Many pharmaceutical companies’ production assets are supported by time-based maintenance programs. Such programs preserve the equipment but usually not its function, and they do not mitigate equipment failure for the balance of a machine’s life cycle. In addition, most time-based systems do not collect data that are needed for failure analysis or criticality analysis. It is thus impossible to determine whether components that fail critically (i.e., shut the equipment or process down) affect the quality of the product. Consequently, all process components are treated as critical, and this approach exposes the organization to needless regulatory scrutiny.

Misperceptions about what makes maintenance compliant lead organizations to adopt exclusively time-based programs. In such programs, the maintenance organization is only allowed to add new preventive maintenance (PM) inspection steps or increase PM frequency. But a vast majority of failures are not effectively addressed by time-based maintenance. After several PM cycles, the equipment often fails again, the process starts over, and the failure is not clearly defined or mitigated. Despite this inefficiency, the maintenance program meets the requirements of 21 CFR Part 67.211. Yet companies are still exposed to avoidable audits, fines, and recalls.

A proactive maintenance approach
In the past four years, publications such as the Good Automated Manufacturing Practice 5 guidelines and the US Food and Drug Administration’s process analytical technology initiative have promoted alternatives to time-based maintenance programs. These publications are intended to encourage innovation and enable greater regulatory flexibility through scientific study and analysis. They suggest that asset maintenance be managed proactively using methods that address known and potential failure modes and improve asset reliability.

The proactive approach originated in 1978, when Nowlan and Heap published Reliability Centered Maintenance (RCM), the results of a 10-year study in the aviation industry (1). This analysis examined equipment function, common failures, the results of those failures, and the controls necessary to mitigate the consequences of failure. RCM showed equipment breakdown cycles to be radically different from what they were thought to be. Before this report, industry assumed that the probability of failure increased as equipment aged and that reliability could be ensured by equipment replacement or overhaul. Nowlan and Heap identified six different equipment behavior curves for equipment failure and found that most failures have no relation to age (see Figure 1).

Figure 1: Equipment breakdown cycles.

The outline of a proactive maintenance approach
Because most equipment fails at random ages, equipment should be examined regularly using predictive technologies such as thermography, vibration, ultrasonics, and oil analysis. For example, all laminar-flow blower units should be subject to a vibration-analysis program, and maintenance should only shut down a cleanroom when it finds a condition that creates failure. Time-based PMs should primarily be used to measure and identify age-related deterioration. A well-managed planning and scheduling process should ensure that the corrective actions are performed in an efficient manner.

This kind of program minimizes exceptions and investigations for equipment failure. The maintenance group sends the quality and operations groups weekly or daily reports, and important findings for each piece of equipment are stored in the maintenance-management system database. This program increases the equipment’s capacity and reduces its costs of operation.

A proactive maintenance model fits perfectly into the corrective and preventive action (CAPA) process. Many companies from other industries have already adopted proactive maintenance programs. The programs’ key principles are the following:

1. Start at the beginning. When qualifying a new asset, start planning the maintenance program for it during validation. Develop a maintenance program to support each machine’s expected life cycle and anticipated failure modes. This technique enables a reliable maintenance program that supports the CAPA process.
2. Develop measurable assessments of asset health. Make sure the tasks supporting production assets address the failure modes. This information comes from RCM or failure modes and effects analysis. 
3. Apply technology to the CAPA process. More information can be obtained through predictive technologies than through human observation.

Educate and collaborate
To develop an effective maintenance program, a maintenance organization must involve the quality and operations groups in the process, educate them about active maintenance concepts, and prepare a plan of action for the entire company. When quality and operations organizations are involved, educated, and trained in the methods that will be employed within the reliability program, the company will see a return on its investment. Because the pharmaceutical industry relies so much on its production assets, it has a strong incentive to adopt a focus on reliability, which can increase efficiency, reduce costs, and minimize regulatory problems.

Reference
1. F.S. Nowlan and H.F. Heap, “Reliability-Centered Maintenance,” DOD report number A066-579, (US Department of Defense, Washington, DC, Dec. 29, 1978).

Jason T. Price is a consultant at Management Resources Group, 555 Heritage Rd., Southbury, CT  06488, tel. 203.264.0500, fax 203.267.1493, pricej@mrginc.net.