Does continuity matter? Measuring the quality of ambulatory care among chronically ill elders in the United States

Objective. To examine how continuity is related to quality measures for chronically ill elderly patients; to the specialty and caseload of the physicians they predominantly visit; and to the regional context of where their ambulatory care is delivered.;Methods. Beneficiaries from the Medicare Part B 20 percent sample with congestive heart failure, chronic obstructive pulmonary disease, diabetes, or hypertension were identified in 2005. The number of visits made and physicians seen over the course of the following year were used to calculate each patient's continuity of care (COC) score and to assign each patient to the physician whom he or she predominantly visited. Documentation of recommended services and the occurrence of hospital utilization were determined from claims data, as were physician specialty and an estimate of the proportion of a physician's assigned patients---or primary patients---relative to his or her Medicare caseload. Each patient was also matched via residential ZIP Code to a hospital referral region (HRR) characterized by its per-beneficiary total Medicare spending.;Results. Eligible patients with higher COC scores were, at best, no more likely to receive recommended services. Patients with higher COC scores were generally at lower risk of preventable hospitalization, readmission within 14 and 30 days of discharge from a preventable or other index hospitalization, and hospitalization within 14 days of discharge from a post-acute care facility and had a reduced rate of care transitions and emergency department visits after adjusting for patient characteristics as well as local hospital bed supply and regional managed care penetration. COC scores tended to be highest among patients assigned to primary care physicians (PCPs) who mostly saw primary patients. In tests for trend, higher HRR spending was negatively associated with the COC score after controlling for patient characteristics, although assignment to a PCP was associated with higher continuity across regions.;Conclusions. Higher continuity of care among the chronically ill elderly is not associated with the receipt of recommended services but does decrease the risk of hospital utilization. Chronically ill elders who predominantly visit PCPs who mostly see primary patients experience higher continuity, as do those who receive care in lower-spending regions.