Article Text
PDF
Abstract
Objective To determine the sensitivity, specificity, and positive and negative predictive values of a cervical cancer screening program based on visual inspection with acetic acid and Lugol’s iodine using a smartphone in a sub-urban area of very low resources in Kinshasa (Democratic Republic of Congo).
Methods This cross-sectional validation study was conducted at Monkole Hospital and it included women between the ages of 25–70 years after announcing a free cervical cancer screening campaign through posters placed in the region of our hospital. Questionnaires collected sociodemographic and behavioral patients characteristics. In the first consultation, we gathered liquid-based cytology samples from every woman. At that time, local health providers performed two combined visual inspection techniques (5% acetic acid and Lugol’s iodine) while a photograph was taken with a smartphone. Two international specialists evaluated the results of the smartphone cervicography. When a visual inspection was considered suspicious, patients were offered immediate cryotherapy. Cytological samples were sent to the Pathology Department of the University of Navarra for cytological assessment and human papillomavirus (HPV) DNA genotyping.
Results A total of 480 women participated in the study. The mean age was 44.6 years (range 25–65). Of all the patients, only 18.7% were infected with HPV (75% had high-risk genotypes). The most frequent high-risk genotype found was 16 (12.2%). The majority (88%) of women had normal cytology. After comparing combined visual inspection results with cytology, we found a sensitivity of 66.0%, a specificity of 87.8%, a positive predictive value of 40.7%, and a negative predictive value of 95.3% for any cytological lesion. The negative predictive value for high-grade lesions was 99.7%.
Conclusions Cervical cancer screening through combined visual inspection, conducted by non-specialized personnel and monitored by experts through smartphones, shows encouraging results, ruling out high-grade cytological lesions in most cases. This combined visual inspection test is a valid and affordable method for screening programs in low-income areas.
- cervix uteri
- uterine cervical neoplasms
Data availability statement
Data are available upon reasonable request.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Cervical inspection with acetic acid is an efficient method to screen pre-invasive disease of cervical cancer primarily oriented for low resources countries where standard methods of screening are unaffordable.
WHAT THIS STUDY ADDS
This study shows a 99% negative predictive value for high-grade lesions by combining a cervical visual inspection with acetic acid plus Lugol’s iodine performed by non-trained health providers along with a smartphone re-evaluated by an external expert.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Comprehensive maintainable screening campaigns might be accomplished in low resources countries, ruling out cervical preneoplastic lesions and impacting the incidence and mortality of this disease.
INTRODUCTION
Over 300 000 women die of cervical cancer annually, 85% in resource-limited regions of the world.1 In the Democratic Republic of Congo, over 7500 women are diagnosed with cervical cancer yearly, and more than 5500 die from the disease. Cervical cancer ranks as the most frequent cancer among women. Screening and management resources are very limited and therefore, typically, patients are diagnosed with locally advanced tumors with very limited options of treatment.2 Early screening of women at risk for cervical cancer allows clinicians to treat precancerous lesions when they are found. The primary cause of precancerous lesions is high-risk human papillomavirus (HPV) genotypes, and persistent high-risk HPV infection is known as a significant risk factor in cervical cancer progression. HPV is currently the most common sexually transmitted infection, and it is estimated that 80% of women will be infected with HPV at some point in their lifetime.3
The goal of cervical cancer screening is to accurately detect high-grade precursor lesions of the cervix to allow timely treatment of cervical intraepithelial neoplasia. Persistent high-risk HPV infection is the causal agent of virtually all cervical cancers and their precursors.4 Unfortunately, access to screening and treatment of precancerous cervical lesions and management of cervical cancer remains a challenge for many women in low- and middle-income countries, further stressing inequities in women’s healthcare.
Implementing cytology-based screening programs has considerably reduced disease burden in high-income countries.5 However, economically poor geographic areas fail to provide efficient cervical cancer screening services due to a deficiency of human and material resources6 Alternative tests for cytology, such as visual inspection of the cervix after application of 3–5% acetic acid or Lugol’s iodine, have been assessed.7 These techniques are affordable and particularly suitable for large-scale screening of cervical lesions in low-resource settings since the test results are obtained immediately, allowing same-day treatment. In addition, this approach provides options for therapy after a positive screening test.8 Visual inspection methods are more sensitive but less specific than cervical cytology. However, visual inspection methods vary widely in the reported sensitivity and specificity.7
Smartphones have been used in cervical screening for visual inspection after acetic acid (VIA) or Lugol’s iodine (VILI) application to capture and share images to improve the sensitivity and interobserver variability of visual inspection. In addition, it has been shown that visual inspection using smartphones may provide additional support to healthcare providers delivering care in low-resource settings.9 The objective of this study was to determine the sensitivity, specificity, and positive and negative predictive values of a sustainable cervical cancer screening system (see and treat) based on visual inspection with acetic acid/Lugol’s iodine (VIA+VILI). We used smartphone technology designed to be applied in an area of very few resources in a sub-urban zone in Kinshasa (Democratic Republic of Congo), supported by international expert monitoring.
METHODS
Study Sample and Design
This cross-sectional study was conducted at Monkole Hospital around Mont-Ngafula in Kinshasa and its outpatient clinics (Kimbondo, Eliba, Moluka, and Gombe) between July 2017 and August 2018.
It included women aged 25 to 70 years old who attended Monkole Hospital to receive medical care following the poster announcement of a free cervical cancer screening campaign or, in some cases, attending human immunodeficiency virus (HIV) voluntary counseling and testing. For 2 weeks, we placed posters announcing the free screening campaign at outpatient clinics and parishes close to Monkole hospital (see online supplemental Figure S1). In addition, this campaign was promoted in the Mont-Ngafula area, facilitating a contact phone number if they wanted to sign up for the program. Pregnant women or those who had undergone a hysterectomy for any reason, and patients with previous cervical cancer history, were excluded from the study.
Supplemental material
Doctors and volunteers organized the participants’ selection. Women who met the inclusion criteria were invited to participate in the study. We informed them in detail about the goals and the development of the study. Those who accepted participating in the study were asked to sign the informed consent (see online supplemental material).
Data Collection
We collected participants’ data through a questionnaire developed for the study. We compiled data on the sociodemographic, clinical, and behavioral characteristics of all these women on their arrival at the screening program. The questionnaires included questions, among others, about tobacco use, long-term use of oral contraceptives, immunosuppression, age at first sexual intercourse, or number of sexual partners (online supplemental material).
Sample Collection and Smartphone-Based Cervicography after VIA+VILI
First, we performed a visual cervical inspection using a disposable speculum. Patients with suspected cervical carcinoma were biopsied and excluded from this study. Then, for every woman, we took a cervical sample for liquid-based cytology and HPV DNA testing to be conducted at an external laboratory in Spain. After taking the cervical sample, we applied a 5% acetic acid embedded swab for 1 min, and immediately we took a photograph with a smartphone. Likewise, we dyed the cervix with a Lugol’s iodine-embedded swab (online supplemental Figures S2A and S2B). Again, we took a new cervical photograph with the smartphone without delay. The test was considered positive when the observer found a suspicious lesion showing concordance between the visual inspection with acetic acid (aceto-white areas) and the visual inspection with Lugol’s iodine (iodine-negative area). An international expert observer re-evaluated both smartphone photographs when positive or inconclusive results were observed. The final diagnosis relied on the expert assessment (Figures 1 and 2). During the first 100 cases, the international experts confirmed every case and thus trained the volunteers.
Design of the study.
Smartphone cervicography after acetic acid during the expert evaluation.
Patient Management After Suspicious Cervicography
After the combined visual inspection, we informed those patients with a positive lesion during the same visit. After the test, we offered them the option of being immediately treated with cryotherapy by applying liquid nitrogen to the cervix, following the available WHO guidelines in 2017.10
Pathology Analyses and HPV Genotyping
When the local campaign was completed, we sent all the cytological samples to the Pathology Department of the University of Navarra Clinic in Spain for cytological evaluation and HPV DNA genotyping. All anonymized results were shared with the physicians involved in the campaign at Monkole Hospital. All patients with a high oncogenic risk of HPV or high-grade cytology lesions were reviewed, and again, cryotherapy was offered in those cases when it was not previously applied.
HPV DNA detection was performed using the Cobas HPV test with the Cobas 4800 platform (Roche Diagnostics GmbH)
Statistical Analysis
The sample size was calculated considering a theoretical visual inspection sensitivity of 67% and a specificity of 84% for a prevalence of preneoplasic lesions of 5%. Therefore, we considered that a sample size of 450 women would be suitable following this criterion. We included the characteristics of all patients, the results of visual inspection techniques, the cytology, and HPV genotypes results in a database to conduct statistical analyses through the STATA program. We calculated sensitivity, specificity, positive predictive value, and negative predictive value to detect any positive and high-grade lesions compared with cytology. For doing this, valid positive tests for high-grade lesions were considered when a positive result in the combined visual inspection was associated with a high-grade lesion in the cervical cytology. Conversely, false positives for high-grade lesions showed a positive test when the cytology was normal or low-grade.
Likewise, valid positive tests for any-grade lesions were considered when a positive result in the combined visual inspection was associated with a high-grade, low-grade, or atypical squamous cells of undetermined significance (ASCUS) in the cervical cytology. Contrarily, false positives for any grade lesions were those who had tested positive when the cytology was normal.
For the data analysis, we conducted descriptive research of the characteristics of patients, the signs, symptoms, and factors associated with HPV infection, to define the study scenario. Additionally, we compared all these characteristics between HPV-infected and uninfected women; a Student’s t-test was performed to compare means for quantitative variables and a χ2 test to reach the proportions between the two groups for categorical variables. In addition, logistic regression analyses were conducted to analyze the association between participants’ characteristics and behaviors and HPV infection.
Finally, sustainability was defined as the project’s ability to be maintained at a specific rate or level based on three goals: increase the use and maintenance of the screening program, training implementation of Monkole Hospital’s staff, and development of staff skills in data collection and monitoring to dictate future management. All authors contributed meaningfully to the conception or design of the work or the study’s acquisition, analysis, or interpretation of data. The authors confirm the completeness and accuracy of the data and analysis, the adherence of the study to the original design, and the final approval of the version for publication.
RESULTS
Patient Characteristics
The mean (SD) age of our population was 44.6 (10.9) years old. There were no differences in age between patients infected and not infected with HPV. Seventy-one percent of women were married, and 70% had attended high school or higher education, although only 34.6% had a paid job. Overall, 18.7% of women were infected with HPV, and 3.5% were positive for HIV. Nearly one-third of women were menopausal, with a mean (SD) age at menopause of 48.1 (3.9) years old.
Regarding known factors associated with HPV infection, the mean (SD) age at first intercourse was 19.5 (4.5) years. Patients had a mean of 4.7 (3.1) pregnancies and reported a mean of 2.1-lifetime sexual partners. Seven percent had cervical cancer family history, and nearly 20% reported hormonal contraception and intravaginal herbal and chemical drugs. Although only 1.3% were smokers, almost one-third (31.4%) stated that they drank alcohol regularly (Table 1).
Sociodemographic, clinical and behavioral characteristics (n=480)
Cytological evaluation
Eighty-eight percent of patients had a normal result on the cervical cytology, 9.2% had low-grade lesions, and 1.8% had high-grade lesions. Ninety-nine percent of HPV-negative patients had a normal cytology report. In comparison, cytology was reported with abnormal findings in 55.6% of the 90 HPV-infected patients, showing either high-grade results (nine cases), low-grade (29 cases), and ASCUS (12 cases) (Table 2).
Results of cytology, low- and high-grade lesions
Combined Visual Inspection
The two combined visual inspection techniques were performed in all patients and compared with the cytology results, which was our standard test for comparison. No biopsies were taken. The combined techniques showed a sensitivity of 60.3%, a specificity of 87.6%, a positive predictive value of 40.2%, and a negative predictive value of 95.3% (Table 3). Additionally, the high-grade lesions sub-analysis showed that combined techniques demonstrated a sensitivity of 88.9%, a specificity of 83.1%, a positive predictive value of 9.3%, and a negative predictive value of 99.7% (Table 3).
Results VIA+VILI compared with the cytology, sensitivity, specificity, and predictive value for the detection of any lesion and for high-grade lesion detection
HPV Infection
There were 18.7% of women infected by HPV (90/480), 75% of them by high-risk genotypes. In 11 patients, more than one genotype was identified, up to four different ones. Fifty percent of women were infected by genotypes 61 (14.4%), 16 (12.2%), 52 (8.8%), and 33, 35, and 83 with a 7.7% frequency each.
DISCUSSION
Summary of Main Results
Our study demonstrated that a local health team composed of general practitioners and medical students could perform an accurate cervical cancer screening program after a short period of training by a gynecologic oncologist. The main result of our study is that a negative observation after the combination of the direct visual inspection with acetic acid and Lugol’s iodine (VIA+VILI) predicts a negative cytological report for high-grade lesions in 99% of cases. This finding is significant, especially considering it is an inexpensive, reproducible screening test that can be performed after a brief training period. During the first 100 cases, the international expert confirmed every case and thus, trained the local healthcare providers. In addition, a remote expert in cervical cancer screening can rectify the results because of smartphone technology which exists in most low-resource countries.
Results in the Context of Published Literature
Cervical cancer screening programs in sub-Saharan African countries are rare or non-existent and, because of this, the prevalence of cervical cancer in those countries is extremely high compared with developed (high-income) countries.11 The absence of structured programs and dedicated gynecologists and pathologists is the rule. Additionally, the cost of a simple Pap smear, a biopsy, or an HPV test is not affordable for most populations. The usual charge of cervical cytology in the Democratic Republic of Congo is unreachable to the average population. Nowadays, the cost of a Pap smear in Kinshasa varies between US$30 and US$70. Most women in this underprivileged region make US$1–2 a day, so cannot pay this amount. This high cost is partly related to the lack of trained individuals to read and interpret Pap smears. Therefore, our goal was to design an affordable and sustainable algorithm to allow us to discard the true negatives and focus on treatment and follow-up of the patients with positive tests at almost no cost.12 13
Several studies have shown different ways of correcting those deficiencies by implementing affordable cervical cancer screening programs that can influence the rate of preclinical cancer diagnosis in women under these circumstances. It has been shown that screening protocols based on visual inspection are efficient and affordable. Some authors have demonstrated an impact in diminishing the rate of cervical cancer in specific countries, such as India.14 Adding a second dye with Lugol’s iodine may confer robustness to a positive result.15
The option of consulting the positive or doubtful results by sharing the cervical smartphone images with an expert through the internet theoretically increases the test’s prediction.9 Finally, demonstrating that our protocol is achievable with non-expert observers after a short training period confirmed to our team the hope of sustainability. Several studies have shown similar results with visual inspection in many sub-Saharan countries, Honduras, Vietnam, Thailand, and India among others.15–18 In all those studies, the negative predictive value after using visual inspection with acetic acid is high, but in some, the addition of Lugol’s iodine increases the rate of negative predictive value. Beyond that, the documentation of the cervicography evaluation with a smartphone photograph allows the healthcare providers to further consult with an expert when needed. A recent systematic review suggests that smartphone-based visual inspection has accuracy in the detection of cervical dysplasia and may provide additional support to healthcare providers delivering care in low-resource settings.9
Strengths and Weaknesses
The strength of this study is the design of a cervical cancer screening strategy in a very-low resource area validated with an external liquid-based cytological evaluation in our European hospital. The pathologists who studied the liquid-based cytology had no information about the patient’s cervical visual inspection. Additionally, the option of a visual inspection final assessment by gynecologic oncologists, using smartphone technology, increases the possibility of improving the accuracy of this technique. In addition, it allows for testing of the options to replicate this type of campaign in the future, with the assistance of non-present expert reviewers using an internet connection. The main weakness of our study is that we compared the visual inspection with cytologic evaluation instead of a colposcopic-directed biopsy. Cervical cytology is just another screening method rather than a final diagnostic tool. The practice of a routine biopsy after a positive visual inspection test was not possible at the time of the study design due to a budget limitation. This could lead to overtreating some patients without any cervical disease. We have incorporated the biopsy for coming campaigns after obtaining further funds. Despite this limitation, the visual inspection only overlooked one case of high-grade lesion of all the high-grade cases captured by the cervical cytology with a negative predictive value of 99.7%. Another study limitation is the potential bias of introducing inexperienced medical providers into the cervical cancer screening campaign. However, this tries to prove that these campaigns might be feasible in the context of amateur observers after short training.
Implications for Practice and Future Research
The most important implication of our study for practice and future research is the demonstration that the combination of two different techniques of cervical visual inspection done by non-expert observers may accurately offer physicians the opportunity to rule out pre-invasive disease. Moreover, in those countries where HPV vaccine availability is non-existent, where many affected women do not even enter the age range of vaccination, and where access to any cervical cancer screening is impossible, this result offers an opportunity for visual inspection as a sustainable pathway to decrease cervical cancer incidence and mortality.
The following campaign has been organized for the summer of 2022, and several additional features will be introduced after our previous experience. For instance, we have planned to add cervical biopsy and thermal coagulation after a positive test. In addition, we will follow the patients with a positive test and thermal ablation with cytology and HPV testing after 3, 6, and 12 months to investigate the impact of thermal ablation on the negativization of HPV infection.
Conclusions
We may affirm that cervical visual inspection with a combined technique using acetic acid and Lugol’s iodine provides a negative predicting value of over 99% for high-grade cytologic lesions, ruling out, in fact, the existence of pre-invasive cervical disease in a women’s population. Furthermore, this screening design is inexpensive, allowing the local health providers to develop more comprehensive campaigns to screen the female population for cervical cancer.
In conclusion, visual inspection with acetic acid plus Lugol’s iodine performed by untrained health providers monitored by an external expert through smartphone technology is highly efficient in ruling out the pre-invasive disease of cervical cancer.
Data availability statement
Data are available upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by the Ethics Committee of University of Navarra 2017.096, and the Ethics Committee of Monkole Hospital N/Réf . O- I2I CEFA-MONKOLE I CELI2O T7. Participants gave informed consent to participate in the study before taking part. All data and samples were coded and confidentially managed.
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
SC and LC are joint senior authors.
Twitter @GabReina, @lchiv4
CT and MF-M contributed equally.
Contributors All authors contributed meaningfully to the conception or design of the work or the acquisition, analysis, or interpretation of data for the study. The authors confirm the completeness and accuracy of the data and analyses, the fidelity of the study to the protocol, and the final approval of the version to be published. LC and SC are both Senior Authors. LC is the corresponding author. LC is the author acting as gurantor.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.