Automated Screening of Pap Smears Improves Accuracy

Over the past several years, a number of biotechnology companies have developed enhanced screening methods designed to improve Pap smear accuracy. Until recently, however, these screening methods were used only as quality control devices, ie, as a way for laboratories to check their work. Since federal guidelines only require that 10% of all slides be rescreened, there is still an opportunity for error.

In May 1998, the AutoPap Primary Screening System, manufactured by NeoPath, became the first, and to date only, FDA-approved computer-assisted primary Pap smear screening device for use in the early detection of cervical cancer. Because this is a new technology, a brief discussion about automated screening and how it works is warranted before a discussion of how it might help reduce malpractice claims.

Computers are especially suited to Pap smear screening because, unlike human beings, they do not suffer from stress or fatigue, which can result in a missed reading. The AutoPap System uses high-speed video microscopes (see figure) and specialized visual intelligence software to render, analyze, and classify slides in the order of probability that abnormal cells are present. Once the system detects the slide, it reads the bar code and touches the slide to understand the orientation of its surface. The system scans the entire specimen at low power to build a 3D map of the specimen.

Information from each scanned field is then sent to computers for processing, and these images are run through a series of interpretation algorithms to determine the final results. At completion, cytotech-nologists receive a printout of the results and remove those slides that require further review.

In premarket tests, a study of 25,000 Pap smears reviewed under standard laboratory practice and with the AutoPap System found that the computer system helped catch 33% more abnormal cells, including two possible cancers that technicians missed.[1] The study also showed a 16% improvement in the detection of false positives. (While no medical malpractice claim has ever been filed due to a false-positive result, a patient could theoretically claim emotional distress and point to any unnecessary testing or procedure costs.)

Moreover, the AutoPap System uses two unique slide classification algorithms to identify those Pap smears with the highest likelihood of an abnormality. As an initial screener, the AutoPap System identifies up to 25% of slides that receive AutoPap review only and may be archived as normal. The remaining 75% are ranked in order of highest probability of containing abnormal cells. This provides a common benchmark from which cyto-technologists can make a more accurate diagnosis, and allows them to spend more time reviewing those slides that are most likely to contain abnormal cells.

As for errors due to insufficient sampling, the AutoPap System automatically identifies and rejects unsatisfactory slides that need to be returned to physicians, thus reducing the subjective dilemma of whether to let an unsatisfactory slide go through the system.

It should also be noted that the AutoPap Primary Screening System is not intended to replace laboratory slide review processes for "high-risk" slides. High-risk slides are those for which a health care provider has requested special handling because of a specified concern (such as a prior abnormal gynecologic history; previous diagnosis of HPV, dysplasia, or HIV infection; or post-menopausal or vaginal bleeding) or those for which the clinical laboratory, through its own procedures, has identified a need for a high-level screening.[2]

Will the Payers Pay?

While an increasing number of insurance companies are beginning to take notice of this technology and raise reimbursement levels accordingly, many are not yet funding enhanced Pap smear screening methods.

Thus, it can be argued that a physician who does not recommend the use of automated screening has not violated the current standard of care. However, physicians do not have the right to decide what the patient can afford or is willing to pay. Increasing pressure from women’s health advocates, physician groups, and the biotech industry for insurance providers to cover enhanced screening may soon resolve this doctor’s dilemma.

In fact, Congress is currently debating legislation that would amend the Social Security Act to increase the amount of payment under the Medicare program for Pap smear screening effective January 1, 1999.[3]

Although this legislation does not specifically include coverage for enhanced screening options, it might be an important first step in opening the door to that possibility and could help establish a precedent for private insurers to follow.

Alternative Methods

Another automated screening device, PapNet, manufactured by Neuromedical Systems Inc, has been FDA approved only for use in quality control. PapNet uses neural network technology to detect suspicious-appearing cells and digitally enlarge the images of these cells for focused review by cytologists.

While cost should not be the sole determining factor when recommending an enhanced screening method, it should be noted that laboratories generally charge patients an extra $30 to $40 per Pap smear for rescreening with PapNet.[4] In contrast, most Pap smears slides screened with the AutoPap System, either as a primary screen or for quality control, cost an additional $8 to $12, depending on the laboratory.[5]

The cost is lower partly because AutoPap-assisted screening can be performed directly in the laboratory, while PapNet slides must be sent back to the manufacturer’s facilities for review. Sending the slide back to another lab may also jeopardize the integrity of the sample.

Since the AutoPap Primary Screening System is a new technology, it is not yet clear what role it may play in Pap smear litigation cases. Yet, its ability to provide greater accuracy and increased capacity will likely lead to increased use. As that transpires, the standard of care may shift, leading to a change in jury decisions.

A physician’s failure to recommend the use of an enhanced screening method, in and of itself, probably will not result in a win for the plaintiff, although it could be a factor in the overall outcome of the claim. On the other hand, if the physician can show that he or she ordered the most accurate and up-to-date screening method for the patient’s Pap smear, it could increase the chances of an outcome favorable to the defense.

With this in mind, the last article in this series (to appear next month) will provide some recommendations for reducing the risk of a malpractice claim regarding Pap smears.

References:

1. Wilbur DC, Prey MU, Miller WM, et al: The AutoPap System for primary screening in cervical cytology. Acta Cytol 42:214-220, 1998.

2. Product Insert, AutoPap Primary Screening System, NeoPath, Inc, 1998.

3. Yaukey J: Bill would double Medicare reimbursement for Pap smears. Gannett News Service, June 18, 1998.

4. PapNet website, Neuromedical Systems, Inc.

5. "Computer System OK’d for Pap Smears," Los Angeles Times, May 6, 1998.