Original articles
Cervical Screening in Africa: Discordant Diagnosis in a Double Independent Reading

https://doi.org/10.1016/S0895-4356(99)00081-5Get rights and content

Abstract

Interobserver variation in the cytological diagnosis of cervical lesions poses a problem for public health screening programs. This study assessed the frequency of discordant diagnoses between two independent cytopathologists in the screening of African women. In Abidjan, Côte d'Ivoire, 2157 women were recruited from three outpatient gynecology clinics and screened for cervical abnormalities and genital and human immunodeficiency virus (HIV) infections. The degree of agreement between the cytopathologists was assessed by kappa statistics. The overall agreement was poor (kappa = 0.33); however, the degree of agreement increased with the severity of the lesions and was fairly good (kappa = 0.53) for high-grade and invasive lesions requiring curative treatment. Discordance was associated with HIV infection but not with genital infections. For a prevention program of cervical cancer in this African context, strategies must be developed to minimize errors in cervical screening. Particularly, HIV-infected women require a systematic rereading to reduce false-negative results.

Introduction

Cancer of the cervix is an important public health problem in Africa with regard to its high prevalence; it is the most common female malignancy 1, 2 and the leading cause of cancer death in African women [3]. In many developing countries, cervical screening programs are almost nonexistent or at the development stage. In Côte d'Ivoire, West Africa, following a prevalence study on cervical dysplasia-neoplasia in relation to human immunodeficiency virus (HIV) infection [4], discussion on a cervical screening program is in progress, including the issue of quality control monitoring.

The current concept in the field of cervical dysplasia, introduced with the Bethesda system [5], separates low-grade squamous intraepithelial lesions (LSILs), which are manifestations of productive human papillomavirus infection and serve as markers for women who are at risk of developing de novo high-grade lesions, and high-grade squamous intraepithelial lesions (HSILs), which are truly precancerous lesions [6]. The distinction between LSILs and HSILs is of great importance because the management of these two lesions is very different [7]. Interobserver variation in the cytological [8] or histopathological 9, 10, 11 diagnosis and grading [12] of dyskaryosis in cervical specimens justifies the setting up of a quality control in public health screening programs [13], in order for the screening strategy to minimize, in particular, false-negative results [14].

We assessed the frequency of discordant diagnosis or grading between two independent cytopathologists during a prevalence study on cervical dysplasia in gynecology clinics in Abidjan.

Section snippets

Methods

Women were recruited from three outpatient gynecology clinics of Abidjan: One is a community-based clinic, and the two others are located within university teaching hospitals [4]. Enrollment was continuous in these centers during the study period. Nonpregnant women between 20 and 50 years of age, without a history of lower genital tract neoplasia and accepting HIV test were included. A signed informed consent was obtained. The study was approved by the Ethical Committee of the Ministry of

Results

Between April 1995 and February 1996, 2281 women were asked to participate in the study; 83 of them (3.6%) were not included because they refused the HIV test. Among the 2198 women enrolled, a Papanicolaou smear could not be performed in 28 instances (1.3%) because of cervical bleeding at the time of enrollment (menses in 17 cases, intermenstrual bleeding in 11 cases), and these 28 women failed to return to the clinic for screening. Thus, a cervical screening was performed for 2170 women; a

Discussion

This study has shown a poor interobserver agreement in the cervical screening of dysplasia-neoplasia performed in the African context of gynecology clinics. Consistency of cytological reporting between two cytopathologists, assessed by kappa statistics, was fairly good for the most serious cervical lesions requiring curative treatment (high-grade and invasive lesions) but poor for less pejorative lesions (low-grade lesions) and null for lesions of undetermined significance (ASCUS). This

Acknowledgements

We are indebted to Roger Salamon for advice and criticism in the preparation of the manuscript. Supported in part by the Agence Nationale de Recherches sur le SIDA (France) and the French Ministry of Cooperation.

References (43)

  • R.J. Kurman et al.

    Interim guidelines for management of abnormal cervical cytology

    JAMA

    (1994)
  • J.P. O'Sullivan et al.

    Interobserver variation in the diagnosis and grading of dyskaryosis in cervical smearsSpecialist cytopathologists compared with non-specialists

    J Clin Pathol

    (1994)
  • S.M. Ismail et al.

    Observer variation in histopathological diagnosis and grading of cervical intraepithelial neoplasia

    BMJ

    (1989)
  • T. Creagh et al.

    Pathologist variation in reporting cervical borderline epithelial abnormalities and cervical intraepithelial neoplasia

    J Clin Pathol

    (1995)
  • W.G. McCluggage et al.

    Inter- and intra-observer variation in the histopathological reporting of cervical squamous intraepithelial lesions using a modified Bethesda grading system

    Br J Obstet Gynaecol

    (1998)
  • P.A. de Ruı́z et al.

    Diagnostic reproducibility of Pap testing in two regions of MexicoThe need for quality control mechanisms

    Bull Pan Am Health Organ

    (1996)
  • C. Bergeron

    Improving the screening of cervical precancerous lesions by improving laboratory quality assurance in cytology

    J Lower Genital Tract Dis

    (1998)
  • National Cancer Institute Workshop

    The Bethesda system for reporting cervical/vaginal cytologic diagnoses

    Acta Cytol

    (1993)
  • J.R. Landis et al.

    The measurement of observer agreement for categorical data

    Biometrics

    (1977)
  • E.C. Lazcano-Ponce et al.

    Reproducibility study of cervical cytopathology in MexicoA need for regulation and professional accreditation

    Diagn Cytopathol

    (1997)
  • D. Palli et al.

    A quality control system involving peer review of abnormal cervical smears

    Cytopathology

    (1993)
  • Cited by (0)

    The DYSCER-CI Group comprises S. Anongba, I. Bobo, M. Koné, N. Messou, G. Moket, S. Mothebesoane-Anoh, Z. Sidibé, M. Sissoko, K. Touré-Coulibaly and C. Welffens-Ekra (gynecologists, Abidjan, Côte d'Ivoire); M. Diomandé, A. Ehouman, and I. Mensah-Ado (pathologists, Abidjan); C. Bergeron and J. Rivel (pathologists, France); D. Bonard, P. Combe, M. Dosso, H. Faye-Ketté, F. Sylla-Koko, C. Montcho, and B. You (microbiologists, Abidjan); G. Orth (virologist, France); F. Dabis and V. Leroy (epidemiologists, France); A. Boka-Yao, I.M. Coulibaly, G. La Ruche, and R. Ramon (epidemiologists, Abidjan).

    View full text