Scheduling for better healthcare

How analytics- and O.R.-driven tools help healthcare organizations move from “tracking” mentality to “delivery and logistics.”

healthcare scheduling

By Randolph Hall and Janice Partyka

I’m new to a city, sick and don’t know where to turn. I could drop in at the nearest emergency room, likely wait hours to be seen, and end up paying hundreds or thousands of dollars. Or, perhaps I could look for an urgent care center, taking a chance on how long I’d wait and whether they have the specialist I need, but knowing that my cost would be more reasonable. Or I could try ZocDoc.

ZocDoc is an Internet tool that allows me to not only compare different doctors (including specialists) and dentists in my area, but also look into their schedules, pick a time that best suits me, and book my appointment all by myself. ZocDoc is but one example of the scheduling tools available today that empower healthcare access, enhancing service for patients while improving utilization for healthcare providers.

Digitization

The Affordable Care Act (ACA), passed by Congress in 2010, will be a hot issue this election season because of its mandate for purchase of insurance coverage. But the ACA has many more provisions, including incentives for healthcare providers to acquire electronic health record systems (EHRs). The EHR provision aims to pull the healthcare industry into the 21st century in its use of computers to improve the delivery of care. And while attention has gone to managing and using health data, the use of computers to automate all aspects of care delivery has also become a huge opportunity.

Scheduling aims to improve the match between healthcare resources (doctors, nurses, rooms, equipment, medicines) and patient needs. A good scheduling system reduces waits for patients while also improving the utilization of critical resources. It does this by tracking the availability of resources, projecting future demands for service and automating the assignment of resources to needs. “Utilization is the number one incentive for buying systems,” say Barry Banek of Max Systems. But scheduling systems also produce needed data, such as how many surgeries were performed and of what sort, how long it takes per procedure and differences between doctors.

Healthcare providers look for cost savings when they purchase scheduling software. “It is very easy to make a quick ROI,” says Matt DeWolf of Runzheimer. “It’s expensive to hire nurses, so if you can get more visits per nurse and maintain quality of patient care, you’ve helped make the organization sustainable.”

Scheduling systems do much more than the traditional “white board” on the wall. Because they run on computer platforms, they can utilize and generate data that produce better schedules in a more flexible format, as well as link scheduling to the health needs of patients. Peter Legor of Brickmed foresees big changes as a result. “With ongoing centralization through health information exchange, we will see national aggregation of analytic systems that will spawn improvement in patient care,” he says. This will include the ability to identify patient risks and engage them in preventive care.

Personalization

While computer software is increasingly common in healthcare scheduling, optimization has made fewer in-roads than other applications of operations research, such as manufacturing and logistics. Patients differ in their genetic make-up, environment from which they come, demographic characteristics (such as age, language and mobility), particular symptoms that they exhibit and their own health behavior, so patient care must always be individualized. Also, unlike manufacturing, healthcare usually requires the presence of the customer (i.e., the patient), often within a stressful or uncomfortable environment.

Patients usually cannot leave until the job is done, making waiting that much more costly. But the consequences of waiting go beyond simple loss of time. The patient may experience pain, or conditions might worsen. For some diseases, death may result if not treated with sufficient speed. In emergency rooms, complications may occur when patients become frustrated with long waits and leave without being seen or leave against medical advice.

Unfortunately, many caregivers lack the skills to systematically improve service by creating schedules that better match resources to patient needs. The person in charge of a private practice is typically the doctor, someone with no managerial or engineering training. Even in larger clinics or health centers, it is uncommon to have someone with an advanced analytical education, such as operations research or industrial engineering, even though a large center may be a multi-billion dollar enterprise and much larger than a typical manufacturer. While healthcare as an industry is driven by deep research on the safety and efficacy of therapies, it often lacks sophistication in the processes by which care is delivered to patients.

healthcare scheduling

Scheduling Challenges

The keys to good scheduling in healthcare are data, analytics, systems, software, culture and management. Data combines with analytics to track historical trends and forecast the future, answering questions such as:

  • How long will a particular procedure take for a given patient, with a given doctor, on a particular day?
  • How many patients can we expect to present for care in an emergency department on a given day of the week, time of day and day of year?
  • How likely is it that a particular patient will be a no-show for a scheduled appointment made a set number of weeks in advance?

In all of these examples, the need is to maximize the precision by which healthcare is delivered to match demonstrated patterns of need.

Analytics, combined with systems, enable a schedule to be optimized against defined objectives related to the cost of offering service, the quality of the service provided and health outcomes, while also meeting defined constraints. In surgery, the underlying system may be one of block scheduling, where particular times and rooms are reserved for specific doctors or specialties. Patients may then be assigned to dates based on the criteria defined by each surgeon. Computer algorithms may be used to optimally assign blocks to particular surgeons and to fill surgical appointment slots.

Software is the tool to implement a schedule and accompanying analytics. The software provides interfaces to: schedulers who set appointments and assign resources, doctors who may wish to input preferences and constraints, patients who sometimes book their own appointments and management who monitor and control performance as well as allocate resources. The software can also support the automated acquisition and recording of data. And software can provide a tool for communication among and between departments, so that the arrival of patients and allocation of resources can be anticipated with greater accuracy.

Lastly, supportive culture and management are essential to implementing any new system for healthcare scheduling. Without support from the top and from the people who rely on the schedule, needed changes will not occur.

As discussed in the “Handbook of Healthcare Systems Scheduling” [1], key applications of scheduling include: capacity planning, patient appointments, workforce scheduling, surgical scheduling, bed management and medical supply logistics. These tools can be used in an ambulatory setting, such as an individual doctor’s office, a clinic or a specialty imaging center. Scheduling may also be used within the hospital, such as the emergency department, “operating theater” or in-patient wards. Scheduling is also needed for patients residing in their homes, such as planning for delivery of medical devices or routing and scheduling of mobile nurses and therapists.

Each scheduling application presents somewhat different issues. In surgery, it is critical that all of the needed resources (people, equipment, implants, carts, etc.) are present when surgery begins. Because these resources move around, operating room scheduling is a synchronization and coordination challenge. In workforce scheduling, the challenge is ensuring that staffing levels track needs, as predicted from patient census and acuity estimates. For appointments, the issue is avoiding holes in the schedule while minimizing the number of patients in the waiting room.

Software Products

Scheduling software empowers better health systems by capturing and analyzing data in an electronic format and presenting it to the scheduler in a user-friendly graphical display. Vendors selling scheduling software range from companies with just a few employees to multi-billion-dollar conglomerates. Smaller companies tend to focus on single applications such as appointment scheduling for the private practice. At the other end of the spectrum, scheduling may be one component of a suite of enterprise-level offerings. McKesson, for instance, offers scheduling alongside a distribution system for pharmaceuticals and medical equipment. Varian’s tools for tracking and scheduling oncology patients are coupled with the equipment it sells for delivering radiation therapies. And Epic and Cerner provide scheduling tools that couple to their EHRs.

For this article, we spoke with software vendors offering systems for clinics and practices (Brickmed, Max Systems), operating rooms (Max Systems, PICIS, SurgiScheduler), mobile equipment and clinicians (SurgiScheduler’s Mobile Medical Device Operator, Runzheimer) and health enterprises (API HealthCare, Cerner). In our discussions, three themes emerged: (1) healthcare software must be responsive to the vicissitudes and preferences of people, (2) regulations, privacy and security impose industry specific requirements, and (3) system and data integration are only nascent capabilities.

Coping with Human Uncertainty and Preferences

Keeping people on track – both patients and clinicians – is often the first problem tackled by scheduling software. “The people creating schedules often spend much of their day calling patients with reminders for appointments,” says Legor from Brickmed. Appointment scheduling systems automate reminders and make it easier to fill in the voids, improving utilization, which suffers when patients simply don’t show up.

Joe Smith of PICIS adds, “In the operating room, each surgeon takes a different amount of time for a procedure, and other factors are important, such as the patient’s medical condition.” Scheduling software provides a data resource to analyze trends and identify surgeons who mis-estimate surgery times or habitually don’t keep to schedule. Allen Kent at SurgiScheduler notes that the “surgeon is more apt to be on time to the operating room” when scheduling software is employed.

With respect to preferences, it is difficult to algorithmically optimize the assignment of individual patients to appointment times because of the complexity of each person’s time demands. As Banek from Max Systems notes, “Some clinics are allowing patients to book their own appointments.” Even when that is not provided, patients are normally offered alternative times and are booked on the spot (rather than the type of batch processing that is used in vehicle routing or production scheduling).

Surgeons are accommodated through the system of block scheduling. “This is a block of operating room time that is regularly set, maybe one day a week between certain hours,” according to Banek. Surgeons compete for the best block times, but once these are assigned, software is used to fill in the slots of each individual surgeon, rather than globally optimized.

Industry Specific Requirements

Healthcare is subject to numerous regulations. For instance, software and data storage systems in the United States must operate with the rules laid down by the Health Insurance Portability and Accountability Act’s (HIPAA) privacy rule, which regulates the release of personal health information. Data security is critical. “If you put data in the cloud or put your data on a third-person server, you need to be certain of their processes,” says Legor.

DeWolf from Runzheimer, which offers CareViz to schedule mobile health workers, sees challenges on linking patient systems into GPS-type devices. “There are HIPAA issues if you are handling patient information. You have to be HIPAA compliant.” Software companies are working to solve problems like these, adding scheduling functionality while remaining HIPAA compliant.

On the other hand, the ACA’s “meaningful use” regulations are spawning the growth of products that link scheduling to the electronic health record. Paul Gorup, chief of innovation and Cerner co-founder, says, “Without integration, you won’t do reporting properly.” This means having consistent data for analysis as well as certified systems. There is a strong pull toward integrating scheduling with the electronic health record, rather than focusing on administrative systems.

System and Data Integration

Traditionally, healthcare systems have been driven by the need to bill and receive payments quickly and accurately, and the ability to track the medical history of individual patients. Operational and quality improvements have been secondary, meaning that systems have been less adept at improvement, efficiency and overall patient outcomes. As stated by Gorup, healthcare organizations need to move from the mentality of “tracking” to that of “delivery and logistics,” just as FedEx had done in the 1990s.

API Healthcare offers an integrated data-driven platform for staff scheduling According to API President JP Fingado, their tool “predicts how much nursing time each patient will need, based on live feeds from the clinic, from the ICU and demographics.” Their software also accounts for the “multitude of payroll classifications,” automating their use. Cerner’s recent acquisition of Clarvia, which offers health workforce scheduling, points toward future integration of scheduling with health records. “We saw opportunity through Clarvia’s expertise in generating acuity requirements of patients,” says Gorup.

Cerner’s customers outside the United States have already demanded larger, more integrated, systems. For instance, in the United Kingdom, patients can schedule appointments across the entire health system, not just in an individual office or clinic. While the United States may never have a national health system, larger “accountable care organizations” that provide a full range of patient services will certainly demand greater integration than in the past.

Filling the Software/Research Gap

Though scheduling software is destined for growth, the available tools fall short from the operations research perspective and don’t utilize all the capabilities in the healthcare scheduling literature. Schedules frequently rely on humans, rather than algorithms, to make choices, and randomness/variability tend not to be modeled with great precision. On the other hand, research falls short on accounting for the diversity of objectives and constraints that healthcare providers must fulfill.

Improvements in data capture systems (for instance radio frequency ID tied to patients and equipment), as well as data integration across systems, are likely to change that trajectory. Good models depend on accurate data, which are only now becoming available. And true optimization is likely to become more prevalent in solving the most complex problems, particularly scheduling the workforce and scheduling multi-stage pathways of care. Also, as care providers become larger and more integrated, the need for scheduling will head toward enterprise-wide systems rather than the individual doctor office.

Even without advanced optimization and stochastic models, scheduling is empowering patients and providers to improve the match between healthcare resources and needs. As data capture becomes more automatic and systems become more integrated, operations research has the potential to make these systems even more effective, providing more immediate patient access at lower cost

Randolph Hall (rwhall@usc.edu) is vice president of research for the University of Southern California, as well as a professor in the university’s Epstein Department of Industrial and Systems Engineering. Janice Partyka (jpartyka@jgpservices.net) is principal for JGP Services and contributing editor for GPS World. In the interest of full discloser, JGP Services had a small engagement with Runzheimer in the past but does not currently work for them, and Cerner is installing the electronic health record system at USC Medical Center in Los Angeles, but Hall has no personal involvement in the project.

Reference

  1. R.W. Hall, ed., 2012, “Handbook of Healthcare System Scheduling,” Springer, New York.