The Trade-off between Productivity and Quality of Care in Primary Care Services
Discussant: John Romley
In delivery systems with a fixed budget like the Veterans Health Administration (VHA), theory suggests a trade-off between productivity and quality of care, however the trade-off has not been quantified. At the VHA resources are scarce, demand is growing and, as a consequence, hospital managers and providers are pushed to be more productive and thereby potentially reduce the quality of care. Quantifying the trade-off between productivity and quality of care is difficult since they are simultaneously determined. This study uses market factors—such as the veteran unemployment rate, which influences the demand for VHA care—as an instrument to identify the effects of exogenous changes in quality of care.
One hundred and twenty-seven VHA facilities provide primary care services throughout the United States. The total number of visits per month per VHA facility between fiscal years 2013 and 2016 was used as an output measure in a production function controlling for quality of care, number of primary care provider FTEs, support staff, enrollee, and facility fixed effects.
Quality of care was measured using patient experience surveys administered to a random sample of VHA patients who had at least one primary care appointment. Patient responses to the question “Did your provider seem to know the important information about your medical history?” were collected on a 4-point Likert scale, which was dichotomized into 1 (i.e. “satisfied”) or 0 (i.e. “not satisfied”). In the 1st stage of an instrumental variables model, the probability of a patient being satisfied with their provider’s knowledge of their medical history was instrumented by the veteran unemployment rate. The 2nd stage measured the effect of facility average quality of care on facility average number of primary care visits.
The veteran unemployment rate was negatively associated with patient satisfaction with their provider with F of 10.6. A statistically significant trade-off between number of visits (i.e. productivity) and quality of care while controlling for staffing parameters at each VHA facility was observed. On average, every percentage point increase in patient satisfaction at a VHA facility would reduce primary care visits by 123 per month. The average number of primary care full-time equivalent physicians was 38 per facility, so a 1% increase in the quality of care measure was equivalent to reducing the length of an appointment by 16% (e.g. 5 minutes for a 30-minute appointment).
Hospitals are under pressure to do more with the same or less resources. This will come at a cost of decreasing the quality of the patient experience. Patient satisfaction measures are used as performance measures for hospitals and providers, and affect value-based payments. This is the first study to quantify the trade-off between productivity and quality of care so that hospitals and providers can predict the quality implications of higher productivity targets.