Reducing COPD Emergency Department Visits in a Small Clinic
You are the medical director of a three-physician primary care clinic serving 4,200 adults. Data from the local health information exchange show that 38 patients with COPD had 61 emergency department visits in the past 12 months; 19 visits were for COPD exacerbations, and 11 of those patients had no follow-up appointment within 14 days of discharge. The clinic has no embedded care manager, one part-time respiratory therapist, and a budget of $8,000 for the next six months. The largest insurer offers a shared-savings contract that pays a bonus if COPD ED visits fall by 20% within one year. Staff are already busy, and two physicians are skeptical that the clinic can influence ED use. You have access to the EHR, a patient portal, and a list of the 38 patients.
1. Design a focused quality improvement intervention using a PDSA cycle. What will you measure, and how will you address the skeptics and budget?
Read the worked answer
I would start with a small, testable change rather than a clinic-wide overhaul. The aim: within six months, increase 7-day follow-up after a COPD emergency department visit from the current baseline to at least 70% for the 38 identified patients. The first PDSA cycle would test a 48-hour post-discharge telephone call plus a scheduled 7-day visit for five patients. The part-time respiratory therapist would deliver inhaler technique teaching and a written COPD action plan; the front desk would flag emergency department discharge notifications; the patient portal would send reminders. Measures would include process (calls completed, visits attended within 7 days, action plans given), outcome (COPD emergency department visits per patient per year), and balancing (staff overtime, no-show rates). Budget: $8,000 could cover spacers, printed materials, and a small amount of interpreter or transportation support; the rest uses existing staff. To address skeptics, I would share the baseline data, start with volunteers, and show run charts at monthly meetings. I would recruit one skeptical physician as a co-lead so the project is not seen as top-down.
2. How would you evaluate whether the intervention is working, and what confounders might mislead you?
Read the worked answer
I would use a run chart or statistical process control chart of monthly COPD emergency department visits and 7-day follow-up rates, with annotations for each PDSA cycle. I would compare the intervention patients with a concurrent comparison group from the same clinic or a nearby clinic, if available, and I would risk-adjust for age, COPD severity, prior emergency department use, and comorbidities. I would track process measures, because a fall in emergency department visits without improvement in follow-up might be due to seasonality or coding changes. Confounders include influenza season, changes in insurance or emergency department access, regression to the mean (the 38 patients were selected because they had high prior use), and outside emergency department visits not captured by the health information exchange. I would also watch for unintended consequences, such as increased urgent care use or staff burnout. I would report results transparently, including negative findings, and revise the intervention rather than declaring success prematurely.
3. One physician says, 'This is social work, not medicine.' How do you respond using systems-based practice principles?
Read the worked answer
I would acknowledge that the comment reflects real time pressure, then reframe the work as part of medical care. COPD exacerbations are driven by inhaler technique, vaccination, smoking, and timely follow-up; those are clinical interventions, not social extras. Under a shared-savings contract, the clinic is accountable for total cost and quality, and emergency department visits are a measurable outcome. I would point out that a 20% reduction is achievable with a small number of high-risk patients, and that the same infrastructure can help with heart failure and asthma. I would invite the physician to choose one piece, such as reviewing inhaler technique or signing action plans, so the work is shared. I would also bring data to the next meeting and recognize early wins. Systems-based practice means coordinating care across settings, not adding social work to a busy schedule.