LEARNING OBJECTIVES
Introduce the concept of a control chart as a quality improvement tool
Understand how a control chart helps in monitoring variation in healthcare settings
Demonstrate how a control chart was used to monitor emergency outpatient department delivery variations and guide improvement strategies
INTRODUCTION
In quality improvement, understanding variation is essential for improving healthcare processes. A common question in the Model for Improvement asks, “What change can we make that will result in improvement?” However, generating ideas for change requires tools that help analyze variation in a system. A control chart is a key tool for monitoring and understanding process variation. In healthcare, it helps identify whether variations are due to common causes (inherent in the system) or special causes (signifying the need for corrective action) (Table 1).
Table 1.
Types of data and appropriate control charts
| Type of Data | Decision Criteria | Control Chart Type | Purpose |
|---|---|---|---|
| Variables data | More than one observation per subgroup | X̄ & S chart | Monitors average and standard deviation of continuous data |
| Only one observation per subgroup | XmR chart | Tracks individual measurements over time | |
| Attributes data | Occurrences vs. nonoccurrences? | P-chart | Tracks the percentage of defective units in a sample |
| Are defects counted in unequal areas of opportunity | U-chart | Measures defect rate when sample sizes vary | |
| Defects counted in equal area of opportunity | C-chart | Counts the number of defects per unit |
UNDERSTANDING HOW RUN CHARTS AND CONTROL CHARTS HELP IN QUALITY IMPROVEMENT
A run chart is an essential tool in the early stages of improvement projects. It allows for the visualization of data trends over time and the identification of signals of improvement. It relies on the median and simple probability rules to detect shifts or patterns but does not quantify variation precisely, as it lacks control limits. Run charts do not assume any specific data distribution, and we need at least 10–12 data points for probability-based rules to be applied in a run chart. If there are fewer than 10 data points, it is better to use the median before and after the intervention or rely on visual analysis to assess trends and patterns.[1–3] However, run charts are less sensitive to subtle process changes than control charts.[4]
A control chart is needed when a process requires more advanced monitoring, particularly to determine whether variations are due to common causes (natural variation) or special causes (unexpected factors requiring intervention). Unlike run charts, control charts rely on statistical control limits (typically ±3 SDs from the mean) and assume the data follow a normal distribution. Control charts require at least 20–25 data points to provide meaningful insights and are more sensitive to subtle process changes. The transition from a run chart to a control chart typically occurs when a sufficient number of data points have been collected, allowing for a more precise analysis of variation and long-term process stability. Additionally, whereas run charts revise the median after a shift or noticeable data change, control charts adjust the mean and control limits when major changes occur.[2,3].
Advantages of a Control Chart
The control chart offers several benefits in healthcare quality improvement
Clarifies Variation: It helps distinguish between normal (common) and abnormal (special) variations in healthcare processes.
Visualizes Trends: The chart provides an easy-to-understand visual representation of trends over time.
Improves Decision-Making: Healthcare teams can make data-driven decisions, identifying when changes are needed.
Facilitates Continuous Monitoring: Regular updates of the control chart ensure that the team can track progress and respond quickly to emerging issues.
Enhances Collaboration: The chart fosters a shared understanding among all stakeholders involved in the process.
HOW TO CREATE A CONTROL CHART
The control chart consists of the following elements:
Center line: Represents the average or expected value of the process.
Upper and lower control limits: These boundaries define the range of acceptable variation. Values that fall outside these limits indicate the presence of special cause variation.
Data points: Represent actual observations plotted over time to show trends.
Steps to create and use a control chart in a healthcare setting:
Define the process to monitor: For example, the number of deliveries in the emergency department.
Collect data over time: Record the number of deliveries at regular intervals (e.g., daily or weekly).
Plot the data: Use the control chart to plot each data point over time.
Analyze the chart: Look for points outside the control limits, indicating special cause variation.
Take action: When special causes are identified, investigate further to find the root cause and develop improvement strategies.
In analyzing the control chart, several rules helped in identifying special cause variations[1]:
Data points outside control limits: Any data points falling outside the established upper or lower control limits signal a special cause of variation, indicating that the process is out of control and requires attention.
Shift: A shift is observed when eight or more consecutive data points are either above or below the center line, suggesting a significant change in the process that needs investigation.
Trend: A trend is identified when six or more consecutive data points show a consistent upward or downward movement, indicating a systematic change that could affect delivery rates and require an improvement strategy.
Two of three: If two of three consecutive data points are within the outer one-third of the control limits, there may be a potential issue with the process requiring further scrutiny.
15 consecutive points: When 15 consecutive data points fall within the inner one-third of the control limits, it suggests that the process may have become too stable, potentially limiting improvements and requiring a reassessment of the system’s dynamics.
Control charts can be created with various tools, such as QI Macros and Excel add-ins, which integrate easily into Excel for quality improvement tasks. Minitab offers advanced features like Real-Time SPC for statistical analysis and process monitoring. These tools help visualize and analyze data to maintain process control and identify areas for improvement.
CASE STUDY AT GANDHI MEMORIAL HOSPITAL EMERGENCY DEPARTMENT
At Gandhi Memorial Hospital, a control chart was used to monitor the number of child births in the emergency department (ED) from March 2023 to January 2025, because the hospital observed fluctuations in number of deliveries. (Fig. 1) Using the control chart, the team was able to identify these deviations as special cause variations, as evidenced by a shift consisting of nine data points above the center line rather than normal fluctuations. Upon investigation, it was discovered that certain factors, such as staffing shortages during peak hours and a lack of coordination between departments, were contributing to the variations. Armed with this information, the hospital implemented targeted strategies to address these issues.
Figure 1.
Control chart showing emergency delivery numbers at Gandhi Memorial Hospital. C-chart: control chart; LCL: lower control limit; OPD: outpatient delivery; UCL: upper control limit.
Improved staffing: The hospital adjusted staffing schedules to ensure sufficient coverage during high-demand times.
Enhanced coordination: Communication between departments was improved to streamline patient flow and reduce bottlenecks.
Continuous monitoring: The control chart was updated regularly to ensure that these interventions were having the desired effect.
As a result of these strategies, the number of deliveries in the ED became more consistent, with fewer extreme fluctuations. This case highlights how a control chart can provide insight into variation, helping healthcare teams take appropriate improvement actions.
SUMMARY
The control chart is a powerful tool for understanding variation in healthcare processes, especially in emergency settings like child birth. This tool helps healthcare teams identify the root causes of variations and implement targeted improvement strategies. At Gandhi Memorial Hospital, applying the control chart reduced unwanted variation in emergency deliveries, resulting in improved care and outcomes. Regular monitoring and data-driven decision-making tools are crucial for maintaining quality and safety in healthcare systems.
References
- 1.Provost LP, Murray SK.. The Health Care Data Guide: Learning from Data for Improvement. 2nd ed. Jossey-Bass; 2022. [Google Scholar]
- 2.Moen RD, Nolan TW, Provost LP.. Quality Improvement Through Planned Experimentation. McGraw-Hill; 2012. [Google Scholar]
- 3.Anhøj J, Olesen AV.. Run charts revisited: a simulation study of run chart rules for detection of non-random variation in health care processes. PLoS One. 2014;9:1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Langley GJ, Moen RD, Nolan KM, et al.. The Improvement Guide: A Practical Approach to Enhancing Organizational Performance. Qual Manag J. 1997;4:85–86. [Google Scholar]