A lot of money has been spent on information technology in health care with little to show for it. To understand why we must pay a visit to the hospital.
It only takes 10 minutes of direct observation of a nurse in a hospital to understand care-delivery processes are not standardized and are dependent on individuals, not systems. This lack of reproducibility leads to errors. Since every caregiver does it his or her own way, it’s difficult to improve anything. Stable systems that are reproducible are required to deliver consistently high quality. Industrial companies figured this out 50 years ago. The writings of manufacturing gurus Imai and Shingo provide insight into how quality is built into processes. A process must first be stabilized then standardized before being improved. Because few standardized processes exist in care delivery there are many possibilities for error. That’s why simply making a poor process electronic by implementing an electronic health record (EHR) doesn’t lead to better quality or cost.
When it comes to change, the technology is the easiest part. Most health systems in America have or are implementing the EHR. And the vendor processes for implementation have become very good. The hard part is to get the doctors, nurses, and administrators to agree on what is the best way to deliver the care. Since the doctors control most care decisions, the rest of the provider team follows the doctors’ lead. If the doctor wants to do things a certain way, that’s what is done. The problem is the next doctor wants it his way and so on. Eventually, we end up with a hopeless mess in which no one knows how anything should be done on any given day. And good luck to a new nurse or technician coming into the system who must learn a multitude of work processes and remember the doctor-dependent differences.
Health care technology is very effective when it is used to support a well-designed care process. The design of new standard care processes need to be owned and driven by the people doing the work, not by some outside consulting firm that brings a 100-page playbook as the answer. As the frontline workers create new designs, they need certain systems that can help them deliver the improved care. Examples of these systems include electronic alerts for medication interactions and reminders to ensure all steps in the care process for the pneumonia patient are followed.
There are two types of improvement systems needed to create a well-designed care process. One is an improvement approach that brings members of an existing clinical team members together to improve an existing care process. They use proven improvement methods such as the principles, systems, and tools of the Toyota Production System (TPS). The second is an innovation process aimed at radically redesigning care. It’s associated with TPS and employs design thinking.
In both cases, the initial effort where rapid experimentation occurs might be an ambulatory clinic or an ER. It becomes a place for others in the organization to learn. It is an inch-wide, mile-deep change in practice that incorporates new processes not only for care delivery but also management. It should result in the systems necessary for sustaining improvement over time. As the model line achieves 50% to 80% improvement over baseline performance, the learning should be spread to other parts of the organization. This new way becomes the new best-known way to deliver care.
One example of a radical innovation is the attempt of HealthEast (now part of Fairview Health Services), which serves the Minneapolis-Saint Paul area, to create the clinic of the future. The leaders brought the vendors in their extended supply chain to the table to help in the design process. This included Epic, an EHR company; Herman Miller, an office furniture company; Boldt, a construction company; and HGA, an architectural firm. Together, the team began redesigning the care-delivery model. Each vendor had the opportunity to deeply understand the needs of the HealthEast providers. By the end of the design phase a new process supported by electronic records, architecture, furniture, and building was integrated to create a unique patient experience.
Before HealthEast formed the model clinic, a group of 11 clinicians had over 11 preferred ways for “their” clinic assistant to do just about everything. One key process, screening the patient for health risks such as cancer and hypertension, resulted in over seven places in the EMR for the provider to look for relevant information. Not only is that time-consuming (contributing to physician burnout), but it also greatly increases the chances of missing important information.
The multi-disciplinary team created a single screening process. Now, clinicians have just two places to look in the EMR for information on whether patients have had screens like mammograms and colonoscopies for cancer, staff can remind patients about what screening tests they need, and leaders are able to support the development of standardized clinical processes. The leader’s standard work is to audit the process and monitor the data. If the process stops being followed or the data shows deteriorating results, leaders will know that immediately.
In the first three months after its introduction, the redesigned process reduced provider search time per patient by 23 minutes. The overall screening rate went from 60% compliance to 72% compliance, meaning over 500 more individuals were appropriately screened over baseline. Perhaps more telling are the changes in patient comments. They went from comments such as “I do not feel my medication list was reviewed,” to “My doctor and medical assistant are always timely, thorough, and reassuring.” These results would not have happened unless all parties were working to build a better process.
Technology now exists to support disruptive innovation in health care. It is an important enabler, but the process must precede the technology. For example, Hospital at Home is an innovation that may well cut the cost of care significantly by reducing the need for inpatient beds. It couldn’t happen without the technology, which allows 24-hour monitoring of patients, real-time electronic communication between providers, and complex equipment to be rapidly set up in the patient home. But it still requires a nurse and a doctor.
What that nurse and doctor do and how they do it are still what will determine successful outcomes of care. Building the care process through careful understanding of what each process step delivers is critical. The medical team can then leverage the technology for data and communication and other needs that support the steps in the process.
Again, this requires standardized work. Every nurse and doctor does not get to do it his or her own way. Standards are established about how the work is performed, and those standards are followed by all until a better way is determined collectively by the team. New innovative care models such as Hospital at Home are based on clear and reproducible standards and will obsolete the old ways of the non-standardized care delivered in most hospitals.
It takes more design time to create a care model that builds in quality and efficiency, but without that work upfront, the technology doesn’t matter and, in fact, only increases costs. This thinking is not new. Many industries from aviation to automotive to nuclear power have been applying this concept of “process before technology” for a long time. The safety and quality results in those industries is second to none. It’s about time health care catches up. Our lives may depend on it.