Special Feature

Computerization of Pap Smear Screening

By Kenneth L. Noller, MD

During the past several months, both the lay public and physicians have become aware that there are new methodologies being proposed (and in some cases already implemented) for changes in the laboratory techniques used for the interpretation of cervical cytology (Pap smear) specimens. Unfortunately, much of the information in the lay press has been incorrect or misleading, and many physicians have not yet been fully informed of the actual "state of the art." The three currently available computerized techniques are highlighted in this article. However, the field is changing so rapidly that this information will likely be out-of-date within a few months.

Slide Preparation

The Cytyc Corporation has developed a technique called Thin Prep. The purpose of this technique is to provide a smear to cytotechnologists that can be interpreted more easily and correctly than with the standard preparation. The technique has now been approved by the federal regulation agencies and is available for use by physicians.

If a physician chooses to use this technique, the specimen must be obtained with a "paddle" supplied by Cytyc. The paddle is then placed in a solution (again, supplied by the manufacturer), and the vial is sent to the laboratory.

Most of the mucus is dissolved by the preparation technique, and individual cells can be visualized more clearly than with the standard preparation. Cytyc is currently authorized to state that more cervical precancers can be detected through the use of this system than the standard technique.

Theoretically, the vial of fluid might be useful for Chlamydia, Neisseria, or HPV testing.

While the technique is not yet available nationwide, some laboratories have begun to obtain the equipment necessary to be able to prepare Thin Prep Pap smear slides. Of course, there is an additional cost for the preparation of slides in this manner.

Computerized Slide Sorting

The Neopath Corporation has developed a computerized technique for the sorting of cervical cytology specimens. Stacks of Pap smears are presented to the "machine," which rapidly scans them. A set percentage of slides (usually 10-30%) is interpreted by the computer as being completely normal, and these receive no further screening. The remaining slides are reviewed by a cytotechnologist in the routine manner.

This technology may be very important to very large Pap smear laboratories. With this technology, a larger number of slides can be handled by the same number of cytotechnologists. Because the computerized technology is expensive, it is likely that only relatively large laboratories will invest in it. As of today, Neopath has not yet received approval for use of this technique for primary screening. However, they have applied to the FDA for such approval.

Computerized Abnormal Cell Identification

Neuromedical Systems Incorporated and PapNet Incorporated have developed a computerized system that uses neuronetwork technology to identify abnormal cells on a Pap smear. With this technique, a slide is prepared in the routine manner and is then screened by computer. The computer identifies the 128 most abnormal appearing cells on the slide and saves these images in memory. A cytotechnologist later reviews the images and determines whether the cells represent true disease. By carefully indexing the slide, the cytotechnologist can find the actual cells on the slide relatively easily. Because of the neuronetwork programming, the computer actually "learns" to identify abnormal cells more accurately the more samples it reviews.

So far, the PapNet system has been approved only for re-review of slides that were originally read as negative by standard cytology laboratory techniques. Thus, it is not possible to send a slide from a physician’s office to PapNet for interpretation. Rather, a physician must send the slide through usual channels and, if it is negative, the slide can then be retrieved and sent to PapNet for rescreening. Of course, there is a significant additional charge for PapNet interpretation.

Unfortunately, PapNet has chosen to market its service directly to the lay public, primarily through advertisements in periodicals that are primarily purchased and read by women. While the wording in the ads is correct and carefully states that PapNet may only be used for repeat screening, it still has confused many patients and physicians. I have been asked many times by patients to send their slides to PapNet for interpretation instead of a routine laboratory. They are usually upset with the company when I explain the actual regulations.

This technology seems to hold promise for the future if it ever becomes available for primary screening at a reasonable cost. However, as long as it remains a technique used only for re-review, the cost-effectiveness will remain unclear.


It is interesting to note that, despite all of the media hype, all three of the currently available "computerized" technologies still require that a cytotechnologist review slides. So far, human input is still essential. We are very far from having a system that works like most laboratory processors—sample in one end, result out the other. Indeed, it never may be possible to fully automate the interpretation of cervical cytology specimens.