PAPNET-assisted rescreening of cervical smears: cost and accuracy compared with a 100% manual rescreening strategy

Evolution of Pathological Techniques for the Screening of Cervical Cancer: A Comprehensive Review

The evolutionary journey of cervical cancer screening has been a major medical success story, considering the substantial role it has played in dwindling the disease burden. Through sustained collaborative efforts within the medical community, significant advances have been made from the humble yet path-breaking conventional Pap smear to the current automated screening systems and human papillomavirus (HPV) molecular testing. With the integration of artificial intelligence into screening techniques, we are currently at the precipice of circumventing the pitfalls of manual cytology readings and improving the efficiency of the screening systems by a significant margin. Despite the technological milestones traversed, the high logistics and operational cost, besides the technical know-how of operating the automated systems, can pose a major practical challenge in the widespread adoption of these advanced techniques in cervical cancer screening programs. This would suggest the need to adopt strategies that are tailored to the demands and needs of the different settings keeping their limitations in mind. This review aims to take the reader through the entire evolutionary journey of cervical cancer screening programs, highlight the individual merits and demerits of each technique, and discuss the recommendations from the major global guidelines.

The evolution of cervical cancer screening

There are few medical success stories in history as significant as the reduction in cervical cancer incidence. Through the collaborative efforts of dedicated scientific pioneers, the past century has witnessed remarkable advancement that began with the detection of exfoliated cancer cells through cytologic examination to widespread implementation of cervical cancer screening programs to the discovery of the link between cervical cancer and human papillomavirus (HPV). Current screening methods apply HPV-based testing, and artificial intelligence-based screening systems utilizing digitalized cytology images are being used in a continuous effort to optimize the accuracy and efficiency of the Papanicolaou test. This review summarizes the major milestones in cervical cancer screening history to emphasize its evolution as the World Health Organization aims for the global elimination of cervical cancer.

BestCyte® primary screening of 500 ThinPrep Pap Test thin-layers: 3 Cytologists' Interobserver diagnostic concordance with predicate manual microscopy relative to Truth Reference diagnoses defining NILM, ASCUS+, LSIL+, and ASCH+ thresholds for specificity, sensitivity, and equivalency grading

Background

The BestCyte® Cell Sorter Imaging System (BestCyte) facilitates algorithmic discrimination of clinically relevant cells in Pap test cytopathology by classifying and projecting images of cells in galleries based on cytomorphology. Warranted is awareness of potential BestCyte advantages as measured through 3 cytologists' interobserver diagnostic concordance, specificity and sensitivity differentials, and equivalency grading relative to manual microscopy (MM).

Objectives

Using 500 MM-reported ThinPrep thin-layers, analyze: (1) cytologists' blinded BestCyte screening to raise Bethesda diagnoses; (2) correlate BestCyte and MM diagnoses (i.e., predicate) to establish Truth Reference Diagnoses (TRDx) from concordance between 4 possible diagnoses; (3) analyze cytologists' and MM predicate diagnoses through 4 diagnostic thresholds defined by TRDx: NILM (Negative) for specificity, and ASCUS+, LSIL+, and ASCH+ (Positive) for graded sensitivity (with abnormal cells decreasing in size with increasing dysplasia); and, (4) statistically determine cytologists' equivalency grading to MM using 95% Confidence Interval (CI) ranges.

Results

500 TRDx breakdown (n/%): NILM (241/48.2), ASCUS (79/15.8), ASCH (9/1.80), AGUS (2/0.40), LSIL (86/17.2), HSIL (68/13.6), CA (2/0.40), UNSAT (13/2.60). TRDx breakdown (n/%) per 4 of 4, 3 of 4, 2 of 4 diagnostic concordances: 264 (52.8%), 182 (36.4%), 54 (10.8%), respectively. No cases of discordant diagnoses were recorded. HSIL TRDx were established from 66.2% of 4 of 4 concordances, followed by NILM (59.3%), LSIL (46.5%), ASCUS (41.8%); antithetically, from 4.40% of 2 of 4 concordances. Specificity for MM predicate (NILM): 67.08%; for Cytologists 1, 2, and 3: 89.71%, 82.30%, 97.53%, respectively. For NILM threshold, cytologists revealed Significantly Superior equivalency to MM. Sensitivity for ASCUS+, LSIL+, and ASCH+ thresholds: MM (91.36%, 86.67%, 74.36%); Cytologist 1 (95.88%, 96.97%, 94.87%); Cytologist 2 (95.47%, 95.76%, 93.59%), Cytologist 3 (94.65%, 95.15%, 98.72%), respectively. Cytologists revealed Significantly Superior equivalency to MM for graded Positive thresholds; with Cytologist 3 for ASCUS+ being: Superior.

Conclusions

BestCyte detects and efficiently displays abnormal cells in strategic galleries standardizing objectivity by systematizing mosaics of cell-types for cytologists' consideration. BestCyte fosters consistent, enhanced cytologists' sensitivity values for the ASCUS+, LSIL+, and ASCH+ Positive thresholds relative to MM. Also, BestCyte facilitates improved specificity and superior equivalency grading to MM reflecting efficient screening, and reduced labor. Confident interpretations of small dysplastic epithelial cells characteristic of HSIL led to exceptional interobserver diagnostic concordance inferring BestCyte is primed for effective cervical cancer screening practice.

Inter- and intraobserver agreement in whole-slide digital ThinPrep samples of low-grade squamous lesions of the cervix uteri with known high-risk HPV status: A multicentric international study

BACKGROUND

High-risk human papilloma virus (HR HPV) testing and liquid-based cytology are used for primary cervical screening. Digital cytology, based on whole-slide scanned samples, is a promising technique for teaching and diagnostic purposes. The aim of our study was to evaluate the interobserver and intraobserver variation in low-grade squamous lesions, HR HPV status bias, and the use of whole-slide scanned digital cervical cytology slides.

METHODS

Fifteen expert cytopathologists evaluated 71 digitalized ThinPrep slides (31 atypical squamous cells of undetermined significance [ASC-US], 21 negative for intraepithelial lesion or malignancy, and 19 low-grade squamous intraepithelial lesion cases). HR HPV data were accessible only in the second round.

RESULTS

In interobserver analysis, Kendall's coefficient of concordance was 0.52 in the first round and 0.58 in the second round. Fleiss' kappa values were 0.29 in the first round and 0.31 in the second round. In the ASC-US category, Fleiss kappa increased from 0.19 to 0.22 in the second round and the increase was even higher expressed by Kendall's coefficient: from 0.42 to 0.52. In intraobserver analysis, personal scores were higher in the second round.

CONCLUSIONS

The interobserver and intraobserver variability in low-grade squamous lesions was within fair agreement values in the present study, in line with previous works. The comparison of two rounds showed that expert cytopathologists are generally unbiased by the knowledge of HR HPV data, but that being informed of the HR HPV status leads to a better agreement. Stain quality and back discomfort were highlighted as factors affecting digital cytopathology use.

Detection of False-Negative Pap Smears using the Papnet System

Aims and background

False-negative cytological diagnoses represent the critical point of a screening program for early detection of cervical cancer. Computer-assisted reading using neural network technology has been suggested as a possible approach to manage the problem. The study assessed the performance and the cost-outcome ratio of computer-assisted versus conventional manual Pap smear reading.

Methods

One thousand routine smears, seeded with 81 false-negative smears, were independently interpreted by two readers by conventional and PAPNET-assisted reading. Results of both readings were compared in terms of: a)sensitivity for false-negative smears, b)specificity, and c) cost-outcome (cost per CIN2+ lesion detected).

Results

PAPNET-assisted reading showed a small increase in sensitivity only for one reader. Including the cost of PAPNET, the cost per detected lesion would be $7,543 and the cost per additional detected lesion would be $25,748.

Conclusions

The present study provides further evidence that PAPNET-assisted screening may allow the detection of a few extra cases of CIN2+ lesions with respect to conventional reading, though at a very high cost.

Dr. Bibbo's Presidential Address on Automation in Cytology: Were Her Predictions Right, Wrong, or Somewhere in the Middle?

In 1983, Dr. Marluce Bibbo gave the Presidential Address at the Annual Meeting of the American Society of Cytology in Denver, CO, USA. The lecture was entitled "Analytic and Quantitative Cytology," a field in which Dr. Bibbo was intimately involved. In the presentation, she included a summary of 30 years of work already accomplished, the present state of the art, and musings about issues encountered, potential resolutions, progress that needed to be made, and her perception of how the field needed to evolve in order to become ultimately successful as a clinical service. This commentary looks back 34 years, with observations about Dr. Bibbo's predictions and how the field of cytology automation did actually evolve in the decades following her address. New challenges are identified and possible paths forward are discussed.

Artificial Neural Networks as Decision Support Tools in Cytopathology: Past, Present, and Future

OBJECTIVE

This study aims to analyze the role of artificial neural networks (ANNs) in cytopathology. More specifically, it aims to highlight the importance of employing ANNs in existing and future applications and in identifying unexplored or poorly explored research topics.

STUDY DESIGN

A systematic search was conducted in scientific databases for articles related to cytopathology and ANNs with respect to anatomical places of the human body where cytopathology is performed. For each anatomic system/organ, the major outcomes described in the scientific literature are presented and the most important aspects are highlighted.

RESULTS

The vast majority of ANN applications are related to cervical cytopathology, specifically for the ANN-based, semiautomated commercial diagnostic system PAPNET. For cervical cytopathology, there is a plethora of studies relevant to the diagnostic accuracy; in addition, there are also efforts evaluating cost-effectiveness and applications on primary, secondary, or hybrid screening. For the rest of the anatomical sites, such as the gastrointestinal system, thyroid gland, urinary tract, and breast, there are significantly less efforts relevant to the application of ANNs. Additionally, there are still anatomical systems for which ANNs have never been applied on their cytological material.

CONCLUSIONS

Cytopathology is an ideal discipline to apply ANNs. In general, diagnosis is performed by experts via the light microscope. However, this approach introduces subjectivity, because this is not a universal and objective measurement process. This has resulted in the existence of a gray zone between normal and pathological cases. From the analysis of related articles, it is obvious that there is a need to perform more thorough analyses, using extensive number of cases and particularly for the nonexplored organs. Efforts to apply such systems within the laboratory test environment are required for their future uptake.

The impact of digital imaging in the field of cytopathology

With the introduction of digital imaging, pathology is undergoing a digital transformation. In the field of cytology, digital images are being used for telecytology, automated screening of Pap test slides, training and education (e.g. online digital atlases), and proficiency testing. To date, there has been no systematic review on the impact of digital imaging on the practice of cytopathology. This article critically addresses the emerging role of computer-assisted screening and the application of digital imaging to the field of cytology, including telecytology, virtual microscopy, and the impact of online cytology resources. The role of novel diagnostic techniques like image cytometry is also reviewed.

Comparison of computer-assisted and manual screening of cervical cytology

OBJECTIVE

The Pap smear, introduced over 50 years ago, has significantly contributed to the reduction of mortality due to cervical cancer. The shortage of skilled cytotechnologists to screen and diagnose Pap slides has always been a concern, thus driving the goal to develop an automated system. This study evaluated the diagnostic performance of an automated computer imaging system for routine cervical cancer screening in a high-volume independent laboratory.

METHODS

Validation and training were conducted upon installation of the computer imaging system. Following validation, data were evaluated comparing cytologic detection rates of a six-month cohort of slides screened with computer imaging assistance versus a historic control of manually screened slides.

RESULTS

For each cytologic abnormal category, the Imager-assisted detection rates were significantly greater than the manually screened historic cohort. The Imager increased the detection of HSIL+ by 38% and LSIL by 46% compared to manual screening. There was an increase in the rate of ASC in the Imager cohort (6.5%) compared to manual screening (4.1%), however, the ASC rate decreased during the time of the study period suggesting learning affect.

CONCLUSIONS

The results indicate that computer-imaging-assisted screening significantly increased the cytologic detection of cervical abnormalities compared to manual screening. The initial increase in ASC rates is partially due to a new stain protocol that may be corrected with additional experience. The implementation of the Imager, however, did not adversely affect the ASC:SIL ratio.

Biomedical Research on Health and Performance of Military Women: Accomplishments of the Defense Women's Health Research Program (DWHRP)

In 1994, Congress provided dollar 40 M for biomedical research on issues of importance for military women. This supported 104 intramural and 30 extramural studies and launched an era of research to narrow the knowledge gap on protection and enhancement of health and performance of military women. Projects addressed issues specific to female physiology (e.g., gynecological health in the field, maternal malaria), problems with higher prevalence for women (e.g., marginal iron deficiency, stress fracture), and issues of drug and materiel safety that had only been extrapolated from studies of men (e.g., chemical agent prophylaxis, fatigue countermeasures). Several important assumptions about female physiology and occupational risks were found to be astoundingly wrong. Hormonal changes through the menstrual cycle were less important to acute health risks and performance than predicted, exercise did not increase risk for amenorrhea and consequent bone mineral loss, and women tolerated G-forces and could be as safe as men in the cockpit if their equipment was designed for normal size and strength ranges. Data on personal readiness issues, such as body fat, physical fitness, nutrition, and postpartum return to duty, allowed reconsideration of standards that were gender appropriate and not simply disconnected adjustments to existing male standards. Other discoveries directly benefited men as well as women, including development of medical surveillance databases, identification of task strength demands jeopardizing safety and performance, and greater understanding of the effects of psychosocial stress on health and performance. This surge of research has translated into advances for the welfare of service women and the readiness of the entire force; relevant gender issues are now routine considerations for researchers and equipment developers, and some key remaining research gaps of special importance to military women continue to be investigated.

Assessing the annual economic burden of preventing and treating anogenital human papillomavirus-related disease in the US: analytic framework and review of the literature

The anogenital human papillomavirus (HPV) is estimated to be the most commonly occurring sexually transmitted infection in the US. Comprehensive estimates of the annual economic burden associated with the prevention and treatment of anogenital HPV-related disease in the US population are currently unavailable. The purpose of this paper is to (i) outline an analytic framework from which to estimate the annual economic burden of preventing and treating anogenital HPV-related disease in the US; (ii) review available US literature concerning the annual economic burden of HPV; and (iii) highlight gaps in current knowledge where further study is particularly warranted. Among eight US studies identified that describe the annual economic burden pertaining to one or more aspects of anogenital HPV-related disease, three met the review eligibility criteria (published between 1990 and 2004, examined multiple facets of annual anogenital HPV-related economic burden, and clearly articulated the data and methods used in the estimation process). All costs were adjusted to 2004 US dollars. Estimates of the annual direct medical costs associated with cervical cancer were comparable across studies (range 300-400 million US dollars). In contrast, there was a wide range across studies for estimates of the annual direct medical costs associated with cervical intraepithelial neoplasia (range 700 million US dollars-2.3 billion US dollars). Only one study reported direct medical costs for anogenital warts (200 million US dollars) and routine cervical cancer screening (2.3 billion US dollars). No studies examined direct medical costs attributable to HPV-related anal, penile, vaginal or vulvar cancers, or the work and productivity losses resulting from time spent receiving medical care, morbidity or mortality. Current economic burden estimates would suggest annual direct medical costs associated with the prevention and treatment of anogenital warts and cervical HPV-related disease of at least 4 billion US dollars. This figure would likely rise to at least 5 billion US dollars per year if direct medical costs associated with other disease entities caused by the sexual transmission of HPV were included, with further additions to the economic burden totalling in the billions of dollars if work and productivity losses were incorporated, a research priority for future studies.

State of the Art Symposium: Prescreening and rescreening

The effectiveness of prescreening and rescreening cervical cytology preparations-whether used as quality assurance measures or in the attempt to improve sensitivity-has engendered broad discussion within the cytology community. This paper forms an introduction to some of the issues surrounding prescreening and rescreening in cervical cytology.

The use of morphological characteristics and texture analysis in the identification of tissue composition in prostatic neoplasia

Quantitative examination of prostate histology offers clues in the diagnostic classification of lesions and in the prediction of response to treatment and prognosis. To facilitate the collection of quantitative data, the development of machine vision systems is necessary. This study explored the use of imaging for identifying tissue abnormalities in prostate histology. Medium-power histological scenes were recorded from whole-mount radical prostatectomy sections at x 40 objective magnification and assessed by a pathologist as exhibiting stroma, normal tissue (nonneoplastic epithelial component), or prostatic carcinoma (PCa). A machine vision system was developed that divided the scenes into subregions of 100 x 100 pixels and subjected each to image-processing techniques. Analysis of morphological characteristics allowed the identification of normal tissue. Analysis of image texture demonstrated that Haralick feature 4 was the most suitable for discriminating stroma from PCa. Using these morphological and texture measurements, it was possible to define a classification scheme for each subregion. The machine vision system is designed to integrate these classification rules and generate digital maps of tissue composition from the classification of subregions; 79.3% of subregions were correctly classified. Established classification rates have demonstrated the validity of the methodology on small scenes; a logical extension was to apply the methodology to whole slide images via scanning technology. The machine vision system is capable of classifying these images. The machine vision system developed in this project facilitates the exploration of morphological and texture characteristics in quantifying tissue composition. It also illustrates the potential of quantitative methods to provide highly discriminatory information in the automated identification of prostatic lesions using computer vision.

The health care costs of cervical human papillomavirus--related disease

OBJECTIVE

The purpose of this study was to examine the health care costs of cervical human papillomavirus-related disease in a US health care setting.

STUDY DESIGN

We conducted an observational cohort study using 1997 through 2002 administrative and laboratory records from 103,476 female enrollees of the Kaiser Permanente Northwest health plan (Portland, Ore). We examined the cost per case and annual cost per 1000 enrollees for cervical human papillomavirus-related events.

RESULTS

A cervical examination with a normal routine papanicolaou smear incurred costs of 57 dollars (95% CI, 57-57). Costs that were associated with abnormal routine screening diagnoses ranged from 299 dollars for atypical squamous cells (95% CI, 245-352) to 2349 dollars for high-grade squamous intraepithelial lesion (95% CI, 1,047-3,650). The costs of histologically confirmed cervical intraepithelial neoplasia ranged from 1026 dollars for cervical intraepithelial neoplasia 1 (95% CI, 862-1191) to 3235 dollars for cervical intraepithelial neoplasia 3 (95% CI, 2051-4419); a cost of 376 dollars (95% CI, 315-436) was associated with false-positive test results. At the level of the health plan, overall annual cervical cancer prevention and treatment costs were 26,415 dollars per 1000 female enrollees, with routine cervical cancer screening accounting for expenditures of 16,746 dollars per 1000 female enrollees, cervical intraepithelial neoplasia accounting for expenditures of 4535 dollars per 1000 female enrollees, cervical cancer accounting for expenditures of 2629 dollars per 1000 female enrollees, and false-positive test results accounting for expenditures of 2394 dollars per 1000 female enrollees.

CONCLUSION

These are the first direct estimates of both individual and population level costs of cervical human papillomavirus-related disease in a general US health care setting. Routine cervical cancer screening comprises nearly two thirds of total annual cervical human papillomavirus-related health care costs, with 10% of expenditures dedicated to the treatment of invasive cervical cancer, 17% to the management of cervical precancers, and 9% to dealing with false-positive Papanicolaou test results.

Cost-effectiveness analyses in gynecologic oncology: methodological quality and trends

OBJECTIVES

To evaluate methodological quality and trends of cost-effectiveness analyses (CEA) published in gynecologic oncology.

METHODS

A medical literature search of articles from 1966 through 2002 was performed to identify original, English-language articles that included economic analyses in gynecologic oncology. We included articles that were cost-effectiveness or cost-benefit analyses or performed these analyses as part of their study. Ten methodological principles that should be incorporated in CEAs were assessed for each study. Each article was given a score of 0, 1, or 2 for each of the 10 methodological principles (max score = 20). Data were analyzed using the Student t test, ANOVA, and linear regression.

RESULTS

We screened 693 articles to identify 68 that met our inclusion criteria. The articles focused on cervical cancer (n = 53; 78%), ovarian cancer (n = 11; 16%), uterine cancer (n = 2; 3%), and general perioperative care (n = 2; 3%). The mean (+/-SD) methodological principle score was 16.1 (+/-4.1) and we observed a significant improvement in the total score over time (P = 0.01). Primary CEA's (CEA identified as the objective of the study) were of higher quality than secondary CEA's (primary objective of the study was not CEA but CEA was included in the study; total scores 18.2 vs. 11.6, respectively; P<0.0001). Studies with at least one investigator in public health or healthcare economies also had higher quality (mean total score 17.7 vs. 15.2; P=0.006). The most common limitations of published CEAs were in methodology or presentation of incremental analyses, sensitivity analyses, and discounting.

CONCLUSIONS

Cost-effectiveness analyses in gynecologic oncology showed significant improvement in quality over the last two decades. Despite this progress, methodological improvement is still needed in the areas of incremental comparisons and sensitivity analysis. Understanding the methodology of cost-effectiveness analysis is critical for researchers, editors, and readers to accurately interpret results of the growing body of CEA articles.

Incremental and average cost-effectiveness ratios: will physicians make a distinction?

Physicians are increasingly asked to use cost-effectiveness information when evaluating alternative health care interventions. Little is known about how the way such information is presented can influence medical decision making. We presented physicians with hypothetical screening scenarios with multiple options, varying the type of cost-effectiveness ratios provided as well as whether the scenarios described cancer screening settings that were familiar or unfamiliar. Half the scenarios used average cost-effectiveness ratios, as commonly reported, calculating benefits and costs relative to a no-screening option. The other half used the preferred incremental cost-effectiveness ratios, with each option's benefits and costs calculated relative to the next best alternative. Relative to average cost-effectiveness ratios, incremental cost-effectiveness information significantly reduced preference for the most expensive screening strategies in two of three unfamiliar scenarios. No such difference was found for familiar scenarios, for which physicians likely have established practice patterns. These results suggest that, in unfamiliar settings, average cost-effectiveness ratios as reported in many analyses reported in the literature can hide the often high price for achieving incremental health care goals, potentially causing physicians to choose interventions with poor cost effectiveness.

Systemic therapy for gynecological neoplasms: ovary, cervix and endometrium

Early-stage endometrial carcinomas should be treated by total abdominal hysterectomy and bilateral salpingo-oophorectomy. In a small subset of patients, who wish to have children, conservative treatment (hormonal therapy) could be considered. The most effective agents for palliation of advanced disease are doxorubicin plus cisplatin.

Cost of screening for cancerous and precancerous lesions of the cervix

This paper is part of the cost-effectiveness study of cervical cancer screening conducted by the French Society of Clinical Cytology (SFCC). It describes the evaluation of costs of conventional smear tests, thin-layer smear tests (ThinPrep 2000 system), and viral typing by the HCS test. For 100,000 examinations per year, the average cost of a conventional smear test is 11.53 dollars in a private anatomo-pathology clinic. The cost of the thin-layer test for the same number of examinations and in the same type of clinic is 13.93 dollars. For 20,000 annual tests, the average cost of human papillomavirus (HPV) is 23.43 dollars in the public sector and 23.48 dollars in the private one. The higher price of the thin-layer method is only justifiable if this screening technique outperforms the conventional method. Furthermore, the high cost of the HPV test means that its integration into a population-based screening program must be carefully defined.

Predicting the presence of acute pulmonary embolism: a comparative analysis of the artificial neural network, logistic regression, and threshold models

OBJECTIVE

The objective of this study was to determine whether an artificial neural network, a new data analysis method, offers increased performance over conventional logistic regression in predicting the presence of a pulmonary embolism for patients in a well-known data set.

MATERIALS AND METHODS

Data from the 1064 patients who received an angiographically based diagnosis of pulmonary embolism in the Prospective Investigation of Pulmonary Embolism Diagnosis study were encoded using a previously described method. The 21 input variables represented abnormalities identified on each patient's ventilation-perfusion scan and chest radiograph. Two methods-an artificial neural network with one hidden layer and a multivariate logistic regression-were compared for accuracy in predicting the presence or absence of pulmonary embolism on subsequent pulmonary arteriography.

RESULTS

No significant difference was observed between the two methods. Areas under the receiver operating characteristic curves +/- standard deviation were 0.78 +/- 0.02 for the artificial neural network model and 0.79 +/- 0.02 for the logistic regression model. Furthermore, use of these two methods resulted in no more diagnostic accuracy than did the use of a simple threshold model based only on the number of subsegmental perfusion defects, which was the dominant input variable.

CONCLUSION

In the study population, the usefulness of data from ventilation-perfusion scans as predictors of the presence of a pulmonary embolism was similar for the three analytic methods, a finding that reinforces the importance of making comparisons to simpler or more established methods when performing studies involving complex analytic models, such as artificial neural networks.

Cost effectiveness in new technology in cervix cancer screening

The new techniques in cervix cancer screening are reviewed. Literature relevant to the performance and cost effectiveness of these technologies is reviewed. Literature relevant to cost effectiveness of other techniques used for cervix cancer screening is reviewed. Prospective trials of competing technology are proposed to resolve the question utilizing contemporaneous methods of assessing cost effectiveness.

A review of evidence of health benefit from artificial neural networks in medical intervention

The purpose of this review is to assess the evidence of healthcare benefits involving the application of artificial neural networks to the clinical functions of diagnosis, prognosis and survival analysis, in the medical domains of oncology, critical care and cardiovascular medicine. The primary source of publications is PUBMED listings under Randomised Controlled Trials and Clinical Trials. The rĵle of neural networks is introduced within the context of advances in medical decision support arising from parallel developments in statistics and artificial intelligence. This is followed by a survey of published Randomised Controlled Trials and Clinical Trials, leading to recommendations for good practice in the design and evaluation of neural networks for use in medical intervention.

Cost-effectiveness of monolayers and human papillomavirus testing compared to that of conventional Papanicolaou smears for cervical cancer screening: protocol of the study of the French Society of Clinical Cytology

The French Society of Clinical Cytology is conducting a study to compare the cost-effectiveness of monolayers and human papillomavirus (HPV) testing with that of conventional Papanicolaou (Pap) smears for cervical cancer screening. The protocol of this study is presented. It includes 3,000 women who will be evaluated by the three methods (conventional Pap smears, or monolayers with or without HPV testing) and by the reference method: colposcopy followed, in cases with abnormalities, by cervical biopsy. Efficacy or performance of the methods will be compared on the basis of sensitivity. Cost comparisons and cost-effectiveness modeling will be based on the costs associated with methods themselves and also the costs of "false positives." This will require specific collection of data concerning the costs of the three methods, as these costs have not previously been accurately documented. Patient recruiting and data collection started in September 1999 and will be complete in June 2000. The first results are expected to be available in spring 2001.

Controversies in the management of abnormal Pap smears

Cytology is not 100% accurate. Overdiagnoses and underdiagnoses occur. Changes in classification systems and new devices have not changed the reliability of the Pap smear. The majority of squamous intraepithelial lesions will remit spontaneously and less than 1% will progress to invasive cancer. HPV-DNA testing has not provided a foolproof way to triage lesions. Therefore, the management of abnormal Pap smears is dependent on the personal experience of the physician, who must consider the following questions. What is the accuracy of your cytology laboratory? Does it overdiagnose or underdiagnose? How often has an ASCUS or LSIL become HSIL or invasive cancer on histologic evaluation? Will your patients come back for follow up? Will their health insurance cover frequent follow-up Pap smears or colposcopies? Do you have the resources to follow every patient who needs it? Do you want to avoid the development of every invasive cancer that may develop during the period of follow up?

The impact of information technology on histopathology

Pathology is one of the most computer intensive areas of medicine and as a result diagnostic pathologists in histopathology have often been at the cutting edge of computer literacy. The majority of laboratories use laboratory information systems to issue and store pathology reports. Many of these systems provide the diagnostician with the ability to retrieve reports and cases using coding systems such as SNOMED, but more advanced computer facilities that might assist the pathologist in the diagnosis or interpretation of a case are often lacking. In recent years advances in computer technology have begun to have a much wider impact on the practice of medicine and newer technologies are beginning to find their way into the reporting room. In this review, I cover some of the recent and emerging advances in IT that have the potential to revolutionize the practice of diagnostic histopathology in the next 5 years. The major area of telepathology has been a subject of several recent reviews and will not be covered here.

Human papillomavirus infection in adolescents

The recent gain in knowledge of the genomic function of HPV has led to a greater understanding of the natural history of HPV infection in women, starting from infection to the development of invasive cancer. LSIL is reflective of a benign process associated with HPV replication, and in the majority of women, HPV is eradicated or put into some type of immunologic control so that it remains undetected. In contrast, in the minority of women who have persistent infection, HSIL and invasive cancer are more likely to occur. These findings can be translated clinically to suggest that LSIL can be followed up for a defined period of time and that HPV testing in older women may be useful to identify persistent HPV infections and subsequent risk for invasive cancers.

Papanicolaou smear: can we make a good test better? Technical and interpretive challenges for the practitioner

The Papanicolaou smear has proven to be an excellent screening test for cervical cancer and has resulted in a significant decrease in mortality from invasive cervical cancer. A relatively high false-negative rate of Pap smears might cause failure to diagnose preinvasive disease, especially when coupled with an increased screening interval. This review focuses on the causes of and methods to reduce the number of false-negative Pap smears. In addition, we address the clinical significance of ambiguous Pap smear diagnoses, such as atypical squamous/glandular cells of undetermined significance, and suggest a scheme for management of patients with these diagnoses. Although efforts to reduce false-negative Pap smears are important, the most effective way of reducing mortality from invasive cervical cancer is universal, organized, population-based screening of all women.

Automated cervical cytology: meta-analyses of the performance of the PAPNET system

Our objective was to assess current knowledge regarding PAPNET (Neuromedical Systems, Inc.) automated cervical cytology screening methods and to assess the performance of this automated system in comparison with manual screening. To this goal, studies published in the English language regarding the PAPNET system, identified from a MEDLINE search through August 1998 were selected. Performance of the PAPNET system was assessed with various meta-analysis techniques, using the method of Mantel-Haenszel. In the primary screening modality, meta-analysis of the performance of the PAPNET system indicates that when compared with manual screening, the odds of obtaining a positive result were significantly greater. The Mantel-Haenszel odds ratio for combined studies was 1.19 (95 percent CI = 1.13 to 1.26, P < .001), corresponding to 20 percent greater odds of positive or suspicious slides with PAPNET system. The PAPNET system performs with almost two-fold less false-negative results. The Mantel-Haenszel odds ratio for combined studies was 0.41 (95 percent CI = 0.25 to 0.67, P < .005). Applied as a quality control modality rescreening all consecutive previously manually screened negative slides, depending on study design, the PAPNET system reclassified as abnormal between 0.1 and 5 percent. However, when the PAPNET system was used to rescreen known false-negative slides, PAPNET system rescreening can correctly identify between 20 and 90 percent of manually screened false-negative slides with an average reduction of 33 percent of the manually screened false-negative slides. We conclude that compared with manual screening, PAPNET identifies 20 percent more abnormal, has two-fold less false-negative, and reclassifies as abnormal one third of manually screened false-negative slides.

TARGET AUDIENCE

Obstetricians & Gynecologists, Family Physicians.

LEARNING OBJECTIVES

After completion of this article, the reader will be able to understand how the PAPNET system works and what is its approved use by the FDA, and to understand the associated benefits and shortcomings of the PAPNET system when compared with the traditional screening method.

Automated Papanicolaou smear analysis as a screening tool for female lower genital tract malignancies

The cervical Papanicolaou (Pap) smear is the most effective and inexpensive cancer screening test in medical history for the identification of premalignant cervicovaginal lesions. Despite this fact, false-negative rates ranging from 6% to 55% have been sporadically reported, resulting in an increase in litigation. With the development of computer-based automated technology, a significant proportion of such false-negative results have been reduced by rescreening of negative smears. This article reviews the automated technology approved by the USA Food and Drug Administration and its use, including performance, accuracy and cost-effectiveness compared with the results of conventional manual screening.

Identification of clustered microcalcifications on digitized mammograms using morphology and topography-based computer-aided detection schemes. A preliminary experiment

RATIONALE AND OBJECTIVES

A mathematical morphology-based computer-aided detection (CAD) scheme for the identification of clustered microcalcifications was developed and tested. The potential for improving either sensitivity or specificity by combining the results with those previously reported was investigated.

METHODS

The CAD scheme presented here is based on mathematical morphology and a series of simple rule-based criteria for the identification of clustered microcalcifications. A database of 105 digitized mammograms was used for training and rule setting of the scheme. A test set of 191 digitized mammograms was used to evaluate its performance. The same test set had been used to evaluate a multilayer, topography-based scheme. The results obtained by the two schemes were then combined using logical OR and AND operations.

RESULTS

The morphology-based and topography-based CAD schemes performed at sensitivities of 82.9% and 89.5%, with false-positive detection rates of 1.3 and 0.4 per image, respectively. A logical OR operation resulted in 95.4% sensitivity. An AND operation achieved 76.2% sensitivity, with no false identifications on 93% of images.

CONCLUSIONS

By combining the results of the morphology-based and the topography-based schemes, either sensitivity or specificity can be improved.

Accuracy comparison between PAPNET diagnoses and conventional diagnoses in an Italian cervical cytology laboratory

In a population-based cervical screening laboratory with a controlled workload, a retrospective PAPNET review of 1,654 Pap smears was performed. The series included a random sample of smears originally and conventionally classified as negative (no. 1,309) and unsatisfactory (no. 43), and 299 consecutive smears originally classified as abnormal. The PAPNET diagnoses and the original diagnoses were compared for accuracy. For smears with discordant classification, a majority diagnosis as obtained on manual review was used as a gold standard. The PAPNET testing showed a greater proportion of negative smears classified as unsatisfactory (3.7% vs. 0.2%; ratio, 16.7; 95% CI 12.4-22.0); a reduced proportion of ASCUS/LGSIL cases correctly identified as positive (i.e., a reduced sensitivity for such lesions) (76.3% vs. 97.9%; ratio, 0.78; 95% CI 0.66-0.91); an equal proportion (96.2%) of AGUS/HGSIL cases correctly identified as positive (i.e., an equal sensitivity for such lesions); and an equal proportion of negative smears correctly identified as such (i.e., an equal specificity). In conclusion, the PAPNET testing did not improve the diagnostic performance of the manual screening.

Managed care and assessment of clinical outcomes

While managed care strategies have been associated with reductions in the utilization of clinical resources, their impact on health care outcomes in general, and women's health services, in particular, remains unclear. This review summarizes recent literature regarding the impact of managed care on clinical resource use, outcomes of women's health services, and cost effectiveness of women's health care processes. Implications of these findings for women's health providers, women's health services and policy, and health services research are discussed.