Patient Complexity, Prescriber Continuity and Acute Care

Author’s Reply—We thank Drs. Yunyongying and Gove for their interest in our article on prescriber continuity and health care utilization.¹ We agree with the authors that one should not infer that having more prescribers of cardiometabolic medications causes increased emergency room visits and hospital admissions. We were careful to use the term “association” throughout the paper and specifically noted in the limitations that the “observational study design used in our study introduces the possibility of unobserved confounding, which precludes causal inference…”

We also agree that not all providers and not all visits are equal because some have a greater impact on patient health than others. In future work, we intend to refine this claims-based measure of Continuity of Medication Management (via counts of cardiometabolic prescribers) to try to account for this fact.

Although we agree with the authors on these points, we disagree that disease complexity was not addressed. We adjusted for patient-level disease complexity via a validated set of chronic condition indicators from the Gagne score² and the number of cardiometabolic medications. We also examined cardiometabolic-specific utilization in addition to all-cause utilization, because these outcomes might more plausibly be related to coordination of care among cardiometabolic prescribers.

Finally, the criticism that we failed to examine potentially more predictive provider-level factors, such as extent of care coordination between providers, is unreasonable for claims-based analyses, because there are no validated measures of this construct. We agree that this construct may be critical to understand the quality of care coordination interventions, but its measurement may require qualitative assessment or survey measurement. Construction of a claims-based measure of care coordination is not currently feasible.

It is important to determine how to provide optimal care to patients, particularly those with multiple chronic conditions. Our manuscript represents the start of this challenging research agenda by presenting a “proof of concept” analysis, which suggested that having multiple prescribers is associated with risk for avoidable health care utilization. We aim to conduct additional research to determine how best to measure the construct of continuity of medication management, to understand its comparative utility over traditional continuity of care measures, and to further validate these initial findings.