Clinical Value of Negative-Pressure Endometrial Sampling Technique in Screening Endometrial Tuberculosis

Introduction

Tuberculosis remains a significant global public health issue characterized by high morbidity and mortality. Approximately 25% of the global population is susceptible to latent infection by Mycobacterium tuberculosis, and 5–10% of these individuals are at risk of progressing to active tuberculosis during their lifetime. In 2021, the disease affected approximately 10.6 million people in the world and led to 1.6 million deaths [1,2]. Although pulmonary tuberculosis is the predominant form worldwide, the incidence rate of extrapulmonary tuberculosis has steadily increased. Among these, female genital tuberculosis significantly impacts reproductive health, accounting for about 5–30% of infertility cases globally [3]. Endometrial tuberculosis, specifically, often presents asymptomatically in the early stage, progressing to severe complications such as menstrual disorders, endometrial atrophy, dysmenorrhea, pelvic pain, tubo-ovarian masses, and infertility. These conditions severely affect women’s quality of life and impose substantial burdens on healthcare systems [4–7].

Traditionally, endometrial tuberculosis is diagnosed by dilatation and curettage, also termed fractional curettage, in which the endocervical canal and uterine cavity are scraped sequentially and submitted as separate specimens for histopathological evaluation. The procedure is invasive, usually performed under anesthesia, and has risks such as patient discomfort, uterine perforation, and infection [8]. It has recently been reported that negative-pressure aspiration using disposable endometrial samplers enables a minimally invasive alternative, enabling flexible collection of adequate tissue samples from various regions of the uterine cavity, thus potentially improving the diagnostic accuracy and patient compliance [9]. Although molecular assays such as PCR can detect Mycobacterium tuberculosis in genital specimens, their availability for endometrial samples remains limited in many low- and middle-resource settings, and costs may be prohibitive for routine use [10]. Histopathological confirmation remains the most accessible and widely accepted diagnostic approach in our institution. Endometrial aspiration or curettage provides sufficient tissue for both histological examination and microbiological testing, which is particularly valuable in atypical or smear-negative cases.

However, data specifically supporting the use of negative-pressure endometrial sampling techniques for the diagnosis of endometrial tuberculosis remain limited. Therefore, this study aimed to evaluate the diagnostic efficacy of the negative-pressure aspiration method compared to traditional fractional curettage in detecting endometrial tuberculosis. Through tissue culture and histopathological biopsy, this study provides essential clinical data for improving patient outcomes and reducing procedural complications associated with diagnosis.

Material and Methods

SUBJECTS:

This study was approved by the ethics committee of our hospital (not registered), and written informed consent was obtained from all patients. From April 2021 to April 2023, subjects (n=84) of this study were enrolled from among patients suspected of having endometrial tuberculosis receiving treatment in our hospital. This was a non-randomized, prospective observational study. Participants were assigned to either the negative-pressure aspiration group (n=48) or the fractional curettage group (n=36) based on their personal preference after receiving standardized counseling by the attending gynecologist, which included an explanation of each procedure’s invasiveness, potential risks, and diagnostic yield. This preference-based grouping reflects real-world clinical decision-making. No formal sample size calculation was performed prior to study initiation. Instead, a convenience sample was used. The sample size was determined by the total number of eligible cases during this predefined enrollment period.

INCLUSION AND EXCLUSION CRITERIA:

The inclusion criteria were: (1) patients with suspected endometrial tuberculosis, (2) those undergoing first examination, and (3) those whose family members were informed of the study and agreed to follow-up. Exclusion criteria were: (1) pregnant patients, (2) patients with other endometrial diseases, (3) those infected with other pathogens, and (4) those complicated with other organ diseases.

SAMPLING METHODS:

All procedures were performed in an outpatient setting under local anesthesia. After confirmation of uterine position and aseptic preparation, a paracervical block with 2% lidocaine was administered (typical total volume 10 mL; ~5 mL at the 4- and 8-o’clock positions), with up to an additional 5–10 mL permitted if discomfort persisted, not exceeding standard maximum dose limits. A waiting period of approximately 2–3 minutes was observed before instrumentation. No intravenous sedation or general anesthesia was used. To reduce the risk of procedure-related infection, all patients received a single prophylactic antibiotic dose 30–60 minutes before sampling (eg, first-generation cephalosporin if no β-lactam allergy). No routine post-procedure antibiotics were prescribed unless clinically indicated.

Patients in the fractional curettage group underwent fractional curettage for sampling, a conventional gynecological procedure involving sequential scraping of the endometrial lining using a metal curette for histopathological examination. Before sampling, the position of the uterus was confirmed, the vulva and metal curette were disinfected, and the patients were placed in the lithotomy position. Tissues were scraped from the cervical canal first and then from the cervix uteri and the endometrium. In the negative-pressure aspiration group, samples were acquired using a negative-pressure endometrial sampler (GYNEMED GmbH & Co. KG, Germany). The preparatory steps were identical to those in the fractional curettage group. The tip of the negative-pressure aspirator was inserted into the cervix, and gently and slowly pushed until reaching the uterine fundus, which was adjusted so that the mouth of the sampler was tightly adhered to the endometrial wall. Next, the sampler was pumped until it was in a negative-pressure state, and the endometrial tissues at the uterine fundus and the surrounding part were slowly aspirated.

The operations in both groups were performed by the same physician, and all the collected endometria were submitted to a specialized pathologist. For the criterion standard diagnosis, surgical samples obtained during subsequent operative management were subjected to histopathological examination (including hematoxylin-eosin staining and Ziehl-Neelsen acid-fast staining) and mycobacterial culture. Molecular assays such as PCR were not routinely performed for endometrial specimens at our institution during the study period. A diagnosis of endometrial tuberculosis was confirmed if any of the following were present in the surgical sample: histopathological evidence of granulomatous inflammation with caseous necrosis, detection of acid-fast bacilli, or positive mycobacterial culture. The diagnostic consistency of the sampling methods was assessed by comparison with the criterion standard using Cohen’s kappa coefficient. The kappa value was calculated based on the agreement between the results from each sampling method and the criterion standard (surgical samples).

Patients with histopathological or microbiological findings consistent with genital tuberculosis were referred to the hospital’s tuberculosis clinic and started on standard anti-tuberculosis therapy (treatment decisions were independent of sampling method). We did not administer empiric anti-tuberculosis therapy prior to tissue diagnosis.

EVALUATION OF PAIN DEGREE AND SAMPLING TIME:

The Face Rating Scale (FRS), a widely used and validated visual analog tool in clinical settings, is employed to assess patients’ pain perception during sampling [11]. The pain sensation of the 2 groups of patients during sampling was rated using the FRS, including 6 faces scored with even numbers 0, 2, 4, 6, 8, and 10, respectively. More faces signified stronger pain sensation; that is, zero faces indicated no pain, and 10 faces represented the severest pain [12]. The sampling time in both groups was recorded.

ASSESSMENT OF SATISFACTION WITH SAMPLING:

The satisfaction with sampling of the 2 groups of patients was recorded through the self-made satisfaction questionnaire of the hospital, with a total score of 100 points and 3 grades: very satisfied (>85 points), satisfied (70–85 points), and dissatisfied (<70 points), with the satisfaction (%) = (very satisfied cases + satisfied cases)/total cases×100%. Prior to the study, the questionnaire underwent reliability and validity testing, demonstrating a Cronbach’s alpha coefficient of 0.85, indicating high internal consistency, and acceptable content validity through expert consultation and pilot testing.

COMPARISON OF EXAMINATION RESULTS:

The examination results of the 2 groups of patients obtained through different sampling methods were recorded and then compared with the pathological biopsy results for their consistency. Later, the sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of the examination results in the 2 groups acquired using different sampling methods were calculated based on the following formulas: Sensitivity (%)=true positive cases/(true positive cases + false negative cases)×100%, specificity (%) = true negative cases/(true negative cases + false positive cases)×100%, accuracy (%) = (true positive cases + true negative cases)/total cases ×100%, positive predictive value (%) = true positive cases/(true positive cases + false positive cases), negative predictive value (%) = true negative cases/(true negative cases + false negative cases) [13].

OPERATIONAL SAFETY EVALUATION:

The presence of hemorrhage ≥10 mL, uterine perforation, and cervical laceration during operations was recorded for both groups. The absence of these conditions was defined as operational safety, while their presence was defined as a lack of operational safety [14]

STATISTICAL ANALYSIS:

Statistical analysis was performed using SPSS 22.0 software. Categorical variables, including menstrual status, satisfaction, pathological examination results, sensitivity, specificity, accuracy, positive predictive value, negative predictive value, and content of operational safety evaluation, are presented as n (%). Comparisons between groups for these categorical variables were conducted using the chi-square (χ²) test. The numerical value of satisfaction grading results was presented as n (%) and the Mann-Whitney rank sum test was employed for comparison. Patient age, course of disease, endometrial thickness, body mass index, FRS score, and sampling time are presented as mean±standard deviation. These variables were tested for normal distribution using the Shapiro-Wilk test prior to analysis. If normally distributed, between-group differences were assessed using the independent samples t test. 95% Confidence intervals (CIs) for kappa values, sensitivity, and specificity were reported. P<0.05 indicates statistical significance.

Results

BASELINE CLINICAL DATA:

No significant differences were found in the baseline clinical data of the 2 groups (P>0.05) (Table 1).

PAIN DEGREE AND SAMPLING TIME:

Patients in the negative-pressure aspiration group had a significantly lower FRS score and a significantly shorter sampling time than those in the fractional curettage group (P<0.05) (Table 2).

PATIENTS’ SATISFACTION WITH SAMPLING:

There was a significant difference in the distribution of populations with different satisfaction between the 2 groups (P<0.05), and the satisfaction was significantly higher in the negative-pressure aspiration group compared with that that in the fractional curettage group – 87.50% (42/48) vs 66.67% (24/36) (P<0.05) (Table 3).

EXAMINATION RESULTS OF PATIENTS:

The kappa value for diagnostic agreement with the criterion standard was 0.692 (95% CI: 0.503–0.881) in the negative-pressure aspiration group and 0.667 (95% CI: 0.462–0.872) in the fractional curettage group, indicating substantial agreement in both methods (Table 4).

DETECTION VALUE OF DIFFERENT SAMPLING METHODS:

The sensitivity of the negative-pressure aspiration group was 89.66% (95% CI: 75.8–97.1%), higher than that of the fractional curettage group (80.00%, 95% CI: 63.1–91.6%) (P<0.05), while the specificity was lower, at 78.95% (95% CI: 54.4–93.9%) vs 87.50% (95% CI: 61.7–98.4%) (P<0.05). However, no significant differences were observed between the 2 groups regarding the accuracy, positive predictive value, and negative predictive value (P>0.05) (Table 5).

RESULTS OF OPERATIONAL SAFETY EVALUATION OF DIFFERENT SAMPLING METHODS:

The operational safety of the sampling methods was assessed, showing that complications such as hemorrhage ≥10 mL, uterine perforation, and cervical laceration occurred in both groups, but there were no significant differences in the number of patients with each complication (P>0.05). Moreover, the total complication rate in the negative-pressure aspiration group was significantly lowered compared with that in the fractional curettage group (10.42% vs 27.78%, P<0.05) (Table 6).

Discussion

Endometrial tuberculosis significantly contributes to infertility and various reproductive tract complications, typically spreading from pulmonary tuberculosis via lymphatic or hematogenous routes [15]. Many patients exhibit atypical symptoms such as chronic lower abdominal and pelvic pain, pelvic cyst, menstrual dysfunction, dysmenorrhea, and postmenopausal bleeding [16]. Although imaging methods like chest X-ray and ultrasonography can identify suggestive features, these techniques alone lack diagnostic specificity and reliability [17,18]. Laparoscopic surgery can provide direct visualization and definitive diagnosis through biopsy, but it remains invasive and unsuitable for early or asymptomatic cases [19]. Therefore, effective and less invasive biopsy methods are critical for improving diagnostic accuracy and patient outcomes [20,21].

Traditionally, fractional curettage combined with transvaginal ultrasonography has been widely used for diagnosing endometrial tuberculosis. However, it poses substantial risks such as uterine perforation, infection, and patient discomfort [22]. In contrast, endometrial samplers have been widely utilized due to their simplicity, safety, and convenience [9]. This study compared fractional curettage with negative-pressure aspiration sampling, revealing significant advantages of the latter method, including shorter sampling time, reduced pain (as indicated by lower FRS scores), and increased patient satisfaction. This is consistent with a study by Saito et al [23], which demonstrated that negative-pressure sampling yielded high-quality specimens without complications, verifying its potential as a safer alternative to fractional curettage.

The diagnostic agreement (kappa value) and sensitivity of negative-pressure aspiration (kappa=0.692, sensitivity=89.66%) surpassed those of fractional curettage (kappa=0.667, sensitivity=80.00%), although the specificity was slightly lower. Similar results have been reported in previous studies. For instance, Hwang et al [24] found strong diagnostic agreement between aspiration sampling and surgical pathology in cases of endometrial hyperplasia. Sirimai et al [25] also supported the reliability of negative-pressure sampling, reporting a high kappa value of 0.86. During fractional curettage, surgeons cannot observe the lesion location and clearly determine the lesion site, so they may miss some tiny hidden lesions during lesion localization, leading to inaccurate diagnostic results [26,27].

Regarding procedural safety, our study showed fewer complications such as uterine perforation with negative-pressure aspiration compared to fractional curettage (2.08% vs 5.56%), although these results should be interpreted cautiously due to the relatively small sample size, but a previous study similarly identified a lower risk profile associated with aspiration sampling [28].

Nevertheless, this study has limitations. First, we used a non-randomized design with patient preference-based group allocation, which may have introduced selection bias despite standardized pre-procedure counseling. Second, no formal sample size calculation was conducted, and a convenience sample was used based on case availability during the study period, which may affect generalizability and statistical power. Third, although FRS was used to assess procedural pain due to its simplicity and practicality in outpatient settings, it lacks sufficient validation in adult gynecologic populations and may be influenced by individual emotional state and pain tolerance. Future studies should consider combining the FRS with more objective or validated tools such as the Visual Analog Scale. Additionally, this study did not evaluate endometrial thickness, which may be altered in endometrial tuberculosis and could affect sampling yield. Finally, the relatively small sample size limited subgroup analysis, particularly regarding the occurrence of rare complications. Larger-scale, randomized studies with extended follow-up are needed to further confirm the safety, diagnostic accuracy, and patient acceptability of negative-pressure aspiration in screening for endometrial tuberculosis.

Conclusions

Negative-pressure endometrial sampling technique shows clear advantages over traditional fractional curettage in clinical screening for endometrial tuberculosis, particularly by reducing patient discomfort, shortening sampling time, and increasing patient satisfaction. Although diagnostic accuracy is comparable between the 2 methods, negative-pressure aspiration notably increases sensitivity. Therefore, it is recommended for broader clinical application.