Point-of-care CRP matters: normal CRP levels reduce immediate antibiotic prescribing for acutely ill children in primary care: a cluster randomized controlled trial

Figures & data

Figure 1. Study rationale.

Figure 2. Flow chart representing the number of acute infectious episodes included in the study.

Table 1. Observed immediate antibiotic prescribing rates differ depending on (1) performance and/or result of POC CRP test and (2) EBM practice guideline advice.

	POC CRP test
	CRP not tested	Normal CRP level (<5 mg/L)	Elevated CRP level (≥5 mg/L)	Total
Advice
Prescribe antibiotic	88/149 (59.1%)	24/74 (32.3%)	146/257 (56.8%)	258/480 (53.6%)
Withhold antibiotic	71/824 (8.6%)	17/521 (3.2%)	87/585 (14.9%)	175/1930 (9.1%)
No advice	44/152 (28.9%)	9/98 (9.2%)	75/184 (40.7%)	128/434 (29.4%)
Total	203/1125 (18.0%)	50/693 (7.2%)	308/1026 (30.0%)	561/2844 (19.7%)

Figure 3. Estimated marginal means of the immediate antibiotic prescribing rate (with 95% confidence interval) according to EBM guideline advice and POC CRP testing (partially adjusted analysis). EBM: Evidence Based Medicine; POC CRP: point-of-care C-reactive protein.

Table 2. Influence of POC CRP (normal or elevated versus not measured) on immediate antibiotic prescribing according to EBM guideline advice.

	Partially adjusted analysis^b		Fully adjusted analysis^c
	OR	95%CI	OR	95%CI
Advice: “Prescribe antibiotic (A)” (480 illness episodes)
Normal CRP level	0.23	0.12–0.45	0.24	0.11–0.50
Elevated CRP level	0.67	0.41–1.11	0.69	0.40–1.17
CRP not tested^a
Advice: “Withhold antibiotic (B)” (1930 illness episodes)
Normal CRP level	0.28	0.16–0.51	0.31	0.17–0.57
Elevated CRP level	1.43	0.95–2.15	1.23	0.80–1.89
CRP not tested^a
No advice (C) (434 illness episodes)
Normal CRP level	0.22	0.10–0.50	0.26	0.11–0.64
Elevated CRP level	1.73	1.01–2.96	1.82	1.01–3.27
CRP not tested^a

^aReference category.

^b2844 illness episodes analyzed. Adjusted for performance of the BISNA-intervention, risk for serious infection estimated by a clinical prediction rule, practice type.

^c2804 illness episodes analyzed (40 episodes (1.4%) excluded because missing information about temperature). Adjusted for applying the BISNA-intervention, risk for serious infection estimated by a clinical prediction rule, practice type, region, mean annual antibiotic prescription rate, perceived parental expectation regarding antibiotics, child’s age, temperature.

Table

Preliminary diagnosis	Required clinical symptom	Required criteria for rational prescribing
Acute tonsillitis	Pus on tonsils	Appearing seriously ill or less eating/drinking
Acute otitis media	Clinical signs of acute otitis media	1. < 6 months OR 2. 6–24 months and appearing seriously ill, fever lasting at least 2 days or bilateral AOM OR 3. >24 months and persistent fever lasting at least 3 days or appearing seriously ill OR 4. Persistent otorrhea
Acute sinusitis	No clinical symptom required	1. Appearing seriously ill OR 2. Fever ≥39 °C OR 3. Red swollen face or eyelid (referral)
Pneumonia	No clinical symptom required	No criteria required
Acute bronchitis	No clinical symptom required	1. < 3 months (referral) OR 2. < 24 months and insufficient fluid intake or breathing rate >70/’ (referral) OR 3. Suspected for serious infection (insufficient fluid intake and vomiting, breathing rate >50/’, oxygen saturation ≤92%, sleepy, waking up hard, irritable, drowsy, reduced consciousness, moaning, nasal flaring or chest wall retracting) (referral)
Pertussis	No clinical symptom required	No criteria required
Impetigo	No clinical symptom required	Fever ≥38.5 °C, adenopathy or failure local therapy
Erysipelas	No clinical symptom required	No criteria required
Urinary tract infection (cystitis or pyelonephritis)	No clinical symptom required	No criteria required
Gastrointestinal infection with diarrhea	Diarrhea	Fever ≥38.5 °C, bloody diarrhea or appearing seriously ill

BAPCOC: Belgian Antibiotic Policy Coordination Committee.

Table

Characteristic	Included FP (133)	Excluded FP (34)	p
	n (%) / mean (SD)	n (%) / mean (SD)
Gender (man)	55 (41.4%)	14 (41.2%)	.99^1
Age	39.9 (10.7)	38.7 (13.2)	.57^2
Region (urban)	53 (39.8%)	8 (25.0%)	.18^1
Graduated (resident)	23 (17.3%)	7 (21.2%)	.60^1
Years of experience	13.3 (10.8)	12.3 (13.0)	.66^2
Practice type (exclusive residents)			.14^1
• Solo	19 (14.3%)	8 (23.5%)
• Duo	45 (33.8%)	6 (17.6%)
• Group	69 (51.9%)	20 (58.8%)

¹Chi-square test

²Student t test

FP: family physician; SD: standard deviation.

Table

Characteristic	Included infectious episode (2844)	Excluded infectious episodes (223)	p
	n (%)	n (%)
Age^$
Infant	966 (34.0%)	69 (31.5%)	.16
Preschool child	1077 (37.9%)	75 (34.2%)
Child	801 (28.2%)	75 (34.2%)
Temperature^$
No fever	626 (22.3%)	76 (36.5%)	<.001
Elevation	1127 (40.3%)	76 (36.5%)
High fever (39° or more)	1051 (37.5%)	56 (26.9%)
CRP
Tested	1719 (60.4%)	129 (57.8%)	.45
Not tested	1125 (39.6%)	94 (42.2%)
EBM guideline advice
Prescribe antibiotic (A)	480 (16.9%)	8 (3.6%)	<.001
Withhold antibiotic (B)	1930 (67.9%)	151 (67.7%)
No advice (C)	434 (15.3%)	64 (28.7%)

^$ 4 cases no information about age (4 excluded infectious episodes), 55 cases no information about temperature (40 included infectious episodes, 15 excluded infectious episodes).

Table

Characteristic	CRP tested (1719)	CRP not tested (1125)	p
	n (%)	(%)
Age
Infant	581 (33.8%)	385 (34.2%)	.002
Preschool child	615 (35.8%)	462 (41.1%)
Child	523 (30.4%)	278 (24.7%)
Temperature^$
No fever	400 (23.3%)	226 (20.1%)	.001
Elevation	634 (36.9%)	493 (43.8%)
High fever (39° or more)	664 (38.6%)	387 (34.4%)
EBM guideline advice
Prescribe antibiotic (A)	331 (19.3%)	149 (13.2%)	<.001
Withhold antibiotic (B)	1106 (64.3%)	824 (73.2%)
No advice (C)	282 (16.4%)	152 (13.5%)
Top 10 preliminary diagnoses
Upper respiratory tract infection, acute	482 (28.0%)	401 (35.6%)	NA
Acute otitis media	276 (16.1%)	158 (14.0%)
Viral disease, other (NOS)	197 (11.5%)	119 (10.6%)
Gastroenteritis, presumed infection	160 (9.3%)	93 (8.3%)
Bronchitis (no bronchiolitis)	183 (10.6%)	67 (6.0%)
Tonsillitis	119 (6.9%)	96 (8.5%)
Influenza	111 (6.5%)	56 (5.0%)
Pneumonia	67 (3.9%)	15 (1.3%)
Urinary tract infection	42 (2.4%)	15 (1.3%)
Bronchiolitis	23 (1.3%)	9 (0.8%)
No preliminary diagnosis registered	103 (6.0%)	63 (5.6%)
Diary return	777 (45.2%)	510 (45.3%)	.945

^$ 40 infectious episodes were excluded because of missing information about temperature.

Table

	Non-severe acute infections		Illness episodes at risk for serious infection
Characteristic	CRP tested	CRP not tested	CRP tested	CRP not tested
	(1178)	(1049)	(541)	(76)
Age
Infant	330 (28.0%)	351 (33.5%)	251 (46.4%)	34 (44.7%)
Preschool child	404 (34.3%)	431 (41.1%)	211 (39.0%)	31 (40.8%)
Child	444 (37.7%)	267 (25.5%)	79 (14.6%)	11 (14.5%)
Temperature^$
No fever	349 (30.1%)	219 (21.2%)	51 (9.5%)	7 (9.5%)
Elevation	495 (42.7%)	473 (45.8%)	139 (25.8%)	20 (27.0%)
High fever (39° or more)	315 (27.2%)	340 (32.9%)	349 (64.7%)	47 (63.5%)
EBM guideline advice
Prescribe antibiotic (A)	161 (13.7%)	130 (12.4%)	170 (31.4%)	19 (25.0%)
Withhold antibiotic (B)	850 (72.2%)	788 (75.1%)	256 (47.3%)	36 (47.4%)
No advice (C)	167 (14.2%)	131 (12.5%)	115 (21.3%)	21 (27.6%)
Top 10 preliminary diagnoses
Upper respiratory tract infection, acute	384 (32.6%)	383 (36.5%)	98 (18.1%)	18 (23.7%)
Acute otitis media	193 (16.4%)	152 (14.5%)	83 (15.3%)	6 (7.9%)
Viral disease, other (NOS)	147 (12.5%)	115 (11.0%)	50 (9.2%)	4 (5.3%)
Gastroenteritis, presumed infection	101 (8.6%)	80 (7.6%)	59 (10.9%)	13 (17.1%)
Bronchitis (no bronchiolitis)	89 (7.6%)	60 (5.7%)	94 (17.4%)	7 (9.3%)
Tonsillitis	84 (7.1%)	88 (8.4%)	35 (6.5%)	8 (10.5%)
Influenza	84 (7.1%)	50 (4.8%)	27 (5.0%)	6 (7.9%)
Pneumonia	18 (1.5%)	9 (0.9%)	49 (9.1%)	6 (7.9%)
Urinary tract infection	22 (1.9%)	11 (1.0%)	20 (3.7%)	4 (5.3%)
Bronchiolitis	5 (0.4%)	8 (0.8%)	18 (3.3%)	1 (1.3%)
No preliminary diagnosis registered	50 (4.2%)	54 (5.3%)	53 (9.8%)	9 (11.8%)

^$ 40 infectious episodes were excluded because of missing information about temperature (36 in children with non-serious acute infections, 4 in children at risk for serious infection).

Table

	Fully adjusted analysis^$
	OR	95%CI
Advice: “Prescribe antibiotic (A)” (480 illness episodes)
Normal CRP level	0.24	0.11–0.50
Elevated CRP level	0.69	0.40–1.17
CRP not tested*
Advice: “Withhold antibiotic (B)” (1930 illness episodes)
Normal CRP level	0.31	0.17–0.57
Elevated CRP level	1.23	0.80–1.89
CRP not tested*
No advice (C) (434 illness episodes)
Normal CRP level	0.26	0.11–0.64
Elevated CRP level	1.82	1.01–3.27
CRP not tested*
Practice type
Group	1.22	0.76–1.97
Duo	0.88	0.54–1.42
Solo*
Clinical prediction rule
Children at risk for serious infection	2.08	1.54–2.81
Children with non-severe acute infections*
BISNA
BISNA performed	1.29	0.91–1.83
BISNA not performed*
Region
Urban	0.95	0.66–1.37
Rural*
Mean annual antibiotic prescription rate
Resident/Early career FP	1.78	1.12–2.82
High prescriber	2.21	1.48–3.31
Low prescriber*
Perceived parental expectation regarding antibiotics
Unknown	2.70	1.95–3.73
Present	12.10	8.60–17.00
Absent*
Child’s age
Infant	0.72	0.51–1.01
Preschool	0.99	0.72–1.36
Child/Adolescent*
Temperature
High fever	2.09	1.37–3.19
Elevation	1.58	1.05–2.37
No fever*

*Reference category.

^$ 40 infectious episodes were excluded because of missing information about temperature.

Table

	Partially adjusted analysis^$1		Fully adjusted analysis^$2
	OR	95%CI	OR	95%CI
Advice: “Prescribe antibiotic (A)” (291 illness episodes)
Normal CRP level	0.28	0.12–0.64	0.27	0.11–0.66
Elevated CRP level	0.55	0.29–1.01	0.65	0.34–1.25
CRP not tested*
Advice: “Withhold antibiotic (B)” (1638 illness episodes)
Normal CRP level	0.26	0.13–0.54	0.28	0.13–0.60
Elevated CRP level	1.58	0.95–2.63	1.44	0.86–2.40
CRP not tested*
No advice (C) (298 illness episodes)
Normal CRP level	0.11	0.03–0.40	0.14	0.04–0.56
Elevated CRP level	1.28	0.64–2.57	1.38	0.65–2.92
CRP not tested*

*Reference category

^$12227 illness episodes analyzed. Adjusted for performance of the BISNA-intervention, risk for serious infection estimated by a clinical prediction rule, practice type.

^$22191 illness episodes analyzed (36 episodes (1.6%) excluded because missing information about temperature). Adjusted for performance of the BISNA-intervention, practice type, region, mean annual antibiotic prescription rate, perceived parental expectation regarding antibiotics, child’s age, temperature.

Table

	Partially adjusted analysis^$1		Fully adjusted analysis^$2
	OR	95%CI	OR	95%CI
Advice: “Prescribe antibiotic (A)” (189 illness episodes)
Normal CRP level	0.25	0.07–0.93	0.46	0.11–1.92
Elevated CRP level	1.30	0.45–3.79	1.74	0.54–5.58
CRP not tested*
Advice: “Withhold antibiotic (B)” (292 illness episodes)
Normal CRP level	0.25	0.07–0.86	0.32	0.09–1.14
Elevated CRP level	1.02	0.40–2.64	0.96	0.35–2.65
CRP not tested*
No advice (C) (136 illness episodes)
Normal CRP level	0.50	0.13–1.91	0.45	0.10–1.98
Elevated CRP level	2.69	0.93–7.81	2.29	0.69–7.65
CRP not tested*

*Reference category

^$1 617 illness episodes analyzed. Adjusted for performance of the BISNA-intervention, risk for serious infection estimated by a clinical prediction rule, practice type.

^$2 613 illness episodes analyzed (4 episodes (0.6%) excluded because missing information about temperature). Adjusted for performance of the BISNA-intervention, practice type, region, mean annual antibiotic prescription rate, perceived parental expectation regarding antibiotics, child’s age, temperature.