This was an observational, patient-based study conducted in the Department of Obstetrics and Gynaecology, Regional Hospital Center of Saint-Louis. Participants were recruited from women attending the Obstetrics and Gynaecology outpatient department during the study period from 01/04/2024 to 31/03/2025. Eligible women were enrolled after assessment against the predefined inclusion and exclusion criteria and after obtaining written informed consent. The study was conducted following approval from the Research & Ethics Committee on 22/03/2024 (Letter No. DMCH/P/2024/239; Ref. No. DMCH/476-2023(4580)).

A screening for eligibility was carried out among women attending the gynaecology outpatient department during the study period. The protocol targeted the recommended population for HPV-based cervical cancer screening, women aged 30 - 65 years (see Inclusion and Exclusion Criteria for the 27 participants older than 65 years who were nevertheless enrolled). Written informed consent was obtained, and women who met the criteria were enrolled. A detailed history was taken, and general, systemic, and gynecological examinations were done. A structured data collection schedule was used to record demographic characteristics, reproductive history, clinical symptoms, vaccination status, and pertinent risk factors.

The sample size was calculated based on an expected prevalence of high-risk HPV of 50%, with a 95% confidence level and an absolute precision of 6.2%, using the single-population proportion formula: n = Z 2 p ( 1 − p ) / d 2 , where n is the minimum required sample size; Z is the standard normal deviate corresponding to the selected confidence level (Z = 1.96 for a 95% confidence level); p is the expected prevalence of high-risk HPV (0.50); ( 1 − p ), or q, is the complementary proportion (0.50); and d is the absolute precision or margin of error (0.062). The level of significance was α = 0.05. Substitution of these values yielded a minimum required sample size of approximately 250 participants (1.96² × 0.25/0.062² = 249.8). No additional allowance for potential non-response or incomplete data was applied.

The protocol inclusion criterion was non-pregnant women aged 30 - 65 years. Exclusion criteria were pregnancy, delivery within the previous 6 weeks, age over 65 years, and a history of cervical cancer or hysterectomy. Although age over 65 years was a protocol exclusion criterion, 27 enrolled participants (23 HPV-negative and 4 HPV-positive; Table 1) were older than 65 years. These participants were retained in all analyses so that the results describe the population actually tested. These criteria were used to ensure that women in the study were a representative sample of women who could receive routine cervical cancer screening, and that none of the women in the study was confounded by factors that could affect the cervical assessment, HPV testing, colposcopy findings, or treatment decision.

The cervical screening was performed with the Truenat HPV DNA test from Molbio Diagnostics. The assay was used as a point-of-care molecular test for HPV genotypes 16, 18, 31, and 45; other high-risk genotypes are not detected, so a negative result does not exclude infection with them. Following sampling, the test results were analysed, and women with HPV-positive results were provided with a screen-and-treat approach. Participants requiring further evaluation based on the screening pathway and clinical indications underwent colposcopy as needed; this included HPV-positive participants and other participants in whom colposcopic assessment was clinically indicated. Eligible patients were offered thermal ablation at the screening visit and, based on clinical examination and colposcopic findings, underwent excisional treatment or biopsy when indicated.

Colposcopy. Colposcopy was classified as positive when we founding abnormal cells on the cervix, vagina or vulva that require further investigation, typically via biopsy, and as negative otherwise.

Histopathology. Histopathology was classified as positive when CIN1 or worse (CIN1+) was reported, and as negative when the specimen showed normal epithelium or acute cervicitis. This is the definition consistent with the counts: the 12 true positives correspond to the HPV-positive participants with CIN1 (n = 4), CIN2 (n = 6) or CIN3 (n = 2); the false positive to the HPV-positive participant with acute cervicitis; the false negative to the HPV-negative participant with CIN2; and the true negative to the HPV-negative participant with acute cervicitis. Because CIN2 or worse (CIN2+) is the usual treatment threshold and the target condition in cervical screening accuracy studies, performance against histopathology was additionally calculated with CIN2+ as the target condition (CIN1, acute cervicitis and normal classified as negative). The target condition for the colposcopy comparison is a colposcopic abnormality, not histologically confirmed disease.

A pre-tested and structured schedule was designed based on the study objectives and was used to collect data. Demographic profile, presenting symptom, HPV test result, time between sampling and reporting, colposcopy result, treatment modality, biopsy status, and histopathology result were documented. The collected data were entered into the study dataset and analysed using the predefined statistical methods described below. Two intervals were recorded to examine operational feasibility: the sampling-to-result interval (from specimen collection to availability of the Truenat result) and the screening-to-treatment interval (from the screening visit to the date of treatment). The response to the treatment and the clinical outcome were evaluated at the 6-week follow-up visit.

The main outcome was the clinical performance of point-of-care (POC) HPV DNA testing for screening and management of cervical precancerous lesions. The sensitivity, specificity, positive and negative predictive value, likelihood ratios, and general accuracy of the tests were used to evaluate their clinical performance. Where applicable, the test was compared against colposcopy and histopathology results. Secondary outcomes were the sampling-to-result and screening-to-treatment intervals, feasibility of same-day management, response to treatment following thermal ablation or surgical resection, and the clinical value of the screen-and-treat approach in a resource-poor clinical setting.

The clinical performance of the Truenat HPV DNA test was evaluated separately against colposcopy and histopathology using 2 × 2 contingency tables. Results were classified as true positive (TP), false positive (FP), true negative (TN), and false negative (FN). Sensitivity was calculated as TP/(TP + FN), specificity as TN/(TN + FP), positive predictive value (PPV) as TP/(TP + FP), negative predictive value (NPV) as TN/(TN + FN), positive likelihood ratio (LR+) as sensitivity/(1 − specificity), negative likelihood ratio (LR−) as (1 − sensitivity)/specificity, and overall accuracy as (TP + TN)/(TP + TN + FP + FN). Diagnostic estimates were reported with 95% confidence intervals (exact Clopper-Pearson intervals for sensitivity, specificity, prevalence and accuracy; logit intervals for predictive values; log intervals for likelihood ratios). Agreement between Truenat HPV DNA testing and the reference methods was additionally assessed using Cohen’s kappa coefficient. Reference tests were not applied to all participants: they were applied to almost all HPV-positive women but only to HPV-negative women with a clinical indication. The estimates therefore apply only to the verified subsets and are subject to partial verification bias, which tends to overestimate sensitivity and underestimate specificity; no correction for verification bias was applied.

Categorical variables were summarised as frequencies and percentages, whereas continuous variables were summarised as mean ± standard deviation or median with interquartile range, as appropriate. Normality of continuous variables was assessed using the Kolmogorov-Smirnov test or the Shapiro-Wilk test. Categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate, with Fisher’s exact test applied when expected cell frequencies were less than 5. Agreement between Truenat HPV DNA testing and the reference methods was assessed using Cohen’s kappa coefficient. All statistical analyses were performed using IBM SPSS Statistics for Windows, version 26.0 (IBM Corp., Armonk, NY, USA). A two-tailed p-value <0.05 was considered statistically significant.

A total of 250 non-pregnant women were enrolled and categorised according to Truenat HPV DNA results. Participants in Group A were HPV negative (n = 219, 87.6%), and Group B were positive for HPV 16, 18, 31 and/or 45 (n = 31; 12.4%, 95% CI: 8.6% - 17.1%). Twenty-seven participants (10.8%) were older than 65 years (see Methods). The flow of participants through HPV testing, colposcopy, biopsy, histopathology and treatment is shown in Table 1. No significant differences in age distribution were observed between the groups, suggesting that age did not significantly affect HPV status in this group. The baseline sociodemographic and behavioural characteristics (age group, socioeconomic status, schooling, sexual debut, and sexual partners) are given in Table 2.

Table 1. Participant flow.

Table 2. Comparison of socio-demographic and behavioural characteristics between Group A and Group B. Group A: Truenat HPV DNA-negative participants (n = 219); Group B: Truenat HPV DNA-positive participants (n = 31). Pearson’s chi-square test was used for comparisons. A p-value <0.05 was considered statistically significant.

Socio-economic status, level of education, and number of sexual partners were significantly associated with HPV positivity. HPV-positive participants were concentrated in the lower-middle socioeconomic category (58.1% vs 13.7% of HPV-negative participants) and among those with undergraduate education (61.3% vs 17.4%). These associations did not follow a consistent gradient across categories and should be interpreted with caution given the small number of HPV-positive participants. HPV positivity was linked to an increased number of sexual partners, thus reinforcing the importance of sexual behaviour in the acquisition of high-risk HPV. No significant associations were found between age at sexual debut and HPV status. These results suggest that there were closer associations between social and behavioural determinants and HPV positivity than age-related factors.

Abnormal uterine bleeding and post-coital bleeding were significantly associated with HPV positivity and could be useful to consider proper molecular screening and further evaluation in these cases. There was no significant association between other genital symptoms (vaginal discharge, lower abdominal pain, or difficulty with penetration) and HPV status. The groups were very different in terms of HPV vaccination status, as HPV vaccination was only seen in HPV-negative women. Table 3 shows the comparison of genital symptoms and HPV vaccination profile between the two study groups.

Table 3.Comparison of genital symptoms and vaccination profile between Group A and Group B. AUB: abnormal uterine bleeding. Group A: Truenat HPV DNA-negative participants; Group B: Truenat HPV DNA-positive participants. Pearson’s chi-square test was used for comparisons. A p-value <0.05 was considered statistically significant.

The participants were tested with HPV DNA testing (Truenat). There was no significant difference between women who were HPV negative and HPV positive when the findings of routine per speculum examination were compared; this suggests that visual clinical examination is of limited value for identifying an HPV infection. HPV-positive women were more likely to undergo colposcopy (30/31, 96.8% vs 10/219, 4.6%), which largely reflects the study pathway, since HPV-positive women were referred for colposcopy by design. HPV positive participants had higher Swede scores, but clinical examination and score patterns were not good substitutes for molecular screening.

HPV-positive women were more likely to need therapeutic interventions. Both excision and thermal ablation were significantly associated with HPV positivity, reflecting a process of clinical triage that occurred after molecular screening and colposcopy. Among the 31 HPV-positive women, 14 excisions and 16 thermal ablations were recorded (30 procedures), whereas 28 women were recorded as treated and 1 as not treated (29 women). The treatment status of the remaining 2 HPV-positive women is therefore not accounted for. A favourable response at 6 weeks was observed (Table 4).

Table 4 shows the distribution of treatment methods and response to treatment.

Table 4. Comparison of treatment method and response between Group A and Group B. Group A: Truenat HPV DNA-negative participants; Group B: Truenat HPV DNA-positive participants. The rows added in yellow must be completed so that the treatment and response sections each account for all 31 HPV-positive women; the two “as originally reported” rows should then be deleted. A chi-square of 7.093 with 1 degree of freedom corresponds to p = 0.008, not 0.124; with counts of 0 and 1, Fisher’s exact test should be used.

Women with HPV were more likely to have an immediate biopsy performed (14/31 vs 2/219), and this is likely due to the greater clinical suspicion and requirement for tissue confirmation within this group. Sixteen immediate biopsies were performed, and evaluable histopathology was reported for 15 participants: 2 HPV-negative (acute cervicitis 1, CIN2 1) and 13 HPV-positive (acute cervicitis 1, CIN1 4, CIN2 6, CIN3 2. These 15 participants constitute the paired histopathology analysis. CIN of any grade was found in 12 of the 31 HPV-positive participants (38.7%) and CIN2+ in 8 (25.8%). The row labelled “Normal” (18 HPV-positive participants, 0 HPV-negative participants) exceeds the number of HPV-positive participants biopsied and therefore cannot represent normal histopathology in all 18. The results of biopsies and histopathology are displayed in Table 5.

Of the 250 participants, colposcopy was performed in 40 (16.0%): 30 of the 31 HPV-positive women and 10 of the 219 HPV-negative women (4.6%), the latter selected on clinical grounds. The analysis against colposcopy was therefore restricted to these 40 participants. The 2 × 2 table (Table 6) shows 28 true-positive, 2 false-positive, 8 true-negative, and 2 false-negative results, giving a sensitivity of 93.3% (95% CI: 77.9% - 99.2%), a specificity of 80.0% (44.4% - 97.5%) and a negative likelihood ratio of 0.08 (0.02 - 0.33). These estimates must be interpreted with caution. First, colposcopic abnormality is a clinical impression, not a reference standard for HPV infection or for CIN. Second, because nearly all HPV-positive women but fewer than 5% of HPV-negative women were examined, the comparison is subject to partial verification bias, which is expected to overestimate sensitivity; the true number of colposcopic abnormalities among the 209 unexamined HPV-negative women is unknown. Third, the disease prevalence in this subset (75.0%) reflects selection and not the screening population, so the predictive values cannot be transferred to population screening.

Table 5. Comparison of biopsy and histopathological findings between Group A and Group B. CIN: cervical intraepithelial neoplasia. Group A: Truenat HPV DNA-negative participants; Group B: Truenat HPV DNA-positive participants. Pearson’s chi-square test was used for comparisons. A p-value <0.05 was considered statistically significant.

Table 6. Two-by-two contingency table for Truenat HPV DNA testing compared with colposcopy. TP: true positive; FP: false positive; TN: true negative; FN: false negative.

Histopathology was available for 15 participants (13 of 31 HPV-positive and 2 of 219 HPV-negative) who underwent tissue sampling on clinical grounds. With CIN1+ as the positive result, the 2 × 2 table (Table 7(a)) shows 12 true-positive, 1 false-positive, 1 true-negative, and 1 false-negative result: sensitivity 92.3% (95% CI: 64.0% - 99.8%) and specificity 50.0% (1.3% - 98.7%). With only two histology-negative participants, specificity is essentially uninformative, and the prevalence of disease in this subset (86.7%) reflects selection. With CIN2+ as the target condition (Table 7(b); derived from the histopathology categories in Table 5), there were 8 true-positive, 5 false-positive (4 CIN1 and 1 acute cervicitis), 1 true-negative, and 1 false-negative result: sensitivity 88.9% (51.8% - 99.7%), specificity 16.7% (0.4% - 64.1%), PPV 61.5%, and overall accuracy 60.0% (32.3% - 83.7%) (Table 7(b)). These data are insufficient to estimate the performance of the assay for CIN2+ in a screening population and do not confirm its accuracy against histopathology.

The clinical performance measures derived from these 2 × 2 contingency tables, including sensitivity, specificity, likelihood ratios, predictive values, disease prevalence in the verified subset, and overall accuracy, are presented in Table 7(b).

Table 7. (a) Two-by-two contingency tables for Truenat HPV DNA testing compared with histopathology (n = 15). TP: true positive; FP: false positive; TN: true negative; FN: false negative. Panel (b) is derived from the histopathology categories reported in Table 5. (b) Clinical performance of the Truenat HPV DNA test compared with colposcopy and histopathology, value (95% CI). Estimates are restricted to the selectively verified subsets and are subject to partial verification bias; they are not population screening estimates. The CIN2+ column was calculated from Table 7(b) with the same methods that reproduce the published colposcopy and CIN1+ columns (exact Clopper-Pearson for proportions, logit for predictive values, log method for likelihood ratios).

Agreement between Truenat HPV DNA testing and the reference methods was assessed using Cohen’s kappa coefficient. Agreement with colposcopy was κ = 0.733 (p < 0.001), while agreement with histopathology was κ = 0.423 (p = 0.101) for CIN1+ and κ = 0.06 (p = 0.76) for CIN2+. The corresponding agreement results are presented in Table 8.

Table 8. Agreement between Truenat HPV DNA testing and reference methods using Cohen’s kappa coefficient. κ: Cohen’s kappa coefficient. A p-value <0.05 was considered statistically significant. The CIN2+ row was calculated with the same asymptotic test that reproduces the published p-value for CIN1+ (0.101).