Patients with persistent medically unexplained physical symptoms: a descriptive study from Norwegian general practice

Norway has 5, 05 million inhabitants and 4 189 GPs. A list system includes 99.5% of the population and connects every Norwegian citizen to a specific GP. The GPs issue 80% of all sickness certificates [22]. Health care is mostly publicly financed, with a fixed amount of maximum personal charge per year for the patient. Vest-Agder county has approximately 180 000 inhabitants and 160 GPs located in 44 different practices. It has a mix of rural and urban areas, and covers a range of socioeconomic status among the inhabitants.

For the purpose of this study, we defined persistent MUPS as “physical symptoms with no identified organic cause, lasting for at least three months and leading to a loss of function”. Loss of function was explained as sickness absence or disability, or withdrawal from social activities like sports, social events or other leisure activities. In case of co-morbid diseases, the unexplained symptoms should outweigh function loss caused by medically explained symptoms. Our case definition was carefully presented in the invitation letter, e-mails and personal meetings with the GPs. Patients ≥ 18 years were included.

The GPs were asked to register data from every eligible patient on a one-page form with simple tick-off questions during a four-week period. The registration form, which was completed in less than three minutes (see Additional file 1), included relevant demographic information, symptom pattern (duration and localization), work participation and the GPs management strategy in the actual consultation.

The data were analyzed using IBM SPSS version 20 for Windows with descriptive statistics for all variables and split analysis on gender and age. We used independent-samples-t-tests for comparing means on continuous variables for two different groups of participants, and Chi-square test for independence to explore the relationship between categorical variables. Possible relationship between number of registered patients and different GP-characteristics were investigated using simple bivariate correlation. The level of statistical significance was set to p < 0.05.

The response rate was 53% (84/160). The invited GPs were representative for Norwegian GPs, although with an excess of men and GP specialists (Table 1). Participating male GPs were significantly older (p = 0.011), and had worked more years in general practice than the females (p = 0.002) (Table 2). There was a variation in the number of recorded consultations with patients with persistent MUPS (MUPS consultations) per GP (range 0–26, mean = 6.38, SD = 4.97). No significant associations were found between number of recorded MUPS consultations and gender, age, specialist status, years in practice, or the total number of consultations during the registration period among the GPs (Table 2).

The GPs registrated 526 MUPS consultations out of a total of 17 688 consultations, which gave a consultation prevalence rate of 3%. Of these were 399 (75.9%) with female patients. The educational level among the patients did not differ from the distribution in the population in the county as a whole [23]. 234 (44.7%) of the patients were working, significantly more men (p = 0.047) (Table 3). Among the 289 not working, 91 (31.5%) were on sick leave, 119 (41.2%) had a disability pension, and 76 (26.3%) had other social benefits, mostly work assessment allowance. There was no significant gender difference in patients’ age, education level or duration of symptoms (Table 4), or symptom pattern (Table 3). Musculoskeletal pain and asthenia/fatigue were predominant symptoms, with 68.1% and 57.0% of all patients respectively. Headache or dizziness was reported by 31.9% of the patients and gastrointestinal symptoms by 20.5%.

The majority of the GPs reported that they had offered supportive counseling in the actual MUPS consultations. Medication was prescribed in 21.6% of the consultations. About 20% of the consultations included lab tests or referrals. There were no significant gender differences in GPs management strategy (Table 3).

This study presents for the first time cross-sectional descriptive data on consultations with patients with persistent MUPS in Norwegian general practice. A consultation prevalence rate of 3% suggests that patients with persistent MUPS are common when compared to other reasons for encounter. Musculoskeletal symptoms were the most frequent symptom pattern, followed by asthenia/fatigue. Almost 45% of the patients were working. About 2/3 of the GPs offered supportive counseling at the actual consultation.

Our study is strengthened by the naturalistic setting and the rigorous case definition, which was carefully described, also through a clinical vignette during personal meetings. Our case definition resembles definitions used in previous research [12, 13, 15]. The prevalence in this study is on MUPS patients as perceived by Norwegian GPs and not a prevalence of MUPS in the general public, and due to the study design one patient may have been registered several times, which may influence the results. We believe however this not to be very likely and does not affect the prevalence rate, as this is still a prevalence of MUPS consultations in general practice.

Our design did not allow for inter-rater-reliability checks of the GPs’ assessments. As mean duration of a GP-patient relationship in Norway is 7.7 years [24], we may assume GPs to easily recognize eligible patients due to their knowledge about their patients. Data collection based on digital patient records or interviews may represent larger sources of bias, due to lack of a specific MUPS-diagnosis, incomplete medical records or interviewer’s lack of personal knowledge of the actual patients.

Sample bias was reduced through an overall representative sample of GPs-, indicating that the results may be generalised to a comparable setting. However, more participating male GPs may underestimate the consultation prevalence since female GPs see more female patients, and, due to gender effects on communication, they may see more patients with MUPS [25]. GP based prevalence estimates are found to be consequently underestimated [15, 26], but this may also indicate registration of symptoms of less clinical significance, which we did not intend to include in this study.

A response rate of 53% is modest, although higher than other surveys of GPs on MUPS [27, 28]. Furthermore, our sample seemed to be fairly representative for Norwegian GPs.

Possible source of information bias may be due to incomplete recordings due to GPs busy workday [29, 30] resulting in ignored eligible patients (consultation prevalence underestimated), or a non-compliant attitude to the actual instruction due to an eager to participate because of professional relationships with the first author (skewing possible for both directions). The GPs in the area received various reminders and information about the study. How this may have influenced the registration is unknown [31]. On the other hand, personal and extensive study information may have raised the actual quality of the collected data.

Only 32% of the patients in our sample were on sick leave. This finding seems to oppose previous research [32] by suggesting that previous assumptions of high level of disability among patients with persistent MUPS needs revision. However, as loss of function was a criterion for eligibility in the study, we can only assume that that these patients met this criterion otherwise than being out of work.

A consultation prevalence rate of 3% suggests that patients with persistent MUPS are common in general practice when compared to other frequently occurring conditions like hypertension, atrial fibrillation, diabetes mellitus, and depression, each accounting for 1.3-4.9% of the consultations in Norwegian general practice [33]. A recent retrospective study from primary and secondary care in US [12] presented exactly the same consultation prevalence for persistent MUPS by use of a similar case definition, and a Dutch study from primary care found a prevalence of 2.45% on persistent MUPS [11].

More women (76%) than men in our sample suffered from MUPS, consistent with prior studies [34]. However, we should keep in mind that doctors tend to categorize female patients more easily as somatizers than male ones [35] and that this may be the same for patients with MUPS as we do not see persistent MUPS as synonymous with ‘severe somatization disorder’ [36].

In contrast to findings from a similar study we found linearity between number of patients with persistent MUPS and total number of consultations at the individual GPs [12]. Musculoskeletal symptoms were the most frequent MUPS symptom pattern in our study, followed by asthenia/fatigue. These findings diverge from a recent study where asthenia/fatigue was leading among patients with persistent MUPS [34]. However use of different case definitions may contribute to this difference, and with knowledge of substantial co-occurrence for these symptoms, this finding may be of less clinical importance.

An important overall finding challenges the stereotypes of patients with persistent MUPS. In our study, this group of patients is heterogenous - a quarter are men, a quarter has higher education, a third is below age 40, and nearly half of them work. This highlights the need to refrain from unsubstantiated stereotyping of “MUPS-patients”.

Our finding of a scarce prescription rate complies with recommendations that medication should play a minor role in medical management of MUPS [37, 38]. Our case definition implied prior appropriate diagnostic evaluations, but still we found that one out of five patients received further referrals. This may illustrate GPs feeling of shortcoming in managing these patients, and a lack of knowledge of what these patients’ actual want and need [39, 40]. Furthermore, it may illustrate a limited awareness among the GPs about potentially iatrogenic harms trough additional interventions [41]. However, actual lab test and referrals in the registered MUPS consultations may represent adequate management of other co-existing diseases or presentation of new symptoms, which may represent new disease.

Levels of supportive counseling at the GP encounter will probably vary according to structure of health care system. In a study from US, 37% were referred to different mental health care center [12], reflecting possible differences in therapeutic cultures. It seems like Norwegians GPs more often provide supportive care themselves rather referring to secondary care.