Author: Ahmed Taher
Welcome to another HiQuiPs post, where we discuss the concept of variability and its relation to QI processes.
You are finishing another busy shift in the ED, and you notice that wait times have increased, patient beds are now set up in the hallway, and extra team members have been called in to help. You are frustrated with the length of time it takes to get routine blood work back on your patients. You feel that these delays may be related to a recent health information system (HIS) update. You discuss this at the next department business meeting, where people have many different opinions on the reasons for these recent delays. Your ED Chief sets up a core change team to investigate this further. As part of this team, you start to wonder about this process variation and what elements could affect it.
Why variation?
Why is this focus on variation so important with regards to healthcare QI? Well, variation is often seen as a source of errors or issues with the system.(1) This concept stems from the manufacturing world, where Deming’s management theory highlighted the importance of reducing variation to ensure quality.(2) You can imagine that if you are producing a line of cars (or any product for that matter), the specifications should be almost identical to ensure a high quality product. Healthcare processes can be viewed in a similar fashion. When looking at clinical outcomes, there may be best practices or guidelines to conform with to ensure appropriate care delivery quality.(3) For example, if one ED doc in your department performs CT scans in 90% of their patients presenting with minor head injuries and another in 10%, this variation is problematic to ensure high-quality care (provided they see similar patients). Hospitals’ laboratory management systems are especially attuned to quality control processes. Indeed, medical biochemistry labs need to have machines calibrated with appropriate reference ranges – and this is a good thing!.
Causes of Variation
Healthcare leaders and change teams need to understand (and sometimes decrease) variations in care by monitoring processes and outcome measures.(3) The underlying rationale is to ensure that processes are safe and effective, and it helps answer questions such as “how is care today?” and “how is care compared to yesterday?”(4)
Process variation is the spread of process output over time. For example, your daily commute to work may take you an average of 45 minutes, but on any given day, it may take you anywhere from 40 to 50 minutes – this is the spread. There is variation in every process, including those in the ED.(1) There are two main types of variation that are important to consider:
Common Cause Variation: This type of variation is produced by random variation, inherent to the process itself.(1) For example, your daily commute time will vary by 10 minutes based on the weather conditions, traffic that day or the number of red lights you encountered.
Special Cause Variation: This type of variation is seen due to identifiable causes, that are outside the core work processes.(1) For example, if there was a big accident on your route to work, or your bike had a flat and you needed to take the bus.
Coming back to your ED problems, common cause variation in the time it takes from ordering routine blood work to obtaining the results includes random variation along the process. This includes placing the order (and getting distracted momentarily), nurse acknowledgement (and being busy with competing orders), blood collection (and whether repeat ‘pokes’ are needed), sending samples to the lab (and whether the porter waited longer than usual for the elevator), and more. This type of variation is said to be ‘stable’ and has a spread that can be anticipated. For example, before your hospital implemented a new HIS, it may have consistently taken 45-60 minutes to receive blood work results from the time of your order.
A process with only common cause variation is said to be a ‘stable process’. On the other hand, special cause variation may be the result of a large change in the usual variation due to a specific event. For example, a machine breakdown in the lab, or HIS changes completely altering the way orders are made and processed. A process with special cause variation is said to be an ‘unstable process’. This type of variation can be identified and then improved or eliminated. This terminology can also be applied to systems (i.e. being a stable or an unstable system based on the type of variation identified).
How to intervene
In general, the first step to take when you identify that there is indeed variation is to identify whicpe of process variation exists, since each type of variation requires a different type of response. If there is only common cause variation – that is, the process is stable but suboptimal, then the process itself needs to be adapted or changed.(1) If special cause variation is identified – that is, something happened that is making the process different in a good or bad way, then this needs to be investigated to determine the appropriate response through QI methodology (as discussed in our previous posts). Additionally, any changes made to the process in response to the identified variation need to be monitored until the system becomes stable again.
Finally, as you may have realized, to enable monitoring of process variation, a robust data collection and reporting approach needs to be constructed. The variables need to be defined deliberately, they need to be measurable, and as always stakeholders need to be engaged throughout the process.
Now we have a better understanding of variation and its relationship to healthcare quality. Join us next time when we will discuss one of the main tools of capturing the data involved and delineating variation as we learn how to construct run charts.
Senior Editor: Lucas Chartier
Copyedited by: Edward Mason
References
1. James C. Manufacturing’s prescription for improving healthcare quality. Hosp Top. 2005;83(1):2-8. https://www.ncbi.nlm.nih.gov/pubmed/16092632.
2. Deming W, Edwards D. Quality, Productivity, and Competitive Position. Cambridge, MA: Massachusetts Institute of Technology, Centre for Advanced Engineering Study; 1982.
3. Bowen ME, Neuhauser D. Understanding and managing variation: three different perspectives. Implement Sci. 2013;8. doi:10.1186/1748-5908-8-S1-S1
4. Neuhauser D, Provost L, Bergman B. The meaning of variation to healthcare managers, clinical and health-services researchers, and individual patients. BMJ Qual Saf. 2011;20 Suppl 1:i36-40. https://www.ncbi.nlm.nih.gov/pubmed/21450768.