Why Using Full-Scale Deflection Introduces Data Integrity Risks
In pharmaceutical manufacturing, Accuracy is a critical ALCOA+ principle that can sometimes get lost, particularly when assessing the holistic system. The field instrument level needs to be adequately integrated into the overall assessment, especially in aseptic manufacturing environments where sterilization is critical (and accurate data, such as chamber temperature, is critical). One common (but flawed) practice in instrument calibration is using full-scale deflection (FSD) as the basis for determining accuracy. While it may seem straightforward, this method can introduce significant inaccuracies, particularly in systems like autoclaves where temperature control is vital for achieving microbial inactivation. Omission of the measuring instrument when performing Data Integrity Risk Assessments (DIRAs) can create significant blind spots.
Understanding Full-Scale Deflection
FSD refers to the maximum value an instrument can measure. When accuracy is expressed as a percentage of FSD, the allowable error remains constant across the entire measurement range. For example, if a temperature sensor has a full scale of 200°C and an accuracy of ± 2% FSD, the permissible error is ± 4°C regardless of whether the reading is 50°C or 150°C.
The Problem with FSD-Based Accuracy
The issue arises when measurements are taken at lower values within the instrument’s range. A ± 4°C error at 50°C represents an 8% relative error, which is significantly higher than the 2% error at full scale. This disproportionate error can lead to misleading results, especially in scenarios requiring tight control, such as sterilization.
Autoclaves and the Importance of Temperature Accuracy
Autoclaves are an essential part in aseptic pharmaceutical manufacturing. The standard sterilization temperature for moist heat processes is 121.1°C, which is a critical data element. This measurement is often used to calculate the F₀ value, which measures the cumulative lethality of the sterilization process based on time and temperature.
Since the F₀ calculation is a cumulative measurement over time, even small deviations can significantly affect the calculated lethality. For instance, a sensor with ± 4°C error due to FSD-based calibration could report 121°C when the actual temperature is 117°C, potentially miscalculating the lethality and compromising sterility.
Other Examples
Consider a pressure gauge rated at 200 psig with 1% FSD accuracy. This means the gauge has a potential error of ±2 psi across its range. At full scale (200 psig), the relative error is 2%. But at 10 psig, the same ±2 psig error becomes a 20% relative error, which may be unacceptable for critical data. The DIRA needs to consider the criticality of this data.
Conclusion
FSD for calibration may seem like the right practice, and it is a common one in instrument specifications; however, it’s a practice fraught with risk in accuracy-dependent environments in pharmaceutical manufacturing. It is critical that the instrument calibration assessment process be integrated with the DIRA process to ensure the completeness and accuracy associated with ALCOA+.
Concerned that your DIRA programs aren’t holistic? Contact Lachman Consultants at LCS@LachmanConsultants.com for help assessing whether your firm’s programs are sufficiently integrated or they exist in silos, creating blind spots for your data governance program.
