Mycoplasma genitalium: An Emerging Cause of Pelvic Inflammatory Disease

Table 1

Studies evaluating the relationship between M. genitalium and pelvic inflammatory disease.

Citation	Sample size, population, setting	Study design	Methods: M. genitalium test PID diagnosis	Findings	Validity
Uno et al. [13]	200 patients aged 19 to 49 years visiting the OB Gyn department of Kizawa Memorial Hospital and Hayasaki Ladies Clinic.	Cross-sectional	M. genitalium: PCR of endocervical specimens.; PID: clinical criteria.	5.7% of PID patients versus 0% of pregnant controls tested positive for M. genitalium (𝑃-value or OR not reported, one patient co-infected with C. trachomatis).	Strengths: Control group of patients without signs and symptoms of PID.; Limitations: No laparoscopic or histologic confirmation of PID.; Although C. trachomatis was assessed, sample size too small to determine independent effect of M. genitalium
Cohen et al. [6]	115 patients presenting with pelvic pain ≤ 14 days presenting to a sexually transmitted diseases clinic, Nairobi, Kenya between 2000–2003.	Cross-sectional	M. genitalium: PCR of cervical and endometrial samples.; PID: histologically confirmed endometritis.	M. genitalium detected in 16% of patients with endometritis versus 2% of patients without endometritis (𝑃=0.03).; M. genitalium identified in the endometrium was associated with endometritis after excluding women with gonococcal or chlamydial infection (𝑃=0.03, percentages not presented in the paper).	Strengths: PID defined histologically.; Adjustment for C. trachomatis and N. gonorrhoeae allows for independent association between M. genitalium and PID to be examined.; Limitations: No control group of women without clinically suspected PID.; Cross-sectional design does not allow for a temporal association to be proven
Simms et al. [43]	45 patients with a clinical diagnosis of PID (ages 16–43) and 37 patients undergoing tubal ligation (ages 21–45).	Case-control study	M. genitalium: PCR of endocervical swabs.; PID: clinical criteria.	M. genitalium detected in 13% of patients versus 0% of controls.	Strengths: Control group of patients without signs and symptoms of PID (although not confirmed histologically or laparoscopically.; Limitations: No upper genital tract specimens collected.; PID was not confirmed laparoscopically or histologically.; No adjustment for confounders.; Cross-sectional design does not allow for a temporal association to be proven.
Cohen et al. [41]	123 women aged 18–40 with laparoscopically confirmed PID treated at Kenyatta National Hospital, 2000–2003.	Cross-sectional study	M. genitalium: PCR of cervical, endometrial, and fallopian tube samples.; PID: laparoscopically diagnosed salpingitis, graded as mild, moderate, or severe.	M. genitalium detected in the fallopian tube of one patient.; 6% of women with mild, 11% of women with moderate, and 6% of women with severe salpingitis tested positive for M. genitalium in one or more site.	Strengths: PID verified by laparoscopy.; Limitations: No control group of women without PID.; Cross-sectional design does not allow for a temporal association to be proven.
Jurstrand et al. [44]	194 inpatients with PID aged 15–50 and 83 inpatients with ectopic pregnancy (EP) aged 18–42 treated in the OBGyn department of Örebo University Hospital, Örebo, Sweden, 1984–1986.; 246 healthy pregnant women being screened for rubella were matched to ectopic pregnancy cases by age.	Case control study	M. genitalium: antibodies assessed using a lipid-associated membrane protein-enzyme immunoassay (LAMP-EIA).; PID: clinical criteria.	M. genitalium and PID: Crude OR 1.3 (0.7–2.2).; Adjusted OR 1.0 (0.6–1.7); M. genitalium & EP: Crude OR 1.3 (0.7–2.5).; Adjusted OR 1.0 (0.5–2.0).; ^(Adjusted for age and C. trachomatis antibodies.)	Strengths: LAMP-EIA covers antigenic variation of different genotypes of M. genitalium with no cross-reactivity with other Mycoplasma species.; Limitations: PID not laparoscopically or histologically confirmed.; Limited adjustment for confounders.; Unable to determine timing of M. genitalium infection in relation to the acute PID episode.
Haggerty et al. [4]	682 women with clinically suspected PID aged 14–37 years recruited from ER, OB/Gyn, STD clinics, and private practice from 13 U.S. urban clinical sites, 1996–1999.	Prospective	M. genitalium: PCR of cervical and endometrial samples.; PID: histologically confirmed endometritis assessed at baseline and at a 30-day follow-up clinic visit.	Baseline comparison of M. genitalium (endometrium) and endometritis: Adjusted OR 3.0 (1.5–6.1).; Prospective evaluation of baseline M. genitalium and incident endometritis (30-days follow-up visit): Adjusted RR 13.4 (2.4–75.2).; ^(Adjusted for age, race, C. trachomatis, and N. gonorrhoeae.)	Strengths: Large sample size.; Histologic confirmation of PID.; Prospective analysis of baseline M. genitalium infection and incident endometritis at the 30 day follow-up visit supports a temporal association.; Adjustment for C. trachomatis and N. gonorrhoeae allows for independent association between M. genitalium and PID to be examined.; Limitations: No control group of women without clinically suspected PID
Bjartling et al. [45]	2079 women aged 15–40 presenting for termination of pregnancy at Malmö University Hospital, Sweden, 2003–2007.	Prospective	M. genitalium: PCR of urine, vaginal, and cervical samples.; PID: clinical criteria.	M. genitalium & posttermination PID: Adjusted OR 6.3 (1.6–25.2).; ^(Adjusted for age and C. trachomatis.)	Strengths: Prospective design allows for temporal inference.; Adjustment for C. trachomatis allows for independent association between M. genitalium and PID to be examined.; Limitations: Generalizability limited to post-abortal PID.; No upper genital tract specimens tested.; PID not confirmed laparoscopically or histologically.
Oakeshott et al. [46]	2378 sexually active female students (mean age 21 years) participating in a chlamydia screening trial, London, 2004–2006.	Prospective	M. genitalium: PCR of self-collected vaginal swabs.; PID: Self-reported PID and PID symptoms, and medical records/clinical diagnosis for a subset of women not completing follow-up questionnaire, assessed over 12 months.	M. genitalium & PID: RR 2.4 (0.7–7.5).	Strengths. Large sample size.; Prospective design allows a temporal relationship to be explored.; Limitations: PID diagnosis based on self-report and limited medical record/clinical diagnosis; no laparoscopic or histologic confirmation.; Asymptomatic PID not captured.; Despite large sample, the study was underpowered to detect a prospective association between M. genitalium and PID.; Selection bias may have caused an underestimate of M. genitalium.