The Trade-off between Follow-ups and New Visits in a Capacity-Constrained Primary Care System

Background. In a capacity-constrained health care system, there is a trade-off between patients with chronic conditions who need regular follow-ups and patients demanding care for new complaints. We quantify this trade-off for patients receiving care from the Veterans Health Administration (VHA).

Methods. We use patient satisfaction data collected between 2014 and 2016 by the SHEP survey, a random sample of patients receiving care in the VHA. We restrict the data to patients with at least one in-person primary care visit who have provided information on their ability to access urgent and routine care. We match SHEP responses to the facility and month where the respondent’s last visit occurred. We calculate monthly measures that characterize facility-level service provision, including the average time between successive primary care visits for patients with multiple visits; the average primary care visit length at the facility; the percentage of primary care appointments at the facility that are overbooked; the percent of visits that are unscheduled (i.e., walk-ins), and a measure of empaneled patients relative to the facility’s provider FTEs. We estimate hierarchical linear regression models with 1-month lagged continuous facility-level predictors. We also estimate logistic models with the lagged measures as categorical predictors (i.e., quartiles, sextiles, and octiles) in order to assess the relationship across the distribution of the predictors. We also control for patient demographic characteristics and the length of the patient-provider relationship.

Results. We find that higher average follow-up times are associated with better access for patients needing urgent or routine care. Although the effects on the continuous predictors are moderate, we find stronger effects at the higher end of the follow-up time distribution. For example, receiving care at a facility in the highest octile of the distribution is associated with a 12% higher odds of reporting better access to both routine and urgent care. Lower average appointment lengths are associated with better access to urgent care, but not to routine care. In all models, the length of the patient-provider relationship was a strong predictor of patient satisfaction with access.

Conclusion. Primary care providers in the VHA have a relatively wide bandwidth for increasing time between visits, but facilities with unusually frequent follow-up visits fare worse in terms of access for new medical complaints.