The top 10 use cases for analytics in high-growth health technologies


Healthcare IT News recently published an article on 18 health technologies poised for big growth, a list culled from a HIMSS database. The database is used to track an extensive list of technology products that have seen growth of 4-10 percent since 2010, but have not yet reached a 70 percent penetration rate.

It strikes me that advanced analytics could have a significant impact on the greater adoption of these technologies and to further their capabilities and deliver increased value.

Because some of health technologies are already based on analytics – forecasting or optimization (e.g. cost/utilization analytics) – I didn’t include them here. This post instead takes a look at 10 use cases that can benefit from the greater use of analytics:

  1. Bed management – Remaining first-time buyers: 49.7 percent

Bed management technologies provide visibility into how beds are used and when they’re in use. What if you applied forecasting to this data so that you could see peaks and troughs of use across the hospital and network and the forecasted demand for specific skill sets to staff those beds? This information could improve planning for capital purchases of new beds, allow for preventive maintenance scheduling on beds and improve staffing across hospital and network operations.

  1. Business intelligence – Remaining first-time buyers: 40.3 percent

Traditional business intelligence is a historical assessment to determine what happened and why it happened. Forecasting analytics and scenario analysis can provide a forward-looking view to what is likely to happen based on various courses of action. Optimization analytics provides the insight to deliver the ideal outcome based upon a specific course of action. If you’re not using forecasting, scenario analysis and optimization, you are missing a significant opportunity to improve operations and clinical care. Forecasting enables you to understand what is most likely to occur and allows you to assess various scenarios to choose the optimal course of action.

  1. Data warehouse – Remaining first-time buyers: 39.7 percent

Modern data warehousing can be distilled to one word – Hadoop. And this is not a what-if option (as with the other technologies). This is a you-must-embrace-it-now solution. It’s a data store environment that is low-cost, flexible, scalable, fault tolerant and can process analytic models using huge volumes of data within the data store.

  1. Dictation with speech recognition – Remaining first-time buyers: 44.4 percent

You can now use text analytics to move natural language processing closer to unstructured text and also use it to update patient records. Neuro-linguistics and medical ontology can be used look for inaccuracies or contra-indicated treatment activity so that they could be identified sooner rather than later. Ultimately, this could mean identifying the appropriate patient record to automatically append the new dictation or halt an inappropriate procedure or treatment.

  1. Enterprise master patient index (EMPI) – Remaining first-time buyers: 39.6 percent

EMPI can use existing data management tools (instead of a purpose-built application with limited utility) to create a master patient record. The benefit of this approach is that the data management software has more utility across the enterprise. Additionally, robust data management tools allows you to make better use unstructured data to enhance match rates and verify (or improve) data accuracy for match decisioning. This results in easier record review and reconciliation.

  1. Enterprise resource planning (ERP) – Remaining first-time buyers: 65 percent

ERP enables inventory optimization. By optimizing your inventory, you can keep just enough supplies to help ensure there are no stock outs that could affect the quality of care or operations. It also means you minimize carrying costs, required shelf space and real estate and can help you avoid holding items beyond their shelf life.

  1. Infection surveillance system – Remaining first-time buyers: 49.3 percent

Infection surveillance systems use analyses to identify areas of operational improvement for preventing infection, monitoring infection control initiatives and assessing the impact of infection control programs. These systems can be further enhanced by using a Hadoop data store to access a significantly larger data set for correlation analyses, forecasting and prediction analysis. Imagine what it would mean for your organization if data such as patient longitudinal views, staffing levels, patient volumes, procedures, complications, event locations, time of day (and more) could be used for uncovering insights.

  1. Laboratory (outreach services) – Remaining first-time buyers: 41.8 percent

Imagine if you could use analytics to stratify patients based on their likelihood to adhere to treatment and then have laboratory outreach services focus their attention on those patients likely to benefit the most (e.g., those most likely to be non-adherent). Outreach services could orchestrate patient compliance programs and use resource allocation models to determine support resources needed at the individual level.

  1. Medical necessity checking – Remaining first-time buyers: 32 percent

Medical necessity checking has been referred to as "next-level clinical decision support." This ability is largely rules-based. What if you could use advanced analytics (in real time) to assess risk and receive recommendations when additional tests are warranted; or, that a patient is at low risk and testing is not warranted? This concept also applies to prescription treatments and other procedures. It provides another source of information and context for clinical recommendations.

  1. Staff scheduling – Remaining first-time buyers: 42.2 percent

It is interesting to note that this technology has less than 60 percent penetration in the market. The reason could that many smaller hospitals do not see a need for it. Or, perhaps the benefits of collecting the data are unclear because of limits of their existing technology. But what if, in addition to the traditional forecasting staffing demands, they also used analytics to assess staffing levels based on patient outcomes? Often, staffing levels are based upon patient volume, not patient outcomes. But what if you had access to recommendations (and the associated cost forecasts) comparing the use of full-time vs. part-time employees to cover demand peaks, weekends and holidays? This enables you to build the case for different staffing models in the future so that there is time to act rather than react.

Analytics can further increase the value of each of the above health care technologies by using the underlying data to provide insights and forecasts about current operations and the advantages and disadvantages of alternate courses of action. This analytical insight helps enable an organization to be proactive and realize additional savings and better patient outcomes.


About Author

Brad Sitler

Principal Industry Consultant Center for Health Analytics & Insights, SAS

I am a strategist focused on relationship marketing and promoting innovation in the biopharmaceutical industry. Prior to joining SAS, I was managing my own consulting practice focused on commercial life sciences. My background also includes leading multiple large pharma client engagements for leading database marketing service provider. I have completed a Six Sigma Belt with Johnson & Johnson Consumer Products Companies while directing the implementation of new technologies into their customer care centers and managing the execution of CRM programs. My experience also extends overseas to AsiaPacific where I worked in the telecommunications industry leading CRM technology enablement for call centers across multiple countries.

Comments are closed.

Back to Top