Barbara A. Porter

October 31, 1997

Critically-appraised topic: Chamydial screening to prevent pelvic inflammatory disease¹

Scenario:

An asymptomatic 22 year old female presents to clinic for a routine physical exam. She is sexually active, has had two sexual partners in the last year, and has never been treated for an STD. Will screening her for chlamydia at today’s visit decrease her risk of developing pelvic inflammatory disease in the future?

Clinical Bottom Line:

In women thought to be at increased risk for chlamydial infection, cervical screening reduced the risk of subsequent PID by 58% relative to women who were not screened.

Eighty-three women at increased risk for chlamydial infections need to be screened by cervical testing in order to prevent one case of pelvic inflammatory disease.

The Evidence:

Randomized controlled trial of 2607 women identified as being at increased risk for chlamydial infection. Women were assigned to receive screening for cervical chlamydial disease or usual care. The diagnosis of pelvic inflammatory disease within one year of onset of study was the outcome measured, and was determined by patient report, review of inpatient and outpatient diagnostic databases, and medical record review.

Randomization into groups occurred before inclusion/exclusion criteria sought.

Except for marital status the two groups were similar in baseline characteristics.

*Calculations based on rates expressed as cases of PID per number of women; paper's rates are per women-months.

Comments:

1. Results are strained by several flaws in study design:

• randomization style, i.e. because of study design, subjects randomized to screening group were more intensely recruited, which may have introduced behavior changes in this population.
• 36% of women in the screening group did not receive the intervention, which seems like a large percentage; however, data was analyzed by "intention to screen".
• subjects and health workers were not blinded.
• Were the 7 women diagnosed with chlamydia in the screened population also diagnosed with PID during the screening test, or were these subsequent diagnoses?

2. Although this study set out to evaluate the usefulness of screening for chlamydia in preventing PID, the paper did not provide a detailed evaluation of the screening tool. Is the prevalence of chlamydia known in this population? What are the sensitivity and specificity of the two screening tests? Why did they use two diagnostic tests? Is one a "gold standard"? The fact that screening required a pelvic exam makes cervical testing less attractive as a screening tool.

3. First-void urine testing with LCR assay for chlamydia may be a more appropriate screening tool², and its utility in the prevention of PID should be studied.

4. Some PID is caused by gonorrhea. Were subjects with chlamydia treated for gonorrhea as well?

5. Besides decreasing the risk of PID secondary to indolent infection, does making a diagnosis of chlamydia lead to education and behavioral changes? Is there a way to measure this?

_________________________________________

¹ Scholes, D. Stergachis A, et al. Prevention of pelvic inflammatory disease by screening for cervical chlamydial infection. New Engl J Med 1996; 334:1362-6.

²Marrazzo JM, White, CL, et al. Community-based urine screening for Chlamydia trachomatis with a ligase chain reaction assay. Ann Intern Med 1997; 127:796-803.