Factors influencing antibiotic prescribing for respiratory tract infections in primary care – a comparison of physicians with different antibiotic prescribing rates

Figures & data

Figure 1. Flow chart showing the inclusion process and division into prescriber groups based on change of antibiotic prescription rate for respiratory tract infections (RTI) between the first period 2006–2009 and the third period 2012–2014.

Table 1. Background information on primary care physicians, respiratory tract infections, and antibiotic prescriptions for included and excluded physicians during the study period 2006–2014.

		Physicians
		Included	Excluded
		n = 161	n = 852
Physicians		n (%)	n (%)
Sex	Female	63 (39)	169 (20)
	Male	88 (55)	194 (23)
	Unknown	10 (6)	489 (57)
Birth year	≤1940s	34 (21)	43 (5)
	1950s	45 (28)	43 (5)
	1960s	45 (28)	65 (8)
	1970s	24 (15)	88 (10)
	>1980s	3 (2)	124 (15)
	Unknown	10 (6)	489 (57)
Training level
Specialist in family medicine during	100% of visits	70 (43)	273 (32)
	50 to <100% of visits	41 (25)	38 (4)
	1 to <50% of visits	30 (19)	32 (4)
Junior physician or other specialist		20 (12)	509 (60)
Continuity – number of primary health care centres at which each physician has worked	1	28 (17)	564 (66)
	2	43 (27)	186 (22)
	3–4	50 (31)	80 (9)
	5+	40 (25)	22 (3)
Respiratory tract infections		n (% of RTI visits)	n (% of RTI visits)
Total visits		262,503 (100)	135,291 (100)
Out-of-hours visits		40,431 (15)	11,757 (9)
Index visits		220,979 (83)	114,469 (87)
Antibiotic prescriptions		107,767 (41)	56,509 (42)
Antibiotic prescriptions at index visits		91,859 (35)	48,463 (37)
Point-of-care testing for respiratory tract infections
C-reactive protein test		104,361 (40)	60,665 (45)
Rapid antigen detection test for Group A Streptococci		50,253 (19)	33,584 (25)
Number of respiratory tract infections visits per physician
		RTI visits, n	RTI visits, n
	10th percentile	259	6
	median	1630	52
	90th percentile	2925	406

RTI: respiratory tract infection.

Table 2. Comparison of characteristics of the primary care physicians in the prescriber groups (the Low Prescribing Group, the Decreasing Prescribing Group, and the High Prescribing Group).

		Prescriber groups
		Low Prescribing Group	Decreasing Prescribing Group	High Prescribing Group	p-Value
Primary care physicians’ characteristics		n (%)	n (%)	n (%)
Physicians’ sex	Female	16 (36)	31 (46)	16 (41)	0.53
	Male	29 (64)	36 (54)	23 (59)
Physicians’ birth year	≤1940s	13 (29)	12 (18)	9 (23)	0.28
	1950s	9 (20)	26 (39)	10 (26)
	1960s	13 (29)	17 (25)	15 (38)
	≥1970s	10 (22)	12 (18)	5 (13)
Training Level
Specialist in Family Medicine	100% of visits	20 (38)	36 (53)	14 (34)	0.24
	50 to <100% of visits	14 (27)	15 (22)	12 (29)
	1 to <50% of visits	8 (15)	11 (16)	11 (27)
Junior physician or other specialist		10 (19)	6 (9)	4 (10)
Physicians’ continuity – number of primary health care centres at which each physician has worked	1	12 (23)	10 (15)	6 (15)	0.3
	2	16 (31)	21 (31)	6 (15)
	3-4	14 (27)	23 (34)	13 (32)
	≥5	10 (19)	14 (21)	16 (39)
Physicians’ out-of-hours rate – Out-of-hours visits per total number of visits per physician	0%	13 (25)	5 (7)	7 (17)	0.084
	0.1 to <10%	13 (25)	16 (24)	13 (32)
	10 to <20%	19 (37)	32 (47)	11 (27)
	≥20%	7 (13)	15 (22)	10 (24)
Physicians’ activity level - number of respiratory tract infection visits in total per physician	≤900	23 (44)	9 (13)	8 (20)	0.006^*
	901–1600	8 (15)	19 (28)	12 (29)
	1601–2300	14 (27)	20 (29)	12 (29)
	>2300	7 (13)	20 (29)	9 (22)

The variables are based on data for all nine years and assigned to the physician of the index visit.

*The comparison between the Low Prescribing Group and the Decreasing Prescribing Group was significant at p < 0.001. No other pairwise comparison was significant.

Table 3. Comparison of the prescriber groups (the Low Prescribing Group, the Decreasing Prescribing Group, and the High Prescribing Group) by physicians’ use of point-of-care tests, their choice of diagnosis, and how often the patients returned and received antibiotics.

		Prescriber groups			p-Value
		Low Prescribing Group n (%)	Decreasing Prescribing Group n (%)	High Prescribing Group n (%)	All	Low vs. Decreasing Prescribing Group^a	Decreasing vs. High Prescribing Group^a	Low vs. High Prescribing Group^a
CRP test use at index visits (% of index visits)	<26%	17 (33)	17 (25)	7 (17)	0.463
	26 to <40.2%	12 (23)	19 (28)	9 (22)
	40.2 to <52.7%	9 (17)	17 (25)	14 (34)
	≥52.7%	14 (27)	15 (22)	11 (27)
Median CRP test result at index visits per physician (mg/L)	<9.5	14 (27)	22 (32)	14 (34)	0.881
	9.5–11.5	17 (33)	25 (37)	16 (39)
	12	6 (12)	7 (10)	4 (10)
	≥13	15 (29)	14 (21)	7 (17)
Median CRP test result at index visits per physician where antibiotics were prescribed (mg/L)	<24	6 (12)	14 (21)	22 (54)	<0.001	0.022	0.001	<0.001
	24–36	6 (12)	20 (29)	14 (34)
	37–46	17 (33)	21 (31)	3 (7)
	≥46.5	23 (44)	13 (19)	2 (5)
Median CRP test result at index visits with pneumonia where antibiotics were prescribed per physician (mg/L)	<59	11 (22)	12 (18)	17 (42)	0.059
	59–75	10 (20)	19 (28)	10 (25)
	75.5–96.5	16 (32)	19 (28)	4 (10)
	≥97	13 (26)	17 (25)	9 (22)
RADT use for GAS at index visits	<13%	19 (37)	18 (26)	4 (10)	0.001	1.000	0.021	0.001
	13 to <19%	11 (21)	19 (28)	10 (24)
	19 to <25.9%	15 (29)	18 (26)	7 (17)
	≥25.9%	7 (13)	13 (15)	20 (49)
Negative RADT result for GAS at index visits where antibiotics were prescribed (% of tested)	<23%	20 (38)	19 (28)	2 (5)	<0.001	1.000	<0.001	<0.001
	23 to <34%	15 (29)	19 (28)	5 (12)
	34 to <48%	9 (17)	20 (29)	12 (29)
	≥48%	8 (15)	10 (15)	22 (54)
Negative RADT result for GAS at index visits with pharyngotonsillitis where antibiotics were prescribed (% of tested)	<11%	22 (42)	13 (19)	5 (11)	0.001	0.153	0.099	0.003
	11 to <23%	13 (25)	22 (32)	5 (11)
	23 to <33%	9 (17)	18 (26)	17 (39)
	≥33	8 (15)	15 (22)	17 (39)
The likelihood of selecting a diagnosis with potential bacterial aetiology^b (% of index visits)	<33%	27 (52)	12 (18)	2 (5)	<0.001	<0.001	<0.001	<0.001
	33.0 to <37.3%	17 (33)	22 (32)	1 (2)
	37.3 to <43.5%	4 (8)	27 (40)	9 (22)
	≥43.5%	4 (8)	7 (10)	29 (71)
Return visits within 30 days (% of index visits)	<13.0%	15 (29)	15 (22)	11 (27)	0.029	0.696	0.015	1.000
	13.0 to <14.3%	13 (25)	19 (28)	7 (17)
	14.3 to <15.3%	11 (21)	24 (35)	6 (15)
	≥15.3%	13 (25)	10 (15)	17 (41)
Antibiotics at return visits within 30 days (% of index visits)	<5.4%	15 (29)	16 (24)	10 (24)	0.419
	5.4 to <6.1%	10 (19)	22 (32)	8 (20)
	6.1 to <6.8%	16 (31)	15 (22)	9 (22)
	≥6.8%	11 (21)	15 (22)	14 (34)

The variables are based on data for all nine years and assigned to the physician of the index visit. The variables are divided into quartiles.

CRP: C-reactive protein; GAS: Group A Streptococci; RADT: rapid antigen detection test.

^ap-Value adjusted for pairwise comparison with Bonferroni correction by multiplying with three.

^bDiagnoses with potential bacterial aetiology: acute media otitis, pharyngotonsillitis, pneumonia, and sinusitis.

Table 4. Association between physician characteristics and belonging to the high prescribing group as compared to the decreasing prescribing group.

	Crude model		Adjusted model
	OR	95% CI	OR	95% CI
Male	1.24	0.56–2.75	0.53	0.11–2.48
Birth year	0.99	0.96–1.03	0.96	0.89–1.03
Number of PHCC	1.28	1.06–1.54	1.20	0.86–1.66
Median CRP level if antibiotics (mg/L)	0.92	0.88–0.96	0.99	0.93–1.06
Negative RADT treated with antibiotics (%)	1.08	1.04–1.11	1.08	1.02–1.14
Potential bacterial diagnoses (% of index visits)	1.31	1.18–1.46	1.32	1.15–1.52

The variables are based on data for all nine years and assigned to the physician of the index visit. Crude and adjusted odds ratio (ORs) with 95% confidence intervals (95% CI). Adjusted ORs were calculated using multiple logistic regressions with a full model.

Number of PHCC – the number of primary health care centres the physician has worked at. Median CRP level if antibiotics – the median C-reactive protein level for the physicians’ patients at index visits who were prescribed antibiotics. Negative RADT treated with antibiotics – the proportion of patients where rapid antigen detection test for Group A Streptococci was performed with negative results and still prescribed antibiotics. Potential bacterial diagnoses – acute media otitis, pharyngotonsillitis, pneumonia, and sinusitis.

Data availability statement

The datasets used and analysed in the current study are available with the corresponding author upon reasonable request.