Terms Of Positive Predictive Value Research Articles

Detecting protein complexes based on a combination of topological and biological properties in protein-protein interaction network.

Protein complexes are known to play a major role in controlling cellular activity in a living being. Identifying complexes from raw protein protein interactions (PPIs) is an important area of research. Earlier work has been limited mostly to yeast. Such protein complex identification methods, when applied to large human PPIs often give poor performance. We introduce a novel method called CSC to detect protein complexes. The method is evaluated in terms of positive predictive value, sensitivity and accuracy using the datasets of the model organism, yeast and humans. CSC outperforms several other competing algorithms for both organisms. Further, we present a framework to establish the usefulness of CSC in analyzing the influence of a given disease gene in a complex topologically as well as biologically considering eight major association factors.

Finding optimum width of discretization for gene expressions using functional annotations

Discretizing gene expression values is an important step in data preprocessing as it helps in reducing noise and experimental errors. This in turn provides better results in various tasks such as gene regulatory network analysis and disease prediction. A supervised discretization method for gene expressions using gene annotation is developed. The method is called “Gene Annotation Based Discretization” (GABD) where the discretization width is determined by maximizing the positive predictive value (PPV), computed using gene annotations, for top 20,000 gene pairs. The method can capture the gene similarity better than those obtained using original expressions. The performance of GABD is compared with some existing discretization methods like equal width discretization, equal frequency discretization and k-means discretization in terms of positive predictive value (PPV). The utility of GABD is also shown by clustering genes using k-medoid algorithm and thereby predicting the function of 23 unclassified Saccharomyces cerevisiae genes using p-value cut off 10−10. The source code for GABD is available at http://www.sampa.droppages.com/GABD.html.

Protein complex finding and ranking: An application to Alzheimer's disease.

Protein complexes are known to play a major role in controlling cellular activity in a living being. Identifying complexes from raw protein-protein interactions (PPIs) is an important area of research. Earlier work has been limited mostly to yeast and a few other model organisms. Such protein complex identification methods, when applied to large human PPIs often give poor performance. We introduce a novel method called ComFiR to detect such protein complexes and further rank diseased complexes based on a query disease. We have shown that it has better performance in identifying protein complexes from human PPI data. This method is evaluated in terms of positive predictive value, sensitivity and accuracy. We have introduced a ranking approach and showed its application on Alzheimer's disease.

Optimization of the French cystic fibrosis newborn screening programme by a centralized tracking process.

ObjectivesTo evaluate the French cystic fibrosis newborn screening algorithm, based on data tracked by a centralized monitoring process, from 2002 to 2014. The programme aimed to attain European Standards in terms of positive predictive value, sensitivity, the ratio of screen positive patients diagnosed with cystic fibrosis to infants who screen positive but with inconclusive diagnosis (CFSPID), and time to diagnosis.MethodsRetrospective analysis of programme performance, compliance with the algorithm, and changes in screening strategy.ResultsModifications in the flow chart protocol improved the positive predictive value to 0.31 while maintaining the sensitivity at 0.95. Among infants diagnosed with cystic fibrosis, or identified as CFSPID, sweat test results were obtained for 94%, and two mutations were identified after exhaustive screening for the gene, when applicable, in 99.6%. The rate of pending diagnosis was very low (0.5%). The ratio of infants with cystic fibrosis:CFSPID was 6.3:1. Age at initial visit at the CF centre was ≤ 35 days, respectively, in 53%/26%.ConclusionPerformances were in agreement with European standards, but timeliness of initial visit needed improvement. Our data complement an accumulating body of evidence demonstrating that attention must be paid to such ethical considerations as limiting carrier detection and inconclusive diagnosis. Newborn screening programmes should have a rigorous centralized monitoring process to warrant adjustments for improving performance to attain consensus guidelines.

Altools: a user friendly NGS data analyser.

BackgroundGenotyping by re-sequencing has become a standard approach to estimate single nucleotide polymorphism (SNP) diversity, haplotype structure and the biodiversity and has been defined as an efficient approach to address geographical population genomics of several model species. To access core SNPs and insertion/deletion polymorphisms (indels), and to infer the phyletic patterns of speciation, most such approaches map short reads to the reference genome. Variant calling is important to establish patterns of genome-wide association studies (GWAS) for quantitative trait loci (QTLs), and to determine the population and haplotype structure based on SNPs, thus allowing content-dependent trait and evolutionary analysis. Several tools have been developed to investigate such polymorphisms as well as more complex genomic rearrangements such as copy number variations, presence/absence variations and large deletions. The programs available for this purpose have different strengths (e.g. accuracy, sensitivity and specificity) and weaknesses (e.g. low computation speed, complex installation procedure and absence of a user-friendly interface). Here we introduce Altools, a software package that is easy to install and use, which allows the precise detection of polymorphisms and structural variations.ResultsAltools uses the BWA/SAMtools/VarScan pipeline to call SNPs and indels, and the dnaCopy algorithm to achieve genome segmentation according to local coverage differences in order to identify copy number variations. It also uses insert size information from the alignment of paired-end reads and detects potential large deletions. A double mapping approach (BWA/BLASTn) identifies precise breakpoints while ensuring rapid elaboration. Finally, Altools implements several processes that yield deeper insight into the genes affected by the detected polymorphisms. Altools was used to analyse both simulated and real next-generation sequencing (NGS) data and performed satisfactorily in terms of positive predictive values, sensitivity, the identification of large deletion breakpoints and copy number detection.ConclusionsAltools is fast, reliable and easy to use for the mining of NGS data. The software package also attempts to link identified polymorphisms and structural variants to their biological functions thus providing more valuable information than similar tools.ReviewersThis article was reviewed by Prof. Lee and Prof. Raghava.Open peer reviewReviewed by Prof. Lee and Prof. Raghava. For the full reviews, please go to the Reviewers’ comments section.Electronic supplementary materialThe online version of this article (doi:10.1186/s13062-016-0110-0) contains supplementary material, which is available to authorized users.

Vessel transform for automatic optic disk detection in retinal images

Precise localisation of an optic disk (OD) in the retinal images is one of the most important problems in the ophthalmic image processing. Although a considerable progress has been made towards a computerised solution of the problem, the numerical algorithms often fail on retinal images characterised by poor quality. Therefore, the authors propose a new method suitable for low-quality images based on exploiting the convergence of the blood vessels to the OD. The novelty of the proposed techniques includes clustering the vessels endowed with a novel correction procedure and the vessel transform (VT) which measures the distance to the main clusters. The algorithm is integrated into the scale-space (SS) analysis to detect the boundary of the OD. The integrated method is called SS algorithm with VT (SSVT). SSVT has been tested on retinal images from two databases with fair and poor images against the fuzzy convergence (FC) method and a modification of the circular transform proposed by Lu. The absolute improvement on sensitivity of SSVT against FC and Lu's are up to 12.37% and 8.18%. Bigger improvements of SSVT in terms of positive predictive value are up to 37.46% and 30.84% against FC and Lu's, respectively.

Gender Differences in Fecal Immunochemical Test Performance for Early Detection of Colorectal Neoplasia

Fecal immunochemical tests (FITs) are used widely in colorectal cancer screening. Programs use the same fecal hemoglobin threshold for colonoscopy referral for men and women, but it is unclear whether FIT performs equally in both sexes. We therefore assessed FIT performance in men and women. A prospective cohort study was performed, in which a total of 10,008 average-risk subjects (age, 50-74 y) were invited for first-round screening and 8316 average-risk subjects (age, 51-74 y) were invited for second-round screening with a single FIT. Subjects with a hemoglobin (Hb) level of 10 μg hemoglobin (Hb)/g (or ≥50 ng/mL) feces or higher were referred for colonoscopy. The test characteristics were assessed by sex for a range of FIT cut-off values. In total, 59.8% of men and 64.6% of women participated in the first round (P < .001). At a cut-off level of 10 μg Hb/g feces, the positivity rate was significantly higher among men (10.7%)compared with women (6.3%; P < .001) in the first round. The detection rate of advanced neoplasia was 4.4% for men and 2.2% for women (P < .001) in the first round. The positive predictive value for advanced neoplasia in the first round was 42% for men and 37% for women (P= .265). A significantly higher false-positive rate in men (6.3%) than in women (4.1%; P < .001) was found. Similar differences in these test characteristics were seen in the second round. At a cut-off level of 10 μg Hb/g feces the FIT positivity rate was higher in men, reflected by both a higher detection rate and a higher false-positive rate. The use of the same cut-off value in men and women in FIT screening is recommended based on equal test performance in terms of positive predictive value.

Benchmarking mutation effect prediction algorithms using functionally validated cancer-related missense mutations

BackgroundMassively parallel sequencing studies have led to the identification of a large number of mutations present in a minority of cancers of a given site. Hence, methods to identify the likely pathogenic mutations that are worth exploring experimentally and clinically are required. We sought to compare the performance of 15 mutation effect prediction algorithms and their agreement. As a hypothesis-generating aim, we sought to define whether combinations of prediction algorithms would improve the functional effect predictions of specific mutations.ResultsLiterature and database mining of single nucleotide variants (SNVs) affecting 15 cancer genes was performed to identify mutations supported by functional evidence or hereditary disease association to be classified either as non-neutral (n = 849) or neutral (n = 140) with respect to their impact on protein function. These SNVs were employed to test the performance of 15 mutation effect prediction algorithms. The accuracy of the prediction algorithms varies considerably. Although all algorithms perform consistently well in terms of positive predictive value, their negative predictive value varies substantially. Cancer-specific mutation effect predictors display no-to-almost perfect agreement in their predictions of these SNVs, whereas the non-cancer-specific predictors showed no-to-moderate agreement. Combinations of predictors modestly improve accuracy and significantly improve negative predictive values.ConclusionsThe information provided by mutation effect predictors is not equivalent. No algorithm is able to predict sufficiently accurately SNVs that should be taken forward for experimental or clinical testing. Combining algorithms aggregates orthogonal information and may result in improvements in the negative predictive value of mutation effect predictions.

Stress echocardiography has superior accuracy and is more cost effective for predicting coronary artery disease in patients with suspected angina compared to exercise ECG

Purpose: Exercise ECG (ExECG) is widely used to assess patients with suspected coronary artery disease (CAD). However stress echocardiography (SE) is a well-established alternative technique for the assessment of these patients. We hypothesised that SE, due to its greater accuracy and feasibility, may be superior to ExECG, in terms of positive predictive value and cost effectiveness, when used as the initial test for the assessment of patients with no previous history of CAD presenting with suspected stable angina. Methods: Patients referred with recent onset chest pain during 2011, with no known history of CAD and with a pre-test likelihood of CAD greater than 10%, who underwent first line ExECG or SE were identified. The tests were classified as positive, negative or inconclusive for ischaemia. Coronary angiography (CA) was performed, with the knowledge of the functional test results, as indicated clinically. CAD was defined as the presence of >50% stenosis in at least one major epicardial vessel on CA. Cost to diagnosis of CAD was determined for each functional test by adding the initial cost, of the test to the cost of subsequent tests leading to and including CA. Follow up data on hard events, cardiac death and acute myocardial infarction (AMI), were collected on all patients for a period up to one year after the presentation of the last study patient. Results: We identified 457 patients who underwent ExECG (225 (49%) negative, 94 (21%) positive, 138 (30%) inconclusive) and 257 who underwent SE (213 (83%) negative, 38 (15%) positive, 6 (2%) inconclusive) as first line. The mean pre-test probability of CAD was 43±26% vs 51±28% respectively. Of 74 patients referred for CA on the basis of ExECG, CAD was present in 35 (47%) which was significantly (p=0.04) lower than in the SE group which predicted CAD in 28 out of 42 patients referred for CA (67%). The mean cost to diagnosis was £460 for the ExECG versus £385 for the SE group (p=0.02). No cardiac deaths were found in either group. There was no difference in event rate, (p=0.8 by log rank test) of AMI in patients who were discharged on the basis of a negative ExECG (4 (2%)) versus those with a negative SE (4 (2%)) over a period of 18±6 months. Conclusions: In a population with intermediate pre-test probability of CAD, SE was more accurate and more cost effective for predicting the presence of CAD compared to ExECG. This study suggests that SE rather than ExECG should be the initial test for the assessment of suspected stable angina in this population.

Comparing the accuracy of two secondary food environment data sources in the UK across socio-economic and urban/rural divides

BackgroundInterest in the role of food environments in shaping food consumption behaviours has grown in recent years. However, commonly used secondary food environment data sources have not yet been fully evaluated for completeness and systematic biases. This paper assessed the accuracy of UK Points of Interest (POI) data, compared to local council food outlet data for the county of Cambridgeshire.MethodsPercentage agreement, positive predictive values (PPVs) and sensitivities were calculated for all food outlets across the study area, by outlet type, and across urban/rural/SES divisions.ResultsPercentage agreement by outlet type (29.7-63.5%) differed significantly to overall percentage agreement (49%), differed significantly in rural areas (43%) compared to urban (52.8%), and by SES quintiles. POI data had an overall PPV of 74.9%, differing significantly for Convenience Stores (57.9%), Specialist Stores (68.3%), and Restaurants (82.6%). POI showed an overall ‘moderate’ sensitivity, although this varied significantly by outlet type. Whilst sensitivies by urban/rural/SES divides varied significantly from urban and least deprived reference categories, values remained ‘moderate’.ConclusionsResults suggest POI is a viable alternative to council data, particularly in terms of PPVs, which remain robust across urban/rural and SES divides. Most variation in completeness was by outlet type; lowest levels were for Convenience Stores, which are commonly cited as ‘obesogenic’.

Optimization of a quantitative signal detection algorithm for spontaneous reports of adverse events post immunization

To optimize the efficiency of signal detection by maximizing the proportion of true positive (TP) signals among signals detected by a disproportionality algorithm. We compared 176 different combinations of stratification factors, sex (S), age (A), region (R) and year of report (Y), and cut-off values of a Multi-Item Gamma Poisson Schrinker (MGPS) algorithm. Spontaneous adverse event reports of eight vaccines from the GlaxoSmithKline Biologicals safety database were used. Defining events listed in the Product Information as proxy of true safety signals, we compared each algorithm performance in terms of positive predictive value (PPV). For each vaccine, each algorithm was ranked according to PPV. Median rank and overall PPV were computed across vaccines. For a standard cut-off of 2, the optimal stratification factors differed by vaccine and led to a set of algorithms with a median rank of 34.5 (PPV = 0.22; 34 TP). Keeping the original SARY stratification led to optimal cut-offs that differed by vaccine and a set of algorithms with a median rank of 1.75 (PPV = 0.20; 142 TP). The optimal combination of cut-off and stratification led to different algorithms by vaccine with a median rank of 1 (PPV = 0.19; 139 TP). The best unique algorithm parameterization across vaccines was 0.8-SARY (cut-off-stratification), with a median rank of 3 (PPV = 0.20; 195 TP). The original 2-SARY was one of the worst algorithms, with a median rank of 150.75 (PPV = 0.13; 8 TP). Within the scope of this study, a unique MGPS algorithm across vaccines with the original full stratification but a lowered cut-off provided major performance improvement.

Wavelet-based detection of transcriptional activity on a novel Staphylococcus aureus tiling microarray

BackgroundHigh-density oligonucleotide microarray is an appropriate technology for genomic analysis, and is particulary useful in the generation of transcriptional maps, ChIP-on-chip studies and re-sequencing of the genome.Transcriptome analysis of tiling microarray data facilitates the discovery of novel transcripts and the assessment of differential expression in diverse experimental conditions. Although new technologies such as next-generation sequencing have appeared, microarrays might still be useful for the study of small genomes or for the analysis of genomic regions with custom microarrays due to their lower price and good accuracy in expression quantification.ResultsHere, we propose a novel wavelet-based method, named ZCL (zero-crossing lines), for the combined denoising and segmentation of tiling signals. The denoising is performed with the classical SUREshrink method and the detection of transcriptionally active regions is based on the computation of the Continuous Wavelet Transform (CWT). In particular, the detection of the transitions is implemented as the thresholding of the zero-crossing lines. The algorithm described has been applied to the public Saccharomyces cerevisiae dataset and it has been compared with two well-known algorithms: pseudo-median sliding window (PMSW) and the structural change model (SCM). As a proof-of-principle, we applied the ZCL algorithm to the analysis of the custom tiling microarray hybridization results of a S. aureus mutant deficient in the sigma B transcription factor. The challenge was to identify those transcripts whose expression decreases in the absence of sigma B.ConclusionsThe proposed method archives the best performance in terms of positive predictive value (PPV) while its sensitivity is similar to the other algorithms used for the comparison. The computation time needed to process the transcriptional signals is low as compared with model-based methods and in the same range to those based on the use of filters. Automatic parameter selection has been incorporated and moreover, it can be easily adapted to a parallel implementation. We can conclude that the proposed method is well suited for the analysis of tiling signals, in which transcriptional activity is often hidden in the noise. Finally, the quantification and differential expression analysis of S. aureus dataset have demonstrated the valuable utility of this novel device to the biological analysis of the S. aureus transcriptome.

Retraction

Listeria monocytogenes poses a serious threat to public health, and the majority of cases of human listeriosis are associated with contaminated food. Reliable microbiological testing is needed for effective control of this pathogen by the food industry and competent authorities. The aim of this study was to determine the performance of three culture media commonly used for detecting L. monocytogenes in foods. Minced pork meat samples (n = 100) were subjected to microbiological testing for L. monocytogenes according to International Organization for Standardization methods 11290-1:1996 and 11290-2:1998 using PALCAM, ALOA, and RAPID'L. mono culture media in parallel. Presence of the pathogen was confirmed by conducting biochemical and molecular tests on the presumptive L. monocytogenes colonies. Performance attributes of sensitivity, specificity, positive and negative predictive values, positive and negative likelihood ratios, diagnostic odds ratios, error odds ratios, receiving operating characteristic (ROC) curve, and area under this curve were calculated from the presence-absence microbiological test results by combining the results obtained from the culture media and confirmative tests. PALCAM had the best performance in terms of positive predictive value (i.e., a positive result indicates high probability of L. monocytogenes presence) but not in terms of sensitivity (i.e., the ability of the medium to detect the pathogen when present). RAPID'L. mono was the most sensitive medium. None of the culture media were perfect for detecting L. monocytogenes in minced pork meat alone. The pathogen was detected in 16, 19, and 26% (apparent prevalence) of the samples by PALCAM, ALOA, and RAPID'L. mono, respectively, although the true prevalence of the pathogen was 22%. These findings indicate that the use of a single culture medium may lead to erroneous determination of the prevalence of L. monocytogenes.

Coverage and Performance of Colorectal Cancer Screening with the Faecal Occult Blood Test in Finland

Mortality from colorectal cancer has been shown to decrease by repeated screening using faecal occult blood (FOB) testing in randomized screening trials. This report presents coverage and performance of organized screening among the general population in Finland. In 2004-2007, people aged 60-69 years were randomized into biennial screening and control arms. The screening test was a guaiac-based FOB test (Hemoccult) with dietary restriction and three test cards for six consecutive samples. Test positives were referred for full colonoscopy. The programme was launched in 2004 and subsequently it expanded over regions and age-cohorts. In 2007, the programme covered one-third of the target population and 74,592 people had been invited for screening, of them 26,866 for the second round. Uptakes for the first and second rounds, respectively, were 62% and 68% in men and 77% and 80% in women. The proportion of test positives increased from 2.4% to 2.9% from the first to the second round and the positive predictive value for cancers decreased from 7.5% to 4.3%. By 2007, organized colorectal cancer screening covered one-third of the target population in Finland. Implementation of screening measured with response rate was successful and met the criteria for a public health programme, but performance in terms of positive predictive value needs monitoring.

Institutional experience of PTH evaluation on fine-needle washing after aspiration biopsy to locate hyperfunctioning parathyroid tissue

Assaying parathyroid hormone (PTH) in the washing liquid after fine-needle aspiration biopsy (FNAB) seems to be a valid approach to locate parathyroid tissue. PTH-FNAB was evaluated in 47 patients with a clinical picture of primary hyperparathyroidism (PHP) and ultrasonography (US) suggestive of parathyroid lesion. The patients were subdivided into two groups on the basis of the absence or presence of US thyroid alterations. The result of PTH-FNAB was compared with those of cytology, scintigraphy and, in 24 patients, surgical outcome. PTH-FNAB samples with a value higher than that recorded in the serum and higher than our institutional cut-off were deemed to be probable samples of parathyroid tissue. Cytology proved diagnostic for benign thyroid lesions, non-diagnostic for thyroid lesions, hyperplastic parathyroid tissue, undetermined or malignant thyroid lesions and other lesions in 45%, 30%, 17%, 4%, and 4% of cases, respectively. In 47% of cases, PTH-FNAB indicated that the sample had been taken in parathyroid tissue. In patients without US alterations, the diagnostic accuracy of PTH-FNAB was greater than that of scintigraphy. After surgery, comparison between the results of PTH-FNAB and scintigraphy, in terms of positive predictive value (PPV), revealed the superiority of PTH-FNAB; PPV was 94% for FNAB and 71% for scintigraphy, while sensitivity was 83% and 69%, respectively. PTH-FNAB evaluation after FNAB appears to be more diagnostic than cytology and scintigraphy. Of all the procedures used, PTH-FNAB appears to be the method of choice when the target is US suggestive and reachable. PTH-FNAB appears to be a useful method of guiding surgical intervention.

Multilingual chief complaint classification for syndromic surveillance: An experiment with Chinese chief complaints

Purpose Syndromic surveillance is aimed at early detection of disease outbreaks. An important data source for syndromic surveillance is free-text chief complaints (CCs), which may be recorded in different languages. For automated syndromic surveillance, CCs must be classified into predefined syndromic categories to facilitate subsequent data aggregation and analysis. Despite the fact that syndromic surveillance is largely an international effort, existing CC classification systems do not provide adequate support for processing CCs recorded in non-English languages. This paper reports a multilingual CC classification effort, focusing on CCs recorded in Chinese. Methods We propose a novel Chinese CC classification system leveraging a Chinese-English translation module and an existing English CC classification approach. A set of 470 Chinese key phrases was extracted from about one million Chinese CC records using statistical methods. Based on the extracted key phrases, the system translates Chinese text into English and classifies the translated CCs to syndromic categories using an existing English CC classification system. Results Compared to alternative approaches using a bilingual dictionary and a general-purpose machine translation system, our approach performs significantly better in terms of positive predictive value (PPV or precision), sensitivity (recall), specificity, and F measure (the harmonic mean of PPV and sensitivity), based on a computational experiment using real-world CC records. Conclusions Our design provides satisfactory performance in classifying Chinese CCs into syndromic categories for public health surveillance. The overall design of our system also points out a potentially fruitful direction for multilingual CC systems that need to handle languages beyond English and Chinese.

Mining for Single Nucleotide Polymorphisms in Expressed Sequence Tags of European Sea Bass

Abstract As a multitude of sequence data are published, discovering polymorphisms bioinformatically becomes a valid option. In silico Single Nucleotide Polymorphism (SNP) detection is based on the analysis of multiple alignments. Each column of an alignment is considered a slice containing one base of every sequence aligned. If a mismatch is detected, the slice is further analysed and the mismatch may be reported as a candidate SNP.About 30,000 Expressed Sequence Tags (ESTs) of the fish European sea bass have been sequenced and processed. Since ESTs are redundant, they provide a resource for in silico SNP discovery. To prevent the detection of sequencing errors, a redundancy of two is chosen in order for a mismatch to be considered a candidate SNP. Among the various tools available to detect candidate SNPs, three software packages were tested: SNPServer, PolyBayes and PolyFreq. Candidate SNPs were validated in the laboratory by cloning and sequencing. From preliminary results PolyFreq outperforms both PolyBayes and SNPServer in terms of positive predictive value and SNPServer is the most sensitive tool. PolyFreq and SNPServer non-default identify respectively the fewest and highest number of candidates. Considering candidates detected by several tools seems to enhance both positive predictive value and sensitivity. Out of the 69 loci sequenced, only four were monomorphic, leading to a total of 91.3% polymorphic loci. Randomly chosen contigs will be sequenced to know whether SNP discovery tools tend to predict polymorphic fragments. Polymorphisms will be mapped, used for selection in aquaculture and the study of adaptation in natural populations.

Risk Assessment and Fracture Discrimination by Ultrasound: The Debate Continues

Risk Assessment and Fracture Discrimination by Ultrasound: The Debate Continues

Routine neonatal examination: effectiveness of trainee paediatrician compared with advanced neonatal nurse practitioner

OBJECTIVETo compare the effectiveness of routine neonatal examination performed by senior house officers (SHOs) and advanced neonatal nurse practitioners (ANNPs).DESIGNA prospective study of all infants referred to specialist orthopaedic, ophthalmology,...

Expanded Cytological Referral Criteria for Colposcopy in Cervical Screening: Comparison with Human Papillomavirus Testing

Objective. The goal of this study was to investigate whether expanded cytologic referral criteria for colposcopy or the addition of human papillomavirus (HPV) testing on cervical screening could improve the rates of detection of cervical intraepithelial neoplasia (CIN).Methods. HPV testing by semiquantitative polymerase chain reaction/ELISA was performed in 1000 women who were self-referred for routine Pap smear. They underwent colposcopy following an abnormal smear result or a positive HPV test. As abnormal smear results were considered reports of low- or high-grade squamous intraepithelial lesion, atypical squamous cells of undetermined significance, and even HPV-associated reactive cellular changes (mild koilocytosis, mild dyskeratocytosis, hyperchromatic nuclei, bimultinucleation, and cleared cytoplasm). Loop excision of the transformation zone was performed in women with cytology and colposcopy indicative of CIN, as well as in women with normal cytology but positive HPV test and colposcopic impression of CIN.Results. The Pap test was abnormal in 89% of the cases of CIN 1 (34/38) and 96% of CIN 2/3 (27/28) diagnosed in our population. HPV testing picked up four additional cases of CIN 1 (11%) and one case of CIN 2/3 (4%). Overall the HPV test detected 95% of the cases of CIN 1 (36/38) and 89% of the cases of CIN 2/3 (25/28).Conclusion. HPV testing does not appear to add significantly to cytology in terms of positive predictive value or detection rate, if extended cytologic indications for colposcopy are used.