Background
Chlamydia trachomatis infection is common and largely asymptomatic in women. If untreated, it can lead to sequelae such as pelvic inflammatory disease and infertility. It is unknown whether a patient’s self-reported history of Chlamydia trachomatis infection is a valid marker of past infection.
Objective
Our objective was to evaluate the validity of women’s self-reported history of Chlamydia trachomatis infection compared with Chlamydia trachomatis serology, a marker for previous infection.
Study Design
We analyzed data from the Fertility After Contraception Termination study. We compared participants’ survey responses with the question, “Have you ever been told by a health care provider that you had Chlamydia?” to serological test results indicating the presence or absence of antibodies to Chlamydia trachomatis as assessed by a microimmunofluorescence assay. Prevalence of past infection, sensitivity, specificity, predictive values, and likelihood ratios were calculated. The Cohen’s kappa statistic was computed to assess agreement between self-report and serology.
Results
Among 409 participants, 108 (26%) reported having a history of Chlamydia trachomatis infection, whereas 146 (36%) had positive serological test results. Relative to positive microimmunofluorescence assay, the sensitivity and specificity of self-reported history of Chlamydia trachomatis infection were 52.1% (95% confidence interval, 43.6–60.4%) and 87.8% (95% confidence interval, 83.3–91.5%), respectively. The positive predictive value of the self-report was 70.4% (95% confidence interval, 60.8–78.8%), and the negative predictive value was 76.7% (95% confidence interval, 71.6–81.4%). The likelihood ratio was found to be 4.28. Agreement between self-report and serology was found to be moderate (kappa = 0.42, P < .001).
Conclusion
Self-reported history of Chlamydia trachomatis infection commonly yields false-negative and false-positive results. When definitive status of past Chlamydia trachomatis infection is needed, serology should be obtained.
Chlamydia trachomatis (CT) infection is the most common notifiable sexually transmitted infection (STI) in the United States and is largely asymptomatic. Up to 80% of women with CT infections experience no symptoms. If CT infections are not diagnosed and treated in a timely manner, these infections can progress and lead to pelvic inflammatory disease, chronic pelvic pain, and infertility. Health care providers may ask about previous CT infection while taking a medical history, but it is unclear to what degree a self-reported history of CT infection is a valid measure of past infection.
Although the literature contains several studies assessing the validity of self-reported STIs, few studies specifically evaluate the validity of self-reported CT infection. One randomized controlled trial for an STI intervention program found that only 68% of African-American female teenagers with a laboratory-confirmed CT infection correctly reported their history of infection 1 month after learning of their diagnosis.
A study performed by Niccolai et al compared female adolescents’ self-reported CT diagnoses with a composite reference standard defined as positive if either the participant’s medical record or a state health department report showed the patient to have a history of CT infection. This study found a high specificity of 97.3% and a sensitivity of 69.1%.
The purpose of this analysis was to evaluate the validity of women’s self-reported history of CT infection compared with CT serology as assessed by microimmunofluorescence, a sensitive marker for previous infection with CT. Self-report and serology may not agree for many reasons, including a lack of understanding of test results, a desire to not disclose a history of positive test results, and the absence of prior testing, either because of the asymptomatic nature of the infection or a lack of access to testing. Given all these complexities surrounding a self-reported history of CT infections, we hypothesize that women will significantly underestimate their history of past infection and that a reported history of CT infection is not a reliable marker of past infection.
Materials and Methods
We performed a cross-sectional analysis of the Fertility After Contraception Termination (FACT) study. FACT is a prospective cohort study developed primarily to assess the role of infection and contraceptive use (specifically intrauterine device use) on fertility. We compared self-reported CT infection with serological results obtained at the time of baseline interview and evaluation.
Many of the FACT participants were recruited from the Contraceptive CHOICE Project, a prospective cohort study of 9256 women in the St Louis area who were provided with no-cost contraception for 2–3 years. A more in-depth methodological description of the CHOICE project has been previously described.
We augmented our sample size with recruitment from 4 additional clinical research programs (University of Pennsylvania, University of Colorado, University of Utah, and University of Southern California). Institutional review boards at all participating institutions approved this study, and written informed consent was obtained from all participants prior to study involvement.
English- or Spanish-speaking women between the ages of 18 to 35 years were eligible for the study if they were discontinuing a contraceptive method to attempt pregnancy. Women were excluded if they met the following criteria: (1) were pregnant at the time of enrollment; (2) did not have a male partner at the time of enrollment; (3) were going to be physically separated from their male partner for 6 months or longer; (4) were with a male partner who has a history of vasectomy, infertility, or abnormal semen analysis; (5) had a history of infertility, tubal reconstructive surgery, or sterilization; or (6) had medical problems known to affect fertility (eg, cancer therapy, thyroid problems, Cushing’s disease, sickle cell disease, kidney disease, and diabetes).
Baseline investigations included a questionnaire, a clinical examination, blood samples for serological testing for CT, Mycoplasma genitalium (MG), and Trichomonas vaginalis (TV), and nucleic acid amplification testing (NAAT, APTIMA, Gen-Probe) for current STIs ( Neisseria gonorrhoeae , CT, TV, and MG). Although both serology and nucleic acid amplification testing were available for CT, MG, and TV, we decided to focus our study on CT, given that it is the most common nationally notifiable sexually transmitted disease and it is likely a more well-known STI compared with MG or TV.
Baseline data collection included demographic, historical, and medical/surgical history and reproductive characteristics. As part of the baseline questionnaire, participants were asked whether they have ever been told by a health care provider that they had chlamydia. Participants’ yes or no answers to this question represented their self-reported history of CT infection.
The serological samples taken at baseline were used to assess current and/or previous CT infection. The samples were clarified by centrifugation, aliquoted, and frozen at –80°C. Samples were then transported on dry ice to the University of Washington Chlamydia Laboratory (Seattle, WA) for analysis using the microimmunofluorescence assay developed by Wang as modified by Hanna and Keshishyan.
All serological assays were performed by the same experienced technician. Based on a previous receiver-operating characteristic curve analysis of CT antibody titers and laparoscopically confirmed tuboperitoneal abnormalities, we considered a titer of 1:16 as evidence of a positive history of CT infection. Whereas previous studies have used cutoff dilutions ranging from 1:8 to 1:640, not all used the same technique or staff to perform this technically complicated assay.
Only those participants whose serological results were returned by the time of analysis were included. Dates of data collection for this analysis ranged from Sept. 8, 2011, to Feb. 25, 2016. Baseline demographic characteristics of this sample were compared using χ2, Fisher exact, and Student t tests, as appropriate. Prevalence of past infection was calculated using the serological data. Sensitivity, specificity, predictive values, and positive likelihood ratio for self-reported history of infection were also calculated.
Cohen’s kappa statistic was calculated to assess the agreement between self-report and serology. Bivariate logistic regression analysis was performed to assess the association between the baseline demographic characteristics and self-reported CT history and serological CT results. All statistical analyses were performed using SAS 9.4 software (SAS Inc, Cary, NC). The significance level of alpha was set at 0.05.
Results
At the time of data analysis, 440 women had enrolled in FACT. A total of 432 participants responded to the question, “Have you ever been told by a health care provider that you had chlamydia,” and 420 had serological data available. Four hundred nine participants had both data points available and are included in our data analysis.
The baseline demographic, reproductive, and behavioral characteristics by a self-reported history of CT infection and serological status are provided in Table 1 . The mean age of participants was 28 years. Forty-two percent of participants were black, 8% reported Hispanic ethnicity, 57% were married, 22% had a high school education or less, 38% were of low socioeconomic status, 36% smoked, 14% used drugs, and 24% had no insurance.
| Positive self-reported CT history (n = 108) | Negative self-reported CT history (n = 301) | P value | Positive serology (n = 146) | Negative serology (n = 263) | P value | |
|---|---|---|---|---|---|---|
| Age, y a | 26.3 ± 4.6 | 28.4 ± 3.8 | < .001 | 26.7 ± 4.4 | 28.5 ± 3.8 | < .0001 |
| Race | < .001 | < .0001 | ||||
| Black | 78 (73.6) | 90 (30.2) | 104 (71.7) | 64 (24.7) | ||
| White | 20 (18.9) | 186 (62.4) | 30 (20.7) | 176 (68.0) | ||
| Other/multiracial | 8 (7.6) | 22 (7.6) | 11 (7.6) | 19 (7.3) | ||
| Hispanic ethnicity | 9 (8.3) | 24 (7.9) | .9132 | 7 (4.8) | 26 (9.9) | .0685 |
| Marital status | < .001 | < .0001 | ||||
| Single/divorced/widowed | 40 (37.0) | 37 (12.3) | 49 (33.6) | 28 (10.7) | ||
| Living with partner | 38 (35.2) | 60 (20.0) | 52 (35.6) | 46 (17.6) | ||
| Married | 30 (27.8) | 203 (67.7) | 45 (30.8) | 188 (71.8) | ||
| Education | < .001 | < .0001 | ||||
| HS diploma, GED, or less | 41 (38.0) | 47 (15.7) | 49 (33.6) | 39 (14.9) | ||
| Some college | 52 (48.2) | 79 (26.3) | 70 (48.0) | 61 (23.3) | ||
| College or graduate degree | 15 (13.9) | 174 (58.0) | 27 (18.5) | 162 (61.8) | ||
| Low SES b | 68 (63.0) | 87 (29.2) | < .001 | 94 (64.4) | 61 (23.5) | < .0001 |
| Birth control method | < .001 a | < .0001 a | ||||
| Non-LARC | 20 (18.6) | 95 (32.4) | 25 (17.2) | 90 (35.3) | ||
| IUD | 48 (44.9) | 153 (52.2) | 76 (52.4) | 125 (49.0) | ||
| Implant | 35 (32.7) | 40 (13.6) | 38 (26.2) | 37 (14.5) | ||
| DMPA | 4 (3.7) | 5 (1.7) | 6 (4.1) | 3 (1.2) | ||
| Insurance | < .001 | < .0001 | ||||
| Private | 35 (32.7) | 213 (71.0) | 51 (35.2) | 197 (75.2) | ||
| Public | 28 (26.2) | 32 (10.7) | 36 (24.8) | 24 (9.2) | ||
| None | 44 (41.1) | 55 (18.3) | 58 (40.0) | 41 (15.6) | ||
| General health c | < .001 | < .0001 | ||||
| Excellent to very good | 48 (44.4) | 216 (71.8) | 72 (49.3) | 192 (73.0) | ||
| Good | 42 (38.9) | 77 (25.6) | 56 (38.4) | 63 (24.0) | ||
| Fair to poor | 18 (16.7) | 8 (2.7) | 18 (12.3) | 8 (3.0) | ||
| Current smoker | 20 (43.5) | 34 (32.7) | .2044 | 28 (45.2) | 26 (29.60) | .0498 |
| Current drug use | 22 (19.4) | 36 (12.2) | .0316 | 32 (21.9) | 26 (9.9) | .0008 |
| Gravidity | < .001 | < .0001 | ||||
| 0 | 21 (19.4) | 128 (42.7) | 26 (17.8) | 123 (47.0) | ||
| 1–2 | 57 (52.8) | 115 (38.3) | 74 (50.7) | 98 (37.4) | ||
| ≥3 | 30 (27.8) | 57 (19.0) | 46 (31.5) | 41 (15.7) | ||
| Age at first pregnancy d | 18.8 ± 3.8 | 20.7 ± 4.9 | .0012 | 18.7 ± 3.7 | 21.2 ± 4.9 | < .0001 |
a Data are shown as mean and SD
b Defined as receipt of public assistance or report of difficulty paying for basic necessities
c Assessed with the following question: “in general would you say your health is excellent, very good, fair, or poor?”
Seventeen participants (4.2%) had positive CT test results at the baseline encounter. Participants with a positive self-reported or serologically confirmed history of CT were more likely to be younger, of higher gravidity, nonwhite, unmarried, of lower educational level and socioeconomic status, former implant or depomedroxyprogesterone acetate users, uninsured, and a current drug user.
Table 2 compares self reported CT infection with serological status. Twenty-six percent of participants reported having a history of CT infection, whereas 36% of participants had positive serological test results. Of the 146 women who had antibodies to CT, 76 reported a history of CT infection. Conversely, of the 263 women who had no antibodies to CT, 231 reported never having a CT infection in the past.
| Positive serology (n = 146) | Negative serology (n = 263) | |
|---|---|---|
| Positive self-report (n = 108) | 76 | 32 |
| Negative self-report (n = 301) | 70 | 231 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
