Efficiency at work: Algorithms advancing the Quadruple Aim of Healthcare

BACKGROUND 

Return on Investment (ROI) measures the amount of return (profit) an investment generates relative to its cost.  In finance, the ROI is useful when making future business decisions.

Efficiency is the quality of being efficient, a person working in a well-organized and competent way.

Many of the health information technology platforms in which organizations invest can improve how they collaborate and communicate. These improvements can save time, reduce costs, and enhance performance. 

While it might be easy to measure the impact of a less costly surgical instrument or a new generic medication on ROI, the benefits of efficiency are harder to measure.

To do it, you must first find a metric that measures time. For example, how much time do your clinicians spending looking for results or other documents? Multiply that by the number of staff and their average salary, and then you can place a monetary value on the efficiency of document look-up.

­­­Time is one valuable component of efficiency and performance is another. We want people to save time, certainly, but we also want to know that people are doing an even better job with the time they have. Measuring changes in performance is harder than measuring time savings. To do it, you must select a key performance indicator - something that tells you if performance has improved. In healthcare, that might mean fewer hospital acquired infections, less ICU bed days, or a shorter length of stay. Next, you have to benchmark the baseline and re-measure when your implementation is complete. Finally, you generalize the performance change across the organization or business unit. This process will give you some insight into how performance changes over time.

We believe DECISIOInsight helps clinicians work faster, perform better, and thus be more efficient through:

  • Real-Time data availability

  • Early identification of high-risk patients

  • Automated triage tools

Let us show you how!

DECISIOInsight AT WORK

DECISIOInsight is a clinical surveillance data visualization platform. It is an application that, once configured, collects data effortlessly in real time, surveilling every patient every moment of every day. DECISIOInsight does this work without any manual data entry on the part of clinicians.

DECISIOInsight monitors information from diverse information systems such as vital and ancillary monitors, laboratory systems, and medication systems, in addition to the EHR. It aggregates information in your hospital’s data center and interprets it in real time.

The Decisio clinical team works closely with your physicians, nurses, and other clinicians to understand and implement your own hospital's protocols and physiological parameters on a unit by unit basis.

DECISIOInsight presents customized displays to the right person at the right time, helping to decrease time to intervention and the variability in care.

Clinical protocols, or algorithms, represent the steps we take to accomplish a goal or solve a problem.

Clinical protocols, or algorithms, represent the steps we take to accomplish a goal or solve a problem.

DECISIOInsight embeds clinical protocols, chosen by our clinical partners, at the hospital and/or unit level. Broken down, clinical protocols are algorithms – steps taken to accomplish a goal or solve a problem. Algorithms exist for basic things, like making a grilled cheese sandwich, and complex things, like asking Google maps to take you from Houston to Los Angeles using the fastest route. Computer algorithms are useful in part because they are really fast. The algorithms in Microsoft Excel, for example, perform calculations and sort data much quicker than you could on your own. Algorithms act like building blocks to help us mathematically solve for complex and critical problems. 

An example of a health care algorithm is the National Early Warning Score (NEWS) algorithm. NEWS was designed to detect deterioration in acute care patients, helping clinicians know which patient needs an ICU admission, intervention for potential cardiac arrest, or is at increased risk for death1. Clinicians follow clinical protocols to provide care once NEWS is calculated. To calculate a patient’s NEWS, you could pull out a worksheet that looks something like this, circle the corresponding values and compute the patient's score manually.

Screen Shot 2019-02-06 at 6.48.21 PM.png

In the middle of a busy clinical day, this would be cumbersome, and you would have to make critical decisions about which patient to assess because this clinical protocol is time consuming. Additionally, after making the decision on which patients to evaluate with NEWS, the score is not updated automatically, and you would have to recalculate often on those patients you designated.

It would be better if you used a MDCalc’s website, like https://www.mdcalc.com/national-early-warning-score-news, to hand enter the values and calculate the result faster. The website’s algorithm would at least do the math for you, making it faster, but you would still have to decide when to use the tool. 

A better solution is to use a computer algorithm, such as the one configured into DECISIOInsight. Your clinical algorithms are monitoring patients continuously, in real-time; no worksheets, data entry, or calculations required. The behind-the-scene computer algorithm runs on all patients, all the time, so you don’t have the pressure of deciding when it should be used. The algorithm fires an intelligent alert, unique to your role-based clinical workflows, signaling the right clinicians when it has reached pre-determined thresholds. The result: this helps clinicians achieve better protocol adherence and improve clinical outcomes.

3 - Surveillance of STICU w:positive Sepsis Screen (cards FRONT).jpg

Spence (2015) argues that we should measure things that matter so that you don’t, for example, increase speed while sacrificing employee and customer satisfaction.

Algorithms can help us achieve the complex Quadruple Aim2,3 objectives: 

  1. Improve the health of populations

  2. Enhance the patient care experience

  3. Reduce the per capita cost of healthcare

  4. Improve the work lives of healthcare clinicians

Let’s take a look at how the algorithms do that.

After a recent implementation at a large, 800-bed academic medical center, we decided to survey a segment of our users because we were interested in quantifying the ways in which the DECISIOInsight tools were helping hospital clinicians to become more efficient through the use of algorithms. We are pleased to share our early results with you from the anonymous responses of 24 clinicians (nurses: 54%, doctors: 30%, other: 16%) working in the hospital’s most critical units. The majority of respondents worked in the emergency department (67%) and the intensive care unit (25%).

1. IMPROVE THE HEALTH OF POPULATIONS

The algorithms at work in DECISIOInsight, chosen by our clinical partners, seek to prevent harm through early detection. We know that high NEWS results are correlated with increased admissions to the ICU and decreased 30-day survival. If you know which members of your population are at risk, you can ensure the patient is seen by the right person at the right time. When an algorithm notifies a clinician to take action, patients stay healthier.  A clinical trial with over 2000 patients demonstrated that early intervention among hospitalized patients with high NEWS scores reduced ICU admissions and inpatient mortality4. This is an example of an algorithm working to improve the health of populations.

2. ENHANCE THE PATIENT CARE EXPERIENCE

The health care professionals we surveyed told us that DECISIOInsight helped them to make better decisions faster as a result of the algorithms. Bedside nurses drew labs earlier because they were alerted to a potential problem, thereby hastening treatment to reverse unstable conditions. Physicians told us that the algorithms helped them to decide which of many patients to see first. In one instance, deciding to first see a patient waiting in the ED with a high NEWS score helped the patient to receive critical treatment right away, bringing his vital signs down to a normal range.

Simply having the DECISIOInsight dashboard in the patient room can enhance the patient’s experience. One nurse told us that the dashboard “generates a lot more questions from families. They ask what every number is and what it means…It makes them more knowledgeable.” Over sixty percent of our respondents told us that patients and families were complimentary about DECISIOInsight.

Algorithms help clinicians make better decisions faster and this, along with a healthy dose of education and communication, improves the patient experience.

3. REDUCE THE PER CAPITA COST OF HEALTHCARE

Avoiding the ICU, where patients most at risk are likely to be, is one of the best ways to reduce the cost of hospitalization.  Even though only about a quarter of all US hospital discharges include ICU services, they account for almost half of total hospital costs. Hospital stays with ICU bed days are at least 2.5 times higher than stays without a visit to the ICU; on average $61,800 versus $25,2005,6. When algorithms detect abnormalities even before an experienced clinician can, the resulting early intervention reduces complications and costs.

4. IMPROVE THE WORK LIVES OF HEALTHCARE CLINICIANS

final.png

Perhaps the best part of computer algorithms is that they run secretly in the background of our work life, 24/7/365. They don’t require conscious thought; they just quietly do their job. In settings like the ED or ICU, where DECISIOInsight runs constantly on a dedicated monitor at the patient's bedside or securely at the clinician's station, staff don’t have to waste time logging into the application to see data. In fact, 75% of the clinicians we surveyed logged in only once or not at all to the DECISIOInsight platform during their shift. Furthermore, because the platform automatically displays results that clinicians would otherwise have to find elsewhere, 50% of our respondents estimated that they further saved at least an hour per shift by not having to navigate through the EHR (22% saved at least 15-30 minutes, 13% saved at least 45 minutes, and 40% saved more than an hour). 

Clinicians tell us that they use DECISIOInsight to check algorithm-generated metrics for NEWS and sepsis screening, to view the latest lab results, and to continuously monitor patients. In doing so, DECISIOInsight helped improve patient safety by:   

Asset 12.png

That’s a big win for clinicians who are dedicated to improving patient outcomes.

IN THEIR OWN WORDS 

Take a look at the clinician's feedback about DECISIOInsight in our survey and how it improved their work lives:

Screen Shot 2019-02-08 at 7.14.07 PM.png

EFFICIENCY MATTERS

We know that small changes in workflow – like seeing labs or alerts at a glance instead of having to walk to a computer station and login into the EHR and view a patient’s record - can make a big difference. Changes like these can help front line clinicians become more efficient and enhance their performance. This is especially true when clinicians are able to identify and triage high-risk patients faster than without the assistance of algorithms and clinical protocols. Scholarly literature repeatedly tells us that early intervention saves lives. The DECISIOInsight platform, when configured to your organization’s highest priority objectives, surveils patients in real time and presents important physiological data using an innovative visualization platform so you too can have more efficiency at work. 

REFERENCES

1 Bilben, Grandal, Sovik. Scad J Trauma Rescusc Emerg Med. 2016; 24: 80.

2 Institute for Healthcare Improvement. IHI Triple Aim Initiative; http://www.ihi.org/Engage/Initiatives/TripleAim/Pages/default.aspx

3 Bodenheimer, Sinsky. Ann Fam Med. 2014; 12(6): 573. 

4 Mitchell, McKay, Van Leuvan, Berry, McCutcheon, Avard, Slater, Neeman, Lamberth. Resuscitation. 2010; 81(6): 658.

5 Halpern, Pastores. Crit Care Med; 43(11): 2452.

6 Barrett, Smith, Elixhauser, Honigman, Pines. Healthcare Cost and Utilization Project. 2014; https://www.hcup-us.ahrq.gov/reports/statbriefs/sb185-Hospital-Intensive-Care-Units-2011.pdf

ADDITIONAL INFORMATION 

Copyright Information: © Decisio Health, Inc.  2019

Colorafi, K., & Quiñones, E. (2019). Efficiency at work:  Algorithms Advancing the Quadruple Aim of Healthcare. Houston, Texas:  Decisio Health.

Available at www.decisiohealth.com