The CCM has been promoted as a template of care for the chronically ill, aiming to substantially improve QoL [18
]. Our study showed that certain patients rated aspects of their care that were consistent with the CCM more favourably. Being younger and less depressed increased the chance of a higher score on the PACIC for both COPD and CVD patient groups.
Evidence that interventions containing at least one CCM element could improve clinical outcomes as well as patient-relevant outcomes exists [20
]. Adams et al.
reported in a recent review that COPD patients who received interventions with two or more CCM components had lower rates of hospitalisations and emergency/unscheduled visits, and shorter hospital stays compared with control groups [23
]. The studies, however, were conducted in hospital settings or HMOs and cannot easily be transferred to primary care settings. So far, only one study is available, showing that CCM elements can be implemented in small independent practices and result in improved care for diabetics [24
]. The PACIC scores in our study were similar to a study in a German primary care setting for patients with osteoarthritis (men
]. The Dutch and German primary care setting PACIC scores were substantially lower than those of Glasgow et al.
whose data was from an HMO setting [8
In line with the findings of Rosemann and colleagues [10
] PACIC scores were not correlated with disease severity in the multivariate analyses. MCS and PCS, reflecting different aspects of QoL of CVD and COPD patients, did not predict PACIC scores, suggesting that care delivered to chronically ill patients is not dominated by the severity of the chronic condition itself. Our results show COPD patients report care more congruent with the CCM compared to CVD patients. This may be explained by the stage of chronic care in the Dutch primary care setting: the COPD care standard (based on the CCM) was implemented in early 2010, while the recently-developed care standard for CVD patients has not yet been implemented in every health care practice. In addition, the CVD population included at-risk patients as well as patients with established disease. The at-risk patients may have had fewer interactions with their care teams and the teams may have put less effort into chronic care, which may also explain the lower average PACIC scores for the CVD population.
While Glasgow et al.
could not reveal significant differences in the PACIC scores regarding patient characteristics in the HMO setting in the US [8
], Rosemann et al.
identified significant differences based on age, education, and depressive symptoms in the primary care setting in Europe [10
]. We also found that younger and less depressed patients reported higher PACIC scores, indicating that their care better aligns with the CCM. Unlike Rosemann et al.
we did not find a significant relationship between education and PACIC scores. This may be explained by disease duration. Patients in the Rosemann study had had osteoarthritis for about 14
years; most of our patients had been recently diagnosed. Different levels of education are most likely to result in differences in coping with a chronic condition over time. Educated people are expected to be better at self-management, getting necessary care, and compliance [10
]. We thus expect to find significant relationships between education and PACIC score over time. The finding that younger, less depressed patients are more likely to report high PACIC scores could reflect differences in physician behaviour towards different patient groups and that such patients more actively seek CCM-compliant care, but the association is non-conclusive. The information is in any case valuable, since it suggests that ensuring that all patient groups benefit to the same extent from advances in chronic illness care is important in implementing CCM.
Our study is not without limitations. Most importantly, the data collected were cross-sectional and causal relationships could not be inferred. Depressive symptomatology may lead to a more negative appraisal of chronic care delivery, however, if patients receive high-quality chronic care this may also lead to less depressive symptoms among chronically-ill patients. Longitudinal data is necessary to disentangle the dynamic relationship between depressive symptomatology and high-quality chronic care delivery. We also expect to find a dynamic relationship between QoL and chronic care delivery. Since we included patients recently enrolled in newly-implemented DMPs we investigated the influence of QoL on patient’s assessment of chronic illness care delivery. There is, however, also evidence that higher levels of chronic care delivery results in improved QoL [18
]. Again, longitudinal data is necessary to disentangle the dynamic relationship between QoL and chronic care delivery. Finally, our sample of CVD and COPD patients limits generalizing study findings to other diseases. Our findings do, however, confirm those of Rosemann et al.
among patients with osteoarthritis [10
]. The strength of our study is its reasonably large and representative sample of primary care practices.
Based upon the work of Glasgow and colleagues [8
] there is adequate evidence to support the use of the survey to measure the CCM. However, further development and refinement of its psychometric properties is needed and some studies point to possible limitations of the PACIC instrument. We used the PACIC as a reflective measure to assess patients’ assessment of chronic care delivery. In accordance with the findings of Glasgow and colleagues [8
] who developed the instrument, we assumed the PACIC to reflect the underlying construct of chronic care delivery. Spicer and colleagues [25
], however, argue that the PACIC is a formative measure and scores on the items cause or form the respondent’s status with respect to the construct. Following their reasoning this may indicate that the level of chronic care delivery also emerges or is formed as a result of patients’ responses. Gugiu and colleagues [26
] argued that the PACIC is actually unidimensional instead of the subscale construct and scoring of the PACIC should be changed to an 11-point scale ranging from 0% to 100% by units of 10% instead of using a 5-point response scale. More research is necessary focusing on the instrument’s validity and reliability.