Sexually transmitted infections are a major global cause of acute illness, infertility, long-term disability and death, and they continue to present significant medical, social, and economic problems worldwide.1 The World Health Organization (WHO) has estimated that 340 million new cases of syphilis, gonorrhoea, chlamydia and trichomoniasis occurred globally in 1999 in men and women aged 15–49 years.1 Effective and early treatment of sexually transmitted diseases (STD) has become a priority in HIV prevention programmes worldwide,2 as successful STD prevention can decrease HIV transmission rates.3

There is a strong consensus that control of treatable STD can be achieved by case detection, management and by behavioural change interventions. In industrialised countries,4 case management is usually based on laboratory-confirmed aetiological diagnoses. Individuals are sent away after testing and are treated after the results of tests come back. In some settings, however, laboratory tests are frequently unavailable as a result of laboratory equipment deficiencies, limited clinical training of laboratory personnel and poor financial resources.5 Consequently, syndromic criteria have provided a rational approach to ensure prompt management of STD.6 The syndromic approach, which has been promoted and implemented worldwide, has served as a major step towards improving the management of these infections in developing countries.6 7 Although studies in several countries have demonstrated that syndromic management is effective,8 the approach has yet to be evaluated in STD clinics in developed countries such as Taiwan. Current clinical practice in Taiwan combines presumptive treatment and laboratory tests. Providers do not defer treatment until after laboratory tests are available. The purpose of the present study was to assess the cost-effectiveness of syndromic management protocols for the treatment of STD in comparison with both the current methods used in Taiwanese STD clinics and the aetiological approach.

The national health insurance programme of Taiwan, which requires mandatory enrollment, was implemented in 1995. By 1999, approximately 96% of Taiwan’s population was covered by the programme.9 The national health insurance programme provides comprehensive coverage, including inpatient care, ambulatory care, laboratory tests, prescription drugs, certain non-prescription drugs, dental services, traditional Chinese medicine and certain preventive services. A co-payment is required for ambulatory care, inpatient care and obtaining pharmaceuticals. Services for catastrophic diseases, child birth and preventive healthcare and medical services are, however, offered without co-payment for low-income households and veterans residing in specific mountain areas and offshore islands.

METHODS

Study population

The study site and population was described in detail in an earlier paper.10 Briefly, during the period between July 2004 and June 2005, all male patients attending two outpatient clinics were invited to participate in this study, regardless of the reason for their visit; one group was investigated at the STD clinic of the Taipei City STD Control Center (northern Taiwan) and another group was studied at a genitourinary outpatient clinic at a Kaohsiung primary care service (southern Taiwan). Taipei and Kaohsiung are two large metropolitan areas in Taiwan. Only patients who met the following criteria were included in the study: male patients with genitourinary symptoms, aged 17 years or older, sexually active, able to provide urine and blood specimens, who reported no use of antibiotics in the preceding 15 days. After an informed consent process was completed, patients were given a self-administered questionnaire regarding limited demographic information and genitourinary symptoms (dysuria, painful urination, urethral discharge, genital ulcer, sores and skin rash).

Subsequently, a genital examination was performed by a doctor and blood samples were tested for Treponema pallidum. Each participant provided a first void urine specimen that was tested for Chlamydia trachomatis and Neisseria gonorrhoea.

Laboratory tests

The Amplicor PCR test for C trachomatis/N gonorrhoeae (Roche Diagnostic Systems, Branchburg, New Jersey, USA) was performed on all individual samples according to the manufacturer’s instructions for processing urine specimens.11 Either the rapid plasma reagin (RPR; Spinreact Reactivos, Spinreact, Spain) or Venereal Disease Research Laboratory tests were used for syphilis screening, followed by the T pallidum haemagglutination assay (TPHA; Gesellschaft fur Biochemica und Diagnostica, Wiesbaden, Germany), following the manufacturer’s instructions. Sera positive in both the RPR/Venereal Disease Research Laboratory test (at 1 : 8 dilution) and TPHA tests indicated the presence of an active syphilis infection. All tests were conducted in the national Taipei City STD laboratory. Additional specimens from 10 participants were sent to the Taiwan Center for Disease Control to confirm a subset of the laboratory tests (RPR, TPHA and PCR for gonorrhoea and chlamydia). The Taiwan Center for Disease Control findings confirmed our results.

Cost and effectiveness analyses

In this study, only the direct expenditures for medical care, including costs and charges (as proxy for cost) for outpatient services and medications, were analyzed.12 The cost of the current and aetiological approach included the amount paid for laboratory tests, physical examinations (check-ups) by doctors, drugs, materials for health education and condoms. Costs were obtained directly from the patient’s payment records. The costs of syndromic management consisted of expenses for physical examinations, drugs, materials for health education and condoms.6 Costs presented have been converted from their original figures to 2005 US dollars using the Taiwan Consumer Price Index from the year in which the original cost was incurred.

Statistical analysis

SAS software (version 8.01) was used for data analysis. Additional cost calculations were carried out using Excel spreadsheets. The sensitivity, specificity and positive predictive value (PPV) of the syndromic approach were calculated, along with corresponding values for the two other conditions of interest.

Definitions

Correctly treated men were defined as those whose conditions were correctly diagnosed and who were given appropriate drugs. Men who were free of C trachomatis, N gonorrhoeae or T pallidum infection but were mistakenly diagnosed and treated for infection were classified as overdiagnosed and overtreated. If men with C trachomatis, N gonorrhoeae or T pallidum infection were mistakenly treated for another infection with drugs not active against the actual infection, they were classified as incorrectly diagnosed and treated.13 The cost of treatment, however, was considered in the cost analysis.

Sensitivity was defined as the proportion of individuals with the disease whose assessments for the disease were positive. Specificity was defined as the proportion of individuals without the disease whose assessment yielded negative results. PPV was defined as the probability of disease in a patient with a positive assessment. Negative predictive value was defined as the probability of not having the disease when the assessment was negative. The predictive value is determined by the sensitivity and specificity of the particular assessment method and the prevalence of the disease in the population being tested. The predictive value is the most relevant parameter to clinicians when they are interpreting test results.14

RESULTS

Study population

During the period from July 2004 to June 2005, a total of 498 male patients who presented to the facilities were eligible for the study. Among these, 473 (95%) patients met all inclusion criteria and were enrolled in the study. The patients’ age range was 17–50 years (median 28.0) and most (55%) patients were under 30 years of age. A total of 335 patients (71%) had urethral symptoms (discharge or dysuria or painful urination) and 138 (29%) had genital ulcers or sores with skin rash. Sixty-six per cent (106/160) of patients with N gonorrhoeae and/or C trachomatis presented only with dysuria/painful urination.

Validity analysis

Validity analysis was performed on 473 study subjects. Of the 241 patients with urethral discharge, 136 were positive for N gonorrhoea and/or C trachomatis. Of 94 patients without urethral discharge, 24 were positive for N gonorrhoea and/or C trachomatis (fig 1). For the currently used approach, 160 men would have been correctly diagnosed, yielding 100.0% sensitivity (160/160), 40.0% specificity (70/175) and 60.4% PPV (160/265). According to the WHO syndromic algorithm, 136 men would have been correctly diagnosed, yielding 85.0% sensitivity (136/160), 40.0% specificity (70/175) and 56.4% PPV (136/241). Twenty-four patients without discharge would not have been diagnosed (table 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Outcome of patients with urethral discharge symptom algorithm.

Among the 112 patients with genital ulcers, 26 were positive for syphilis. Of the patients without genital ulcers, seven were positive for syphilis (fig 2). For the current approach, 33 syphilis patients would have been correctly diagnosed, yielding 100% sensitivity (33/33), 86.7% specificity (91/105) and 70.2% PPV (33/47). If the syndromic management approach was used, 26 syphilis patients would have been correctly diagnosed, yielding 78.8% sensitivity (26/33), 18.1% specificity (19/105) and 23.2% PPV (26/112) (table 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 Outcome of patients with genital ulcer disease algorithm.

Effectiveness analysis

With the current approach, among patients with urethral discharge, 60% (160/265) were correctly treated for gonorrhoea and/or chlamydia and 40% (105/265) without C trachomatis/N gonorrhoeae infection were overtreated. If the syndromic management approach had been used, 136 (56%) men with C trachomatis and/or N gonorrhoeae would have been correctly treated and 105 (44%) patients would have been overtreated.

Among the patients with ulcers, the current approach correctly treated 70% (33/47) of patients with syphilis; 30% (14/47) of patients without syphilis were overtreated. If the syndromic management approach had been used, 23% (26/112) of syphilis patients would have been correctly treated and 77% (86/112) of patients without syphilis would have been overtreated.

Cost analysis

With the current approach, the median cost per correct treatment for a patient with urethritis was US$31.40–36.54 (table 2). The average cost per correct treatment for urethritis was US$54.27. For the aetiological approach, the average cost per correct treatment for urethritis was US$32.83 (table 3). According to the WHO protocol,6 two drugs are to be used that are effective for both N gonorrhoeae and C trachomatis. Cefixime (400 mg in a single oral dose) is prescribed to treat gonococcal urethritis and doxycycline (100 mg by mouth twice a day for 7 days) is used to treat chlamydial urethritis. The cost of the two drugs per patient is US$0.79. The fee for a physical examination is US$0.60. The syndromic management approach also requires health education and condom provision. Therefore, US$0.79 per patient should be added to cover physical examination (US$0.60), health education materials (US$0.17) and 10 condoms (US$0.62), yielding a total of US$2.18 per patient treated. The average cost per correctly treated patient with urethritis would thus be US$3.86.

With the current approach, the median cost either per correctly treated or per overtreated patient with syphilis was US$21.58 (table 2). The average expenditure per correctly treated syphilis patient was US$30.74. For the aetiological approach, the average expenditure per correctly treated syphilis patient was US$21.58 (table 3). For the treatment of syphilis by syndromic management, the WHO recommends benzathine penicillin G (2.4 million units intramuscularly). The cost was US$1.93 plus US$0.60 for a physical examination and US$0.79 for educational materials and condoms. The total cost was US$3.32, regardless of whether the patient received correct treatment or overtreatment. The average cost per correct treatment for all patients with genital ulcers by syndromic management would be US$14.30.

DISCUSSION

This study indicates that a modified syndromic management approach would demonstrate higher cost-effectiveness in the treatment of men with STD compared with the current and aetiological management approaches in Taiwan. For the current approach, the average costs per correct treatment were US$54.27 and US$30.74 for urethritis and syphilis, respectively. For the aetiological approach, the average costs per correct treatment were US$32.83 and US$21.58 for urethritis and syphilis, respectively. When considering only patients with one symptom (either genital ulcers or urethral discharge), the average cost of syndromic treatment was US$3.86 and US$14.30 for urethritis and syphilis, respectively, which is well below the cost of both the current and the aetiological approaches.

Although Taiwan’s clinicians have been trained in aetiological diagnosis, most patients who complained of urogenital symptoms were initially treated while they awaited the results of diagnostic testing. This suggests that clinical diagnosis without laboratory determination is an important approach to STD management in resource-rich areas such as Taiwan. Therefore, current management still has high rates (30–40%) of overtreatment for STD in Taiwan.

The syndromic approach presents a number of problems and challenges. It does not address the issue of asymptomatic or subclinical STD.15 16 In addition, some infected individuals with symptoms do not seek healthcare, which increases the likelihood that they will infect their sexual partners.16 The number of overtreated patients with genital ulcers was greater with syndromic management than with the current approach in this study. Because the current approach also used RPR and TPHA to diagnosis syphilis, however, the syndromic and current approaches have almost the same ability to diagnose syphilis. Laboratory test results cannot usually be obtained immediately. As a result, patients have to return for treatment 1 or 2 days later and some patients may not return at all. From the viewpoint of STD prevention, the syndromic approach is still a valuable management approach, even for use in resource-rich areas.

Another major concern raised by the syndromic management of STD is that high costs may lead to underutilisation of services by patients.17 The cost per treatment for urethritis by the current approach was much higher than that associated with syndromic management. The expenses associated with the current approach include fees for doctor’s examination and the use of multiple pharmaceuticals, which occasionally leads to the overprescription of drugs.

The results of this study have important health implications. First, the results should allay the fears that syndromic management results in a higher level of overtreatment. Second, use of the syndromic management approach can be implemented easily by primary healthcare units without highly trained staff or laboratories.

It has been reported that syndromic management protocols provide adequate treatment for more than 90% of patients with urethral discharge syndrome.18 Moreover, the reported PPV of the syndromic approach for gonococcal and/or chlamydial urethritis ranged from 75% to 97% in developing countries.19 When analyzing the specificity, sensitivity and PPV for syndromic management of laboratory-confirmed STD in this study, the false negative rate for the urethral discharge syndrome-based syndromic management was approximately 26% (24/94) for cases of genital infection involving gonococcal, chlamydial or combined species. This strategy, however, led to false positives in 44% (105/241) of cases. Compared with syndromic management, the current approach had higher sensitivity (100%) and PPV (60.4%) for urethritis in this study. That is, the current method correctly treats more infections.

Chancroid, syphilis and herpes simplex virus 2 (HSV-2) are the main causes of genital ulcer disease (GUD).6 Chancroid has, however, become extremely rare in Taiwan.15 Although with syndromic management the sensitivity was high for GUD (78.8%), 21.2% (7/33) of active infections were missed. This lack of diagnostic efficacy often deprives asymptomatic individuals of therapeutic intervention when it is required. In addition, previous studies have reported that the HSV-2 seroprevalence in the general population and in high-risk groups ranged from 30% to 80%.20 21 The PPV was low for GUD (23.2%) in this study, and 76.8% (86/112) of patients with GUD were negative for syphilis. We did not perform the HSV test for the patients with GUD in this study. Some of the patients with HSV-2 infection would be missed in this study. In the future, a proper GUD syndromic analysis should require an evaluation of herpes infection.

Even though syndromic management for sexually transmitted infections has been a major step towards improved treatment efficiency in developing countries,7 the present study reveals that its efficacy is limited in urethral discharge and GUD cases. Incorporating risk assessment and rapid diagnostic testing can improve the effectiveness of syndromic management.

The main weakness of this study is the fact that only patients who presented with genital ulcers and urethral discharge were considered. Patients with such clinical manifestations only account for approximately half of all new STD cases treated at the study sites during the period of observation.10 In terms of the therapeutic setting, our urban STD clinic may not adequately reflect patient profiles at other clinics. Furthermore, it is possible that patients with severe STD are less likely to visit these specialist centres than are individuals with milder STD. This bias could well be corrected in future surveillance by including selected local gynaecology clinics and hospitals as sites.

In conclusion, the use of syndromic management for men with urethral discharge and ulcers is a simple, cost-effective approach in developed countries such as Taiwan.22 Overtreatment of patients is not any greater of a problem with the syndromic approach than it is with the currently used approach. Health authorities may consider implementing this approach in resource-rich areas, especially at the primary healthcare level in rural settings.