Sickness certification in primary care – the physician ’ s role

Background: Sickness certification is a routine task of primary care (PC) physicians which has an impact on patients’ health, the health care system and the economy. As sickness certification is poorly studied, we quantified sickness certification and explored qualitatively the sickness certification process by Swiss PC physicians. Methods: PC physicians participating in the Swiss Sentinel Surveillance Network (SSSN) recorded the frequency and duration of absence from work of each related consultation and certificate during 2005. Patients’ age and gender, reason for sick leave, psychosocial cofactors, problems at the workplace, type of employment, type of occupation, duration of absence (weeks) and type of certificate were registered. Physicians’ views on the procedure and their suggestions for change were gathered before and after the study by means of a questionnaire containing four open-ended questions. Results: Of the 223 SSSN physicians 73% participated. A total of 24,676 forms issued by 150 physicians were analysed. An average of four certificates was issued per 100 consultations; somewhat fewer by internists than by general practitioners and less in rural areas than urban areas. Psychosocial or work-related factors were mentioned in 20% of the certificates and were more often associated with longer absences from work. These factors were seen as inseparable from the somatic factors. Recommendations for change included the prolongation of self declaration time, a uniform declaration form, availability of an authority to which complex cases can be referred and the use of case management models. Conclusions: Sickness certificates were issued in 4% of GP consultations. This task has been assessed by physicians as part of their function. The certification process should be improved through better coordination and communication between all parties involved: patients, employers, insurers, physicians and politicians.

Background: Sickness certification is a routine task of primary care (PC) physicians which has an impact on patients' health, the health care system and the economy.As sickness certification is poorly studied, we quantified sickness certification and explored qualitatively the sickness certification process by Swiss PC physicians.
Methods: PC physicians participating in the Swiss Sentinel Surveillance Network (SSSN) recorded the frequency and duration of absence from work of each related consultation and certificate during 2005.Patients' age and gender, reason for sick leave, psychosocial cofactors, problems at the workplace, type of employment, type of occupation, duration of absence (weeks) and type of certificate were registered.Physicians' views on the procedure and their suggestions for change were gathered before and after the study by means of a questionnaire containing four open-ended questions.
Results: Of the 223 SSSN physicians 73% participated.A total of 24,676 forms issued by 150 physicians were analysed.An average of four cer-tificates was issued per 100 consultations; somewhat fewer by internists than by general practitioners and less in rural areas than urban areas.Psychosocial or work-related factors were mentioned in 20% of the certificates and were more often associated with longer absences from work.These factors were seen as inseparable from the somatic factors.Recommendations for change included the prolongation of self declaration time, a uniform declaration form, availability of an authority to which complex cases can be referred and the use of case management models.
Conclusions: Sickness certificates were issued in 4% of GP consultations.This task has been assessed by physicians as part of their function.The certification process should be improved through better coordination and communication between all parties involved: patients, employers, insurers, physicians and politicians.

Key words: sickness absence; sickness certification; Swiss Sentinel Surveillance Network
Sickness certification is a routine and frequent task of primary care (PC) physicians.It has a significant impact on sick or injured patients, on the health care system and on the economy.Except for the UK and Scandinavian countries, the issue has received little research attention by the medical profession [1,2].Certification is usually based on subjective information and symptoms without necessarily a precise diagnosis [3][4][5].Additionally, physicians have a conflicting role, being simultaneously the patient's advocate and the guarantor of an honest declaration to the employer or insurance company.Consequently certification for sick leave is the result of a negotiation between doctors and patients [6,7].Indeed, a recent qualitative study suggests that general practitioners in the UK implement strategies such as acquiescence, negotiation and challenging patients when issuing sickness certificates.They consider that their responsibility to patients outweighs their responsibility to the state and that challenging patients would endanger the doctor-patient relationship.The burden of sickness certification on general practitioners (GPs) in the UK is so onerous that one half of the GPs wished that their certification role was removed [8].

Introduction
certification in Switzerland [9].In the present study we took advantage of the information generated during the year 2005 by the Swiss Sentinel Surveillance Network (SSSN) [10] to get more representative information.The SSSN serves as a monitoring system for infectious diseases and collects data relevant to health problems in the primary health care sector.Data are collected from a volunteer network of primary care physicians (GPs, general internists and paediatricians) and transmitted routinely each week to the Swiss Federal Office of Public Health (SFOPH) for processing.We aimed to measure the frequency of sickness certification, the duration of certified absence from work, and to explore qualitatively how primary care physicians viewed sickness certification and their suggestions for change.

Methods
Physicians with a primary care activity participating in the SSSN recorded all sickness certificates issued between January 1 st and December 31 st , 2005.Characteristics of physicians participating in the study were ascertained: age, gender, location of practice and number of consultations per week.Around 90% of the Sentinel participants are regularly reporting (at least 39 weeks per year).In this study only regularly reporting participants were included.Sentinel physicians account for about 3% of all primary care physicians in Switzerland.The turnover of participants is about 30% each year.Sentinel practices are stratified by geographic area, socio-demographic characteristics and physician speciality [11].Most participants were GPs (57%), followed by internists (29%) and paediatricians (14%) in 2005.The respective proportions of these three specialty groups were 42%, 44% and 14% according to the Swiss Medical Association.Age distribution and postgraduate education of SSSN physicians and of all Swiss physicians having a primary care activity do not differ much.However, SSSN physicians practising in the large cities are underrepresented compared to their colleagues in rural areas.
Patient specific details included age, gender, reason for sick leave (illness, accident, surgical intervention), psychosocial cofactors, problems at the workplace, type of employment (employee or self-employed), type of occupation (manual/other), certified duration of absence from work (<1, 1-3, >3 weeks), type of certificate (initial, continuation, final).

Physician questionnaire
PC physicians completed a questionnaire (four open questions) at the beginning and at the end of the survey (pre-/post-survey).The questions were: 1.How do you, as a primary care practitioner, view the act of certifying absence from work?Describe positive and negative aspects. 2. To which extent do psychological and social factors influence your decision?3. Which changes to the system do you suggest? 4. Which additional comments do you wish to make?

Data analysis
Descriptive data concerning the PC physicians and the certificates issued were examined in relation to physicians' characteristics.Rates of issues were calculated per 100 consultations.Case available analyses were used to cope with missing data.Because the sample was non-random and clustered, an inferential approach would have been inappropriate.Thus proportions were given without confidence intervals and without p-value for group comparisons.As multivariate modelling may have led to erroneous interpretation, this approach was disregarded.
The responses to the physician questionnaires were transcribed and analysed by the first author (UB) according to the method described by Mayring [12].Frequently mentioned terms or concepts were categorised and linked with annotations in a matrix-like structure so that complex phenomena hidden in the answers could be uncovered.The same procedure was repeated with the information obtained at the end of the survey in order to check whether categories were appropriate.Paired questionnaires, i.e. those which had been answered twice (pre-survey and post-survey) were compared to find out whether physicians had changed their views markedly over the survey year.

Results
The quantitative analysis was based on 24,847 certificates issued by 162 participating physicians, out of the total of 223 SSSN physicians in 2005.The average age of participating physicians was 52 years (50 for the non-participants), of which 74% was male (70% among non-participants).Participation was highest among general practitioners (81%) and internists (75%), but very low among paediatricians (32%).The proportion of physicians working less than 100% was higher among non-participants (43 vs 30%).Two physicians (52 certificates) were excluded because of missing val-ues of key physician characteristics.Ten paediatricians (119 certificates) were excluded because only 10/31 SSSN paediatricians participated and because they were concerned solely with the certification of children's absence from school.Eventually, 24,676 certificates from 150 GPs and general internists were analysed.Complete information on both the number of certificates and the total number of consultations was available from 139 participants, 23,875 certificates during 591,791 consultations (ie an average of four certificates per 100 consultations).
Patients' characteristics are shown in table 1. Between 16 and 65 years of age, an average of 5000 certificates was prescribed for each age group except in the 55-64 category.Of the certificates 58% were issued for men, 58% for manual workers, and 95% for employees.Less than one fifth (18%) of the certificates were issued in relation to accidents.Over one half (57%) were initial certificates and about half (53%) were issued for 7 days or less.
Overall, psychosocial and work-related cofactors were mentioned on 19% of all certificates.Illnesses were accompanied by psychosocial and work-related cofactors more often than was reported with accidents (22% versus 4%).Longer absences from work (>3 weeks) were associated with a higher proportion of psychosocial and work-related cofactors (table 1).GP's issued 75% of all certificates.They issued 4.3 certificates per 100 consultations compared to 3.6 certificates per 100 consultations issued by internists (table 2).Neither physicians' age nor gender influenced the certification rate, whereas slightly fewer certificates were issued by internists compared to GPs and physicians working in rural areas.

Qualitative aspects
For analysis 78 pre-and post-test questionnaires were available, and their results are described in Boxes 1-3.Responses to the preand post-test questionnaires were very similar.-"No problem: the incapacity to work and the date of resuming work are discussed between the doctor and the patient."-"Employees are cleared of the company's/employer's suspicion that the absence from work is unjustified."-"Certification may offer some relief to a patient in a difficult situation."-"In general, the family physician has known the patient for a long time which makes judgement of incapacity to work easier."

Negative aspects of sickness
At the same time, many PC physicians considered the task of certifying as "a cumbersome task" which poses problems of various kinds.-The system is open to be misused by all parties involved (employees and employers, insurers and doctors).-Certification for absences of long duration is often problematic.-There is a risk for unnecessary consultations for the sole act of issuing a certificate.Examples: -"The doctor is at the patient's 'mercy', as he has to 'believe and certify' what the patient tells him, which is a highly unsatisfying situation."-"The doctor can be put under pressure by the employer: 'When the order situation is bad and there is little work to do, write him off sick and when there is a lot of work to do, send him back to work as soon as possible.'"-"The doctor has insufficient knowledge about the patient's occupation and hesitates to contact the employer."-"Absences of long duration warrant a more detailed certification form."-"Sometimes the patient is consulting only for a certificate and not because of his illness."

Influence of psychological and social factors
The interaction between a somatic illness, or an accident, and psycho-social aspects was seen as normal rather than as a cofactor.
Examples: -"All these factors play a role, of course.Good medicine is built on the bio-psycho-social paradigm."-"All these factors are present at varying degrees.The following well known situations pose a particular problem: backache, depression, burn-out/stress."

Suggested changes
Suggestions for changing the certification system: -The time of self-certification should be extended from three days to seven days.-Employers' and employees' attitudes toward absenteeism must improve.-More attention should be paid to the development of a healthy working environment.-PC physicians must have an easy access to an authoritative agency / professional for delegating complex cases.-Medical education must include the topic on all levels.Free comments mainly focussed on the present study.According to 13 participants the study did not change their views on sickness certifica-tion, and/nor did they find it time consuming.Conversely, 22 participants felt that the study helped them to become more conscious of their certifying task.

Discussion
To our knowledge this study of sickness certification in primary care conducted during the year 2005 represents the largest sample in Switzerland.Most of the findings of this study correspond with those reported from other European countries, such as higher rates of sickness certification among men than women, among manual workers than others, among employees than employers, higher rates for sickness than for accidents, and lower rates in rural areas than in urban areas [13].It is however difficult to compare these data, given the differences in work organisation, healthcare services, periods compared and methods used.One noticeable difference is the rate of certification (4 per 100 consultations) which was much lower than in other European countries where a range of 11-35 per 100 has been reported [13].Primary care physicians in Switzerland have a clear view on the positive and negative aspects of sickness certification and suggested a number of possible improvements: self-declaration for short periods of absence from work, authoritative support and case management.

Quantitative aspects
Many of the findings in this study confirm those from other studies and are logically explained by differing employment patterns by age and gender etc.We believe the data on duration of certificates are new and, though not quantified in days, provide some information on the way physicians cope with this social task: half of the certificates are prescribed for less than 1 week and half of the certificates were declared "continuation" or "final".The latter figure implies that very often several encounters are needed for these cases.Because follow-up certificates were not linked to initial certificates in many cases, we are not able to comment on the total duration of certified absence.In addition, as a patient may receive certificates of absence from various physicians, it would not be possible to summarise all information.Finally, the threshold for issuing sickness certificates varies with the disease or injury, according to its severity, the age of the sufferer and the type of employment; and sometimes also with the opinion of the physician consulted.Psychosocial and workrelated cofactors were mentioned more often in the context of absence from work caused by disease, rather than by accident.The longer the absence from work the more often cofactors were part of the health problem.

Qualitative aspects
To some of the physicians, the overall level of certification appeared to conflict with their preconceived ideas about the workload involved in certification.However, since conflict with patients can arise over the question of fitness for work, sickness certification is sometimes an unwelcome task which easily leads to the assumption that the act of sickness certification is a frequent and demanding task for practitioners.It was widely felt by recording physicians that short-term sickness absence (7 days or less) should be self-certified and dealt with at the workplace, even if it involved spot checks to avoid abuse.Payment for the issue of certificates was considered by some as a means of minimising certification for trifles.The certificate should give a clear indication of the expected duration of absence and should allow flexibility so that persons could return initially on a defined part-time basis.Difficult cases with long periods of sickness absence should be reviewed by an independent physician specifically appointed and trained for his purpose.Case management plans with defined goals relating to the return to work could become an essential part of certification practice.
There are some limitations to this study.Although representative in regional and socio-demographic terms, the SSSN comprises highly motivated PC physicians, especially those prepared to participate in this rather demanding 12 month duration study and may thus not be representative for the primary care physicians in general.The scope of the study had to be limited because of the routine workload in practice and likely attrition by participants if the number of questions were excessive.This explains the simple classification of morbidity by cause (illness or accident) and by type of occupation (manual and other).These factors have been studied extensively by others [7,9,14].Patients with a high probability of long-term sickness absence were those with circulatory system diseases, musculoskeletal/connective tissue diseases, neoplasms, endocrine/nutritional/metabolic diseases, and mental disorders.Sickness certification is not a prerogative of primary care physicians in the Swiss health system.Hospital based and practice based specialists also provide sickness certificates, often in collaboration with family physicians, which partly explains the low issuing rate reported here.Finally rates are based on consultations rather than on the population at risk.They provide a measure of the workload occasioned to the physicians by the certification task, not an estimate of the proportion of workers who received sickness certificates [13].

Conclusions
Most physicians considered sickness certification to be an appropriate part of their function.However our findings suggest that the role of the certifying physician could be strengthened by a clearer definition of the purpose of certification, especially in relation to returning to work.This may call for appropriate training as part of postgraduate medical education and an improved dialogue between physicians and other parties involved.Changes such as extending the self-certification period would have an obvious impact on firms from an economical perspective and, therefore, would have to be discussed with public health officers, insurers, employers and politicians.