What Are the Steps to Quality Improvement in Cancer Care?

The past 25 years have brought a revolution in the attention to quality in health care, dating from the 1999 Institute of Medicine Report, “To Err Is Human: Building a Safer Health System.” After the first wave of change focused on reducing medical errors, work shifted toward quality standards that produce the best outcomes. In cancer care, this has given rise to efforts such as the American Society of Clinical Oncology’s Quality Oncology Practice Initiative, or QOPI.

But how does an institution update its metrics, or institute a new quality improvement project? What happens when a giant health system acquires a smaller one, and the quality metrics must be integrated? How are processes changed or added to the workflow without driving down morale or causing staff to quit?

Christina Southey | Image credit: LinkedIn

There are known processes for this, said consultant Christina Southey, who on Wednesday led a group of oncology care professionals through the preconference workshop, “Adopting New Innovations in Oncology,” which offered an outline for how to bring a quality improvement idea to fruition. The session took place ahead of the 50th Annual Meeting and Cancer Center Business Summit of the Association of Cancer Care Centers with ACCC’s Oncology Practice Transformation and Integration Center, or OPTIC, which helps practices implement new treatment options or clinical guidelines.

Southey started by asking about barriers to change, and participants offered plenty:

• Physicians may have big ideas about QI tools, but the staff who have to keep the hospital or clinic running are pressed for time and may be less interested in learning to use them. As one participant said, “people like to stay in their own lane.”
• There is agreement that the current focus on disparities is warranted, but less consensus on what to do once disparities are identified.
• Standardizing workflows is very difficult; there is variation from clinic to clinic, even within the same organization.
• Agreeing on what is the right data and getting it can be a challenge.
• It is difficult to align incentives.

Adopting an Innovation

Southey explained the concept of “diffusion of innovation,” developed by Everett Rogers, which categories different types of people based on the pace at which they adopt innovation. Understanding this pattern is important as health care considers the pace of adopting elements of artificial intelligence (AI) into the workflow.

Innovators are the risk takers who will try a new technology or system when it is brand new and very expensive, before the problems are ironed out. Early adopters tend to be younger opinion leaders, but more selective about which innovations they try. Those in the early majority come next; this group is larger and adopts an innovation when it is clear the innovation will become part of the workflow. Next come the late majority, who are slower to accept the innovation and do so with some resistance. Finally come the laggards, who may refuse to adopt the innovation unless there is a penalty.

Each part of the population plays a role, Southey said. “Our innovators are the ones who are going to be helping us figure out what some of the key barriers are within the system,” she said. They are “helping the public advocate for some of the larger changes that are needed, so that the earlier adopters have a slightly smoother road to be able to adopt this new technology.”

Attributes. Southey outlined Rogers’ attributes of innovation, which should be considered when deciding whether to pursue a project. What is the relative advantage over the current method? Does it offer compatibility with the current work pattern? In terms of complexity, is it difficult to understand and apply? Does the innovation offer trialability—can it be attempted in advance at low risk? What is the observability; in other words, can it be seen in advance among early adopters?

In health care especially, Southey said, very few want to be the first to adopt an innovation, but many are willing to be among an early wave.

Approaching an Innovation Adoption—or Lack of Adoption

Key questions to ask before deciding to pursue an innovation are: What is the nature of the problem you are trying to solve? What are you trying to achieve? And each must be broken down into discrete steps:

• Observation to gain understanding
• Speak with staff, patients, and caregivers to gain all perspectives
• Diagram or brainstorm solutions
• Design measures to evaluate success

Quality improvement projects should be evaluated to see how they align with the “Model for Improvement,” a process developed by the Associates for Process Improvement that includes 2 parts:

• It asks 3 fundamental questions, which can be addressed in any order
• It uses the “plan, do, study, act” or PDSA method to test changes and then adapt based on results.

Setting and Achieving Goals

Southey took breaks during the workshop for participants to discuss challenges in their own clinics or workplaces to setting and achieving QI goals. One project involved getting nurse navigators involved earlier in the process when patients with stage I to III breast cancer ended up in the emergency department and needed follow-up care. Another discussed the feeling of being overwhelmed with the wave of data—of different types—and trying to figure out where to start. Still others said that within a large health system, the QI priorities of different geographic locations will not be the same, and this must be addressed. Some suggestions:

• Use data to identify which subgroups are not benefiting from current approaches
• Ensure that activities make sense for the patient population—work with a trusted community organization to confirm this. What the health system or cancer center thinks should be the top priority may not be the community’s priority.
• Have small group conversations with diverse groups of 10-15 patients.
• Have a sense of what is a reasonable amount of time for change to occur.

Approaches to Measurement

Southey explained both the different types of measures and how important measurement is to the success of QI initiatives. Process measures involve steps that physicians and staff perform, while outcomes measures assess the clinical results that occur as a result of the processes; necessarily, outcomes take longer to measures.

“Process measures are the things that are going to be our bread and butter with improvement projects, because those are the things that we have more control over,” Southey said. “Measuring the steps in the process and how reliable they are, helps us to know that if we're doing these things, then we have a relatively high degree of confidence that our outcomes will eventually be able to move, even if it's going to take longer to see that data over time.”

Outcomes measures do matter, however. Southey said 4 to 8 were optimal to get the right amount of data without collection being overwhelming.

Balancing measures are consequences of a QI project. They can be positive or negative and include things such as increases in staff workload, additional costs, savings, or movement in measures such as patient satisfaction.

Selecting the right types of data matters greatly—as much as possible, Southey said, using pre-existing data points that are already collected is preferred. Timeliness counts. Using data points collected only once a year will not allow a clinic team to see if their efforts are working, as data need to be gathered monthly or even weekly—and displayed graphically in a place that everyone can see.

And, the best possible outcome is if positive results can connect to a federal or state program that can reward the institution with a better rating or increased revenue.

Southey concluded by reviewing some sample tools that clinics can use for brainstorming and QI project development. All, she said, should develop a prediction and a question that the team is trying to answer.

QI projects, she said, should be integrated into the normal course of the day. The more frequently teams are engaging with the data, the better. “You want this data to be an active part of your discussion at every single meeting,” she said.