Antimicrobial stewardship in Australian hospitals: how does compliance with antimicrobial stewardship standards compare across key hospital classifications?

BACKGROUND: There is little information on the prevalence and type of antimicrobial stewardship (AMS) activities that are currently occurring in Australian hospitals.
OBJECTIVES: To determine what AMS activities are currently occurring in Australian hospitals, identify gaps in compliance with the Australian Commission on Safety and Quality in Health Care (ACSQHC) standards and determine perceived barriers and enablers for implementing AMS programmes.
METHODS: A cross-sectional survey open to all Australian hospitals, conducted online and available to hospitals between November 2016 and July 2017.
RESULTS: Responses were received from 254 hospitals. Compliance with ACSQHC AMS essential activities was high, except for essential activity 3 (post-prescription reviews), which was conducted by only 39% of respondent hospitals. Importantly, compliance varied by hospital remoteness classification for all activities except essential activity 1 (availability and endorsement of guidelines) and additional activity 4 (publishing antimicrobial susceptibility data annually), with major city hospitals having the highest compliance across all activities. The three most frequently reported barriers to implementing AMS programmes were a lack of training and education, lack of pharmacy resources and a lack of willingness from medical officers to change.
CONCLUSIONS: Due to low response rates from certain hospital groups, the survey results are not generalizable to all Australia hospitals. This survey has identified that several gaps in compliance still exist and outlines the need to address lower AMS compliance in hospitals located outside major cities. The key barriers and enablers for AMS programme implementation identified should be used to inform future strategies.

Introduction

As the pipeline for new antibiotics is running dry and resistance to currently available antimicrobials is increasing, the need to reduce the development and spread of antimicrobial resistance has become a pressing global public health concern. A major factor contributing to the increasing rate of antimicrobial resistance is the overuse and misuse of antibiotics., Antimicrobial stewardship (AMS) programmes encourage the appropriate use of antimicrobials and have been shown to decrease antimicrobial consumption as well as reduce infections due to specific antimicrobial pathogens., In 2011, the Australian Commission on Safety and Quality in Health Care (ACSQHC) developed guidelines for AMS programmes in Australian hospitals. These guidelines outline five essential activities (implementing clinical guidelines, antimicrobial formularies, post-prescription reviews, performance monitoring and selective reporting of susceptibility testing) and four additional AMS activities (prescriber education, point-of-care interventions, electronic decision support and publishing antimicrobial susceptibility data annually) (Figure 1). In 2013, these guidelines became part of the accreditation standards for all Australian hospitals.

Figure 1.

The Australian Commission of Safety and Quality in Health Care’s Antimicrobial Stewardship strategies for Australian hospitals.

AMS programmes have been implemented in Australian hospitals to varying degrees. The evidence suggests that, amongst Victorian hospitals, AMS activities are poorly implemented in private hospitals and in public regional and remote hospitals compared with public metropolitan hospitals. These findings align with two similar surveys in the USA, which found that rural hospitals were less likely to have AMS policies and had the lowest rates AMS programme uptake compared with larger urban hospitals. Other studies have shown that public hospitals in the state of Queensland and Australian children’s hospitals have some key AMS activities implemented; however, additional resources are required to meet all accreditation standards., These studies provide some insight into AMS programmes within their specific study settings, outlining that AMS activities vary between different hospital types.

The existing literature suggests that the two most commonly perceived barriers to achieving and implementing a fully compliant AMS programme in Australia are a lack of education and a lack of resources., These barriers are consistent with studies in the USA, Canada and New Zealand.,, In regional and rural settings, a lack of specialist support was found to be a barrier to implementing AMS programs. The literature regarding enablers for implementing AMS programmes in Australian hospitals is more limited. Regional hospitals report that flatter governance structures, a greater sense of pride and good telehealth services are key enablers. Acknowledging the need for assistance and readiness to consult national guidelines were key enablers for Australian tertiary hospitals.

At present, there is no information regarding the prevalence of current AMS activities occurring in hospitals at a national scale across Australia. Moreover, there is insufficient understanding of variation in compliance by key hospital classifications, such as geographical (hospital jurisdiction and remoteness classification) and financial (hospital funding type). Identifying how AMS activities vary across hospital classifications at a national level is crucial to improving the quality of AMS programme implementation.

The National Centre for Antimicrobial Stewardship (NCAS) conducted a national AMS survey to address this gap. The aims of this study were to: (i) outline AMS activities occurring in Australian hospitals; (ii) identify gaps in compliance across key hospital characteristics; and (iii) identify key barriers and enablers to meeting hospital accreditation standards for AMS.

Methods

A cross-sectional, point-prevalence survey design was used. The study focused on Australian hospitals and was made available online between November 2016 and July 2017. Jurisdictional AMS leaders (i.e. clinical leads for AMS at the state departments of health) were notified about the survey and helped recruit participants by disseminating the survey to hospitals within their jurisdiction. Participants were also recruited via e-mail [e.g. registrants on the National Antimicrobial Prescribing Survey (NAPS) contact list] and on online discussion groups for infectious diseases physicians, microbiologists and pharmacists. Survey respondents were recruited through multiple platforms to improve response rates. To ensure the appropriate person completed the survey, the information letter specified that the respondent should be the best person at the hospital to respond to a survey regarding AMS practices. Due to the varying resources at different hospitals the ‘best person to respond’ will differ, thus no one profession could be targeted. Respondents were required to insert their hospital name and answer on behalf of their hospital. All hospitals in Australia, as defined by the Australian Institute for health and Welfare (AIHW), who utilize antibiotics were eligible to participate and there were no exclusion criteria. The survey was conducted through SurveyMonkey®, and all surveys were completed voluntarily.

The national AMS survey was developed by NCAS and consisted of 36 closed- and 4 open-ended questions; of these, 30 questions were compulsory and 6 were optional. Questions covered hospital geographic location, funding type, types of AMS activities, governance structures, resources and workplace capacity, as well as perceived barriers and enablers to implementing AMS (the survey is available as Supplementary data at JAC-AMR Online). The survey was initially designed through an iterative process with end users for a 2013 study of Victorian Hospitals. Following this study, further feedback was received and changes were made accordingly to maximize the surveys validity for its use in this study.

Data cleaning was undertaken to remove respondents who registered for the survey but provided no data, multiple respondents from the same hospital and respondents who represented more than one hospital (a hospital network response). Following cleaning, data were coded to generate a binary variable (compliant and non-compliant) for the majority of ACSQHC AMS activities, using 13 survey questions (Table S1). AMS essential activity 5 and additional activity 2 were not assessed as no questions covered this content. For essential activity 5, information was required from microbiology laboratories. For additional activity 2, this was reliant on essential activity 3 and more in-depth questions relating to this were not asked.

Respondent hospitals were categorized by geographical (hospital jurisdiction and remoteness), financial (hospital funding type) and peer group classification type. Jurisdictions refer to the Australian state or territory in which the hospital is located. Remoteness classifications were derived from the Australian Bureau of Statistics’ classification system. Funding type was defined as either public or private, as recorded by the AIHW. Hospital peer groupings refer to a classification system where hospitals with similar characteristics, such as bed number, case mix and funding type, are grouped together. Hospital peer groups were defined in accordance with the AIHW classification system, and hospitals were sorted into 12 peer groupings (public principal referral, public acute A, public acute B, public acute C, public acute D, public other, public women’s and children’s, private acute A, private acute B, private acute C, private acute D and private other).

Response rate was calculated by dividing the number of respondents by the number of eligible hospitals. Hospital compliance was measured using AMS activities outlined by the ACSQHC. Statistical analyses were conducted to determine if compliance varied by hospital remoteness classification and funding type using Fisher’s exact test, with the significance level set at P < 0.05. Perceived barriers and enablers were calculated using descriptive statistics. This study was approved by the Melbourne School of Population and Global Health Human Ethics Advisory Group (Ethics ID: 1851295.1).

Results

A total of 357 survey responses were received and 103 were excluded from analysis. Fifty-nine were excluded due to being secondary responses from a responding hospital, 36 were excluded as they represented a hospital network and 8 were excluded as no responses to the survey questions were received, leaving 254 hospitals available for analysis. The survey had a response rate of 27%, with 951 Australian hospitals eligible for the survey. There was large within-group variation in response rates across all hospital classifications; the characteristics of respondents are outlined in Table 1. The most common respondent profession type was AMS/infectious disease pharmacist (36%) (Table S2).

Table 1.

Characteristics of survey respondents

Characteristic	Number of respondents (% of total respondents)	Response rate (% respondents of all eligible Australian hospitals within subgroup)
Hospitals, by jurisdiction
Australian Capital Territory	3 (1)	33
New South Wales	111 (44)	38
Northern Territory	1 (1)	17
Queensland	35 (14)	19
South Australia	6 (2)	5
Tasmania	4 (2)	14
Victoria	80 (32)	38
Western Australia	14 (6)	13
Hospitals, by remoteness classification
major city	105 (41)	30
inner regional	79 (31)	33
outer regional	60 (24)	27
remote and very remote	10 (4)	8
Hospitals, by funding type and peer group
public
principal referral	23 (9)	79
public acute A	40 (16)	61
public acute B	27 (11)	61
public acute C	54 (21)	38
public acute D	33 (13)	17
very small	19 (8)	14
women’s and children’sa	6 (2)	43
otherb	16 (6)	24
all public hospitals	218 (85)	31
private
private acute A	8 (3)	36
private acute B	9 (4)	25
private acute C	8 (3)	16
private acute D	4 (2)	6
private otherb	7 (3)	10
all private hospitals	36 (14)	14
Total	254 (100)	27

Women’s and children’s hospitals include hospitals classified as: combined women’s and children’s hospitals; women’s hospitals; and children’s hospitals.

Other hospitals include hospitals classified as: mixed sub-acute and non-acute; other acute specialized hospitals; drug and alcohol hospitals; other acute specialized hospitals; other day procedure hospital; private rehabilitation hospitals; public rehabilitation hospitals; psychiatric hospitals; and mixed day procedure hospitals.

Respondent hospitals’ compliance with the essential and additional ACSQHC AMS activities and strategies by funding type and remoteness classification are presented in Table 2. Due to poor response rates from some jurisdictions and peer groups this analysis was not presented in Table 2.

Table 2.

Number and percentage of hospitals compliant with Australian Commission of Safety and Quality in Health Care’s antimicrobial stewardship guidelines by hospital funding type and remoteness classification

Hospital classification	Compliant hospitals, n (%)
	Essential activity 1(availability and endorsement of guidelines)	Essential activity 2(antimicrobial formulary)	Essential activity 3(post-prescription reviews)	Essential activity 4(performance monitoring)	Additional activity 1(AMS staff education)	Additional activity 3(electronic decision support)	Additional activity 4(antimicrobial susceptibility data)
All hospitals	218 (89)	193 (77)	95 (39)	204 (85)	145 (60)	40 (17)	127 (53)
Funding type
public	188 (89)	171 (79)	83 (39)	175 (84)	133 (64)	39 (19)	104 (50)
private	30 (91)	22 (67)	12 (36)	29 (91)	12 (38)	1 (3)	23 (72)
Fisher’s exact test	near 1	0.13	near 1	0.43	0.006	0.04	0.02
Remoteness classification
major city	91 (93)	94 (92)	58 (60)	86 (90)	72 (74)	30 (31)	59 (61)
inner regional	70 (91)	60 (77)	27 (35)	65 (86)	42 (55)	7 (9)	39(51)
outer regional	48 (81)	32 (53)	9 (15)	48 (83)	24 (41)	2 (3)	24 (41)
remote and very remote	9 (90)	7 (70)	1 (10)	5 (50)	7 (70)	1 (10)	5 (50)
Fisher’s exact test	0.14	<0.001	<0.001	0.02	<0.001	<0.001	0.13

Compliance with essential activities

Essential AMS activity 1 (availability and endorsement of ‘Therapeutic Guidelines: Antibiotic’) was undertaken in almost all hospitals (89%) (Table 2). This high level of compliance was consistent across all hospital subgroups.

A large proportion, 77% of respondents, were compliant with AMS essential activity 2 (having an antimicrobial formulary in place that included restrictions on broad-spectrum antimicrobials) (Table 2). Compliance did not vary by hospital funding type (P = 0.13) but did vary by remoteness (P < 0.001), with the highest compliance being seen in major city hospitals (92%) and lowest in outer regional hospitals (53%).

Compliance with essential AMS activity 3 (post-prescription reviews) was poor: only 39% of respondents reported conducting these reviews. Compliance with this activity varied by hospital remoteness (P < 0.001), but not by funding type (P value near 1) (Table 2). Compliance decreased with increasing remoteness, with only 15% of outer regional and 10% of remote and very remote hospitals conducting post-prescription reviews.

Compliance with essential AMS activity 4 (performance monitoring through regular antimicrobial audits and feedback to prescribers) was high, with this activity occurring in 85% of respondent hospitals (Table 2). This level of compliance was consistent across hospital funding types (P = 0.43); however, it did vary by hospital remoteness (P = 0.02). Again, compliance with performance monitoring decreased with increasing remoteness.

Compliance with additional activities

Most hospitals undertook additional AMS activity 1 (education on antimicrobial prescribing to medical officers, pharmacists and nurses), with 60% of hospitals reporting compliance with this activity (Table 2). Compliance varied by hospital funding type (P = 0.006) with only 38% of private hospitals conducting education on antimicrobial prescribing compared with 64% of public hospitals. Compliance also varied by remoteness (P < 0.001), with the highest compliance being seen in major city hospitals (74%) and the lowest in outer regional hospitals (41%).

Additional activity 3 (electronic decision support) had poor compliance, with only 17% of respondents reporting compliance with this activity (Table 2). Compliance varied by remoteness (P < 0.001) with major city hospitals having a much higher compliance compared with hospitals outside major cities. Compliance also varied by funding type (P = 0.04) with only 3% of respondent private hospitals having electronic decision support, compared with 19% of public hospitals.

Approximately half of respondents, 53%, were compliant with additional activity 4 (publishing antimicrobial susceptibility data annually) (Table 2). Compliance did not vary by remoteness (P = 0.13) by did vary by funding type (P = 0.02), with 72% of respondent private hospitals publishing antimicrobial susceptibility data annually, compared with only 50% of public hospitals (Table 2).

Barriers and enablers

Note that respondents were able to select multiple barriers and enablers. The three most frequently reported barriers were a lack of training and education on antimicrobial use, a lack of willingness from medical officers to change their prescribing practices and a lack of pharmacy resources, with 60%, 59% and 54% of respondents selecting these barriers, respectively (Table 3).

Table 3.

Perceived barriers to implementing antimicrobial stewardship

Barriers to implementing AMS	Respondents selecting barrier (%)
Lack of training and education in antimicrobial use	60
Lack of willingness from medical officers to change their prescribing practices	59
Lack of pharmacy resources	54
Lack of an electronic medication management system	49
Lack of dedicated funding for an antimicrobial stewardship physician	46
Lack of dedicated funding for an antimicrobial stewardship pharmacist	44
Lack of infectious diseases or microbiology services	37
Lack of enforcement by facility management / executive	35
Lack of support from senior clinicians at the facility	34
High level of transient or seconded staff	33
Lack of leadership to promote antimicrobial stewardship at the facility	30

The three most frequently reported enablers were the availability of ‘Therapeutic Guidelines: Antibiotic’; the National Safety and Quality Health Service Standards 2012, Standard 3.14; and the National Antimicrobial Prescribing Survey, with 88%, 69% and 68% of respondents selecting these enablers, respectively (Table 4).

Table 4.

Perceived enablers to implementing antimicrobial stewardship

Enablers to implementing AMS	Respondents selecting enabler (%)
The Therapeutic Guidelines: Antibiotic, Therapeutic Guidelines Limited	88
National Safety and Quality Health Service Standards 2012, Standard 3.14; AMS	69
The National Antimicrobial Prescribing Survey (NAPS)	68
Antibiotic Awareness Week	58
AMS in Australian Hospitals 2011, The ACSQHC	55
The Antimicrobial Stewardship Clinical Care Standards 2014, ACSQHC	55
Resources provided by ACSQHC	44
The National Prescribing Service (NPS)	43
The National Antimicrobial Utilisation Surveillance Program (NAUSP)	41
The National Centre for Antimicrobial Stewardship (NCAS)	32
Resources provided by the department of health in your state or territory	26
Discussion forums within your professional group	24
Resources provided by your professional society	19
Resources provided by international professional organizations	13

Discussion

This survey outlined key gaps in AMS programmes across Australian hospitals. Only 39% of respondent hospitals were identified as conducting post-prescription reviews, essential activity 3. This low compliance should be the focus of future interventions and may be due to a lack of pharmacy resources, which was a frequently reported barrier. An international survey across 67 countries reported similarly low rates of prescription reviews, with 43% conducting outcome reviews at day 3 post prescription. Compliance with additional AMS activities (AMS staff education, electronic decision support and publishing antimicrobial susceptibility) is another key gap. This may be due to hospitals focusing resources towards achieving essential activities. Additional activity 3 had particularly low compliance, with only 17% of hospital having electronic decision support available. This is likely due to the high cost of electronic decision support. An international survey across 67 countries identified similar rates of electronic prescribing.

With the exception of essential activity 3 (post-prescription reviews), compliance with essential activities (availability and endorsement of guidelines, antimicrobial formulary and performance monitoring) was high. Importantly, compliance with essential activities 2 and 4 (antimicrobial formulary and performance monitoring) was lower in hospitals outside major cities.

Compliance varied by remoteness for measured AMS activities except essential activity 1, availability and endorsement of guidelines (P = 0.14) and additional activity 4, publishing antimicrobial susceptibility data annually (P = 0.13) (Table 2). For AMS activities that varied by remoteness classification, compliance was consistently highest in major cities. This finding is consistent with other literature that indicates that more regional hospitals tend to have lower rates of conducting AMS activities compared with metropolitan hospitals. Due to the low response rate from remote and very remote hospitals the external validity of results from these hospitals is limited.

Funding type only had an impact on compliance for additional activities (Table 2). These finding are inconsistent with the 2013 Victorian study that found private hospitals had lower compliance across all AMS activities in comparison to public metropolitan hospitals. Due to the limited responses from private hospitals in this study, the comparability of these study results is limited.

When the compliance of Victorian hospitals in this survey (Table S3) was compared with the results of the 2013 Victorian study, it was found that compliance was much higher for essential activity 2 (antimicrobial formulary) and additional activity 4 (antimicrobial susceptibility data published annually). This indicates an improvement on these activities over time, likely due to the incorporation of AMS guidelines into hospital accreditation standards in 2013, which occurred after the Victorian study was completed.

The three most frequently reported barriers—a lack of training and education, lack of pharmacy resources and a lack of willingness from medical officers to change—are consistent with the wider literature.,, Due to the inconsistency in methodology in the wider literature, the enablers identified in this study were not easily comparable to other literature. Further investigation using consistent methodology should be undertaken to understand enablers, especially in hospital groups with lower compliance.

This survey has several limitations. As with any self-reported survey, the response of one person might not reflect the true nature of what is happening in the hospital. A further limitation is that this survey only captures whether the AMS activity occurs or not and is not able to comment on the quality of the AMS activities occurring in the responding hospitals. Low response rates from private and ‘remote and very remote’ hospitals as well as hospitals from South Australia, Western Australia, Tasmania, Northern Territory and Queensland may limit the generalizability of study results and therefore should not be considered representative of all Australian hospitals. It is likely the low response rates in remote and very remote hospitals reflects a lack of AMS resources or engagement. AMS within these hospitals was a relatively new practice at the time this survey was completed compared with some larger major city hospitals, who have been undertaking some form of AMS for some years allowing for strong relationships with AMS communities and resources to be developed. Future efforts should focus on strengthening the engagement of remote hospitals and other low responding hospital groups to better connect them with appropriate resources. Where variation in response rates was too large and likely marred by confounding, this analysis was not included, such as for hospital jurisdictions and hospital peer groups.

As AMS has only been mandated in Australian hospital accreditation since 2013, it is important to understand how compliance with these standards has changed over time. Similar surveys should be repeated at regular intervals in the future to monitor uptake and to measure compliance with the newly updated 2018 standards. Future national surveys should focus on better engaging ‘hard to reach’ hospitals, including remote, very remote, smaller and private hospitals as well as aiming to understand the quality of programme and activity implementation.

This survey has identified that while there have been some improvements over time, several gaps in AMS compliance still exist. This survey has also identified a specific need to address and further investigate lower AMS compliance in hospitals located outside major cities. Furthermore, this survey has identified key barriers and enablers for Australian AMS programme implementation, this will assist in informing future improvements to AMS and strategies. Continued research into AMS strategies within all healthcare settings is required to understand the changing nature of how these different strategies are being utilized. Identification of the key gaps in AMS compliance as well as how compliance varies by hospital classification is important for deciding how resources are directed and uptake of AMS activities is improved, so that the use of antimicrobials is optimized.

Click here for additional data file.