[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
November 3, 2008

Subsequent Sexually Transmitted Infection After Outpatient Treatment of Pelvic Inflammatory Disease

Author Affiliations

Copyright 2008 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2008

Arch Pediatr Adolesc Med. 2008;162(11):1022-1025. doi:10.1001/archpedi.162.11.1022

Objective  To determine the frequency of recurrent sexually transmitted infections (STIs) and/or pelvic inflammatory disease (PID), the average time until subsequent infection following a baseline PID diagnosis, and age- and insurance-related associations with subsequent diagnoses.

Design  This study used prospective longitudinal follow-up of STI and/or PID outcome data from electronic medical records.

Setting  An urban academic hospital system.

Participants  A total of 110 adolescent girls treated for PID as outpatients in pediatric ambulatory sites.

Main Exposure  Electronic medical records used to assess subsequent PID diagnoses and/or infections with Neisseria gonorrhoeae or Chlamydia trachomatis during the study window.

Main Outcome Measures  Demographic, health care use, and STI and/or PID outcome data were examined. Incidence of an STI and/or PID was calculated as incident cases per person-months of exposure. Cox proportional hazard modeling was performed to evaluate the incidence of STI by age or insurance status.

Results  The mean (SD) age was 16.8 (1.9) years, 89% of patients were black, and 39% had laboratory results that were positive for N gonorrhoeae or C trachomatis at baseline. Thirty-four percent of patients had an additional diagnosis of an STI during the 48-month follow-up window (incidence, 3.1 per 100 person-months) and the mean (SD) time to a subsequent STI and/or PID was 377 (297) days. Of those patients, 67% (n = 18) had chlamydia, 11% had gonorrhoeae, and 44% had PID. There were no differences based on age or insurance status.

Conclusions  Adolescents treated for PID are at risk for subsequent STI and/or PID for a 48-month period. Given the need to prevent future infections in these vulnerable youths, efforts to explore the value of ongoing strategies for risk reduction after diagnosis are warranted.

Randomized controlled trials have demonstrated that outpatient treatment can be as efficacious as inpatient therapy for the treatment of mild to moderate pelvic inflammatory disease (PID).1,2 However, no translational work has demonstrated effectiveness in adolescent clinical or public health practice. Our previous work3 has demonstrated that adolescents have poor adherence to outpatient PID treatment recommendations, putting them at increased risk for acute and chronic sequelae. Furthermore, adolescents often receive PID care in acute care settings,3-7 limiting the time that can be devoted to nurturing, patient-centered, adolescent-specific care and risk-reduction counseling.

Additional behavioral research has demonstrated that adolescents diagnosed with PID are engaging in high-risk sexual behaviors.8 Without substantial behavioral change following a PID diagnosis, these girls are at significant risk for subsequent sexually transmitted infections (STIs) and PID, a major risk factor for the development of chronic pelvic pain, ectopic pregnancy, and tubal infertility.9-12 Furthermore, adolescents from communities with high STI prevalence may be more vulnerable to reinfection, not only owing to developmental factors (behavioral and biological) common to adolescence,13 but also to sex partner selection, relationship patterns,14,15 and limited ability to accurately assess risk because of the community burden of disease.16

To date, there have been no longitudinal studies exploring subsequent STIs and/or PID in adolescents treated for PID as outpatients in pediatric ambulatory settings. The aim of this study is to determine the frequency of subsequent STIs and/or PID, the average time to subsequent infection following a baseline PID diagnosis, and potential age- and insurance-related associations with subsequent diagnoses.

Design and setting

Prospective longitudinal follow-up was conducted for 48 months during which adolescents were treated for PID in an urban pediatric center located in a large STI-prevalent city in the United States.17-19 This institution has an established outpatient PID quality improvement intervention for outpatient management of PID designed to improve provider adherence to the Centers for Disease Control and Prevention Sexually Transmitted Infection Treatment Guidelines.20,21 The quality improvement intervention has been previously described3 and includes (1) a management algorithm and clinical practice guidelines initially based on the 2002 guidelines that were updated with each new version released by the Centers for Disease Control and Prevention,20,21 (2) standardized discharge instructions, (3) a full course of medication dispensed at discharge, (4) standardized follow-up procedures, and (5) hospital admission for adolescents younger than 5 years.

Clinical records of PID consisting of completed clinical pathway assessment and follow-up forms are maintained in the general pediatrics, adolescent medicine, and pediatric emergency departments for quality assurance purposes. These records were used to identify patients. International Classification of Diseases, Ninth Revision coding through an administrative data set was used to identify missed cases. Electronic medical and laboratory records (EMR) were used to verify the diagnosis, track patterns of health care use, and determine STI outcomes. A single reviewer extracted data to a scannable form for electronic database entry. To ensure that data was extracted appropriately, the principal investigator (M.T.) reviewed 10% of selected medical records. This project was approved by the Johns Hopkins institutional review board.


Extracted data included demographic information, visit data, and STI and/or PID outcome data from the visit. Tracked STI outcomes using the laboratory feature of the EMR included subsequent infections by Neisseria gonorrhoeae or Chlamydia trachomatis, as these are the 2 most common pathogens associated with PID.9,12 Nucleic acid amplification testing via cervical and urine collection are standard institutional practice. Subsequent PID outcomes were also tracked in the EMR. An extraction form was completed for each patient per visit. All patients were assigned a unique identifier at the time of extraction that was used to monitor duplicate visits.

Data analysis

Descriptive data analysis was used to assess the demographics of the sample, health care patterns of use, and frequency of STI and/or PID for the entire cohort. The primary outcome measure was STI and/or PID during the 48-month study window. Incidence of STI and/or PID was calculated as the rate of incident cases per person-months of exposure. Only patients with at least 1 follow-up visit in the EMR were used to calculate the incidence, indicating use of the site for periodic and/or routine care visits. Sensitivity analyses were performed to determine the effect of these inclusion criteria on the incidence calculations. Patients who were diagnosed with the same STI within 30 days of the previous diagnosis were not considered to have a repeat infection given the use of nucleic acid amplification testing as the primary method of diagnosis and the persistence of positive test results after infection has cleared. Patients diagnosed with another STI during this time interval were considered to have a new STI. Time was measured in days from the date of first PID diagnosis to the first STI and/or PID diagnosis during the study period. Cox proportional hazard modeling was performed to evaluate the incidence of STI by age and insurance status.


A total of 110 patients were identified for evaluation in the study. The mean (SD) age of patients at the time of initial PID diagnosis was 16.8 (1.9) years. Most patients were black (89%) and covered by health insurance. Baseline laboratory results for N gonorrhoeae and C trachomatis were available for 85% (n = 94) of patients, and diagnosis was confirmed for all patients using medical records. Thirty-nine percent of patients had laboratory test results that were positive for N gonorrhoeae or C trachomatis at baseline and 73% (n = 80) of patients had at least 1 additional follow-up visit (Table).

Selected Demographics
Selected Demographics
Recurrent infection

Of the 80 patients who had any type of follow-up visit in the EMR, 34% (n = 27) had at least 1 additional visit at which an STI and/or PID were diagnosed. Of those 27 patients, 67% (n = 18) were diagnosed with chlamydia, 11% (n = 3) with gonorrhea, and 44% (n = 12) with PID (these categories are not mutually exclusive, as a patient could have more than 1 diagnosis at the visit). Furthermore, 42% (n = 5) of those diagnosed with PID had laboratory tests positive for N gonorrhoeae or C trachomatis. While 70% (n = 19) of patients had only 1 recurrent STI, 30% had 2 or more subsequent STIs and/or PID (Figure 1).

Figure 1.
Number of subsequent sexually transmitted infection (STI) and/or pelvic inflammatory disease (PID) diagnoses in girls with at least 1 subsequent diagnosis of STI and/or PID during the study window (n = 27). Second bar denotes those who had at least 1 repeat episode of PID.

Number of subsequent sexually transmitted infection (STI) and/or pelvic inflammatory disease (PID) diagnoses in girls with at least 1 subsequent diagnosis of STI and/or PID during the study window (n = 27). Second bar denotes those who had at least 1 repeat episode of PID.

The incidence of recurrent STIs (including PID) in this cohort was 3.1 per 100 person-months. Sensitivity analyses revealed that inclusion of patients who did not seek additional visits during the 48-month study window reduced the incidence of subsequent PID to 2.0 per 100 person-months.

Time to subsequent infection and/or pid diagnosis

The mean (SD) time to a recurrence of PID and/or an STI was 377 (297) days (range, 15-1300 days) (Figure 2). There was no difference in the risk of subsequent infection based on insurance status (hazard ratio, 1.31; 95% confidence interval, 0.57-3.0; P = .52) or age (hazard ratio, 0.96; 95% confidence interval, 0.79-1.18; P = .72) in hazard models.

Figure 2.
Survival curve depicting time to recurrent diagnosis of sexually transmitted infections and/or pelvic inflammatory disease (N = 110).

Survival curve depicting time to recurrent diagnosis of sexually transmitted infections and/or pelvic inflammatory disease (N = 110).


Adolescents treated for PID in ambulatory pediatric settings are at significant risk for subsequent STI and/or PID in the 48 months following a baseline diagnosis. While earlier research from this health care system has identified major barriers to adherence for adolescents treated as outpatients,3 there is limited knowledge about the behavior, needs, and support for girls in the immediate interim after a PID diagnosis. This suggests that without behavioral intervention surrounding the baseline PID visit, the potential risk for STI and/or PID–related sequelae is significant.

The findings from this study must be considered in light of several limitations. International Classification of Diseases, Ninth Revision coding and patient medical records were used to identify patients. While this method likely captures all patients seen, there still may be patients missed using this method. Our team also relied on medical record data from a single hospital system with an established PID management protocol, which limits our ability to generalize these findings. Given the community burden of STIs in this city,17-19 the public health department has provided for alternative venues for testing and evaluation including 2 clinics devoted to management of STIs, a family planning clinic, a free-standing teen health clinic, and 8 comprehensive school-based health centers.22 Although we used the most conservative estimate for calculating incidence in those who had a least 1 follow-up for any reason, our method still likely underestimates STI-related disease experienced by this cohort of girls. However, these numbers represent the actual burden of disease observed in pediatric ambulatory sites and are noteworthy.

Sensitivity analyses indicate a lower incidence for the entire cohort (N = 110). The fact that patients excluded in the core analysis did not have a subsequent visit of any type in the system given the expectations for follow-up and rescreening for this cohort suggests that the excluded participants are not regular or periodic users of this institution for care. Use of either estimate, however, indicates that adolescents with PID in this STI-prevalent urban community are at significant risk for subsequent disease in the short-term following diagnosis.

While these findings suggest similar patterns of subsequent infection as seen in adolescents and adults diagnosed with infections by N gonorrhoeae or C trachomatis,23-25 they are somewhat more concerning given the risks to future reproductive health after PID. Given that this is the first published data with longitudinal follow-up of a cohort of girls with upper reproductive tract infections in a reality setting, it adds to the literature.

Adolescents treated for PID are at significant risk for STI and/or PID recurrence in the 48 months following initial diagnosis. Given the need to prevent future episodes of STI and PID in this population of vulnerable youth, exploring the value of behavioral intervention focused on risk reduction and designed to complement outpatient management strategies is critical.

Correspondence: Maria Trent, MD, MPH, Johns Hopkins School of Medicine, 200 N Wolfe St, Office 2064, Baltimore, MD 21287 (mtrent2@jhmi.edu).

Accepted for Publication: March 20, 2008.

Author Contributions:Study concept and design: Trent and Ellen. Acquisition of data: Trent, Chung, and Forrest. Analysis and interpretation of data: Trent and Chung. Drafting of the manuscript: Trent, Chung, and Ellen. Critical revision of the manuscript for important intellectual content: Trent and Forrest. Statistical analysis: Trent and Chung. Obtained funding: Trent. Administrative, technical, and material support: Trent and Forrest. Study supervision: Trent and Ellen.

Financial Disclosure: None reported.

Funding/Support: This study was funded by the Thomas Wilson Sanitarium Foundation for the Children of Baltimore City; the Robert Wood Johnson Generalist Faculty Scholars Program; and grant 1K01DP001128-01 from the Centers for Disease Control and Prevention.

Ness  RBSoper  DEHolley  RL  et al.  Effectiveness of inpatient and outpatient treatment strategies for women with pelvic inflammatory disease: results from the pelvic inflammatory disease evaluation and clinical health (PEACH) randomized trial.  Am J Obstet Gynecol 2002;186 (5) 929- 937PubMedGoogle ScholarCrossref
Ness  RBTrautmann  GRichter  HE  et al.  Effectiveness of treatment strategies of some women with pelvic inflammatory disease: a randomized trial.  Obstet Gynecol 2005;106 (3) 573- 580PubMedGoogle ScholarCrossref
Trent  MJudy  SLEllen  JMWalker  A Use of an institutional intervention to improve quality of care for adolescents treated in pediatric ambulatory settings for pelvic inflammatory disease.  J Adolesc Health 2006;39 (1) 50- 56PubMedGoogle ScholarCrossref
Trent  MEllen  JMWalker  A Pelvic inflammatory disease in adolescents: care delivery in pediatric ambulatory settings.  Pediatr Emerg Care 2005;21 (7) 431- 436PubMedGoogle ScholarCrossref
London  RSGoldthorn  JFBerman  JI Treatment of pelvic inflammatory disease in an urban emergency service.  Health Serv Rep 1973;88 (6) 550- 554PubMedGoogle ScholarCrossref
London  RSGoldthorn  JFBerman  JI P.I.D. treatment: a growing problem for urban EDs.  RN 1975;38 (8) OR6, OR8PubMedGoogle Scholar
Banikarim  CChacko  MR Pelvic inflammatory disease in adolescents.  Adolesc Med Clin 2004;15 (2) 273- 285, viiiPubMedGoogle ScholarCrossref
Chandra  AMartinez  GMMosher  WDAbma  JCJones  J Fertility, family planning, and reproductive health of U.S. women: data from the 2002 national survey of family growth.  Vital Health Stat 23 2005; (25) 1- 160PubMedGoogle Scholar
Washington  AESweet  RLShafer  MA Pelvic inflammatory disease and its sequelae in adolescents.  J Adolesc Health Care 1985;6 (4) 298- 310PubMedGoogle ScholarCrossref
Westrom  L Effect of pelvic inflammatory disease on fertility.  Venereology 1995;8 (4) 219- 222PubMedGoogle Scholar
Weström  LJoesoef  RReynolds  GHagdu  AThompson  SE Pelvic inflammatory disease and fertility: a cohort study of 1,844 women with laparoscopically verified disease and 657 control women with normal laparoscopic results.  Sex Transm Dis 1992;19 (4) 185- 192PubMedGoogle ScholarCrossref
Westrom  LV Chlamydia and its effect on reproduction.  J Br Fer Soc 1996;1 (1) 23- 30PubMedGoogle Scholar
Lee  VTobin  JMFoley  E Relationship of cervical ectopy to chlamydia infection in young women.  J Fam Plann Reprod Health Care 2006;32 (2) 104- 106PubMedGoogle ScholarCrossref
Andrinopoulos  KKerrigan  DEllen  JM Understanding sex partner selection from the perspective of inner-city black adolescents.  Perspect Sex Reprod Health 2006;38 (3) 132- 138PubMedGoogle ScholarCrossref
Lenoir  CDAdler  NEBorzekowski  DLTschann  JMEllen  JM What you don't know can hurt you: perceptions of sex-partner concurrency and partner-reported behavior.  J Adolesc Health 2006;38 (3) 179- 185PubMedGoogle ScholarCrossref
Hallfors  DDIritani  BJMiller  WCBauer  DJ Sexual and drug behavior patterns and HIV and STD racial disparities: the need for new directions.  Am J Public Health 2007;97 (1) 125- 132PubMedGoogle ScholarCrossref
Jennings  JMCurriero  FCCelentano  DEllen  JM Geographic identification of high gonorrhea transmission areas in Baltimore, Maryland.  Am J Epidemiol 2005;161 (1) 73- 80PubMedGoogle ScholarCrossref
Turner  CFRogers  SMMiller  HG  et al.  Untreated gonococcal and chlamydial infection in a probability sample of adults.  JAMA 2002;287 (6) 726- 733PubMedGoogle ScholarCrossref
Burstein  GRGaydos  CADiener-West  MHowell  MRZenilman  JMQuinn  TC Incident Chlamydia trachomatis infections among inner-city adolescent females.  JAMA 1998;280 (6) 521- 526PubMedGoogle ScholarCrossref
Workowski  KABerman  SMCenters for Disease Control and Prevention, Sexually transmitted diseases treatment guidelines, 2006.  MMWR Recomm Rep 2006;55 (RR-11) 1- 94PubMedGoogle Scholar
 Sexually transmitted diseases treatment guidelines 2002: Centers For Disease Control and Prevention.  MMWR Recomm Rep 2002;51 (RR-6) 1- 78PubMedGoogle Scholar
Baltimore City Health Department, School Health Services. http://www.baltimorehealth.org/schoolhealth.html. Accessed March 22, 2007
Orr  DPJohnston  KBrizendine  EKatz  BFortenberry  JD Subsequent sexually transmitted infection in urban adolescents and young adults.  Arch Pediatr Adolesc Med 2001;155 (8) 947- 953PubMedGoogle ScholarCrossref
Fortenberry  JDBrizendine  EJKatz  BPWools  KKBlythe  MJOrr  DP Subsequent sexually transmitted infections among adolescent women with genital infection due to Chlamydia trachomatis, Neisseria gonorrhoeae, or Trichomonas vaginalis.  Sex Transm Dis 1999;26 (1) 26- 32PubMedGoogle ScholarCrossref
Peterman  TATian  LHMetcalf  CA  et al.  High incidence of new sexually transmitted infections in the year following a sexually transmitted infection: a case for rescreening.  Ann Intern Med 2006;145 (8) 564- 572PubMedGoogle ScholarCrossref