What Are Healthcare Data Analytics?

Science and healthcare have always gone hand in hand, but in recent years a new form of science has entered the field: data science. Perhaps this isn’t too surprising considering that every day, humans generate 2.5 quintillion bytes of data.1 Healthcare analytics typically includes the analysis of data generated from four different areas: claims and cost data; pharmaceutical and R&D data; clinical data collected from electronic medical records; and patient behavior and sentiment data, which includes retail purchases. All of this is analyzed to determine patterns and figure out how healthcare organizations can improve care, run more efficiently and reduce spending.

Organizing and analyzing this amount of data can be a large undertaking, and some may question whether it’s worthwhile. However, there are numerous benefits to healthcare data analytics—and, of course, some inherent challenges—as well as platforms and systems that can help ease organization’s transition to this new model of operating.

The Challenges

The sheer quantity of data available presents a challenge in and of itself. Processing, organizing and cleaning data can be time-consuming, as is correctly identifying which data is usable and which may skew results. Traditionally, this type of work required the expertise of data scientists—an expense many organizations are either unable or unwilling to take on.

Privacy and security are another major concern. In addition to HIPAA, some states have individual privacy laws to protect individuals’ health information. The lines between state privacy laws and HIPAA are not always clear and can make it difficult to determine how to obtain data while still upholding privacy. In addition, healthcare organizations need to feel like their analytics products and systems are safe and secure. These concerns may cause facilities to reconsider whether pursuing healthcare data analysis is worth the effort.

The Benefits

While there are challenges to overcome with healthcare analytics, the potential benefits typically far outweigh them. The correct application of healthcare data analytics can:

  • Help cut down administrative costs.
  • Support clinical decision-making.
  • Cut down on fraud and abuse.
  • Allow for better care coordination.
  • Improve overall patient wellness—perhaps the most important benefit of all.2

Cutting down on administrative and operational costs is a major potential benefit. Running a hospital takes a seemingly endless number of small steps that must all work together seamlessly; if one process or procedure gets out of sync, it can bog down the whole system. Data analytics can help find the kinks in the system, processing vast quantities of data and providing a “command center” overview to quickly identify problems. The best analytics platforms go even further, using predictive learning (artificial intelligence and machine learning) to more proactively identify issues and even recommend solutions. Major large healthcare organizations successfully use data analytics to better facilitate resource allocation, staff schedules, and patient admittance and discharge.3

One example of this is operating-room scheduling. This traditionally has been a cumbersome, inefficient process requiring phone calls, emails and faxes. UCHealth

in Colorado began using a scheduling app that employed data analytics and predictive analysis to allow surgeons and schedules to request block times with one click. The results were fairly dramatic. Surgeons released their unneeded blocks 10% faster than with manual scheduling techniques, and the median number of blocks released by surgeon per month increased by 47%. This all resulted in a 4% increase in per-OR revenue—an additional $15 million in revenue annually.3

NewYork-Presbyterian Hospital used similar techniques to improve infusion scheduling, which can be an extremely complex process. Using predictive analysis and machine learning, the hospital was able to optimize schedules for a 50% drop in patient wait time, which helped optimize nurses’ workloads as well.3

Predictive data analytics can also be used to forecast the likelihood of a patient needing to be admitted, as well as an immediate estimate of which units can accommodate the patient. Sharp HealthCare in San Diego was able to reduce its admit order-to-occupy time by more three hours through the use of a data-driven analytics system.3 On the flip side of this coin, analytics can also be used to facilitate the discharge process. Predictive programs can scan mountains of data to determine which avoidable factors were responsible for past discharge delays (such as insurance verification problems, lack of transportation or post-discharge care) and then apply that data to determine which current patients are likely to be delayed. This allows case managers and social workers to proactively work with these patients to help avoid discharge delays. Using these techniques, MedStar Georgetown University Hospital in Washington, DC, was able to increase its daily discharge volume by 21%, reduce patients’ length of stay by half a day, and increase morning discharges to 24% of all daily discharges.3

Excitingly, healthcare analytics can also help patients avoid hospital stays altogether. For patients with chronic conditions, predictive algorithms can identify potential risk factors, allowing clinicians to provide proactive management. Sanford Data Collaborative partnered with University of North Dakota School of Medicine’s Population Health Department to develop an algorithm that can predict with nearly 80% certainty the likelihood that a given diabetic patient will incur a costly and unwanted unplanned visit. Healthcare providers are then able to intervene and help patients avoid such visits—a benefit both for the patient and for the healthcare facilities.4


Conclusion

Despite inherent challenges, healthcare data analytics are certainly the way of the future; they allow healthcare institutions to provide better care to patients while improving efficiencies and lowering costs. Unfortunately, a lack of training and expertise—as well as overly complicated systems and programs—often discourage institutions from implementing such analytics.

Quartz from Quartz Clinical utilizes predictive learning to create an intuitive, powerful healthcare data analytics platform. Its user interface is simple and straightforward, making it easy to use without any specialized training.

Quartz allows users to access clinical, quality, financial, operational and supply metrics on any device. Its analytics do a deeper dive for monitoring inpatient and outpatient volume, patient disposition, length of stay and more—even allowing users to drill down by clinical department, service line, site of service, procedure, provider or insurance product.

Here are some examples of what can be achieved with Quartz:

  • Track 30-day return to ER and readmission rates to identify outliers and implement changes. 
  • Automatically analyze inpatient mortality, helping you meet value-based purchasing goals.
  • Dissect financial data by costs, reimbursement and contribution margin, resulting in an overall decreased cost of care. 
  • Identify outliers in procedures and with providers.
  • Track patient disposition by provider to avoid bundled payment penalties.
  • Optimize staffing hours (and hours of operation) to decrease costs.

Healthcare data analysis can seem overly complicated, but it doesn’t need to be. Implementing platforms such as Quartz allows healthcare organizations to minimize the challenges of such analysis while enjoying the benefits, ultimately leading to more efficient facilities and better patient care.

References:

  1. https://www-01.ibm.com/common/ssi/cgi-bin/ssialias?htmlfid=WRL12345USEN
  2. http://www.ajmc.com/journals/issue/2016/2016-vol22-n2/the-promise-and-perils-of-big-data-in-healthcare?p=2
  3. https://hbr.org/2017/10/why-hospitals-need-better-data-science
  4. https://hbr.org/2018/03/making-better-use-of-health-care-datas