Abstract

Background: Several studies have shown great differences in physicians’ way to sick list. The roles of physician-related factors and local structural factors on the length of the sick leaves have been ambiguous. The aim was to examine the variation in short-term sick-listing practices among primary care physicians. Methods: A questionnaire study with 19 hypothetical patient cases was conducted among 300 Finnish primary care physicians. The effects of both physician related and local structural background variables on sick leave prescribing were studied using univariate and multiple linear regression models. Economic consequences of the variation in sick leave prescribing were estimated. Results: On an average, the overall number of sick leave days prescribed for the entire group of the 19 patient cases was 97.4, varying between 42 and 165 days. The economic consequences to the society of the sick leaves prescribed to them would be €29 442 on average, varying between €11 837 and €51 613. Clinical specialists prescribed shorter sick leaves than general practitioners, with estimated costs of €27 888 and €30 789, respectively. More days of sick leave was prescribed in smaller municipalities than in larger ones. Conclusion: There is a lot of variation in physicians’ sick leave prescribing practices and it depends both on physician-related factors and local structural factors. The speciality status of a physician was the most significant single factor affecting the variation. Notable savings for the society might be possible to achieve by increasing sick-listing education and training.

Introduction

Several studies have shown great differences in physicians’ way to sick list.1–5 In general practice, certification for sickness absence has been estimated to be the single most expensive item, even more expensive than drug prescriptions.1,6 When comparing the costs of prescribed tests, procedures, medication, sick leave and referrals as measures taken by a physician, the costs for sick leave constituted 59% of the total costs, being significantly the most cost-creating measure.1

In a case-simulation study, there was no correlation between the sex of the physician and the sick leave prescribed in primary care.1,7 However, in another case-simulation study, male physicians sick listed 43% and female physicians 58% of their patients, and the cost of sick listing would have been significantly higher among female than male physicians.2 In case of long sick leaves, no differences between male and female physicians’ sick-listing practices were found in a Swedish study,8 but in another study in UK, male physicians reported issuing a higher average number of certificates than female physicians.9

The age of the physicians and the duration of sick leaves have shown positive1,6,10 or no correlation.8 Instead, the age of the patients has been shown to be associated with longer sick leave episodes.3,11

In Finnish health-care setting, all physicians after graduation are called general practitioners (GPs). Additional qualifications, also called specialities, can be obtained through 6-year training programmes. One of these programmes concentrates on family medicine and those who have completed this training have the degree of clinical speciality (CS) in family practice. Although EU has standardized medical training in a large part of Europe, notable differences still exist between European countries.

The effect of the physician's speciality or additional qualifications on sick listing is ambiguous. In some studies, physicians with speciality degree have been shown to prescribe shorter or less sick leaves,2,6,10 and some other studies have found no difference1 or the opposite.2,3

Local structural factors may also have influence. Physicians who practiced farther away from hospitals tended to prescribe more sick leaves than those closer to a hospital.1 However, in another study, physicians in municipalities without a hospital issued shorter certificates.4 Furthermore, the longest certificates of sick leave were issued in the smallest and largest municipalities.4

In Finnish health-care system, all physicians are equally qualified to prescribe sick leaves and it is a part of every physician's daily clinical practice. National Social Insurance system covers sick leave costs from the ninth sick leave day onwards. Before this, the costs are covered by employers and individuals. Thus, no registers or readily available data exists on shorter sick leaves.

The purpose of this study was to examine how much variation there is in short-term sick-listing practices of primary care physicians and which physician-related factors and local structural factors affect the variation.

Methods

From the physician register kept by the Finnish Medical Association, 300 physicians working in primary health care in Finland were randomly selected, representing ∼10% of all such physicians. The sample was proportional to the number of practicing physicians in the five districts of Finland.

A questionnaire form together with a prepaid return envelope was sent in the spring of 2009 to the selected physicians. The forms were marked to detect participants in order to enable repeated mailings to non-participants only. Otherwise, the questionnaires were anonymous. Three attempts were made to contact the non-participants by phone. If reached, they were motivated to participate, and new questionnaire forms with prepaid return envelopes were mailed. The final number of returned questionnaires was 165, of which two were unanswered and a few were partly filled.

In the beginning of the questionnaire form, the purpose of the study was described and the recipients were told that participation was voluntary. The first part of the questions covered socio-demographic background such as age, sex and the physician's working municipality. This was followed by questions dealing with years worked as a physician, whether under training or having a degree in a clinical speciality, area of speciality and whether employed directly by the health authority (table 1).

Table 1

Descriptive characteristics of the sample physicians

Mean Sample All Finnish physicians 
Age, mean (SD) 45.2 (11.2) 45 
Years practiced, mean (SD) 18.0 (11.1) NA 
Women 74.8 56.9 
Specialists 50.6 60.1 
Family medicine 40.1 9.4 
Other 10.5 50.7 
Residents 28.0 NA 
Region of Finland   
    Southern 29.5 38.0 
    Western 17.3 13.9 
    Central 24.4 21.0 
    Eastern 16.0 14.2 
    Northern 12.8 12.8 
Municipal population <100 000 56.4 NA 
Employed directly by health authority 83.6 74 
Central hospital in working municipality 59.4 NA 
Mean Sample All Finnish physicians 
Age, mean (SD) 45.2 (11.2) 45 
Years practiced, mean (SD) 18.0 (11.1) NA 
Women 74.8 56.9 
Specialists 50.6 60.1 
Family medicine 40.1 9.4 
Other 10.5 50.7 
Residents 28.0 NA 
Region of Finland   
    Southern 29.5 38.0 
    Western 17.3 13.9 
    Central 24.4 21.0 
    Eastern 16.0 14.2 
    Northern 12.8 12.8 
Municipal population <100 000 56.4 NA 
Employed directly by health authority 83.6 74 
Central hospital in working municipality 59.4 NA 

Values are represented as percentages unless otherwise specified.

NA = not available

Next followed 19 hypothetical case simulations representing typical patients in primary health care. The cases were created with experienced specialists in family practice and a pilot study was carried out before the main study was performed. This was done to ensure that the cases were easy to understand and that the answering was effortless. All cases included information about patient's age, sex, occupation, symptoms in short and the diagnosis. Cases were described in short so that they would widely represent the patients of primary care (Supplementary data). The physician was asked to consider each case as a typical non-complicated patient attending primary visit on a Monday morning and to evaluate how many days of sick leave he/she would prescribe to each patient according to his/her usual practice. Eighteen different diagnoses were chosen for the cases, two patients were diagnosed for depression: a 26-year-old woman and a 45-year-old man. In a few cases, some respondents had answered with a range of days, and in these cases the mean value was used.

For statistical analyses, additional variables were formed: The population of the physician's working municipality was collected from the Statistics Finland on ground of the postal code. This was used as such and also a dichotomy was formed: (0) municipality under 100 000 inhabitants and (1) 100 000 inhabitants and over. For each municipality, the type of available health-care facilities were determined according to national hierarchy and classified as: (0) only a health centre; (1) a health centre hospital or a district hospital; (2) central hospital; and (3) university hospital. Also, a dichotomy was formed: (0) no central hospital or university hospital and (1) central hospital or university hospital. The physician's speciality status was classified as: (0) no clinical speciality degree, i.e. GP and (1) clinical speciality degree. The physician's clinical working experience in years was used both as a continuous variable and dichotomized as: (0) <20 years and (1) ≥20 years. The dichotomy of working experience was used in the final analyses. The physician's employment status was asked and a dichotomy was formed: (0) not employed directly by a health authority and (1) direct employment.

Diagnoses were also classified into groups. The musculoskeletal group included the cases of lumbago, tension neck, tennis elbow and sprain of the ankle. The psychiatric group included two depression cases, burn out and hangover. The trauma group included wound, sprain of the ankle and burn cases. The infection group included common cold, pneumonia, gastroenteritis, shingles and urethritis. The chronic diseases group included tension neck, vertigo, eczema and COPD. To depict a hypothetical 1-day work load of a primary health-care physician, an overall variable was formed by summing up the sick leave days of all 19 cases.

To estimate the possible economic consequences of varying sick leave prescribing practices, the age, sex and occupation-specific mean salaries from the official statistics were used for each patient case. As all cases required relatively short sick leaves, possible social security reimbursements were not applicable. The production losses of varying sick leave prescribing practices were estimated applying the computation formula of The Finnish Employers’ Association: Total productivity loss of a sick leave day equals three times the daily gross salary.

The statistical evaluation of the data was based on Student's t-test and chi-squared test for means and proportions, respectively. The distributions of sick leave days were close to normal. Univariate linear regression models were used for each patient case, and both fixed and stepwise ordinary least square multiple linear regression models with selected background variables were fitted for the overall number of sick leave days.

Some of the studied background variables were significantly correlated. In order to control for the possible effect of multicollinearity, various combinations of independent variables were also fitted in several regression models. These did not, however, produce estimates that would be significantly different from the complete set of explaining variables, and the final models were based on the full set of independent variables.

Results

A majority of the physicians who participated in the study was female. Over a half of the physicians were clinical specialists and family medicine was the most common speciality (table 1).

On an average, the overall number of sick leave days prescribed for the entire group of the 19 patient cases was 97.4, varying between 42 and 165 days. Using the evaluation formula of The Finnish Employers’ Association, the economic consequences to the society of the sick leaves prescribed to all 19 cases by one physician would be €29 442 on an average, varying between €11 837 and €51 613. The variations of all 19 cases and all diagnoses together were larger than random (P < 0.001).

CSs prescribed shorter sick leaves than GPs. The differences were statistically significant for six cases. The same pattern, although non-significant, was observed also in most other cases. The variation in sick leave length between physicians was smaller among CSs than GPs in most diagnoses (table 2).

Table 2

Mean amount of sick leave days in the different diagnoses from GPs, CS and altogether

Diagnosis GP (SD) CS (SD) All (SD) 
M54.5 Lumbago 5.2 (2.2) 4.6 (1.6) 4.9 (1.9)a 
G44.2 Tension neck 2.3 (1.1) 2.2 (1.1) 2.3 (1.1) 
M77.1 Tennis elbow 7.7 (3.5) 6.6 (2.8) 7.2 (3.2)a 
J06.9 Common cold 2.4 (0.9) 2.5 (1.0) 2.4 (0.9) 
J18.9 Pneumonia 9.7 (3.4) 8.8 (2.8) 9.3 (3.1) 
A08.4 Gastroenteritis 3.0 (1.0) 2.7 (0.7) 2.8 (0.9)a 
B02.9 Shingles 5.7 (2.7) 5.5 (2.5) 5.6 (2.6) 
H81.1 Vertigo 3.0 (1.8) 3.1 (1.6) 3.1 (1.7) 
S61.8 Wound of the hand 7.1 (2.6) 7.0 (2.2) 7.1 (2.4) 
S93.4 Sprain of the ankle 6.6 (2.4) 6.6 (2.7) 6.6 (2.6) 
T24.1 Burn of the thigh 4.8 (2.8) 4.8 (2.9) 4.8 (2.8) 
L20.0 Atopic eczema 2.0 (3.2) 1.0 (1.8) 1.5 (2.6)a 
O47.0 Uterine contractions 4.2 (2.6) 4.0 (2.7) 4.1 (2.6) 
J44.0 COPD 4.4 (1.8) 4.6 (2.1) 4.5 (1.9) 
F32.1 Depression, female 13.8 (7.2) 11.6 (6.7) 12.7 (7.0)a 
F32.1 Depression, male 15.7 (7.9) 13.0 (6.8) 14.4 (7.5)a 
F43Z73 Burnout 2.5 (3.6) 2.1 (3.1) 2.3 (3.4) 
R11 Hangover 0.8 (0.9) 0.8 (0.9) 0.8 (0.9) 
N39.0 Urethritis 1.0 (1.0) 1.0 (0.8) 1.0 (0.9) 
All Diagnoses 101.6 (27.0) 92.5 (22.3) 97.4 (25.3)a 
Diagnosis GP (SD) CS (SD) All (SD) 
M54.5 Lumbago 5.2 (2.2) 4.6 (1.6) 4.9 (1.9)a 
G44.2 Tension neck 2.3 (1.1) 2.2 (1.1) 2.3 (1.1) 
M77.1 Tennis elbow 7.7 (3.5) 6.6 (2.8) 7.2 (3.2)a 
J06.9 Common cold 2.4 (0.9) 2.5 (1.0) 2.4 (0.9) 
J18.9 Pneumonia 9.7 (3.4) 8.8 (2.8) 9.3 (3.1) 
A08.4 Gastroenteritis 3.0 (1.0) 2.7 (0.7) 2.8 (0.9)a 
B02.9 Shingles 5.7 (2.7) 5.5 (2.5) 5.6 (2.6) 
H81.1 Vertigo 3.0 (1.8) 3.1 (1.6) 3.1 (1.7) 
S61.8 Wound of the hand 7.1 (2.6) 7.0 (2.2) 7.1 (2.4) 
S93.4 Sprain of the ankle 6.6 (2.4) 6.6 (2.7) 6.6 (2.6) 
T24.1 Burn of the thigh 4.8 (2.8) 4.8 (2.9) 4.8 (2.8) 
L20.0 Atopic eczema 2.0 (3.2) 1.0 (1.8) 1.5 (2.6)a 
O47.0 Uterine contractions 4.2 (2.6) 4.0 (2.7) 4.1 (2.6) 
J44.0 COPD 4.4 (1.8) 4.6 (2.1) 4.5 (1.9) 
F32.1 Depression, female 13.8 (7.2) 11.6 (6.7) 12.7 (7.0)a 
F32.1 Depression, male 15.7 (7.9) 13.0 (6.8) 14.4 (7.5)a 
F43Z73 Burnout 2.5 (3.6) 2.1 (3.1) 2.3 (3.4) 
R11 Hangover 0.8 (0.9) 0.8 (0.9) 0.8 (0.9) 
N39.0 Urethritis 1.0 (1.0) 1.0 (0.8) 1.0 (0.9) 
All Diagnoses 101.6 (27.0) 92.5 (22.3) 97.4 (25.3)a 

aP < 0.05

Male physicians prescribed significantly (P < 0.05) longer sick leaves to wound diagnose [7.7 (2.2)] and burn diagnose [5.7 (3.8)] than female physicians {wound [6.8 (2.5) and burn [4.5 (2.3)]}. Female physicians did not prescribe significantly shorter or longer sick leaves to female patients than male patients and male physicians did not prescribe significantly shorter or longer sick leaves to male patients than female patients.

The number of sick leave days was higher and the variation between physicians was greater in psychiatric diagnoses [30.2 (14.9)] than in other groups [19.2 (4.3)] (P = 0.001). In smaller municipalities, the number of prescribed sick leave days would have been greater in all diagnostic groups than in larger municipalities; however, the difference was significant only in trauma diagnoses.

The greatest differences between CSs [27.0 (13.4)] and GPs [33.1 (16.3)] occurred in psychiatric diagnoses (P < 0.02), which was the group with the longest sick leaves overall. CSs prescribed less sick leave days than GPs to all other diagnostic groups too, but these differences were non-significant. Male physicians prescribed significantly (P < 0.05) longer [20.1 (6.6)] sick leaves to trauma diagnoses than female physicians [17.9 (5.4)].

Regarding the overall sick leaves to all 19 patient cases, CSs prescribed significantly shorter sick leaves than GPs. The estimated average cost to the society of the overall sick leaves prescribed by CSs would be €27 888, while that of GPs would be €30 789. More days of sick leave was prescribed in smaller municipalities than in larger ones (table 2). The overall-length of the sick leaves was not affected by the physician's gender.

In a set of univariate regression models, the physician's speciality and the studied structural factors were significantly associated with the number of prescribed sick leave days (table 3). However, the presence of a central hospital in the municipality and a larger population in the municipality were highly significantly (P < 0.001) associated. In several multivariate regression models, the size of the municipality better explained the variation in the number of sick leave days (table 4).

Table 3

Univariate linear regression analyses of selected background variables on overall days of sick-leave for all 19 cases

Variable β SE t-value P-value R2 
Age 0.083 0.205 0.404 <0.687 0.001 
Sex 4.472 4.985 0.897 <0.371 0.006 
Experience ≥20 years 1.189 4.298 0.277 <0.783 0.001 
Central hospital in working municipality −8.887 4.452 −1.996 <0.048 0.030 
Municipal population >100 000 −9.189 4.350 −2.112 <0.037 0.033 
Clinical speciality −9.064 4.229 −2.143 <0.034 0.032 
Employed directly by health authority 14.525 5.760 2.522 <0.013 0.044 
Variable β SE t-value P-value R2 
Age 0.083 0.205 0.404 <0.687 0.001 
Sex 4.472 4.985 0.897 <0.371 0.006 
Experience ≥20 years 1.189 4.298 0.277 <0.783 0.001 
Central hospital in working municipality −8.887 4.452 −1.996 <0.048 0.030 
Municipal population >100 000 −9.189 4.350 −2.112 <0.037 0.033 
Clinical speciality −9.064 4.229 −2.143 <0.034 0.032 
Employed directly by health authority 14.525 5.760 2.522 <0.013 0.044 
Table 4

Multiple linear regression models of all selected background variables and final stepwise model of three statistically significant variables on overall days of sick-leave for all 19 cases

Variable All background variables
 
Final stepwise model
 
 β SE t-value P-value β SE t-value P-value 
Constant term 87.554 14.484 6.045 <0.001 95.234 6.145 15.498 <0.001 
Age 0.250 0.363 0.688 <0.493     
Sex 3.149 5.033 0.626 <0.533     
Experience 20 years or more −1.536 7.444 −0.206 <0.837     
Central hospital in working municipality −3.986 5.794 −0.688 <0.493     
Municipal population >100 000 −6.891 5.737 −1.201 <0.232 −9.195 4.247 −2.165 <0.032 
Clinical speciality −11.015 4.911 −2.243 <0.027 −8.908 4.204 −2.119 <0.036 
Employed directly by health authority 11.127 6.199 1.795 <0.075 12.422 6.003 2.069 <0.041 
R2 0.111     0.0   
Variable All background variables
 
Final stepwise model
 
 β SE t-value P-value β SE t-value P-value 
Constant term 87.554 14.484 6.045 <0.001 95.234 6.145 15.498 <0.001 
Age 0.250 0.363 0.688 <0.493     
Sex 3.149 5.033 0.626 <0.533     
Experience 20 years or more −1.536 7.444 −0.206 <0.837     
Central hospital in working municipality −3.986 5.794 −0.688 <0.493     
Municipal population >100 000 −6.891 5.737 −1.201 <0.232 −9.195 4.247 −2.165 <0.032 
Clinical speciality −11.015 4.911 −2.243 <0.027 −8.908 4.204 −2.119 <0.036 
Employed directly by health authority 11.127 6.199 1.795 <0.075 12.422 6.003 2.069 <0.041 
R2 0.111     0.0   

Discussion

We found a lot of variation in physicians’ sick leave prescribing practices. The variation was related both to physician-related and local structural factors. A previous study of six written patient cases has shown a 6-fold difference in total cost of physicians’ practices. Sick leaves formed the majority of those costs, although other costs too were included. The almost 4-fold difference in our study seems to corroborate their finding of the magnitude of the variation.1

Because Finnish National Social Insurance registers only sick leaves from the ninth day onwards, case simulation study design was chosen to cover shorter sick leave prescribing practices also. The selected 19 patient cases were expected to represent the patient load of 1 working day for a physician in Public Health Service. Also, the sample was drawn from those physicians registered to work in the Public Health Service in Finland. According to our finding, the estimated costs of the sick leave prescribing practice to the society can be considerable. The difference between the most ‘cost-saving’ and the most ‘spendthrifting’ physician in the public health service would have been almost €40 000/day. This estimate corroborates the earlier finding that sick leave prescribing causes significant economic consequences to the society.1,6

In Finland, the physicians have no guidelines which could be even indicative to direct sick leave prescribing. All sick leave prescribing solely depends on each physician's impressions and customs. Particularly younger physicians in the beginning of their clinical career may face difficulties when deciding how many days of sick leave could be relevant in each patient case. For them, as well as for others with a less risk taking attitude, it may be tempting to work on the safe side and prescribe longer sick leaves.12 However, in the light of our cost estimates, this kind of attitude or uncertainty may have devastating economic corollaries. The patients of those physicians who initially prescribe shorter sick leaves may return to work earlier than medically appropriate and they may end up needing more sick leave days on the long run.

The finding that clinical specialists prescribe shorter sick leaves than GPs has been shown also in previous studies.6,10 However, the physician's experience did not seem to affect the length of the sick leave. Increasing the training of sick leave prescribing both at under- and postgraduate levels may reduce the economic burden of sick listing.

According to a previous study, the longest episodes of sick leaves were prescribed in the smallest and largest municipalities.4 In our study, the physicians working in smaller municipalities prescribed longer sick leaves than those working in larger municipalities. The physicians working in larger municipalities have more opportunities to consult various specialties. In dense-populated areas, patients live nearer the health centres, have better access to care and possible reappointments are easier for the patients to arrange. Also emergency health care is more readily available in larger municipalities. Such structural factors could influence the physician's practice to prescribe sick leaves in larger municipalities.

The effect of physician's sex on sick-listing practices has been ambiguous in previous studies. However, most of the former reports have not been able to show the association between sick leave lengths and physician's sex.1,6,8 In this study, no gender-related trend was found, but a significant difference between male and female physicians occurred in a few patient cases. Our study corroborates the view that physician's gender has only a minor possible effect on the length of sick leaves. However, this is not necessarily controversial to the finding in UK, that male GPs issued a higher number of certificates.9

With the exception of the presence of a central hospital in the working municipality, the effects of the independent variables in various regression models remained practically the same. This can be interpreted as evidence for stability and reliability of the effects. Although we found several factors that influenced the physicians' sick-listing practices, these factors explain only 11.1% of all the variation. One factor which could explain part of the remaining variation could be the personal risk taking willingness and attitudes of the physicians.10,12 However, such factors were not included in the present study.

We used hypothetical patient cases to be able to compare physicians’ practices more reliably. There is not a consensus on how well the results of the case studies represent the actual behaviour of physicians.13 The results of a study of patients with osteoarthritis support the use of written cases as a valid proxy measure of trends in the actual practices of physicians.14 It has been proposed that this method makes the physicians do their best, when the results represent the ideal situation without distractions.13,15 The respondents were asked to think each case as a typical, non-complicated patient at their first visit on a Monday morning, so that the bases for the assessments would be as similar as possible. These actions were taken to diminish the variation and to increase the reliability in depicting the sick-listing practices of physicians.

There is a lot of variation in physicians’ sick leave prescribing practices and it depends both on physician-related factors and local structural factors. The effect of additional qualifications and speciality training on sick-listing practices would require further investigation. Our findings indicate that physicians with speciality training would prescribe shorter sick leaves, but some earlier studies suggest that those with additional qualifications or longer experience in Family Medicine issue more sickness certificates.7,9 Notable savings for the society might be possible to achieve by increasing sick-listing education and training.

Supplementary data

Supplementary data are available at Eurpub online.

Funding

The Hospital District of Southwest Finland.

Conflicts of interest: None declared.

Key points

  • There is a large variation in primary care physicians’ sick-listing practice.

  • The aim of this study was to examine whether the variation might be influenced by physician-related or local structural factors.

  • The speciality status of a physician was the most significant single factor affecting the variation.

  • This means that notable cost savings could be achieved by increased education and training.

References

1
Peterson
S
Eriksson
M
Tibblin
G
Practice variation in Swedish primary care
Scand J Prim Health Care
 , 
1997
, vol. 
15
 (pg. 
68
-
75
)
2
Englund
L
Tibblin
G
Svärdsudd
KF
Variations in sick-listing practice among male and female physicians of different specialities based on case vignettes
Scand J Prim Health Care
 , 
2000
, vol. 
18
 (pg. 
48
-
52
)
3
Arrelöv
BE
Borgquist
L
Ljungberg
D
Svärdsudd
KF
Do GPs sick-list patients to a lesser extent than other physician categories? A population-based study
Fam Pract
 , 
2001
, vol. 
18
 (pg. 
393
-
8
)
4
Arrelöv
BE
Borgquist
L
Svärdsudd
KF
Influence of local structural factors on physicians’ sick-listing practice: a population-based study
Eur J Public Health
 , 
2005
, vol. 
5
 (pg. 
470
-
4
)
5
Wahlström
R
Alexanderson
K
Chapter 11. Physicians’ sick-listing practices
Scand J Public Health
 , 
2004
, vol. 
32
 
Suppl 63
(pg. 
222
-
55
)
6
Rutle
O
Forsen
L
Allmenpraksis-tema med variasjoner. [Inter-doctor variation in general practice.] Oslo: National Institute of Public Health, Department of Health Services Research
  
Report No: 8,1984
7
Norrmén
G
Svärdsudd
K
Andersson
D
Impact of physician-related factors on sickness certification in primary health care
Scand J Prim Health Care
 , 
2006
, vol. 
24
 (pg. 
104
-
9
)
8
Karlsryd
E
Det nya särskilda läkarutlåtandet fr o m den 29:e dagen. [New separate physician's statement beginning from 29th day. Authors’ translation.]
Socialmedicinsk tidskrift
 , 
1997
, vol. 
8–9
 (pg. 
426
-
9
)
9
Wynne-Jones
G
Mallen
CD
Main
CJ
Dunn
KM
Sickness certification and the GP: what really happens in practice?
Fam Pract
 , 
2010
, vol. 
27
 (pg. 
344
-
50
)
10
Tellnes
G
Sandvik
L
Torbjorn
M
Inter-doctor variation in sickness certification
Scand J Prim Health Care
 , 
1990
, vol. 
8
 (pg. 
45
-
52
)
11
Reiso
H
Nygård
JF
Brage
S
, et al.  . 
Work ability and duration of certified sickness absence
Scand J Public Health
 , 
2001
, vol. 
29
 (pg. 
218
-
25
)
12
Mabeck
CE
Kragstrup
J
Is variation a quality in general practice?
Scand J Prim Health Care
 , 
1993
, vol. 
11
 
Suppl 1
(pg. 
32
-
5
)
13
Jones
TV
Gerrity
MS
Earp
J
Written case simulations: do they predict physicians’ behaviour?
J Clin Epidemiol
 , 
1990
, vol. 
43
 (pg. 
805
-
15
)
14
Holt
W
Jr
Mazzuca
SA
Prescribing behaviors of family physicians in the treatment of osteoarthritis
Fam Med
 , 
1992
, vol. 
24
 (pg. 
524
-
7
)
15
Rethans
JJ
Validity of case simulations
J Clin Epidemiol
 , 
1991
, vol. 
44
 (pg. 
1119
-
20
)

Comments

0 Comments