have provided further insight into the clinical decision-making process in cases for which a urine culture was not sent. Finally, because of the relatively small sample size and the fact that this was a single-center study, these results may not be generalizable to similar populations.
Conclusions
Less than half of Ghanaian adults treated for UTIs had their diagnoses confirmed by urine culture. Few had their treatment appropriately tailored to pathogen susceptibilities. Given the global threat of antimicrobial resistance, the use of clinical microbiology laboratory testing and results can improve the appropriate use of antimicrobial agents for adult patients with UTIs, which is one of the most common causes of hospitalization.
Conflict of interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Acknowledgments
The authors acknowledge the ongoing efforts of the clinical and microbiology laboratory staff at Komfo Anokye Teaching Hospital.
References
- Assessment of empirical antibiotic therapy optimisation in six hospitals: an observational cohort study.Lancet Infect Dis. 2014; 14: 1220-1227
- Frequency of alcohol use among injured adult patients presenting to a Ghanaian emergency department.Ann Emerg Med. 2016; 68: 492-500
- International clinical practice guidelines for the treatment of acute uncomplicated cystitis and pyelonephritis in women: a 2010 update by the infectious diseases society of America and the European society for microbiology and infectious diseases.Clin Infect Dis. 2011; 52: e103-e120
- The race against drug resistance.(Available at:) Center for Global Development, 2010www.WhenMedicinesFail.org
- Antibiotic use in an accident and emergency department in a teaching hospital in Ghana.Open Forum Infect Dis. 2017; 4 (S247-S247)
- The diagnosis of urinary tract infection: a systematic review.Dtsch Arztebl Int. 2010; 107: 361-367
World Health Organization. Antimicrobial Resistance. World Health Organization, Available at: World Health Organization www.who.int/mediacentre/factsheets/fs194/en/.
Article info
Publication history
Published online: October 29, 2020
Accepted:
October 23,
2020
Received in revised form:
October 18,
2020
Received:
September 10,
2020
Identification
Copyright
© 2020 The Authors. Published by Elsevier Ltd on behalf of International Society for Infectious Diseases.
User license
Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0) | How you can reuse 
Elsevier's open access license policy
Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0)
Permitted
For non-commercial purposes:
- Read, print & download
- Redistribute or republish the final article
- Text & data mine
- Translate the article (private use only, not for distribution)
- Reuse portions or extracts from the article in other works
Not Permitted
- Sell or re-use for commercial purposes
- Distribute translations or adaptations of the article
Elsevier's open access license policy
