The need for simple, cost-effective screening approaches for cervical cancer prevention in low-resource countries has led to the evaluation of visual screening with 3–5% acetic acid. The low reproducibility and wide variation in accuracy reflect the subjective nature of the test. Pooled sensitivity, specificity, positive and negative predictive values were 80%, 92%, 10% and 99%, respectively, for detecting cervical intraepithelial neoplasia grade 2 or worse lesions. Realistic sensitivity of a quality- assured single visual inspection with acetic acid is around 50%. A single round of visual inspection with acetic acid screening has been associated with a 25–35% reduction in cervical cancer incidence and the frequency of cervical intraepithelial neoplasia grade 2 or worse lesions in randomised-controlled trials. Despite all its limitations, implementing visual inspection with acetic acid screening in low-resource countries may provide a pragmatic approach to building up human resources and infrastructure that may facilitate the highly anticipated low-cost, rapid human papilloma virus testing in the near future.
Introduction
Cervical cancer is a major public health problem in many developing countries, and the absolute burden will increase in the future if effective prevention measures are not undertaken. The global estimates for cervical cancer burden in the world around the year 2008 indicated that there were 530,232 new cancer cases, 275,008 deaths, with four-fifths of the estimated global burden occurring in the low- and middle-income countries (LMIC) of South and South East Asia, sub-Saharan Africa, and South and Central America. Cytology screening is a time-tested, effective approach to reduce the incidence of cervical cancer through early detection and treatment of high-grade cervical intraepithelial neoplasia (CIN), particularly CIN 3 lesions.
Regularly repeated Pap smear screening linked with treatment has prevented millions of women from developing cervical cancer in high-income countries with well-equipped and resourced healthcare services. Most LMIC in Africa and Asia lack screening programmes. Widespread opportunistic screening and the large-scale national or regional cytology screening programmes in Brazil, Cuba, Costa Rica, Chile, Mexico, among others, in Latin America and the Caribbean have been largely ineffective in reducing the cervical cancer burden compared with high-income developed countries.
The diagnostic and treatment algorithms after a positive Pap smear include three visits for colposcopy, diagnosis and treatment. The sensitivity of a single Pap smear to detect CIN 2–3 lesions ranges between 30 and 60% in most settings. Repeated screening at 3–5-year intervals ensure that lesions missed in a given round may be detected in subsequent rounds. Such resource-intensive regimens are not feasible in most LMIC. The low-to-moderate sensitivity and the constraints in implementing and sustaining quality-assured cytology screening in LMIC have led to the evaluation of alternative screening tests such as visual and human papilloma virus (HPV) tests, and new paradigms in low-resource settings. Visual inspection with 3–5% acetic acid (VIA) is the most widely evaluated visual test with a large evidence base from a range of field studies in sub-Saharan Africa, China, India, Bangladesh, Thailand, the Philippines and Latin America. Data for visual inspection with Lugol’s iodine (VILI) are rather limited. We discuss whether visual screening can be a pragmatic and effective public health approach for cervical cancer prevention in LMIC by reviewing the feasibility, acceptability, safety, accuracy, efficacy and cost-effectiveness in preventing cervical neoplasia.
New screening paradigms
In recent years, new paradigms have been proposed to maximise participation of women in screening and treatment, cost-effectiveness and efficiency of screening and treatment in low-resource countries. These include a low-intensity screening involving a single screen targeted at women aged 30–59 years or 30–49 years or screening at 10-yearly intervals, with emphasis on covering a large proportion of targeted women with a highly sensitive test; providing screening, colposcopy, directed biopsies and treatment with cryotherapy or loop electrosurgical excision procedure in one or two sittings ; and a single visit ‘screen-and-treat’ approach when screen-positive women, without evidence of invasive cancer, are treated with cryotherapy or cold coagulation, without triaging procedures such as colposcopy and biopsy. ‘Screen-and-treat’ eliminates investigations to confirm a diagnosis before treatment and minimises loss to follow up, delay in treatment and missed disease. A major concern with ‘screen-and-treat’ cervical cancer prevention strategies is that a large number of women without precursor lesions will undergo cryotherapy or cold coagulation, although available data do not suggest that overtreatment is harmful; on the other hand, it may provide some marginal benefit by protecting women against future HPV infection and by reducing cervical ectopy and targeting the transformation zone where cervical neoplasia occur. Current evidence suggests that ‘screen-and-treat’ interventions are safe, well accepted by women, and are effective in preventing cervical neoplasia.
New screening paradigms
In recent years, new paradigms have been proposed to maximise participation of women in screening and treatment, cost-effectiveness and efficiency of screening and treatment in low-resource countries. These include a low-intensity screening involving a single screen targeted at women aged 30–59 years or 30–49 years or screening at 10-yearly intervals, with emphasis on covering a large proportion of targeted women with a highly sensitive test; providing screening, colposcopy, directed biopsies and treatment with cryotherapy or loop electrosurgical excision procedure in one or two sittings ; and a single visit ‘screen-and-treat’ approach when screen-positive women, without evidence of invasive cancer, are treated with cryotherapy or cold coagulation, without triaging procedures such as colposcopy and biopsy. ‘Screen-and-treat’ eliminates investigations to confirm a diagnosis before treatment and minimises loss to follow up, delay in treatment and missed disease. A major concern with ‘screen-and-treat’ cervical cancer prevention strategies is that a large number of women without precursor lesions will undergo cryotherapy or cold coagulation, although available data do not suggest that overtreatment is harmful; on the other hand, it may provide some marginal benefit by protecting women against future HPV infection and by reducing cervical ectopy and targeting the transformation zone where cervical neoplasia occur. Current evidence suggests that ‘screen-and-treat’ interventions are safe, well accepted by women, and are effective in preventing cervical neoplasia.
Visual inspection with 3–5% acetic acid
Visual inspection with acetic acid involves naked eye inspection of the cervix, using a bright torch light or a halogen focus lamp, one minute after the application of 3–5% dilute acetic acid using a cotton swab or a spray. Most of the reported experience with VIA involves 4 or 5% acetic acid, which is well tolerated by women. The changes that occur in the cervix in relation to the squamocolumnar junction (SCJ) after acetic acid application visible to the naked eye can be categorised as negative or positive tests for cervical neoplasia. A negative test is characterised by one or more of the following: no acetowhite lesions, ill-defined, faint, translucent acetowhite lesions, acetowhitening of endocervical polyps, nabothian cysts, prominent acetowhitening of the SCJ, dot-like acetowhitening scattered all over the cervix, geographic satellite acetowhite lesions not touching the SCJ. A positive test is characterised by well-defined, opaque acetowhite lesions in the transformation zone close to the SCJ or to the external orifice or the entire cervix turning acetowhite ( Fig. 1 ). Invasive cancer is suspected when a cervical growth turns acetowhite.
Immediate results make VIA a real time test permitting a ‘single visit’ approach. It does not require a laboratory infrastructure and acetic acid is widely available. Although simple, feasible and affordable to carry out, interpretation of test results is difficult in postmenopausal women with brittle, degenerating cervical epithelium, and unsatisfactory in women with partially visible or invisible SCJ, as the visibility of the transformation zone is critical for VIA accuracy. It has all the drawbacks of a provider dependant, subjective test, such as wide variability in test-positive rates, detecting high-grade CIN and low reproducibility.
Visual inspection with acetic acid can be rapidly taught to a range of test providers (screeners) such as primary-health workers, midwives, nurses, paramedical workers and doctors in short competency-based training courses of 4–10 days with the help of manuals and hands-on training. A learning period of few weeks to few months has been observed following the initial training. Periodic short retraining sessions have been found to be useful in reinforcing skills. The VIA positivity by newly trained screeners ranges between 20 and 35%, which decline to less than 15% after the learning curve.
Test performance of visual inspection with acetic acid
The test performance in detecting disease is assessed by accuracy and predictive values. Sensitivity of a test is a measure of probability of correctly identifying people with true disease, whereas specificity is a measure of correctly identifying people without disease. Sensitivity is defined as the proportion of people with a given disease who had a positive screening test, and specificity as the proportion of those without disease who tested negative on screening. They are measures of accuracy. The predictive values of a test reflect its clinical utility. Positive predictive value (PPV) refers to the proportion of people with positive screening test diagnosed with disease, whereas the negative predictive value (NPV) indicates the proportion of people with negative screens with no disease.
The performance of VIA in detecting cervical neoplasia has been addressed in more than 50 cross-sectional studies of variable quality involving both asymptomatic and symptomatic women in urban and rural populations in countries in Africa, Asia and Latin America. Almost all studies measured the true disease at one cross-section in time without longitudinal follow up. The age range of included women, test result nomenclature and criteria, the type of screeners and the training methods varied between the studies. Reference standards used for diagnosis and the thresholds used for defining true disease differed between the studies. Reference standard investigations included both colposcopy and histology by directed biopsy for most studies, whereas some used histology by four quadrant biopsy or Pap smear results or colposcopy only; reference standards were provided to all women irrespective of test results in some and to test positive women with or without a fraction of test negative women in others. The reference standards were provided by newly trained clinicians and pathologists in some studies. The possibility of wide variation in the proportion of all women with true disease actually diagnosed by the reference standards used in the studies cannot be excluded. Some studies included both CIN and invasive cancers as true disease, whereas some did not include cancer. The threshold used to define true disease was CIN 2 in most studies, whereas few studies used CIN 1 or CIN 3. Hence, the accuracy of VIA reported in studies should be interpreted with caution.
The sensitivity of VIA to detect CIN 2 and 3 lesions and invasive cervical cancer varied from 31–95%, and the specificity varied from 49–97% in the different cross-sectional studies. The PPV ranged from 5–20% and the NPV ranged from 92–99%. Results from a recent meta-analysis of 77 cross-sectional studies assessing the accuracy of VIA since 1990 until 2009 are instructive. It categorised the studies into different groups depending upon whether asymptomatic general population of women or symptomatic women in clinical settings were included; the threshold used for defining true positive disease; whether the reference diagnostic investigations for final diagnosis was applied to the entire women included in the studies irrespective of screening test results (thereby minimising verification bias in establishing true disease in test positive and test negative women, allowing a conservative estimate of accuracy) or only to the test positive women with or without a sample of test negative women. On the basis of 60 studies, the sensitivity of VIA to detect CIN 2 and 3 lesions and invasive cervical cancer varied from 14–95%; the specificity varied from 14–98%; PPV ranged from 2–73% and the NPV ranged from 92–99% ( Table 1 ). The wide range in accuracy parameters of VIA in individual studies reflects the subjective, provider-dependent nature of the test and the varying quality of reference standards used to establish the true disease.
| Characteristic | % |
|---|---|
| Range in sensitivity | 14–95 |
| Range in specificity | 14–98 |
| PPV | 2–73 |
| NPV | 92–99 |
| Pooled sensitivity a | 80 |
| Pooled specificity a | 92 |
| Pooled PPV a | 10 |
| Pooled NPV a | 99 |
a Based on 26 studies with minimal verification bias; CIN 2, cervical intraepithelial neoplasia grade 2; NPV, negative predictive value; PPV, positive predictive value; VIA, visual inspection with acetic acid.
On the basis of 26 studies meeting the criteria, such as inclusion of asymptomatic women who all had confirmatory reference standard investigations, the pooled estimates of sensitivity, specificity, PPV and NPV to detect CIN 2 or worse lesions were 80%, 92%, 10% and 99%, respectively ( Table 1 ). Another meta-analysis using random-effect methods of 11 studies reported pooled estimates of sensitivity, specificity, PPV and NPV as 72%, 79%, 17% and 99%, respectively.
The reported sensitivities for VIA in most studies are likely to be overestimated, given the wide variation in the quality of reference diagnostic standards; a high correlation with colposcopy and VIA may result in overestimation of VIA sensitivity in studies relying on colposcopically directed biopsy to establish the final diagnosis. It seems that a quality-assured single VIA has an average sensitivity around 50% and specificity around 85% to detect high-grade CIN in controlled study settings.
When conventional cytology was concurrently evaluated in these studies, the sensitivity of VIA was found to be higher than or similar to that of cytology, but had lower specificity. Concurrent HPV testing had consistently higher sensitivity and specificity than VIA. The longitudinal sensitivity of VIA and HPV testing in detecting CIN 2–3 lesions diagnosed over a follow-up period of 36 months from screening were 53% and 90%, respectively in a South African study. Low-level magnification using 2–3X hand-held lens does not improve the test performance of naked eye visualisation after acetic acid application.
Quality assurance of visual inspection of acetic acid
Quality assurance procedures for VIA remain to be standardized ( Table 2 ). Assuring consistent high performance under field conditions and in routine programmes can be challenging. Constant monitoring and frequent re-training of test providers are needed to ensure consistency in performance. Acetic acid should be applied gently to avoid bleeding, which interferes with interpretation of the test results. The useful quality assurance procedures include use of 5% freshly prepared acetic acid on daily basis, use of a stop watch or a clock in the clinic to ensure screeners wait for one minute after the application of acetic acid to allow its effect to take place before reporting the test results, digital documentation and review of VIA findings using digital cameras, periodic re-training of providers and reproducibility studies using both photographs of acetic acid soaked cervix and VIA findings live in women, and monitoring the following process measures regularly: test positivity rates (acceptable limits 7–15%), PPV for any grade of CIN (around 20%), PPV for CIN 2–3 lesions (around 10%), and inter-observer agreement (exceeding 70%, kappa > 0.4).
|
Follow up of women positive for visual inspection with acetic acid
Management of women positive for VIA in studies and programmes involved one of the following options: (1) immediate treatment with cryotherapy in the same setting as screening (single visit ‘screen-and-treat’ approach) after clinically excluding women with large lesions that cannot be adequately covered by the cryoprobe or suspected invasive cancer (who are referred for further investigations and treatment); (2) colposcopy and treatment based on colposcopic findings; (3) magnified visual inspection (VIAM) using a 2–5X hand-held lens and histological sampling and treatment with cryotherapy after histological sampling (with retrospective access to histology); (4) colposcopy and histological sampling by directed biopsies from abnormal areas and treatment based on colposcopic findings (with retrospective access to histology results); (5) colposcopy and histological sampling by directed biopsies and treatment based on results of histological diagnosis.
A single-visit approach has been widely used in programme settings in Thailand and in study settings in some sub-Saharan African countries. Colposcopy and directed biopsies have been predominantly used in triaging women tested positive by VIA in Bangladesh, India, some sub-Saharan countries and Latin American countries in both programme and study settings in India. VIAM has been used in Peru and in India to triage women tested positive by VIA.
Strengths and limitations of visual inspection by acetic acid
The major strengths of VIA include its affordability and feasibility in LMIC, possibility for repeating the test and to train screeners quite rapidly, wide availability of dilute acetic acid, immediate test results, high potential for linkage with diagnosis and treatment and minimal infrastructure needs. The limitations include the subjective nature of the test, poor accuracy in postmenopausal and old women, low reproducibility, low specificity and high false positivity, leading to unnecessary anxiety, over-investigation and over-treatment, and lack of standardised methods for training, competency evaluation and quality assurance. The possibility that cost savings from low-test costs may be offset by excess expenditures caused by over-investigations and over-treatment cannot be ruled out. False negative VIA results may lead to delayed diagnosis and treatment.
Visual inspection by acetic acid screening among women tested positive for human immunodeficiency virus
Women infected with human immunodeficiency virus (HIV) have a higher prevalence, incidence, and persistence of HPV infection, higher frequency of high-grade lesions and larger lesions and a higher incidence, persistence, and progression of CIN than in HIV-negative women. Highly active anti-retroviral therapy (HAART) has increased the life expectancy of HIV-positive women, but has little beneficial effect on the natural history of cervical neoplasia, and more of these women will develop cervical cancer unless effective intervention is available.
Visual inspection by acetic acid has been evaluated in HIV-positive women in few cross-sectional studies in sub-Saharan Africa and in India. In a study in India, at the CIN2 disease threshold, the sensitivity, specificity, PPV and NPV estimates of VIA were 80%, 83%, 48%, and 95%, respectively, compared with 60%, 60%, 22%, and 89% for cytology at atypical squamous cells of undetermined significance threshold, and 61%, 65%, 25%, and 89% at low-grade squamous intraepithelial lesion threshold. Studies in sub-Saharan Africa indicate that platforms within the HIV care services constitute a pragmatic health service entry point to provide sustainable cervical screening services. The cervical cancer prevention efforts in Zambia, embedded in the existing HIV and AIDS programme, has provided VIA ‘screen-and-treat’ services to over 58,000 women over the past 5 years. In a study involving 3642 HIV-positive women within HIV care and treatment clinics in Kisumu, Kenya, VIA followed by colposcopy and directed biopsies led to the detection and treatment of 259 women with CIN 2–3 lesions.
Recent evidence from a randomised trial in South Africa indicated that VIA ‘screen-and-treat’ approach is an effective method of reducing high-grade CIN in HIV-positive women; a significant 49% reduction in CIN 2 or worse lesions after VIA screening and treatment was observed among 322 HIV-positive women compared with 312 HIV-negative women in the control group who received delayed evaluation at 6 months; 18 women were diagnosed with CIN 2–3 lesions in the VIA group compared with 36 women in the control group ( Table 3 ). Human papilloma virus screen-and-treat in 322 HIV-positive women was more effective than VIA and treat, showing a significant 80% reduction in high-grade lesions compared with the control group.
