Evaluation of antibiotic appropriateness at an outpatient oncology centre

Abstract

Current evidence supporting antimicrobial stewardship programs focused largely in inpatient setting. With the shift in cancer management from inpatient to ambulatory setting, it is crucial to examine the prevalence and predictors of inappropriate antibiotics prescribing. This is a retrospective cross-sectional study conducted at the National Cancer Centre Singapore (NCCS). Patients at least 21 years, with an active or past cancer diagnosis and prescribed with at least one oral antibiotic by a NCCS physician from 1st July to 30th September 2019 were included. Antibiotic appropriateness was assessed using institutional antibiotic guidelines or published clinical practice guidelines. For cases where antibiotics appropriateness cannot be ascertained using these guidelines, an independent three-member expert panel was consulted. A total of 815 patients were screened; 411 (59.4% females) were included with mean age of 62.4 years. The top three cancer diagnoses were breast (26.5%), lung (15.6%) and head and neck (13.6%). More than half (58.6%) received appropriate antibiotic choice. Of which, 235 (97.5%), 238 (98.8%) and 194 (80.5%) received appropriate dose, frequency and duration respectively. The presence of non-oncologic immunosuppressive comorbidities (OR 4.890, 95% CI 1.556-15.369, p-value = 0.007), antibiotic allergy (OR 2.352, 95% CI 1.178-4.698, p-value = 0.015) and skin and soft tissue infections (OR 2.004, 95% CI 1.276-3.146, p-value = 0.003) were associated with a higher incidence of inappropriate antibiotic choice. This study highlighted that inappropriate antibiotic prescribing is prevalent in the ambulatory oncology setting. Predicators identified can aid in the design of targeted strategies to optimise antibiotic use in ambulatory oncology patients.

Keywords

antibiotics appropriateness ambulatory oncology predictors

Introduction

Overprescribing antibiotics creates selective pressure that contributes to increased antimicrobial resistance (AMR) that poses as a significant threat to public health.^1–5 According to the Centers for Disease Control and Prevention (CDC), in the United States alone, the number of people infected with drug-resistant bacteria has increased from at least 2 million in 2013 to 2.8 million in 2019.^6,7 Similarly, the number of deaths resulting from AMR has also increased from 23, 000 to more than 35, 000.^6,7 Within oncology, it had been reported that nearly 70% of ambulatory patients were prescribed with broad-spectrum antibiotics unnecessarily, resulting in an increased risk of AMR.⁸ Overprescribing antibiotics is of particular concern in cancer patients as they may be receiving immunosuppressive chemotherapy regimens, may become neutropenic and prone to infections.^8–10 With the rise in AMR due to the overprescribing of antibiotics, it may limit the options of effective antibiotics for the treatment of infections. Moreover, infections caused by AMR organisms can result in a delay in chemotherapy treatment, leading to an increased in morbidity and mortality.^11,12

To address AMR, antimicrobial stewardship programs (ASPs) were established to promote the judicious use of antibiotics. ASPs have demonstrated to reduce duration of antimicrobial therapy, length of hospitalisation, healthcare associated infections, and healthcare costs.^13–15 For instance, it was observed that ASPs resulted in a 9% reduction in healthcare-associated antibiotic resistant organisms and a 13% reduction in healthcare-associated multi-drug resistant organism (MDRO)¹³ as well as a reduction in the duration of antimicrobial therapy by 2 days and length of hospitalisation by 5 days in the acute management of bacterial skin infections.¹⁵ Likewise, implementation of ASPs in oncology and specifically haematopoietic stem cell transplant settings had also demonstrated a decrease in antibiotic use without compromising on patient clinical outcomes.¹⁶ Currently, ASPs for oncology patients mainly focused on hospitalised patients (inpatients). With the paradigm shift in oncology care from inpatient to outpatient^17–19 coupled with the high prevalence of inappropriate antibiotics prescribing observed in ambulatory settings,¹⁷ it is crucial to gain insights into the prevalence of inappropriate antibiotic prescribing in an ambulatory oncology setting.

Method

Study setting and design

This is a retrospective cross-sectional study conducted at National Cancer Centre Singapore (NCCS), the largest ambulatory cancer centre in Singapore, that serves up to 65% of the country's cancer population.²⁰ This study was approved by SingHealth Centralised Institutional Review Board (CIRB reference number: 2020/2058) with waiver of informed consent.

Study objectives

The primary objective of this study was to assess the prevalence of inappropriate antibiotic choice in an ambulatory cancer centre. The secondary objectives were to determine the appropriateness of dose, frequency and duration of appropriately prescribed antibiotics and to identify predictors for inappropriate antibiotic choice.

Study participants

Patients who were dispensed with at least one oral antibiotic from NCCS pharmacy between 1^st July to 30^th September 2019 were screened for eligibility. Inclusion criteria were patients of at least 21 years with an active or past cancer diagnosis. Exclusion criteria were patients with haematological malignancies or those who had received antibiotics prescribed by a non-NCCS physician, antibiotics prescribed for prophylaxis, standby or topical use. For patients who were prescribed with oral antibiotic(s) at more than one encounter during the study time period, only the first encounter was included. For patients included in the study, data were collected via retrospective review of electronic clinical charts and physicians’ clinical notes.

Algorithm for assessment of antibiotics appropriateness

Appropriateness of antibiotic choice was determined using a three-step algorithm as shown in Figure 1. Firstly, all antibiotic prescriptions were assessed to determine if the prescribed antibiotic was for empiric or culture-directed therapy. If the antibiotic was prescribed for empiric therapy, the institutional guideline; Singapore General Hospital (SGH) antibiotics guidelines (2017 version) were used to assess for appropriateness of antibiotics. The SGH guidelines were selected as references for assessment as NCCS patients are usually admitted to SGH in the event of complications arising from cancer treatment. These guidelines were developed based on international clinical practice guidelines, contextualised to the institution's antibiogram and formulary and are updated every 3 years. The 2020 version was not available at the point of this study. Secondly, in the absence of SGH antibiotic guidelines, the following guidelines were utilised to assess antibiotic appropriateness in descending order namely the National Comprehensive Cancer Network (NCCN) – Prevention and Treatment of Cancer-related Infections guideline followed by the Infectious Diseases Society of America (IDSA) guidelines and other internationally recognised clinical practice guidelines.^21–31

For culture-directed antibiotic, the definition of appropriate antibiotic choice was adapted from the World Health Organisation (WHO) AWaRe classification of antibiotics.³² The Duke Antimicrobial Stewardship Outreach (DASO) Network³³ was used to assess the appropriateness of culture-directed antibiotic therapy and took into account the oral antibiotics available in NCCS formulary (Table 1). In cases where cultures demonstrated bacterial susceptibility to multiple antibiotics, the antibiotic with the narrowest spectrum will be considered as appropriate, in the absence of antibiotic allergy and infection with MDRO in the previous 3 months.^34,35

Table 1.

Classification of Spectrum of antibiotics.^34,35

Narrow spectrum	Broad spectrum	Extended spectrum (include MDRO) and anti-pseudomonal
Amoxicillin	Amoxicillin/clavulanic acid	Ciprofloxacin
Cloxacillin	Azithromycin	Levofloxacin
Cefalexin	Cefuroxime
Clindamycin	Clarithromycin
Dapsone
Doxycycline
Metronidazole
Nitrofurantoin
Penicillin V
Sulfamethoxazole/trimethoprim
Tetracycline

For cases with appropriate choice of antibiotics and does not require consultation with the expert panellist, further assessment for appropriateness of dose, frequency and duration was performed. Based on the intended indication, the dose, frequency and duration of antibiotics were respectively assessed for appropriateness using either the SGH antibiotic guidelines or international clinical practice guides.

Lastly, for cases where antibiotic appropriateness cannot be ascertained using the criteria outlined above, an independent three-member expert panel comprising of an infectious disease (ID) physician, an ID specialist pharmacist and a medical oncologist was consulted. Relevant clinical information pertaining to these patients were consolidated and presented to the expert panel for evaluation. Each panellist evaluated these patients independently and was allowed to seek clarification or request for additional patient information. An overall rating of appropriate antibiotic choice was accorded when at least two panellists rated the patient as having received the appropriate antibiotic choice. Similarly for cases that were deemed as having received appropriate antibiotics, the expert panellists will also assess for appropriateness of dose, frequency and duration of prescribed antibiotics.

Statistical analyses

Descriptive statistics were used to report patients’ demographics, clinical and treatment information, characteristics of antibiotic prescriptions, appropriateness of antibiotics and all other outcomes. Sample size calculation was not performed as this was a baseline study intended to examine the antibiotic prescribing patterns at an outpatient oncology centre. The Cohen's Kappa (κ) test was used to analyse any inter-rater variability among the responses of the three-member expert panel. A κ-value of 0.01–0.2 will be interpreted as none to slight, 0.21-0.4 as fair, 0.41-0.6 as moderate, 0.61-0.8 as substantial and 0.81-1 as almost perfect agreement.³⁶ A univariate analysis was performed using either Fisher's exact or Chi-square for two-group categorical parameters including age (≥65 years vs. <65 years), gender, presence of non-oncologic immunosuppressive comorbidities, current cancer treatment intent (curative/palliative), total white cells (≥10 × 10⁹/L vs. <10 × 10⁹/L), procalcitonin (≥0.5mcg/L vs. <0.5mcg/L)³⁷ and C-reactive protein (≥10 mg/L vs. <10 mg/L)³⁸ while the two-way ANOVA was used for categorical parameters with more than two groups such as reasons for inappropriate antibiotics choice, cancer diagnosis, antibiotic indication and physician designation. Factors with a p-value of less than 0.25 in univariate analysis were subsequently included in the logistic regression analysis to determine predictors of inappropriate antibiotic choice.³⁹ The Hosmer-Lemeshow method was used to test for goodness of fit for the model. An a priori alpha (α) of 0.05 was used to define statistical significance. All statistical analyses were performed using IBM SPSS version 25 (SPSS Inc, Chicago, IL).

Results

A total of 815 patients were screened for eligibility; among which 404 were excluded due to various reasons stated in Figure 2. Four hundred and eleven patients were included in the study for appropriateness of choice of antibiotics.

Patients’ demographics and clinical information

Patient demographics and clinical information were as described in Table 2. The mean age of patients was 62.4 ± 12 years. There were slightly more females (59.4%) as compared to males (40.6%). The top three cancer diagnoses were breast (26.5%), lung (15.6%) and head and neck (13.6%) cancers. Nearly half of the patients (45.7%) received either chemotherapy alone or combination chemotherapy with immunotherapy, targeted or radiotherapy.

Table 2.

Patients’ demographics, clinical and treatment information.

Demographics	n = 411 (%)
Age (mean) ± SD, years	62.4 ± 12.0
Gender
Male	167 (40.6)
Cancer diagnosis ^*
Breast	109 (26.5)
Colon	50 (12.2)
Gastrointestinal^†	37 (9.0)
Genitourinary^†	13 (3.2)
Gynaecological^†	33 (8.0)
Head and neck	56 (13.6)
Hepatobiliary	31 (7.5)
Lung	64 (15.6)
Lymphoma/Sarcoma	13 (3.2)
Prostate	20 (4.9)
Others^‡	17 (4.1)
Presence of non-oncologic immunosuppressive comorbidities ^§	17 (4.1)
Presence of antibiotic allergy	51 (12.4)
Presence of MDRO infection in the last 3 months	6 (1.5)
Presence of any documented infection in the last 3 months	106 (25.8)
Current cancer treatment intent
Adjuvant/Neoadjuvant	172 (41.8)
Palliative	239 (58.2)
Current cancer treatment^\|\|
Active surveillance	52 (12.7)
Chemotherapy	153 (37.2)
Combination chemotherapy^¶	35 (8.5)
Hormonal therapy	17 (4.1)
Combination hormonal and immunotherapy	10 (2.4)
Immunotherapy	19 (4.6)
Radiotherapy	6 (1.5)
Targeted therapy	47 (11.4)

MDRO: Multi-drug resistant organisms.

Numbers do not add up to the total number of patients and percentages add up to more than 100% as some patients had multiple cancer diagnosis.

^†

Gastrointestinal include stomach and rectal. Genitourinary include bladder and renal. Gynaecological include cervical, endometrium, fallopian tube, ovarian and vaginal.

^‡

Bone, glioblastoma, melanoma and neuroendocrine tumours.

^§

Systemic lupus erythema, myasthenia gravis, Sjogren's syndrome, Crohn's disease, hepatitis B infection, rheumatoid arthritis, essential thrombocytosis and autoimmune myopathy.

^||

Numbers do not add up to the total number of patients and percentages add up to less than 100% as some patients were not on any current cancer treatment.

^¶

Combination chemotherapy refers to chemotherapy with immunotherapy/targeted therapy/radiotherapy.

Characteristics of prescribed antibiotics

The top 3 sites of infections were skin and soft tissue (32.8%), lower respiratory tract (16.1%) and genitourinary tract (14.8%). Majority of the patients were prescribed with a beta-lactam (71.5%), among which 286 (97.3%) received a penicillin and 8 (2.7%) received a cephalosporin (Table 3). Approximately equal number of antibiotic prescriptions were prescribed by senior physicians (senior consultants, consultants and associate consultants) (48.0%) and junior physicians (medical officers, resident physicians and registrars) (52.0%).

Table 3.

Characteristics of prescribed antibiotics.

Characteristics	n = 411 (%)
Types of infection
Gastrointestinal	21 (5.1)
Genitourinary	61 (14.8)
Respiratory (Lower)	66 (16.1)
Respiratory (Upper)	52 (12.7)
Skin and soft tissue	135 (32.8)
Unspecified source	30 (7.3)
Others^*	46 (11.2)
Antibiotics class ^†
Beta-lactam	294 (71.5)
Fluoroquinolone	64 (15.6)
Lincomycin	11 (2.7)
Macrolide	37 (9.0)
Sulfonamide	2 (0.5)
Tetracycline	24 (5.8)
Others^‡	19 (4.6)
Seniority of physicians
Medical officers	88 (21.4)
Resident physicians	50 (12.2)
Registrars	76 (18.5)
Associate consultants	31 (7.5)
Consultants	42 (10.2)
Senior consultants	124 (30.2)

Others include sinusitis, neutropenic fever and bacteraemia.

^†

Numbers do not add up to the total number of patients and percentages add up to more than 100% as some patients were prescribed with multiple antibiotics.

^‡

Others include nitrofurantoin and metronidazole.

Antibiotics choice

Appropriate

A total of 241 patients (58.6%) received appropriate choice of antibiotics. Inter-rater analysis was performed for the 58 patients referred to the expert panel, showed moderate agreement (κ = 0.539, p < 0.0001) between an ID consultant and an ID specialist pharmacist while fair agreement was observed between ID consultant/oncologist (κ = 0.268, p = 0.074) and ID specialist pharmacist/oncologist (κ = 0.230, p = 0.104). Among patients who received appropriate choice of antibiotics, majority also had the appropriate antibiotic dose [235 (97.5%)], frequency [238 (98.8%)] and duration [194 (80.5%)].

Inappropriate

Among patients who received inappropriate choice of antibiotics, commonly reported reasons for inappropriateness include the selection of a broader spectrum of antibiotic that was not recommended by either the SGH or international clinical practice guidelines (45.3%), absence of clear source of infections (31.8%) and presentation of viral instead of bacterial infection (16.5%). Subgroup analysis showed that among patients who received inappropriate antibiotics due to non-adherence to guidelines, majority received a broader spectrum antibiotic than required (Table 4).

Table 4.

Reasons for inappropriate antibiotic choices.

Reasons for Inappropriate Antibiotic Choices	Assessed based on Guidelines n = 142 (%)	Assessed by Expert Panel n = 28 (%)	p-value
Not adhering to SGH/international guidelines			0.072
Broader spectrum	61 (43.0)	5 (17.9)
Narrower spectrum	3 (2.1)	1 (3.6)
Overlapping coverage	4 (2.8)	3 (10.7)
Absence of a clear source of infection	40 (28.2)	14 (50.0)
Viral infection	23 (16.2)	5 (17.9)
Inappropriate choice based on list adapted from AWaRe and DASO	11 (7.7)	-

Predictors for inappropriate antibiotic choice

A univariate analysis was conducted and factors that met the pre-defined p < 0.25 for inclusion into the multivariate logistic regression modelling were the presence of non-oncologic immunosuppressive comorbidities, antibiotic allergy, types of infection, cancer diagnosis, presence of any documented infection in the last 3 months and presence of urine catheters (Table 5).

Table 5.

Univariate analysis on factors predicting inappropriate choice of antibiotics.

Variables	Appropriate Antibiotic Choice n = 241 (%)	Inappropriate Antibiotic Choice n = 170 (%)	p-value
Age (years)
≥65 years	111 (46.1)	71 (41.8)	0.388
Gender
Male	93 (38.6)	74 (43.5)	0.315
Presence of non-oncologic immunosuppressive comorbidities ^*	4 (1.7)	13 (7.6)	0.003
Cancer diagnosis ^{* †}
Breast	67 (27.8)	42 (24.7)	0.173
Colon	20 (8.3)	30 (17.6)
Gastrointestinal	27 (11.2)	20 (11.8)
Genitourinary	10 (4.1)	3 (1.8)
Gynaecological	17 (7.1)	16 (9.4)
Head and neck	33 (13.7)	23 (13.5)
Hepatobiliary	19 (7.9)	12 (7.1)
Lung	40 (16.6)	24 (14.1)
Prostate	10 (4.1)	10 (5.9)
Skin, soft tissue and bone	9 (3.7)	5 (2.9)
Others^‡	5 (2.1)	5 (2.9)
Treatment intent
Adjuvant/Neoadjuvant	101 (41.9)	71 (41.8)	0.977
Palliative	140 (58.1)	99 (58.2)
Presence of antibiotic allergy ^*	37 (15.4)	14 (8.2)	0.031
Presence of MDRO infection in the last 3 months ^*	2 (0.8)	4 (2.4)	0.236^\|\|
Presence of documented infection in the last 3 months ^*	68 (28.2)	38 (22.4)	0.181
Types of infection ^*
Gastrointestinal	8 (3.3)	13 (7.6)	<0.0001
Genitourinary	33 (13.7)	28 (16.5)
Respiratory (Lower)	28 (11.6)	38 (22.4)
Respiratory (Upper)	25 (10.4)	27 (15.9)
Skin and soft tissue	94 (39.0)	41 (24.1)
Others^§	53 (22.0)	23 (13.5)
Presence of hypotension	2 (0.8)	1 (0.6)	1.0^\|\|
Presence of fever ^¶	49 (20.3)	33 (19.4)	0.9
Presence of neutropenia ^#	177 (73.4)	123 (72.8)	0.881
Presence of febrile neutropenia ^¶	1 (0.4)	2 (1.2)	0.572
Presence of urine catheter ^*	2 (0.8)	4 (2.4)	0.236
Presence of central/peripheral lines	46 (19.1)	39 (22.9)	0.342
Presence of UFEME performed	25 (10.4)	22 (12.9)	0.42
Pyuria	16 (6.6)	11 (6.5)	0.642
Presence of chest x-ray performed	35 (14.5)	31 (18.2)	0.658
Consolidation	23 (9.5)	19 (11.2)	0.714
Presence of abdominal symptoms ^**	4 (1.7)	4 (2.4)	0.616
Seniority of physicians
Medical officer	57 (23.7)	31 (18.2)	0.528
Resident physician	32 (13.3)	18 (10.6)
Registrar	43 (17.8)	33 (19.4)
Associate consultant	19 (7.9)	12 (7.1)
Consultant	25 (10.4)	17 (10.0)
Senior consultant	65 (27.0)	59 (34.7)
Presence of concurrent radiotherapy	7 (2.9)	2 (1.2)	0.317
Total white cells (x 10⁹/L)
≥10	27 (11.2)	25 (14.7)	0.339
Procalcitonin (mcg/L) ^††
≥0.5	11 (4.6)	4 (2.4)	0.317
C-reactive protein (mcg/L)
≥10	17 (7.1)	8 (4.7)	0.404

FEME: Urine full examination microscopic examination MDRO: Multi-drug resistant organisms.

Variables included in the multivariate logistic regression modelling.

^†

Numbers do not add up to total patients and percentages add up to more than 100% due to some patients having multiple cancer diagnosis.

^‡

Glioblastoma, thymus and neuroendocrine tumours.

^§

Others include sinusitis, neutropenic fever and bacteraemia.

^||

Fisher exact.

^¶

Febrile as documented in patient's clinical notes while febrile neutropenia defined as presence of absolute neutrophil count less than 1 × 10⁹ cells/L and single temperature of >38.3°C or sustained temperature of ≥38^oC.¹

Neutropenia defined as absolute neutrophil count less than 1.0 × 10⁹ cells/L.¹

^**

Abdominal symptoms refer to any reported symptoms including pain, tenderness, nausea, vomiting and diarrhoea.

^††

Procalcitonin readings available only in 41 patients (26 patients with appropriate antibiotic choice) and percentages were expressed over the total patient population.

The logistic regression modelling had identified several significant predictors for inappropriate antibiotic choice. Factors identified were the presence of non-oncologic immunosuppressive comorbidities (OR = 4.890, 95% CI 1.556-15.369, p = 0.007), presence of antibiotic allergy (OR = 2.352, 95% CI 1.178-4.698, p = 0.015) and skin and soft tissue infections (SSTI) (OR = 2.004, 95% CI 1.276-3.146, p = 0.003) (Table 6). In contrast, lower respiratory tract infection was found to be significantly associated with appropriate antibiotic choice (OR = 0.380, 95% CI 0.214-0.656, p = 0.001). The regression model had an R² of 0.137 and the Hosmer-Lemeshow Chi square for the final model was 6.273, p-value = 0.617.

Table 6.

Multivariate analysis for predictors for inappropriate choice of antibiotics.

Variables	Odds Ratio	95% CI	p-value
Non-oncologic immunosuppressive comorbidities	4.890	1.556-15.369	0.007
Cancer diagnosis	1.026	0.959-1.097	0.454
Antibiotic allergy	2.352	1.178-4.698	0.015
Presence of MDRO infection in the last 3 months	0.398	0.066-2.408	0.316
Presence of documented infection in the last 3 months	1.503	0.914-2.469	0.108
Antibiotic indication
Genitourinary	0.890	0.500-1.585	0.694
Respiratory (Lower)	0.375	0.214-0.656	0.001
Respiratory (Upper)	0.593	0.326-1.08	0.088
Skin and soft tissue	2.004	1.276-3.146	0.003
Presence of urine catheter	0.435	0.074-2.557	0.357

Discussion

Our study demonstrated that at least 40% of antibiotics dispensed at an ambulatory oncology centre in Singapore were inappropriate. Majority of patients who received inappropriate antibiotics were due to discordance with clinical practice guidelines, prescribed with antibiotics without clear sources of infections and for viral infections. Subgroup analysis also did not reveal any statistical significance between patients assessed based on expert panel and respective guidelines. The high prevalence of inappropriate antibiotic prescribing observed in our study was congruent with other studies published in the literature. For instance, studies conducted in the general ambulatory patients in China and in the ambulatory oncology patients in the United States reported prevalence of inappropriate antibiotics to be 51% and 30% respectively.^17,40 Possible reasons for inappropriate antibiotics prescribed include the “just-in-case” mentality where antibiotics are prescribed given any possibilities that there may be an infection,⁴¹ physicians’ perceptions on patients’ demand for antibiotics,⁴² lack of physician awareness towards AMR, the increase societal affluence and thus, accessibility to broader-spectrum antibiotics as well as the potential knowledge gap of physicians towards alternative antibiotic options in managing patients with antibiotic allergy.⁴³

A higher rate of antibiotic inappropriateness was observed for their duration compared to dose and frequency. Possible explanation includes the presence of clinical decision support with pre-set dose and frequency within the prescribing system in NCCS, thus increasing the likelihood of physicians selecting the appropriate dose and frequency. However, for antibiotic duration, it is often less straightforward as appropriate antibiotic duration is influenced by indications and many other patient-specific factors. Physicians may prescribe a longer duration of antibiotics in immunocompromised cancer patients as there are concerns of having an increase in mortality with partially treated or unresolved infections⁴⁴ and the impracticability of a daily review in an ambulatory setting.

Potential solutions to improve appropriateness in antibiotic selection can include not just the implementation of periodic educational talks and case reviews together with feedback/audits on the prescribing patterns of antibiotics but also to adopt behavioural interventions such as having an automatic alert sent by the e-prescribing system prompting prescribers to justify indication when antibiotics are prescribed for patients for with diagnoses of acute respiratory infections.⁴⁵ This will educate and raise awareness towards inappropriate antibiotic prescribing among physicians and pharmacists.^46,47 Other longer-term solutions can include the adopting and implementing of ASPs for ambulatory setting. To facilitate ASPs in ambulatory setting, we can leverage on available technology to provide closer patient monitoring. Monitoring of patients’ vitals such as temperature, blood pressure and heart rate at home using various tracking devices such as smart watches cum a tele-consult can aid physicians in the assessment for improvement in patients’ clinical status. With such technology, closer patient surveillance can be achieved. Thus, physicians may be more reassured to de-escalate to a narrow-spectrum empiric antibiotics or to prescribe a shorter course of antibiotic therapy.^48,49

Our study had identified several patient-related factors such as having non-oncologic immunosuppressive comorbidities, presence of antibiotic allergy and SSTIs to be associated with inappropriate antibiotic choice. However, given the relatively poor pseudo-R² reported, it suggests the possibility that there were many other physicians’,^50–54 pharmacists’⁵² and patients’⁴⁷ related factors were not explored due to the nature of the study design involving a retrospective clinical chart review. Hence, we would recommend for future studies to be conducted to validate these predictors identified in our study and to identify other possible predictors. Nevertheless, these factors identified are still relevant at NCCS and can still aid in the design of several targeted strategies to optimise antibiotic use, thus reducing AMR in our cancer centre. To start off, periodic educational talks can be customised and tailor for physicians and pharmacists. These sessions can include how to access and clarify a drug (antibiotic) allergy, guidance on alternative therapies for patients with true antibiotic allergy as well as the incorporation of antibiotic guidelines or order sets for SSTIs into clinical decision support.⁵⁵ In order to reduce the prevalence of inappropriate antibiotic prescribing, a stepwise multi-pronged approach would be more beneficial due to prescriber buy-in and the ability to perform quality improvement assessments on each intervention.

Due to the paucity of evidence on the trend of antibiotic prescribing in the oncology patient population, our study focused on the antibiotics prescribing in this vulnerable population. Half of our patients received conventional chemotherapy (either as a single agent or in combination) that may potentially be myelosuppressive, resulting in a higher risk of infection. Infections can possibly delay chemotherapy treatment, thereby compromising patients’ treatment outcomes especially in an adjuvant setting.⁵⁶ Inappropriate antibiotics use may breed resistance and lead to infections that are harder to treat. This is evident in the rising number of infections especially MDROs and Clostridioides difficile, adverse effects, increased healthcare costs and the dearth of effective antibiotics against these pathogens. Thus, ASPs have also been recommended and established in numerous oncology centres to optimise the use of antibiotics.^{16,40,44,57–59}

Some strengths of our study include this being the first published study reporting on the prevalence and predictive factors for inappropriate choice of antibiotics in an ambulatory oncology setting. Thus, it contributes to the current limited evidence on antibiotic prescribing in the oncology patient population. Secondly, our study adopted a three-pronged step-wise approach involving institutional antibiotic guidelines, published international clinical practice guidelines as well as an independent three-member panel to assess appropriateness of antibiotic choice. This is in contrast to several outpatient studies,^17,40,57 where appropriateness was defined either based on a consolidated list of pre-defined guidelines for each diagnosis by an expert panel using available international guidelines¹⁷ or the matching of diagnosis from each outpatient visit to a pre-defined list of diagnosis established using the Chinese International Classification of Diseases (ICD), 10^th revision code that “almost always”, “sometimes” and “almost never” justify for antibiotics use.⁴⁰ Possible discrepancies in the choice of antibiotic between the pre-defined guidelines and the guidelines used by physicians in their daily prescribing may be present leading to inaccurate rating of appropriateness of antibiotics.

Secondly, our study had only included patients receiving the first encounter of antibiotic within the study period. Thus, it avoided multiple encounters of the same patient being reviewed for the primary infection which can falsely lead to a higher prevalence of inappropriate choice of antibiotics as seen in other study.⁴⁰ While our cross-sectional and systematic approach could possibly contribute to the observed lower prevalence of inappropriate antibiotic, at present, there is no gold standard which one can adopt to evaluate appropriateness of choice of antibiotics.

Nevertheless, our study had several limitations. Firstly, this was a retrospective cross-sectional study where the assessment of antibiotic appropriateness was solely based on physicians’ clinical documentation and patients’ history. The evaluation of antibiotic choice was highly dependent on the accuracy of the clinical documentation without the ability to verify. Secondly, it was a single centre study, thus it is uncertain if comparable prevalence of inappropriate choice of antibiotics will be observed in other ambulatory oncology centres. Thirdly, our definition of antibiotic appropriateness (narrowest spectrum in the absence of antibiotic allergy and infection with MDRO in the past 3 months) does not account for other factors that may affect antibiotic choice, e.g. drug interactions, co-infections (without microbiology results), patient tolerance and adherence issues. Given the retrospective nature of this study, such information were either not documented or not accessible for research purposes. Lastly, this was a baseline study focusing on the prevalence of inappropriate choice of antibiotics. Thus, it is not feasible to extensively investigate other physician-, pharmacist- or patient-related factors that potentially predicts for inappropriate antibiotics prescribing. Future studies looking specifically at predictors for inappropriate antibiotic choice may be considered to bridge the current knowledge gap.

Conclusion

Our study showed that inappropriate antibiotic prescribing is prevalent in an ambulatory oncology setting. Several predictive factors identified from our study can aid in the design of targeted interventions to optimise antibiotic prescribing in this vulnerable population, the oncology patients.

Footnotes

Acknowledgements

The authors would like to acknowledge the generous support of the three panellists, Consultant, Dr Benjamin Cherng from the Department of Infectious Disease, Singapore General Hospital, Associate Consultant, Dr Ryan Shea Tan from the Division of Medical Oncology, National Cancer Centre Singapore and ID Specialist Pharmacist, Ms Liew Yi Xin from the Department of Pharmacy, Singapore General Hospital.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Cindy Chew

References

Spellberg

Powers

Brass

, et al. Trends in antimicrobial drug development: implications for the future. Clin Infect Dis 2004; 38: 1279–1286.

Ribeiro da Cunha

Fonseca

Calado

CRC

. Antibiotic discovery: where have we come from, where do we go? Antibiotics (Basel) 2019; 8: 45. Published 2019 Apr 24.

Spellberg

Guidos

Gilbert

, et al. The epidemic of antibiotic-resistant infections: a call to action for the medical community from the infectious diseases society of America. Clin Infect Dis 2008; 46: 155–164.

Lieberman

. Appropriate antibiotic use and why it is important: the challenges of bacterial resistance. Pediatr Infect Dis J 2003; 22: 1143–1151.

Biggest Threats and Data: Centers for Disease Control and Prevention [Available from: https://www.cdc.gov/drugresistance/biggest-threats.html?CDC_AA_refVal=https%3A%2F%2Fwww.cdc.gov%2Fdrugresistance%2Fbiggest_threats.html.] Last accessed: 24 August 2021.

Antibiotic Resistance: A Global Threat. Centers for Disease Control and Prevention, National Center for Emerging and Zoonotic Infectious Diseases: Division of Healthcare Quality Promotion; 2020. [Available: https://www.cdc.gov/drugresistance/solutions-initiative/stories/ar-global-threat.html] Last accessed: 24 August 2021.

Antibiotic Resistance Threats in the United States, 2013. Centers for Disease Control and Prevention; U.S. Department of Health and Human Services. [Available: https://www.cdc.gov/drugresistance/pdf/ar-threats-2013-508.pdf] Last accessed: 24 August 2021.

De Silva

Jackson

Steer

. Infections, resistance patterns and antibiotic use in patients at a regional cancer centre. Intern Med J 2018; 48: 323–329.

Gudiol

Carratalà

. Antibiotic resistance in cancer patients. Expert Rev Anti Infect Ther 2014; 12: 1003–1016.

10.

Abbo

Ariza-Heredia

. Antimicrobial stewardship in immunocompromised hosts. Infect Dis Clin North Am 2014; 28: 263–279.

11.

Bodro

Gudiol

Garcia-Vidal

, et al. Epidemiology, antibiotic therapy and outcomes of bacteremia caused by drug-resistant ESKAPE pathogens in cancer patients. Support Care Cancer 2014; 22: 603–610.

12.

Samonis

Vardakas

Kofteridis

, et al. Characteristics, risk factors and outcomes of adult cancer patients with extensively drug-resistant Pseudomonas aeruginosa infections. Infection 2014; 42: 721–728.

13.

Peragine

Walker

SAN

Simor

, et al. Impact of a comprehensive antimicrobial stewardship program on institutional burden of antimicrobial resistance: a 14-year controlled interrupted time-series study. Clin Infect Dis 2020; 71: 2897–2904.

14.

Al-Omari

Al Mutair

Alhumaid

, et al. The impact of antimicrobial stewardship program implementation at four tertiary private hospitals: results of a five-years pre-post analysis. Antimicrob Resist Infect Control 2020; 9: 95

15.

Loo

Liew

Lee

, et al. Impact of antimicrobial stewardship program (ASP) on outcomes in patients with acute bacterial skin and skin structure infections (ABSSSIs) in an acute-tertiary care hospital. Infect Dis Ther 2015; 4: 15–25.

16.

Pillinger

Bouchard

Withers

, et al. Inpatient antibiotic stewardship interventions in the adult oncology and hematopoietic stem cell transplant population: a review of the literature. Ann Pharmacother 2020; 54: 594–610.

17.

Fleming-Dutra

Hersh

Shapiro

, et al. Prevalence of inappropriate antibiotic prescriptions Among US ambulatory care visits, 2010-2011. JAMA 2016; 315: 1864–1873.

18.

Steve Paulson. Cancer care migrates to outpatient setting. The Journal of Healthcare Contracting. May 2011. [Available from: https://www.jhconline.com/cancer-care-migrates-to-outpatient-setting-2.html] Last accessed: 24 August 2021.

19.

Bertucci

Le Corroller-Soriano

Monneur-Miramon

, et al. Outpatient cancer care delivery in the context of E-oncology: a French perspective on “cancer outside the hospital walls”. Cancers (Basel) 2019; 11: 19. Published 2019 Feb 14.

20.

National Cancer Centre Singapore. About Us. [Available from

https://www.nccs.com.sg/giving/about-us.] Last accessed: 24 August 2021.

21.

NCCN Clinical Practice Guidelines in Oncology. Prevention and Treatment of Cancer-Related Infections. Version 1.2020.

22.

Mermel

Allon

Bouza

, et al. Clinical practice guidelines for the diagnosis and management of intravascular catheter-related infection: 2009 update by the infectious diseases society of America. Clin Infect Dis 2009; 49: 1–45.

23.

Liver. EAftSot. EASL Clinical Practice Guidelines. The diagnosis and management of patients with primary biliary cholangitis. J Hepatol 2017; 67: 145–172.

24.

Stevens

Bisno

Chambers

, et al. Practice guidelines for the diagnosis and management of skin and soft tissue infections: 2014 update by the infectious diseases society of America. Clin Infect Dis 2014; 59: 147–159.

25.

Rubin

Ananthakrishnan

Siegel

, et al. ACG Clinical guideline: ulcerative colitis in adults. Am J Gastroenterol 2019; 114: 384–413.

26.

Feuerstein

Isaacs

Schneider

, et al. AGA Clinical practice guidelines on the management of moderate to severe ulcerative colitis. Gastroenterology 2020; 158: 1450–1461.

27.

Belum

Cercek

Sanz-Motilva

, et al. Dermatologic adverse events to targeted therapies in lower GI cancers: clinical presentation and management. Curr Treat Options Oncol 2013; 14: 389–404.

28.

Lieberthal

Carroll

Chonmaitree

, et al. The diagnosis and management of acute otitis media. Pediatrics 2013; 131: e964–e999.

29.

Martin DHNaJT. Common dental infections in the primary care setting. Am Fam Physician 2008; 77: 797–802.

30.

Punit Vaibhav Patel SKG, Amrita Patel. Periodontal abscess: a review. J Clin Diagn Res 2011; 5: 404–409.

31.

Krishnamoorthy

Relias

Sebastian

, et al. Management of regorafenib-related toxicities: a review. Therap Adv Gastroenterol 2015; 8: 285–297.

32.

World Health Organization. (2019) The 2019 WHO AWaRe classification of antibiotics for evaluation and monitoring of use.. World Health Organization; 2019. https://apps.who.int/iris/handle/10665/327957. License: CC BY-NC-SA 3.0 IGO.

33.

Duke Center for Antimicrobial Stewardship and Infection Prevention. Developing patient safety outcome measures and measurement tools for antibiotic stewardship programs. Metrics Guide 2017. A joint initiative between the CDC Foundation and Merck & Co., Inc., Kenilworth, NJ, USA.

34.

Lin

Huang

Pan

, et al. Duration of colonization with and risk factors for prolonged carriage of multidrug resistant organisms among residents in long-term care facilities. Antimicrobial Resistance & Infection Control 2017; 6 1–7 https://doi.org/10.1186/s13756-017-0246-4

35.

van Bijnen

den Heijer

Paget

, et al. The appropriateness of prescribing antibiotics in the community in Europe: study design. BMC Infect Dis 2011; 11: 293. doi: 10.1186/1471-2334-11-293.

36.

McHugh

. Interrater reliability: the kappa statistic. Biochem Med (Zagreb) 2012; 22: 276–282.

37.

Bouadma

Luyt

Tubach

, et al. Use of procalcitonin to reduce patients’ exposure to antibiotics in intensive care units (PRORATA trial): a multicentre randomised controlled trial. Lancet 2010; 375: 463–474.

38.

Pearson

Mensah

Alexander

, et al. Markers of inflammation and cardiovascular disease: application to clinical and public health practice: a statement for healthcare professionals from the centers for disease control and prevention and the American heart association. Circulation 2003; 107: 499–511.

39.

Phungoen

Kraisriwattana

Apiratwarakul

, et al. Predictors of appropriate antibiotic Use in bacteremia patients presenting at the emergency department. Antibiotics (Basel) 2019; 8: 42. Published 2019 Sep 9.

40.

Zhao

Wei

, et al. Appropriateness of antibiotic prescriptions in ambulatory care in China: a nationwide descriptive database study. Lancet Infect Dis 2021; 21: 847–857.

41.

Goff

File

. The risk of prescribing antibiotics “just-in-case” there is infection. Seminars in Colon and Rectal Surgery 2018; 29: 44–48.

42.

Kohut

Keller

Linder

, et al. The inconvincible patient: how clinicians perceive demand for antibiotics in the outpatient setting. Fam Pract 2020; 37: 276–282.

43.

Yam

ELY

Hsu

Yap

, et al. Antimicrobial resistance in the Asia pacific region: a meeting report. Antimicrob Resist Infect Control 2019; 8: 202

44.

Aitken

Nagel

Abbo

, et al. Antimicrobial stewardship in cancer patients: the time is Now. J Natl Compr Canc Netw 2019; 17: 772–775.

45.

Persell

Doctor

Friedberg

, et al. Behavioral interventions to reduce inappropriate antibiotic prescribing: a randomized pilot trial. BMC Infect Dis 2016; 16: 73.

46.

Shively

Buehrle

Clancy

, et al. Prevalence of inappropriate antibiotic prescribing in primary care clinics within a veterans affairs health care system. Antimicrob Agents Chemother 2018; 62: e00337–18.

47.

King

Fleming-Dutra

Hicks

. Advances in optimizing the prescription of antibiotics in outpatient settings. Br Med J 2018; 363: k3047.

48.

Dias

. Paulo silva Cunha J. Wearable health devices-vital sign monitoring. Systems and Technologies. Sensors (Basel) 2018; 18: 2412. Published 2018 Jul 25.

49.

Soon

Svavarsdottir

Downey

, et al. Wearable devices for remote vital signs monitoring in the outpatient setting: an overview of the field. BMJ Innov 2020; 6: 55–71.

50.

Zetts

Stoesz

Garcia

, et al. Primary care physicians’ attitudes and perceptions towards antibiotic resistance and outpatient antibiotic stewardship in the USA: a qualitative study. BMJ Open 2020; 10: e034983.

51.

Walsh

Taffe

Sacca

, et al. Risk factors for unnecessary antibiotic prescribing for acute respiratory tract infections in primary care. Mayo Clin Proc Innov Qual Outcomes 2020; 4: 31–39.

52.

Dempsey

Businger

Whaley

, et al. Primary care clinicians’ perceptions about antibiotic prescribing for acute bronchitis: a qualitative study. BMC Fam Pract 2014; 15: 94.

53.

Sanchez

Roberts

Albert

, et al. Effects of knowledge, attitudes, and practices of primary care providers on antibiotic selection, United States. Emerg Infect Dis 2014; 20: 2041–2047.

54.

Fleming-Dutra

Bartoces

Roberts

, et al. Characteristics of primary care physicians associated with high outpatient antibiotic prescribing volume. Open Forum Infect Dis 2018; 5: ofx279.

55.

Barlam

Cosgrove

Abbo

, et al. Implementing an antibiotic stewardship program: guidelines by the infectious diseases society of America and the society for healthcare epidemiology of America. Clin Infect Dis 2016; 62: e51–e77.

56.

Taha

Vinograd

Sakhnini

, et al. The association between infections and chemotherapy interruptions among cancer patients: prospective cohort study. J Infect 2015; 70: 223–229.

57.

Krantz

Zier

Stohs

, et al. Antibiotic prescribing and respiratory viral testing for acute upper respiratory infections Among adult patients at an ambulatory cancer center. Clin Infect Dis 2020; 70: 1421–1428.

58.

Laxminarayan

Duse

Wattal

, et al. Antibiotic resistance-the need for global solutions. Lancet Infect Dis 2013; 13: 1057–1098.

59.

Bernatz

Safdar

Hetzel

, et al. Antibiotic overuse is a major risk factor for clostridium difficile infection in surgical patients. Infect Control Hosp Epidemiol 2017; 38: 1254–1257.