Table 5.3GRADE profile: Is primary prophylaxis with antibiotics more effective than no primary prophylaxis at improving outcomes in patients at risk of neutropenic sepsis

Quality assessmentNo of patientsEffectQuality
No of studiesDesignRisk of biasInconsistencyIndirectnessImprecisionOther considerationsAntibioticsNo primary prophylaxisRelative
(95% CI)		Absolute
Mortality (quinolone studies)
10randomised trialsno serious risk of biasno serious inconsistencyno serious indirectnessserious2none32/1295
(2.5%)		57/1286
(4.4%)		RR 0.615 (0.4 to 0.946)17 fewer per 1000 (from 2 fewer to 27 fewer)MODERATE
Infection related mortality (quinolone studies)
6randomised trialsno serious risk of biasno serious inconsistencyno serious indirectnessvery serious2,3none19/1295
(1.5%)		36/1286
(2.8%)		RR 0.58
(0.336 to 1.001)		12 fewer per 1000 (from 19 fewer to 0 more)LOW
Febrile neutropenia (quinolone studies)
10randomised trialsserious1,4serious5no serious indirectnessno serious imprecisionnone419/1339
(31.3%)		594/1341
(44.3%)		RR 0.727
(0.62 to 0.852)		121 fewer per 1000
(from 66 fewer to 168 fewer)		LOW
Febrile neutropenia (ciprofloxacin studies)
2randomised trialsno serious risk of biasno serious inconsistencyno serious indirectnessvery serious2,6none19/56
(33.9%)		26/56
(46.4%)		RR 0.95 (0.66 to 1.35)23 fewer per 1000 (from 158 fewer to 163 more)LOW
Febrile neutropenia (levofloxacin studies)
3randomised trialsno serious risk of biasno serious inconsistencyno serious indirectnessno serious imprecisionnone347/1160
(29.9%)		460/1160
(39.7%)		RR 0.76 (0.7 to 0.82)95 fewer per 1000 (from 71 fewer to 119 fewer)HIGH
Febrile neutropenia (ofloxacin studies)
4randomised trialsno serious risk of biasserious5no serious indirectnessserious2,6none34/111
(30.6%)		70/106
(66%)		RR 0.35 (0.1 to 1.23)429 fewer per 1000
(from 594 fewer to 152 more)		LOW
Febrile neutropenia (TMP-SMZ studies)
16randomised trialsno serious risk of biasserious5no serious indirectnessno serious imprecisionnone367/713
(51.5%)		473/711
(66.5%)		RR 0.80 (0.69 to 0.92)133 fewer per 1000
(from 53 fewer to 206 fewer)		MODERATE
Infection with bacteria resistant to the antibiotic used for prophylaxis
15randomised trialsno serious risk of biasno serious inconsistencyno serious indirectnessserious2none74/1680
(4.4%)		50/1654
(3%)		RR 1.43 (1 to 2.03)13 more per 1000 (from 0 more to 31 more)MODERATE
Infection with bacteria resistant to quinolones (quinolone studies)
4randomised trialsno serious risk of biasserious inconsistency5no serious indirectnessserious2none44/1185
(3.7%)		39/1180
(3.3%)		RR 0.87 (0.18 to 4.15)4 fewer per 1000 (from 27 fewer to 104 more)LOW
Colonisation with bacteria resistant to quinolones (quinolone studies)
2randomised trialsserious risk of bias7no serious inconsistencyno serious indirectnessserious2none6/47
(12.8%)		2/46
(4.3%)		RR 3.30 (1.03 to 12.56113 more per 1000 (from 1 more to 503 more)LOW
Length of hospital stay - not reported
0-----none----
Quality of life - not reported
0-----none----
1

Most studies did not have clear allocation concealment or double blinding.

2

Low number of events.

3

Confidence interval of the pooled estimate crosses both no effect and significant benefit.

4

9/25 had adequate allocation concealment and 13/25 double blinding

5

Statistically significant heterogeneity

6

95% confidence interval around the pooled estimate of effect includes both no effect and appreciable benefit or appreciable harm.

7

Both trials had unclear allocation concealment

From: 5, Reducing the risk of septic complications of anticancer treatment

Cover of Neutropenic Sepsis: Prevention and Management of Neutropenic Sepsis in Cancer Patients
Neutropenic Sepsis: Prevention and Management of Neutropenic Sepsis in Cancer Patients.
NICE Clinical Guidelines, No. 151.
National Collaborating Centre for Cancer (UK).
Copyright © National Collaborating Centre for Cancer, 2012.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.