Abstract and Introduction
Sample adequacy is an important aspect of overall fine-needle aspiration cytology (FNAC) performance. FNAC effectiveness is augmented by an increasing number of needle passes, but increased needle passes are associated with higher costs and greater risk of adverse events. The objective of this study was to compare the impact of several different sampling policies on FNAC effectiveness and adverse event rates using discrete event simulation. We compared 8 different sampling policies in 12 different sampling environments. All sampling policies were effective when the per-pass accuracy is high (>80%). Rapid on-site evaluation (ROSE) improves FNAC effectiveness when the per-pass adequacy rate is low. ROSE is unlikely to be cost-effective in sampling environments in which the per-pass adequacy is high. Alternative ROSE assessors (eg, cytotechnologists) may be a cost-effective alternative to pathologists when the per-pass adequacy rate is moderate (60%–80%) or when the number of needle passes is limited.
Fine-needle aspiration cytology (FNAC) is a widely used sampling technique. FNAC is a complex multistep process, and the overall success depends on a wide range of factors such as the anatomic location, characteristics of the lesion, size of the lesion, skill of the aspirator, use of ancillary imaging techniques, and the use of rapid on-site evaluation (ROSE). Studies often find wide variation in FNAC performance across study sites; however, because of the variability in non–FNA-related factors between study sites, it is often difficult to make direct comparisons between studies. It is often difficult to design controlled studies with sufficient power to detect small changes in performance. In addition, most studies on FNAC performance are observational or retrospective rather than randomized controlled trials, so confounding factors are usually not randomly distributed across study arms.
Sample adequacy is an important component of FNAC performance and is frequently reported in FNAC studies. In general, sample adequacy is expected to increase with the number of needle passes, and the effect of the number of passes has been investigated in a small number of studies. However, the risk of adverse events also increases with the number of needle passes, and the benefits of an adequate sample must be balanced against the potential costs of an adverse event (both in terms of patient harms and monetary costs). The trade-off between adverse events and successful FNAC can be modified by the use of ROSE, but the relationship between effectiveness and adverse events has not been demonstrated in clinical studies. It is unlikely that a single study could obtain a sufficient number of cases to fully reveal these relationships.
Sampling is a process with uncertain outcomes (a sample has 2 possible outcomes, adequate or inadequate, and the result is not known until after the sample is taken) that are governed by the laws of probability. It is therefore amenable to mathematical modeling approaches. Modeling has the advantage that it is not subject to the site-to-site variation that complicates clinical studies. Modeling studies can reveal the relationships that could not be otherwise examined. In particular, simulation modeling provides the opportunity to explore complex processes that are not amenable to simple closed-form mathematical solutions. We therefore used simulation modeling to compare the trade-off between adequacy and adverse events for several different methods of FNAC sampling.