Patient-reported quality of outpatient healthcare in patients with chronic back or arthrosis pain with long-term opioid therapy in Germany | BMC Primary Care

The cross-sectional survey cohort comprised 661 individuals insured by “DAK-Gesundheit” with long-term opioid therapy and chronic back and/or arthrosis pain. The aim of this analysis was to analyze quality of outpatient healthcare according to the CCM from the patient’s perspective. Survey participants rated the quality of healthcare according to the CCM rather low. Patient characteristics showed minimal to no influence on PACIC-5A scores, while patient’s health situation demonstrated only partial correlations, including a negative correlation with psychological distress. Conversely, there was a consistently positive correlation observed with the delivery of aspects of pain treatment, such as the setting of therapy goals and a comprehensive treatment concept. Selection criteria resulted in 0.7% (first selection) and 0.6% (second selection) of insured adults receiving long-term opioid therapy and chronic back and/or arthrosis pain.

PACIC-5A

As initially assumed, the cohort under consideration predominantly (85%) exhibited pain related high disability, moderately to severely limiting according to the GCPS. Therefore, the utilization of the PACIC-5A questionnaire in this cohort of highly burdened chronically ill individuals can be deemed appropriate. The psychometric quality of the PACIC-5A assessment, as measured by Cronbach’s α, demonstrated acceptable to good internal consistency, with the exception of the “advise” subscale, which narrowly missed the threshold of 0.7 for acceptable internal consistency (Cronbach’s α = 0.66).

In terms of the overall evaluation of healthcare quality according to the CCM, the mean subscale values were lowest for the subscales on follow-up of treatment plans “arrange” (1.90) and on aid for patients in achieving set goals “assist” (1.91). Furthermore, pronounced floor effects for these two subscales suggest that these aspects of healthcare according to the CCM were entirely absent for almost a quarter of respondents. The two aforementioned aspects represent crucial pillars of healthcare according to the CCM, thus offering significant potential for enhancing the quality of healthcare for chronic pain patients with long-term opioid therapy. Though, these results need to be interpreted in regards of the German healthcare system focusing on medical treatment. Therefore, preventive and activating aspects outlined in the CCM might not be met by health services due to personnel, organizational, and financial constraints. In these terms low sores in PACIC-5A subscales need to be addressed by health policy especially since they cover aspects recommended by the LONTS guideline [6]. The improvement of the “assist” aspect could involve encouraging of patients to join (self-help) groups or facilities for their illness. To enhance the rating of the “arrange” subscale, regular provision of health advice and nutritional counseling could be implemented. Patients could also be encouraged to participate in public programs and contacted more frequently between appointments to check on their condition. Additionally, they could be educated on how to seek support from family and friends. Furthermore, while the “advise” (2.42), “assess” (2.67), and “agree” (2.75) subscales were rated higher in mean values within a possible range of PACIC-5A rating from 1 to 5, there is still the possibility of improvement toward optimal healthcare according to the CCM. Specifically, the “advise” subscale could be improved by providing patients with better information on the support of their treatment by other specialist physician visits.

To our knowledge, this is the first time that the PACIC-5A survey has been conducted in a German cohort of patients experiencing chronic back and/or arthrosis pain and receiving long-term opioid therapy. Several studies examined the quality of outpatient healthcare according to the CCM using the PACIC-5A in patients with different medical indications in Germany. These studies mirrored our survey’s findings, with the “assist” and “arrange” scales consistently showing the lowest values [27,28,29,30]. A survey conducted on patients after an acute inpatient stay due to colorectal carcinoma (n = 816) evaluated similar values, with the present survey rating the subscale “advise“ 0.3 points lower and “assess“ 0.3 points higher [27]. Patients surveyed on the quality of outpatient healthcare in accordance with the CCM after an acute inpatient stay for mental disorders (n = 2,289), showed an average of 0.4 points higher result across all scores. Individual assistance (“assist”) in particular deviated downward by -0.7 points in the present cohort [28], which is one of the lowest-rated subscales in the current study. A PACIC-5A survey among persons with osteoarthritis (n = 1021) also showed higher scores across all (sub)scales of the PACIC-5A compared to the present analysis [29]. A further survey in a cohort of individuals with obesity (n = 117) described scores of comparable magnitude to the present survey [30].

Patient characteristics

We assumed that patient characteristics could possibly influence healthcare quality as specified by the CCM. However, our results indicate minimal to no correlations. A younger age and a higher educational level showed weak positive, partly statistically significant correlation with healthcare quality ratings according to the CCM. This trend regarding age [28, 29] and educational level [29] also appeared in literature. The independence of the PACIC-5A scores from sex in terms of statistical significance was also shown in other surveys [21, 28, 29]. As it pertains to aspects of equity in health, patient characteristics should not influence the quality of healthcare according to the CCM [31]. For migration background, pain diagnosis, and sex, no deviation from this normative principle could be shown in our analysis. However, older and less educated individuals appear to exhibit a slight tendency towards poorer-rated healthcare quality in terms of the CCM. Patient-centered healthcare, with the integration of shared decision-making, is to be understood as a process that requires productive communication between physician and patient. This communication is influenced by patient preferences and physicians’ competencies to communicate in a manner appropriate to the individual target group. Furthermore, even if we assumed that patient characteristics act as independent variables, it cannot be excluded that patients valued treatment differently based on their characteristics. Future research should investigate the extent to which certain aspects of healthcare according to the CCM are not offered, for example, to older and less educated individuals, or are not preferred by them. Recommendations for physician-patient communication tailored to subgroups should be developed.

Patient’s health situation

The second group of variables, patient’s health situation, assumed to be influenced by healthcare quality according to the CCM, correlated only partially. A higher degree of anxiety and depression symptoms, as well as a higher degree of opioid addiction problems, tended to correlate slightly with a poorer quality of healthcare as specified by the CCM. Similarly, another evaluation of the PACIC-5A score found an association between higher depressive severity levels on the PHQ-9 instrument and lower PACIC-5A scores [29]. This trend may indicate a protective effect of good quality healthcare according to CCM on risks commonly associated with chronic pain or opioid treatment, such as psychological distress, particularly depression [32], or opioid addiction [6]. It should be noted that the observed effects are small, and causal relationship cannot be verified by this cross-sectional survey. Furthermore, the GCPS [24] is a crucial variable for assessing the treatment success of long-term opioid therapy. The corresponding subgroup analysis revealed only one statistically significant correlation between healthcare quality according to the CCM and the insured individuals’ pain experience. The lack of a clear trend might be attributed to many possible causes: results of the GCPS show strong ceiling effects as across the 5 levels of the GCPS, 85% are distributed to the two highest classifications. This indicates long-term opioid therapy to be prescribed to a highly burdened patient group suffering from pain for a long time. Therefore, a more sensitive instrument than the GCPS would be desirable. Lastly, the causal relationship of a change in healthcare quality as specified by the CCM on the patient’s health situation cannot be demonstrated in a cross-sectional design. Therefore, future longitudinal studies should investigate the influence of the quality of healthcare according to the CCM, particularly on pain intensity and pain-related impairment, in greater detail.

Pain treatment aspects

The last category of subgroup analysis, pain treatment aspects, revealed the strongest correlations with the assessment of healthcare quality according to the CCM, particularly in the domain of guideline-compliant [6] pain treatment aspects, namely therapy goals and treatment concept. Although there was partial content overlap between PACIC-5A subscores and these variables, the non-overlapping subscales also exhibiting the effect in the same direction. Given that only a third of respondents in our analysis reported a comprehensive treatment concept, this area appears to offer great potential for enhancing the quality of healthcare as specified by the CCM. It should be noted that the subjective perception of patients to the setting of therapy goals and the development of a comprehensive therapy concept may differ from the objective implementation in healthcare. However, since concepts need to be remembered by patients to facilitate a behavioral change, this subjectivity does not appear to be disadvantageous. Additionally, the implementation of different procedures of an interdisciplinary pain therapy, such as psychotherapy, remedies or complementary procedures, which can form part of a comprehensive treatment concept, also demonstrated a positive correlation with PACIC-5A scores. Hence, health participation in diverse procedures is associated with a higher healthcare quality according to the CCM. Moreover, the slightly higher rating of the healthcare quality as specified by the CCM among insured individuals who received outpatient pain therapy treatment could be attributed to the specialized expertise of the providers [33]. Specialized pain physicians in Germany can bill more time-intensive consultations per patient through SHI, potentially resulting in increased resources for healthcare in accordance with the CCM. It can be concluded that treatment according to the guideline [6] recommendations constitutes an important contribution to the quality assurance of outpatient pain management according to the CCM. Following the assumption of a bidirectional relationship, the guideline [6] recommendations seem to align with the CCM, and aspects of treating chronically ill persons seem to be addressed correctly in the guideline [6]. Therefore, it seems sensible to promote guideline [6] adherence, such as increasing physician awareness of the guideline [6] and improving communication with patients about the importance of multimodal therapy. Pain therapists also appear to play an important role in delivering high-quality treatment according to the CCM. However, there still appears to be potential for improvement in the quality of healthcare as specified by the CCM provided by pain therapists.

Chronic care model in Germany

Developed as an ideal healthcare model for chronically ill patients, the CCM aims to promote behavioral changes and optimize disease management. However, implementing the CCM in the German healthcare system is challenging and seems more feasible in settings with sufficient resources, such as specialized pain therapy. The root of the problem is not just with individual physicians but also at multiple levels, such as the current healthcare system structure, where time resources for each patient are often severely limited. Additionally, some of the aspects gathered by the PACIC-5A questionnaire are not covered by the SHI system and, as such, are ineligible for reimbursement. This might be the reason, why our results seem to be in line with other German studies on other indications. Nevertheless, guidelines recommend this several aspects of the CCM perceiving pain in regards of the biopsychosocial disease model [6].

External validity

The survey is representative for the German SHI “DAK-Gesundheit”. Compared to the entire SHI population in Germany (which covers 88% of the population [34]), “DAK-Gesundheit” has a slightly older population with a higher proportion of women. Since the benefit package, contractual relations between health insurers and providers and other legal requirements for all insured persons in the SHI population are essentially the same, it can be assumed that the results can be extrapolated to the entire population of insured persons. Due to the slight shift in the structure of the insured, it can be presumed that the rating of the PACIC-5A might be slightly better in the entire SHI population, as the female gender and a higher age tended to be associated with poorer ratings of the PACIC-5A.

Limitations

The current analysis offers valuable insights into outpatient healthcare quality according to the CCM for patients with chronic back and/or arthrosis pain who received long-term opioid therapy. However, some limitations must be considered. The bivariate analysis only establishes correlation, not causality, and potential possible confounding variables are not accounted for. Additionally, the assumed direction of effects cannot be definitively determined. The cross-sectional design precludes the demonstration of long-term effects of the level of healthcare quality as specified by the CCM, such as on the GCPS. Individuals with the same quality of healthcare may have responded differently based on their level of education. Conversely, it is possible that patients are treated differently by physicians depending on their education level, or that patients themselves approach medical treatment with varying expectations. Moreover, while the ultra-brief PHQ-4 questionnaire, consisting of 4 items, is a validated instrument [17], some loss of information in measuring psychological distress is conceivable compared to the more detailed PHQ-9 questionnaire. Additionally, the survey results may be affected by recall bias due to the survey period. Though most of the question addressed health services during the last 3 to 12 months. Furthermore, it remains unknown whether all respondents fully understood the questions. Finally, a net response rate of 21.8% suggests the possibility of a non-response bias; however, non-response analysis revealed a comparable age and gender distribution.