Seroprevalence and risk factors on Syphilis among blood donors in Chengdu, China,from 2005 to 2017

The blood samples were collected from blood donors in Chengdu from January 2005 to December 2017. Donors with seropositivity of syphilis alone were selected as cases. Controls were matched to cases to control confounding. For each positive case, two syphilis-negative, age- and sex-matched donors were selected as controls. Between January 2005 and December 2017, 368 positive cases and 736 controls were included in the prevalence study. This study was approved by the Medical Ethics Committee of Sichuan University (NO:2014015–02).

Blood specimens from the blood donors were tested for HBsAg, anti-HCV, anti-HIV (types 1 and 2), syphilis, and ALT according to procedures stipulated in the Chengdu blood center. The equipment of anti-TP testing: FAME24/20 automatic enzyme immunoassay system (Hamilton, USA). Reagents: anti-TP diagnostic kit (Beijing Wantai, lot number N20110405, N20110708; Beijing Huada love, lot number 20110616, 20110915). Blood specimens were carried out using two different test kits that were approved by the Chinese Food and Drug Administration. Each donation was screened by two ELISA kits regardless of the result of the first one. Each initial reactive sample was further tested in duplicate, according to the manufacturer’s instructions. If two out of three results were positive (Repeat Reactive), the unit was discarded and the sample was classified as positive.

Donors with anti-TP infection alone were selected as cases. Controls were matched to cases to control confounding. All the selected donors were contacted by telephone obtained from donor registration forms. After explanation of the study aim some blood donors agreed to participate in the study. A risk assessment questionnaire (Additional file 1) was given to every eligible donor in order to investigate relevant information. The main contents of the questionnaire include: the information consisted of facial shaving by sex, age, nationality, occupation, educational level and marital status; related behaviors including sharing razor, tattooing, dental treatment, ear piercing, transfusion, acupuncture, injection, dental treatment, body fluid contact, condom use, paid sex, sexual contact with syphilis patients. The investigation was conducted by professional medical personnel in order to ensure the accuracy of data collection. All the information was confidential.

Calculated the rate of TP-positive, seroprevalence and serological incidence based on the test results. The rate of TP-positive refers to the proportion of the total number of syphilis-positive in the total blood donations, that is, the total number of syphilis-positive as a molecule, and the total population of blood donations as the denominator. Seroprevalence was calculated using the total number of syphilis-positive donors as the numerator in the initial blood donation, the total number of initial blood donors at the same period as the denominator. Serological incidence (%): The total number of syphilis-positive in the repeat blood donors in that year was the numerator, all repeat blood donors during the same period was the denominator. Database was established using Epidata3.1. Data were analyzed using SPSS version 22.0. Possible risk factors for screening of meaningful variables were analyzed by using conditional logistic regression and estimates of the odds ratios (ORs) with their corresponding 95% confidence interval (CI). The crude odds ratios (ORs) and 95% CI were estimated by the univariate conditional logistic regression analysis. Potential interactions between razor sharing and tattooing, ear piercing, condom and number of sex partner were assessed. Correlation analyses, such as Spearman rank correlation were also performed for testing collinearity among independent variables. Forward stepwise methods were used to determine which variables significantly contributed to syphilis infection. All of statistical tests were two-sided and a level of P < 0.05 was used to indicate statistical significance. The proportion of all the TP-positive cases in the population of the voluntary blood donors attributable to multiple risk factors (population attributable risk, PAR) was estimated by using Bruzzi’s formula [18] to the observed OR.

A total of 2,100,071 voluntary blood donors were selected in the Chengdu blood center during 2005–2017 for the study. It is worth mentioning that all participants in the study have normal levels of transaminase, and HIV, HBV and HCV tests are negative. Of these, 20,510 blood donations were positive for syphilis (Table 1). The rate of anti-TP positivity was 0.88% corresponding to 815 repeatedly reactive ELISA tests out of 92,610 blood donations in 2005 and the rate of TP-positivity was 0.98 in 2017. The seroprevalence of syphilis in blood donors from 2005 to 2017 in Chengdu also showed an overall upward trend (Fig. 1), while the serological incidence was fluctuated. The seroprevalence of the first-time blood donors in the study period were higher than the serological incidence (Table 1).

During the case-control study, 368 (33.3%) of the 1104 eligible anti-TP-positive donors were enrolled in the study. After obtaining informed consent for the study, 368 anti-TP-positive cases and 736 controls were investigated. Reasons for non-response included contact lost and refusing to participate in the study. Overall and subgroup-specific prevalence of T.pallidum seropositivity and active syphilis infection were calculated using case definitions described above. The demographic characteristics associated with all samples were shown in Table 2. Demographic variables included marital status, educational attainment, sex, and occupation. In the study, no significant differences were found between the cases and the demographic and socioeconomic characteristics of controls.

By conditional logistic regression, occupation as a risk factor significantly associated with syphilis infection was student, the education was bachelor. Razor reuse, ear piercing, tattoo, dental operation, paid sex, number of sex partner, condom and sexual contact with syphilis (P < 0.05; Table 3). After the forward stepwise method was used for variable selection in the multiple conditional logistic regression, interaction terms were checked in the final model. It is shown that no statistically significant interactions were found among these variables: occupation and education. Dental operation and paid sex did not show significant associations with syphilis infection (P > 0.05; Table 4).

Multiple condition logistic regression model denoted the following risk factors for increasing rates of syphilis infections: multiple sexual partners (OR = 7.1; 95% CI = 3.685, 13.599), razor reuse (OR = 1.7; 95% CI = 1.252, 2.261); ear piercing (OR = 2.7; 95% CI = 1.877, 3.916); tattoo (OR = 3.3; 95% CI =1.492, 7.167); condom occasionally (OR = 2.8; 95% CI = 1.802, 4.399). The estimates of the PAR indicated that number of sex partner for syphilis infection accounted for 22.5% of all TP-positive cases (Table 5). Razor reuse, ear piercing, tattoo, condom, and Sexual contact with syphilis accounted for 14.4, 14.7, 1.8, 12.9, 1.8% respectively (Table 5). However, The estimates of the PAR about uncertain sexual contact with syphilis in TP-positive cases was 8.1%. In fact, 68.78% of all TP-positive cases occurring in subjects were closely related to these risk factor. The PAR was used to assess the importance of a risk factor among subjects because it is a function of both the relative risk of exposure to that factor and the prevalence of exposure within the population.