INTRODUCTION

Clinical laboratory test results are used in many clinical settings including screening, case finding, teaching and training, plus research and development.

However, most are used in monitoring individuals over time and in the diagnostic process. There is considerable evidence that the performance characteristics of the tests do affect clinical outcomes in both of these situations.

Performance characteristics involve both practicability and reliability characteristics.

The former include skills required, speed of analysis, plus volume and type of sample required.

The latter include precision, bias, limit of detection, and measuring range. It is often suggested that, for POCT, turnaround time is the most important performance characteristic.

However, quality specifications for the reliability performance characteristics, particularly precision and bias, are vital for creation and management of the analytical quality that impinges on test result interpretation.

However, it is given that quality specifications should be firmly based upon medical requirements, useable in all laboratories irrespective of size, type or location – including POCT, generated using simple-to-understand models, and generally supported by professionals in the field.

For about 40 years, there has been a steady stream of publications concerned with the generation and application of quality specifications [1].

There appeared to be a real conflict about how to set quality specifications, but a decisive recent advance was that a consensus was reached in 1999 on global strategies to set quality specifications in laboratory medicine [2].

This consensus was based upon a hierarchical approach published just prior to the consensus conference [3]. 

The hierarchy is shown below (Table I). All the approaches have advantages and disadvantages, but quality specifications based on components of biological variation (strategy 2A) seem very widely favored and will be discussed in detail in this contribution.

THE EFFECT OF ANALYTICAL PERFORMANCE ON GENERAL CLINICAL OUTCOMES

The second strategy in the hierarchy is the creation of quality specifications based on components of biological variation – within-subject (CV_I) and between-subject (CV_G) variation.

These are directly related to outcomes in monitoring and diagnosis.

Monitoring

Considering monitoring first, probably no one would disagree that analytical random variation must be kept low so that any changes seen in test results in an individual over time are clinically interesting.

In other words, we need to see the “signal” (real change) and not just the “noise” (analytical random variation – precision). For POCT, this is really important because, historically, the analytical performance achieved in alternate sites was not so good as in laboratories and, in consequence, the signal-to-noise ratio was rather low. In consequence, results obtained in POCT settings were of less than desirable quality.

This is important, because it is often said that an advantage of POCT is that patients can be monitored closely and frequently.  

Monitoring involves comparison of serial test results from an individual over time. In the simplest model, changes in serial results can be due to:

• the patient improving
• the patient deteriorating
• preanalytical variation
• biological variation (within-subject) and
• analytical variation – mostly inherent random variation, measured as precision (CV_A)

If preanalytical sources of variation are made as small as possible, then, to assess whether change has occurred, it must exceed the inherent variation due to biological and analytical variation which is now best termed the “reference change value”(RCV) which can be calculated as:

RCV = 2^½ · Z · (CV_A² + CV_I²)^½

where Z is the number of standard deviates appropriate to the probability selected (for example, 1.96 for P < 0.05 and 2.58 for P < 0.01).

Calculation of the effect of precision on medical decision-making is straightforward. If we investigate cholesterol (CV_I ~ 6 %) as one example of a widely done POCT procedure in pharmacies, clinics, physician’s offices and other alternate sites, the change required for significance (at P < 0.05) increases with precision as shown in Table II.

The quality specification advocated for precision is that the analytical variation should be less than one-half the average within-subject biological variation [4]. The rationale for this was expounded by Harris who showed that, if CV_A < 0.50CV_I, then the amount of variability added was about 10 % – said to be “reasonable” [5]. 

This proposal has been very widely accepted by professionals. This concept has been expanded more recently. Three classes of analytical quality (optimum, desirable, and minimum), based upon different fractions of within-subject biological variation, have been proposed as shown in Fig. 1 [6].