Introduction

Chronic back pain (CBP) is a common condition as evidenced by literature estimating that approximately 20% of Europeans report chronic pain, with chronic back pain responsible for the majority of cases [1]. CBP decreases not only the quality of life for the individual, but is also a relevant societal challenge for the health care and economic system [1, 2]. Back pain is the major cause for loss in productivity [3] and sick leave [4], and is a major cause for disability pension [4–6] and presenteeism [3, 7]. In addition, mental health problems have been shown to impact the extent of sick leave taken by individuals [8]. Literature recognizes that depression is strongly associated with chronic pain [9], especially with increased prevalence of back pain. However, evidence is still conflicting with regard to the timing and causality of this relationship [10]. Moreover, both depression and other nonspecific and sub-acute psychological conditions such as psychological distress (PD) are associated with an increased prevalence of musculoskeletal problems. Dorner and colleagues found, for example, that social stress in the workplace more than doubled the probability for musculoskeletal problems in males, and sexual dissatisfaction was correlated with a three-fold increase in reporting of musculoskeletal pain [11]. PD, as it is defined in this publication, is a sub-acute condition which results from a persons´ non-ability to cope with personal, social, or occupational demands [8, 12].

It is well known that individuals with chronic pain make more contact points with the health care system [13, 14]. In contrast, literature suggests that persons with psychosocial conditions, such as partnership problems and social stress are less likely to visit the primary health care system for those psychosocial reasons, but more for physical complaints [15–17]. Primary health care is recognized as a focal point for customized and effective management of psychosocial conditions, especially those in combination with chronic pain [18–23]. In the United States, a country without a gatekeeping system like Austria, those with PD that have no primary care physician, or those who cannot afford one, are more likely to use emergency services for non-emergency reasons [24].

There is limited evidence for an additive effect of depression and pain on the utilization of specialist care. For example, one study in patients with spinal injuries in the United States showed that the additive effect of pain and depression were associated with a higher utilization of specialist care than the sum of both effects alone [25]. Similar observations were found in rural settings in Canada, where the presence of both chronic pain and depression led to a higher primary health care utilization [26]. In addition, a larger Swedish cohort study recently showed a synergistic effect of back pain and common mental disorders that lead towards disability pension [27]. However, little is known about CBP and PD on the utilization of the different levels of the ambulatory health care system in a European country without a gatekeeping system, and even less about the possibilities of a synergistic effect.

In this study, the term primary care is used to reference ambulatory general practitioner (GP) offices. Meanwhile, the secondary level of care includes those that are composed of specialist practitioners such as ophthalmologist, dermatologist, orthopaedist, internal medicine specialist, ENT specialist, gynaecologist, or urologist. The reason we were interested in both levels of ambulatory care was that in Austria, with some exceptions, patients have free access and financial coverage to both the primary and secondary level of care. Patients make individual determinations as to which level of care she or he consults [28]. Therefore, Austria is an ideal setting to observe a possible synergistic effect of PD and CBP on the utilization of different levels of care in the ambulatory sector.

It was the aim of this analysis to assess the impact of CBP and PD on the utilization of the primary and secondary levels of care in the ambulatory health care sector. Additionally, the intent was to evaluate if the joint effect of CBP and PD is higher than the impact of the sum of the two single conditions.

Materials and Methods

Results

Sample

Data from 8,021 women (51.8%) and 7,453 men (48.2%) were analyzed. Of all women surveyed, 40.8% (n = 3,275) were considered as having PD and 34.5% (n = 2,766) responded as having CBP. In contrast, 44.4% (n = 3,311) of all males were considered as having PD and 30.5% (n = 2,269) experienced CBP.

Table 1 shows the distribution of the demographic and health variables for the whole sample.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of respondents (n = 15474), values are numbers (%) unless stated otherwise.

https://doi.org/10.1371/journal.pone.0134136.t001

Impact of PD and CBP on GP and specialist visits

Table 2 shows the different models for the impact of PD and CBP on the utilization of the primary and secondary level of ambulatory care. Both PD and CBP showed a significant positive association with GP visits and specialist consultations, even after the full adjustment in model IV. This means that, after the full adjustment for other chronic diseases and socio-demographic variables, the probability of persons with both CBP and PD, respectively, consulting a specialist was increased significantly, with an OR of 1.68. Further, the probability of a person with only CBP consulting a specialist showed an OR of 1.19, compared to subjects without CBP and without PD (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. The different models for the impact of PD and CBP on the utilization of GP and specialist visits.

https://doi.org/10.1371/journal.pone.0134136.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Synergistic effects for the four different CBP/PD groups for GP visits.

https://doi.org/10.1371/journal.pone.0134136.t003

Synergistic effects for PD and CBP on GP and specialist visits

Tables 3 and 4 show that the respondents that reported to have had contact with GPs or specialists were nearly twice as high in the simultaneous PD and CBP group than in the group without PD or CBP. Additionally, they describe two findings each: First that the likelihood of consulting a physician is significantly higher for individuals with both conditions compared to individuals without these conditions, for both the primary and the secondary levels of care. Secondly, statistically significant synergistic effects (S) were found for the group of subjects with PD and CBP related to the primary, as well as the secondary level of ambulatory care (Tables 3 and 4). However, for the primary level of care the synergistic effect was found to be significant until the full adjustment. The synergistic effect for PD and CBP for visiting a specialist was nearly twice as high as it would have been expected on the basis of the mere addition of the effects of the single PD and CBP variables. Further, the PAF shows that 20% of the specialist consultations can be explained by the synergistic effect.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Synergistic effects for the four different CBP/PD groups for specialist visits.

https://doi.org/10.1371/journal.pone.0134136.t004

Discussion

These findings represent the first analysis of associations and synergistic effects of PD and CBP on the utilization of both the primary and secondary level of ambulatory care in a country without a gatekeeping system.

The finding that both single factors had a positive association with the utilization of the health care sector would be expected. However, it is interesting that a nonspecific emotional factor, such as PD, significantly increased the likelihood of consulting the secondary level of care in a country without gatekeeping, even after the full adjustment with an OR of 1.24 (Table 2).

The synergistic effect revealed that if both factors of PD and CBP occur together, the need of patients for care at all levels was much higher than the mere addiction of the effect of the individual factors. This was nearly double the effect that the addition of the single factors displayed, with a synergistic index of 1.16 for the primary level of care and 1.99 for the secondary level of care (Tables 3 and 4). Particularly for the secondary level of care, this finding is remarkable; indicating that if CBP occurs together with an unspecific psychological condition, about 20% of those specialist consultations could be explained by the interaction effect of CBP and PD alone. In addition, both PD and CBP showed a significant positive association with GP visits and specialist consultations. This means that CBP and an unspecific condition such as PD increased the probability of presenting to a GP or specialist significantly (Table 2).

After adjustment for other chronic diseases, the significance of the synergistic effect vanished for the primary level of care. This suggests that a large proportion of primary health care utilization in these patients was probably mediated alongside other chronic diseases. CBP and PD patients very often had other chronic conditions, especially, those with both conditions simultaneously, as well as those patients with CBP only.

Our findings implicate consequences for transforming the primary level of care. The synergistic effect should be seen as a special window of opportunity, due to the fact that the likelihood to visit a GP with these conditions is very high (Tables 2 and 3). Recent findings suggest that PD as a non-specific psychological condition, and CBP as a chronic condition, could be best managed at the primary level of care. This is due to the fact that primary health care is centered around core competences of person-centered care, comprehensiveness, continuity, and coordination of care, which could provide an ideal framework for addressing these bio-psycho-social problems and better coordinate the complex care strategy needed [38–40]. Stein and colleagues found that the most important element for care of chronic pain patients in Austria is a comprehensive discussion with their physician; which was observed as more important than a reduction in pain intensity. More efforts have to be made to educate and inform chronic pain patients adequately from the doctor's side considering their psycho-social context [41]. One other important issue seems to be to ask and apply relevant diagnostics and evaluation for co-morbidities. To be able to make use of this window of opportunity the primary level of health care should be organized and supported to strongly enable the application of the core competencies of primary care. Unfortunately, this is not completely the case in the health care delivery system in Austria, especially with regard to the core competencies for coordination of care and comprehensiveness [28, 36, 40, 42].

Our results could implicate additional consequences for the secondary level of ambulatory care. It seems that in a country without a gatekeeping system, where the access to the secondary level of care is free, persons with CBP and/or PD have a higher probability to present to the secondary level of care (Tables 2 and 4). This raises several questions, including as to whether the secondary level of care is the best point of care for persons with complex psychological and chronic conditions as well as multi-morbidities [43, 44]? Further, what is the underlying reason for consultation of specialists? Is it an integrated care concept, or just the perception of the patients that believe that the secondary level of care is a more sophisticated level of care? Unfortunately, the study design and questionnaire does not allow such analyses. However, it could be speculated that at least some part of the consultations at specialists were due these patient perceptions, as Austria is a secondary care focused country, and the overall utilization of the second level of health care is quite high [36, 45]. In this case it could be of benefit for patients to be guided through the complex health care system with the support of a primary care provider, preferentially through structural incentives or a gatekeeping system.

The strengths of the present analysis were the analytical design and the large sample size, which allowed adjustment for possible confounders. Additionally, the large number and random selection of participants offers a high external validity of the results, and in particular for Austria. The utilization of a consistent survey-interview-team and a standardized questionnaire increased the probability for data consistency. One major limitation is the challenge of diagnosing PD: for measuring psychological distress we used the “general-mental-health”-questions of the Short-Form-36 Health Survey which covers psychological distress. However, this tool cannot distinguish from more severe or acute forms of distress or psychopathologies or from a mixture of somatic and psychological symptoms. It would have been of benefit to additionally use other tools like the Four-Dimensional-Symptom Questionnaire (4DSQ) to be able to better distinguish the more severe and acute forms of mental health disorders in distressed GP patients as recommended by Terluin et al. from the sub-acute psychologically distressed patients [46]. This challenge could have led to an overestimation of persons with sub-acute PD. Another important limitation which could have influenced the results is that different time frames were used throughout the given questionnaire design; with a 12 month cutoff assessed for the existence of CBP and a four week one for PD in conjunction with utilization of the two levels of the ambulatory health care system. This cutoff may have underestimated the number of patients seeking care for PD, or otherwise captured a larger number of patients that sought care for CBP, in comparison. On the other hand, PD has the same timeframe as the health care utilization question and CBP is by definition chronic which implies a longer period than 4 weeks to determine the presence meaning the influence of the results may be a limited one. Moreover, the analysis is cross-sectional and, therefore, no causal relationships can be drawn. Additionally, results are based on descriptive and self-reported survey data rather than administrative data [47]. Furthermore, the cut-off value for the two PD groups was based on the median, as no normative values for psychological distress could be identified in the literature. Finally, and very important, we had no knowledge about the reasons for and appropriateness of the consultations.

Conclusion

Synergistic effects and associations for CBP and PD towards the utilization of both the primary and secondary level of the ambulatory health care sector were identified. In addition, persons with CBP and/or PD had more co-morbidities than persons without CBP and PD. Our results could yield consequences for the organization of the health care system. Particularly for primary level of care, these results could be seen as a window of opportunity, where the complex care strategy needed for psychosocial as well as chronic diseases and multi-morbidities could be emphasized.

This is particularly important, as patients have the freedom to access either level of care based on their individual preference. The results have profound implications on steering health system organization, where priorities can be made to improve avenues for patients to access care at the best point of entry into the system. Specifically, secondary levels of ambulatory care are often more costly, and may not lead to positive differences in outcomes. Counter, it has been observed that, for patients with concomitant PD and CBP, that the best point of care is most often a primary care setting.

To better direct patients towards the most appropriate health services, it would be valuable to avoid indiscriminate self-referral to secondary levels of care. To meet these goals, efforts must additionally be made to strengthen the quality of care at the primary level of care as well as align to meet recent recommendations. The findings of this study support recommendations for health care governance and health systems to improve coordination of person-centered care in Austria and similar countries.

Supporting Information

Acknowledgments

We would like to thank Statistics Austria and the Austrian Ministry of Health for giving us the opportunity to analyze the data.