Using Predictive Analytics To Better Manage Capacity In Healthcare

Hospitals are turning to technology for leaner processes


As the Baby Boomer generation has entered their retirement, pressure on the healthcare system has increased exponentially, and many are at breaking point. Also on the rise is the cost of healthcare. In an April 2016 survey carried out by personal finance site GOBankingRates, 18% cited rising health care costs as their biggest financial burden - above taxes, retirement savings and higher education. There has been a 3600% increase in healthcare spending since 1970, and despite efforts to quell the tide, the rise shows no signs of abating.

While there is a certain degree of profiteering in the healthcare industry, it is still largely the case that the cost of providing treatment is simply too high. Costs must be cut in order to deliver the kind of service needed at a price that can be afforded, while at the same time ensuring that the greater volume of patients coming into hospitals are taken care of. The key to solving these problems is improving the effective capacity of a healthcare system, thereby opening up access, see more patients in a more timely manner, and patients wait less. To do this requires predictive analytics.

Hospitals are now encouraged to gather and use as much data as they possibly can. And it seems many are already appreciating the benefits. An online survey of 136 health professionals conducted in August by Health Catalyst, a data warehouse, analytics and outcomes vendor, found that almost 80% of hospital executives believe healthcare could be improved by using predictive analytics more in their daily operations.

Healthcare provision is slow and waiting lists are long because assets held in the healthcare system - beds, waiting rooms, equipment, surgeons - are so scarce. However, utilization of these assets is shockingly low. Speaking at the recent Big Data & Analytics in Healthcare Summit, Mohan Giridharadas from LeanTaaS, noted that ’utilization of operating rooms is often less than 50%, utilization of infusion chairs is often less than 60%.’ He compared the situation to some queuing in line for a Super Bowl ticket for several hours, only to turn up to the game itself and find the stadium half empty.

The most common solution to increasing capacity is hiring and training new staff. This has the problem of greatly increasing costs. Since 2006, Fitch Rating reported that personnel costs have consumed an average of 50% of hospital revenue. In 2012, they comprised 54.2% of hospital operating revenue. It also takes time to train employees, and allow them to settle in to their role and get up to the required competency.

In order to circumnavigate such high labor costs and encourage more lean behavior, hospitals are increasing turning to predictive analytics. Predictive analytics allows you to anticipate demand for resources and take out any non-value adding time, ensuring lean operations and minimal wastage. It looks at time savings in every node, ensuring that everything is working optimally on a micro level, entering the next frontier of capacity.

Jamie Bachman of UCHealth notes how his hospital analyzed inpatient stay to make better use of operating rooms. By looking at the historical data of inpatients suffering from similar problems and in similar circumstances to a current patient, you can understand when they will come in and how long they stay, using this information to predict where they are likely to end up during their stay so you can allocate resources according. You can even ultimately match surgeons with patients according to demand, in a similar way that, for example, Uber operates when it comes to matching drivers with passengers, thus ensuring less downtime and lower wasted costs. Machine learning algorithms are then applied so that models are constantly learning and adapting according to new data, and can make better, smarter decisions.

Despite these benefits, just 31% of respondents to the Health Catalyst survey said their organizations have used the technology for more than a year, while 19% claim their organizations have no plans to use it. The amount of data created by the influx of extra patients as the Baby Boomers enter their twilight years, as well as technologies such as IoT and wearables as these become more commonplace, provide even more opportunities for hospitals to collect data. Politicians and healthcare providers need to understand the opportunity that proper analysis of this information can bring, and how a moderate investment in redesigning health services to incorporate the technology and leverage the insights can bring huge cost savings in the long run. 

University lecture small

Read next:

How Are Higher Education Institutions Using Analytics?