Background

Prognosis is the forecast of an outcome, and in a disability evaluation this outcome mainly concerns the functional abilities needed for work [1]. For example, will a 50-year-old nurse with degenerative knee arthritis be able to climb stairs in the future? [2, 3]. The evaluation of the prognosis is an important aspect of disability evaluations [4, 5] and has financial, personal, and legal consequences. Prognosis assessment is internationally considered to be one of four main tasks in disability evaluations [6], but physicians find the assessment of prognosis challenging [7, 8]. Prognostic questions make up 39% of physicians’ information needs during disability evaluation [9].

A Dutch study identified 23 aspects for consideration in the prognosis evaluation during disability assessment [8]. These aspects fell into six areas: the specific disease or disorder, any treatment or potential treatment, the course of the disease, external information from other specialists or scientific evidence, patient-related aspects (e.g., coping), and physician-related considerations (e.g., role, empathy). However, it is not known how these aspects are valued and dealt with in practice. Although this qualitative study provided us with a range of aspects considered during prognosis assessment, it did not provide quantitative information on how important physicians rate these aspects or why some are considered more or less important. More insight into the importance of the aspects considered during prognosis assessment will also guide the development of a prognostic tool in the future.

Therefore, the purpose of this study was to evaluate how physicians rate the importance of six prognosis areas considered during work disability evaluation and to explore what kind of support they would like during prognosis assessment.

Methods

Design

In this study, we asked participants in a workshop to complete an anonymous paper questionnaire.

Participants

The study participants includes mainly social insurance physicians and occupational physicians, who assess a person’s work capacity in the context of a disability benefit claim or with respect to reintegration, respectively. The participants attended an annual congress of the professional association for social insurance medicine in the Netherlands. The programme comprised several lectures and workshops, e.g. on prognosis assessment. This workshop involved a plenary presentation, followed by a case demonstration, discussions and the completion of an anonymous questionnaire. All procedures followed were in accordance with the ethical principles for medical research involving human subjects: before participants attended the workshop, they were informed about the study aim and procedure. Participation was voluntary; participants had the opportunity to discontinue at any time.

The questionnaire

Two authors (RK, JH) developed the questionnaire specifically for this study, it consisted of three parts. The first presented the 23 prognosis aspects found by Kox et al. [8]. The physicians were asked to rate the importance for the prognosis on a Likert scale ranging from 1 (not important at all) to 10 (of utmost importance), both for the presented case vignette (see Fig. 1) and “in general,” that is, for any kind of patient. Each aspect was followed by an open question asking participants to motivate their answer, for example: “How important, in your opinion, is the severity of the disease for the prognosis? Please give a rating on a scale of 1 (not important at all) to 10 (of utmost importance) for this case vignette and in general.” The second part consisted of five questions. Four were about encountered needs and solutions for the assessment of prognosis. The fifth asked whether the physician judged that the assessed limitations in functioning for the case vignette were permanent. The last part of the questionnaire consisted of questions concerning the respondent’s characteristics, for example, age, sex, job area (e.g., social insurance medicine or occupational medicine), years of experience performing disability evaluations, and number of hours worked per week.

Fig. 1
figure 1

Case vignette and examples of evidence provided to physicians

Full size image

Statistics and analyses

The answers were stored and analyzed in IBM SPSS Statistics for Windows®, version 25.0 (SPSS Inc., Chicago, IL, USA) using descriptive statistics. As most of the dependent variables from both the Likert scale questions and the demographic questions were not normally distributed, we calculated the median and interquartile range (IQR) to describe the distribution of the scores. The open answers were read and entered into SPSS by LJ; two researchers (SK and BD) double-checked these data. For the qualitative analyses, one researcher (SK) explored the answers and grouped them into themes; this process was then checked by three other researchers (JH, FS, BD, and LJ.)

Results

Demographics

Of the 78 physicians attending the workshop, 76 completed the questionnaire and handed it in after the workshop. The demographic questions in the last section of the questionnaire were completed by 53–55 of all responders. Most of the participating physicians worked in social insurance medicine (n = 35) or in occupational medicine (n = 15). Three participants worked in both these specialties and two worked in another medical specialty. A slight majority of the physicians were over 55 years old (n = 31, versus n = 23 less than 55 years), most had been in the profession for over 15 years (n = 42, versus n = 12 less than 15 years), and most worked 32 h or more per week (n = 40, versus n = 13 less than 32 h).

Ratings of prognosis aspects relevant for the assessment of prognosis

When the physicians were asked to rate the importance of the prognosis aspects “in general,” their median scores varied between 6.0 and 8.0 (see Table 1). The median importance scores concerning the same prognosis aspects for the case vignette (person with MSA parkinsonism) were between 4.0 and 9.0. Interquartile ranges (IQRs) varied between 1.0 and 4.8 “in general” and between 1.0 and 5.0 for the case vignette.

Table 1 Importance scores of prognosis aspects (n = 76), “in general” and for the case vignette
Full size table

For the prognosis aspects scored “in general,” the first four prognostic areas (disease, treatment, course, and information) had median scores of 8.0, with IQRs of between 1.0 and 2.0. The remaining two areas—namely patient- and physician-related considerations—had more variable scores (i.e., medians ranged between 6.0 and 8.0), with IQRs of between 2.0 and 4.8.

In comparison, the case vignette had a somewhat higher importance score for severity of the disease (9.0), while prognosis aspects—such as disease-maintaining factors and vocational rehabilitation—had a somewhat lower score (7.0). From the interquartile ranges (IQRs), it can also be seen that physicians varied more in their assessment of, in particular, the patient- and physician-related considerations. Scores in the first four areas ranged between 7.0 and 9.0 with IQRs ranging from 1.0 to 3.0, whereas scores in the latter two varied between 4.0 and 7.0 with IQRs of between 3.0 and 5.0.

Open questions on the motivation for the ratings

Illustrative answers are shown in Table 2.

Table 2 Motivations accompanying the scores of importance (open questions) [Translated by the authors. Original source consisted of Dutch handwritten texts.]
Full size table

Disease

Some physicians stated that the nature and severity of the disease are a starting point for the assessment of the prognosis (e.g., #42). Several physicians made it explicit that the degenerative, progressive nature of the disease is essential for the prognosis (e.g., #32). It was also noted that in the case vignette, the progressive nature of the disease was most important for prognosis, whereas in other cases other prognosis aspects might also be relevant (e.g., #73).

Treatment

Considering treatment, many physicians evaluated the effect of the current one and looked for future options that could bring about a change (e.g., #6, #27). It was also pointed out that one should distinguish short-term treatment effects from long-term disease outcomes. In addition, physicians distinguished between relief of symptoms and curation (e.g., #26, #52).

Course

Regarding the course of the disease, physicians considered that the rate of progression of the disease and the results of treatment up to the moment of the disability evaluation forecast future developments (e.g., #4). For the case vignette with a progressive medical disease, it was mentioned that the course of the illness (amount of deterioration over time) was of main importance (e.g., #27).

Information and evidence

Information from the treating physician is considered important to get an insight into the patient’s unique course and progressive symptoms (e.g., #15). Some physicians stated that they preferred research evidence (e.g., #23). Others pointed out that you may not need scientific literature in all cases (e.g., #19), but you may need more evidence when a patient claims a large number of ongoing disabilities (e.g., #36).

Patient-related considerations

Several patient-related factors, such as coping and the way patients perceive their illness, were considered important (e.g., #23, #45). Others spontaneously mentioned age as a patient-related ADP that either guides the decision (#51) or does not (#18). One physician (#26) stated that patient-related factors can be important for vocational rehabilitation, but that they inform the physician conducting a disability evaluation to a lesser degree. Physicians also mentioned that it was important to assess whether patient-related prognosis aspects can be changed (e.g., #46).

Physician-related considerations

Regarding the importance of physician-related factors, differing opinions were found, ranging from “very important” (e.g., #72) to “as little as possible” (e.g., #32), and some mentioned them to be case-dependent (e.g., #50, #73). Some physicians said that ethical aspects were considered (e.g., #50) and that the physicians’ own impressions were important (e.g., #72), while others indicated that these should not be a part of the evaluation as they could unintentionally influence their judgment (e.g., #73, #32) or could bias the physician (e.g., #45), and efforts should be made to ensure that a client receives similar outcomes from different assessors (e.g., #42). However, it was also stated that these prognosis aspects are often hard to recognize, given their often subconscious nature (e.g., #24). They are interrelated with a physician’s “blind spot” (#29), but it is important to be aware of those (e.g., #27).

Specific support that physicians suggest would aid them during the prognostic assessment

Several answers illustrate additional needs, suggestions, and support for physicians during the assessment of prognosis (see Table 3). Principles of evidence-based medicine (EBM) were suggested as an integral part of the support (e.g., #45, #73), which could include evidence-based information on the nature (e.g., #45) or course of a disease (e.g., #73), treatment options, (e.g., #45, #73) and prognostic factors (e.g., #8, #16, #32, #77) that could be supported by figures, facts, and references (e.g., #46). Some physicians mentioned that a potential tool should provide information about relevant prognosis aspects (e.g., #58). The case vignette used might illustrate relevant prognostic information (e.g., #41) as an example. A diagrammatic way of presenting information (such as checklists, programmed instructions, or flow charts) was also suggested (e.g., #17).

Table 3 Potential elements of supportive aid for prognosis assessment (open and checkbox questions)
Full size table

Type of aid

As a preferred medium, digital aids were most frequently chosen (e.g., “website” n = 38, “app” n = 17). A desk pad format (a leaflet summarizing key remarks) was also considered useful (n = 16).

Discussion

In this study we evaluated how physicians working in the field of disability evaluation rated the importance of six areas of prognosis, namely disease, treatment, course of the disease, external information, patient-related aspects, and physician-related aspects. Although all six areas were considered important, there was more consensus among physicians concerning the three medical areas (disease, treatment, and course). The scores and verbatim remarks regarding the patient- and physician-related considerations (non-medical areas) reflected a more varied appreciation of importance among physicians.

Our use of a clear and severe medical case vignette may have influenced the physicians’ more limited appreciation for non-medical prognosis aspects during the prognosis assessment. Verbatim remarks were made to the effect that medical considerations sufficed, and the importance of medical prognosis aspects appeared relatively high. In contrast, physicians’ remarks regarding prognosis assessment of cases “in general,” suggested that non-medical prognosis aspects could become more important in cases with a less clear medical background. Physician-related considerations appeared not to be given an explicit role in the assessment. Some physicians actually mentioned this and scores within this area appeared lower. However, it was also mentioned that they do play a relevant, though often implicit or subconscious, role.

As a form of support, physicians mentioned some kind of overview of prognosis aspects and relevant scientific evidence. A digital form of support was preferred.

Comparison with the literature

The importance of the six areas of prognosis: diverging opinions on non-medical prognosis aspects

In disability evaluations, the framework of the International Classification of Functioning, Disability and Health (ICF) [10] has been adopted by and used in several countries [11, 12]. This classification system can also be used to describe work functioning, taking contextual and personal factors into account. However, there is some criticism that the ICF scheme is suggestive of the dominance of a medical perspective rather than a biopsychosocial one. Therefore, suggestions for a revision of the ICF have been made [13, 14]. For prognosis assessment in work disability evaluations, there is another problem with the ICF: The dynamic aspect of functioning and health over time is not addressed, nor are any consequent changes in activities or participation over time. As such, the use of the ICF during social–medical history taking and prognosis assessment is limited [11]. These two issues with the ICF (i.e., suggestive dominance of the medical perspective and absence of a time frame) leave room for diverging opinions [15, 16], especially regarding non-medical aspects, which corresponds to our findings.

The importance of the six areas of prognosis: non-medical prognosis aspects

Our study shows that the perceived relevance of patient-related considerations varies in prognosis assessment. In the case of a clear medical condition, they might not all be necessary or might make a smaller contribution to the prognosis evaluation, whereas for medically unexplained physical symptoms (MUPS) or in vocational rehabilitation settings, physicians may attribute more value to them.

In a study on arguments used in disability prognosis [17], medical clarity contributed to the type and number of arguments used. In less clear medical cases, more arguments were used, including non-medical aspects such as coping or education. In addition, Ankersmit et al. [17] found that the expected outcome reception played a role: when disability claims were substantial or the chances of appeal were considered high, physicians preferred a more comprehensive evaluation of all potentially relevant aspects, including patient-related considerations (a preference that was also mentioned by physicians in our study). Another study [18], which concerned aspects for consideration in disability evaluation, suggested that the timespan covered in the evaluation (e.g., 5 days, 3 months, or 5 years) determined the value of the aspects used: For longer timespans, less relevance was attributed to patient-related factors. Some physicians in our study made explicit comments that in vocational rehabilitation settings, they would attribute more value to those patient-related prognosis aspects and that these could be targeted by interventions. Prognostic systematic reviews have shown that, regardless of context, patient-related factors, such as coping and self-efficacy, have more prognostic value for participation in work than factors of a medical nature [19,20,21,22] and could inform prognosis assessment and re-evaluations over time.

The importance of the six areas of prognosis: implicit role of physician-related considerations

Physician-related considerations were regarded by the physicians in our study as often subconscious factors that influence the physician’s judgment. The consequences of those influences for prognosis assessment are mentioned in various medical studies within several fields [8, 23,24,25]. For example, physicians tend to express the prognosis in a way that is too optimistic. This could originate from, for example, a tendency to provide hope or to stimulate healthy rehabilitation and recovery behavior [8] and “not to harm” by taking those away [8, 24]. For example, an earlier study found that physicians performing disability assessments did not want to permanently deny young adults any hope or chance of future work participation [15], given the positive aspects of work. To overcome some of those consequences, physicians in our study said that the important thing is to be aware of these influences.

The way to support physicians: EBM as core ingredient, covering relevant prognosis aspects

The physicians in this study wanted support, including evidence-based prognostic information, preferably pre-appraised. Evidence should ideally be tailor-made, as suggested by earlier studies [7, 8]. Even if prognostic evidence is present, it requires skills to find, appraise, and apply it in a particular case and within the country-specific legislative context [26,27,28]. However, studies have demonstrated that training in evidence-based medicine may improve the quality of disability evaluations, prognosis assessment, and job satisfaction [4, 29]. Some physicians in our study mentioned that useful prognostic search strings might be provided, thus meeting the demand for help in finding prognostic evidence. This demand was also reported in other studies [26, 30], which led to research providing potential search strategies, filters, or strings regarding themes such as prognosis and work participation [30,31,32]. The desire for user-friendliness, simplicity, and help in overseeing the various prognostic aspects was also identified by Kox et al. [8] and Louwerse et al. [33], both of whom were exploring possible prognostic tools. In contrast, when presented with possible prognostic tools, physicians also stressed the importance of preserving their professional autonomy to make unique, tailored evaluations. [34, 35] Moreover, they needed to become acquainted with them and they wanted to estimate their validity. [35]

Strengths and limitations

A strength of our study is that it combined insights from the quantitative data with qualitative data from corresponding remarks from physicians. However, our questionnaire was not suited for an in-depth exploration of the reasons why the importance was scored higher or lower. Also, we cannot exclude a selection effect, as the physicians attending the workshop may be more interested in this topic, but it is not clear how or in which direction this could have influenced the results.

The fact that the case vignette concerned a clear, severe medical condition might explain why the physicians did not elaborate much on the functional abilities of the patient. Some commented that this patient had no abilities for work at all and referred to the medical condition. It would be useful to evaluate the importance of prognostic aspects in a similar study that includes a case vignette with a less severe, chronic health condition (e.g., rheumatoid arthritis) or a condition with less medical clarity (e.g., MUPS). However, we tried to partially counter this disadvantage by asking for opinions on the prognosis for “general” cases, although we acknowledge that what is “general” could mean different things for the participants in this study.

Conclusions

This study demonstrated that all six areas of prognosis are important and that their individual contribution during prognosis assessment may vary from case to case. There is a need for evidence-based prognostic decision-making as well as tools to assist physicians in searching for, appraising, and applying prognostic evidence to substantiate their prognosis assessments.