Due to their high cost and time-consuming nature, invasive diagnostic methods are obviously not convenient for every patient presenting to various clinics with gastro-intestinal complaints. Instead, non-invasive methods are more commonly preferred as they are faster and more economical alternatives. In the present study, we aimed to compare two serological test methods, rapid cassette test and micro ELISA method, using patient serum samples.
Several authors have examined non-invasive methods for diagnosis of H. pylori infections using various body secretions (saliva, stool, etc.). For example, Felz et al.  applied invasive (upper gastrointestinal endoscopy and biopsy) and non-invasive methods (rapid urease test and [14C] urea breath test) for diagnosing of H. pylori in 26 patients with chronic gastritis. All (100%) of the 20 patients that had histologically confirmed diagnosis of H. pylori infection were found to have strong positive results from urea breath test. Rahman et al.  evaluated the performance of immunoblot test method and immunochromatographic (ICT) tests including CIM and ELISA. They measured anti-H pylori IgG antibodies using ELISA, ICT (H. pylori rapid test), and immunoblot methods in totally 82 serum samples, of which 61 were obtained from patients with confirmed H.pylori infection using endoscopy. They calculated sensitivity, specificity, positive predictive value, and negative predictive value for these methods, and found that ELISA method had a sensitivity (96.7%) that was close to immunoblot test method (98.3%), and was less expensive than the latter as well. Demiray et al.  measured anti-Helicobacter pylori IgG antibodies using URINELISA, RAPIRUN, and anti-Helicobacter pylori ELISA methods in urine and serum samples of 124 patients that had dyspeptic complaints and underwent upper gastrointestinal system endoscopic examination. Of these patients, 69 patients had positive results from both URINELISA and RAPIRUN, while 109 patients had positive results from anti-H. pylori IgG ELISA. The sensitivities of these tests were found as 74.4%, 73.2% and 100%, respectively, and specificities were 81.0%, 78.6% and 35.7%, respectively. In our study, we examined serum samples of 224 patients that had gastrointestinal complaints, using three non-invasive serological test methods (rapid diagnostic test and micro ELISA IgA and IgG). Among these methods, micro ELISA IgG had the highest positivity rate and yielded positive results in 139 (62.1%) patients. In comparison to endoscopy results, IgG cassette test, ELISA IgA test, and ELISA IgG tests had sensitivity rates as 77.4%, 85.7% and 85.7%, and specificity rates as 76.9%, 52.0% and 44.0%, respectively.
In one similar study, Tsongo et al.  measured serum samples from 174 patients with gastroduodenal ulcer symptoms, using rapid diagnostic test, and stool samples of the same patients using ELISA method, and they compared the two methods. Accordingly, the prevalence of H. pylori was found as 29.9% (52/174) with ELISA method and 37.4% (65/174) with the rapid test, respectively. In addition, they applied a questionnaire to assess the socioeconomic levels and life styles of their patients. Accordingly, smoking, poor sanitation and lack of formal education were found as predisposing factors to infection (p<0.05). The two methods yielded the same results in 87.9% of the patients. In our study, we rather measured serum samples with three serological methods, and we found prevalence rates as 29.5% (66/224) with IgG cassette test, 50.9% (114/224) with ELISA IgA test and 62.1% (139/224) with ELISA IgG test. As observed, prevalence rates were somewhat higher with ELISA methods. This may be related to the selection of the patient group that we included in this study.
In a 2013–14 study with 160 students, Yo et al.  obtained 0.5 ml of saliva samples and analyzed them within 5 minutes using a H. pylori Saliva Test Cassette. They found positive results in 82 of the subjects and negative results in 78, and calculated the oral H. pylori infection rate as 51%. In addition, the subjects were questioned in terms of smoking, dietary and dental care habits and family history. Of the 82 subjects with positive test results, 74 had poor dental care. We did not apply the questionnaire to our patients; however, the results we obtained from the serological tests (positive results for IgG cassette, ELISA IgA, and ELISA IgG were 54 (61.4%), 66 (75.0%) and 68 (77.3%), respectively) were consistent with the findings of the aforementioned study.
Agbor et al.  conducted a prevalence analysis to determine the epidemiological profile of H. pylori infection by obtaining blood and stool samples from 500 patients with gastric complaints between 2013 and 2015. They used a one-step H. pylori antibody device for serum analysis. Three drops of serum was put into a well in the device and the result was obtained after 10 minutes. For stool analysis, a one-step H. pylori antigen test device was also used; 50 mg solid or two drops of watery stool sample was transferred to a tube and mixed with buffer. Two drops of this mixture was put into a well in the device and the result was obtained after 10 minutes. Of the 500 stool samples examined, 237 (43%) were positive for the H. pylori antibody test. Seropositivity rates were found to be similar between females and males (42% and 45%, respectively). They calculated the sensitivity and specificity of the antibody test (90% and 98%, respectively) in reference to the antigen test, which yielded higher positive rates. Twenty four samples were positive with the antigen test but negative with the antibody test. In contrast, 4 samples were positive with the antibody test but negative with the antigen test. Both tests yielded positive results for 213 of the samples. As a result, the authors noted that there was no significant difference between the two test methods (p = 0.204). In our study, we did not measure antigens in the stool; however, we used three different antibody tests and compared the results to the endoscopy results. The positivity rates that we obtained from the antibody tests were close to the results of the study mentioned above: 29.5%, 50.9%, and 62.1% for IgG cassette test, ELISA IgA test and ELISA IgG test, respectively. However, females had a little higher positivity rate in our study (63.5% in females vs. 36.5% in males, according to the endoscopy reports).
She et al.  retrospectively reviewed 4,722 samples (58% female and 42% male) measured between the years 1998–2009, and compared the results of serum H. pylori IgG, IgA and/or IgM with the results of stool H. pylori antigen test (HpSA). The positivity rate of HpSA (12.1%) was significantly lower compared to IgG (35.6%) and IgA (32.7%) tests (p<0.001), whereas the positivity rate of IgM (4.3%) test was lower than the other three tests (p = 0.001). When HpSA was taken as the gold standard, sensitivity, specificity, positive predictive value (PPV)m and negative predictive value (NPV) in all age groups were 87.6%, 61.0%, 22.8% and 97.4% for IgG test, 63.4%, 67.6%, 17.6% and 94.4% for IgA test, and 6.8%, 95.8%, 13.6% and 91.2% for IgM test, respectively. IgG test was found to have better correlation with HpSA than IgA and IgM. Additionally, IgM was reported to have little diagnostic value in H. pylori infections. In our study, although the number of samples was lower, our rates were similar to that study. On the other hand, we compared serological test results with endoscopy instead of the HpSA results.
Kazemi et al.  examined 94 patients with dyspeptic complaints who underwent endoscopic examination and evaluated rapid urease test, C-urea breath test, histological examination reports, serum antibody and stool antigen tests. The authors concluded that the stool antigen test was more appropriate than UBT for diagnosing H. pylori infection in untreated patients. Similar to that study, we compared the results of serological tests with the gold standard endoscopy reports, and our sensitivity, specificity, PPV, NPV and accuracy values were 77.4%, 76.9%, 88.9%, 58.8%, and 77.3% for IgG cassette test, respectively.
In another study, Veijola et al.  recruited their study participants via a newspaper advertisement, and included 1,574 adult subjects that did not receive antibiotic treatment for the last 2 months, or H2-receptor antagonists or bismuth or proton pump inhibitors for the last 2 weeks, or receive H. pylori eradication treatment for the last 5 years, or have history of gastric operation, chronic GIS disease, pregnancy or lactation. The subjects were tested with rapid whole blood antibody (IgG) test for diagnosis of H. pylori infection, and the 300 subjects with positive test results were confirmed with UBT and an in-house EIA-based serological assay (IgG and IgA). Of the 300 subjects, 196 were confirmed positive with both methods; however, since 11 subjects did not meet the inclusion criteria, 185 positive subjects were left. With the addition of 97 subjects who had positive results from confirmatory tests despite having negative screening test results, the total number of subjects enrolled in the study was 282 (186 females and 96 males). One hundred eighty five subjects who had positive results from all three methods were enrolled in the eradication program, and they were retested after 4 months with serological methods. The success criterion for eradication therapy was defined as at least 40% reduction in IgG antibody level. The performance of the three stool antigen tests, HpSA (polyclonal antibody-based), HpStAR (mAb-based Amplified IDEIA) and ImmunoCard (based on monoclonal H. pylori antibody and a lateral flow chromatography technique), which were applied to the subjects both before and after the eradication treatment, was evaluated in comparison to UBT and serology. Accordingly, pre-eradication sensitivity, specificity, PPV and NPV values of the three stool antigen tests were found as 91.9%, 95.9%, 97.7% and 87.7% for HpSA, 96.2%, 95.9%, 97.8% and 93.0% for HpStAR, and 93.0%, 88.7%, 94.0% and 86.9% for ImmunoCard, respectively. Post-eradication values were 81.3%, 97.0%, 76.5% and 98.2% for HpSA, 100%, 97.6%, 80.0% and 100% for HpStAR, and 93.8%, 97.0%, 75.0% and 99.4% for ImmunoCard, respectively. In our study, we included patients presenting to various clinics. Our female to male ratio was similarly high (63.5%), and another interesting finding was that we found lower sensitivity, specificity, PPV, NPV and accuracy values (85.7%, 44.0%, 79.4%, 55.0% and 73.9%, respectively, for ELISA IgG). This difference might be related to the patient selection.
In the study of Abu Shady , 100 pediatric patients with an age range of 4–10 years, who were referred to endoscopic examination due to upper GIS complaints were tested with rapid urease test (RUT) and biopsied for histological examination. For RUT, the result was obtained after adding biopsy sample to the urea solution (NaCl, KH2PO4 and NaOH). A change in the color of the urea solution from yellow to red, due to increase in pH induced by H. pylori, was accepted as a positive test result. Histological examination was performed after staining with hematoxylin and eosin. Additionally, a microplate enzyme immunoassay (EIA) and an antibody detection kit were used to detect antibodies against H. pylori (IgG) in patient serum samples. The analysis was performed per manufacturer’s instructions and cutoff threshold was 10U/mL The gold standard for diagnosis of H. pylori infection was accepted as positive results from both histological examination and the rapid urease test. Accordingly, while standard test result was positive in 57% and negative in 43% of patients, serological test was positive in 60% and negative in 40% of patients. In addition, sensitivity and specificity values of anti-H. pylori IgG antibody test were found as 96.5% and 93%, respectively. The authors concluded that IgG antibody test was a good alternative to invasive diagnostic tests such as urea breath test, for diagnosis of H. pylori infection, and that IgG type of antibodies produced against H.pylori had higher diagnostic value. In contrast to that study, our study included adult patients. Although the study designs were similar, we had lower sensitivity and specificity values. This was due to the fact that our study sample did not include children.