Bachmann, Laura H. MD, MPH*; Richey, Chanty M. MPH*; Waites, Ken MD†; Schwebke, Jane R. MD*; Hook, Edward W. III MD*
From the University of Alabama at Birmingham, Departments of *Internal Medicine and †Pathology, Birmingham, Alabama
The authors thank Sharron Hagy for secretarial support and Nikki Jordan for technical support.
Correspondence and reprint requests: Edward W. Hook III, MD, The University of Alabama at Birmingham, Division of Infectious Diseases, THT 229, 1900 University Boulevard, Birmingham, AL 35294‐0006.
Objective:: Although testing for Chlamydia trachomatis is encouraged and increasingly practiced at sexually transmitted disease (STD) and family planning clinics, patterns of testing and follow‐up in other settings are not well described. To begin to address these issues, we performed a chart review of patients with a positive laboratory test for C. trachomatis at a major university medical center.
Methods:: Chart review of medical records for all patients with positive laboratory tests for C. trachomatis during calendar year 1996.
Results:: Of 326 patients with positive tests, 95% were female and 5% were male. Median age was 22 for females and 25 for males. Most positive C. trachomatis test results were from the emergency room (ER)/walk‐in clinic (55%) or patients receiving obstetric/gynecologic (OB/GYN) care (31%). While most C. trachomatis tests performed were on patients who had symptoms, patterns of treatment varied between sites. Fifty‐seven percent of ER/walk‐in patients received empiric antibiotics at the initial visit versus 36% of patients under OB/GYN care. Among patients with positive screening tests seen in the ER/walk‐in clinic, 32% of patients had no treatment documented versus 14% of OB/GYN patients. Four percent of women with positive tests who did not receive therapy at the time of their initial evaluation developed pelvic inflammatory disease in the interval between testing and return to the medical center.
Conclusions:: Of the patients with positive chlamydial screening tests, the proportion not treated was similar to that found in studies performed in STD clinics.
CHLAMYDIA TRACHOMATIS GENITAL infections are the most common bacterial sexually transmitted diseases (STDs) in the United States, with approximately 4 million new cases occurring annually.1 Chlamydial infections are often asymptomatic, and irrespective of the presence or absence of symptoms, upper genital tract involvement in women may lead to complications and sequelae. Furthermore, in men and women, asymptomatic infections may serve as a reservoir for further transmission of infection. For these reasons, screening for early diagnosis and treatment of C. trachomatis is an important component of chlamydia control efforts. Screening and diagnostic testing practices vary and in some settings not all untreated patients with positive chlamydial screening tests receive timely treatment for infection. Several studies conducted in STD and family planning clinics have demonstrated that the median interval between testing and therapy for women with positive screening tests ranges from 11 to 21 days.2–4 In addition, in these studies, up to 20 to 30% of untreated patients with positive test results for C. trachomatis did not return to clinic for treatment or were unable to be notified about positive test results within 1 month of initial evaluation.2–4 To determine whether the problem of failure to notify patients of positive STD test results is generalizable beyond STD clinics, and to review chlamydia testing practices in a nonreproductive health setting, we performed a chart review of all patients who had positive chlamydia tests at the University of Alabama at Birmingham (UAB) Hospital and affiliated clinics during 1996.
The University of Alabama at Birmingham Hospital is a 812‐bed, tertiary‐care urban university hospital with multiple affiliated primary care and subspecialty clinics. The population consists of a heterogeneous ethnic and socioeconomic patient mix. In 1996, specimens for chlamydia testing within the UAB complex were sent to one of three laboratories for evaluation where they were tested using enzyme immunoabsorbance assay (EIA) or cell culture.
Using data collected from the three UAB Hospital laboratories that perform chlamydia testing, patients with positive C. trachomatis tests from January to December 1996 were identified. Efforts were then made to review all available hospital charts and clinical records. Medical records were reviewed using a standardized data collection form that recorded demographic characteristics, the site of care where patients were tested, indications for testing and, after testing, whether there was evidence that patients were notified or treated, the interval between testing and treatment, and clinical findings at the time of treatment. This study was approved by the institutional review board of the University of Alabama at Birmingham.
After removal of identifiers, data were entered into an Epi Info database (Centers for Disease Control and Prevention, Atlanta, GA) for analysis.5 To determine significant differences between groups with regards to testing and treatment strategies, odds ratios were calculated using Mantel‐Haenszel chi‐square analysis. Patients with two positive chlamydia tests within 45 days of each other were considered as a single episode (only data from the first visit episode were used for analysis), whereas patients with two positive chlamydia tests greater than 45 days apart within the year of 1996 were treated as two separate clinical events. In evaluation of treatment, antibiotics considered appropriate for chlamydia therapy included azithromycin, erythromycin, doxycycline, tetracycline, ofloxacin, and amoxicillin.6
In 1996, of 5,162 chlamydia tests performed at this hospital, 399 (7.7%) were positive. Of the 399 possible clinical episodes, medical records were available for 326 (82%) patients, consisting of 15 males and 311 females with median ages of 25 and 22 years, respectively. Most patients were black (93% of males and 84% of females). Seventy‐nine (25%) of the females with positive chlamydia tests were pregnant, of whom 39 (49%) women had documentation of prior visits for prenatal care on their charts (Table 1).
Overall, most of the positive chlamydia tests were obtained from patients seen in the emergency/walk‐in clinic (55%) followed by patients seen for obstetric or gynecologic care (31%) and in the family practice clinic (10%). Eleven (73%) men and 146 women (47%) with positive tests for C. trachomatis received appropriate treatment at the time of chlamydial testing. Of the 165 women and 4 men who were not treated initially, 57% (94 females and 2 males) were treated at a follow‐up visit, whereas for 43% (71 females and 2 males) no documentation of treatment or notification of test results could be found (Table 1).
At the time of initial evaluation, 91% of the females had at least one sign or symptom suggestive of genital tract infection recorded (i.e., complaints of discharge or dysuria, pelvic abdominal pain, vaginal bleeding, cervical discharge, cervical friability, cervical motion tenderness, or adnexal tenderness). Patients receiving obstetric or gynecologic care were somewhat less likely to have documented symptoms or physical signs on the chart at the time of initial evaluation (82%) when compared to females evaluated at other sites (emergency room [ER]/walk‐in, 95%; family practice clinic, 92%; medicine clinic, 100%; nonobstetric/gynecology inpatient services, 100%). Patient characteristics present at the time of initial evaluation were compared in terms of the likelihood that appropriate treatment was provided at the time chlamydia tests were performed. Ninety percent of infected female patients (73/81) initially treated in the ER/walk‐in clinic were documented as having a discharge at presentation compared with 74% of women (32 of 43) who were not treated initially (OR = 3.14; 95% CI = 1.04, 9.58; P = 0.02). Sixty‐five percent of the patients (57/88) treated at the time of testing had cervical motion or adnexal tenderness present on examination compared with 25% of patients (15/60) who were not treated empirically (OR = 5.52; 95% CI = 2.51, 12.27; P < 0.0001). Among patients receiving obstetric and gynecologic care who were treated initially, 86% (19/22) had a discharge and 40% (6 of 15) had cervical motion or adnexal tenderness present on examination versus 79% and 0%, respectively, of patients with positive chlamydia test results who were not treated initially.
When patterns of treatment for possible chlamydia infection were examined by site of care (Table 2), women were more likely to be treated empirically in the ER/walk‐in clinic than at any other site (OR = 2.51; 95% CI = 1.54, 4.10; P < 0.0001). Women with positive chlamydia test results seen in the emergency department or walk‐in clinic who did not receive treatment at the time of initial evaluation were also more likely to have no treatment documented on the chart when compared to women receiving obstetric and gynecologic care (OR = 10.55; 95% CI = 4.49, 25.25; P < 0.0001). Patients seen in family practice clinics, internal medicine clinics, and in obstetric or gynecology clinics were also more likely to receive treatment after the availability of test results (Table 2). While 75% of nonobstetric or gynecologic inpatients were treated at the time of their initial evaluation and during their hospital stay, treatment was usually based on chlamydia test results that were available before hospital discharge. Of 67 women documented to have been examined at follow‐up for their positive chlamydial tests, 3 (4.5%) were clinically diagnosed with pelvic inflammatory disease (PID). Two of these three women had medical record documentation that a prescription for treatment of their infections had been called into a local pharmacy when the positive chlamydial test result became available. Documentation that patients infected with chlamydia were advised to notify his/her partners was found in only 25% of charts (80/326).
Prompt detection and treatment of C. trachomatis infection is a critical element of efforts to control this disease and prevent serious sequelae. Screening has been found to be effective in reducing the prevalence of chlamydial infection and to reduce subsequent risk for pelvic inflammatory disease.7 Additional benefits of prompt treatment after a positive diagnostic assay include reduction of risk for transmission to others. For chlamydial screening and testing to be maximally effective, patients must be promptly notified of their test results, treated, and educated about the need for partner referral. In many hospitals, chlamydial testing is performed in a number of different settings, each of which may have its own approach to notification of patients of their test results. We believe that in many instances procedures for notification of patients of test results may not receive sufficient emphasis. In this study, we found that, as previously demonstrated in several STD and family planning clinics, more than one fifth of patients with a positive diagnostic test for C. trachomatis had no documented treatment for their infection in the month after testing.
Much chlamydial testing conducted at this hospital is done to evaluate genitourinary signs or symptoms (i.e., as diagnostic testing), rather than as screening of asymptomatic patients. Because of the lack of chart documentation, it is difficult to draw reliable conclusions about the role that specific physical signs play in leading to empiric chlamydial treatment. The limited data available suggest that the presence of a discharge contributes to the decision to treat females at initial presentation and that upper tract signs were present more often in women treated initially compared to those whose treatment was deferred. We were unable to evaluate whether the presence of a cervical (as opposed to vaginal) discharge or friability led more frequently to empiric therapy because few clinicians documented these specific entities. Interestingly, 34% of the pregnant women and 23% of females who were not pregnant or of unknown pregnancy status with positive chlamydial tests had vaginal bleeding. These findings are similar to those of Chan et al. who found that 31% of female patients evaluated in an emergency room with positive chlamydial cultures had vaginal bleeding as one of their chief complaints.8 Unfortunately, it is unknown exactly how many females had vaginal bleeding (or other signs and symptoms such as vaginal discharge) and did not have chlamydia.
The problem of notification difficulties appears to be particularly common for patients seen in the ER/walk‐in clinic. For example, at UAB Hospital, of the 26 pregnant women with no documentation of treatment at the time of testing, 65% (17) were evaluated initially in the ER/walk‐in clinic. Furthermore, 84% (38/45) of the nonpregnant women or women of unknown pregnancy status who had no treatment documented were evaluated initially in the ER/walk‐in clinic. Once a patient leaves the emergency room, it can be difficult to notify him or her of test results and the need for further treatment. A study by Adams et al. found that almost half of 4,741 patients who came to the emergency room and supplied telephone numbers gave phone numbers at which they could not be reached for follow‐up.9
When considered with prior studies,2–4 these data suggest that patients tested for STDs but not treated at that time should be strongly encouraged to find out and act on test results. Such efforts may not only lead to treatment of more infected patients but also, by reducing the interval between testing and treatment, reduce the approximately 4% incidence of interval PID documented in this and a similar Baltimore study.10 Optimal methods for insuring follow‐up are unknown and are an appropriate topic for future study.
A second potential approach to addressing the problem of patients who do not become aware of or act on positive test results is use of rapid, onsite, chlamydial tests, which would allow the patient to receive test results and treatment before discharge, thus avoiding the problematic issue of follow‐up. While available point‐of‐care chlamydia tests are not as sensitive or specific as laboratory‐based tests such as polymerase chain reaction (PCR) and chlamydia culture, in some settings reducing the often many patients with positive results lost to follow‐up may lead to higher numbers of infected patients being treated.11
Several potential limitations of this study should be acknowledged. The lack of follow‐up may be overestimated because of incomplete documentation of therapy on the chart. Hospital and clinic charts are separate. Thus, despite our efforts to review all available medical records, patients who were not treated appropriately in one part of the facility such as the ER may have been treated in a clinic without our knowledge or prescriptions for treatment may have been provided but not documented in the medical record. It is unclear exactly how many untreated patients were ultimately referred to the health department through legally required reporting and, therefore, treated. Finally, the frequency of signs and symptoms of infection and counseling about partner notification may have been underestimated because of lack of documentation on the chart.
In summary, a substantial proportion of patients with positive screening tests for C. trachomatis may not receive timely treatment of their infections. These data suggest that this problem occurs in settings other than STD and family planning clinics and represents a potential target for future STD intervention activities.
1. Eng TR, Butler WT. The Hidden Epidemic. Washington, DC: National Academy Press, 1997.
2. Foglia GM, Goldberg M, St. Louis ME. Assurance of and duration of treatment after screening for Chlamydia trachomatis infections. International Congress of Sexually Transmitted Diseases. Seville, Spain, 1997:732. Abstract.
3. Erickson B, Wasserheit JN, Rompalo AM, Brathwaite W, Glasser D, Hook EW III. Routine voluntary HIV screening in STD clinic clients: characterization of infected clients. Sex Transm Dis 1990;17(4):194–9.
4. Schwebke JR, Sadler R, Sutton JM, Hook EW III. 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(4):181–4.
5. Dean AG, Dean JA, Coulombier D, et al. Epi Info, Version 6: a word processing, database and statistics program for epidemiology on microcomputers. Atlanta, GA: Centers for Disease Control and Prevention, 1994.
6. Centers for Disease Control and Prevention. 1998 Guidelines for treatment of sexually transmitted diseases. MMWR Morb Mortal Wkly Rep 1998;47(No. RR-1):54–6.
7. Scholes D, Stergachis A, Heidrich FE, Andrilla H, Holmes KK, Stamm WE. Prevention of pelvic inflammatory disease by screening for cervical chlamydial infection. N Engl J Med 1996;334(21):1362–6.
8. Chan L, Snyder HS, Verdile VP. A retrospective review of positive chlamydial cultures in emergency department patients. Am J Emerg Med 1996;14(4):406–9.
9. Adams SL, Thompson DA. Inability to follow-up ED patients by telephone: there must be 50 ways to leave your number. Acad Emerg Med 1996;3(3):271–3.
10. Hook EW III, Spitters C, Reichart CA, Neumann TM, Quinn TC. Use of cell culture and a rapid diagnostic assay for Chlamydia trachomatis screening. JAMA 1994;272(11):867–70.
11. Gift TL, Pate MS, Hook EW III, Kassler WJ. The rapid test paradox: when fewer cases detected leads to more cases treated-a decision analysis of tests for C. trachomatis. Sex Transm Dis 1999;26(4):232–40.