Persuasive Interventions

Restrictive Interventions

Structural Interventions

1. General education (education detailing, lectures, posters, newsletters etc) (n=13)

1. Expert approval of restricted drugs (n=14)

1. Rapid identification and susceptibility testing (n=2)

2. Review/recommend changes to antibiotic therapy (n=16)

2. Removal/restriction (n=9)

2. Therapeutic drug monitoring (aminoglycoside dosing optimisation programme, n=1)

3. Reminders (n=8)

3. Compulsory order forms for restricted drugs (n=5)

4. Guidelines (n=5)

4. Cycling/rotation (n=4)

5. Audit and feedback (n=4)

5. Therapeutic substitution (n=3)

6. Care pathway (n=3)

6. Automatic antibiotic stop‐order policy (n=2)

7. Opinion leaders (n=2)

7. Compulsory interactive computer order form (n=1)

8. Review/make changes to antibiotic therapy (n=1)

Deliverer

N

Intervention

Aim¹

Outcome²

Persuasion

Restriction

Mixed

Structural

Increase

Decrease

Drug

Microbial

Clinical

Pharmacist

22

14

64%

5

23%

3

14%

0

0%

0

0%

22

100%

22

100%

1

5%

3

14%

Specialist physician (ID/Microbiology)

17

1

6%

11

65%

2

12%

3

18%

0

0%

14

82%

11

65%

7

41%

3

18%

Multidisciplinary team

11

9

82%

1

9%

1

9%

0

0%

3

27%

8

73%

8

73%

5

45%

2

18%

Antibiotic Policy

7

0

0%

7

100%

0

0%

0

0%

0

0%

7

100%

2

29%

6

86%

1

14%

Departmental physician

4

1

25%

2

50%

1

25%

0

0%

1

25%

3

75%

4

100%

0

0%

0

0%

Computer or written feedback

5

4

80%

1

20%

0

0%

0

0%

2

40%

3

60%

5

100%

1

20%

1

20%

Type of Intervention or Study Design

Risk of Bias

Low

Medium

High

Persuasive

6

23

5

Restrictive

0

23

7

CBA

0

4

2

CCT

0

1

1

Cluster CCT or RCT

2

0

0

ITS

9

33

0

RCT

2

8

2

Study

Intervention

Design

Outcome

Difference (Int‐Con)¹

CI (Lower‐Upper)²

p Value

Relative Risk

CI (Lower‐Upper)²

Effect

Decision to Prescribe an Antibiotic (including timing of first dose)

Chu 2003

Persuasive, Care pathway

CBA

% patients with pneumonia receiving antibiotics within 4h of admission

+17.7%

+11.1% to +24.3%

<0.001

1.34

1.20 to 1.51

Yes

Trenholme 1989

Structural, Rapid culture

RCT

% patients with positive blood cultures started on antibiotics

+9.1%

+3.7% to +14.5%

<0.001

NA³

NA³

Yes

Wyatt 1998

Persuasive, Guideline

Cluster RCT

% patients receiving antibiotic prophylaxis for caesarean section

‐3.1%

‐10.1% to +4.0%

>0.1

0.96

0.87 to 1.06

No

Drug Regimen Prescribed

Burton 1991

Structural, Bayesian dosing program

RCT

% patients with peak aminoglycoside concentration >4mg/l

+22.6%

+8.3% to +36.9%

<0.01

1.37

1.11 to 1.70

Yes

Doern 1994

Structural, Rapid culture

RCT

% antibiotics changed within 24h of receipt of culture results

+9.9%

+1.7% to +18.0%

<0.05

1.22

1.03 to 1.45

Yes

Naughton 2001

Persuasive, Care pathway

RCT

% patients receiving IV antibiotics when appropriate for pneumonia

+12.8%

‐8.6% to 34.1%

>0.1

1.16

0.89 to 1.57

No

Trenholme 1989

Structural, Rapid culture

RCT

% patients changed to more effective antibiotics

+6.4%

+1.3% to +11.5%

<0.05

8.44

1.07 to 66.35

Yes

Zanetti 2003

Persuasive, Computer reminder

RCT

% patients received intra‐operative dose for operations lasting >4 hours

+27.4%

+16.1% to +38.8%

<0.001

1.68

1.33 to 2.12

Yes

Study

Intervention

Design

Outcome

Difference (Int‐Con)¹

CI (Lower‐Upper)²

p Value

Relative Risk

CI (Lower‐Upper)²

Effect

Burton 1991

Structural, bayesian dosing program

RCT

% infections responded to antibiotic treatment (Intended direction: +ve, increase)

+7.4%

‐4.7% to +19.4%

>0.1

1.09

0.94 to 1.26

No

Chu 2003

Persuasive, Care pathway

CBA

% mortality for patients with community acquired pneumonia (Intended direction: ‐ve, decrease)

‐1.4%

‐5.4% to +2.6%

>0.1

0.86

0.55 to 1.33

No

Mean length of stay for patients with community acquired pneumonia (Intended direction: ‐ve, decrease)

‐0.20 days

Not calculable from data in paper

³>0.1

NA

NA

No

Dean 2001

Persuasive, Care pathway

CBA

% mortality for patients with community acquired pneumonia (Intended direction: ‐ve, decrease)

‐3.2%

‐5.8% to ‐0.5%

<0.05

⁴0.69

⁴0.49 to 0.97

Yes

Doern 1994

Structural, Rapid culture

RCT

% mortality (Intended direction: ‐ve, decrease)

‐6.5%

‐11.8% to ‐1.3%

<0.05

0.57

0.36 to 0.91

Yes

Zanetti 2003

Persuasive, Computer reminder

RCT

% post‐operative wound infection, randomised controls (Intended direction: ‐ve, decrease)

‐2.2%

‐7.3% to +2.8%

>0.1

0.62

0.21 to 1.85

No

% post‐operative wound infection, historical controls (Intended direction: ‐ve, decrease)

‐5.6%

‐10.0% to ‐1.3%

<0.05

0.44

0.19 to 1.00

Yes

Dempsey 1995

Persuasive, Care pathway

ITS

Mean charge per case for patients with nursing home acquired pneumonia (Intended direction: ‐ve, decrease)

Change in level: ‐$703

p=0.738 level

>0.5 level & slope

+$218

p=0.812 slope

No

Study

Intervention

Design

Outcome

Difference (Int‐Con)¹

CI (Lower‐Upper)²

p Value

Relative Risk

CI (Lower‐Upper)²

Effect

Bailey 1997

Persuasive, Pharmacist

RCT

Switch to oral in 24h (Intended direction: +ve, increase)

+9.8%

‐7.6% to +27.2%

>0.1

1.15

0.90 to 1.47

No

Burton 1991

Structural, Bayesian dosing program

RCT

Trough aminoglycoside level <2mg/l (Intended direction: +ve, increase)

+4.5%

‐6.2% to +15.2%

>0.1

1.05

0.93 to 1.19

No

Cordova 1986

Persuasive, Pharmacist

CBA

Cefazolin dosing 8hrly or more (Intended direction: +ve, increase)

+54.9%

+33.4% to +76.5%

<0.001

3.93

1.40 to 10.99

Yes

Dranitsaris 2001

Persuasive, Pharmacist

RCT

Appropriate cefotaxime indication+dosing (Intended direction: +ve, increase)

+5.9%

‐4.1% to +15.9%

>0.1

1.09

0.94 to 1.25

No

Appropriate cefotaxime indication (Intended direction: +ve, increase)

+1.9%

‐7.0% to +10.7%

>0.1

1.02

0.92 to 1.14

No

Appropriate cefotaxime dosing (Intended direction: +ve, increase)

+8.1%

+1.4% to +14.9%

<0.05

1.09

1.01 to 1.18

Yes

Fraser 1997

Persuasive, ID fellow and pharmacist

RCT

Mean daily doses of IV antibiotics per patient (Intended direction: ‐ve, decrease)

‐3.43

Not calculable from data in the paper

0.09

Continuous variable

Trend

Mean antibiotic charges per patient (Intended direction: ‐ve, decrease)

‐$387

Not calculable from data in the paper

0.05

Yes

Gums 1999

Persuasive, antimicrobial management team

RCT

Median antibiotic costs (Intended direction: ‐ve, decrease)

‐$605

‐$662 to ‐$548

<0.001

Continuous variable

Yes

Median of total costs (Intended direction: ‐ve, decrease

‐$3,054

‐$3,280 to ‐$2,828

<0.001

Yes

Herfindal 1983

Persuasive, Pharmacist

CBA

Mean antibiotic cost per patient day (Intended direction: ‐ve, decrease)

‐$2.35

‐$3.08 to ‐$1.62

<0.001

Continuous variable

Yes

Mean antibiotic doses per patient day (Intended direction: ‐ve, decrease)

‐0.98

‐1.28 to ‐0.68

<0.001

Yes

Mean antibiotic courses per patient day (Intended direction: ‐ve, decrease)

‐0.35

‐0.46 to ‐0.24

<0.001

Yes

Landgren 1988

Persuasive, Pharmacist

CBA

% appropriate timing of prophylaxis (Intended direction: +ve, increase)

+13.0%

‐4.0% to +30.0%

0.12

Relative risk cannot be calculated from data in paper

No

% appropriate duration of prophylaxis (Intended direction: +ve, increase)

+20.0%

+1.0% to +40.0%

0.04

Yes

Parrino 1989

Persuasive, Feedback letter

CBA

Mean quarterly antibiotic expenditure per physician (Intended direction: ‐ve, decrease)

+$3,818

+$603 to +$7,033

>0.5

Continuous variable

No

Pastel 1992

Persuasive, Pharmacist

CCT

Appropriate modification based on C&S results, all physicians (Intended direction: +ve, increase)

+11.1%

‐5.3% to +27.5%

>0.1

1.16

0.93 to 1.45

No

Same, private physicians only (Intended direction: +ve, increase)

+34.7%

+4.2% to +65.1%

<0.05

1.97

0.94 to 4.15

Yes

Shojania 1998

Persuasive, reminder via computer screen

RCT

Mean vancomycin orders per prescriber (Intended direction: ‐ve, decrease)

‐5.40

‐10.34 to ‐0.46

<0.05

Continuous variable

Yes

Mean vancomycin days per prescriber (Intended direction: ‐ve, decrease)

‐14.70

‐27.31 to ‐2.09

<0.05

Yes

Singh 2000

Restrictive, Stop order

RCT

Patients receiving > 3 days antibiotics for suspected pneumonia in the ICU (Intended direction: ‐ve, decrease)

‐69.2%

‐84.2% to ‐54.3%

<0.05

0.29

0.17 to 0.48

Yes

Solomon 2001

Persuasive, Multi‐disciplinary team

RCT

Mean days unnecessary levofloxacin or ceftazidime (Intended direction: ‐ve, decrease)

‐3.30

‐3.81 to ‐2.79

<0.001

Continuous variable

Yes

Trenholme 1989

Structural, Rapid culture

RCT

Patients changed to less expensive antibiotics (Intended direction: +ve, increase)

+16.4%

+5.1% to +27.8%

<0.01

1.91

1.20 to 3.04

Yes

Walker 1998

Persuasive, Reminder

RCT

Patients switched to oral antibiotics (Intended direction: +ve, increase)

+52.0%

+29.3% to +74.7%

<0.001

2.44

1.42 to 4.20

Yes

Study

Intervention

Outcome

Points (Pre / Post)

Change in Level

SE (p Value)

Change in Slope

SE (p Value)

Effect

Abramowitz 1982

Pharmacist reviewed & recommended changes

Total costs of study drugs

9 / 6

‐$23,912

$12,166 (0.078)

‐$1,422

$3,394 (0.684)

Immediate

Adachi 1997

Vancomycin guideline and reminder through order sheet

Vancomycin cost ($) per 1000 patient days

5 / 7

‐$136.19

$53.12 (0.037)

‐$40.47

$14.66 (0.028)

Immediate + Sustained

Avorn 1988

Guideline and reminders through mail, posters and order sheet

% kinetically incorrect Cefazolin dosing

20 / 18

‐13.6%

‐7.7% (0.128)

‐1.7%

‐0.9% (0.067)

Trend, Sustained

Berild 2002

Guideline and reminders through lectures, meetings and academic detailing of new doctors

Antibiotic use DDD/100 hospital days

3 / 4

‐6.893

0.714 (0.011)

‐6.968

0.298 (0.002)

Immediate + Sustained

Antibiotic costs per 100 hospital days

3 / 4

‐£180.88

£14.08 (0.006)

‐£120.29

£5.90 (0.002)

Immediate + Sustained

Hess 1990

Pharmacists contacted physicians to discuss cefazolin dosing

Cefazolin expenditure per patient day

4 / 4

‐$0.38

$0.06 (0.009)

‐$0.01

$0.02 (0.829)

Immediate

Lee 1995

Pharmacist & ID fellow reviewed & recommended changes

Ceftriaxone grams utilized

8 / 4

‐589.428

139.800 (0.004)

45.685

57.545 (0.453)

Immediate

Patel 1989

Guideline reinforced by ward pharmacists

Augmentin expenditure (£'s)

7 / 5

‐$611.36

$127.62 (0.002)

‐$48.69

$0.61 (0.248)

Immediate

Richardson 2000

Guideline, pharmacist reviewed & recommended changes

% inappropriate episodes of vancomycin

3 / 6

‐20.6%

10.9% (0.131)

‐3.0%

4.6% (0.546)

None

Stevenson 1988

Policy, pharmacist reviewed & recommended changes

Antibiotic expenditure per patient (£'s)

10 / 6

‐£6.80

£1.63 (0.002)

£0.39

£0.42 (<0.001)

Immediate + Sustained

Wilson 1991

Newsletter prepared & distributed by pharmacists

% of amoxycillin daily doses versus amoxicillin + pivampicillin total

5 / 28

‐3.8%

5.9% (0.556)

2.8%

1.7% (0.103)

None

Study

Intervention

Outcome

Points (Pre / Post)

Change in Level

SE (p Value)

Change in Slope

SE (p Value)

Effect

Belliveau 1996

Compulsory order form

Vancomycin doses (per 1000 patient days)

12 / 14

‐22.827

11.206 (0.054)

+5.726

1.464 (<0.001)

Immediate + Significant Decay

Bradley 1999

Compulsory antibiotic policy, switching from ceftazidime to piperacillin tazobactam and back

Ceftazidime use (patient days per month)

4 / 8

‐227.8

21.9 (<0.001)

‐19.3

7.2 (0.037)

Immediate + Sustained

Switch back from piperacillin tazobactam to ceftazidime

8 / 6

‐16.9

30.6 (0.59)

64.4

9.3 (<0.001)

Sustained

Bunz 1990

Therapeutic substitution by pharmacist

% metronidazole prescribed <Q12h

6 / 6

‐80.7%

7.2% (<0.001)

+1.8%

2.2% (0.445)

Immediate

Gupta 1989

Therapeutic substitution by pharmacist

% cefazollin prescribed <Q8h

3 / 8

‐61.5%

4.0% (<0.001)

+1.2%

1.7% (0.502)

Immediate

Himmelberg 1991

Use of 9 antibiotics required approval of ID physician

Total antibiotic expenditure

6 / 6

‐$42,414

$8,601 (0.001)

‐$1,883

$2,369 (0.453)

Immediate

Huber 1982

Cephalexin prescriptions had to be signed by the Chief of Staff

Number of cephalexin doses per year

3 / 5

‐88.430

28.415 (0.053)

‐46.327

12.949 (0.037)

Immediate + Sustained

Lautenbach 2003

Use of Vancomycin required approval by hospital antibiotic management program

Vancomycin use (DDD/1000 bed‐days)

3 / 7

‐15.9

26.49 (0.574)

+4.832

11.64 (0.695)

None

McElnay 1995

Consultant signature required for restricted drugs

Total expenditure on antibiotics

12 / 12

‐£859.29

£559.69 (0.141)

+£103.67

£89.15 (0.259)

None

McGowan 1976

Use of chloramphenicol required authorisation by ID consultant

Chloramphenicol use, G per hospital admission

4 / 4

‐0.620

0.081 (0.004)

+0.005

0.030 (0.881)

Immediate

McNulty 1997

Restriction of IV Cefuroxime and removal of oral form from stock

Cefuroxime cost

7 / 16

‐£501.78

£186.86 (0.015)

+£51.78

£37.73 (0.187)

Immediate

Meyer 1993

Restriction of ceftazidime

Number of patients receiving ceftazidime

22 / 6

‐26.440

7.865 (0.003)

‐10.210

2.240 (<0.001)

Immediate + Sustained

Richards 2003

Removal of cefotaxime and ceftriaxone from stock in general wards + consultant approval required

Ceftriaxone/cefotaxime use (DDD/1000 bed‐days)

8 / 16

‐32.54

4.314 (<0.001)

+0.058

0.859 (0.947)

Immediate

Sirinavan 1998

Compulsory order form for restricted drugs with review of inappropriate use by ID consultant

Total restricted drugs cost (million Baht per 200,000 patient days)

4 / 4

‐4.044

0.579 (0.006)

‐1.521

0.216 (0.006)

Immediate + Sustained

Tolzis 1998

Restriction of ceftazidime requiring microbiology approval

Ceftazidime doses

7 / 12

‐176.672

23.200 (<0.001)

‐13.351

4.617 (0.012)

Immediate + Sustained

Woodward 1987

Restriction of target drugs through approval of initial prescription by ID physician + 72h stop order

Average cost per antibiotic day

25 / 17

‐$1.46

$0.63 (0.026)

$0.01

$0.06 (0.928)

Immediate

Young 1985

Therapeutic substitution of gentamicin by amikacin

Gentamin usage (% of aminoglycoside usage)

15 / 22

‐19.7%

6.25% (0.003)

‐1.1%

1.5% (0.465)

Immediate

Study

Intervention

Outcome

Points (Pre / Post)

Change in Level

SE (p Value)

Change in Slope

SE (p Value)

Effect

Everitt 19990

Guideline disseminated by department leaders plus removal of restricted drugs

% caesarean sections receiving <5g cefoxitin

9 / 25

‐13.4%

7.4% (0.080)

0.2%

1.2% (0.896)

Trend, Immediate

Perez 2003¹

Restrictive for aminoglycoside & cephalosporin (compulsory structured prescription form), supported by education (guideline, lectures, posters and targeted messages). Persuasive only for prophylaxis (reminders)

% incorrect aminoglycoside prescriptions

79 / 66

‐0.2%

0.1% (0.858)

‐47.7%

6.4% (<0.001)

Sustained

% incorrect cephradine & cephalexin prescriptions

79 / 66

NS

NS

NS

NS

None

% incorrect ceftazidime and cefotaxime prescriptions

79 / 60

‐4.2%

12.0% (0.726)

‐7.3%

3.0% (0.017)

Sustained

% incorrect prescriptions for antibiotic prophylaxis

103 / 42

+24.1%

13.2% (0.070)

‐19.9%

6.9% (0.004)

Sustained

Inaraja 1986

Restrictiion of cephalosporin use, method not clear supported by pharmacist reviewing all antibiotic prescriptions and recommending change

Cost of cephalosporins as % total antibiotics

9 / 3

‐11.6%

5.9% (0.093)

+4.0%

3.7% (0.316)

Trend to Immediate

Mercer 1999

Removal of 16 drugs from stock in ER or operating theatre + consultant approval required, supported by pneumonia guideline + reminder in notes

Antibiotic expenditure ($'s)

12 / 12

‐$13,687

$7,170 (0.071)

‐$79.76

$1,004 (0.938)

Immediate

Richardson 2000

Guideline, pharmacist reviewed and recommended changes

% inappropriate use of vancomycin

3 / 6

‐20.6%

10.9% (0.131)

‐3.0%

4.6% (0.546)

None

Saizy‐Callaert 2003

Compulsory order form plus audit & feedback

Mean expenditure on all anti‐infectives per hospital patient, including anti‐retroviral drugs

3 / 4

$0.01

$0.50 (0.981)

‐$0.29

$0.21 (0.299)

None

Salama 1996

Compulsory order form plus automatic three‐day stop order supported by guidelines, lectures, reminders (posters)

Vancomycin usage (units)

13 / 29

+7.341

102.613 (0.943)

‐40.594

11.661 (0.001)

Sustained

Suwangoal 1991

Restriction by requiring authorisation by ID consultant supported by educational guideline

Total restricted drugs cost (Baht per month)

6 / 12

‐238,253

(0.054

‐51,498

(0.07)

Immediate + Sustained

Study

Intervention

Planned

Outcome

Points (Pre / Post)

Change in Level

SE (p Value)

Change in Slope

SE (p Value)

Effect

Bradley 1999b

Restriction by antibiotic policy. Intervention 2 years after resistance emerged

Yes

Probability of being free of VRE

4 / (8/6)

Authors' Cox regression analysis: patients were significantly more likely to be VRE free in the 8 months after the policy change and then significantly less likely to be VRE free the 6 months after the policy reverted to Ceftazidime

Sustained

Calil 2001

Restriction by removal. Response to increasing Enterobacter cloacae

No

% monthly incidence of E.cloacae infections

4 / 7

‐15.51%

1.31% (0.054)

‐2.73%

0.6% (0.138)

Immediate

Carling 2003

Persuasive by Antimicrobial Management Team. Response to 32 % increase in use of expanded spectrum cephalosporins and aztreonam

Yes

Clostridium difficile cases per 1000 patient days

3 / 7

‐1.470

0.160 (<0.001)

‐0.814

0.052 (<0.001)

Immediate + Sustained

Resistant Enterobacteriaceae cases per 1000 patient days

4 / 7

‐2.340

0.798 (0.032)

‐1.340

0.340 (0.011)

Immediate + Sustained

% MRSA

3 / 8

+0.55%

3.66% (0.885)

+1.33%

1.27% (0.335)

No

% VRE

3 / 3

+0.40%

9.47% (0.968)

+1.56%

4.54% (0.742)

No

Climo 1998

Restriction by prior approval. Response to increasing cases of Clostridium difficile associate diarrhoea

No

Cases of C.difficile associated diarrhoea per quarter

9 / 11

‐26.326

3.424 (<0.001)

‐3.791

0.570 (<0.001)

Immediate + Sustained

de Champs 1994

Restriction by removal. Response to outbreak of E.cloacae

No

Number of multiresistant E.cloacae

7 / 12

‐7.468

1.367 (<0.001)

‐1.000

0.272 (0.002)

Immediate + Sustained

Gerding 1985

Restriction by ID approval, cycling from amikacin to gentamicin to amikacin. Intervention 5 years after resistance emerged

Yes

% Gram‐negative bacilliary resistance to gentamicin

4 / 26

‐4.69%

3.10% (0.590)

1.64%

1.11% (0.149)

No

26 / 12

‐3.08%

1.87% (0.109)

‐0.04%

0.28% (0.899)

No

12 / 12

‐3.85%

1.70% (0.035)

0.09%

0.28% (0.754)

Immediate

Khan 2003

Restriction, cefotaxime to ceftriaxone. Response to increasing drug costs

Yes

Cases of C.difficile associated diarrhoea per quarter

6 / 13

+19.680

10.19 (0.074)

+4.660

2.410 (0.073)

Immediate + Sustained

Restriction, ceftriaxone to levofloxacin. Response to C.difficile associated diarrhoea

No

13 / 5

‐5.780

8.860 (0.525)

‐5.810

3.090 (0.080)

Sustained

Landman 1990

Restriction by prior approval. Response to increase in VRE

Yes

Ceftazidine resistant Klebsiella pneumoniae per 1000 discharges

29 / 23

‐2.893

2.253 (0.205)

+0.040

0.155 (0.798)

No

MRSA per 1000 discharges

29 / 23

‐4.769

5.372 (0.379)

‐0.412

0.396 (0.304)

No

Cefotaxime resistant Acinetobacter spp. per 1000 discharges

29 / 23

‐0.052

1.613 (0.974)

+0.337

0.112 (0.004)

Sustained INCREASE

Lautenbach 2003

Restriction by prior approval. Response to emergence of VRE

No

% VRE

3 / 7

‐14.52%

6.65% (0.081)

‐8.96%

2.86% (0.026)

Sustained

Leverstein 2001

Restriction by prior approval & restriction. Response to outbreak

No

% intestinal colonization by gentamicin resistant enterobacteriaceae

4 / 8

‐2.43%

2.09% (0.284)

0.34%

0.69% (0.637)

No

McNulty 1997

Restriction by removal. Response to increasing cases of C.difficile associated diarrhoea

No

Cases of C.difficile associated diarrhoea per month

7 / 16

‐3.270

2.000 (0.120)

‐0.500

0.404 (0.230)

No

Meyer 1993

Restriction by prior approval. Response to outbreak of ceftazidime resistant K pneumoniae

No

Ceftazidime resistant K pneumoniae per 1000 average daily census

14 / 11

‐38.563

8.559 (<0.001)

‐6.172

1.470 (<0.001)

Immediate + Sustained

Pear 1994

Restriction by prior approval. Response to outbreak of C.difficile associated diarrhoea

No

Number of Clostridium difficile associated diarrhoea cases per month

40 / 14

‐3.683

1.754 (0.041)

‐0.323

0.212 (0.134)

Immediate

Intervention

Outcome

¹Good Evidence

²Weak Evidence

³No Evidence

Restriction of third generation cephalosporins

Resistant gram‐ve bacteria

Carling 2003

de Man 2000

Calil 2001

Landman 1999

Meyer 1993

Leverstien van Hall 2001

CDAD

Carling 2003

McNulty 1997b

Khan 2003

MRSA

Carling 2003

Landman 1999

VRE

Bradley 1999

Restriction of aminoglycosides

Resistant gram‐ve bacteria

de Champs 1994

Gerding 1985

Restriction of clindamycin

CDAD

Climo 1998b

Pear 1994

Restriction of vancomycin

VRE

Lautenbach 2003

Reduced duration of antibiotics in ICU

Colonisation or infection by resistant gram‐ve bacteria

Singh 2000

Cycling of antibiotics in ICU

Toltzis 2002

Study

Intervention

Design

Outcome

Difference (Int‐Con)¹

CI (Lower‐Upper)²

p Value

Relative Risk

CI (Lower‐Upper)²

Effect

Bailey 1997

Persuasive, pharmacist

RCT

Restart of IV antibiotics

0.0%

‐9.1% to +9.1%

>0.5

1.00

0.21 to 4.72

No effect

30 day readmission

+19.6%

+4.7% to +34.5%

<0.05

3.00

1.18 to 7.64

Adverse

Infection related readmission

+2.0%

‐7.8% to +11.8%

>0.5

1.33

0.31 to 5.66

No effect

In‐hospital mortality

0.0%

‐9.1% to +9.1%

>0.5

1.00

0.21 to 4.72

No effect

Burton 1991

Structural, Bayesian dosing programme

RCT

Aminoglycoside toxicity

‐3.8%

‐12.2% to +4.7%

>0.1

0.60

0.18 to 1.95

No effect

DeMan 2000

Restrictive, compulsory antibiotic policy

Cluster CCT

Mortality

+4.6%

‐0.9% to +10.1%

>0.1

1.63

0.90 to 2.94

No effect

Fraser 1997

Persuasive, ID fellow and pharmacist

RCT

Clinical response

‐1.1%

‐11.6% to +9.4%

>0.5

0.99

0.87 to 1.12

No effect

Restart of antibiotics in 7 days

‐8.5%

‐16.2% to ‐0.9%

<0.05

0.36

0.14 to 0.90

Improved

Readmission

+4.8%

‐3.8% to +13.4%

>0.1

1.47

0.72 to 3.02

No effect

Mortality

+2.2%

‐6.4% to +10.8%

>0.5

1.20

0.59 to 2.44

No effect

Gums 1999

Persuasive, antimicrobial management team

RCT

Mean length of stay (days)

‐3.3

‐3.5 to ‐3.1

<0.001

NA

NA

Improved

Herfindal 1983

Persuasive, pharmacist

CBA

Mean length of patient stay (days), all patients

+3.4

+0.5 to +6.3

<0.05

NA

NA

Adverse

Mean length of stay (days), patients receiving antibiotics

+8.4

+3.9 to +12.9

<0.01

NA

NA

Adverse

Singh 2000

Restrictive, stop order for low risk patients

RCT

30 day mortality

‐18.1%

‐35.6% to ‐0.7%

<0.05

0.41

0.16 to 1.05

Improved

Solomon 2001

Persuasive, guideline and academic detailing by multi‐disciplinary team

RCT

Readmission in 30 days

+0.7%

‐0.3% to +1.7%

>0.1

1.21

0.91 to 1.61

No effect

ICU transfer

+0.2%

‐1.2% to +1.6%

>0.5

1.03

0.84 to 1.27

No effect

In‐hospital mortality

+0.1%

‐0.7% to +0.9%

>0.5

1.05

0.73 to 1.50

No effect

Mean length of stay (days)

‐0.2

‐0.5 to +0.1

>0.1

NA

NA

No effect

Walker 1998

Persuasive, reminder in case notes placed by pharmacist

RCT

Total readmissions

‐4.0%

‐23.2% to +15.2%

>0.1

0.75

0.19 to 3.01

No effect

Study

Intervention Cost

Savings or Costs Achieved by Intervention

Abramowitz 1982

80 hours of clinical pharmacists' time per month (2.5 hours per week by 8 pharmacists).

$16,000 per year at 1982 prices.

The authors estimate savings of $156,756 per year based on an uncontrolled before and after analysis. Our segmented regression analysis showed a sudden change in level of ‐$23,913 (p=0.078) and a non significant change in slope by $1,423 per year (p=0.684).

Bailey 1997

Labour costs (pharmacists' time) were estimated to be $15,000 per year at Hospital A and $7,000 per year at Hospital B at 1997 prices.

Extrapolating the average postrandomisation costs to 200 patients per year at Hospital A, the estimated annual saving was $1,600 per year (95% CI from $3,100 increase to $6,300 saving).

Extrapolating the average postrandomisation costs to 100 patients per year at Hospital B, the estimated annual saving was $4,200 per year (95% CI from $700 increase to $9,000 saving).

30 day readmission rates were significantly increased in the intervention patients at Hospital A.

Gums 1999

For the 125 patients in the Intervention group, the time required was 15.6 hours (7.5 minutes per patient) for Infectious Diseases physician consults and 10.4 hours (5 minutes per patient) for Microbiology consults, adding up to a total of $1,092 at 1999 prices, or $8.74 per patient. Pharmacist time was 3.5 days per week or approximately $21,000 per year.

The difference in median antibiotic costs was ‐$605 per patient (CI from ‐$548 to ‐$662) so that savings greatly exceeded costs.

Landgren 1988

The total cost of both campaigns, including the audits and analysis of results was Australian $71,950.

The estimated annual saving was $69,434 for the first intervention and $55,636 for the second intervention.

Solomon 2001

Estimated annual cost of the intervention was $21,750.

Formal economic analysis was not performed but the institution "plans to continue and expand antibiotic counter‐dealing."

Woodward 1987

The expenses incurred in setting up the program (computer costs and consultant time) were paid by the hospital but details are not provided.

Programme required 76h per month from pharmacists, 24h per month from ID fellows and 10h for ID faculty. These hours were "absorbed within the working days of each person" and were not costed.

Our segmented regression analysis shows a sudden change in level of average cost per patient day by ‐$1.46 (SE $0.63, p=0.026).

Wyatt 1998

Fixed cost of preparing the video was £5,000. Variable cost per visit £445 (travel £25, hotel £60, staff time £330). Overall mean cost per visit £860, 1995 prices.

The intervention had no significant impact on practice.