Unsanctioned techniques for having sickness certificates accepted: a qualitative exploration and description of the strategies used by Swedish general practitioners.

OBJECTIVES: To explore informal and unsanctioned techniques general practitioners (GPs) employ as a means to increase the likelihood of sickness certificate approval, following the Swedish Social Insurance Agency's (SSIA's) consolidation of the gatekeeping role in sickness benefit evaluation.
DESIGN: Qualitative semi-structured interviews with 20 GPs working in Swedish primary care. A thematic analysis of the transcribed material was carried out to map different techniques employed by the practitioners.
RESULTS: Eight techniques were identified, particularly with respect to the way in which the sickness certificate is written to ensure approval by the SSIA. The identified techniques were most commonly adopted when the patient's case was perceived to be at high risk for rejection by the SSIA (such as psychiatric illnesses, chronic pain etc.).
CONCLUSIONS: The findings imply that the informal and unsanctioned techniques are complex and ambiguous. They are used intentionally and covertly. The study also suggests that, while the consolidation of SSIA's gatekeeping role may have resolved some sickness absence issues, a consequence may be that GPs develop unsanctioned techniques to ensure compliance.

Introduction

Since the 1990s, sickness absence in Sweden has fluctuated [1]. A surge in absenteeism beginning in the late 90s prompted public debate. In mid-2000, a discursive shift in the debate introduced new arguments for reforming the social insurance system: social security was overused, and perhaps even subject to fraudulent claims. It was argued that the controls intended to regulate eligibility for sickness benefits had failed [2]. In this old system, the physician was the de facto gatekeeper, deciding eligibility for sickness benefits. Research suggested that physicians themselves were dissatisfied as guardians of social resources, and that they had difficulties handling it. They took certain liberties when completing sickness certificates—exaggerations, ambiguous and vague wording, writing in favour of the patient etc.—partially due to pressure placed on them from the patients [3-5].

In response to the fluctuations, and the debates and research surrounding it, substantial changes to the social security system have been continuously implemented since the second half of the 2000s. Two changes stand out as fundamental to the shift in dynamics between the SSIA and physicians:

In 2008, the rehabilitation chain (‘rehabiliteringskedjan’) was implemented as a means to achieve sick leave process efficiency. It stipulates that the SSIA should conduct more rigorous assessment of a person’s work ability after 90 and 180 days of sick leave. After 90 days, work ability is assessed with respect to possible assignments the employer might offer, and after 180 days, work ability is assessed in relation to any assignment or job on the labor market [6].

In 2009, an official governmental report introduced a model for evaluating work ability. This so called DFA-chain (‘DFA-kedjan’) highlights three essential dimensions for determining eligibility for sickness benefits on the basis of sickness certificates: diagnosis (e.g. depression), functional impairment (e.g. an inability to concentrate, impaired memory, tiredness and rating scale scores) and lastly how this affects work ability (e.g. to be attentive or understand instructions). An acceptable sickness certificate should adequately describe these dimensions, as well as establish a reasonable link between them [7].

Sickness certificates are required after one week of sick leave. Physicians are expected to reason in accordance with the DFA-chain when issuing certificates, which are filled out with the use of a standardized form. At the time of conducting the interviews, the form explicitly requested a description of ‘objective findings on an organ level’, regardless of type of disease in question [8]. The form is then submitted to the SSIA for evaluation. The SSIA can accept the certificate, ask for further specification or reject it outright.

In recent years, the SSIA has become more rigorous in assessing eligibility in accordance with the DFA-chain and some aspects of the rehabilitation chain [1,9]. As a result, the task of controlling eligibility for social insurance has been further transferred from physicians to the SSIA. The reforms seem to have mitigated the potential conflicts between patient and physician during the sickness certification process. For example, disagreeing with patients who demand sickness certificates from their GPs is less of a problem when the SSIA are responsible for making the final decision [10]. However, the SSIA’s more rigorous assessment of sickness certificates has, between 2014 and 2016, resulted in a threefold increase in the proportion of rejected sickness certificates [11]. The characteristics of GPs’ working conditions (time constraints, the high number of complicated and unclear sickness certification cases, exclusively dealing with outpatients), make their relationship to the SSIA particularly vulnerable: between 2004 and 2017, the proportion of GPs who reported that their medical judgments were questioned by the SSIA rose from 10% to 57%. The number of GPs who experienced that the SSIA requested unnecessary corrections to the sickness certificates increased from 48% to 72% between 2012 and 2017. In 2017, 72% of all surveyed GPs felt that the SSIA requested ‘objective findings’ in cases where objective signs are notoriously difficult to identify in the clinical setting (e.g. psychiatric disability, chronic pain, etc.) [12].

Several overviews regarding physician sickness certification practices have been compiled during the last two decades. Aspects relevant to this study, such as managing conflicting roles as a practicing physician, and difficulties in cooperation with other actors, have spurred some research interest [13]; however, most of it has focused on GP attitudes concerning sickness certification rather than actual praxis. There are studies on the theme of conflict between stakeholders related to sickness absence, but they have to a large extent targeted conflicts between physician and patient, as opposed to conflicts between physician and external institutions [14-16]. Earlier research has hinted at the existence of ways for circumventing the expectations of insurance agencies. Emphasis has been placed on the physicians’ views and attitudes towards sickness certification [10,17,18], rather than descriptions of the actions themselves, or in some cases focused on understanding the actions as textual incompetence [19] rather than as skillful goal-oriented activities.

The main assumption of this article is that the experienced difficulties of GPs’ in their day-to-day interaction with the SSIA could in some way affect praxis, especially in terms of informal solutions to these difficulties. The purpose of this study is to explore the informal and unsanctioned techniques employed by GPs as a means to increase the likelihood of sickness certificate approval, following SSIA’s consolidation of the gatekeeping role in sickness benefit evaluation.

Material and methods

Study design

A qualitative and descriptive approach has been adopted for the purposes of exploring the informal practices of GPs regarding sickness certification. 20 in-depth interviews constitute the data corpus for the study.

Participants

GPs were strategically sampled to ensure a balanced composition regarding gender, age, location, professional experience and educational background. Twenty interviews were conducted before thematic saturation was reached. Informants were recruited through direct contact with health centers, as well as through advertisement in Facebook groups and Internet forums for physicians. Additional informants were recruited through snowball sampling. Eleven informants were female, nine were male. Thirteen of them worked in the Stockholm region, while the remaining seven worked in other Swedish regions. The age range was 29–65 years. The average experience in primary care was nine years (ranging between 0.5 and 34). Thirteen had received their medical degree at Swedish faculties, and seven from Eastern and Central European countries.

Data collection

The data were gathered between December 2017 and February 2018 through qualitative in-depth interviews with GPs working in primary care. The interviews were semi-structured, roughly following a preset interview guide. Topics covered during the interviewing process included: general attitudes towards, and difficulties regarding, sickness certification and the SSIA; examples of difficult sickness certification cases, and (both sanctioned and unsanctioned) ways of solving them; lastly, unsanctioned ways of solving difficult cases were broached with informants, through the use of both direct and indirect forms of questioning (see Appendix). The interviews were recorded and transcribed. One-on-one interviews were chosen over focus groups, as the topic was considered sensitive and that peer pressure could inhibit the informants’ willingness to report compromising information [20]. Similarly, being a medical doctor myself was a double edged sword. GPs seem to be more open and spontaneous when interviewed by fellow professionals, switching between treating the interviewer as professional peer and a private confidante. However, respondents can also feel judged and distance themselves [21]. To the extent possible, I welcomed the two roles ascribed to me as interviewer, while avoiding judgmental reactions to the informants’ accounts.

Data analysis

A thematic analysis, as devised by Braun and Clarke, was used to interpret the data. This entailed an identification and explication of recurring patterns of meaning across the data set [22]. After a preliminary familiarization with the transcripts, coding of units of meaning relevant to the research question of this article was undertaken; this meant placing greater emphasis on the description of techniques adopted by GPs, rather than, for example, focusing on GPs’ beliefs and attitudes regarding the SSIA. The preliminary coding was then checked by the researcher’s supervisor to ensure fidelity to the transcripts. Categories were inductively developed on the basis of the codes. In the following phase of reviewing the categories, the secondary supervisor aided in sharpening distinctions and minimizing overlap. For example, the preliminary category of ‘generalization’ was removed due to excessive overlap with other categories. Longer illustrative quotes were attributed to informants by letters between (A) and (T).

Results

Most informants reported that the last years of practice were marked by increasing difficulties surrounding the authoring of sickness certificates. The GPs were more often caught in a frustrating dilemma: either write transparent sickness certificates, thereby risking failure to satisfy SSIA standards, or complete them in a less transparent way in order to meet the standards set by the SSIA.

A minority of the informants were fine with any decision SSIA would reach. The majority was not, at least not in practice. Although a broad variation of views were found among them regarding the work of the SSIA, ranging from ‘they do a great job’ to the conviction that ‘they make people sicker’, most of them (with the exception of four) admitted that they occasionally adopted informal instrumental techniques when authoring sickness certificates for the purpose of securing SSIA approval. The GPs referred to the instrumental techniques with terms such as ‘embellishment’, ‘uglifying’, ‘exaggerations’, ‘amplification in order to illuminate the total impression’, which demanded ‘street smartness’, ‘tactics’ and ‘maneuvering around’. The reasons for doing it varied, and often several interrelated reasons were given. The four categories of inductively developed reasons (in the best interest of the patient, in defence of professional autonomy, to save time, or simply to bridge the gap between complex medical reality and the conceptual toolbox at their disposal), did not seem to affect preference for any particular technique.

A repertoire of techniques in dealing with the sickness certification process

Eight categories of techniques used to maximize the likelihood of having sickness certificates accepted by the SSIA were identified: exaggeration, quasi-quantification, omission, depersonalization of the patient voice, adjustment of disease progression, buzzwords, communication off the record and production of redundant somatic data.I define them as informal, and unsanctioned, ways of maximizing the likelihood of sickness certificate acceptance by the SSIA.

Dealing with quantities: exaggeration and quasi-quantification

A widespread conception among the informants was that quantification of patient symptoms is crucial when writing a successful sickness certificate. Consequently, informants feared that the SSIA case workers risked missing the overall picture and deny patients sickness benefit, sometimes solely due to insufficient quantification. The problem of insufficient quantity was informally resolved in mainly two ways: exaggeration and quasi-quantification. A young GP said the following:

I have exaggerated. I exaggerate the problems, so it’ll pass […] I’ve even told the patient: maybe this wasn’t exactly what you told me, but to be certain that you’ll get your sick pay, I will have to write it like this. For example […] wakes up ten times per night, although in reality maybe it is five times per night (I)

In this context, exaggeration seems to denote statements that represent quantities as more or less than they actually are. Yet, the informants who exaggerated were careful to differentiate exaggeration from lying or fabrication. One informant stated: ‘I amplify to illuminate the total impression’ (G).

Another way of solving the problem of insufficient quantity was to derive quantities from qualities, which I call quasi-quantification. This operation differs from exaggeration because there is no quantity to begin with. Instead, a quantity is created. One informant depicted the difficulties of assessing extent and frequency of a (psychiatric) patient’s disabilities:

[No patient] can answer that, no patients reason in those terms. So I ask a bit about what they do during the days, when do you wake up, when do you eat breakfast, what do you do after that. And then I transform it to minutes or hours or something like that (C)

Informant (C) felt that patients tend to have difficulties providing quantitative data directly, and therefore preferred to extrapolate them out of the patient narrative. Also, some informants felt that the encounter with the patient could become too inquisitorial when focusing on quantities, thereby undermining the patient–physician relationship.

Omission

Several GPs recounted how they sometimes deliberately excluded facts when issuing sickness certificates, in order to influence SSIA’s decision. The cases could comprise of omitting references to small but remaining work ability, or not mentioning leisure activities. One GP gave the following example:

For example, if the patient says that she’s still active in her local sports club, and has occupational burnout, which actually was a case I had this week. It wasn’t mentioned [in the sickness certificate]. She was being reintroduced to work at the time, but I didn’t write it in the certification document […] things that speak against sick listing tend to be omitted (B)

This GP felt responsible for ensuring that the sickness certificate went through after he had made the decision to issue it, and that the considerable risk of rejection mandated the use of this technique.

Depersonalization of the patient voice

Several informants reported that the SSIA did not seem to be interested in the patient’s own narrative. This was particularly problematic in disability cases with few or no objective findings, when the patient’s own narrative is the most important source of information. Consequently, a perceived devaluation of the patient’s voice was sometimes circumvented by objectivizing it in various ways. On the most superficial level, some cosmetic changes made to the written document could erase the presence of the patient:

[…] so you need to ask the patient, ‘what is it you can’t do?’ You ask, but of course you don’t write that you asked the patient, because it could lead to a rejection by the SSIA. I’ve seen this happen a number of times, that they motivate it: ‘well, what the patient says doesn’t mean a thing’ (K)

On a deeper level it sometimes entailed transposing first person accounts onto something more tangible, such as objective clinical signs. For example, informant (J) recalled how he sometimes wrote that a depressed patient without visible signs of depression ‘looks sad’ or has a ‘flat facial expression’, rather than opting for a ‘neutral mood’. The informant also expressed frustration with sanctioned ways of providing SSIA with objective data, such as rating scales. Other informants had also experienced that rating scales were not always accepted in isolation, particularly if they did not accord with the mental status examination.

Adjustment of disease progression

Some of the informants believed that both expected and unexpected disease progression could influence patients’ eligibility for sickness benefit. To prevent premature termination, GPs could decide to either understate or overstate how fast the patient was recovering when renewing sickness certificates. A GP gave an example of understating when sick listing a depressed patient:

The first certification was easy, because she really had a flat facial expression and a monotone voice, […] then we started treatment, she went to a psychologist, was given medication and started to feel better clinically, and she didn’t have the signs anymore. But it was still out of the question that she should be exposed to work yet. So we see the progression in this patient, and other patients with these diseases. […] That’s when one has to tell white lies sometimes, such as ‘she looks tired’, although that’s not the case (G)

At other times, informants felt compelled to overstate the rate of progress and ability. Another GP told the story of a recent patient he was dealing with, nearing retirement age with occupational burnout, and rather severe hypertension. Working part time was not going well.

[H]e would actually need a total sick leave. But he is even more stressed out about being on total sick leave, because it increases the risk that SSIA just removes him from work and puts him in some goddam’ vocational training programme. That would break him, totally. From a purely medical standpoint he needs to be totally off work (J).

Instead of insisting on 100% disability, the GP and his patient decided to lower his disability to 50%, in order to avoid putting the patient at the disposal of the whole labor market, as dictated by regulations.

Buzzwords

Standardized words and phrases were widely employed by the informants when writing sickness certificates. They did it reluctantly, worried that it partially dissociated the words from what they are supposed to signify, or what some GPs called ‘catchwords’, or ‘code words, platitude’. One GP illustrated with using a particular phrase when describing her clinical findings in psychiatric patients:

I write ‘cognitively impaired’, because I’ve learnt that they want to hear that particular phrase. It’s not enough to write memory and concentration loss; for some reason the word ‘cognitive’ must be used. It has become such a routine, you use a few keywords (M)

The informants used buzzwords in two related ways: (1) As a device on a purely textual level, using words and phrases they know that the SSIA wants (as the case above); (2) As a way of directing the actual conversation with the patient. For example, one informant recounted how she could ask leading questions using buzzwords, in order to get an affirmative answer to it, just to be able to write it in the sickness certificate.

Techniques beyond the text: communication off the record

Techniques for conveying views of a patient’s disabilities were not limited to techniques of writing the sickness certification document in a particular way. When requested to clarify sickness certification documents, some GPs would sometimes phone SSIA caseworkers. Beyond the completely sanctioned purposes of phone calls (such as giving a more thorough account of a patient, sorting out misunderstandings, and finding out what additional information the SSIA needs) it was a pathway for subjective and affective modes of persuasion. As one GP put it: ‘If you have a caseworker you trust, […] it could become a little “off the record”, things you don’t want to write’ (E). This so called off the record communication concerned matters the informants both did not want to write, but also things they could impossibly write because it did not meet the standards of writing objective certificates. For example, the GP’s subjective impressions could be communicated, as one of the informants explained:

I don’t write ‘I was genuinely worried for the patient’, but it is something I can say to the caseworker, and to add my thoughts and values to the whole, in a way I can’t do [in text]. That makes it extra obvious for the caseworker, that the [patient’s] narrative is very credible, and I think they are more inclined to [say]… ‘oh, was it that bad? Ok’, and then you can cooperate a bit more. (G)

Another sub-category was to involve oneself rather than one’s subjective impressions of a particular patient. It could involve talking about personal experiences of psychiatric disease, in order to make the caseworker understand the severity of the condition (one GP told a caseworker of his own depression, for example). It could also take the form of asserting one’s own medical authority, for example through categorical statements about the state of the patient: ‘sometimes I notice that it helps to say “there is no doubt about it”‘ (K). Occasionally, some informants could remind the caseworker of their superior medical knowledge.

Techniques beyond the text: production of redundant somatic data

The standards for establishing links between diagnosis, impairment, and disability was by some informants considered inflexible to the degree that GPs must be redundantly clear about them. Hence, when attempting to make a solid case for sick listing, they could feel compelled to add information despite significant nuisance, or perhaps even potential harm, to the patient. An illustrative example was provided by one informant, who recounted a case of vertebral compression fracture. Although the GP did not believe that a radiological confirmation of the diagnosis was medically motivated, she ordered it anyway, solely to acquire objective findings. ‘She will be exposed to unnecessary radiation for the sake of the SSIA’ (D).

Discussion

This qualitative interview study has aimed to explore and thematize the informal and unsanctioned techniques GPs employ as a means to increase the likelihood of sickness certificate approval. The findings provide a non-exhaustive overview of these techniques used for meeting the stricter standards of sickness certification imposed by the SSIA. The eight categories of techniques identified were exaggeration, quasi-quantification, omission, depersonalization of the patient voice, adjustment of disease progression, buzzwords, communication off the record and production of redundant somatic data.

They were mostly utilized when GPs were convinced that a patient had a medically motivated work disability, but may have had a hard time satisfying the criteria of objective findings, as well as establishing links between the triad of diagnosis, (objective) impairments and work ability. Typical cases primarily involved patients with unclear prognoses that were on sick leave for longer periods (especially exceeding 90 and 180 days), and patients lacking objective findings to link their subjective illness with an objective disease (most often psychiatric problems such as depression and occupational burnout, but also patients experiencing chronic pain).

The techniques are quite heterogeneous, but a common denominator is their role in the struggle over objectivity. Some techniques aim at producing quasi-objectivity (for example when transforming the patient’s narrative into an objective finding made in the examination room); others at evading objectivity (for example when trying to convince caseworkers over the phone); a few techniques produce objectivity at a price (for example when referring a patient to a redundant radiological examination to secure objective findings, at the price of unnecessary radiation exposure).

The distinction between sanctioned and unsanctioned ways of communicating with the SSIA is not always clear-cut. Rather, the techniques fit on a continuum. For example, when issuing sickness certificates (or any document, for that matter), physicians must always fish out relevant facts from the stream of unordered clinical impressions. Leaving out particular facts is essential in producing meaningful text. Yet, under some circumstances, the omissions of facts can be considered unsanctioned.

The techniques are intentionally used, and are therefore not simply mistakes. They display a certain level of sophistication, and go to great lengths to (at least superficially) comply with imposed rules and standards. Consequently, the techniques are difficult to identify as either compliant or non-compliant. Are they overzealous implementations of the standards of objectivity, or ways to wiggle out of them? That organizational misbehavior, even resistance, can straddle the line between compliance and non-compliance has been pointed out in earlier sociological research concerning ‘everyday resistance’ among airline pilots [23], office workers [24] as well as in spheres outside of the workplace [25]. This suggests that the use of informal techniques could be understood in similar terms. However, understanding GPs’ informal sickness certification techniques through the lens of everyday resistance transcends the scope of this exploratory study.

According to Michael Lipsky, public policy does not simply encompass political and legislative stipulations, rather it is something made in every single street-level encounter between professionals and their clients [26]. The cunning use of objectivity can be understood as ways for GPs to control public policies by modulating their implementation, thereby retaining their own discretionary power. Needless to say, the discretionary powers of street-level bureaucrats (such as GPs) are not unproblematic. Advocacy of individual patients can be incompatible with an organizational perspective.

The mere existence of non-compliance might have implications from an organization perspective. Further research is needed regarding the prevalence of non-compliant actions, as well as potential causal links between the implementation of external control mechanisms and the development of sophisticated forms of non-compliance in the sickness certification praxis. The relation between them may be counterintuitive, in the sense that more control not always entails less deviation: interventions often produce an appearance of conformity, cloaking new and deeper forms of misbehavior that can be even harder to identify and counteract [27]. Similarly, consolidating SSIA’s gatekeeping role may have solved some sickness absence issues (revolving around patient-driven demands for sick listing [10]) at the cost of inadvertently stimulating new forms of underground praxes that are necessary to keep the head of clinical practice above water.

Strengths and limitations

There are some potential sources for bias in the collected data. Firstly, no informants have an educational background in countries outside of the EU. Secondly, a majority of the informants work in Stockholm. What problems does this pose for a qualitative study? Techniques that are unique for GPs outside of the EU and Stockholm could have been missed in this study. As this is not a quantitative study, the preference for one technique over any other is not sought for, and therefore the slight overrepresentation of one geographical area is a minor problem. Neither is the ratio between GPs that use techniques, and those who abstain from them, an overarching concern in this study.