Validation of ICD-10 codes for identifying cases of chlamydia and gonorrhea

Background

Public health researchers seek to use administrative health data captured in digital health systems to examine outcomes for individuals with sexually transmitted infections (STIs). Yet the International Classification of Diseases, Tenth Revision (ICD-10) codes used to identify cases of chlamydia and gonorrhea have not been validated.

Objective

We sought to assess the validity of using ICD-10 codes to identify cases of chlamydia and gonorrhea.

Methods

We utilized data from electronic health records gathered from private and public health systems from 1 October, 2015 to 31 December, 2016. Patients were included if they were aged 13-44 and received either 1) laboratory testing for chlamydia or gonorrhea or 2) an ICD-10 diagnosis of chlamydia, gonorrhea, or an unspecified STI. To validate ICD-10 codes, we calculated positive and negative predictive values, sensitivity, and specificity based on the presence of a laboratory test result, or any STI laboratory test results in case of unspecified STI. We further examined the timing of clinical diagnosis relative to laboratory testing.

Results

A total of 238,876 individuals (16.0% of population) were either tested for chlamydia or gonorrhea, or diagnosed with an ICD-10 code of interest, during the study period. For cases in which a patient was diagnosed with chlamydia or gonorrhea, 82% and 78% of cases were confirmed, respectively. The positive predictive values for chlamydia, gonorrhea, and unspecified STI ICD-10 codes were 87.6%, 85.0%, and 32.0%, respectively. Negative predictive values were high (>92%). Sensitivity for chlamydia diagnostic codes was 10.6% and gonorrhea was 9.7%. Specificity was 99.9% for both chlamydia and gonorrhea.

Conclusions

Disease specific ICD-10 codes accurately identify cases of chlamydia and gonorrhea. However, low sensitivities suggest that most gonorrhea and chlamydia cases could not be identified in administrative data alone without laboratory test results.

Key messages

Disease specific ICD-10-CM codes accurately identify cases of chlamydia and gonorrhea. Low sensitivities suggest that most gonorrhea and chlamydia cases could not be identified in administrative data alone without laboratory test results.

Men with broken condoms: who and why?

OBJECTIVES

To identify (1) the prevalence of condom breakage, and demographic and sexuality-related differences among young men who have sex with women reporting and not reporting this event; (2) condom-specific behaviours associated with breakage.

METHODS

Young men (n = 278) attending a clinic for treatment of sexually transmitted infections (STIs) responded to an anonymous questionnaire aided by a CD recording of the questions. The samples were screened to include only men who had used a condom during penile-vaginal sex at least three times in the past 3 months. Condom-specific behaviours (including breakage) were assessed using these last three acts of condom use as the recall period. Correlates achieving bivariate significance were subjected to multivariate analysis.

RESULTS

Nearly one third (31.3%) of the men reported recent breakage. The breakage rate was 15%. Three correlates significantly distinguished between men who did and did not report breakage. Men who had past STIs were more likely to report breakage (adjusted odds ratio (AOR) 2.08), as were men who also reported condom slippage (AOR 2.72). Less self-efficacy for correct condom use was also significantly associated with breakage (AOR 1.07). Further, three condom-specific behaviours were significantly associated with breakage: allowing condoms to contact sharp objects (AOR 2.6), experiencing problems with the "fit or feel" of condoms (AOR 2.3) and not squeezing air from the receptacle tip (AOR 2.0).

CONCLUSIONS

Breakage may be common and may occur in a larger context of difficulties with condoms. STI clinics could potentially benefit some men by providing instructions on the correct use of condoms.

Identification of T cells that respond to serovar-specific regions of the Chlamydia trachomatis major outer membrane protein in persons with serovar E infection

Protection from Chlamydia trachomatis infection is serovar-specific and T cell-dependent; however, the T cell epitopes identified to date have not been serovar-specific. The chlamydial major outer membrane protein (MOMP) contains serovar-specific B cell epitopes in four regions of the molecule whose amino acid sequence varies among serovars, the variable segments (VS). Serovar-specific T cell epitopes were sought by examining proliferation of blood mononuclear cells (PBMC) from Chlamydia-infected patients in response to VS peptides of serovar E MOMP. Serovar E-specific peptides from VS1, VS2, and VS4 stimulated PBMC to a greater extent in serovar E-infected than in non-E-infected subjects. Peptides containing constant regions of MOMP were recognized equally by all infected persons. The observed responses were attributable to T cells. T cell recognition of serovar-specific regions of MOMP is common and may contribute to the serovar-specific protection previously observed.

Chlamydia trachomatis and febrile complications of postpartum tubal ligation

OBJECTIVE

Our goal was to determine whether chlamydia-infected women have a higher rate of febrile complications after postpartum tubal ligation.

STUDY DESIGN

Cross-sectional analysis of 1447 women tested for chlamydial infection within 2 weeks of delivery and who underwent postpartum tubal ligation was performed. Subjects were identified with the Regenstrief Institute for Health Care database. Infected subjects were compared with uninfected subjects for incidence of fever not explained by nongynecologic sources.

RESULTS

Women infected with Chlamydia trachomatis at delivery were more likely to experience febrile postoperative complications after tubal ligation (p < 0.0001, relative risk 9.5, 95% confidence interval 4.5 to 20.1).

CONCLUSION

Women undergoing postpartum tuba ligation may benefit from prompt diagnosis and preoperative treatment of chlamydial infection.

Haemophilus ducreyi elicits a cutaneous infiltrate of CD4 cells during experimental human infection

Human subjects were experimentally infected with Haemophilus ducreyi for up to 2 weeks. Bacterial suspensions were delivered into the epidermis and dermis through puncture wounds made by an allergy-testing device. Subjects developed papular lesions that evolved into pustules resembling natural disease. Some papular lesions resolved spontaneously, indicating that host responses may clear infection. Bacteria were shed intermittently from lesions, suggesting that H. ducreyi may be transmissible before ulceration. Host responses to infection consisted primarily of cutaneous infiltrate of polymorphonuclear leukocytes, Langerhans cells, macrophages, and CD4 T cells of alpha beta lineage. Expression of HLA-DR by keratinocytes was associated with the presence of interferon-gamma mRNA in the skin. There was little evidence for humoral or peripheral blood mononuclear cell responses to bacterial antigens. The cutaneous infiltrate of CD4 cells and macrophages provides a mechanism that facilitates transmission of human immunodeficiency virus by H. ducreyi.

Serologic responses of infertile women to the 60-kd chlamydial heat shock protein (hsp60)

OBJECTIVE

To determine whether women with Chlamydia trachomatis-associated tubal infertility are more likely than other infertile women to have antibodies to a particular region of the 60-kd chlamydial heat shock protein, hsp60.

DESIGN

Serologic responses to the chlamydial hsp60 were examined in 43 infertile women seropositive for Chlamydia trachomatis, including 21 women with tubal infertility, 13 women with endometriosis, and 9 women with other causes of infertility. Antibody responses were localized to regions of hsp60 using five nonoverlapping recombinant polypeptides.

RESULTS

Sixteen women with tubal infertility had anti-hsp60 antibodies compared with seven women with endometriosis and two women with other causes of infertility. Antibodies of 11 women with tubal infertility reacted predominantly with a region of hsp60 containing amino acids (201 to 300) compared with 1 women without tubal infertility. In contrast, antibodies that localized to the carboxyl terminus, amino acids (401 to 544), were seen equally in all groups.

CONCLUSIONS

Among seropositive infertile women, antibodies that localized to amino acids (201 to 300) were immunodominant in those with tubal infertility but not in those with infertility due to other causes.

Age and clinical immunity to infections with Chlamydia trachomatis

BACKGROUND AND OBJECTIVES

The prevalence of chlamydial infection decreases with age possibly in part because of increasing immunity.

GOAL OF THIS STUDY

To determine whether increased age is an independent predictor of decreased chlamydial infection and whether chlamydia-specific antibody titer and blastogenesis increase with age.

STUDY DESIGN

Data from all patients cultured for Chlamydia trachomatis between January 1984 and August 1989 were examined and multiple logistic regression models were used to identify the independent predictors of culture positivity. Antichlamydial antibody titer and chlamydia-specific blastogenesis were examined for a subset of patients for correlation with age.

RESULTS

Young age was found to be predictive of chlamydial infection independent of all factors examined in men and women. Antibody titers had no relation to age (n = 245) whereas the level of blastogenesis correlated only weakly with age (n = 155).

CONCLUSIONS

Assays of systemic immunity do not reflect the protection from chlamydial infection associated with age.

Interferon-gamma in endocervical secretions of women infected with Chlamydia trachomatis

Although interferon-gamma has been associated with control of chlamydial infections in mice, no direct evidence links it to human chlamydial infections. Therefore, interferon-gamma was assayed by ELISA in endocervical secretions and plasma of women cultured for Chlamydia trachomatis. Women with positive endocervical chlamydial cultures had increased levels of interferon-gamma in endocervical secretions (6.7 +/- 2.8, mean +/- SE, n = 47) compared with uninfected women (1.4 +/- 0.4, n = 52) (P = .002). Interferon was also present in secretions of women with gonorrhea. Higher levels were seen in secretions from older women with positive chlamydial cultures. Interferon levels in secretions did not correlate with simultaneous plasma levels, the number of organisms recovered in tissue culture, or clinical correlates of inflammation. These data suggest that interferon-gamma is present at the site of chlamydial infection; however, further experiments are needed to determine whether interferon is specifically involved in protection or is a nonspecific indicator of inflammation.