Characteristics of sick-listing cases that physicians consider problematic – analyses of written case reports

Abstract

Objective

The aim was to discern common characteristics in the sick-listing cases that physicians in general practice and occupational health services find problematic.

Design

Descriptive categorization within a narrative theoretical framework.

Setting

Sickness-insurance course for physicians in general practice and occupational health services.

Subjects

A total of 195 case reports written by 195 physicians.

Main outcome measures

Categories of features regarding medical, work, and social situation as well as medical interventions.

Results

Beside age and sex, the following information was often provided: family situation, stressful life events, occupation, problem at work, considerations concerning diagnoses, medical investigations, treatments, and vocational rehabilitation measures. Two-thirds of the patients had been sickness absent for more than a year. The most common type of case reports concerned women, employed in non-qualified nursing occupations, and sick listed due to mental disorders. The most common measures taken by the physicians were referrals to psychotherapy and/or physiotherapy, and prescribing antidepressants (SSRI). Facts about alcohol habits were rarely provided in the cases.

Conclusions

Some of the circumstances, such as prolonged sick-listing, are likely to be more or less inevitable in problematic sick-listing cases. Other circumstances, such as stress-full life events, more closely reflect what the reporting physicians find problematic. The categories identified can be regarded as markers of problematic sick-listing cases in general practice and occupational health service.

A previous systematic review of published studies established that physicians experience sick-listing tasks as problematic [1], especially GPs, for instance regarding conflicts with patients over sickness certification [2–7]. In Sweden, as in many Western countries, the vast majority of sickness absence is due to musculoskeletal or mental diagnoses [1,8,9.] Many physicians find sickness certification more problematic when the decision has to be made practically only on the basis of the patient’s description of his or her symptoms [2,10] and physicians tend to be highly influenced by how the patient gives this description [11,12]. Also, specific dilemmas experienced by physicians in their sick-listing practice have been reported [10,13].

Although there is much knowledge about risk factors for sickness absence regarding for example sex, age, work, and psychosocial aspects [14–16] knowledge is lacking about the factors that contribute to problems in the physician’s actual sick-listing practice [1].

The majority of patients had been sickness absent for more than a year.

• The most common type of cases concerned women, often employed in non-qualified nursing occupations and sick-listed due to mental disorders.

• The most common measures taken by the physician were referrals to a psychotherapist and/or a physiotherapist and prescribing antidepressants.

Physicians’ education is based on biomedicine, whereas practice is about actually meeting patients. It has been proposed that medicine is an art that relies on interpreting the patient’s story, and that “the case report is meant to establish a claim to understand what it is all about” [17]. Case reports may, then, give rich material for research with a narrative approach. This approach is often descriptive, and without aiming to find causalities. Nevertheless, cases can also be studied with the aim of identifying main causal conditions. One way to do this is to try to find main characteristics shared by cases of a certain quality [18]. The way factors combine is something the researcher then can discern using her or his in-depth knowledge of cases [18]. In terms of problematic sick-listing cases, a search for such main characteristics in their corresponding case-reports may give insights into circumstances that tend to make a sick-listing situation problematic for the physician. There are few published studies of case-reports of sick-listing [1] and to our knowledge none using this approach.

The aim was to discern common characteristics of case reports of sick-listing cases that physicians in general practice and occupational health services find problematic.

Material and methods

Analyses of 195 sick-listing case reports by 195 physicians were conducted. The case reports were written as a pre-assignment for participation in a five-day course for physicians to improve their sick-listing skills [13]. The participants were asked to choose a case where they “didn’t feel comfortable with their sick-listing role”, and write it down on about one page. Case reports were collected from nine consecutive courses arranged in different parts of Sweden from March 2004 to June 2005, in all 265 participants. Some 148 (55%) were employed in occupational health service, 117 (45%) were employed as GPs, and 114 (43%) were women. Most of the occupational physicians were also board certified specialists in general practice, some in psychiatry, orthopaedics, or internal medicine. CER and ME taught on the courses. Written information was given regarding the study. Of the 70 participants who did not join the study, 18 rejected the invitation, while 52 (19.6%) from the first five courses never answered the invitation that, in their case, was sent by e-mail after the course.

Two dimensions of content appeared in most of the case reports. One concerned facts, the other the physicians’ experiences and understanding. In this study we examined the first dimension: factual information given by the physician about the patients’ life history, their social and medical situation, and also descriptions of measures taken by the physician. The method used was a stepwise descriptive categorization and quantification. First, all the case reports were scrutinized to identify and categorize facts that were always or often provided, and that in themselves are clearly defined categories (such as sex, age, occupation, and type of sick-leave diagnosis; see Table I). Then, the case reports were analysed in more detail and depth to find categories of facts that also were directly occurring but less frequent (such as country of birth and physical examination). In the last step we made categories out of circumstances that were given in a more narrative way. The circumstances that were included here were based on our pre-understanding of both frequency and relevance [13]. The process of categorization here was a one-step, straightforward qualitative analysis, developed in response to the character of the material. It has traits in common with the “editing style” described by Crabtree and Miller [19]. Our search for main characteristics, with possible causal implications, was also guided by Ragin [18].

Table I.

Categories of facts used to describe patients and categories of measures taken by physicians, numbers and percentages, in 195 case reports and findings for the categories.

	Cases in which the category occurred
Category	No.	%	Findings for the category
Sex	195	100	Women: 134; Men: 61
Age (both sexes)	195	100	Mean: 44.1; Median: 45
Marital status	140	72	Married: 94; Single: 46
Having children at home1	105	54
Disease or resent death in close relative1	36	18
Other stressful life event1	47	24
Alcohol consumption	29	15	Yes: 12; No: 17
Occupation or profession	191	98	See Table II
Work-related problems1	73	37
Sick-leave diagnosis	195	100
Having dog or horse1	6	3
Country of birth (not Sweden)1	18	9
Physical examination1	55	28
Prescribe antidepressants1	62	32
Work-oriented rehabilitation measures1	149	76
Referrals	153	78	See Table III

Note: ¹In none of the cases was information reported about not having any of the categories, such as not having children, dogs, or work-related problems.

Sick-leave periods certified by physicians other than the reporting physicians were included in the calculation of the duration of sick-leave, as was part-time sick-listing. The cases were categorized according to duration of sick-leave into three groups; < 3 months; > 3 months but < 1 year; and > 1 year, respectively. All categories were imported into an Excel sheet for calculations of occurrences.

The apparent conformity of the material induced a search for typical cases. The extracted data made it possible to describe a frequently occurring female and male case respectively, based on occurrence of the categories identified.

The primary reading and categorization was done by ME and CER. The further categorization, and inclusion of only implicitly given facts (such as sick-leave periods and sick-leave diagnoses), was done by ME and continuously discussed with CER and KA until agreement was reached.

The Regional Ethics Committee of Stockholm approved the study.

Results

Twelve characteristics of the patient and his/her situation and four categories of measures taken by the physicians were identified (Table I).

Life situation

Of the patients in the 195 case reports, 134 (69%) were women. Mean age for women was 43.9 and for men 44.4 (median 45 years for both).

A total of 36 (18%) of the case reports included descriptions of a relative who was severely ill or had recently died, in nine cases a child. In 10 of these, a relative with fibromyalgia or other chronic pain was mentioned, in six there was a relative with alcoholism or other type of substance abuse. Among the case reports, 47 (24%) contained descriptions of other stressful life events such as traffic accident, abuse, rape, imprisonment, divorce, or custody dispute. Only 29 (15%) of the reports mentioned the level of alcohol consumption of the patient. In 12 cases it was stated that alcohol was part of the problem, including those reporting high levels of alcohol blood markers. In 17 cases the patient had stated no excessive alcohol consumption. No case reports described any other type of substance abuse by the patient, such as painkillers, sedating drugs, or narcotics.

From earlier experience we expected country of birth or having dogs or horses to turn up quite often. These categories were in fact scarce and therefore not analysed.

Work

In 183 (94%) of the cases it was possible to categorize the patient’s sex and occupation according to SSYK [20] (Table II). The clearly largest group (service workers and shop sales workers) contained almost only women. Of the 49 women in this category, 46 had a social or nursing occupation, which corresponds to 34% of all the women in the material.

Table II.

Categories of occupation among the women and men, number and percentages, in the 195 case reports.

	Women		Men		Total
Occupation (SSYK)	No.	%	No.	%	No.	%
1. Legislators, senior officials, and managers	5	4	5	8	10	5
2. Professionals	20	15	7	11	27	14
3. Technicians and associate professionals	8	5	5	8	13	7
4. Clerks	14	10	4	7	18	10
5. Service workers and shop sales workers	49	37	6	10	55	28
6. Skilled agricultural and fishery workers	1	1	0	0	1	1
7. Craft and related trades workers	4	3	10	16	14	7
8. Plant and machine operators and assemblers	5	4	12	20	17	9
9. Elementary occupations	19	14	9	15	28	14
Unemployed, student, on social security	5	4	3	5	8	4
Occupation not stated	4	3	0	0	4	2
All	134	100	61	100	195	100

Work-related problems

In 73 (37%) of the cases, the physician reported that factors at work contributed to the need for sickness absence. Such factors were conflicts, bullying, harassment, or offensive behaviour towards the patient. For the majority, the reported events had happened long ago. In 17 cases, a high workload was reported; others concerned reorganization, bad working environment, or a work accident.

Symptoms and diagnoses

The texts concerning medical problems, symptoms, and diagnosis were a mixture of professional language and the physician’s own interpretation. Despite this, in most cases it was easy to categorize the symptoms presented and also easy to decide what was the likely main sick-leave diagnosis put on the sickness certificate by the physician.

The main sick-leave diagnoses were classified into three groups: related to musculoskeletal disorders, to mental disorders, or to “other”, respectively. Of the diagnoses 114 (59%) were mental disorders; mainly stress reaction, depression, and crisis. In 44 (23%) cases, the physician reported sleeping disorders. One case report described an anxiety disorder, another described bulimia. Musculoskeletal sick-leave diagnoses comprised 71 (36%) of the cases and for all these patients a mental disorder was also reported. Most common was lumbago and neck pain. Some specific diagnoses such as epicondylitis, arthritis, and Morbus Bechterew were also reported. Six cases reported whiplash-associated disorder as the cause for sick listing.

There were only 10 (5%) “other” diagnoses, e.g. vertigo, multiple sclerosis, ulcerous colitis, and breast cancer. All these cases also included descriptions of musculoskeletal and/or mental disorders.

Duration of sickness absence

The sick-leave duration was in the majority of the cases quite long; in 15 (8%) cases less than three months, in 57 (29%) more than three months but less than one year, and in 123 (63%) the sick-leave had lasted for more than one year. In 73 (37%) cases the sick-leave had been initiated by a physician other than the reporting physician.

Measures taken by the physician

In 62 (32%) of the case reports there were comments on prescription of antidepressants, by far the most common being SSRI. There were many examples of having tested several kinds of SSRI, and almost none of the cases described this as an effective treatment; rather, side effects and bad compliance were reported. Sleeping pills were reported in only two cases.

Some kind of referral was reported in 153 (78%) of the cases (Table III), referral to psychotherapy and physiotherapy being the two most common. In 90 (46%) of the case reports there was more than one referral, in some three or more

Table III.

Type of referrals reported by the physicians in 153 case reports, number and percentages, for female and male patients.

	Women		Men		Total
Type of referral	Number	%	Number	%	Number	%
Psychotherapy	54	28	17	19	71	26
Physiotherapy	37	20	14	16	51	18
MRT/CT	17	9	11	12	28	10
Orthopaedist	13	7	13	15	26	9
Psychiatrist	12	6	9	10	21	8
X-ray	13	7	6	7	19	7
Work capacity assessment	17	10	2	2	19	7
Neurologist	10	5	4	5	14	5
Rheumatologist	4	2	3	3	7	3
Pain investigation	4	2	3	3	7	3
Stress management	4	2	2	2	6	2
Alcohol rehabilitation	2	1	2	2	4	1
Neurophysiologic	2	1	2	2	4	1
All	189	100	88	100	277	100

Work-oriented rehabilitation measures initiated

Amongst the case reports, 149 (76%) included information on measures to promote return to work, mostly meetings with the employer and the social insurance official. In one-quarter of these cases, vocational training was initiated and sometimes adjustment of the work tasks. In several cases, an investigation of the patient’s work capacity was made by a rehabilitation clinic. In 10 cases the social insurance offices decided to withdraw the sick-leave benefits.

Typical female and male cases

A common female patient was sick-listed for more than three months, had a nursing occupation (SSYK 5), and was sick-listed due to mental disorders. The physician prescribed SSRI and/or referred her to a psychotherapist. There were 23 such patients.

A common male patient was sick-listed for more than three months, employed in machinery, building, or non-qualified work (SSYK 7-9), and sick-listed due to a musculoskeletal disorder. The physician initiated investigations including MRT/CT, X-ray, and/or referral to an orthopaedic specialist and/or to a physiotherapist. There were 11 such patients.

Discussion

The vast majority of the patients in the study were long-term sickness absent for more than one year. The most common type of case reports concerned female patients employed in non-qualified nursing occupations, predominantly sick-listed due to mental and/or musculoskeletal disorders. The most common measures reported by the physician were referrals to psychotherapy, physiotherapy, and prescription of antidepressants. Facts regarding alcohol habits were rarely provided.

Strengths of the study were the large number of case reports, and that in total they proved to be rich in information and also that researchers with different clinical and theoretical background took part in the analyses. Weaknesses of the study were the heterogeneity of the material and that some of the case reports provided nothing but the clearly defined categories (such as sex, age, and sick-leave diagnosis).

It is likely that most of the participating physicians were motivated to improve their sick-listing skills. Physicians who did not wish to take the course may have fewer problems or be unaware of problems that they actually have, and as regards these physicians, our results may be of limited relevance. The vast majority of the physicians were board-awarded specialists in general practice and many were also trained and employed in occupational health service, which may have made them more observant about workplace factors than other types of physicians.

In our material, a larger proportion of the cases involved mental disorders compared with long-term sick-leave diagnoses in the whole Swedish population [1,16], indicating that patients with such disorders tend to be experienced as more problematic than others. Furthermore, many of the patients were reported to have both mental and musculoskeletal disorders. A high rate of comorbidity is in line with another recent study of problematic long-term sick-leave cases [21]. Also, physicians seldom tend to view the more “obvious” diseases as problematic when it comes to sick-listing [10,12].

Facts regarding alcohol habits were rarely provided, although several studies have shown an association between sick-leave and alcohol consumption [1,22,23]. Our material is a selection of problematic sick-listing cases in which one could have expected a certain amount of attention to the possible role of alcohol. Other studies have found that many physicians find it difficult to discuss matters of alcohol consumption with their patients [24,25].

Long duration of sick leave was a prominent factor in our material, so prominent that in itself it may be suspected to contribute to making a sick-listing case be considered problematic. To our knowledge there is no previous study that clearly shows this connection. In addition, in a substantial proportion of the long cases, a physician other than the reporting physician had initiated the sickness certification. Having problems managing such cases is in line with findings from a questionnaire study by Löfgren et al. [3].

Our case report material can be regarded as a large sample of narratives, or “Doctors’ stories” [17], in sick-listing practice. Based on our understanding of the context for which the cases were written, we have identified the main characteristics of sick-listing cases physicians find problematic. Here we were inspired by Ragins’s [18] ideas about how main causal conditions, shared by relevant cases, can be identified. Some of the categories provided through the analysis, such as prolonged sick-listing, seem to be more or less inevitable in most problematic sick-listing cases. Other categories reflect what, in varying numbers, the reporting physicians find problematic in their respective cases. Here, the selection is probably more conditional and more dependent on individual characteristics of the physicians. Such characteristics were, however, beyond the scope of this study. Furthermore, circumstances other than prolonged sick-listing may cause problems for most physicians, once they occur. For example, within the diagnostic categories “mental” and “musculoskeletal”, the patient’s own description is often the only ground for assessing the patient’s working capability, as also discussed by Englund [12]. However, stressful life-events were frequently reported in the narratives. They might have been included to illustrate the difficulties. As discussed by Hunter [26] the reporting physicians may have tried to make their written case reports understandable and acceptable, both in the course context and also to themselves.

Many patients in this material were offered SSRI treatment. Furthermore, other types of actions frequently taken by the physicians (such as referrals to psychotherapy and physiotherapy) seem to reflect unsuccessful efforts to promote return to work. Those are actions that might be more relevant when patients are sick-listed for diseases with more clear-cut disabilities.

There was a striking resemblance between the types of sick-listing cases the physicians had sent in. From this, we conclude that the categories identified could also be regarded as markers of problematic sick-listing cases in general practice and occupational health service. Other Swedish researchers have discussed the idea of similar markers or risk factors. Linton et al. [27] found that it was possible to predict outcome of sickness absence for many patients with acute or sub-acute low-back pain by employing five variables (fear-avoidance work beliefs, perceived improvement, problems with work function, stress, and previous sick leave). Kaiser and colleagues [28] found psychosocial markers to be important for the outcome of vocational rehabilitation for patients sick-listed with musculoskeletal disorders (alcohol-related problems dominated for men and private-life-related problems for women). These studies are in line with our findings that there seem to be certain cases that physicians find problematic and that the strategies used often do not seem to solve the problems. From a narrative outlook we learn that it also takes a doctor to make a problematic case.