How do Swiss general practitioners agree with and report adhering to a top-five list of unnecessary tests and treatments? Results of a cross-sectional survey

Figures & data

Table 1. Characteristics of respondents from the Swiss primary care active monitoring network (n = 167).

Characteristic
Sex (women)	50 (30%)
Mean age (±SD)	54 (±8.9)
Language area
German	90 (54%)
French	61 (37%)
Italian	14 (8%)
Country of medical training (Switzerland)	151 (90%)
Years of experience in practice (±SD)	18 (±11)
Practice size
Solo practice	46 (28%)
2 to 4 physicians	105 (63%)
5 or more physicians	16 (10%)
Practice in rural area	43 (26%)
Average consultation <20 min	70 (42%)
In practice:
Pharmacy	69 (41%)
Laboratory	143 (86%)
X-ray equipment	106 (63%)

Table 2. Awareness of campaigns to decrease overuse among general practitioners (n = 167).

Question	n (%)
Have you heard about the campaign ‘Smarter Medicine’
Yes	104 (62%)
No	45 (27%)
I don’t know	11 (7%)
Have you heard about the campaign ‘Choosing Wisely’
Yes	95 (57%)
No	63 (38%)
I don’t know	2 (1%)
Have you heard about the ‘do not do’ lists?
Yes	77 (46%)
No	73 (44%)
I don’t know	8 (5%)

Table 3. Proportion of physicians who encounter clinical scenarios often or very often^a, who rarely or never go against each ‘Smarter Medicine’ recommendation, and mean agreement with recommendation^b.

Clinical scenario	Category	Total (n = 167)
1. Patients with non-specific low-back pain	Encountered often or very often^a	138 (83%)
	Rarely or never get imaging	114 (68%)
	Mean agreement with recommendation not to get imaging for non-specific low-back pain (0–10)^b	9.0 (±1.9)
2. Prostate cancer screening using prostate-specific antigen test (PSA)	Encountered often or very often^a	97 (58%)
	Discuss most of the time prior to screening	116 (69%)
	Mean agreement with recommendation to discuss before screening (0–10)^b	8.9 (±1.7)
3. Upper respiratory tract infections without signs of complications	Encountered often or very often^a	152 (91%)
	Rarely or never prescribe antibiotics for these infections	113 (68%)
	Mean agreement with recommendation not to prescribe antibiotics for these infections (0–10)^b	9.1 (±1.5)
4. Patients without lung pathology for preoperative assessment	Encountered often or very often^a	79 (47%)
	Rarely or never request a chest X-ray	124 (74%)
	Mean agreement with recommendation not to request preoperative chest X-rays (0–10)^b	9.2 (±1.4)
5. Long-term use of proton pump inhibitors without confirmed pathology	Encountered often or very often^a	98 (59%)
	Rarely or never continue medication without lowering dose	56 (34%)
	Mean agreement with recommendation to not continue medication without lower dose (0–10)^b	8.9 (±1.4)

^a Often defined as weekly and very often as very often.

^b On a 10-point Likert scale, from 0 (complete disagreement) to 10 (complete agreement).

Table 4. Reasons why physicians go against the recommendations that received 12 or more positive responses, with several responses possible (n = 167).

Recommendation	Reasons for not following recommendation	GPs responding yes (%)
1. Do not obtain imaging studies for patients with non-specific low-back pain	Request or insistence of patient	89 (67%)^a
	Desired information from imaging	28 (21%)^a
	To increase bond with the patient	17 (13%)^a
	Not applicable as I always follow this recommendation	35 (21%)
2. Do not perform prostate cancer screening without a discussion	PSA already ordered previously	45 (62%)^a
	Patients don’t want to discuss	15 (21%)^a
	I don’t have the time	12 (16%)^a
	Not applicable as I always follow this recommendation	94 (56%)
3. Do not prescribe antibiotics for uncomplicated respiratory tract infections	Prescribed for patients at high risk of complications	91 (73%)^a
	Request or insistence of patient	58 (46%)*
	For infections lasting more than 10 days	54 (43%)*
	Not applicable as I always follow this recommendation	42 (25%)
4. Do not obtain chest radiography in the absence of suspected lung pathology	Requested by surgeon	78 (68%)^a
	Not applicable as I always follow this recommendation	53 (32%)
5. Do not continue long-term use of proton pump inhibitors without titrating to the lowest dose needed	PPIs continued to avoid recurrent symptoms	61 (46%)^a
	I allow patients to decide whether to continue medication	36 (27%)^a
	PPIs started by another physician	30 (22%)^a
	PPIs continued to avoid complications	18 (13%)^a
	Not applicable as I always follow this recommendation	33 (20%)

^a Those responding ‘not applicable’ excluded from totals for percentages to other responses.

Table 5. Factors associated with following fewer of the ‘Smarter Medicine’ Top five recommendations (n = 159)^a.

Characteristic	Univariate incidence rate ratio (95%CI)^b	P	Multivariate incidence rate ratio (95%CI)^b	P
Sex (women)	0.98 (0.77–1.25)	0.86	–
Age, years	1.00 (0.99–1.01)	0.91	–
Language area, German	0.75 (0.62–0.92)	0.01	1.05 (0.84–1.30)	0.69
Country of medical training, outside Switzerland	1.08 (0.73–1.61)	0.69	–
Experience in practice, years	1.00 (0.99–1.01)	0.90	–
Practice size, five or more physicians	0.82 (0.56–1.19)	0.30	–
Practice location, rural	1.01 (0.81–1.26)	0.91	–
In-practice pharmacy, yes	0.70 (0.57–0.86)	<0.01	0.71 (0.57–0.88)	0.002
In-practice laboratory, yes	0.94 (0.66–1.35)	0.74	–
In-practice radiology, yes	0.93 (0.75–1.16)	0.53	–
Knows of ‘Smarter Medicine’	0.91 (0.73–1.14)	0.42	–	–
Mean agreement with recommendations ≤8^c	1.62 (1.34–1.96)	<0.01	1.60 (1.33–1.94)	<0.001

^a Defined as responding ‘nearly always’ for discussing prostate cancer screening prior to testing and ‘rarely or never’ to the other four clinical scenarios.

^b Incidence rate ratio gives the ratio for following one additional recommendation.

^c Mean agreement for individual GPs with the five recommendation statements.