Medical Device Link.

REGULATIONS & STANDARDS

Raymond J. Skelly is a biotechnology industry manager at Camstar Systems Inc. (Charlotte, NC). He can be reached at rskelly@camstar.com.

Closely coupled enterprise manufacturing and quality management computer systems can lower the costs of monitoring and acting on key business risk areas in IVD product manufacturing. Such lowered costs provide greater incentive to engage in quality by design before full-scale production, and in continuous improvement afterward. Such a holistic approach drives down the total cost of quality throughout an IVD product’s life-cycle and allows innovative new products to achieve targeted quality levels at rates and costs that were not previously possible.

Compelling as it is, the journey toward perfect IVD production that is free of nonconformances, production delays, waste, rework, complaints, corrections, and removals can be expensive and appear financially unjustified. A return on investment exceeding a predetermined hurdle rate must be demonstrated in order to make the business case for such changes. When quality improvement initiatives are examined on a case-by-case basis, such clear cost-justifications for change are not often easily conveyed. Resistance to absorbing the up-front costs of preventing failures can result in poor initial design or deferred continuous improvement. An IVD manufacturer soon finds itself continuing to suffer from sustained high failure costs and being listed among the thousands of annual observations and warnings from governmental health regulating agencies.¹ Worse still, the same perceived high costs of prevention can stifle the lifeblood of any IVD manufacturing firm: rapid innovation.

A Holistic Approach to the Cost of Quality

This problem can be easily solved, but unfortunately is not done frequently enough. The answer lies in taking a holistic approach to the cost of quality. The classical definition of cost of quality breaks down the total cost into three components, prevention, appraisal, and failure:

total cost of quality = costs of (prevention + appraisal + failure).²

In this model, prevention costs are incurred in trying to prevent defects and errors from occurring. This includes the cost of preproduction design activities focused on risk-reduction and postproduction continuous improvement activities. Examples of such costs include total quality management; continuous improvement; Six Sigma and variation risk management; and government-regulated quality improvement methodologies, such as Corrective and Preventive Action (CAPA; 21 CFR Part 820.100 subpart J).³

Appraisal costs arise from determining the current quality of the production system (i.e., quality control [QC] testing or measuring). Failure costs are incurred when nonconformities or other undesirable situations are found either before delivery to a customer (internal failure), or after receipt by a customer (external failure). They can include costs associated with scrap, rework, customer dissatisfaction, added inventory carrying costs, lost capacity, and reduced cycle time.³

For IVD manufacturers, this cost-of-quality model can be restated:

total cost of quality = costs of (prevention + QC + failure).

When depicted graphically in terms of quality level, a hypothetical curve illustrates a significant leverage point.

Figure 1. (click to enlarge) The cost of quality curve for diagnostic and therapeutic products illustrates how TCoQ increases at highest levels of quality by rising costs of QC + Prevention.4 CoF = Cost of Failure; (CoQC + CoP) = Cost of QC + Prevention; TCoQ = Total Cost of Quality.

The combined costs of QC and prevention raise the total cost of quality as the quality level increases (see Figure 1). When pursuing improvements at already high quality levels, the QC costs tend to drive the total cost of quality up exponentially. For example, adding 100% inspections or tests at multiple production stages to raise the quality level will be prohibitive and add new permanent costs to IVD product manufacturing. In fact, such testing quality for an IVD product rarely makes sense financially. The total cost of quality rises due to the added cost of testing without commensurate drops in failure costs.

Alternatively, IVD manufacturers can invest in prevention, hiring, or reassigning resources to identify sources of failures, variance, and trends, analyzing root causes, and improving product and process design. Such steps will improve the quality level and reduce the need for QC.

However, comparing the costs and benefits of these opportunities on their own merits will not provide clear justification for action because the failure costs are already low and dropping fast at higher quality levels. Forces internal and external to an IVD manufacturer will work against implementing such seemingly unjustified initiatives. Few observations will be acted upon. All levels of the company will become accustomed to the business risks and await a precipitating event (often in the form of a severe and newsworthy external failure) to spur them into action.

Shifting the Cost of Quality Curves

Economics teaches about the relationship of supply and demand by also using complementing curves.⁵ In the supply and demand curves, price is plotted against quantity supplied for a given product or service. While movement along the curves is defined and predictable, there are conditions under which they can actually be redrawn. For example, demand reacts to advertising. Supply also reacts to investments in new processes or equipment.

In fact, a common means of changing the dynamics of the supply curve is the application of new technologies.

Breakthrough processes enabled by new technologies shift the entire supply curve to the right because greater quantity can be supplied at lower costs. Rules governing production costs change across the entire range of the curve. At the end of the day, the old paradigm is replaced, and cost calculations are conducted against entirely new assumptions. New technologies have the same potential for the total cost of quality curve for IVD manufacturers.

Figure 2. (click to enlarge) The shifted TCoQ curve continues to drop as quality levels rise. CoF = Cost of Failure; (CoQC + CoP) = Cost of QC + Prevention; TCoQ = Total Cost of Quality.

For example, an IVD manufacturer applies a break­through process that is enabled by a new technology to enhance product quality. The resulting cost of quality control plus cost of prevention curve will shift to the right just as in a supply curve. Costs at many points along the curve can be reduced to a fraction of previous values as prevention costs decrease (see Figure 2). QC costs can drop accordingly as prevention initiatives affect changes in the causes of failure, thereby making ongoing QC testing efforts less critical. Events and observations can translate into investigations and actions with clearer financial justifications on a case-by-case basis, generating the support of all levels of management. With its newly shifted curve, total cost of quality continues to fall as quality levels rise to previously unattainable levels.

All that is required to attain such lower cost of quality is applying a breakthrough process enabled by a new technology, which is easier said than done. How can an IVD manufacturer know what processes or technologies to apply in order to achieve the desired results? The following suggestions looks into some best-in-class manufacturers for ideas.

Manufacturing Technology Solutions

Recent research by Aberdeen Group (Boston) into the cost of quality examined the quality management initiatives of 333 manufacturers. The research suggested the following two best-in-class steps to lower the cost of quality.⁶

The first step is to combine automated data collection with value-chain quality data as actionable intelligence. Merely collecting product data along its value chain of sourcing, production, and distribution provides no benefit in and of itself. What matters is the ability to use such data as information, make decisions driven by them, and act upon such decisions (i.e., actionable intelligence).

The second step is to establish real-time interoperability between plant floor automation (PFA), manufacturing execution systems (MES), quality management systems (QMS), and supply-chain management (SCM). Interoperable computer systems work together on common business processes. PFA refers to the collection of computer systems and devices that control machinery and equip­ment used for manufacturing. An MES manages infor­mation and controls production from order launch to the shipment of finished goods.

QMS does not refer to FDA’s quality system regulation (21 CFR 820) but rather to computer applications managing a broad spectrum of quality management information. This system can include atypical event, trend, nonconformance, complaint and CAPA tracking, investigations, and root-cause analyses. SCM describes computer applications specializing in various aspects of SCM and can include customer relationship management (CRM), supplier relationship management (SRM), and internal supply-chain management (ISCM).⁷

Figure 3. (click to enlarge) Common touch points of enterprise quality management architecture interoperability among average and best-in-class manufacturers.

However, selecting and installing manufacturing information systems as if they were lunch plates from a cafeteria line is not sufficient for success in achieving a dramatic shift in cost of quality. According to Aberdeen Group, “In order for any manufacturer to successfully implement an enterprise quality management architecture and eliminate disconnected quality processes, there must be some level of interoperability between quality systems and technologies. When best-in-class are examined, we find that there are very specific touch points of interoperability that [they] are considerably more likely than average industry manufacturers to leverage (see Figure 3).”⁶

Although popular among best-in-class manufacturers, PFA leaves a lot to be desired when it comes to capturing the full spectrum of manufacturing activities involved in producing IVDs. Investments into such specialized hardware and software solutions as PFA alone yield islands of excellence amidst a sea of potential chaos. While adding a QMS can enable some interoperability with the islands, it fails to model, monitor, or control the remainder of the manual processes on the plant floor. Without MES functionality, which is included in the other interoperable options listed by Aberdeen, an IVD manufacturer is still pushing around a paper device history record (DHR). Assuming a manufacturer can find the record when it is needed, the manufacturing data in a DHR are also subject to the double liability of human error as they are written without automated business rule enforcement and as they are read and transcribed for analysis and reporting.

SCM can affect the bottom line, especially where IVD raw-material costs are high and incoming quality levels are critical to success. MES integration with SCM can allow intimate collaboration with IVD suppliers and customers, providing them with deep and real-time visibility into inventory, enabling automatic replenishment contracts in lieu of an endless stream of purchase orders, and safeguarding against supply-chain whiplash effects due to overzealous customer order buffering.⁸ Of course, if nonconformance reports (NCRs) are maintained using a shop-floor paper system that is invisible to suppliers and customers, the value of such inventory intimacy is significantly diminished. The lot of finished product that they see in inventory may actually be destined for the scrap heap.

MES interoperating with QMS can also provide visibility across the production value chain, as does the pairing of MES with SCM, extending real-time inventory information access to IVD suppliers and customers. This mix further maintains a comprehensive view of the IVD man­ufacturing process, injecting automated business rule enforcement and data capture at predetermined steps. New or changed process models can be reviewed and tested in controlled environments before use in production. Employees, materials, and equipment can be checked for training or qualification prior to conducting critical tasks. Along with their imbedded data, NCRs and DHRs are paperless and accessible across a global enterprise. Traceability, statistical process control, quality dashboards, and audit and complaint management are additional features common in modern pairings of MES and QMS.

Table I. (click to enlarge) Quality improvement before and after manufacturing and quality system computer interoperability.

To illustrate the value for IVD manufacturing, Table I compares some contrasting aspects of common quality-improvement scenarios. An IVD manufacturer should consider how much each aspect costs to execute in its current production environment versus how much it could cost with manufacturing and quality computer system interoperability.

Conclusion

An automation solution pairing MES and QMS functionality can make it easier and cheaper to implement quality improvements at all points in an IVD product’s life­cycle. With easy access to manufacturing data and computer-aided trend and root-cause vanalyses, coupled with a practical means for applying controls to future design and production, failure prevention costs are significantly lower than without such tools.

Facing such reduced quality costs in design preproduction and continuous improvement postproduction, IVD manufacturers are likely to justify and implement more quality-improvement initiatives over time, resulting in lowered failure and quality control costs. The net result is a lower total cost of quality, an improved bottom line, and improved opportunities to accelerate the rate of introduction and exploitation of innovative new IVD products.

---