Application of local treatment guidelines/pathways, hospital antiinfective formulary, formulary restriction and approval requirements
The guideline development group recommends:
Developing and updating local treatment guidelines, clinical pathways, and an antiinfective formulary is one of the ABS team’s chief responsibilities. The antiinfective formulary should be based on national and international guidelines as well as on the local/regional pathogen and resistance patterns, and possibly drug costs. Drugs on the antiinfective formulary should be categorised according to recommended versus reserve or special compounds. In addition, these should be tagged with special prescription status and be subject to approval and preauthorisation requirements. The antiinfective formulary is updated at least yearly based on therapy guidelines and whenever necessary and approved by the Therapeutics and Drugs Committee (A).
Adherence to guidelines regarding substance selection, dosing, route and duration of treatment may improve clinical outcome in terms of mortality, as well as treatment duration and length of hospital stay. To ensure adherence, users should be involved in developing the guidelines and be educated through audits of antiinfective use or antiinfective point-of-care chart reviews (A).
Individualising antiinfective prescriptions with or without special approval requirements improves targeted therapy and reduces inappropriate treatment. Various possibilities for implementation have been described and should be used, from simple antimicrobial order forms to highly differentiated antiinfective request forms that may be subject to specific time limits or limited to certain hospital areas (A). Guideline-based antiinfective drug use or use of individual defined substances can be controlled by this means, thus minimising consumption, costs and adverse drug events.
Restricting whole substance classes can—by shifting to an alternative substance—prove to be an effective strategy for controlling nosocomial infections and the development of critical resistance levels; accordingly, antiinfective restriction ought to be targeted (B). At the same time, routine surveillance of antibiotic consumption and locally prevalent pathogens and their susceptibility patterns should be performed to detect possible adverse effects of the strategy in time (A).
Local treatment guidelines and clinical pathways are established and regularly updated by the ABS team with the involvement of the ABS representatives delegated from other clinical departments. National and international guidelines, the patient mix and local microbiology and resistance patterns should be taken into account. The established or revised local guidelines should have institution-wide validity for which consensus must be obtained. The treatment guidelines are presented to the Therapeutics and Drugs Committee and the Hospital Infection Control Committee. It is recommended to provide local treatment guidelines in electronic or pocketbook format and to ensure acceptance among users through training and education [
88]. Without these measures guideline adherence is rather poor, and effects in terms of improving clinical outcomes or other endpoints remain small [
89].
Treatment guidelines or clinical pathways can improve outcomes related to mortality, length of hospital stay and duration of treatment [
90,
91]. High adherence to guidelines or clinical pathways, e.g. for management of community-acquired or nosocomial pneumonia, can be achieved with training and education. Thus, mortality can be decreased and the medium duration of therapy and hospital stay can be shortened by 1.7–6.8 days, while antiinfective usage is reduced by up to 77 % [
92‐
98]. Various strategies of treatment optimisation have been studied for community-acquired or nosocomial pneumonia [
99‐
102] and have partly been addressed in international and national guidelines. Their implementation in local guidelines, guideline adherence assumed, can help to avoid that therapy is either too broad or too long. An American and a French observational study have shown that involving physicians in the development of local guidelines can improve acceptance. When local consensus guidelines were posted on the intranet and regularly distributed to physicians and presented in departmental staff meetings, guideline-conform management of nosocomial pneumonia increased from 46 to 81 %, and 14-day mortality dropped from 23 to 8 % [
103]. In a study of endocarditis, compliance with antimicrobial therapy improved from 23 to 62 % and 1-year mortality significantly decreased from 19 to 8 % [
104] (Table
1). Numerous new investigations on improving guideline compliance have shown that institutionalising guidelines can optimise the quality of therapy in different categories (e.g. dose adjustment to renal function, parenteral-to-oral conversion, timely administration) by about 10 or more percent [
105‐
110].
Table 1
Examples for use of treatment guidelines and clinical pathways
| Observational study (II) | Patients with community-acquired pneumonia (58 patients before intervention, 58 patients after intervention) | Establishment of guidelines for the diagnosis and management of nosocomial pneumonia | Mortality Proportion of patients with guideline-conforming treatment | Lower mortality rate at 14 days (23 vs 8 %, p = 0.03) Increase in the number of patients treated in conformity with guidelines (46 vs. 81 %, p < 0.01) |
Botelho-Nevers et al. [ 104] | Observational study (II) | Patients with infectious endocarditis (173 patients before intervention, 160 patients after intervention) | Establishment of treatment guidelines for management of infectious endocarditis | Mortality Guideline adherence (compound selection, duration of treatment) | Lower 1 year mortality (18.5–8.2 %, HR 0.41; 95 % CI, 0.21–0.79, p = 0.008) Lower hospital mortality (12.7–4.4 %, p = 0.007) Increase in guideline adherence: compound selection (31.6 % auf 95 %, p < 0.001) Compound selection and duration of treatment (22.7 % auf 61.8 %, p < 0.001) |
| Randomised, controlled study (i) | Patients with community-acquired pneumonia in the emergency room of a hospital (nine hospitals with clinical pathway, 10 hospitals without clinical pathway) | Establishment of a clinical pathway for treatment of community-acquired pneumonia in the emergency room of nine hospitals | Mortality Length of hospital stay Duration of treatment Proportion of patients with monotherapy | No difference in mortality Shorter length of hospital stay by 1.7 days (6.1–4.4 days, p = 0.04) Shorter duration of treatment by 1.7 days (6.3–4.6 days, p = 0.01) Increase in the proportion of patients with monotherapy (27–64 %, p < 0.001) |
| Randomised, controlled study (I) | Patients with ventilator-associated pneumonia (39 patients treated in accordance with a risk score-based clinical pathway, 42 patients received standard therapy) | Establishment of risk score-based clinical pathway | Mortality Length of hospital stay (ICU) Detection of MDR pathogens Duration of treatment, costs | No difference in mortality Shorter length of hospital stay (ICU) by 5.3 days (14.7–9.4 days; p = 0.04) Reduced detection of MDR pathogens (38–14 %, p = 0.017 Shorter duration of treatment (9.8–3 days, p = 0.0001) Lower treatment costs (640$–259$, p = 0.0001) |
| Observational study (II) | Patients with ventilator-associated pneumonia (50 patients before intervention, 52 patients after intervention) | Establishment of a treatment guideline for management of ventilator-associated pneumonia | Mortality Length of hospital stay, antibiotic therapy complying with guidelines, Duration of treatment | No difference in mortality Increase in the proportion of antimicrobial therapy conforming to guidelines from 48 to 94.2 % (p < 0.001) Shorter duration of treatment from 14.8 days ± 8.1 days to 8.6 ± 5.1 (p < 0.001) |
Clinical pathways complement local treatment guidelines, often taking into account diagnostic algorithms and risk scores. They are designed as a flowchart to simplify and improve the management of patients with infectious diseases. In a controlled, multi-centre Canadian study a risk score (PSI, pneumonia severity index)-based clinical pathway was instituted, addressing criteria for inpatient admission, sequential therapy and discharge of patients with community-acquired pneumonia. Although patients in the “experimental” arm had more severe disease, hospital stay and duration of parenteral antibiotic therapy was significantly shortened in this patient group, and the patients received monotherapy significantly more often without negative impact on mortality. Within this framework, an Australian study showed an approximately 10 % reduction in the use of broad-spectrum antibiotics [
93,
111,
112]. Similar results were achieved by a more recent observational study in the UK, where introduction of a risk score (CURB-65)-based clinical pathway for treatment of community-acquired pneumonia influenced prescribing behaviour. As expected, CURB65-guided therapy resulted in an overall reduction in the prescription of cephalosporins and macrolides by 19 and 14 %, respectively, without negatively affecting outcome (30-day mortality, clinical response, treatment outcome). There was a corresponding increase in use of aminopenicillin monotherapy, and guideline compliance increased from 25 % to over 60 % [
113].
Acceptance and implementation of treatment guidelines not only improves by involving users in guideline development. Other supplemental ABS strategies such as repetitive education, training and audits of antibiotic prescribing with feedback to the prescriber improve acceptance and adherence [
114]. This is shown by a controlled before-and-after study in which adherence was consistently improved by a combination of interventions involving distribution of information packs to staff, repeated compilation of prescription data and educational sessions followed by reminders in the form of posters [
98]. Implementation of a uniform guideline for perioperative prophylaxis including recommendations for choice of agent, dosage and timing resulted in annual antimicrobial cost savings of approximately USD 112,000 in a 1400-bed hospital [
115].
The institutional antiinfective formulary is established by the pharmacist in the ABS team based on therapeutic efficacy, toxicity and cost. Drugs of the formulary should be categorised into recommended versus reserve or special compounds depending on local treatment guidelines. Graphical overview with alerts (traffic light system), information on daily therapeutic costs or restrictions on use is advisable. Adding information on special prescription or approval requirements is desirable. Besides information on agent and trade names, these lists contain information on the recommended daily dose, including dose adjustments in regard to organ impairment (Table
2). The antiinfective formulary must be passed by the Therapeutics and Drugs Committee. The formulary has an immediate influence on prescribing behaviour [
116].
Table 2
Example of a formulary
Caution should be exercised in controlling antibiotic use via the formulary alone without an indication-based treatment concept and concomitant surveillance of antibiotic consumption and resistance. It was for instance observed that by adding levofloxacin to the antiinfective formulary fluoroquinolone use subsequently increased substantially, resulting in a higher rate of MRSA infection. When an alert was inserted next to the fluoroquinolone selections on the electronic order entry screen, indicating alternative antibiotic agents in accordance with local guidelines, levofloxacin use decreased again by 50 % from 12 to 6 DDD/100 patient-days and the MRSA infection rate decreased again from 1.37 to 0.63 cases per 1000 patient-days [
117]. Similar effects have been observed for other substances and classes and pathogens [
118,
119].
Individualised antiinfective orders with or without approval requirements extend from simple to highly differentiated, computer-assisted order forms with an automatic prescription stop after a defined time (so-called “automatic stop order”). These can be agent, patient or indication based, temporary or limited to certain hospital areas. Individualised antiinfective orders present an effective tool to quickly and effectively influence prescribing behaviour. Special order forms or approval requirements are usually implemented for broad-spectrum antibiotics, new/expensive substances or substances requiring extensive consultation. They require justification for prescription, which must be evaluated prior to approval, and can effectively control use and costs. These substances are separately marked in the antiinfective formulary.
Many older prospective before-and-after trials dating from the 1980s and 1990s documented that restricting use of new and expensive cephalosporins generated cost savings of between 19 and 46 %, and reduced consumption by up to 50 % [
6,
120‐
124]. Significant cost reductions being achieved through an antimicrobial-restriction policy are less commonly observed in recent years, because numerous antibiotics have lost patent protection. Nevertheless, more recent studies showed continuing effectiveness regarding reduction in antibiotic consumption of as much as 54 %. [
23,
125‐
127]. Newer research on restricting use of broad-spectrum antibiotics yielded a monthly reduction from 137 to 72 DDD/100 cases or from 181 to 102 DDD/1000 patient-days, respectively. Overall, after implementation, the ABS programme delivered effective cost savings of USD 300,000 p.a. (corresponding to net savings of USD 2350/100 cases or 2182/1000 patient-days, respectively). [
59].
Use of special order forms limiting antibiotic duration has proved to be particularly effective within the field of perioperative antibiotic prophylaxis. In several prospective, quasi-experimental before-and-after studies the effect of automatic stop order forms on antibiotic consumption, costs and guideline adherence to avoid extended prophylaxis was evaluated. By educational training, an overall 20–30 % improvement in guideline adherence was observed with respect to choice of drug and duration of antimicrobial use, with one study also showing improvement in appropriate timing of perioperative antibiotic prophylaxis before incision. This resulted in a reduction of surgical site infections from 3.2 to 1.9 %, a reduction in cost of USD 3000/100 patient-days and a reduction in consumption of approximately 20 DDD/100 patient-days [
128‐
132]. Other equally effective automated stop orders limiting total duration of treatment (e.g. 14 days) or restricting duration of reserve drugs such as vancomycin or carbapenems (72 h for empiric therapy, 7 days for therapeutic indication) have been described. Treatment beyond was only possible following consultation with the infectious diseases specialist or pharmacist. As a result, consumption of these substances was reduced by 10–25 % [
125,
133,
134].
Specific programmes to restrict antimicrobial use can minimise nosocomial infections (e.g.
C.
difficile) and the increase of resistant pathogens (ESBL, MRSA) by a rapid and marked alteration in consumption. However, such programmes are usually only temporary and lack sustainable efficacy [
22]. Restriction strategies are adopted in coordination with the Therapeutics and Drugs Committee, Hospital Infection Control Committee, the pharmacy, and hospital management. Timely and continuous surveillance of consumption, infectious diseases and resistance data are to be assured, to monitor compliance, but also to be able to rapidly identify possible negative impacts. The importance of instituting a programme for the surveillance of antimicrobial use including unrestricted antibiotics, cost and the development of resistance demonstrated by a prospective quasi-experimental observational study at a 450-bed hospital in Greece. In the study, use of carbapenems, third-generation cephalosporins, and fluoroquinolones was restricted based on a national recommendation in context of growing resistance among Gram-negative microorganisms. As a result, ciprofloxacin and ceftazidime consumption decreased as desired by 28 and 42 %, respectively. Subsequently, susceptibility of
P. aeruginosa (32–45 %) and
E. coli (77–84 %) to ciprofloxacin increased. On the other hand, susceptibility of
K. pneumoniae to ciprofloxacin (80–60 %) and ceftazidime (61–46 %) continued to decrease. Of note, piperacillin/tazobactam use increased by 271 % and overall costs and consumption were 12–13 % higher than before intervention [
135,
136].
Programmes restricting use of cephalosporins and fluoroquinolones have been repeatedly examined for their “ecological” effects [
137‐
140]. Multicenter controlled investigations in France show a 90 % reduction in fluoroquinolone use after introduction of a time-limited restriction, resulting in a significant reduction in MRSA. Reintroduction of fluoroquinolones was associated with a significant increase in MRSA compared to the previous period [
141,
142]. A new study from France shows that even less restrictive fluoroquinolone use (20 % reduction) combined with improved hand hygiene also reduces the rate of MRSA (moderate) and at the same time impacts positively on resistance of
P. aeruginosa to fluoroquinolones [
143]. Other new studies demonstrate effects of changes in fluoroquinolone prescribing practice on
C. difficile-associated diarrhoea [
144‐
148]. The effects, however, are not always due to the fluoroquinolone reduction alone.
Design and implementation of education, training and information
The guideline development group recommends:
Targeted education, training and information are essential elements of any ABS programme. They provide the foundation of knowledge needed to promote more rational use of antibiotics and reasonable microbiological diagnostic, and to improve acceptance of ABS programmes. They have the objective of optimising the therapeutic and diagnostic management of patients with infection through greater adherence to recommendations. They should preferably take place as an active training measure rather than in the form of passive communication of information (A).
Education, training and information in different formats and on various topics should be offered repeatedly as they are not sustainable as a one-off measure. They should be organised in agreement and integration with local ABS programmes (A).
Education, training and information should be independent of commercial interests, whereby the hospital administration is responsible for implementing and financing the measures (A).
Education, training and information are essential elements of every ABS programme. Overall, in a systematic review active clinician education in the form of lectures, seminars, “bedside teaching” demonstrated greater effectiveness than passive education techniques like posters, pocket cards or written prescription recommendations [
149]. Two multicenter, randomised, controlled, and some before-and-after studies demonstrated that an educational intervention improved compliance with guideline-recommended diagnostic, therapeutic and prophylactic measures and resulted in a reduction in the number of non-indicated treatments [
150‐
152]. By educational training of nurses and medical staff inappropriate submission of urine cultures decreased from 2.6 to 0.9 per 1000 patient-days; treatment of asymptomatic bacteriuria was reduced from 1.7 to 0.6 per 1000 patient-days, while in another study a significant reduction from 74 to 17 % was seen [
153,
154]. In Canadian long-term health care facilities, a 1-year long educational intervention involving repeated mailing of antibiotic guidelines with feedback on individual antibiotic prescribing behaviour of urinary tract infection, pneumonia, skin and soft tissue infection and sepsis resulted in a significant 64 % reduction of nonadherent treatment compared to control facilities [
155]. Nonadherent antibiotic prescriptions remained lower during follow-up, although after termination of the educational intervention, the effect was no longer significant compared to control facilities. A similar effect was achieved in a study on an educational programme for guideline-based treatment of respiratory tract infections in emergency departments, in which 1 year post-intervention a 10 %, albeit non-significant reduction in antibiotic consumption was still documented compared to sites without intervention [
156].
With the aim of reducing extended use of perioperative antibiotic prophylaxis by means of information disseminated by e-mail, poster and lectures, 12 Australian hospitals succeeded in rapidly and effectively limiting the duration of antibiotic prophylaxis to maximally 48 h, thus achieving considerably lower costs, which more than outweighed the costs of the 1-year intervention [
157]. However, the effect of the intervention rapidly declined with time, as seen in other studies [
158]. Repeated guideline-based educational interventions are necessary. They were shown to be particularly effective in optimising perioperative antibiotic prophylaxis [
6,
159]. In an Argentine multi-step ABS programme involving training and formulary restriction, antimicrobial consumption could be reduced from 43 to 28 DDD/100 patient-days, resulting in substantial savings (>900.000 USD) over 18 months. During the training period a significant increase in the rate of prescriptions based upon microbiology results (27–63 %) was found, and use of ceftriaxone and carbapenems subsequently more than halved [
30]. The combination of one-on-one education (academic detailing) and special review of orders for either levofloxacin or ceftazidime was also seen as a highly effective method for reducing inadequate antibiotic use. Unnecessary antibiotic prescription was significantly reduced by 41 % with no change in clinical outcome [
28,
160]. However, academic detailing is time-consuming and personnel-intensive [
28,
160]. This can partly be compensated by less time-consuming feedback activities, e.g. in the form of written recommendations placed in the patient chart; however, these are not quite as effective and less sustainable [
157].
Education, training and information should be independent and should not be guided by the commercial interests of the manufacturers of medical and diagnostics products, since this is the only way to ensure that prescribing and professional behaviour are not subject to direct or indirect influence (Table
3). A systematic review examined the impact of various strategies undertaken by the pharmaceutical industry such as visits, funding for travel or lodging, sponsoring educational events, free samples, etc., on prescribing practices. According to the study, industry-sponsored continuing medical education (CME) had the biggest impact on physician prescribing practices compared to other activities, leading to a 6–19 % increase in prescription rates of the sponsor’s medication [
41,
42]. The responsibility of organising, holding and financing educational events should be assumed by hospital management.
Table 3
Examples of the influence of commercial interests on prescribing and formulary design [
41,
42]
Meetings with pharmaceutical representatives | 66 % less likelihood of prescribing generic products |
Travel sponsoring (congresses, etc.) | Requests to add the sponsor’s drugs to the hospital formulary are associated with an odds ratio of 7.9 % (95 % CI, 1.1–55.6) 4.5- to 10-fold increase in hospital prescribing rate (sponsor’s products) as compared to before travel |
Continuing medical education (CME funding) | Prescribing rate, (sponsor’s products) increases by 5.5–18.7 % |
Research funding | Requests to add the sponsor’s drugs to the hospital formulary is associated with an odds ratio of 9.5 (95 % CI, 2.6–35.7) |
Conducting proactive audits of antiinfective use
The guideline development group recommends:
Proactive on-site audits of antiinfective use in the context of antiinfective point-of-care chart reviews are important elements of ABS programmes and should be performed routinely by the ABS team (A). They enhance compliance with guidelines or clinical pathways, improve outcome in patients with infection and improve the quality of prescribing with regard to indication, choice of agent, dosing, dosing interval, administration route and treatment duration.
Depending on the problem and treatment target, besides point prevalence studies, agent-, indication- and/or diagnosis-related audits of antiinfective use should be conducted within the scope of regular antiinfective point-of-care chart reviews either hospital-wide or at the unit level, whereby quality indicators should preferably be applied (A).
Results should be fed back in direct interaction with prescribing physicians and discussed with them (A).
Proactive audit of antiinfective use with review and feedback include the collection and analysis of data on diagnosis, indication, choice of agent, dosing, administration route and treatment duration at patient level. The results are fed back to and discussed with the prescribing physicians (point-of-care interventions). In personal consultation with the prescribing physicians reasons for choice of drug could be asked for and therapy should be optimised based on clinical, laboratory, radiological and microbiological examination results. Concomitant disease, comedication, expected pathogens when microbiology is not yet known and local antimicrobial susceptibility patterns must be taken into account. The guideline of two North American medical societies has described this type of audit as a highly effective interventional tool that provides a core strategy for an antimicrobial stewardship programme—called “prospective audit with intervention and feedback” [
6].
Proactive audit of antiinfective use with review and feedback by an ABS team has been described as effectively increasing the rate of adequate antiinfective use by 20 %; the strategy can reduce the rate of inadequate use by half [
34,
53,
161‐
164]. Methods of feedback can be modified, especially when computer-based assistance is available [
165]. The quality of information is important; however, personal feedback is often more effective [
165,
166]. In addition to a direct improvement in the quality of prescribing, audits of antiinfective use allow to recognise the need for education and training.
Audits of antiinfective use can be agent-, diagnosis- or indication-based and can be performed at the patient level, in individual departments, wards, or hospital-wide. Targeted (e.g. in relation to agent, speciality department or ward) as well as time-restricted antiinfective audits can be highly effective (examples are shown in Table
4). A programme in which an infectious diseases specialist or pharmacist conducted targeted point-of-care chart reviews (3×/week) of patients receiving multiple antibiotics, prolonged or high-cost therapy (120-bed hospital), achieved good acceptance: 69 % of the recommendations were accepted and implemented, of these 38 % were to discontinue therapy due to excessive duration, duplicate coverage or inappropriate use, and 33 % were to switch to oral application. Compared with the previous year, a cost reduction of 19 %, estimated savings of USD177,000 were achieved [
167].
Table 4
Examples for performing targeted proactive audits of antiinfective use
• Perioperative antibiotic prophylaxis in selected surgical fields | |
• Targeted therapy of bacteremic patients hospital-wide | |
• Community-acquired pneumonia in the emergency department | |
• Sequential therapy on general wards with antibiotics of high bioavailability | |
Agent-related proactive audit of antiinfective use can address dose adjustment to organ dysfunction, switch to oral application, discontinuation of therapy or targeted therapy based on microbiology. In a multicenter, randomised, controlled study parenteral antibiotics could be significantly reduced by 1 day when patients who had received parenteral antibiotics for longer than 3 days were reviewed by an infectious diseases physician for possible sequential therapy based on defined clinical and laboratory criteria and a recommendation was made for switch to oral drug application [
38]. The intervention showed lack of effect on length of hospital stay, but reduced mean antibiotic costs per patient significantly from USD 36 to USD 20. Following updated recommendations on aminoglycoside treatment, antiinfective visitations by infectious diseases physicians achieved a significant 11 % reduction in nephrotoxicity by shortening the treatment duration from 6 to 4 days and optimise dosing by monitoring drug levels [
161]. In a study assessing the effects of intervention and feedback by the infectious diseases physician, empiric treatment with levofloxacin, vancomycin and carbapenems was switched to targeted antibiotic treatment in line with guidelines. Consumption subsequently decreased by 20 % and median duration of therapy was reduced from 6 to 4 days in comparison to a control group [
168,
169]. In another intervention, a prospective audit and feedback programme was instituted in a teaching hospital by pharmacists and infectious diseases physicians to counteract a trend towards increasing use of expanded-spectrum antimicrobials. This resulted in a significant reduction in consumption of third-generation cephalosporins and aztreonam within a period of 6 years from 28 to 6 DDD/1000 patient-days. Furthermore, there was a significant decrease over time in infections caused by
C. difficile from 2.2 to 1.4 cases/1000 patient-days [
29]. According to another recent time-series analysis, a significant decrease in fluoroquinolone consumption from 118 to 78 DDD/1000 patient-days over 4 years was achieved by the ABS team following implementation of daily hospital-wide audits of fluoroquinolone use based on individual patient data. At the same time, the rate of fluoroquinolone-resistant
P. aeruginosa continuously decreased from 42 to 26 % [
143]. A similar intervention in intensive care units resulted in a sustained 22 % decrease in the number of days of therapy with extended spectrum antibiotics compared with the control group—without negative impact on mortality [
170] (Table
5).
Table 5
Suggested evaluation categories in local audits of antimicrobial use
1. | Antimicrobial therapy adheres to established institutional guidelines with respect to: |
Choice of agent |
Dose |
Route of administration |
Duration of infusion |
Duration of treatment |
2. | Antimicrobial prophylaxis adheres to established institutional guidelines with respect to: |
Choice of agent |
Dosing |
Route of administration |
Timing of preoperative dose |
Dosing interval |
Duration of administration |
With proactive audit of antimicrobial use focussing on diagnosis and indication for antibiotic treatment by an infectious diseases physician-led ABS team with direct interaction and feedback as well as written documentation of recommendations, length of stay was shortened by 3.3 days and a 6 % decrease in mortality was achieved [
27]. As a result of the intervention, median hospital costs were reduced by USD 2642/intervention. By optimising the process of perioperative antibiotic prophylaxis, appropriate dosing and timely administration significantly increased from 72 to 90 % and 36 to 79 %, respectively [
171]. Within the frame of quality assurance, and for benchmarking purposes with other hospitals, targeted audits of selected process of care indicators for the management of important and frequent infections can also in small acute-care hospitals lead to a significant improvement in adherence to established guidelines. In a quasi-experimental before-and-after study of a total of 36 hospitals (<200 beds) the effect of proactive audit and feedback, by using quality indicators, on the management of pneumonia in the emergency department was investigated. The hospitals demonstrated a 30 % improvement in the performance of microbiological diagnostics (blood/sputum cultures) prior to therapy and antibiotic administration within 4 h of hospital admission. As a consequence, mortality was reduced significantly by 12–6 % [
172]. Proactive audits of antiinfective use based on selected quality indicators should regularly take place (see Sect.
2.2.4).
The ABS team should determine the objective, type, contents and frequency of point-of-care chart reviews in agreement with the wards or departments involved and should give report on its effects. The ABS team should get project-specific to hospital-wide access to the laboratory, radiological and microbiological data needed. Computer-based information technology can facilitate audits of antiinfective use (see Sect.
3.3.4)
Quality indicators
The guideline development group recommends:
ABS programmes should be integrated within the hospital’s quality management. Content overlaps with the Therapeutics and Drugs Committee (drug safety) and Hospital Infection Control Committee (prevention of nosocomial infection) is useful and desired. Appropriate quality indicators to measure prescription practice (process measure), emergence of resistance or trend in consumption (outcome measure) and structure ought to be set and applied in every ABS programme (B). At least three indicators measuring structural quality and at least three indicators measuring process quality should be set regularly (A).
ABS programmes are to be regarded as a strategy to ensure quality and should preferably reside as a standard component within the hospitals’ existing quality management [
25]. It is recommendable to utilise data captured pursuant to the new Infection Protection Act for surveillance of resistant microorganism or antiinfective drug use (IfSG §23 Abs. 4) as well as selected data on infection management provided by external quality assurance sources. Additional quality indicators for local use should be selected and applied regularly. This allows to evaluate and document whether ABS aims can be met [
171]. Owing to the different structures and organisation of hospitals, ABS measures must be evaluated locally and if need be adjusted accordingly [
2,
22].
Ideally, indicators ought to be evidence-based, i.e. guideline-derived, and ought to be supported by a formal consensus process in regard of their relevance and practicability; last but not least, they ought to be also put to the practical test. Indicators have been developed for community-acquired pneumonia and urinary tract infections. However, numerous suggestions have been put forward for indicators whose evidence base is rather small and whose relevance and practicability rests on consensus alone. In Germany, there are catalogues of quality indicators for instance for the Helios Hospital Group (“Initiative of Quality Medicine”) or for the Rhön, Sana und Asklepios Hospital Group (“Quality Hospitals”). They also exist for mandatory external health care quality assurance concepts, for whose development and implementation the German National Institute for Quality Measurement in Health Care (Bundesgeschäftsstelle Qualitätssicherung gGmbH; BQS) till 2009, and since then the AQUA Insitut (the Institute for Applied Quality Improvement and Research in Health Care GmBH) was commissioned by the Federal Joint Committee (Gemeinsamer Bundesausschuss; G-BA). However, only few quality indicators have been set for measurement of antibiotic prescribing of which some are already at goal (e.g. for community-acquired pneumonia, antibiotic prophylaxis for obstetric and gynaecological indications, femur fracture, as well as hip and knee endoprosthesis). Individual more or less plausible and consented catalogues of structural indicators are available outside Germany [
65,
173,
174]. A lot of experience was especially gained in France. Process quality indicators, respective pneumonia and surgical prophylaxis, are available in multiple countries (e.g.
http://www.qualitymeasures.ahrq.gov or
http://www.jointcommission.org or
http://www.ic.nhs.uk).
The guideline committee in collaboration with the ABS Expert Network (
http://www.antibiotic-stewardship.de) and the University Hospital of Freiburg established a catalogue of consensus structural and process ABS quality indicators in a multistage procedure including Delphi survey. The catalogue should facilitate external and internal quality assurance. Clinical, ecological (resistance) and economical (cost, cost-effectiveness) relevance as well as the presumed practicability were assessed separately in several categories. In analogy with the so-called QUALIFY process [
175], a provisional list of potentially suitable structural and process indicators was drafted. It was based on the draft of the Guideline itself, the current literature [
1,
6,
25,
86,
105,
171,
176‐
189], including documents and experience with the former ESAC Group (
http://www.esac.ua.ac.be) [
190] and the former ABS International Group, an initiative of 9 EU member states for the improved use of antiinfectives (
http://www.abs-international.eu) [
191]. Based on the results of a workshop (15 participants) held at the ABS Expert Network meeting in November 2011 in Freiburg, a later questionnaire survey (Delphi methods,
n = 75 ABS experts, i.e. advanced members of the ABS training programme of varying professional background, incl. pharmacy and microbiology) and a further workshop, of the initial 99 potential indicators 67 were put forward for discussion and 21 structure and 21 process indicators subsequently selected as presumably most suitable (see Tables
7,
8 in the
Appendix) [
434]. Suitable indicators on the quality of antimicrobial prescribing (process indicator) and structure (structure indicator) taken from this list (Tables
7,
8 in the
Appendix) ought to be set and used in every ABS programme. At least three indicators on structure and process quality each should be determined regularly.