CHLAMYDIAL AND GONOCOCCAL INFECTIONS are the 2 most common notifiable diseases in the United States and disproportionately affect young women. Women with these infections are often asymptomatic and may not seek medical care, yet untreated chlamydia and gonorrhea can lead to pelvic inflammatory disease (PID) and its sequelae of ectopic pregnancy, chronic pelvic pain, and infertility.1–3 In addition, the longer the duration of infectiousness, the greater the probability that infection will be transmitted to a sexual partner. For women attending sexually transmitted disease (STD) clinics, screening for both Chlamydia trachomatis and Neisseria gonorrhoeae is generally performed4–6; however, identification of infected women does not necessarily result in prompt treatment. Several studies7–9 have shown treatment delay in this population, and at least 2 studies7,10 have described the development of PID in up to 4% of women during the interval between testing and treatment.
Postscreening treatment of women attending STD clinics is often challenging since reliable point-of-care testing for C trachomatis or N gonorrhoeae is not currently available, and patients must obtain their results at a later date. STD clinics are often busy or understaffed, and notification of women who have positive test results may not happen promptly, despite best intentions. Finally, in some STD clinics, patients are not required to present identification, and names, addresses, and telephone numbers may be inaccurate and may lead to difficulties when trying to locate infected patients for treatment.
STD programs and clinics have traditionally addressed these challenges with 2 general approaches in order to improve treatment coverage: (1) presumptive treatment of those suspected of having infection (e.g., signs, symptoms, current sexual behavior, contact of known infected individual, etc); and (2) efficient recall of those who tested positive but were not treated (e.g., having clinicians or health department staff actively notify infected patients). In practice, both approaches are employed by all STD clinics to some degree but with varying success and with no clear standardized guidelines or recommendations.
In order to explore these issues further, we evaluated publicly funded STD clinics from 3 US cities. We analyzed data for women with laboratory-confirmed C trachomatis or N gonorrhoeae infections and qualitatively described and compared the recall methods used by each of the clinics. We hypothesized that recall methods would differ among the STD clinics and that differences in these processes might impact the rates and timeliness of treatment.
STD clinics from Washington, DC; Los Angeles; and San Diego were chosen for this evaluation; all 3 cities have a medical epidemiologist assigned from the Field Epidemiology Unit of the Division of STD Prevention at CDC and offer typical STD services. Washington, DC, has 1 STD clinic, which serves a district population of 570,000 residents11, a proportion of clients are also seen from the surrounding metropolitan areas in Virginia and Maryland. Los Angeles County has 13 STD clinics, serving a population of 9.8 million residents12; 3 clinics with high STD rates were sampled for this study. San Diego County has 1 main STD clinic and 3 part-time clinics for a population of 2.9 million residents4; only the main STD clinic was included in this assessment.
Study Population and Chart Abstraction
For each site, all women with positive laboratory tests (culture or DNA-based testing) for C trachomatis or N gonorrhoeae were identified by reviewing primary laboratory data. Medical charts and health department records were reviewed for all infected women; data were abstracted only for those women, selected sequentially, who had not been treated presumptively at the time of specimen collection.
For each record, the interval between specimen collection and treatment (“time to treatment”) was calculated; additional subintervals were calculated if dates were available in the medical record for specific events preceding treatment (e.g., date test results received by clinic, etc.). Univariate analyses were conducted to identify factors associated with treatment delay, which was defined as >14 days from the time of specimen collection. All data were entered and analyzed using Epi Info Version 6.04 days (Centers for Disease Control and Prevention, Atlanta, GA).
A minimum of 50 consecutive charts was abstracted at each site, based on initial sample size calculations, which estimated that 75% of women would be treated within 30 days (expected precision ± 12%). Time periods for record abstraction varied among the 3 sites. In Washington, DC, records were abstracted from January 2000 to October 2002, and in Los Angeles, from January to December 2002. In San Diego, clinic follow-up procedures changed in April 2001 after a state Infertility Prevention Project (IPP) evaluation documented treatment delay; therefore, records were abstracted pre- and postintervention. Subsequently, these time periods are referred to as “baseline” (January 2000 to March 2001) and “intervention” (April 2001 to December 2002).
For the qualitative assessment, clinic procedures and practices were outlined from the time of testing until either the time of treatment or time of case closure by disease intervention specialists (DISs). Procedures and methods used to recall infected patients were reviewed at each clinic. Interviews were conducted with program and clinic directors, clinicians, laboratory staff, DISs, and data managers. Data flow and data storage systems were observed.
Rates of infection, as illustrated by data from calendar year 2002, varied among the 3 cities (Table 1) and were consistent with data previously reported to CDC. Reportedly, all women were offered testing for C trachomatis and N gonorrhoeae at each site, regardless of age or symptoms. Positivity was affected by the sensitivity of tests13,14 used in each locality. Washington, DC, used a DNA probe test (Gen-Probe PACE 2; Gen-Probe Incorporated, San Diego, CA) to identify both organisms, while Los Angeles and San Diego used more sensitive nucleic acid amplification tests (Gen-Probe Aptima Combo 2 or Abbot LCx; Abbott Laboratories, Abbott Park, IL). Less frequently, gonorrhea cultures were also performed in Washington, DC, and San Diego.
At all sites, most women infected with C trachomatis or N gonorrhoeae had been treated presumptively at the initial visit (range, 58%–93%) and were not evaluated. Overall, presumptive treatment was more common for chlamydia than for gonorrhea. Washington, DC, had the highest percentage who were presumptively treated (91% for chlamydia, 93% for gonorrhea). San Diego had the lowest percentage presumptively treated for chlamydial infections (69%), and Los Angeles had the lowest percentage presumptively treated for gonococcal infections (58%). Differences in presumptive treatment practices among the 3 cities were not assessed in this evaluation.
Characteristics of Study Population
Among infected women who had not been treated presumptively, a total of 327 records were abstracted from the 3 sites: 50 from Washington, DC; 152 from Los Angeles; and 125 from San Diego. Overall, 67% of the records were for chlamydia and 33% were for gonorrhea; 58% of the women were aged 22 years or less. Race/ethnicity data varied by city. In Washington, DC, 92% of women were black; in Los Angeles, 69% of women were black and 23% were Hispanic; and in San Diego, 34% were Hispanic, 27% were white, and 22% were black. History of any STD (lifetime) was reported by 39% in Los Angeles, 41% in San Diego, and 64% in Washington, DC.
Time to Treatment and Clinic Procedures
The sites had different procedures for notifying infected women who had not been treated presumptively (Table 2) and had different time-to-treatment intervals (Table 3). Because time to treatment was similar for chlamydia and gonorrhea at each site, the data are not stratified by infection type.
Patients tested for C trachomatis or N gonorrhoeae were told to return to the STD clinic in 2 weeks to get their test results. Even though chlamydia and gonorrhea results were usually available within 1 week, patients were told to return in 2 weeks because results for syphilis and HIV tests (commonly performed for most women) took longer to process. No test results were given by telephone. When positive STD results were reported to the clinic, a laboratory coordinator assigned follow-up to a DIS, who was responsible for determining if patients were treated and, if not, bringing them back to clinic for treatment. Callbacks were not routinely made by clinicians. Only a small number of women required follow-up for treatment (approximately 1.9 women per month) because over 90% of infected women received presumptive therapy at the time of specimen collection. No logbook was kept detailing efforts made to notify patients of positive test results.
Among untreated infected women, subsequent treatment was confirmed for 82%, but only 34% were treated within 14 days. Among those treated, the median time to treatment was 18 days and was the longest among the 3 cities. The median time from specimen collection to test result was 4 days, the median time from test result to patient notification was 9 days, and the median time from patient notification to treatment was 4.5 days. Age, race/ethnicity, and history of STD were not predictors of delayed treatment >14 days.
Patient notification procedures differed among the 3 Los Angeles STD clinics we evaluated. Because STD and HIV results could not be given by telephone, patients were told to return to clinic to obtain them. Two of the clinics scheduled return appointments 1 week after the initial visit, while the third clinic scheduled return appointments at 2 weeks. Upon receipt of chlamydia or gonorrhea test results, medical charts were reviewed to determine treatment status. This process was performed by a single nurse or nurse manager at 2 of the clinics but was done by many physicians at the third clinic. For one clinic, patient notification of a positive test result was performed by physicians or nurses, but in the other clinics, notification was only performed by DISs. Two clinics maintained logbooks to track treatment compliance among infected women who had not been presumptively treated.
Among untreated infected women, subsequent treatment was confirmed for 92%, and 67% were treated within 14 days. Among those treated, the median time to treatment was 8 days, and the median time from specimen collection to test result was 3 days. Median intervals between test result, patient notification, and treatment could not be accurately calculated from available data. Age, race/ethnicity, and history of STD were not predictors of delayed treatment >14 days.
In San Diego, patients were encouraged to call a dedicated telephone line for test results, but most were also given a 2-week return appointment for HIV results, which were only given in person by an HIV counselor. A nurse practitioner in the clinic reviewed all positive laboratory results. Records of infected patients who had not been treated presumptively were forwarded to DISs for notification. No logbook was kept detailing efforts made to notify patients of positive test results.
Among untreated infected women at baseline, subsequent treatment was confirmed for 88%, and 46% were treated within 14 days. Among those treated, the median time to treatment was 14 days. The average time from specimen collection to test result was reportedly 2 to 3 days, though this interval was not formally measured. The median time from posting of the test result in the medical record to patient notification was 6 days, and the median time from patient notification to treatment was 4 days. Seventy-three percent of untreated infected women were referred to DIS for follow-up. Age, race/ethnicity, and history of STD were not associated with delayed treatment >14 days.
New clinic follow-up procedures were implemented in April 2001. Instead of forwarding medical records of untreated patients to DIS, 2 nurse practitioners assumed responsibility for notifying patients. If neither nurse practitioner reached the patient within 24 hours of receiving laboratory results, the case was forwarded to DISs for further follow-up. One of the nurse practitioners was also responsible for maintaining a logbook of untreated infected patients. Dates of specimen collection, first telephone contact, treatment, and DIS referral (if needed) were recorded in the logbook. Return appointments were scheduled 1 week after the initial visit (due to improved efficiency in HIV testing), and patients were encouraged to call a dedicated telephone line for STD test results.
Among untreated infected women at intervention, subsequent treatment was confirmed for 92%, which was similar to baseline (88%). However, the percentage of women treated within 14 days improved significantly from 46% to 82% (P < 0.05), and among those treated, the median time to treatment decreased from 14 days to 7 days (P < 0.05). The median time from posting of the test result in the medical record to patient notification decreased from 6 days to <1 day, and the median time from patient notification to treatment decreased from 4 days to 2 days. In addition, the proportion of untreated infected women referred to DISs decreased from 73% to 25%.
This evaluation showed that time to treatment was delayed at all 3 sites for women infected with C trachomatis or N gonorrhoeae. However, simple follow-up changes implemented by the San Diego STD clinic markedly decreased time to treatment to a median of 7 days, which was better than the other clinics evaluated. From previous studies, in Birmingham, AL, the median time to treatment was 11 to 13 days,8 and in Philadelphia, only 28% of women with untreated chlamydial infections received therapy within 14 days of screening.9 Both of these studies emphasized the importance of effective patient follow-up; however, our study is the first we are aware of to show the impact that simple procedural changes can have on reducing treatment delay.
Changes in recall methods made by the San Diego STD clinic focused on prompt notification of untreated infected individuals. Important modifications included requiring clinicians to notify patients of positive test results within 24 hours of laboratory receipt, and maintaining a clinic logbook to track progress and treatment compliance. Though the number of women who were treated within 30 days was similar for baseline and intervention periods (85% vs. 89%), there were large differences in the proportion of women treated within 14 days (46% vs. 82%). This improvement in timeliness was manifest primarily by decreasing 2 distinct subintervals: (1) the time between posting of the laboratory result in the medical record and patient notification; and (2) the time between patient notification and treatment. Notably, the modifications made by the San Diego STD clinic did not require any additional personnel; in fact, the number of untreated infected women referred to DIS substantially decreased, allowing DIS to pursue other priority work.
Another important observation from our study was that protocols for patient follow-up and recall varied among the STD clinics in the 3 cities. These differences highlight how practices among STD clinics and programs are largely individual and that protocols may not be routinely shared, even within 1 county (Los Angeles). We recommend that STD programs routinely measure time to treatment to determine if improvements are needed. In some clinics, these evaluations can be performed in collaboration with regional IPP evaluators, as was the case in San Diego, where an annual review resulted in a change in procedure.
Presumptive treatment was more common than we had anticipated. Our findings and recommendations are most directly applicable to patients who are screened for STD and treated based on test results. However, the patient-notification procedures we describe could also benefit presumptively treated patients by providing an efficient mechanism for notifying patients who have positive tests, delivering timely patient education, and ensuring that all their partners are referred for testing and treatment. Besides STD clinics, these notification procedures could also be implemented in family planning, adolescent, and obstetric/gynecology clinics where a more sizeable proportion of patients are screened for STD.
One limitation of our investigation was that the retrospective chart review meant data collection could not be standardized for all sites. Ideally, an analysis of treatment outcomes would outline the individual steps leading to treatment (e.g., specimen receipt by laboratory, receipt of test result by clinic, etc.), document the timeliness of each step, and then propose interventions for how the timeliness of those steps could be improved. None of the clinics we studied consistently recorded dates for all events; therefore, direct comparisons of these subintervals could not be made. Another limitation of our study was the lack of data regarding reinitiation of sex before treatment. If a large proportion of patients resume sex before receiving treatment, reducing treatment delay warrants higher priority as an STD control measure.
Based on our findings, we recommend several steps for STD programs to consider when developing or modifying patient notification procedures. (1) a staff person from the clinic should make the initial callback within 24 hours of laboratory receipt for all untreated infected patients. Having DISs make the initial callback causes delays because time is taken to process paperwork, and DISs may prioritize other tasks (e.g., syphilis case management) above chlamydia and gonorrhea notification. (2) If clinical staff are unable to reach the patient by telephone within 24 hours, the case should be forwarded to a DIS who should initiate a priority follow-up within 48 hours of laboratory receipt. (3) Clinics should maintain a logbook to document follow-up and treatment. Having 1 clinician (with a backup) responsible for the logbook assigns ownership for the task and ensures more complete follow-up. (4) Clinics should allow patients to call for test results (may require a policy change) and/or make the return appointment for results ≤1 week.
Since a major STD program objective is to decrease the duration of infectiousness of bacterial STDs, our findings underscore the continued importance for the development and implementation of a rapid point-of-care test,15,16 which would both ensure prompt treatment and substantially decrease the use of unnecessary antimicrobial therapy. Other strategies for improving timeliness of treatment, such as delivery of medications to patients by DIS (“field-delivered therapy”), have been explored in different settings17 but are still in preliminary stages of evaluation and are not currently endorsed by most STD programs. Until affordable rapid diagnostic tests are available, enhanced awareness and modification of follow-up recall procedures can provide a simple, inexpensive way for STD clinics to ensure that more infected women are treated in a timely manner.
1. Hillis SD, Owens LM, Marchbanks PA, et al. Recurrent chlamydial infections increase the risk of hospitalization for ectopic pregnancy and pelvic inflammatory disease. Am J Obstet Gynecol 1997;1:103–106.
2. Cates W Jr., Wasserheit JN. Genital chlamydial infections: epidemiology and reproductive sequelae. Am J Obstet Gynecol 1991;164:1771–1781.
3. Miettinen A, Heinonen PK, Teisala K, et al. Serologic evidence for the role of Chlamydia trachomatis
, Neisseria gonorrhoeae
, and Mycoplasma hominis
in the etiology of tubal factor infertility and ectopic pregnancy. Sex Transm Dis 1990;17:10–14.
4. Marrazzo JM, Celum CL, Hillis SD, et al. Performance and cost-effectiveness of selective screening criteria for Chlamydia trachomatis
infection in women: implications for a national chlamydia control strategy. Sex Transm Dis 1997;24:131–141.
5. Urban MA, Coury-Doniger P, Reichman RC. Results of a screening program for Chlamydia trachomatis
infection in men attending a sexually transmitted diseases clinic. Sex Transm Dis 1997;24:587–592.
6. Hook EW 3rd, Darroch JE, Landry D, et al. Sexually transmitted disease and reproductive health clinical services: categorical clinics and integrated programs. In: Holmes KK, Sparling PF, Mardh P-A, et al., eds. Sexually Transmitted Disease. 3rd
ed. New York: McGraw-Hill, 1999:1273–1279.
7. Hook EW 3rd, Spitters C, Reichart CA, et al. Use of cell culture and a rapid diagnostic assay for Chlamydia trachomatis
screening. JAMA 1994;272:867–870.
8. Schwebke JR, Sadler R, Sutton JM, et al. Positive screening tests for gonorrhea and chlamydial infection fail to lead consistently to treatment of patients attending a sexually transmitted disease clinic. Sex Transm Dis 1997;24:181–184.
9. Foglia G, Rhodes P, Goldberg M, et al. Completeness of and duration of time before treatment after screening women for Chlamydia trachomatis
infections. Sex Transm Dis 1999;26:421–425.
10. Bachmann L, Richey CM, Waites K, et al. Patterns of Chlamydia trachomatis
testing and follow-up at a university hospital medical center. Sex Transm Dis 1999;26:496–499.
11. US Census Bureau. Population estimates for cities and towns. Available at: http://eire.census.gov/popest/data/cities/tables/SUB-EST2002–01.php
. Accessed June 27, 2004.
12. US Census Bureau. California county population estimates. Available at: http://eire.census.gov/popest/data/counties/tables/CO-EST2002/CO-EST2002–01-06.php.
Accessed June 27, 2004.
13. Centers for Disease Control and Prevention. Screening tests to detect Chlamydia trachomatis
and Neisseria gonorrhoeae
infections: 2002. MMWR Morb Mortal Wkly Rep 2002;51:1–40.
14. Black CM, Marrazzo J, Johnson RE, et al. Head-to-head multicenter comparison of DNA probe and nucleic acid amplification tests for Chlamydia trachomatis
infection in women performed with an improved reference standard. J Clin Microbiol 2002;40:3757–3763.
15. Gift TL, Pate MS, Hook EW, et al. The rapid test paradox: when fewer cases detected lead to more cases treated: a decision analysis of tests for Chlamydia trachomatis
. Sex Transm Dis 1999;25:241–242.
16. Svanas G, Boisen M, Brien J, et al. Preliminary characterization of the GC OIA rapid immunoassay for female endocervical swabs and male urine samples from symptomatic and asymptomatic patients [abstract 472]. In: Program and book of abstracts, International Society of Sexually Transmitted Disease Research, 15th
Biennial Congress. Ottawa, Ontario, Canada: 2003.
17. Steiner KC, Davila V, Kent CK, et al. Field-delivered therapy increases treatment for chlamydia and gonorrhea. Am J Public Health 2003;93:882–884.