Towards Understanding of Determinants of Physicians’ Sick-listing Practice and their Interrelations: A Population-based Epidemiological Study

Full text

(1)Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 1283. Towards Understanding of Determinants of Physicians’ Sick-listing Practice and their Interrelations A Population-based Epidemiological Study BY. BRITT ARRELÖV. ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2003.

(2)

(3)

(4)

(5)

(6)

(7)

(8) ! "!#"$ % & % % '& & ( % )

(9) *+ ,&

(10)

(11) -

(12)

(13) -&+ . /+ !+ , -

(14)

(15)

(16) %

(17)

(18) % '&

(19) 0 1

(20) ' .

(21) & 2

(22)

(23) + '

(24) 1 3 +

(25)

(26) +

(27)

(28)

(29)

(30)

(31)

(32) "4!+ $5 + + 2/6 7"1$$81$5"41$ '&

(33) 1

(34)

(35) & 1 &

(36)

(37) + 3 &

(38)

(39) -

(40) &

(41) .

(42) &

(43)

(44) 1

(45) + 2

(46)

(47) &

(48) 0

(49)

(50) % & %

(51) %

(52) 1

(53) + ) & %

(54) &

(55) 0

(56) & & - %

(57)

(58) + ,& % & -

(59)

(60) % 1

(61) -

(62) &

(63)

(64) &

(65) %

(66) % &

(67) &

(68) 1

(69) &

(70) %

(71) %

(72)

(73)

(74)

(75)

(76) % & &

(77) &

(78) 1

(79) &

(80) %

(81) % %

(82) &

(83)

(84) &

(85) %

(86) % &

(87)

(88) &

(89) 1

(90) + ,& -

(91) 1

(92) % $5$! 9 % %

(93)

(94) 4

(95)

(96) %%

(97) & -&

(98)

(99) %

(100) &

(101) "77$

(102) -

(103) &

(104) "77+ '

(105)

(106)

(107)

(108) &

(109)

(110) % &

(111) &

(112)

(113)

(114) :

(115)

(116) -

(117) %

(118)

(119)

(120)

(121)

(122)

(123) %

(124) % 1

(125) + '&

(126) & &.

(127) %

(128) - - &

(129) %

(130) % 1

(131) + '&

(132) -

(133)

(134)

(135)

(136) %

(137) & % 1

(138) &

(139) & & &

(140) + ;

(141) %

(142)

(143) .

(144)

(145) %

(146)

(147) 1

(148) - &

(149)

(150) -

(151) %

(152)

(153) & %

(154)

(155)

(156) + &

(157) %

(158) & - - & %%

(159) 1

(160) + 2

(161)

(162)

(163)

(164) % % & &

(165)

(166)

(167) &

(168) .

(169) &

(170)

(171) % - %

(172)

(173) %

(174) &

(175) % 1

(176) + 2. & & &

(177) %

(178)

(179) % - & - &

(180) -

(181) ++ &

(182) 1 &

(183)

(184)

(185) & && &

(186) &

(187) +

(188) 1

(189)

(190)

(191)

(192)

(193) 1

(194) &

(195) !" #$ % &

(196) "

(197) #'"

(198)

(199) " ()*+,-.

(200) " < / . ! 26 41585 2/6 7"1$$81$5"41$

(201) #

(202)

(203) ### 1!$$5 (& #==

(204) ++= >

(205) ?

(206) #

(207)

(208) ### 1!$$5* Printed in Sweden by Universitetstryckeriet, Uppsala 2003. !..

(209)

(210) .

(211)

(212) . !" # #$ % & '"( ) *+,,-. & / 01

(213)

(214) 2

(215)

(216) 1

(217)

(218) -3*4.5 6761673. . !" # #$ % & '"( ) *+,,6. 0

(219) 8

(220)

(221)

(222) 1

(223) 965 +:71+93. . !" # #$ '"( )

(224) 8 01

(225)

(226)

(227) 1 * .. ;. !" # #$ '"( ) "

(228) 8 01

(229)

(230)

(231) 1 * ..

(232)

(233)

(234) .

(235) Contents Prologue.......................................................................................................... 1 Introduction ................................................................................................... 3 Physicians as sick-listing agents ............................................................... 3 Costs due to sick-listing ............................................................................ 4 Factors affecting sick-listing decisions ......................................................4 Theoretical considerations ........................................................................ 9 Aims of the study ......................................................................................... 14 Study population and methods ..................................................................... 15 Setting ..................................................................................................... 15 Data collection ........................................................................................ 15 Variables .................................................................................................. 15 Financial co-operation ............................................................................. 17 Sub-studies .............................................................................................. 18 Statistical considerations ......................................................................... 18 Results .......................................................................................................... 19 Characteristics of sick-listing certificates in the four studies ................. 19 Patient level ............................................................................................ 21 Physician level ........................................................................................ 21 Health care organisation level ................................................................ 24 Local society level .................................................................................. 28 National level .......................................................................................... 33 General discussion ....................................................................................... 37 Validity .................................................................................................... 37 Limitations .............................................................................................. 37 Strategy ................................................................................................... 38 Certificates and sick-listed persons ........................................................ 38 Physician-related factors ......................................................................... 39 Variation between physician categories .................................................. 40 Influence of reimbursement and other financial incentives .................... 40 Local sickness absence practice .............................................................. 41 Influence of legislative change ............................................................... 42 Implications for the future ...................................................................... 43 Future research ........................................................................................ 46 Conclusions .................................................................................................. 47 Acknowledgements ...................................................................................... 48 References .................................................................................................... 49.

(236) Prologue My interest in this subject field is two-fold. First, as a former general practitioner and medical advisor in a social insurance office I am interested to know more about physicians’ sick-listing practices, and secondly as an administrator in a purchasing organisation, trying to govern health care, I want to know more about decisionmaking processes and influencing factors. I first came into contact with sick-listing decisions during my clinical training as an undergraduate trainee. Generally, I issued certificates on behalf of a supervisor, but sometimes on my own. The certified sick-listing period often followed a biomedical schedule, for instance appendicitis- three weeks, fracture of the radiusfour weeks etcetera. It really seemed to be clear-cut decisions. But when I reached the level of assistant physician the issue of disparate interests and loyalty dilemmas soon became apparent and I realised that I was no expert in the field. Later as a general practitioner I had to deal with problems related to sick-listing almost every day. Within a short period of time I was in the middle of conflicting interests between clerks from the social insurance office, patients, employers, other physicians, media and myself. It was obvious that my medical training had not given me the skills I needed to handle this dilemma properly. Later as a manager of primary health care units and project leader of a financial co-operation project run by social insurance and primary health care I could no longer manage the task with my insufficient knowledge. I had to learn more about the sick-listing process and the social insurance system. So, I became a medical advisor at a local social insurance office. In that position I learned how to deal with the sick-listing issue from another point of view than the practicing physician’s. I became a representative of the social insurance organisation and I found out how the clerks assessed the sick-listing certificates. They told me that individual physicians or groups of physicians were seen both as good or bad sick-listing agents. With fresh eyes I saw my fellow-physicians sick-listing certificates, an interesting experience, especially the variation of conduct. Among other things I noticed that the introduction of a “Family physician” system in general practice influenced the GPs sick-listing practice. That made me even more interested in the determinants of the sick-listing process and clinical decision-making per se. As a practicing physician I had a patient-oriented view regarding sick-listing. In the work with the financial co-operation project and as a medical advisor at the local social insurance office I had to adopt an administrative-oriented view. During the same period, as a medical expert in an official commission, I met the opinion that physicians overall did not take enough responsibility as gate-keepers to the social insurance system. The various views and opinions about sick-listing that I was confronted with during this period made me really interested in studying this issue.. 1.

(237) My experience was then and still is that most physicians act as responsible as they can while sick-listing. They try to do the best to satisfy all interested parts, just as they do in all clinical decision-making! Sick-listing decisions differ to some extent from clinical decision-making in other fields, but the prerequisites for physicians’ clinical decision-making while acting as experts are generally the same. This study has given me the opportunity to test the hypotheses that physicians’ act as experts when sick-listing, and that their practice are influenced by the prerequisites and the context in which they perform this task. The decision-making will thereby to a large extent reflect the views and values of the society. Through the years my focus of interest has been somewhat modified, due to new knowledge, change of work positions during the years and due to the fact that there has been a shift in the public sickness absence debate. Not only have I gained knowledge about medical decision making and sick-listing; I have also learnt to deal with problems inherent in moving from an idea to realization of a project and the struggle to show endurance and to complete the task.. 2.

(238) Introduction Physicians’ as sick-listing agents Physicians caring for patients are repeatedly faced with the fact that they have the problem of decision-making based on uncertainty [1] and to act as “double agents”, weighing competing allegiances to patients’ medical needs against the monetary costs to society [2]. Decisions about sick-listing belong to this kind of decisions. Certificates issued by physicians play a major role in the distribution process in benefit programs, when illness or disability is used as eligibility criterion [3]. It is a generally accepted fact that physicians are able to determine when patients should abstain from working due to illness or injury and when he or she may safely return to work [4]. The accuracy of a medical certificate depends on the ability of the physician to make such assessments [5]. The certificate is accepted as a legitimate means of control and gives physicians a central role as gate-keepers to the social insurance system. The necessity of gate-keeping is usually underscored with three types of arguments: the need to ensure that patients receive appropriate care, the need for budget control, and the need of justice in the distribution of benefits [6]. Uncertainty during sick-listing is due to the fact that many patients claim illnesses with no or few objective findings, the degree of inability to work may be difficult to assess, the physician may not be aware of the patients job demands, or have insufficient knowledge about the social insurance legislation, or there is defective knowledge about the effects of sick-listing and of the prognosis of the illness [4, 5 ,7-12]. As gate-keepers the physicians decides who will enter the social insurance system and who will not [6]. As street-level bureaucrats they use discretion in determining access to the benefits, even though their discretion is formally circumscribed by rules [3, 13, 14]. Patients and the public usually believe that physicians hold the key to their benefit and that the certificate is a pay-check and not merely a medical report. The primary concern of physicians is the patient’s return of functional ability in contrast to the social insurance office goal that is simply back to work [7]. There is a wide variation of physicians’ practices influenced by the interaction of self-interest, concern for individual patients and regard for well-being of society at large [15]. The only way to prevent the different obligations of physicians from becoming conflicting loyalties would be to provide them with the necessary resources and training to perform their “balance act” [6]. Variation of practice has been demonstrated particularly in fields with a large part of professional uncertainty [16]. Sick-listing activities affect the workload of the physician’s practice and his 3.

(239) or her working environment [17-19]. Negative feelings in relation to the sick-listing role have been reported [7, 9, 18]. A possible explanation might be the lack of consensus regarding sick-listing between medical professional groups, the general population and insurance clerks [20].. Costs due to sick-listing Physicians’ practice regarding sick-listings has been one of the questions in focus of the debate regarding the escalating sick-listing costs. This item has been on the agenda for a long time. Physicians have been accused for acting irresponsible and their gate-keeping ability has been questioned for more than 25 years [4]. Sickness absence costs generates high expenditures for the community and the proportion of the labour force being absent from work has increased in Western Europe since the 1960s [21]. In general practice certification for sick-leave might be the single task causing the highest expenditure in society of all medical actions taken [22-24]. The cost due to variation of physician’s sick-listing practice might vary considerably [25, 26]. The Nordic sickness benefit schemes have many similarities, but differ considerably in degree of compensation, and the length of self certification [27]. However, basic needs for social security are met by public insurance schemes in all Nordic countries. They all use illness or disability as eligibility criterion and certificates issued by physicians play a major role in the distribution process. In spite of the organisational differences all Nordic countries are concerned about the costs and the physicians' role as certifiers. The number of sick days paid by the National Social Insurance in Sweden has been fluctuating over the years [28]. The cost for paid sick-days was in 1996 almost 15 billion Swedish crowns (approximately 1.5 billion euro) for the 9 million residents. From the end of 1980’s until the middle of the 1990’s the number of compensated sick days decreased and during 19941995 there was a reduction of early retirement pensions [29]. Despite all the reforms during the first half of the 1990’s the end of the decade showed a dramatic increase of the amount of paid sick-days. This increase has continued during the first years of the 2000’s.. Factors affecting sick-listing decisions Studies of clinical decision-making have shown that many different and interacting factors influence physicians practice patterns. Only a few studies have focused on sick-listing decisions, but there is no reason to believe that sick-listing decisions differ from other decision-making processes in health care studied more thoroughly. Consequently, the decision of whether or not to put the patient on the sick-list and how to do that is influenced by a complex system of factors, like for example physicians’ prescribing of drugs [30]. In addition to the patient’s illness or disability and the rehabilitation prognosis the following factors have been claimed to influence sick-listing decisions: 4.

(240) 1. Legislation, especially regarding health care and social security [3, 31-42] 2. Health care: structure, organisation and reimbursement system [1, 31, 37, 4371] 3. Labour-market and work-place, including organisation of companies, occupational health care, job demands and absence culture [5, 10, 21, 31, 38, 40, 72-85] 4. Society: socio-economic factors, opinions and attitudes [3, 21, 31, 38, 41, 45, 46, 58, 73, 74, 81, 85-89] 5. Patients: socio-economic factors such as incentives, knowledge, attitudes and beliefs [3-5, 10, 15, 19, 21, 25, 31, 38, 40-42, 45, 48, 65, 71, 76-79, 81, 84, 85, 89-108] 6. Physicians: personal characteristics, incentives, knowledge, attitudes and beliefs [3, 4, 15, 20, 23, 25, 26, 37, 43-50, 54, 55, 64, 65, 67, 68, 70, 71, 78, 86, 104, 105, 109-120]. The focus of this study is not absenteeism but sick-listing practice and we therefore focus our interest on factors related to items 1, 2, 4 and 6. Most previous studies related to sick-listing have studied absenteeism and have therefore focused on factors related to item 3 and 5. Legislation By giving the framework, legislation forms a base for all actions in health care. As legislators Government and Parliament form the conditions for all organisational and structural actors affecting sick-listing decisions, such as health care, social insurance and labour-market. The compulsory sickness cash-benefit insurance was implemented in 1955. The present benefit scheme for absence from work due to sickness and disability in Sweden contains remuneration for sickness days and pension. The National Insurance Act in Sweden covers all people between 16-65 years of age [121]. For employed persons the first two weeks of sickness absence are compensated by the employer and self-employed persons receive sickness benefits the first four weeks if they have supplementary insurance. Thereafter and for unemployed persons during the whole period the sickness benefits are paid by the National Social Insurance. Degrees of incapacity to work are given as fulltime, three quarters of fulltime, halftime or a quarter. Disability pensions are used for persons that do not regain capacity to work fulltime because of their illness. A sickness certificate issued by a physician is needed, in order to obtain sickness benefits, except for the first week when sickness absence may be declared by self-certification. There are three types of standard forms of certificates issued by physicians. The first is used for the first 28 days and the second for the certification periods thereafter. These are used together with a notification of illness from the patient. The third (statement of health) is used as medical document in early retirement cases. All Swedish 5.

(241) physicians are entitled to issue certificates for sickness absence and disability pensions. The sick pay scheme may not in itself constitute a cause of absence, but it may provide facilitating circumstances [38, 40-42]. During the 1990’s there were reforms of the National Social Insurance system almost every year, in order to get a better cost control. In October, 1995 a new legislation was instituted aiming at strengthening the medical criteria for sick-listing, requiring more information on the certificates and the local Social Insurance Offices’ need of information before allowances were paid for was specified. Moreover, all certificates in a sick-leave episode with duration of more than 28 days were to be examined by a consultant physician. Two new standard certificate forms were introduced, one for the first 28 days in a sickness-episode and one for periods beyond that duration. The knowledge of how the performance of physicians as certifiers for sickness absence is affected by legislation is scarce. There are only a few studies of how physicians handle a legislative change. In a Swedish study of the legislative change in October 1995 only small effects were found [32]. The physicians filled in the forms more completely and the proportion of sick-listings made by GPs increased. In a Norwegian study of the narrowing of the medical criteria for disability pension eligibility a majority of the studied GPs thought that they had to adjust the use of diagnoses in order to fit the new requirements for pension benefits [34]. In another study of the same legal change physicians were found to use various coping strategies related to the change. The authors concluded that many certifying physicians seemed to have been under considerable pressure because of the conflict of interest between patients and authorities [33]. Health care The vast majority of health care in Sweden is financed by taxes in combination with patient fees [122]. The county councils are responsible for the structure and allocation of resources for primary health care, geriatric care, psychiatric care, hospital care, most private health centres and some private specialists. The county councils are free to set their level of taxes, patient fees and reimbursement schemes. With few exceptions hospitals, psychiatric care centres, geriatric care centres and most general practices are governed by the county councils, but the number of privately owned surgeries for general practice is increasing. Private health centres and private practitioners in communities are quite rare except in big cities. It is not mandatory for companies to be associated with an occupational health service, but many, especially big companies, have their own health service and others buy that service from contractors. Hospitals are reimbursed with overall budgets in combination with fee for services, and private specialists merely by fee for service. General practice is reimbursed with overall budgets or capitation fees, in combination with patient charges. Private specialists may be contractors to the county councils with 6.

(242) reimbursement through fee-for-service or may rely on patient fees only. There is no general system for reimbursement of occupational health care. The individual physician is commonly paid by salary. Only exceptions are private specialists and some occupational health specialists. In Sweden, there is according to law a freedom of choice of physician. Patients’ choice might depend on his or her age, sex, the kind of disease they suffer from or factors related to other preferences. In addition, physicians have preferences in different areas of medical practice and on their choice of patients. This may influence the kind of patients that consult different physicians and thereby the clinical decision-making [23, 68]. The use of specialist services and practice characteristics varies according to the existence of whether primary care has a gatekeeping function or not [59]. In Sweden this is not thoroughly accomplished, but some counties have tried to introduce a “semi” gate-keeping function, with demand for referrals from primary care for patients who want to see a specialist. Although the physician is the primary decision-maker in the allocation of health care resources, the limited resources and rules for conduct set the boundaries for physicians’ decisions regarding individual patients [1]. As opposed to the costs for health care, the costs for sickness absence do in no way influence the physician himself, his employer or the health care organisation at large since they are paid over the national state budget. Labour-market and work-place The sickness benefits reduce the economic burden when a person is unable to go to work due to illness. How sick a person must be to stay at home depends partly on his or her work tasks. The structure of the local labour-market is important for persons who are disabled or need to change job due to sickness. In studies of absence from work the workplace is one of the most studied objects [31]. When the influence of other factors were considered job satisfaction was shown to be the most important factor in relation to absenteeism [38]. However, other studies have found that absence culture and psychological contracts are more important than job satisfaction and personal characteristics [74]. Only few studies have focused on physician’s practice patterns in relation to the labour-market or the work-places [10, 72, 73, 82]. The physician provides a person with a certificate saying “on sick-leave”, which allow him or her to abstain work [80]. In this way the sick-role status of a person is confirmed. The sick-listing physician is often not familiar with the demands and organisation of the patient’s job, the structure or attitudes in the workplace or the rehabilitation possibilities performed by the company. Occupational health physicians have a better knowledge of this type than other physicians. Other physicians have to rely on the information given by the patient.. 7.

(243) Society Variation in absence from work is influenced by social consensus both inside and outside the workplace [74]. Sickness absence may be understood as a practice in which subjects take into account both their health and the rules of the community in which they are living [88]. Even though there has been an improvement in public health since the mid 1950s, the satisfaction with personal health has declined during the same time [89]. There is no one-to-one correspondence between objective health status and subjective health perception. Regional variation of absence has been found to be associated with socio-economic factors, particularly social class [88]. Local sickness absence practices have been defined as an expression of class cultures, incorporating largely the same elements of way of life, taste and style as the corresponding practices in other fields of social life. Physicians may adjust their certification practices according to this local sickness absence practice. Their possibilities to take these factors into consideration depend on their knowledge of the local community. Patient The influence of the patient on physicians sick-listing practice has been pointed out as most important [19, 93]. Sickness absence is a complex phenomenon combining illness and coping behaviours [85]. A person might be absent from work on a particular day depending on the relative strength of the tendency to appear at work and the barriers to attendance [107]. The causal connections are multi-factorial comprising both medical and non-medical components and much more complicated than purely the existence of illness [92, 99]. The impact of various absence inducing events varies from person to person and depends on their influence on the motivation to attend at work [81]. There is not enough support in the physician’s role to get a person back to work if the patient doesn’t have the desire or motivation to do so [91]. Illness or disability might be seen as a social role [3]. This means that being sick is a status that entails certain behaviour by the sick person and certain behaviour of others towards the sick person. There are rules that specify how sick a person must be to obtain desired benefits and these rules vary in different contexts. Physicians are forced to collude with the patient’s definition of ill-health, which may not be in the best interest of the patient or the society [100]. It is relevant for the certifying physician to have an understanding of the sick role and the factors that influence its development in patients in working age [84]. Certification recognizes neither the scope of the problems encountered nor the degree of incapacity [99]. The role of the physician is to mediate between the claims of individuals and the formal rules of organisations and society. The importance of the doctor-patient relationship and the attitudes that practitioners bring into that relationship is very important [45]. In relation to the patient’s interest the physician takes into consideration the probability of disease, potential clinical benefit and risk for the patient and cost to the patient [15]. 8.

(244) Physician Physicians’ sex, age, specialisation and experience have been shown to have an impact on practice style [23, 26, 114, 115]. Variations in sick-listing practice and assessment of work ability among male and female physicians’ have been shown [25, 78]. Geographic variation among physicians has been found to depend to a large extent on the result of the physicians’ evaluation of the patients’ diagnosis or in their belief in the value of the procedures and practices for meeting patients needs [109]. Physicians acts as agents for their patient and thereby try to give the most appropriate care. The physician’s own benefit is affected not just by his own income but by how he thinks he has carried out this role as an agent for the patient [46, 71]. Financial incentives directed towards physicians in order to reduce costs have been used in various countries. Especially in the managed care system in the United States and in the National Health Service in Great Britain capitation, withholdings, bonuses, profit sharing and financial penalties are in ordinary use [95]. Financial incentives have in different studies been shown to influence clinical practice [15, 43, 46, 5457, 68, 69, 71, 113, 116]. To counteract physicians’ resistance to change there is a trend towards use of strong financial incentives [111]. Physicians themselves do not consider financial incentives to be important to the same extent that managers do [58]. They rather consider the influence of the medical professional system as more important. No incentive package has yet been demonstrated to influence all physicians at the same time or a single physician all the time and the organisation of health care, including the imbedded incentives seems to have less impact on how doctors work than the characteristics of the population and the local economy [70].. Theoretical considerations Throughout the work with the study, especially during the analysis of data and the efforts in understanding the results, support from the theoretical literature about medical and clinical decision-making has been used. The way in which physicians make medical decisions can be viewed from either of two perspectives: a prescriptive theory of how medicine ought to be practiced or a descriptive view of how medicine actually is practiced [65]. The descriptive view is most applicable for the purpose of this study, because it highlights the influence of a range of social factors and puts emphasis on identifying and measuring sources of variability in decision-making. Decision-making in a clinical context The term clinical decision-making is used to describe a systematic way to handle information in order to decide on a best course of action [123]. Medical decisionmaking, concerned with individual patients, can be a function of both who the 9.

(245) patient is as much as what the patient has [65]. Characteristics for decision-making in natural contexts are that the decisions are made under significant time pressure and thereby closely connected to action [124]. It is furthermore characterized by badly structured problems and a dynamic and uncertain environment. The various actors’ goals may be shifting, badly defined or competing, but the outcomes are of high significance for all of them. Value judgements underlie virtually all clinical decisions [1]. Many individuals are actively involved and the physician must balance personal choice and the patient’s preferences with organizational norms and goals. Family physicians, who know patients for many years, can be in a better position to estimate their patients’ preferences. The physician may consciously or unconsciously bias the patient’s preferences, when these are being elicited. Patients are likely to sense the physician’s preferences and may be influenced by them. The utility of the physician, who makes vicarious decisions, may not correspond to the patient’s utility [125]. Health resources are consumed in order to produce health benefits [1]. Society operating through its secondary decision-makers set the boundaries for the physician’s decisions regarding individual patients. The opportunity cost of a resource is the value of benefits forgone by failing to apply the resource to the most productive alternative use. Resources used for one patient or for a group of patients are unavailable for alternative use. Influences on physicians’ decisions and actions The physician’s threshold for action when making medical decisions is based on a complex combination of factors, such as the physician’s perceived probability of disease, potential clinical benefit, risk and cost to the patient, the financial benefit or cost for the physician and the practice organisation, the physicians satisfaction with his practice, potential impact on the physician’s self-image or status in the community, in combination with potential benefits, risks and costs for society at large [15]. According to Eisenberg this is called a “multi-attribute utility problem”, which means that the physician has to simultaneously satisfy a number of goals, some of which are conflicting. According to Wennberg the role as a rational agent for patients and society, which the society has asked the physicians to assume, is one they cannot be expected to perform [126]. It is important to take into account the social context in which the decisionmaking activities take place and within which the decision-maker conceptualises the consequences of his or her actions [127]. The social world in its manifestations as culture and language (macro level) and family, school, work environment, peer group etcetera (micro level), determines the terrain of definitions and the range of understanding an individual can use and perform in his or her decision-making activities. It also provides the person with the arena in which his or her choices and actions will be tried out, justified, interpreted by others and rewarded or sanctioned. People who occupy a certain role or position respond to the expectations which others have of that role and are influenced by the persons he or she associates with 10.

(246) occupying that role [128]. The patients physical appearance and the organisational setting in which medical care is delivered, may have as much influence on medical decisions as the actual signs and symptoms of disease [65]. Physicians are human actors and therefore there is a need to consider social behaviour in any comprehensive theory of medical decision-making. Uncertainty and physicians’ decision-making processes The physician caring for patients is more or less constantly faced with decisions, some of a routine character and some complicated [1]. In some clinical situations the decisions seem fairly straightforward and any well-trained clinician might reach the same conclusions. Other situations may be more ambiguous, and the decisions may be less clear-cut. Medical decisions almost always bear the stamp of uncertainty. They may be stratified by the degrees of uncertainty on which they are based. The causes of uncertainty in clinical decision-making include errors in clinical data e.g. the patient may state one complaint but the physician interprets it as another one, test-results or observations may be inaccurate, or the clinical data may be wrong or non-representative or ambiguous. Observations may be interpreted differently by different observers and observers may differ in ability to detect signs and in their propensity to record them. The relations between clinical signs, symptoms and disease are not the same in every patient. Only a few diagnoses have so called pathognomonic signs, indicating the presence of a particular condition, and that condition only. The effect of any treatment is uncertain in any given patient and the course of any disease, is in itself uncertain in any particular case. According to Weinstein may good decision-making in spite of uncertainty be seen as the art of medicine [1]. There is a need to make value judgements about which risks are worth taking. Value judgements and trade-offs concerning the possible outcomes of treatment are continually made. Clinical strategies are a sequence of decisions that are made over time, with each decision based on the information available at the time the decision is made. The use of probabilities for an individual patient is central to systematic clinical decision-making. Patient characteristics along with the disease prevalence may determine the probability of disease at any point in time. Such assessments may be derived from medical literature, clinical experience, personal opinion or combinations of these. Subjective probabilities are obtained by combining one’s strength of belief that an event will occur with reference events of known probability. Practice and experience can probably improve the ability to correctly assess probability. Variation of physicians’ practice patterns According to Andersen and Mooney could practice variation primarily be seen as a symptom and the variation might be explained by the presence of uncertainty in decisions about medical matters [87]. They consider that there are various ways to make treatment decisions and that many factors might influence clinical decision11.

(247) making processes. Practice variation may reflect genuine differences in the “needs” of different populations, resulting from differences in environment, life-style and genetic endowment, i.e. all the various factors that can be shown or hypothesised to affect human health. It may also reflect cultural differences in expectations or beliefs about health and health care. The main element that regulates professional practice is the professional system [58]. It is a widely accepted, socially legitimated and legally recognised set of shared traditions, norms and values that define good medical practice. Different patterns of medical practice are determined by the social, cultural and economic settings where the service is carried out together with the organisation of the health care system itself. Rationality and physicians’ decision-making processes An action is judged to be rational if it is in line with the values and beliefs of the individual concerned [129]. Rationality is also a question of the preferences and beliefs the person holds. A decision might be consistent and thus formally rational, but the judgments that provide the input for the decision might be very poor and thus the choice might be substantially not rational. The specific choice depends on characteristics of the situation, i.e. the physician’s familiarity with the problem, time pressure, knowledge and ability to handle the subject. The decision-maker might escape decisional conflicts by shifting responsibility to someone else, or construct wishful rationalizations and remain selectively inattentive to corrective information. According to Jungermann the rationality of the selected decisionmaking strategy might be valued by way of a cost/benefit compromise between the decision-maker’s desire to make the best decision and his or her negative feelings about investing time and effort in the decision-making process. With finite time and resources available, it might not be rational to spend infinite effort on the exploration of all potential consequences of all options. The decision costs are weighted against the potential benefits resulting from the application of a decision strategy. Expertise and physicians’ decision-making processes In the certifier position the physician is supposed to act as an expert. An expert is someone who is capable of doing the right thing at the right time and someone who can make an appropriate response to a situation that contains a degree of unpredictability [129]. In naturalistic dynamic settings experts frequently generate and evaluate single options [124]. Because most of the problems are not structured, decision-makers choose an option that is good enough, though not necessary the best. To be able to act as an expert certain skills are necessary [130]. Practice has been shown to be the major independent variable in the acquisition of skill, but mere experience may not be sufficient. People do not always improve judgment by experience, because experience often gives very little information to learn from 12.

(248) [131]. Studies have shown that expert judgments regarding predictive accuracy in most clinical and medical domains are not more accurate than those of lightly trained novices [132]. Experts use less information but more knowledge than novices in auditing. Expert performance depends on the ability to match properties of different practices to the performance criteria relevant in the actual work circumstance. Performance limitations of non-experts in medical decision-making are not knowing what information is relevant and why, the interrelations among variables and difficulties in combining and integrating information [133]. The manner in which physicians use uncertain information in order to arrive at a judgment depends, at least partly, on how specifically the information refers to the event in question. One of the main lessons of decision research is that feedback is crucial for learning [132]. Inaccurate treatment practices may persist because experts who get slow, infrequent, or unclear feedback will not learn that their practices are wrong. Sick-listing decisions are clinical decisions based on both medical decisions regarding individual patients as well as social considerations. Theories about medical and clinical decision-making processes are applicable, taking the social context into account. This study was planned out of a multi-factorial interaction model, figure 1. It was accomplished in an every day setting. The fundamental components are the following; real decisions, high level of professional uncertainty, discretion in judgment, multi-factorial and multi-level interactions and social context.. Figure 1. Analysis model of physicians’ sick-listing practice, its determinants and influencing bodies. 13.

(249) Aims of the study The aims of this study were to gain knowledge about physicians’ acting as gatekeepers to the social insurance system and their possibilities to full-fill their role as expert decision-makers in the sick-listing process. The specific aims were to analyse: – variation of sick-listing practice between physician categories and the influence of physician characteristics on sick-listing practice – the influence of structure, organisation and remuneration of health care on physician sick-listing practice – the influence of local structural factors in the community – the influence of a legislative change on physician sick-listing practice. 14.

(250) Study populations and method Setting The study was conducted as a cross sectional epidemiological study of sick-listing certificates issued by physicians and received by the local social insurance office during February, April, June and October 1995 and April and October 1996 in 28 municipalities located in 8 counties (out of 24) in Sweden. The area covered included urban and rural districts in all the counties. The participating counties and municipalities are presented in table 1. Five offices participated only during some of the study months. The collection of certificates required a large amount of work by the local social insurance offices. A prerequisite for their participation was therefore first of all interest to participate and secondly that at least one city or municipality with a hospital and at least one with no hospital in each county must be included. None of the three largest cities in the country participated. In the various municipalities 1522% of the physicians were general practitioners, 64-81% were hospital based physicians, 2-7% were occupational health physicians and 2-7% were other physicians (mostly private practitioners).. Data collection A total of 57563 sickness certificates were received by the offices during the study period and registered. A photocopy of the certificate with date of arrival added was sent to the study coordinator. The certificates were checked, corrected if necessary, missing data were completed if possible, and then the data were coded and fed into a computer. After these procedures were accomplished less than 1% of the certificate data were missing. Validation of the data-registration was done by a second registration of a two-week subset and by cross-tabulation of variables.. Variables The variables registered were insurance office number, county code, year and month when the certificate was issued, the patient’s age and sex, and diagnoses. The diagnostic classification used is given in table 2. Furthermore, category of certifying physician, crude number of days of the current certified sickness period, degree of sickness absence (100%, 75%, 50% or 25%), number of net days (length of the 15.

(251) Table 1. Participating counties and municipalities. Sick days=number of paid sickness absent days/resident 16-65 years old/year, Disability days=Number of paid sick-days and disability pension days/insured 16-65 years old/year. –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Number Sick Disability Hospital Months Number of of days days in muni- in study certificates residents –––––––––– cipality ––––––––– –––––––– Women Men 1995 1996 1995 1996 –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Kronoberg 180747 37.5 27.2 Ljungby 27642 9.6 36.4 27.3 small 4 2 1781 641 Tingsryd 14434 10.1 43.1 32.3 no 4 2 818 290 Växjö 72432 7.8 32.2 24.3 large 4 2 2808 1039 Kalmar 244057 41.3 33.0 Hultsfred 16991 12.5 54.4 41.4 no 4 2 1109 452 Nybro 20873 10.4 46.3 34.0 no 4 2 1142 596 Oskarshamn 27264 10.7 47.5 32.7 small 4 2 1740 842 Bohuslän 320041 44.8 33.4 Kungshamn 9800 10.1 55.0 33.0 no 4 2 532 245 Kungälv 35918 13.4 43.8 31.1 small 4 0 2961 0 Kållekärr 14650 11.1 41.6 29.5 no 4 2 885 342 Värmland 285498 45.8 35.9 Arvika 27010 13.5 52.2 42.2 small 4 0 2121 0 Filipstad 13009 11.8 64.4 52.6 no 4 2 761 268 Skoghall 14200 8.0 39.9 31.3 no 4 2 882 304 Örebro 276828 42.9 31.7 Hallsberg 16571 11.7 49.1 31.7 no 4 2 1008 429 Hällefors 8909 9.6 51.2 43.3 no 4 2 451 206 Örebro 39226 9.8 38.1 30.1 large 4 2 2148 861 Västmanland 261753 51.8 36.2 Arboga 14646 6.8 36.6 24.9 no 4 0 293 0 Kungsör 8440 9.2 55.1 38.9 no 4 2 453 249 Köping 26331 9.7 52.6 38.1 small 4 2 1330 518 Norberg 6649 9.1 54.0 36.5 no 4 2 467 176 Skinnskatteberg 5264 14.2 63.5 50.8 no 4 2 537 176 Surahammar 11344 10.0 58.5 40.1 no 4 1 740 136 Västerås 122995 11.0 48.8 33.7 large 4 2 8211 3809 Fagersta 13583 10.8 71.0 44.2 small 1 1 171 158 Jämtland 136301 49.7 38.4 Svenstavik 8527 13.5 50.8 40.7 no 4 2 546 173 Östersund 59730 14.7 47.8 36.1 large 4 2 5097 2263 Norrbotten 267648 52.6 44.5 Arvidsjaur 7974 12.1 50.7 49.8 no 4 2 508 142 Boden 30700 10.6 47.3 43.4 small 4 2 1675 539 Kalix 19108 10.5 53.6 46.7 small 4 2 822 402 ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––. 16.

(252) Table 2. Sickness certification diagnoses in whole material ——————————————————————————————————— Diagnoses N % ——————————————————————————————————— Musculoskeletal disorders 24568 42.7 Neck and shoulder 5326 9.3 Thoracic and lumbar spine 6895 12.0 Joint diseases 3349 5.8 Other musculoskeletal disorders 8998 15.6 Injuries 6977 12.1 Neck and shoulder 1167 2.0 Thoracic and lumbar spine 289 0.5 Upper extremities 2033 3.5 Lower extremities 2604 4.5 Miscellaneous 884 1.6 Psychiatric disease 8468 14.7 Alcohol and drug abuse 684 1.2 Psychoses 402 0.7 Neuroses 7382 12.8 Gastrointestinal and urinary tract disorders 3471 6.0 Cardiovascular diseases 3237 5.6 Respiratory diseases 3155 5.5 Miscellaneous diseases 7687 13.4 Diseases during pregnancy and puerperium 1914 3.3 Gynaecological diseases 1573 2.7 Dermatological diseases 865 1.5 Other infectious diseases 463 1.0 Other diseases 2872 5.0 ——————————————————————————————————— Total 57563 100 ———————————————————————————————————. current certified period after adjustment for degree of sickness absence, i.e., number of crude days times degree), the municipality disability rate (mean number of sick days and sick pension days per person, 16-65 years of age), the presence of a hospital in the municipality (none, small, large), and reimbursement and employment form in general practice was registered. In addition, the length of the total sickness, episode was obtained. Data on the physician’s age, sex and clinical experience was obtained for certificates from the county of Västmanland.. Financial co-operation At the time of the study, a project with shared economic responsibility for sicklisting between the primary health care organisation and the county branch of the 17.

(253) national social insurance office, called financial co-operation (FINSAM), was launched in three out of eight studied administrative areas in the county of Västmanland. In this project the primary health care and the local social insurance offices were responsible for the financial outcome of sick-listings in the area and were given a sick-listing budget. The primary health care organisation was responsible for the accomplishment of the project in exchange for profit-sharing. A set of incentives was used in order to influence the general practitioners and the clerks working at the local social insurance offices to become more costconscious. In addition, individual general practitioners were offered a bonus agreement, in the form of a percentage of the cost reduction, in exchange for participation in educational and other co-operational activities. 1138 certificates were issued by the 24 general practitioners working in the financial co-operation area and 13 of these signed a bonus agreement.. Sub-studies All 57563 certificates for sickness absence were used for study I. In study II a subset with 31730 certificates, received during April and October 1995 and April and October 1996 at the offices participating all the months, was used to control for seasonal variation. In study IV a subset of 17424 certificates from the county of Västmanland was used together with a subset of 23512 certificates issued by physicians working in general practice. 8257 certificates issued in general practice in the county of Västmanland, were used to study the impact of employment form and financial co-operation. In 1995 and 1996, 56% and 55%, respectively, of the physicians working in general practice in this county, were publicly employed. 1234 certificates were issued by GPs in the financial co-operation area. A summary of aims, study designs, focus variables and physician categories studied is given in table 3.. Statistical considerations In the data analyses a structured model, shown in figure 1 was used. Patient age, sex and diagnosis were used as patient characteristics. Physician category, age, sex and specialisation were used as physician characteristics. Private or public employment, remuneration of general practice and impact of financial co-operation were used as measures of health care characteristics. County, hospital in municipality, municipality population size and rate of disability were used as measures of the impact of the local society level. Change of legislation was used as measure on the national level. In studies I, III and IV a cross-sectional study design and analytical approach was used. In study II a pre-during-post design was used in order to illustrate possible changes over time. To optimise the statistical power graded variables were used as far as possible. For the same reason statistical tests were performed as trend tests 18.

(254) Table 3. Summary of sub-study characteristics. –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Specific aim study I study II study III study IV Study design study I, III, IV study II Focus variables study I study II study III study IV. variation of sick-listing practice between physician categories influence of legislative change on physician sick-listing practice influence of attitudes and local culture on sick-listing practice influence of organisation and remuneration of health care on sick listing practice cross-sectional pre-during-post. physician category and net days certified net days certified over time local structural factors and net days certified organisational and economical factors and length of sickness episode. –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– whenever possible. Grouping of continuous or ordinal variables was used mainly for illustrational purposes. The analyses were performed with the Statistical Analysis System [134] and the JMP programme packages [135]. Standard methods were used for computing means, standard deviations and confidence intervals. Analyses based on continuous data were performed with Student’s t-test or analysis of variance, and analyses based on ordinal data were performed with the chi-square test. Multivariate analyses were performed with linear or ordinal logistic regression technique in their multivariate forms. All tests were two-tailed. Probability values less than 0.05 were regarded as statistically significant. Very small p values were denoted <0.0001 even when they were much smaller.. Results Characteristics of sick-listing certificates in the four studies Characteristics of the certificates and the sick-listed patients in the different subsets are shown in table 4. Patient mean age was 45 years and 60% were women. Approximately 28% were initial certificates and approximately 80% full-time sicklistings. GPs issued the highest share of the certificates, followed by hospital physicians. The differences between the data sets used in study I, II and III were small. The dataset used in study IV only contain certificates issued by general 19.

(255) 20. Table 4. Characteristics of the study populations. ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Study I and III Study II Study IV –––––––––––––––––––––––– –––––––––––––––––––––––– ––––––––––––––––––––––––– N Mean or % 95% CI N Mean or % 95% CI N Mean or % 95% CI ––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Patient age, mean 57563 44.7 44.6-44.8 31730 44.9 44.8-45.0 23512 44.5 44.4-44.7 Patient sex, % Men 22762 39.5 39.1-39.9 12373 39.0 33.6-44.4 8282 35.2 34.6-35.8 Women 34801 60.5 60.1-60.9 19357 61.0 55.6-66.4 15230 64.8 64.2-65.4 Diagnostic groups, % Musculoskeletal 24568 42.7 42.3-43.1 13502 42.5 42.0-43.1 12580 53.5 52.9-54.1 Injuries 6977 12.1 11.9-12.4 3805 12.0 11.6-12.3 2097 8.9 8.6-9.3 Psychiatric 8468 14.7 14.4-15.0 4665 14.7 14.3-15.1 2627 11.2 10.8-11.6 Gastrointestinal 3471 6.0 5.8-6.2 1921 6.1 5.8-6.3 982 4.2 3.9-4.4 Cardiovascular 3237 5.6 5.4-5.8 1824 5.7 5.5-6.0 928 3.9 3.7-4.2 Respiratory 3155 5.5 5.3-5.7 1793 5.7 5.4-5.9 2002 8.5 8.2-8.9 Miscellaneous 7687 13.4 13.1-13.6 4220 13.3 12.9-13.7 2296 9.8 9.4-10.1 Type of certificate, % Initial 16179 28.1 27.7-28.5 9071 28.6 28.1-29.1 6531 27.8 27.1-28.4 Continuation 41377 71.9 71.5-72.3 22654 71.4 70.9-71.9 16978 72.2 71.6-72.8 Degree of sick-listing, % Full-time 45777 79.6 79.1-79.8 25056 79.0 78.2-79.8 18450 78.5 77.9-79.0 Part-time 11767 20.4 19.9-21.0 6664 21.0 20.6-21.5 5053 21.5 21.0-22.0 Number of days certified, mean Length of sickness episode 57549 193.4 191.0-195.7 31723 197.5 194.2-200.9 23507 173.2 169.6-176.8 Length of current certificate 57536 36.0 35.7-36.2 31716 36.3 35.9-36.6 23507 30.1 29.7-30.4 Net days of current certificate 57535 31.0 30.8-31.2 31716 31.1 30.8-31.4 23506 25.6 25.2-25.9 Issuing physician, % General practitioners 23512 40.9 40.5-41.3 12956 40.8 40.3-41.4 23512 100.0 Hospital physicians 22540 39.2 38.8-39.6 12441 39.2 38.7-39.7 Occup. medicine physicians 6765 11.8 11.5-12.0 3779 11.9 11.6-12.3 Other physicians 4662 8.1 7.9-8.3 2519 7.9 7.6-8.2 Unspecified 84 0.1 35 0.1.

No results found