INFECTIONS CAUSED BY Chlamydia trachomatis (CT) impose a significant social and economic burden on society. Approximately 89 million new cases of CT infections occur each year worldwide,1 3 million of which occur in the United States alone.2 In women, CT may lead to cervicitis, pelvic inflammatory disease (PID), or other syndromes. PID may result in infertility, ectopic pregnancy, and chronic pelvic pain. CT-infected women may infect their newborns during birth, thereby increasing the risk of neonatal conjunctivitis and pneumonia. Furthermore, CT may lead to urethritis and epididymitis in men.3,4 As a result, CT screening of young persons, especially women, has been recommended by different organizations, including the Centers for Disease Control and Prevention.5 Screening programs have been shown to significantly decrease the CT prevalence in some regions of the United States and Sweden.6,7
Several studies have been conducted to evaluate the cost-effectiveness of CT screening programs.8-16 Our study was designed to overcome the limitations of those studies by accounting for screening effects on CT incidence in the population, to address the problem of reinfection due to failed partner referral, and to evaluate a longer screening time frame (> 1 year). We present calculations on the cost-effectiveness of a general practitioner (GP)-based screening program for CT on the basis of results obtained using a dynamic model for the transmission of CT in a heterosexual population.
Setting and Intervention
Currently, some regions of The Netherlands are considering the implementation of a screening program for CT infection. A corresponding pilot study was conducted in Amsterdam in 1996 and 1997.17 If implemented, a CT test on urine would be offered once a year to visitors of a GP if the visit is not related to a sexually transmitted disease (STD) and if the patient indicates heterosexual activity. The refrigerated samples are sent to a central laboratory, where a ligase chain reaction (LCR) test for CT is performed according to the manufacturer's instructions (LCX; Abbott Laboratories, Chicago, IL, U.S.A.). Visitors with a positive test result are asked to revisit the GP for treatment and partner referral. We assumed a single-dose treatment with azithromycin because it leads to higher compliance than multidose treatments with doxycycline. Associated side effects of this treatment are not serious and, therefore, were not considered.18
We estimated the cost-effectiveness of this screening program by using a dynamic model that consists of the following two submodels:
1. An epidemiologic model to estimate the program's ability to decrease the incidence and prevalence of CT infections in the population and to compute the number of treated CT infections per year.19
2. An economic model to estimate the effects of the decreasing CT incidence on the incidence of CT-related complications. Based on these data, the averted complications compared with no screening were estimated. Medical resources and loss of productivity time were linked to the averted complications and valued monetarily to render the averted costs. The investment costs of the program were computed, and the net costs were determined and expressed per major outcome averted (MOA). Symptomatic PID, chronic pelvic pain, ectopic pregnancy, infertility, and neonatal pneumonia were considered major outcomes.
Table 1 shows important probabilities that were used in the models. Baseline conditions were set as 10 years of CT screening for women and men who are sexually active, 15 to 24 years, visit a GP, and have no STD-related symptoms; and partner referral of 65% of partners of treated women and of 44% of partners of treated men. A 10-year screening period was chosen for analysis because our epidemiologic model showed that during the first 10 years of screening, the strongest decrease in CT infections is observed. The time horizon of the analysis was 42 years (Appendix).
Our study corresponds with the guidelines of the Panel on Cost Effectiveness in Health and Medicine,40 except that quality-adjusted life years were not applied. Our analysis also corresponds to the principles of the currently published guidelines for modeling studies.41
Epidemiologic Model: the Population-Based Stochastic Simulation Model
A stochastic simulation model was employed to describe the formation and separation of partnerships in a heterosexual population and the transmission of infectious disease in those partnerships (Appendix). The model describes a population of 10,000 heterosexual women and men with a sex ratio of 1:1 and a uniform age distribution of 15 to 64 years. Persons leave the population at 65 years and are replaced by 15-year-olds who enter the population. The model differentiates between sex, age (5-year age groups), sexual activity (high or low), type of partnership (steady or casual), and CT infection (asymptomatic or symptomatic). Persons with high sexual activity (core group) may have more than one casual partner in addition to their steady partner (at most one) and have a higher probability of forming casual partnerships than persons with low sexual activity. Persons not in the core group have at most one steady or casual partner at a time. The core group comprises 5% of the 15 to 34 years age group.20 The screening program was integrated into the model and the effect of different screening strategies on the yearly CT incidence or prevalence and on the number of treated CT infections was simulated. The prescreening total CT prevalence (prevalence of asymptomatic and symptomatic CT infections) in our model population in the 15 to 64 years age range is 4.1%.
Economic Model: the Decision Analysis Model
Averted complications. To calculate the complications averted by the screening program, data from the epidemiologic model of yearly CT incidence with screening was compared with CT incidence without screening. For example, the incidence of asymptomatic infections in women 15 to 19 years is 11.5% before screening and 10.4% after 1 year of screening (baseline). The difference was used to compute the number of averted complications. Figure 1 displays a flowchart of possible complications in asymptomatically infected persons and the possibility of vertical infection and associated neonatal pneumonia and conjunctivitis. The respective probabilities are presented in Table 1. CT infections as a cause of PID were considered until age 44 years, because onset of menopause commonly begins between 45 and 49 years, and study findings suggest that CT usually spreads during menses from the cervix to the endometrial cavity because of the loss of cervical mucus.31 Symptomatic CT infections (cervicitis, male urethritis) were assumed to be treated without complications, and no progression of disease was taken into account. The infectious period of CT was assumed to last on average 1 year.9Table 2 details medical resources, and Table 3 describes the loss of productivity time attributed to the CT-related complications under consideration.
Costs. All costs are calculated from a societal perspective. The value of healthcare use was determined with unit cost prices (e.g., a GP visit costs US $1646). Fees charged were only used to estimate the costs for medical procedures from hospitals (except medication and intensive neonatal care units). Cost of medication was determined using Dutch average market prices for standard prescriptions.47 Dutch gross domestic product price indices were applied to inflate or deflate all costs to 1997 Dutch guilders.48 Costs were converted to 1997 US dollars using gross domestic product purchasing-power parities.48 All future costs were discounted at 3% per year (Appendix).40
The indirect costs were estimated by applying the human capital approach. To value the productivity loss of paid work, the average labor costs by age and sex were used to approximate the net product per capita due to paid work (Appendix). The productivity loss of unpaid work (e.g., household services, care of dependents) was not considered in the baseline analysis (i.e., no valuation of unpaid work). However, in the sensitivity analysis it was taken into account and valued with the effective labor costs of employed housekeepers (high valuation of unpaid work), or the effective net income of employed housekeepers (low valuation of unpaid work).35 High valuation of unpaid work corresponds to the substitution approach, whereas low valuation of unpaid work corresponds to the minimum opportunity cost approach.
Investment costs. Investment costs consist of the screening and subsequent treatment costs, which include partner referral costs. The number of screened persons tested because of the screening program was computed based on the number of persons eligible for CT screening in the model population (Appendix).
The screening costs per tested person (US $22) were derived from the pilot study and consist of US $5 for the GP, US $2.5 for the transport of the urine specimens to the laboratory, US $7 for the LCR kit and controls, US $5 for labor and instruments, and US $2.5 for overhead. Because the screening test was offered to patients during a regular GP visit (opportunistic screening), the costs of this GP visit do not contribute to the screening costs. Instead, the GP receives US $5 for the time spent explaining the screening and for the collection of urine. According to the pilot study, persons with a positive test result were assumed to need a second GP visit and a single azithromycin prescription. The costs incurred by referred partners correlated with the different treatment schemes (Appendix).
Averted costs.Table 2 shows the direct costs for the considered CT-related conditions. Medical costs plus the daily hospital hotel costs of US $30449 were used to calculate the costs of inpatient treatment. Fees, which were provided by the Academic Hospital Groningen (The Netherlands), were used to cost inpatient and outpatient hospital services, except for the costs for in-vitro fertilization and medication. In-vitro fertilization costs were taken from a Dutch study.33 The estimated cost of an inpatient hospital day for intensive neonatal care was US $1,249.50
All outcomes are shown for the model population (10,000 women and men 15-64 years with a total prescreening CT prevalence of 4.1%). The baseline analysis includes 10 years of screening for women and men who are sexually active, 15 to 24 years, visit a GP, and have no STD-related symptoms; and partner referral of 65% of partners of treated women and of 44% of partners of treated men. The cost-effectiveness ratios (CERs) are presented as costs per MOA. Negative costs indicate savings. MOAs are shown with and without discounting in Tables 4 and 5. Only MOAs discounted at 3% per year are used for the CERs and in the body of the text. The sum of direct and indirect costs are the total costs.
The screening program (baseline) significantly reduces the CT prevalence in the model population.19 In particular, the total CT prevalence in the 15 to 44 years age classes, which is crucial for calculating CERs, decreases by 67% (from 6.3% to 2.1%). A similar effect can also be observed for CT incidence, as shown in Figure 2.
The cumulative net total costs of the CT screening program decrease the longer the CT screening program runs (Figure 3). This decrease is related to the decreasing CT incidence in the population, which is an effect of the screening program (Figure 2). Comparisons of screening with no screening show that the number of MOAs per year and the averted total costs per year significantly increase with screening duration (Figure 3: increasing slope of the line of cumulative number of MOAs and decreasing slope of the line of cumulative averted total costs). In contrast, the yearly total investment costs of the screening program change only slightly (Figure 3: approximately constant slopes of the cumulative total screening and total treatment cost lines). Because of the decreasing net total costs and the increasing number of MOAs, the CER also decreases with time (data not shown).
The overall CER of the screening program for the considered screening time frame of 10 years is −US $492 or −US $1,086 per MOA based on exclusion or inclusion of indirect costs, respectively. The screening program averts 127 major outcomes and reaches the point of saving after 5.0 or 3.3 years, excluding or including indirect costs, respectively. Therefore, the investigated screening program has to run at least 5.0 (3.3) years to render any cost savings. The point of saving is defined as the moment at which the averted costs due to the screening program equal its investment costs. Averted costs in the future were attributed to the year in which they were prevented through screening and subsequent treatment, and were accordingly discounted. This means that the point of saving is reached earlier than the break-even point. In our model population, the screening program required investments of US $175,800 (direct costs) and US $180,100 (total costs) and led to averted costs of −US $238,400 (direct costs) and −US $318,200 (total costs). The net direct costs are −US $62,600 and the net total costs are −US $138,100.
As shown in Figure 4, it is primarily the prevention of PID that produces the savings of the program; 31% of the averted total costs are directly due to symptomatic PID. Another 37% of the averted total costs are due to the prevention of PID related diseases, i.e., ectopic pregnancy, infertility, and chronic pelvic pain.
Table 4 presents the results of the univariate sensitivity analysis that was conducted to test the robustness of our model and to identify the most critical variables. The outcomes are shown as incremental net costs and MOAs compared with no screening. The baseline results are sensitive to variation in the CT prevalence of the target population and the rate of partner referral. Reducing the size of the sexually active core group by 50% results in a 56% reduction (from 4.1% to 1.8%) of the total prescreening CT prevalence in our model population, a 147% increase in the net direct costs, a 90% change in the net total costs, and a 42% decrease in the number of MOAs. Even with a low prescreening CT prevalence of 1.8%, the investigated screening program remains cost saving when total costs are considered. Increasing or decreasing the partner referral rates by 50% changes the net direct costs by up to 93%, the net total costs by up to 55%, and the number of MOAs by up to 26%. With no partner referral, no savings result; instead, costs of US $57,600 (direct costs) or US $16,600 (total costs) arise. Furthermore, the valuation of productivity loss of unpaid work, the CT test price, the probability of developing PID after CT infection, and the discount rate are also important model parameters. However, lowering the test-acceptance rate to 80% has a small effect on the net costs and the number of MOAs.
Table 5 shows the outcomes of the scenario analysis (i.e., the effects that are dependent on program design). Unlike the sensitivity analysis, the outcomes are shown as incremental net costs and MOAs compared with baseline. The inclusion of older age classes into the screening program leads to a higher number of MOAs. It also increases the savings when total costs are considered. Restricting the screening program to women renders higher savings in both cost categories (direct and total). Depending on the included age classes, it prevents fewer MOAs (screening for women 15-24 years) or more MOAs (screening for women 15-29 years or 15-34 years). The exclusion of women from the screening program has no advantages because the program is not cost saving for direct costs only (direct costs, US $411 per MOA), and the number of MOAs averted is the smallest.
The highly infectious character of CT, the lack of an appropriate vaccine, and the failure of the immune system to build up resistance to reinfection makes a complex dynamic approach necessary to determine the CER of any CT screening program. Until now, a decision analysis-based static approach has been the standard for determining the CER of CT screening programs.8-16 However, only a dynamic approach adequately accounts for the CT transmission dynamics and the impact of CT prevention measures on CT incidence at the population level during longer intervals. These are essential points because the bacteriologic eradication of CT in one person decreases the average CT infection risk in the population. Furthermore, the risk of becoming reinfected by CT-infected members of the population, especially via failed partner referral or a new CT-positive partner, can be adequately modeled only by using a dynamic approach. Finally, the static approach can roughly account only for the effects of the first year of screening and is not suitable for analyzing longer intervals, whereas the presented dynamic model can be used to determine the effects of screening programs over the long term. As shown in Figure 3, the CER in the first year of screening does not represent the CER of any long-term screening program. Evaluations based on 1 year in the investigated screening program will show that program to be far from cost saving. However, the picture changes dramatically with each additional year of screening. We conclude that cost-effectiveness of CT screening programs seems best to be determined with dynamic modeling on a population basis.
Our results indicate that the investigated CT screening program would eventually save costs for the society. However, in the early stages costs will outweigh savings, and it will take several years to reach the point of saving and longer to reach the break-even point. This period is primarily caused by the dynamic spread of CT, especially by the rapid reinfection of effectively treated persons because of failed partner referral. However, the longer the screening program runs, the more CT is eradicated in the population, resulting in a decreased risk of infection or reinfection and a greater aversion of related sequelae and costs compared with no screening. This suggests that a CT screening program should run a specific minimum period to achieve full effectiveness on a population basis (i.e., to significantly decrease the CT incidence). However, the cost-effectiveness of each successive year of screening becomes less favorable with increasing screening time because less CT infections are avoided by screening and treatment than in the preceding screening year. We did not separately evaluate the cost-effectiveness of single years of screening because this would have substantially enlarged the model's complexity. Further studies will be necessary to answer the question of which screening strategy should be chosen after reducing the CT incidence to a low level.
Because of improvements in the field of diagnostic and increasing returns to scale in large-scale programs, greater savings may be realized. In two recent studies, it was shown that pooling of urine specimens for LCR or polymerase chain reaction can be a cost-saving strategy.51,52 For screening costs of US $15, net direct savings would almost double. One of the primary determinants of success in CT screening programs is the rate of partner referral. Decreasing partner referral rates produces higher percentage decreases in net direct and net total savings in the baseline program. Therefore, effective partner referral is an essential part of cost-effective CT screening programs. Furthermore, the prescreening CT prevalence has a major impact on the CER of the investigated program. The prescreening prevalence of asymptomatic CT infections in men and women 15 to 39 years in the model population is similar to the one recently found in the 1996 and 1997 Amsterdam study.17,19 However, even with a total prescreening CT prevalence of 1.8% in the total model population, the investigated CT screening program generates savings of US $194 per MOA when considering total costs.
The scenario analysis showed that different screening strategies (inclusion of older persons or exclusion of men) may be dominant, depending on budget restrictions (the program's investment costs are positively correlated with the number of persons eligible for the program) and on whether maximization of direct or total savings or of MOAs is preferred (Table 5).
Our CERs may be conservative for several reasons. First, not all adverse effects of CT infections were taken into account. For example, the higher risk of HIV infection among CT-infected persons was not considered. Furthermore, study findings indicate that CT may be associated with postpartum endometritis, prematurity, and low birth weight.4 Moreover, genital chlamydial infection may increase the risk for chlamydial conjunctivitis in adults.53 Second, the effects of the screening program on CT prevalence or incidence at the end of the program were not included. If the screening program is cancelled after 10 years, it will take a couple of years before the CT prevalence or incidence reaches the level of prescreening.19 Third, the indirect costs of care given to CT-infected persons (e.g., from the mother to her CT-infected child) and direct non-medical costs were not considered.
There are several limitations of our study. Our model is, like other dynamic models, sensitive to changes in specific parameters (sensitivity analysis); therefore, results are profoundly impacted by the precision of the according parameter estimates. There is limited empirical knowledge about sexual behavior (e.g., size of the highly sexually active core group and its distribution over the age classes) and disease-specific parameters (e.g., probability of CT transmission per sexual contact and the fraction of asymptomatically CT-infected persons) that influence the prevalences and incidences generated by the model.19 Assumptions that the CT incidence would remain stable over time without CT screening may be not true. Public health measures aimed at other STDs can also decrease the CT incidence. Our model assumes a closed population, and no migration effects are taken into account. The model population has a uniform age and sex distribution. Emerging antibiotic resistance of CT could negatively impact the success of a screening program. However, the apparent inability of CT to acquire antibiotic resistance54 makes this problem rather theoretical. Technology and relative prices in the medical field and productivity per capita may change during the period considered in our analysis. Gains in quality of life due to the screening were not measured. The determination of these gains seems not to be relevant because our results indicate cost savings for society. Because of the model's complexity, the number of assumptions inherent in the model is high. To avoid substantially increasing the model's complexity further and to accommodate a deficit of necessary data, we did not take into account that the susceptibility for CT infection decreases with age55 and the risk for PID and its sequelae increases with each asymptomatic CT infection in women. Instead, we used conservative average estimates for the CT-related risks.
Despite these limitations, we think that our dynamic approach leads to more realistic CERs of CT screening programs than the commonly used static models because it overcomes their most important limitation: the failure to account for dynamic processes such as chlamydial transmission and the impact of screening on CT incidence in the population. As with other cost-effectiveness analyses, there may be problems directly transferring our CERs to other countries because of possible differences in CT prevalence or incidence, sexual behavior, resource use, and resource valuation. However, the presented methodology can be applied worldwide.
Discounting and Time Frame
It was assumed that PID, epididymitis, and complications in newborns occur in the year of CT infection and that chronic pelvic pain takes place within 5 years of PID onset.14 The probabilities of having an ectopic pregnancy or of being identified as infertile depend on active interest in having a child. Hence, CT-induced damage of a young woman's fallopian tubes is commonly discovered later on, and the costs have to be discounted accordingly. Diagnosis of infertility and subsequent in-vitro fertilization treatment is assumed to occur within 2 years after the first unsuccessful attempt to conceive. The average age of the youngest age group is 17.5 years (range, 15-19 years), and the average age of the oldest childbearing age group is 47.5 years (range, 44-49 years). Subtracting 17.5 from 47.5 and adding 2 years for the diagnosis of infertility and subsequent in-vitro fertilization treatment renders a time horizon of up to 32 years for healthcare use after a CT infection. With an 10-year investigation of CT screening, the time horizon of our analysis is 42 years.
This model has been described in detail elsewhere.56 The parameters of the model were based on the results of a survey about sexual behavior in a representative sample of the Dutch general population performed in 1989.20 To account for recent study findings, some minor adjustments of parameter values were made.19 The types of partnership are distinguished by their average duration (steady partnerships, 6.9 years; casual partnerships, 10 days) and frequency of sexual intercourse (steady partnerships, 0.25/day; casual partnerships, 1/day). The probability of having a partner who is in an older or younger age group was considered. Because the model is stochastic, 100 simulation runs were performed for every set of parameter values, and averages were used to compute prevalence and incidence of asymptomatic and symptomatic infections for the different age classes (Monte Carlo simulation).19 Total prescreening CT prevalence derived from the model in the 15 to 64 years age group is 4.3% for women and 3.9% for men. Prescreening prevalences of asymptomatic CT infections for women and men were 9.9% and 5.9% in the 15 to 19 years age group, 9.3% and 6.2% in the 20 to 24 years age group, 7.8% and 7.0% in the 24 to 29 years age group, 7.2% and 6.7% in the 30 to 34 years age group, and 3.7% and 4.1% in the 35 to 39 years age group, respectively.
The average Dutch labor costs by age and sex were calculated on the basis of the average Dutch gross income by age and sex57 and an analysis of labor costs in The Netherlands.58 Labor costs were multiplied with the employed labor-force participation rate by age and sex,57 yielding the yearly net product per capita due to paid work. The age-specific and sex-specific averages of time spent on unpaid work were derived from a Dutch time budget study59 for calculations of the net product per capita due to unpaid work. Because of the lack of specific Dutch data, estimates of the loss of productivity time due to CT infections and their sequelae were taken from a literature review.9,10,14,44 The most conservative published estimates were used in estimating the loss of productivity time due to PID, chronic pelvic pain, ectopic pregnancy, and epididymitis (Table 3).
Screened and Treated Persons
The eligible CT screening population equals the number of individuals in age-specific and sex-specific categories multiplied by the probabilities that they are sexually active (Table 1). Multiplying the eligible CT screening population by age and sex with the respective probabilities of visiting a GP during 1 year and the sex-specific probability of accepting the test (Table 1) yielded the number of persons per year that were tested because of the screening program. In the baseline analysis, this resulted in 510 tested women and 337 tested men for each year of the program. The calculation of the number of tested women was as follows: EQUATION
- where, N1 is the number of women 15-19 years;
- P1s is the sexual activity of women 15-19 years;
- P1v is the GP visits per year of women 15-19 years;
- N2 is the number of women 20-24 years;
- P2s is the sexual activity of women 20-24 years;
- P2v is the GP visits per year of women 20-24 years;
- P3 is the test acceptance of women.
The number of infected persons found by the screening program and subsequently treated was derived from our epidemiologic model and varies by year. Dividing this number (by age, sex, and screening year) by the respective positive predictive value of the screening test rendered the total number of persons who received a positive test result and were treated via the screening program. The positive predictive values were computed by using the sensitivity and specificity of the screening test (Table 1) and the prevalence of asymptomatic CT infections (epidemiologic model) in the screening eligible population by age, sex, and screening year. Our epidemiologic model also supplied the screening year-specific number of infected persons found via partner referral. On the basis of the different evaluation or treatment options for referred partners, the number of persons that were tested or treated via partner referral was computed per year. Referral of healthy partners was included by assuming that 32% of partners of CT-infected persons were not infected.27
Based on the outcomes of the pilot study, three evaluation or treatment alternatives for referred partners were considered: (1) test-based treatment, which includes a GP visit, an LCR test, and if necessary a single-dose prescription of azithromycin; (2) epidemiologic treatment 1, which involves an azithromycin prescription plus a GP visit; and (3) epidemiologic treatment 2, which includes only azithromycin treatment. On the basis of the pilot study results, it was assumed that of referred partners, 22% received test-based treatment, 37% received epidemiologic treatment 1, and 41% received epidemiologic treatment 2. The compliance with single-dose azithromycin was assumed to be 98% for test-based treatment, 95% for epidemiologic treatment 1, and 85% for epidemiologic treatment 2. Taking into account the compliance, test sensitivity, and azithromycin effectiveness, we calculated that 83% of the referred infected partners were cured.
1. Gerbase AC, Rowley JT, Heymann DHL, Berkley SFB, Piot P. Global prevalence and incidence estimates of selected curable STDs. Sex Transm Infect 1998; 74(suppl 1):12-16.
2. Cates W Jr. Estimates of the incidence and prevalence of sexually transmitted diseases in the United States: American Social Health Association. Sex Transm Dis 1999; 26(suppl 4):2-7.
3. Centers for Disease Control and Prevention. Recommendations for the prevention and management of Chlamydia trachomatis
infections. MMWR Morb Mortal Wkly Rep 1993; 42(RR-12):1-39.
4. Hillis S, Black C, Newhall J, Walsh C, Groseclose SL. New opportunities for chlamydia prevention: application of science to public health practice. Sex Transm Dis 1995; 22:197-202.
5. US Preventive Services Task Force. Screening for chlamydial infection, including ocular prophylaxis in newborns. In: US Preventive Services Task Force, ed. Guide to Clinical Preventive Service. 2nd edition. Alexandria, Virginia: International Medical Publishing, 1996:325-334.
6. Centers for Disease Control and Prevention. Chlamydia trachomatis
genital infections: United States, 1995. MMWR Morb Mortal Wkly Rep 1997; 46:193-198.
7. Hermann B, Egger M. Genital Chlamydia trachomatis
infections in Uppsala County, Sweden, 1985-1993: declining rates for how much longer? Sex Transm Dis 1995; 22:253-260.
8. Paavonen J, Puolakkainen M, Paukku M, Sintonen H. Cost-benefit analysis of first-void urine Chlamydia trachomatis
screening program. Obstet Gynecol 1998; 92:292-298.
9. Buhaug H, Skjeldestad FE, Halvorsen LS, Dalen A. Should asymptomatic patients be tested for Chlamydia trachomatis
in the general practice? Br J Gen Pract 1990; 40:142-145.
10. Genç M, Ruusuvaara L, Mårdh PA. An economic evaluation of screening for Chlamydia trachomatis
in adolescent males. JAMA 1993; 270:2057-2064.
11. Genç M, Mårdh P-A. A cost-effectiveness analysis of screening and treatment for Chlamydia trachomatis
infection in asymptomatic women. Ann Intern Med 1996; 124:1-7.
12. Howell MR, Quinn TC, Gaydos CA 1998. Screening for Chlamydia trachomatis
in asymptomatic women attending family planning clinics: a cost-effectiveness analysis of three strategies. Ann Intern Med 1998; 128:277-284.
13. Howell MR, Quinn TC, Brathwaite W, Gaydos CA. Screening women for Chlamydia trachomatis
in family planning clinics: the cost-effectiveness of DNA amplifications assays. Sex Transm Dis 1998; 25:108-117.
14. Marrazzo JM, Celum CL, Hillis SD, Fine D, Delisle S, Handsfield HH. Performance and cost-effectiveness of selective screening criteria for Chlamydia trachomatis
infection in Women. Sex Transm Dis 1997; 24:131-141.
15. Howell MR, Gaydos JC, McKee KT Jr, Quinn TC, Gaydos CA. Control of Chlamydia trachomatis
infections in female army recruits: cost-effective screening and treatment in training cohorts to prevent pelvic inflammatory disease. Sex Transm Dis 1999; 26:519-526.
16. Randolph AG, Washington AE. Screening for Chlamydia trachomatis
in adolescent males: a cost based decision analysis. Am J Public Health 1990; 80:545-550.
17. Van den Hoek JAR, Mulder-Folkerts DKF, Coutinho RA, Dukers NHTM, Buimer M, Van Doornum GJJ. Opportunistische screening op genitale infecties met Chlamydia trachomatis
onder de seksueel actieve bevolking in Amsterdam I: meer dan 90% deelname en bijna 5% prevalentie. Ned Tijdsch Geneesk 1999; 143:668-672.
18. Lea AP, Lamb HM. Azithromycin: a pharmaeconomic review of its use as a single-dose regimen in the treatment of uncomplicated urogential Chlamydia trachomatis
infections in women. Pharmacoeconomics 1997; 12:596-511.
19. Kretzschmar M, Welte R, Van den Hoek JAR, Postma MJ. A comparative model-based analysis of screening programs for Chlamydia trachomatis
infections. Am J Epidemiol 2000, in press.
20. Zessen G van, Sandfort TGM. Seksualiteit in Nederland: Seksueel Gedrag, Risico en Preventie van AIDS. Amsterdam: Swets & Zeitlinger, 1991.
21. Vogels T, Van Der Vliet RWF, MJ Danz. Regionale Gegevens, Jeugd en Seks. Leiden/Utrecht: NIPG/TNO/NISSO, 1991.
22. Centraal Bureau voor de Statistiek. Health Inquiries 1994. Voorburg: Centraal Bureau voor de Statistiek, 1994.
23. Buimer M, Van Doornum GJJ, Ching S, et al. Detection of Chlamydia trachomatis
and Neisseria gonorrhoeae
by ligase chain reaction-based assays with clinical specimens from various sites: implications for diagnostic testing and screening. J Clin Microbiol 1996; 34:2395-2400.
24. Van Doornum GJJ, Buimer M, Prins M, et al. Detection of Chlamydia trachomatis
infection in urine samples from men and women by ligase chain reaction. J Clin Microbiol 1995; 33:2041-2047.
25. Ridgeway GL, Muntaz G, Robinson AJ, et al. Comparison of the ligase chain reaction with cell culture for diagnosis of Chlamydia trachomatis
in women. J Clin Pathol 1996; 49:116-119.
26. Van De Laar MJW, Termorshuizen F, Van Den Hoek A. Partner referral by patients with gonorrhea and chlamydial infection. Sex Transm Dis 1997; 24:334-342.
27. Quinn TC, Gaydos C, Shepard M, et al. Epidemiologic and microbiologic correlates of Chlamydia trachomatis
infection in sexual partnerships. JAMA 1996; 276:1737-1742.
28. Stamm WE, Cole B. Asymptomatic Chlamydia trachomatis
urethritis in men. Sex Transm Dis 1986; 13:163-165.
29. Cates W Jr, Wasserheit JN. Genital chlamydial infections: epidemiology and reproductive sequelae. Am J Obstet Gynecol 1991; 164:1771-1781.
30. Schachter J, McCormack WM, Sweet RL, Thompson KM. Control of chlamydia: a role for every clinician who sees sexually active patients. J Clin Outcomes Management 1997; 4:60-73.
31. Cates W Jr, Rolfs RT Jr, Aral SO. Sexually transmitted diseases, pelvic inflammatory disease, and infertility: an epidemiologic update. Epidemiol Rev 1990; 12:199-200.
32. Centers for Disease Control and Prevention. Pelvic inflammatory disease: guidelines for prevention and management. MMWR Morb Mortal Wkly Rep 1991; 40(RR-5):1-25.
33. Severijnen AJ. Chlamydia trachomatis
als volksgezondheidsprobleem. Rotterdam: Erasmus University, 1992.
34. Weström L. Incidence, prevalence and trends of acute pelvic inflammatory disease and its consequences in industrialized countries. Am J Obstet Gynecol 1980; 138:880-891.
35. Centraal Bureau voor de Statistiek. Statistical Yearbook of the Netherlands 1997. Voorburg: Centraal Bureau voor de Statistiek, 1997.
36. Weström L, Joesoef R, Reynolds G, Hagdu A, Thompson 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; 19:185-191.
37. Schachter J, Grossman M, Sweet RL, Holt J, Jordan C, Bishop E. Prospective study of perinatal transmission of Chlamydia trachomatis.
JAMA 1986; 255:3374-3377.
38. Hammerschlag MR, Anderlea M, Semine DZ, McComb D, McCormack WM. Prospective study of maternal and infertile infection with Chlamydia trachomatis.
Pediatrics 1979; 64:142-148.
39. Washington AE, Johnson RE, Sanders LL. Chlamydia trachomatis
infections in the United States: what are they costing us? JAMA 1987; 257:2070-2072.
40. Gold MR, Siegel JE, Russel LB, Weinstein MC, eds. Cost-Effectiveness Analysis in Health and Medicine. New York: Oxford University Press, 1996.
41. Nuijten MJC, Pronk MH, Brorens MJA, et al. Reporting format for economic evaluation. Part II: focus on modelling studies. Pharmacoeconomics 1998; 14:259-268.
42. SIG Health Care Information Netherlands. Medische Registratie Gepoolde Gegevens over 1993. Utrecht: SIG, 1994.
43. Centers for Disease Control and Prevention. 1998 Guidelines for treatment of sexually transmitted diseases. MMWR Morb Mortal Wkly Rep 1998; 47(RR-1):1-119.
44. Washington AE, Arno PS, Brooks MA. The economic cost of pelvic inflammatory disease. JAMA 1986; 255:1735-1738.
45. Centers for Disease Control and Prevention. Chlamydia trachomatis
infections: United States, 1995. MMWR Morb Mortal Wkly Rep 1997; 46(9):193-198.
46. Rutten FFH, van Ineveld BM, Van Ommen R, Van Hoet BA, Huijsman R. Kostenberekening bij Gezondheidszorgonderzoek: Richtlijnen Voor de Praktijk. Utrecht: Jan van Arkel, 1993.
47. Ziekenfondsraad. Pharmacotherapheutical Compass 1998. Utrecht: Roto Smeets, 1997.
48. Organization for Economic Cooperation and Development. OECD health data for Windows, Version 1998. Paris: Organization for Economic Cooperation and Development, 1998.
49. Nederlands Ziekenhuis Instituut/Nederlandse Zorgfederatie. De intramurale Gezondheidszorg in Cijfers per 1 Januari 1996. Utrecht: Nederlands Ziekenhuis Instituut/Nederlandse Zorgfederatie, 1996.
50. Hilgerman CN, Koopmanschap MA, Rutten FFH. Kosten van intensive care in Nederland. Rotterdam: Erasmus University/iMTA, 1992.
51. Peeling RW, Toye B, Jessamine P, Gemmill I. Pooling of urine specimens for PCR testing: a cost saving strategy for Chlamydia trachomatis
control programmes. Sex Transm Infect 1998; 74:66-70.
52. Kacena KA, Quinn SB, Howell MR, Madico GE, Quinn TC, Gaydos CA. Pooling urine samples for ligase chain reaction screening for genital Chlamydia trachomatis
infection in asymptomatic women. J Clin Microbiol 1998; 36:481-485.
53. Postema EJ, Remeijer L, van der Meijden WI. Epidemiology of genital chlamydial infections in patients with chlamydial conjunctivitis: a retrospective study. Genitourin Med 1996; 72:203-205.
54. Erbelding E, Quinn TC. The impact of antimicrobial resistance on the treatment of sexually transmitted diseases. Infect Dis Clin North Am 1997; 11:889-903.
55. Stergachis A, Scholes D, Heidrich FE, Sherer DM, Homes KK, Stamm WE. Selective screening for Chlamydia trachomatis
in a primary care population of women. Am J Epidemiol 1993; 138:143-153.
56. Kretzschmar M, van Duynhoven THP, Severijnen AJ. Modeling prevention strategies for gonorrhea and chlamydia using stochastic network simulations. Am J Epidemiol 1996; 144:306-317.
57. Centraal Bureau voor de Statistiek. Statistical Yearbook of the Netherlands 1998. Voorburg: Centraal Bureau voor de Statistiek, 1998.
58. Statistisches Bundesamt, Löhne und Gehälter, Arbeitskostenerhebung, Fachserie 16, Reihe 5, Löhne, Gehälter und Arbeitskosten im Ausland 1996. Stuttgart: Metzler-Poeschel, 1997.
59. De Hart J. Tijdopnamen. Rijswijk: Sociaal en Cultureel Planbureau, 1995.