Evaluating provider-and patient-related drivers of why care is not concordant with guidelines for patients with multiple chronic conditions

Introduction: Patient centred care is often at odds with disease-focused evidence-based clinical practice, particularly for patients with multiple chronic conditions (CCs).

To assess the extent to which care for patients with multimorbidity was not guideline-concordant and the reasons why not, examining reasons stratified by disease.

Methods: Retrospective cohort study conducted within the context of a nurse-primary care physician team-based care management program, the Comprehensive Care for Multimorbid Adults Project (CC-MAP) across 12 primary care clinics in Israel.

This study included CC-MAP patients aged 45+ years who had at least two of eight common CCs. Forty-four guideline-based care processes for these eight CCs and dyslipidemia (‘focus conditions’) were evaluated over one year prior to data collection, through EHR review, including: lifestyle modifications, diagnostic and follow-up tests, medications, non-medication treatments, and specialist referrals. When care was not guideline-concordant, reasons why not were examined by: biomedical, personal and care burden, and system related factors.

Proportions of non-guideline-concordant care and related reasons were assessed overall and stratified by focus condition. Reasons were also stratified: provider-driven if the clinician didn’t refer care, and patient-driven if care was referred but not followed by the patient.

Results: 4515 care processes were evaluated among 204 multimorbid patients, on average aged 72.3 years (SD 9.7) and with 3.8 (SD 1.1) CCs. Overall, 20.8% of care was not guideline-concordant, and when examining disease-specific care for the focus conditions, ranged from 14.7% (for IHD) to 48.9% (for depression).

Of 1071 reasons why care deviated from guidelines, 59.8% were provider-driven and 40.2% patient-driven. Thirty-five percent of all reasons given were biomedical-related, 28.5% personal and care burden-related, 13.3% system-related, 11.9% other and 11.9% unknown. A higher proportion of reasons were provider-driven than patient-driven for atrial fibrillation, congestive heart failure, diabetes, and hypertension specific care; whereas, a higher proportion of reasons were patient-driven than provider-driven for chronic kidney disease, chronic obstructive pulmonary disease, depression, and dyslipidemia. For ischemic heart disease, the provider- and patient-driven reasons were fairly evenly divided.

Discussion: Patients with multimorbidity do not receive about a fifth of guideline recommended care overall, but this proportion varies substantially across diseases. More than a third of non-guideline-concordant care is driven by biomedical reasons, and about a third attributed to patient-related reasons.

Conclusions: Understanding the drivers to deviations from guidelines can inform improvements to guidelines and care delivery for patients with multimorbidity.

Lessons learned: These findings show that many clinical, personal patient, and system-related reasons exist for deviation from guidelines. Quantifying these types of reasons may provide a basis for tailoring assessments of the appropriateness and compatibility in combining disease-specific guidelines for multimorbid patients.

Limitations: An evaluation in a control group or in a usual care setting was not possible for this study and while the extent of guideline-concordance was measured, this was not linked to outcomes.

Suggestions for future research: Future research could examine these patterns of care in a usual care practice setting, as well as aim to evaluate the relationship of guideline-concordant care with clinical outcomes among multimorbid patients.