#402 – IMPORTANCE OF CONTEXT CHANGES IN ROOT CAUSE ANALYSIS – DUKE OKES

One of the more valuable changes in the ISO 9001:2015 edition was the introduction of the phrase Context of the Organization. Having organizations evaluate context when designing a quality management system (QMS) meant they could better justify the degree of rigor of their processes and controls.

Relevant factors include things such as the industry and amount of regulation; markets in which it operates and competitive, economic and social dynamics; and product and/or operating technologies. 

Context isn’t a new idea for organizations. For decades the strategic planning process has included an early activity titled “environmental scan,” which has the intent of evaluating the broader environment in which the organization operates, and setting the stage for factors to consider when making strategic and tactical decisions. For example, Walmart’s foray into Germany failed largely due to a misunderstanding of the difference in cultural values. 

Context is of course dynamic, meaning that as the world, markets, competitors and technologies change, the organization can be positively or negatively impacted. Some shifts in context may occur over a long period of time (e.g., changes in the ecological or social environment in which the product is used) or quite rapidly (e.g., introduction of a competitive product or an economic downturn that obliviates the market). Adjustments to the QMS may then be necessary or useful in order to maintain and/or improve performance of the QMS and/or viability of the products/services produced. 

Contextual factors are also relevant during root cause analysis. One of the seven M’s (causes of variation) is Mother Earth (environment). Some examples might be how a sudden significant change in humidity can affect how well a household refrigerator works, or how an ice storm can lead to a rash of traffic accidents. We can often detect these as causes by doing a change analysis (e.g., what changes in the process might have contributed to the change in performance?). 

Here are some other examples of context shifts and their impacts: 

  • Technology: Different battery technology in electric vehicles (EVs) called for a significant change in how firefighters deal with some automotive fires 
  • Sales Volume Increase: Preventive maintenance is determined by calendar months versus by hours operated, causing excess unplanned downtime in an organization where sales were growing rapidly 
  • Human Factors: An instrument was miscalibrated by a technician whose cognition was impaired by development of obstructive sleep apnea 

However, there are several issues that might impact our ability to identify and respond to shifts in context: 

  • A real change is not detected (Type 1 error) – Examples could be customer “off label” use of a product, changes made by a supplier that they don’t communicate, or a drop in unmonitored pressure on a machine. Tesla’s autopilot mode has demonstrated a weakness in identifying hazards (such as a large white truck lying on its side across both lanes of a highway, a condition likely not included in the training data set or previous experiences). 
  • A change in context is perceived but not accurate (Type 2 error) – An example might be believing customer satisfaction has changed after a service process redesign when in fact the difference resulted from a change in the satisfaction survey. Another might be interpreting a natural change in color of an item when exposed to its normal operating environment (vs. how it looked during manufacturing or while in storage) as indicating that it is eroding. 
  • The change in context is a slow drift that has no effect until a critical tipping point is reached – Examples include the slow degradation of organizational competency as more experienced employees are laid off or retire, or a slippery slope in regulatory compliance resulting from a change in management priorities driven by declining profits. 

So, changes in contextual factors are often the causes of problems, and identifying the specific ones is the role of the root cause analysis process. This identification can be supported by an effective internal change management process and external change/context tracking, as well as taking a view broader than just the seven Ms by thinking about context. 

The 2018 edition of the ISO 31000 risk management guideline also included identification of context as one of the early steps, indicating we should be able to proactively identify the risks related to these factors and implement adequate controls, thereby reducing the number of events for which root cause analysis is required. However, here it is important to also take into account of the scope of the risk management process, such as organizational, product or process. Existing controls themselves create a context that also need consideration, since there will be variances in both design and implementation effectiveness. 

Finally, we should be able to predict some changes in context. For example, changes in demographics can be predicted pretty far in advance, otherwise insurance companies would be out of business. Thru the use of AI organizations should be able to detect/predict changes in customer interests, just as process variables can be monitored to predict when a piece of equipment might fail. The same is true of human performance, albeit perhaps with somewhat less accuracy as well as employee concerns about that monitoring. Proactive responses, though, can significantly reduce the risks related to changes in context. 

Summary: 

With all the publicity about the ChatGPT bot developed by OpenAI I thought it would be interesting to see what it says about context and root cause analysis. So I asked and was somewhat impressed. Makes me wonder if my writing days are limited, unless I decide to try fiction! 

Per ChatGPT: 

There are several ways in which “context” can be used when performing a root cause analysis: 

1. Identifying the root cause: By considering the context of the problem, you can better understand the root cause of the problem and identify potential contributing factors that may not have been immediately obvious. This can help you to identify the true root cause of the problem and take appropriate corrective action. 

2. Identifying potential solutions: By understanding the context of the problem, you can identify potential solutions that are tailored to the specific circumstances and needs of your organization. This can help you to implement more effective and efficient solutions that address the root cause of the problem. 

3. Identifying potential risks: By considering the context of the problem, you can identify potential risks that may arise as a result of implementing a particular solution or taking a particular course of action. This can help you to mitigate these risks and ensure that your solution is successful. 

4. Improving the root cause analysis process: By considering the context of the problem, you can identify opportunities to improve the root cause analysis process itself. For example, you may identify ways to gather more information about the problem or to involve more stakeholders in the process. 

Overall, considering the “context” of a problem when performing a root cause analysis can help you to more accurately identify the root cause of the problem and implement more effective and efficient solutions that address the root cause. 

Addendum: 

As I was finishing this paper on context I read a post on Nextdoor where a local homeowner said a water pipe in their house had frozen and burst. No surprise, given the temperatures during that week. We’ve had a stream of water running from our faucet during the time and have not had a problem. 

The homeowner, who had just recently moved here from a warmer climate, stated they asked the realtor if freezing pipes were a problem, and the realtor said it didn’t get that cold here in East TN. 

Let’s analyze this with a context perspective: 

  • The homeowners recognized the shift in context as possibly creating a potential water pipe freezing event – Good … realized a context shift and identified potential risk. 
  • They requested information (from the Realtor) in order to see if the risk was real – Good … collecting data to assess level of risk. 
  • They accepted the realtor’s response as accurate – Bad … single source of unknown quality. o Perhaps the realtor hasn’t lived here long enough to know the local weather patterns 
    • Perhaps the realtor didn’t want to infer that the potential buyer might need to invest some additional money in the property immediately after buying 
    • Perhaps the realtor didn’t care … they were working for the seller. 
  • They didn’t do any verification of their single data source by checking other sources (e.g., new neighbors, internet search, plumbers, …) – Bad … quality of data should always be considered. 
  • In addition, the cost to install pipe heating tape or insulation, and/or to have run a stream of water, would have been minimal compared to potential cost of damage/repairs to the home. Not good or bad … just a decision they made to accept the risk. 

Life is always providing us with lessons. Thankfully we don’t always have to personally experience them in order to learn. The homeowners no doubt learned about freezing pipe risk in this area (single loop learning), but I wonder if they learned about doing better risk management (double loop learning)? 

ChatGPT’s analysis of the above case study (edited for length): 

If the homeowner was concerned about the risk of frozen pipes, they should have taken steps to prevent the pipes from freezing. There are a few things the homeowner can do: 

1. Keep the thermostat set to a consistent temperature, even when the home is unoccupied. This will help prevent the pipes from getting too cold. 

2. Insulate any exposed pipes, particularly those in unheated areas like basements, crawl spaces, and attics. This will help to keep the pipes warm. 

3. Keep cabinet doors open to allow heat to reach pipes under sinks, particularly on exterior walls. 

4. Allow a slow trickle of water to flow from faucets during very cold weather. This can help to prevent the water in the pipes from freezing. 

5. If the homeowner is going to be away from home for an extended period of time, they should consider shutting off the water supply to the home and draining the pipes to help prevent them from freezing. 

It’s also important for homeowners to take steps to prevent them, even if a realtor or other professional says it is not a concern. In this case, the homeowner may want to consider getting a second opinion or taking additional precautions to protect their home from freezing pipes. 

©2023 Duke Okes  Used with Permission

Bio:

Duke Okes is a writer, speaker, trainer and consultant for quality management. He is the author of “Root Cause Analysis: The Core of Problem Solving and Corrective Action, 2nd ed.” as well as books on performance metrics and internal quality audits. 

YouTube channel: https://www.youtube.com/channel/UCV4arkOVWAbiVfbBDrDD34Q where you will find videos on root cause analysis, risk management, auditing, … 

Website: https://aplomet.com/ where you will find articles, links to online courses, on RCA … 

LinkedIn: https://www.linkedin.com/in/dukeokes/ where you can find posts such as this as well as occasional pithy comments. 

 

Leave a Reply

Your email address will not be published.