Data Integrity and Human Error

Ensuring Data Integrity and Preventing Human Error

Prevention of human error is essential for ensuring data integrity. For instance, which poses a greater risk to patient safety: deliberately falsified data or data inadvertently altered through human error? The answer is that both present equal risk. Regardless of whether data has been intentionally modified or unintentionally changed through error, any data that fails to meet the ALCOA Plus principles (Attributable, Legible, Contemporaneous, Original, Accurate, Complete, Consistent, Enduring, and Available) is considered falsified and may directly impact patient safety.

In the pharmaceutical industry, it is reported that over 80% of incidents threatening data integrity assurance are attributed to human error. This statistic clearly demonstrates that the primary cause of quality deviations and process deviations lies in human factors. Even with the implementation of security features and audit trail functions in systems, these measures alone cannot prevent human error itself. While technical countermeasures are important, they do not completely eliminate errors arising from human factors. Therefore, preventing human error itself is an urgent priority.

Detecting Human Error in Data Integrity and Quality Risk Management

In ensuring data integrity and quality risk management, measures to address failures of facilities, structures, and analytical instruments are important, but detecting human error within processes is equally critical. Modern quality management systems adopt a risk-based approach, requiring the identification of potential error sources throughout the entire process and the implementation of appropriate control measures.

For example, transcription errors may occur in transcription tasks, calculation errors in calculation tasks, and input errors in data entry tasks. Checking procedures must be incorporated into each standard operating procedure to address these potential errors. While it is possible to design forms with dedicated check boxes, for important data and critical processes that directly affect patient safety, it is desirable that an independent second person performs the verification. This “double-check” principle is recommended by regulatory authorities such as the FDA (U.S. Food and Drug Administration), EMA (European Medicines Agency), and WHO (World Health Organization) as a fundamental requirement in pharmaceutical quality control.

Pointing and Calling: An Effective Method for Solo Verification

However, implementing double-checks by separate individuals for all processes is not realistic from the perspective of resources and efficiency. For example, while train conductors check the opening and closing of doors, having two conductors per train car is neither economically nor operationally feasible. This is where “pointing and calling” (also known as “shisa kanko” in Japanese), a method developed and practiced in Japan, comes into play.

“Pointing and calling” is also referred to as “finger-pointing confirmation calling” or “finger-pointing vocalization” depending on the industry or workplace. Although performed by a single individual, this method implements a verification process that integrates three senses: “visual confirmation,” “pointing with the finger,” and “vocalization,” effectively serving as a triple-check function. This technique utilizes multiple sensory modalities—visual, kinesthetic, and auditory—enabling a single operator to achieve high verification accuracy.

Historical Background of Pointing and Calling

The widespread adoption of pointing and calling is attributed to steam locomotive engineers of the Japanese National Railways who practiced this method. While not clearly documented in literature, this practice may have been implemented in the railway industry since the Meiji era. It is said that one engineer who had vision problems required his assistant engineer to repeatedly confirm signals using motions similar to pointing and calling. A supervisor in the locomotive department, recognizing the effectiveness of this technique, praised it and gradually institutionalized it as a standard procedure within the organization.

Subsequently, this method was formally incorporated into official manuals during the Japanese National Railways era, and is now employed across a wide range of industries where safety confirmation is critical, including railway, aviation, transportation, construction, manufacturing, and healthcare. In the medical field particularly, the effectiveness of pointing and calling has been demonstrated in preventing medication errors in nursing medication administration tasks and in verifying prescriptions against medications.

Scientific Validation of Pointing and Calling

Experiments by the Railway Technical Research Institute

The Railway Technical Research Institute (RTRI) began scientific verification of the effectiveness of pointing and calling in 1996. For example, data exists investigating how much pointing and calling can suppress errors when performing a switch operation, comparing whether or not the operator performs the action of pointing at and confirming the target.

In the experiment, the error rate in pressing keys was measured under the following four conditions, with 10 subjects performing 200 repetitions each:

1. Pressing the key without any confirmation
2. Pointing before pressing the key
3. Calling before pressing the key
4. Pointing and calling before pressing the key

The results regarding the error occurrence rate in key-pressing performance are shown in the following table:

Confirmation Method	Error Rate	Error Reduction Effect
No action	2.38%	Baseline
Calling only	1.00%	~58% reduction
Pointing only	0.75%	~68% reduction
Pointing and calling	0.38%	~84% reduction

These results demonstrate that when comparing operations with combined pointing and calling versus operations without any confirmation, the error rate is reduced to approximately one-sixth (more precisely, approximately 1/6.3).

Source: Haga, S., Akatsuka, H., & Shirato, H. (1996). Laboratory experiments for verifying the effectiveness of “finger-pointing and call” as a practical tool of human error prevention. Journal of Industrial and Organizational Psychology, 9(2), 107-114.

Neurophysiological Mechanisms of Pointing and Calling

This effect is achieved through brain activation and enhanced attention resulting from pointing at the target, vocalizing its name, and hearing one’s own voice. Neurological studies have confirmed that implementing pointing and calling increases local blood flow in the frontal lobe, improving cognitive function. It is reported that the level of brain activity increases more than threefold compared to operating without any confirmation.

Additionally, according to the “phase theory” in human factors theory, pointing and calling elevates the operator’s consciousness level to “Phase III” (a state where the brain is actively engaged and thinking is proactive), enhancing alertness and concentration.

Five Error Prevention Functions of Pointing and Calling

Research conducted in 2014 has clarified five specific error prevention functions of pointing and calling:

1. Enhanced Visual Accuracy: Pointing at the target object brings the operator closer to it, allowing stimuli to be transmitted to the retina more accurately and clearly
2. Temporal Buffer Creation: The act of pointing takes time, and this brief interval suppresses rushing and prevents impulsive actions
3. Memory Reinforcement: Recalling and verbalizing the name firmly establishes it in working memory
4. Multimodal Cognition: Combining pointing and vocalization integrates visual and auditory senses, enhancing cognitive accuracy regarding the target object
5. Self-Monitoring: Hearing one’s own voice allows unconscious verification that the vocalized content matches the actual target

Inevitability of Human Error and Countermeasures

However, even when implementing pointing and calling, as the experimental results demonstrate, 0.38% of errors persist. This statistical fact carries important implications. In other words, humans will inevitably make mistakes approximately 4 times out of 1,000 operations (0.4%).

This “inevitability of human error” is an extremely important premise in the design of quality management systems. No matter how well-trained operators are, no matter how carefully work is performed, achieving zero errors is impossible. Therefore, to ensure data integrity, the following multi-layered approach is necessary:

1. Error Prevention Measures: Reduce error occurrence rates as much as possible through pointing and calling, standard operating procedures, training, and other methods
2. Error Detection Measures: Detect errors early through double-checks, automated verification systems, statistical process control, and other means
3. Error Response Measures: Appropriately respond to errors that occur and prevent recurrence through deviation management and Corrective and Preventive Action (CAPA) systems
4. Quality Culture Development: Build an organizational culture that encourages reporting rather than concealing errors, promoting continuous improvement

Latest Regulatory Trends and International Standards

Regulatory requirements regarding data integrity have become increasingly stringent in recent years. The latest guidance from major regulatory authorities includes:

• FDA: “Data Integrity and Compliance with cGMP Guidance for Industry” (2018)
• MHRA (Medicines and Healthcare products Regulatory Agency, UK): “GxP Data Integrity Guidance and Definitions”
• WHO: “Guidance on Good Data and Record Management Practices”
• PIC/S: “Good Practices for Data Management and Integrity in Regulated GMP/GDP Environments” (PI 041-1)

These guidances apply to both electronic and paper records, requiring compliance with ALCOA Plus principles. In particular, specific technical requirements are specified, including ensuring audit trails, access control, data backup and recovery, and management in hybrid systems.

Conclusion

Ensuring data integrity is the foundation for securing patient safety and product quality in the pharmaceutical industry. Given that human error is inevitable, it is essential to proactively introduce proven error prevention methods such as pointing and calling, and to promote the integrated advancement of technical countermeasures, process management, and organizational culture improvement. Only through a comprehensive approach based on the concept of defense in depth, rather than relying on a single countermeasure, can true data integrity be achieved.