Hands on: Is there an association between doing procedures and job satisfaction?

Article Figures & Data

Table 1

Characteristics of sample (potential confounding variables)

POTENTIAL CONFOUNDING VARIABLES	N (%)
Age
• <35	2079 (10.3)
• 35–44	6458 (31.9)
• 45–54	7216 (35.6)
• 55–64	3287 (16.2)
• ≥65	1229 (6.1)
TOTAL^*	20 269 (100)
Sex
• Male	13 283 (65.1)
• Female	7134 (34.9)
TOTAL^*	20 417 (100)
Setting
• Solo practice	5323 (26.2)
• Group practice	15 033 (73.8)
TOTAL^*	20 356 (100)
Population served by practice
• Urban	12 474 (65.0)
• Semiurban	3679 (19.2)
• Rural	3045 (15.9)
TOTAL^*	19 198 (100)
Teaching
• Yes	4500 (22.2)
• No	15 733 (77.8)
TOTAL^*	20 233 (100)
Constraints to medical care services
• No problem	1186 (6.2)
• Minor problem	4168 (21.7)
• Moderate problem	9382 (48.8)
• Severe problem	4493 (23.4)
TOTAL^*	19 229 (100)
Balance of personal and professional commitments
• About right	5130 (25.2)
• Not right	15 195 (74.8)
TOTAL^*	20 325 (100)

↵* Totals vary due to missing data; total data set represented 20 507 family physicians.

Table 2 Percentage of physicians doing various procedures (predictor variable is range of procedures)

Mean number of various procedures done by physicians was 6.85 (± 3.32).

PROCEDURE	% DOING PROCEDURE (N = 20 238)
Pap smear	94.7
Suturing	81.8
Other minor surgery	71.1
Musculoskeletal injections or aspirations	67.6
Skin biopsy	59.8
Intrauterine device insertion	49.3
Casting or splinting	42.5
Needle aspiration for diagnosis or biopsy	40.9
Electrocardiogram interpretation	40.2
Anoscopy	34.3
Other procedures	26.5
Pulmonary function testing	19.7
Lumbar puncture	15.1
Other biopsy	14.9
Other endoscopy	13.9
Dilatation and curettage aspiration	7.5
Audiometry	7.0
Refraction	1.9

Table 3 How confounding variables relate to predictor variable and outcome variable:

Differences noted were based on P < 0.1.

POTENTIAL CONFOUNDING VARIABLE	RELATIONSHIP TO RANGE OF PROCEDURES	P VALUE	RELATIONSHIP TO OVERALL JOB SATISFACTION	P VALUE
Age	Younger physicians did significantly more procedures	.0001	Significant differences based on age	.0001
Sex	Male physicians did significantly more procedures	.0001	Significantly greater for male physicians	.019
Group or solo practice	Physicians in groups did significantly more procedures	.0001	Physicians in solo practice were significantly more satisfied	.0001
Population served by practice (urban, semiurban, rural)	Rural physicians did significantly more procedures than semiurban physicians, who did significantly more than urban physicians	.0001	Rural physicians were more satisfied than semiurban physicians, who were more satisfied than urban physicians	.0001
Number of medical services offered	Physicians who provided more services did significantly more procedures	.0001	No significant difference	.854
Teaching (yes, no)	Physicians who taught did significantly more procedures	.0001	Teachers were significantly more satisfied	.0001
Constraints to medical care services	Significantly different among groups	.0001	Physicians with more constraints were significantly less satisfied	.0001
Balance of personal and professional commitments	Physicians who reported the balance was about right did significantly fewer procedures	.0001	Those with the balance about right were significantly more satisfied	.0001

Table 4 Multiple regression analysis of overall job satisfaction with predictor range of procedures and confounding variables

N = 16 877.*

PREDICTOR VARIABLES	STANDARDIZED BETA	T	P
Range of procedures	.040	4.906	.0001
Age	.028	3.487	.0001
Sex	−.014	−1.736	.083
Group or solo practice	.042	5.419	.0001
Population served by practice:
urban vs rural	−.101	−9.899	.0001
semiurban vs rural	−.024	−2.422	.015
Teaching	.052	6.794	.0001
Constraints to medical care services	−.227	−30.412	.0001
Balance of personal and professional commitments	.140	18.641	.0001

↵* Total varies from the sample size of 20 507 because analysis was conducted only on physicians with complete data for all variables. All the predictor variables correlated with overall job satisfaction (R = 0.302), therefore, all the predictor variables explain 9.1% of the variance (R2 = .91).