Passive surveillance of antimicrobial resistance in Queensland public hospitals: the basis for a national system?

This report published in Communicable Diseases Intelligence, Volume 28, Number 2 details an antimicrobial resistance surveillance system developed by Queensland Health Pathology and Scientific Services with the capacity to handle national data and discusses if this could be used as the basis for national surveillance.

Page last updated: 04 July 2004

Graeme R Nimmo,1 Jonathan Fong2

Introduction | Methods | Results | Discussion | Acknowledgements | References

Abstract

Australia currently has no system of passive surveillance of antimicrobial resistance in spite of the importance of surveillance in identifying and defining emergent resistance being generally accepted. Queensland Health Pathology and Scientific Services have developed flexible software for passive surveillance with the capacity to handle national data. The system imports raw data strings in delimited ASCII text format into a relational database and screens to exclude duplicates before the processing of the cumulative susceptibility data. It allows considerable flexibility in inquiry parameters and has the ability to 'drill down' to individual laboratory results. Examples of analytical output are given for 49,169 unique isolate results obtained in all Queensland Health Pathology Service laboratories from 1 January to 30 June 2003. The system could form the basis of a national system for passive antimicrobial resistance surveillance. Commun Dis Intell 2004;28:230-235.

Top of page

Introduction

The emergence of increasing levels of resistance in a growing number of major pathogens has led to the recognition of antimicrobial resistance as an important public health issue. Government sponsored reports in a number of countries including the United Kingdom, the United States of America and Australia have identified the need for action to deal with this emergent problem.1,2,3 Furthermore, the World Health Organization has acknowledged the global nature of this problem and has recommended a concerted international approach to controlling the emergence and spread of antimicrobial resistance.4 The importance of surveillance in identifying and defining emergent resistance is generally accepted. The Australian report (JETACAR)3 recommended that a comprehensive surveillance system be established in Australia incorporating both active and passive components.3

Prior to the publication of JETACAR passive surveillance of resistance in a broad range of pathogens was conducted for a number of years by the National Antimicrobial Resistance Surveillance Program (NARSP) by collation of results obtained in 29 pathology laboratories both public and private.5 However, due to the laborious nature of data collection and collation, NARSP publications lagged behind testing by several years at least. This tended to limit their utility in identifying emergent problems. The introduction of The Surveillance Network™ (TSN®), an American commercial computerised surveillance system, into Australia in 1998 promised for some time to fill the gap vacated by NARSP and to provide national passive surveillance data in close to 'real-time'.6 However, the decision by TSN® to withdraw from Australia in 2003 leaves us without a national passive surveillance system.

Queensland Health Pathology and Scientific Services (QHPSS) have made the provision of passive antimicrobial resistance surveillance data to clinicians within Queensland public hospitals a high priority. The purpose of passive surveillance is to provide estimates of the prevalence of resistance phenotypes based on specimens submitted to clinical laboratories. The aim of the current project is to provide timely cumulative susceptibility data to interested health care professionals at a local, regional and state level.

This paper describes the creation of comprehensive state-wide network of passive surveillance incorporating all susceptibility data generated in our laboratories and gives some examples of its reporting output.

Top of page

Methods

Queensland Health Pathology Service (QHPS) consists of a hierarchical, networked system of 32 laboratories (Figure) which vary in size from small remote laboratories serving rural communities to large multi-divisional laboratories serving metropolitan tertiary referral hospitals. It provides laboratory services to all public hospitals but one in Queensland. Public hospitals in Queensland form part of health districts that in turn belong to three zones. The Northern Zone serves a population of 596,725, the Central Zone 1,365,076 and the Southern Zone 1,624,094 according to 2001 census data.

Figure. Location of Queensland Health Pathology Service laboratories

Figure. Location of Queensland Health Pathology Service laboratories

A single laboratory information system (LIS) (AUSLAB™ , PJACC, Melbourne) is networked to all laboratories from a central computer. Twenty-four of the laboratories perform antimicrobial susceptibility testing. Since July 2002, all susceptibility testing is performed according to National Committee for Clinical Laboratory Standards (NCCLS)7 or using automated methods based on NCCLS methods (Vitek®, bioMerieux, Missouri or MicroScan®, Dade Behring, Illinois). Specimen, isolate and susceptibility data ('S', 'I' or 'R' calls) are downloaded from the LIS using an 'autodump' function. A data field is included to specify the test method thus allowing data derived from any method to be captured and analysed separately if required. The raw data (strings in delimited ASCII text format) are imported into a relational database (Microsoft™ SQL Server 2000™ ) and processed through a screening algorithm to remove duplicates. Results are stored in separate tables in the relational database. Data are presented with a 3-tiered-architecture web application developed using Microsoft™ Dot Net™ technologies. A web browser is used as the user interface allowing access through the Queensland Health intranet or through remote access secure internet connection. The 'unique' isolate definition used for duplicate removal uses the following parameters: identical patient identifiers (name, patient record number, date of birth), identical organism name, identical susceptibility pattern and isolation within five days of a previous 'identical' isolate. Specimen type is also included in the algorithm from specimen category inquiries. Data from patients with multiple isolates were audited to ensure that the algorithm selected the first isolates of a particular profile and excluded all duplicates.

The database can be queried for all specimens or for particular categories of specimens. Specimens are categorised as blood, cerebrospinal fluid, ear/nose/throat, enteric, genital, respiratory, tissue/fluid/pus/prostheses, urine, infection control screening and other. Ad hoc inquiries are processed according to the following parameters: testing laboratory, health care facility, ward address, inpatient/outpatient status, zone, district year/month, organism, antimicrobial, and specimen type. Data for clinical isolates and infection control screening isolates can be analysed separately. Reports state the number of isolates tested and the percentage susceptible.8 For demonstration purposes the results reported here reflect results as recorded in the LIS. These would not necessarily have appeared in the pathology report.

Top of page

Results

Susceptibility results were available for 52,563 isolates between 1 January and 30 June 2003. The screening algorithm excluded 6.5 per cent of isolates leaving 49,169 in the active database including 48,096 clinical and 1,073 infection control screening isolates. Some of the flexible data analysis capabilities of the system are demonstrated in the tables.

Tables 1 and 2 show the susceptibilities of common gram-negative and gram-positive blood culture isolates from all laboratories respectively. The variation of the proportion of blood culture isolates of Staphylococcus aureus that were methicillin-resistant (MRSA) in the three zones is shown in Table 3. Differences in the susceptibility of inpatient and outpatient MRSA isolates at one Southern Zone hospital are seen in Table 4. The ability to differentiate between clinical and infection control screening isolates is demonstrated by data from the same hospital (Table 5). Summary data for extended spectrum beta-lactamase (ESBL) producing Klebsiella pneumoniae and vancomycin resistant enterococci (VRE) for the three zones are shown in Tables 6 and 7 respectively. Cumulative susceptibility to a variety of antimicrobials and reduced susceptibility to penicillin of Streptococcus pneumoniae isolated from sterile and non-sterile sites are shown in Table 8.

Table 1. Antimicrobial susceptibilities of the 10 most common aerobic gram-negative isolates in blood cultures in all laboratories, January to June 2003

Organism Amikacin Amoxycillin Cefotaxime Ceftazidime Cephalothin Ciprofloxacin Gentamicin Meropenem Timentin
%S n %S n %S n %S n %S n %S n %S n %S n %S n
Acinetobacter baumannii
90.2
41
3.1
32
42.9
42
50.0
38
0.0
32
70.5
44
70.5
44
62.5
33
72.7
44
Enterobacter aerogenes
100.0
13
92.3
13
100.0
13
100.0
10
Enterobacter cloacae
98.1
53
100.0
55
84.2
57
100.0
44
Escherichia coli
100.0
484
54.7
525
99.8
492
100.0
428
69.4
520
99.0
514
99.4
526
100.0
371
89.2
454
Klebsiella oxytoca
100.0
27
0.0
28
96.4
28
100.0
21
63.0
27
100.0
28
96.4
28
100.0
17
84.6
26
Klebsiella pneumoniae
100.0
142
0.0
150
97.9
143
97.5
122
91.0
145
97.2
145
96.7
150
100.0
112
94.0
133
Proteus mirabilis
100.0
30
90.6
32
100.0
31
100.0
25
75.9
29
93.6
31
96.9
32
100.0
21
100.0
30
Pseudomonas aeruginosa
97.4
114
94.0
117
95.0
119
94.2
120
95.0
99
89.2
120
Serratia marcescens
97.4
38
100.0
37
92.1
38
100.0
29
Stenotrophomonas maltophilia
60.0
15
43.8
16
33.3
3
0.0
1
37.5
16
      Agent not recommended or not tested for this species.      

Table 2. Antimicrobial susceptibilities of the 10 most common aerobic gram-positive isolates in blood cultures in all laboratories, January-June 2003 (excluding common skin flora)

Organism Amoxycillin Ciprofloxacin Clindamycin Erythromycin Fusidic acid Gentamicin Penicillin G Rifampicin Vancomycin
%S n %S n %S n %S n %S n %S n %S n %S n %S n
Enterococcus faecalis
100.0
59
21.2
33
100.0 59
Staphylococcus aureus
98.3
342
100.0
242
89.1
367
95.6
296
99.1
349
16.3
368
100.0
342
100.0 367
Staphylococcus aureus (MRSA)
44.9
69
61.5
52
22.2
72
95.8
72
36.6
71
0.0
72
95.8
72
100.0 72
Streptococcus agalactiae (Group B)
100.0
24
88.9
36
100.0
44
100.0 33
Streptococcus milleri group
100.0
14
95.8
24
100.0
36
100.0 26
Streptococcus mitis
100.0
5
80.0
10
73.3
15
100.0 10
Streptococcus oralis
100.0
7
66.7
9
73.3
15
100.0 9
Streptococcus pneumoniae
84.2
101
90.5
105
100.0 93
Streptococcus pyogenes (Group A)
100.0
13
96.9
32
100.0
46
100.0 28
Streptococcus sp. Group G
100.0
10
92.3
26
100.0
32
100.0 27
      Agent not recommended or not tested for this species.    

Top of page

Table 3. Proportion of methicillin-resistant Staphylococcus aureus (MRSA) among blood culture isolates of Staphylococcus aureus and proportion of gentamicin susceptible MRSA, January to June 2003

Zone Total Staphylococcus aureus Methicillin-resistant Staphylococcus aureus MRSA gentamicin susceptible
n % %
Northern
94
11
11.7
72.7
Central
143
24
16.8
17.4
Southern
130
22
16.9
31.8

Table 4. Antimicrobial susceptibility of inpatient and outpatient isolates of methicillin-resistant Staphylococcus aureus at a metropolitan teaching hospital, January to June 2003

Antibiotic Inpatients Outpatients
%S n %S n
Vancomycin
100.0
267
100.0
25
Fusidic acid
89.9
267
80.0
25
Clindamycin
74.9
267
84.0
25
Ciprofloxacin
43.1
267
76.0
25
Gentamicin
38.2
267
72.0
25
Tetracycline
33.7
267
72.0
25
Erythromycin
22.1
267
44.0
25
Rifampicin
78.2
266
92.0
25
Mupirocin
99.2
127
100.0
3

Top of page

Table 5. Comparison of susceptibility of all methicillin-resistant Staphylococcus aureus isolates with those from clinical specimens and those from infection control screening specimens at a metropolitan teaching hospital, January to June 2003

Antimicrobial All isolates Clinical isolates Screening isolates
%S n %S n %S n
Ciprofloxacin
45.9
292
57.6
165
30.7
127
Clindamycin
75.7
292
80.0
165
70.1
127
Erythromycin
24.0
292
31.5
165
14.2
127
Fusidic acid
89.0
292
90.9
165
86.6
127
Gentamicin
41.1
292
48.5
165
31.5
127
Mupirocin
99.2
130
100.0
17
99.1
113
Rifampicin
79.4
291
81.8
165
76.2
126
Tetracycline
37.0
292
39.4
165
33.9
127
Vancomycin
100.0
292
100.0
165
100.0
127

Table 6. Antimicrobial susceptibility of extended spectrum beta-lactamase producing Klebsiella pneumoniae isolated in Queensland Health Zones, January to June 2003

Antibiotic Northern zone Central zone Southern zone
%S n %S n %S n
Amikacin
100.0
36
98.0
51
100.0
12
Amoxycillin
0.0
40
0.0
50
0.0
12
Cefotaxime
20.5
39
7.5
40
81.8
11
Ceftazidime
2.6
38
5.7
35
27.3
11
Cephalothin
2.6
38
0.0
50
0.0
12
Ciprofloxacin
43.6
39
89.8
49
16.7
12
Co-trimoxazole
13.2
38
15.7
51
16.7
12
Gentamicin
5.0
40
2.0
51
0.0
12
Imipenem
100.0
37
97.9
48
100.0
12
Meropenem
100.0
31
97.6
41
100.0
11
Netilmicin
nt
nt
66.7
3
100.0
11
Trimethoprim
9.1
33
9.8
41
27.3
11

nt Not tested

Top of page

Table 7. Vancomycin resistant enterococci isolated from screening specimens in Queensland Health zones, January to June 2003

Zone Enterococcus faecium (van B phenotype) Enterococcus faecalis (van B phenotype)
Northern
0
0
Central
28
2
Southern
1
2

Table 8. Antimicrobial susceptibilities of Streptococcus pneumoniae in all laboratories, January to June 2003

Antimicrobial Sterile sites Non-sterile sites
%S %I n %S %I n
Chloramphenicol
97.3
74
98.5
401
Co-trimoxazole
64.3
56
74.1
325
Erythromycin
82.6
109
85.1
578
Penicillin G
89.3
4.5
112
81.1
9.0
586
Tetracycline
90.9
77
85.0
454
Vancomycin
100.0
101
100.0
501

Top of page

Discussion

We have endeavoured to display the versatility of the antibiogram software by presenting data derived from the entire state, the three Queensland Health zones and from an individual institution. The data are presented as recorded as the system does not at present include software to automatically identify improbable results. The current version of software does however provide a 'drill down' feature which allows individual anomalous results to be identified by laboratory number and testing laboratory for follow up.

We presented data concerning some key endemic and emerging resistant organisms. MRSA is of particular interest in Queensland due to the emergence of non-multiresistant strains causing severe community acquired infections.9,10 Gentamicin susceptibility has been used as a surrogate marker for these strains and Table 3 shows marked differences in the gentamicin susceptibility of MRSA in the three zones. This suggests that community strains are probably most common in the Northern Zone and least common in the Central Zone. Examination of MRSA susceptibilities at one metropolitan teaching hospital in Table 4 demonstrates that outpatient isolates are more susceptible to non-beta-lactam antimicrobials, which is also in keeping with community acquisition of non-multiresistant strains. The varying cumulative susceptibility results displayed in Table 5 demonstrate the importance of separating results of infection control screening isolates from clinical isolates when reporting cumulative susceptibilities, in this case for MRSA.

Comparison of susceptibilities of ESBL producing K. pneumoniae to cephalosporins in the three zones in Table 6 illustrates another potential pitfall in interpreting cumulative susceptibility results from different laboratories. The marked differences in susceptibility recorded here is due to differing reporting practices between laboratories: some record results as tested and suppress them while others record all as resistant and report them. Clearly, knowledge of reporting practices for organism/antibiotic combinations where susceptibility phenotypes are not reliable indicators of clinical utility, is an important element in reporting and interpreting cumulative data.

Analysis of infection control screening for VRE showed a large number of isolates of van B E. faecium in the Central Zone (Table 7). Drilling down revealed that this was due to an outbreak in one institution only. The emergence of resistance to penicillin and other antimicrobials in S. pneumoniae has been evident in Australia for over a decade.11 Table 8 shows that, while isolates from blood and CSF are, as expected, more susceptible than isolates from non-sterile sites, the proportion of resistant isolates (penicillin MIC ≥2 mg/L) from sterile sites is 6.2 per cent, which is cause for concern.

The method of data transfer employed makes this passive surveillance system adaptable to output from any modern LIS. The system also allows for internet access with appropriate security. These features suggest it could be relatively easily adapted to provide the basis for a national system of passive surveillance. National input into specification of such a system would be required and a mechanism for providing this through the Australian Group for Antimicrobial Resistance has been proposed. While the Commonwealth has stated its commitment to the surveillance approach suggested by JETACAR,12 a practical and cost effective solution to the requirement for passive surveillance is yet to be implemented. We suggest that the system described would, with appropriate modification, satisfy the requirement for national passive surveillance of antimicrobial susceptibility.

Top of page

Acknowledgement

We acknowledge the contribution of the Microbiology Discipline Working Party and the Auslab Support Unit of QHPSS.

Top of page

References

1. Health Education and Human Services Division. Antimicrobial resistance: data to assess public health threat from resistant bacteria are limited. Washington, D.C.: United States General Accounting Office, 1999.

2. House of Lords Select Committee Sub-group on Antimicrobial Resistance. Seventh report: Resistance to Antibiotics and Other Antimicrobial Agents. London, UK: Stationery Office, 1998.

3. Joint Expert Technical Advisory Committee on Antimicrobial Resistance. The Use of Antibiotics in Food-producing Animals: Antibiotic-resistant Bacteria in Animals and Humans. Canberra: Commonwealth Department of Health and Aged Care and Commonwealth Department of Agriculture, Fisheries and Forestry-Australia, 1999.

4. World Health Organization. WHO Global Strategy for the Containment of Antimicrobial Resistance. Switzerland: WHO, 2001.

5. Bell J, Turnidge J. National Antimicrobial Resistance Surveillance Program: 1994. Canberra: National Centre for Disease Control, 1998.

6. Turnidge J, McCarthy LR, Master RN, Kepner DE, Weslock J. TSN Database Australia, a new tool to monitor antimicrobial resistance in Australia. Commun Dis Intell 2003;27 Suppl:S67-S69.

7. National Committee for Clinical Laboratory Standards. Performance Standards for Antimicrobial Susceptibility Testing; twelfth informational supplement. NCCLS document M100-S12. Pennsylvania, USA: National Committee for Clinical Laboratory Standards, 2002.

8. National Committee for Clinical Laboratory Standards. Analysis and Presentation of Cumulative Antimicrobial Susceptibility Test Data; Proposed Guideline. NCCLS document M39-P [ISBN 1-56238-422-8]. 940 West Valley Road, Suite 1400, Wayne, Pennsylvania 19087-1898 USA: National Committee for Clinical Laboratory Standards, 2000.

9. Munckhof WJ, Schooneveldt J, Coombs GW, Hoare J, Nimmo GR. Emergence of community-acquired methicillin-resistant Staphylococcus aureus (MRSA) infection in Queensland, Australia. Int J Infect Dis 2003;7:259-264.

10. Nimmo GR, Schooneveldt J, O'Kane G, McCall B, Vickery A. Community acquisition of gentamicin-sensitive methicillin-resistant Staphylococcus aureus (MRSA) in south-east Queensland. J Clin Microbiol 2000;38:3926-3931.

11. Collignon PJ, Bell JM. Drug-resistant Streptococcus pneumoniae: the beginning of the end for many antibiotics? Australian Group on antimicrobial Resistance (AGAR). Med J Aust 1996;164:64-67.

12. Commonwealth Department of Health and Aged Care. The Commonwealth Government Response to the Report of the Joint Expert Technical Advisory Committee on Antimicrobial Resistance (JETACAR). Canberra: Commonwealth of Australia, 2000.

Top of page

Author affiliations

1. Director of Microbiology, Queensland Health Pathology Service, Princess Alexandra Hospital, Brisbane, Queensland

2. Scientist, Queensland Health Pathology and Scientific Services, Coopers Plains, Brisbane, Queensland

Corresponding author: Dr Graeme R Nimmo, Director of Microbiology, Queensland Health Pathology Service, C/-Princess Alexandra Hospital, Brisbane QLD 4102. Telephone: +61 7 3240 2389. Facsimile: +61 7 3240 5786. Email: Graeme_Nimmo@health.qld.gov.au


This article was published in Communicable Diseases Intelligence, Volume 28 No 2, June 2004.

Communicable Diseases Intelligence subscriptions

Sign-up to email updates: Subscribe Now

This issue - Vol 28 No 2, June 2004