This is the third blog in a series highlighting how the Los Angeles County Department of Health Services is helping patients better navigate their care. Read part 1 and part 2.

These days, we’re all pressured to prove our efforts are worth the investment. Essential hospitals, or those that serve a safety net role in their communities, are under particular stress. How do you prove your efforts are effective?

In the case of Olive View-UCLA Medical Center, our wayward traveler (i.e., rescue care patient) now has a guide. His patient-centered medical home (PCMH) care manager now receives automatically triggered emails alerting her of the patient’s “off-course” visit to the emergency department or urgent care clinic.

In light of their new ability to proactively reach out to providers and patients to coordinate care, our group named this project “Prospective Action in Care Transitions (ProACT).” We’re now in the early stages of “go-live” testing and refining the process.

How can we be confident any improved care coordination is related to this innovation?

The health care environment is changing rapidly and simultaneously in many ways, with:

  • Medicaid expansion;
  • PCMH implementation and maturity; and
  • improved case management, care management, and utilization review.

Any of these may result in access to improved care coordination, and the effect may be additive.

How do we know whether ProACT is effective? Perhaps it only appears effective in light of simultaneous changes that may also improve care coordination.

Poise and Equipoise

When ProACT is fully operational, even when “lost” in his medical care, our patient can feel poised. In contrast, we, as project managers and stewards of safety net resources, are in equipoise.

Equipoise is the state of uncertainty about whether an intervention will produce benefits. To determine ProACT’s efficacy, we need to isolate ProACT from potential confounding influences such as PCMH maturation. The best way to do that is to randomize patients so that some participate in ProACT, while others undergo usual (non-ProACT) care coordination.

In this manner, all patients will be “exposed” to usual care. ProACT patients will have what we hope will be enhanced care coordination and improved outcomes, including fewer ED visits, fewer admissions, and fewer readmissions.

If we knew for certain that programs like ProACT were effective, we would not randomize.

Random Chance

By a coin toss (a good way to randomize), we determined that all patients whose medical record number ends in an even number will be ProACT-managed patients, while patients whose medical record number ends in an odd number will receive usual care.

Sounds tricky, but luckily our programmers know which patients have medical record numbers ending in even versus odd numbers, and will selectively transmit email notifications only for those with even numbers.

Stay tuned for the next installment, which will explain how care manager feedback after rollout led to important improvements in ProACT.