Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 330
--> Appendix D Estimates of the Economic Burden of STDs: Review of the Literature with Updates Joanna E. Siegel1 Introduction STDs impose a significant burden of morbidity and mortality in the United States. They range from diseases that, for the most part, cause temporary discomfort and inconvenience to illnesses that impair fertility, result in long-term morbidity, or shorten life. This paper summarizes information in the literature to provide estimates of the economic impact of seven sexually transmitted diseases. These are gonorrhea, chlamydial infection, syphilis, genital herpes, human papillomavirus (HPV), chancroid, and hepatitis B virus infection. This list of STDs is far from comprehensive, but it includes most of the reportable STDs, with the exception of HIV infection, and some additional STDs that potentially impose a large burden of illness. Estimates are presented for the annual costs of these illnesses for the year 1994. Economic Burden Of Illness The economic burden of an illness reflects the opportunity cost of all resources used as a result of the illness, that is, the value of these resources in their 1 Department of Maternal and Child Health, Harvard School of Public Health, Boston, MA. Acknowledgments: Thanks are due to the following individuals for their assistance with this project: James Ida (research assistant), Melinda Flock, Alan Friedlob, Mike St. Louis, Tom Eng, and Gene Washington.
OCR for page 331
--> preferred alternative use, assuming the illness had not arisen. A measure of the economic burden of STDs thus indicates the quantity of resources that would be available for other purposes if they were not used to address STDs. The actual source of funds may be the government, individuals, or other parties such as insurers. In some cases, resources may be used with no accompanying transfer of money. The costs associated with STDs and other illnesses include "direct" costs, both medical and nonmedical, for services and materials. Major categories of direct costs incurred because of illness include the costs of health care workers' services (those of physicians, nurses, technicians, and others), costs associated with hospital admissions, costs of pharmaceuticals, and costs of medical equipment. The costs of transportation, home care, special schooling, and other such resources are also direct costs. Productivity costs are also incurred by illness. These costs, often termed "indirect" costs, reflect the opportunity cost of the productive time the affected individual (and society) lose because of illness. Productivity costs include both time spent sick, when an individual is unable to engage in the activities he or she otherwise would, and premature mortality. The time an individual invests in treatment-traveling to obtain care, waiting in the clinic waiting room, filling a prescription, undergoing a procedure-could be categorized either as a direct cost or a productivity cost. Regardless of how it is categorized, it entails an opportunity cost. The economic burden of a disease can be demonstrated by calculating either its prevalent costs (the annual costs of cases prevalent during a given year) or the lifetime costs of individual cases (incident costs). This review focuses on the former measure of economic burden, often termed the "cost-of-illness." This approach seeks to capture the costs associated with the cross-section of existing cases-including current costs for cases that occurred previously-but does not reflect future costs associated with a disease. Prevalent costs provide an estimate of economic burden that is useful for comparisons in an annual context, such as for federal, state, or institutional budgets. Incident costs, in contrast, demonstrate the full impact of an illness over time on a cost-per-case basis. They include the present value of future costs associated with the cases occurring during a given year. For some illnesses, the distinction between prevalent and incident costs is of minimal importance, as the primary costs of the disease relate to the acute infection and occur during the first year of the disease. For most illnesses, however, costs occur over a longer period of time. This is clearly true for debilitating congenital illnesses, for which costs of care may extend over a lifetime. It is also true for STDs that have long-term consequences, such as pelvic inflammatory disease (PID) and infertility. For this reason, the prevalent costs of STDs cannot be assumed to reflect the full costs per case of these diseases. This effort focuses on the collection and adjustment of existing estimates. It
OCR for page 332
--> estimates prevalent costs (costs-of-illness) for the year 1994, reflecting the most current data available at the time of this research. The emphasis is on direct costs, although productivity loss estimates are reported where they exist. Why The Cost Of Illness Is Assessed Policymakers solicit and researchers undertake cost-of-illness studies to demonstrate the impact of an illness on society. As noted earlier, these studies place the burden of illness in the context of annual spending related to a specific cause. Consumers of this information then can more accurately assess the priority of a problem or the appropriateness of the level of spending. For example, the Senate Appropriations Committee recently requested that the National Institutes of Health compile a table of annual costs of illness for leading causes of mortality. The request specified that a column indicating the fiscal year research funding for each cause of death should be included in the table (NIH, 1995). Annual spending may reflect a minor or a major proportion of the total expenditures related to an illness. In the early 1980s, for example, annual costs for HIV infection represented only a very small part of the costs eventually attributable to existing infection. In addition, the annual burden of an illness gives no indication of the effectiveness of dollars invested in alleviating it. Nonetheless, the annual figure gives an indication of the importance of the problem at a point in time. As a result, federal offices and congressional committees have demonstrated interest in clarifying the methodology for these studies, in conducting them, and in assembling their results (Rice et al., 1985). Challenges To Assessing Economic Impact There are numerous obstacles to accurately accounting for resource use associated with disease and to the assignment of dollar values to these resources. Some problems are specific to assessing the cost of STDs, while other problems apply equally to the assessment of cost of both STDs and other illnesses. A basic problem in assessing the cost of STDs is the difficulty in establishing incidence and prevalence. Counts of cases are critical when estimates are based on an annual cost per case that is multiplied by the estimated total number of cases. Because the United States has no integrated and comprehensive medical data system, incidence of illnesses must be estimated. The reporting of many STDs is required, and therefore their incidence should be easier to ascertain than the incidence of nonreportable diseases. However, the government does not require the reporting of some important (particularly viral) STDs, and even reportable STDs are commonly underreported. The extent of underreporting is not well understood. Surveys completed over two decades ago demonstrated that private physicians reported only 25-50 percent of the gonorrhea cases they treated (IOM, 1985; Moran et al., 1995). Updated information on reporting is currently being
OCR for page 333
--> gathered. Meanwhile, the CDC assumes that only half of all gonorrhea cases are reported. Reported syphilis cases, in contrast, are believed to provide a good indicator of the incidence of that disease (Moran et al., 1995). Other factors complicating the determination of STD counts are errors in the communication of reported statistics, obstacles to clinical diagnosis, and error in laboratory testing (Schmid et al., 1987; Corey, 1994). In addition, some afflicted individuals never seek treatment and therefore are not counted in data systems. This is true, for example, for genital herpes, where primary infection is asymptomatic (or at least unrecognized) in more than half of cases (Johnson et al., 1989; Corey, 1994). Alternatively, patients may seek treatment long after the initial infection is acquired. Sequelae of STDs that motivate treatment, such as PID, may occur years later. The valuation of resources used in the identification and treatment of illness, which links counts of events and medical interventions to final estimates of cost, can also be problematic. In most analyses, opportunity costs are approximated by market prices, which is generally the most tractable method of estimation. According to economic theory, prices are an adequate reflection of the opportunity cost of resources in competitive markets. However, health care markets are generally acknowledged not to meet the standard criteria for competitive markets, so prices may deviate significantly from resource costs. For example, hospital costs for most patients have long been covered by health insurance, creating a situation in which the buyers and sellers of hospital services have had no direct contact. As a result, the consumers have exhibited little price sensitivity. Health care markets have become dramatically more competitive in recent years, owing principally to the buying power of managed care organizations and other large purchasers. However, it is unclear whether these changes have caused "prices"-i.e., fees and charges-to become more reflective of costs. Many of the prices used to estimate costs in the literature are administrative and are likely to be only rough surrogates for cost. It is difficult to obtain full information with which to compute administrative prices. In addition, these prices inevitably lag behind technological changes in the delivery of care. A particular problem affecting the estimation of STD-related costs is that a large proportion of patients are treated in public STD clinics. Little is known about the units of service provided and their costs in these settings because the clinics are allocated budgets and do not generate case-by-case claims records for purposes of reimbursement. A problem common to cost estimates of a range of diseases is that treatment patterns and costs vary geographically and by care setting. Studies using a variety of data sets would be required to accurately assess the distribution of costs. The estimation of the economic burden of illness suffers further from the virtual lack of information on certain categories of resource use. In general, there are few data available describing the "out-of-pocket costs" paid by patients and family members. The units of time lost in pursuing treatment of, and recovery
OCR for page 334
--> from, a particular disease often are not available. The valuation of productive time lost, once it is identified, is complicated both by practical and ethical issues. For example, the wage rates most frequently used to value time lost to illness generally do not reflect the replacement of labor by unemployed workers, and they do not apply readily to some categories of individuals affected by illness, such as the unemployed, children, homemakers, and the retired. The obstacles to obtaining adequate measures of cost are reflected in the estimates of economic impact presented in this literature review. In some cases, the estimates are carefully developed, subject to the broader constraints of the field. In other instances, estimates are rough; these are usually intended to provide policymakers and the research community with a general gauge of costs. Methods Literature on the Economic Impact of STDs This paper reviews the existing literature to identify estimates of the economic impact of STDs in the United States. The MEDLINE and Health databases produced by the National Library of Medicine were the primary means of identifying literature. MEDLINE covers the international biomedical literature, including the allied health fields, the biological and physical sciences, selected humanities, and information services as they relate to medicine and health care. The database from 1966 through December 1995, which includes some 3,600 journals and selected monographs of scientific congresses and symposia, was searched. In addition, the Health database from 1975 through December 1995 was searched. This database contains references from the Hospital Literature Index and other selected journals in addition to MEDLINE information. The following headings were used as keywords in both titles and abstracts: cervical cancer, cost and cost analysis, chlamydial, chancroid, economics, genital herpes, gonorrhea, health expenditures, herpes simplex, hepatitis B, human papillomavirus, pregnancy, sexually transmitted diseases, and syphilis. These headings were used separately and in selected combinations to search for materials to be reviewed for their relevance to the present study. Additional information was obtained from the Centers for Disease Control and Prevention (CDC), from members of the Institute of Medicine (IOM) Committee on Prevention and Control of Sexually Transmitted Diseases, and through citations in journals and textbooks. Adjustments Two primary adjustments are made to cost estimates from the literature in order to present annual costs for 1994. The first is an adjustment for inflation: All cost estimates are converted to 1994 dollars. This adjustment is made using the
OCR for page 335
--> general medical component of the Consumer Price Index (CPI), except in a few cases where another conversion factor is more specific to the cost being considered. In those cases, the specific category is identified (e.g., the hospitals component of the CPI). In most of the literature containing cost estimates, the year of costs is given in the study. When it is not, it is assumed that costs in the original literature were collected for the year prior to the year of publication. The second major adjustment to estimates in the literature is for changes in the incidence or prevalence of illness. Many of the estimates in the literature reflect incidence rates from the early 1980s, which were higher in many cases than 1994 STD incidence rates. In adjusting overall cost estimates for current trends, the assumption is made that the cost per case has remained constant. Therefore, if estimated 1994 incidence was 60 percent as high as an earlier year, the overall cost estimate was adjusted to 60 percent of the original figure. In some cases, estimates for STD treatment costs for an STD are calculated. The recommended treatment regimens are obtained from the CDC STD Treatment Guidelines (CDC, 1993). Costs for a physician visit are estimated at $61, based on a survey of physician fees charged to privately insured patients (AMA, 1994). This is the estimated charge for a new patient office visit for a general/family practice physician. New patient visits for other types of physicians can be much higher (e.g., the comparable fee for an internist is $122). However, fees for established patients are in this range for most relevant physician specialties (general/family: $44; internal medicine: $62; all physician average: $56). In addition, Medicaid fees and costs of physicians at public clinics are likely to be lower. Thus, the $61 fee seems a reasonable approximation for a majority of visits. Costs of drug treatments are obtained from the Drug Topics Red Book (1995). To estimate costs consistently for STDs, a standard set of expenditures should be utilized. Many categorizations would be reasonable. For example, Bowie (1995) divides resource use associated with STDs into the following categories: Routine STD laboratory testing for screening or diagnosis (diagnostics, specimen collection and transport, laboratory resources, reporting) Nonroutine diagnostic testing Physician costs (counseling, acute management, follow-up evaluations, complications) Other health care providers Medication-associated costs for treatment (drug costs, prescribing fees, equipment for packaging) Hospitalization Costs to affected individuals Preventive medications/barrier methods Administration (collection of statistical data, control programs, accounting)
OCR for page 336
--> Ideally, costs should be counted for all STDs in all relevant categories. In this review and in the estimates presented, this type of consistency is not attempted. For example, in some cases, an estimate of treatment costs was conducted that included costs of personnel, laboratory testing, and contact tracing. In other estimates, some of these costs were purposely excluded. In far more cases, there was simply no detailed assessment of many categories of the relevant costs in the literature. In this review, all available information is presented, and the limited number of adjustments listed above are used. Because many components of cost are missing, the totals calculated should be considered a lower bound for estimates of the economic burden of most STDs individually, and STDs as a group. Results Types of Studies There are several types of studies in the literature that bear on economic impact. The first, and most relevant for the purpose of this review, are studies evaluating the overall, or global, annual cost of an illness. In the area of STDs, there are studies on the overall cost of gonorrhea, chlamydial infection, and herpes simplex virus. A second type of study assesses costs for a component of an illness-that is, for some subset of the resource use associated with the broad category of the STD. The articles reviewed here focus either on relatively narrow subsets, such as treatment of an STD, or on larger subsets, such as the costs of congenital syphilis or PID. These studies are useful building blocks for estimates of overall economic burden. Finally, there are cost-effectiveness and cost-benefit studies, often termed "economic evaluations." These studies are of limited use in assessing economic burden, although they may contain information relevant to calculating annual costs. They are often prescriptive, assessing costs for medical technologies that are not yet used or only partially implemented. Their assessment of economic impact is almost always based on an analysis of lifetime costs of an illness. Studies containing cost estimates or data used in the estimates presented in Table D-1 and Table D-2 are reviewed below. This review seeks to be comprehensive in its discussion of overall cost-of-illness estimates. However, only cost-effectiveness analyses from which specific data have been obtained for estimates are reviewed. There are many others in the literature. Studies of Overall Costs of STDs A study on the overall costs of gonorrhea was conducted as part of the IOM report New Vaccine Development: Establishing Priorities (IOM, 1985). The study
OCR for page 337
--> sought to demonstrate the cost of vaccine-preventable gonorrhea, which required, in part, development of an estimate of the total economic burden of disease. The authors of the IOM study estimate annual cases of gonorrhea, adjusting for underreporting and misdiagnosis. They divide symptomatic cases into morbidity categories, assigning costs based on treatment profiles for each of these categories. Costs attributable to sequelae of gonorrhea (PID and ectopic pregnancy) are included, based on estimates of the annual occurrence of these conditions. Costs of infertility resulting from PID are also included; these appear to be calculated using an incident cost-of-illness approach, rather than the annual estimates otherwise used in their calculations. Incidence of epididymitis, ophthalmia neonatorum, and disseminated infection are assumed to be negligible and are not included. The IOM estimates exclude productivity costs. In addition, they exclude the costs of public health measures to prevent further spread of gonorrhea, such as contact tracing. The advantages of these estimates are related to their comprehensiveness: all major components of cost attributable to the illness, from acute infection to sequelae, are included. The disadvantages stem from the lack of detail in cost calculations and lack of data to verify treatment protocols and cost estimates used in the calculations. The IOM estimate is clearly intended to provide a "ballpark" figure rather than a precise estimate. In the same report and using the same general procedures, the IOM estimates costs for herpes simplex viruses (HSV) types 1 and 2. The same basic strengths and weaknesses apply. Due to a striking lack of information on such components of cost as the number of annual recurrent cases of herpes and the frequency and extent of treatment of these cases, important aspects of these costs estimates are based on relatively crude estimates. Similarly, the judgment of experts is used to categorize cases in terms of severity (and likely treatment) and treatment protocols. However, the estimates are ambitious, broad, and comprehensive-and often cited in the literature (Cates, 1988; CDC, DSTD/HIVP, 1995). It should be noted in using the IOM's unadjusted estimate that it is for both HSV-1 and HSV-2. Costs of herpes in the IOM document are estimated separately for primary and recurrent cases of genital and labial herpes as well as for herpes keratitis. A substantial proportion (40 percent) of the total costs calculated for herpes are associated with herpes keratitis, reflecting costs of corneal transplants, blindness, and treatment. Costs for other sequelae reflect encephalitis, immunocompromised and disseminated HSV, and long-term costs associated with central nervous system (CNS) impairment. Neonatal herpes and its prevention are also included, although costs associated with HSV cultures during pregnancy are now outdated as a result of changes in obstetrical practice. Annual costs are cited, with the exception of those associated with CNS impairment, which are available in the report. However, the cost estimates for CNS impairment use total costs per case rather use costs on an annualized basis.
OCR for page 338
--> Washington and others (1987) provide an overall cost estimate for chlamydial infection. Their estimates of chlamydial incidence are based on gonorrhea incidence, using ratios of diagnosed gonorrhea to chlamydial infection among men and women at sentinel sites and in published reports. They obtained estimates of epididymitis from national data sets, and from this information estimated a proportion of physician visits and hospital admissions attributable to chlamydial epididymitis. Similarly, costs of uncomplicated infection among women and of PID are estimated nationally and a proportion assigned to chlamydial. Incidence of conjunctivitis among infants is based on a CDC estimate of births to chlamydial-infected women and an attack rate estimated from published sources (A. Eugene Washington, University of California, San Francisco, personal communication, 1996). Cost data are based on national average physician fees, a survey of bills from two hospitals (for epididymitis), and published estimates. The authors base their estimates for conjunctivitis on assumptions about hospitalization rates, Diagnosis-Related Group rates, and physician fees. Indirect costs are based on assumptions about work loss and loss of homemaking services resulting from illness. Indirect costs for infants are included as the cost of lost work for parents. This is an unusual interpretation of the indirect cost for infants, which ordinarily seeks to capture the value of the infant's time; the parent's care-taking time would be included additionally. The authors note exclusion of a number of cost sources, such as adverse pregnancy outcomes and some sequelae of infection. However, probably the most important sources of uncertainty relate to disease incidence. Cost data, because they reflect only private practice physicians and exclude clinics, may also pose a source of error. In addition, PID costs comprise a substantial proportion of the costs attributed to chlamydial. These estimates are taken from the literature (Washington et al., 1986); their limitations are discussed above. Component Studies De Lissovoy and others (1995) constructed a model to estimate direct medical costs of congenital syphilis cases during 1990. The study generated a case-severity distribution for true and presumptive cases and estimated costs for five treatment protocols, assuming treatment was based on severity. The cost of hospital care was obtained from Maryland hospital per diems; Resource-Based Relative Value Scale rates (using a Maryland conversion factor) were used for outpatient visits and procedures. The estimate included direct, first-year medical costs only. It accounted for neither costs of stillbirths resulting from congenital syphilis nor annual costs associated with long-term disability after the first year of life. The authors tested their model for sensitivity to the assumptions contained in the model, reporting that the cost estimates were significantly sensitive to the assumed severity of cases.
OCR for page 339
--> Hibbs and Gunn (1991) examined an on-site syphilis control intervention conducted during a cocaine-related syphilis outbreak in Chester, Pennsylvania. The program involved screening and treatment, targeting cocaine users and sex workers. The study reports costs and cases identified by the program, as compared to routine syphilis identification and treatment activities at public STD clinics. Its purpose was to assess the viability of this type of targeted intervention, and it does not address issues of generalizability to other populations, situations, or intervention strategies. Washington and others (1986), Washington and Katz (1991), and Curran (1980) have conducted studies of costs of pelvic inflammatory disease. The Curran study is not reviewed here, because it is based on data from the early 1970s, now quite dated. However, estimates from this study appear in Table D-1. The other two studies, Washington and colleagues (1986) and Washington and Katz (1991), used similar methodologies. Incidence and cost data were updated in the 1991 study. Both studies calculate indirect as well as direct costs. Costs were estimated separately for hospitalized cases of PID, outpatient visits, ectopic pregnancies, and infertility. Data on hospitalized cases of PID and ectopic pregnancy are from the Hospital Discharge Survey for both studies. Data for outpatient visits were obtained from the National Ambulatory Medical Care Survey for the earlier study and from the National Disease Therapeutic Index for the later study. Both studies used the National Health Survey of Physician Visits to estimate costs of clinic and emergency room visits. Data were collected to estimate costs of PID hospitalization for both studies. The later study is more detailed, including a longer time period and statewide discharge data as well as individual hospital data. Costs for outpatient physician visits are more approximate. The number of physician visits per case (2.5) does not appear to be based on data. Outpatient expenses included an estimate of lab tests and medications in addition to physician charges, but these estimates were derived from national average private practice data on physician costs and, in the case of the 1991 study, from information on a specific hospital physician group. The alternative to these approximations would have been to survey these costs, as no obviously better sources of data were available. Indirect cost calculations were based on assumptions regarding work loss and assumptions regarding the distribution of occupation for affected women. There are few data available to perform this type of calculation. These studies confronted problems of inadequate data regarding outpatient services, including those for treatment of PID, ectopic pregnancy, and infertility. The extent of bias introduced by the assumptions used is not clear. For example, the ratio of initial physician visits to office-based practices versus clinics or emergency departments is 2.27 to 1 in the 1986 study, while in the 1991 study this ratio was assumed to be the reverse (1:2 physician offices to clinics and emergency departments). As no explanation is given for this difference, the data
OCR for page 340
--> TABLE D-1 Estimates of Overall Cost of Illness in the Literature (millions of dollars) STDa Original Estimate (direct costs) Original Estimate (indirect costs) Original Year Source Estimate in 1994 Dollars Gonorrheab $937 1984 IOM, 1985 $1,980 Chlamydial infectionb $727 $687 1985 Washington et al., 1987 $1,350 (direct); $1,280 (indirect) PIDc $2,730 $1,510 1990 Washington and Katz, 1991 $3,540 (direct); $1,950 (indirect) PIDc $699 $557 1980 Curran, 1980 $2,230 (direct); $1,780 (indirect) Congenital syphilis $12.4 (first-year costs only) 1990 de Lissovoy et al., 1995 $16.1 Herpes simplexd (genital) $84.5 1984 IOM, 1985 $178 NOTE: Cost estimates are reported using 3 significant digits. a There are no published overall estimates of costs for syphilis, human papillomavirus, chrancroid, and hepatitis B virus infection. b The estimates of costs for chlamydial infection and gonorrhea contain costs for PID. c The estimates of costs for PID include cases resulting from chlamydial infection or gonorrhea. d The IOM estimate for all types and sequelae of herpes simplex virus is $452.20 (1984$) or $954.56 (1994$). SOURCES: Curran JW. Economic consequences of pelvic inflammatory disease in the United States. Am J Obstet Gynecol 1980;138:848-51. de Lissovoy G, Zenilman J, Nelson KE, Ahmed F, Celentano DD. The cost of a preventable disease: estimated U.S. national medical expenditures for congenital syphilis, 1990. Public Health Rep 1995;110:403-9. IOM (Institute of Medicine). New vaccine development: establishing priorities; vol. I, Diseases of importance in the United States. Washington, D.C.: National Academy Press, 1985. Washington AE, Johnson RE, Sanders LL. Chlamydial trachomatis infections in the United States: what are they costing us? JAMA 1987;257:2070-2. Washington AE, Katz P. Cost of and payment source for pelvic inflammatory disease: trends and projections 1983 through 2000 [see comments]. JAMA 1991;266:2565-9.
OCR for page 346
--> Washington, 1993) (Table D-4). Assuming that 20 percent of women with PID become infertile each year and that 25 percent of those seek treatment (Washington and Katz, 1991), the incidence of infertility would have been an estimated 232,116 cases in 1993, with 928,464 seeking treatment. Adjusting the 1991 estimates for inflation and for current incidence, the 1994 estimate of the total cost of PID is $3,118.79 million (Table D-4). Chlamydial Infection Reporting of Chlamydial trachomatis is not consistently mandated across the United States. The 451,752 cases reported to the CDC in 1994 therefore understate the incidence of chlamydial infection (CDC, DSTD, 1995). In an article estimating the impact of chlamydial infection, Washington and colleagues (1987) used case ratios from sentinel sites to estimate the incidence of chlamydial infection relative to gonorrhea. If those 1983-1985 rates applied today (1.4 chlamydial cases for each case of gonorrhea for men, and 2.6 for women), the annual incidence would be closer to 1,765,539 adult cases (Table D-5). This annual incidence is about one-half of the 3,570,000 adult cases estimated for 1985 (Washington et al., 1987), which was the basis for an overall cost estimate for chlamydial of $727 million in direct costs (1985$). This estimate includes costs for an estimated 73,800 cases of conjunctivitis and 37,100 cases of TABLE D-5 Calculation of 1,765,539 Cases of Chlamydial Infection Using Estimates of Gonorrhea Incidence 1. Estimate of total gonorrhea CDC total reported cases of gonorrhea: 450,221 (Adds 1993 state estimate for Georgia (31,483) to total (418,738). Total estimated cases of gonorrhea accounting for underreporting: 900,442 2. Estimate of ratio of chlamydial infections in men and women Ratio of men to women cases of gonorrhea: 1.14a Ratio of 1.14 applied to total gonorrhea cases gives: 479,675 cases of gonorrhea among men 420,767 cases of gonorrhea among women Using chlamydial/gonorrhea case ratiob Men: 479,675 × 1.4 = 671,545 cases of chlamydial Women: 420,767 × 2.6 = 1,093,994 cases of chlamydial 3. Estimate of total chlamydial cases: 1,765,539 a CDC/DSTDP (Division of STD Prevention). Sexually transmitted disease surveillance 1994. U.S. Department of Health and Human Services, Public Health Service. Atlanta: Centers for Disease Control and Prevention, September 1995. Tables 12, 13:1994 men (222,718); 1994 women (195,576). b Washington AE, Johnson RE, Sanders LL. Chlamydial trachomatis infections in the United States: what are they costing us? JAMA 1987;257:2070-2.
OCR for page 347
--> pneumonia among infants. Without accounting for changes in costs associated with treatment of chlamydial infection and its complications, but accounting for inflation, the change in incidence would imply that 1994 costs should have been in the range of $668.39 million (1994$). A recent CDC estimate of chlamydial incidence, however, is much higher-4,000,000 cases (CDC, DSTD/HIVP, 1995)-implying 1994 costs of $1,513.87 million (1994$). Syphilis The recent incidence of primary and secondary syphilis peaked in 1990 and has declined consistently since then. In 1994, the CDC reported a total of 83,751 cases of syphilis (all stages). This included 20,947 cases of primary and secondary syphilis; 32,970 cases of early latent syphilis; and 27,597 cases of late and late latent cases. An estimate of incidence from a CDC report is higher: 101,000 cases (CDC, DSTD/HIVP, 1995). No reports of the overall cost of syphilis are available in the literature. Treatment. Hibbs and Gunn (1991) estimated costs for the identification and treatment of patients at STD clinics at $469 per case treated in 1989 dollars ($663 per case in 1994 dollars). This estimate, from an area of high incidence (Chester, Pennsylvania), includes investigator hours, practitioner wages, and diagnostic and treatment costs. If applied to the 1994 incident cases of primary, secondary, and early latent syphilis (53,917 cases), this estimate would amount to some $35.75 million (1994$) in treatment costs. The implied treatment costs would be $66.96 million (1994$) using the higher CDC incidence estimate. Congenital Syphilis. Mirroring the incidence of primary and secondary syphilis, the incidence of congenital syphilis rose during the late 1980s, peaked in 1991, and has since declined. Some 2,224 U.S. cases were reported in 1994 (CDC, DSTDP, 1995). In a model of congenital syphilis costs, de Lissovoy and others (1995) assigned a treatment protocol to patients as a function of the estimated level of the severity of illness. According to their estimates, some 75 percent of patients required hospitalization for treatment, for a median charge of $3,171 (1990$). They estimate direct, first-year medical costs (charges) of $12.4 million in 1990 dollars, based on 3,484 reported cases and an assumed 916 (20 percent) unreported cases. Using the estimates of de Lissovoy and others, and adjusting for current incidence, the current estimated cost of congenital syphilis is $7.8 million in 1990 dollars or $10.1 million in 1994 dollars. This assumes, as did de Lissovoy and his colleagues (1995), that 20 percent of current cases go unreported. Using a higher CDC estimate of 3,400 cases (CDC, DSTD/HIVP, 1995), this figure is $12.4
OCR for page 348
--> million (1994$). These figures do not reflect nonmedical costs and do not reflect annual costs for infants who were infected in earlier years. Genital Herpes Although nationwide statistics for the incidence and prevalence of genital herpes are not available, studies from the early and mid-1980s indicated that an estimated 200,000 to 500,000 primary episodes of genital herpes occur each year (USPSTF, 1996). The IOM (1985) estimate was higher: 724,000 new cases annually. As a result of high annual incidence, HSV prevalence has continued to increase (Corey, 1994). A national study found that the prevalence of HSV-2 (the predominant causative agent for genital herpes) was 16.4 percent among U.S. adults in the late 1970s (Johnson et al., 1989); a decade later, prevalence is estimated to be 23 percent (Corey, 1994). The national survey cited above estimated the number of cases of HSV-2 to be 25.4 million in the late 1970s (Johnson et al., 1989), and the number of all genital herpes cases has been estimated to be 26 million to 31 million (Johnson et al., 1989; CDC, DSTD/HIVP, 1995). Some concurrent estimates are as high as 35 million (Fish, 1992). Recent counts of initial physician visits demonstrate a general upward trend in incidence since the 1970s. In 1993, estimates based on data from the National Disease and Therapeutic Index (NDTI) indicated that there were 171,565 initial visits to private physicians' offices for genital herpes (IMS America, 1993; CDC, DSTDP, 1995). This estimate does not reflect visits to STD clinics and other types of practices. The incidence of herpes simplex is clearly much higher than this figure, because the infection is often asymptomatic; even when symptoms are present, patients may not seek medical treatment. The IOM estimates (IOM, 1985) of the costs of herpes simplex include $84.5 million (1984$) attributable to genital herpes. This figure, adjusted for inflation, is $178.3 million in 1994 dollars. The calculation incorporates costs in the following categories: primary genital herpes, recurrent genital herpes, neonatal HSV, costs associated with culturing during pregnancy, and cesarean section (Table D-6). It does not include costs for keratitis or encephalitis, some cases of which may be due to HSV-2. This IOM cost calculation is based on the estimated 724,000 new infections, plus an estimated 4,826,667 annual recurrences. A simple (proportional) adjustment for incidence is not attempted here. Induced Cesarean Sections. An important cost generated by genital herpes is that of cesarean deliveries performed to reduce the risk of viral transmission from the mother to the newborn. Studies have questioned the necessity and wisdom of the aggressive approach (Chang and O'Keefe, 1977; Randolph et al., 1993); however, cesarean delivery in the presence of herpetic lesions during labor is currently standard practice (USPSTF, 1996). In their study of the cost-effectiveness of cesarean section to prevent neonatal
OCR for page 349
--> TABLE D-6 Costs of Genital Herpes (dollars) Category Costs (1984$) Primary genital Severity A 5,680,000 Severity B 5,287,000 Recurrent genital 25,581,000 Neonatal HSV 7,156,000 Neonatal HSV Mild CNS impairment 702,000 Serious CNS impairment 1,495,000 Very severe CNS impairment 15,982,000 Cultures during pregnancy 5,391,000 Cesarean sections 17,250,000 Total 84,524,000 (1984$) (178.3 million [1994$]) NOTE: This estimate does not include costs of encephalitis or herpes keratitis. Some cases of these complications are due to genital herpes and should be included. It includes present value of future costs attributable to central nervous system (CNS) impairment. SOURCE: IOM. New vaccine development: establishing priorities; vol. I, Diseases of importance in the United States. Washington, D.C.: National Academy Press, 1985. herpes, Randolph and colleagues (1993) estimated the cost of excess cesarean sections in hypothetical cohorts of pregnant women with and without a history of HSV infection. Costs for cesarean among women with recurrent HSV were $22.3 million (1992$) ($24.7 million in 1994$) per million pregnant women, and costs for women with a negative history were $341,000 (1992$) ($378,000 in 1994$) per million pregnant women. These calculations are based on the assumption that the incremental cost of a cesarean section instead of a vaginal delivery is $3,725 in 1992 dollars ($4,135 in 1994$). The estimate of 20 percent prevalence of HSV-2 infection among pregnant women (Randolph et al., 1993) implies that current costs for excess cesareans are about $20.9 million (1994$) annually. The IOM (1985) estimated the total costs of cesareans at a higher value of $17.3 million (1984$) or $36.4 million in 1994 dollars. This total cost estimate would be somewhat lower if Randolph and others' (1993) estimate of the incremental cost of a cesarean were used: $31.3 million (1994$). However, the IOM estimate is still higher than the $20.9 million figure, despite a lower assumed prevalence of HSV-2 of 5 percent. This difference may reflect the number of cesarean sections attributed to herpes infection. Under the current policy, it is no longer recommended to perform a cesarean section based on a series of prenatal viral cultures, but only if a herpetic lesion is present.
OCR for page 350
--> TABLE D-7 Calculation of Annual Costs for Central Nervous System (CNS) Impairment Secondary to Neonatal HSV Infection (dollars) IOM (1985) estimates:a 7 percent of cases mild CNS impairment, $2,000 per year (1984$), for 20 years ($4,220 in 1994$) 6 percent of cases serious CNS impairment, $5,000 per year (1984$) for 20 years ($10,550 in 1994$) 16 percent of cases very severe CNS impairment, 20,000 per year (1984$) for 20 years ($42,200 in 1994$) Assume average incidence of 322 per year (between the IOM's estimate  and current estimates ). Total: $2.47 million (1994$) a SOURCE: IOM. New vaccine development: establishing priorities; vol. I, Diseases of importance in the United States. Washington, D.C.: National Academy Press, 1985. Neonatal HSV. Neonatal herpes infection is not reported to the CDC, so most estimates of incidence depend on surveys. Based on a review of surveys, Chuang (1988) estimated incidence to be between 1/7,500 births and 1/30,000 births, depending on the geographic location and the population included in the survey. Chuang gives 1/15,000 as the most likely incidence. With 3.979 million births in 1994 (Singh et al., 1995), this estimate suggests a 1994 incidence of 265 cases. The 1985 IOM study estimated first-year treatment costs for neonatal herpes to be $7.156 million (1984$). If adjusted to reflect 265 cases, this figure would suggest costs of $10.530 million (1994$). The IOM estimates were computed by determining the present value of future spending for cases involving CNS impairment. Using the IOM estimates of mild, severe, and very severe impairment and their corresponding costs of annual care, the cost for 1994 cases would be an estimated $2.47 million (1994$) (Table D-7). Thus, in 1994, total annual costs of neonatal HSV would have been an estimated $13.00 million (1994$). Human Papillomavirus (HPV) Infection Like many STDs, the incidence of HPV infection rose through the 1970s and 1980s, but appears to have decreased since 1987. In that year, there were 351,370 initial visits to physicians' offices to seek care for HPV, in contrast to 166,796 initial visits to physicians' offices in 1993 (CDC, DSTDP, 1995). However, some estimates of HPV incidence are several times higher than these figures suggest. The CDC estimates an annual incidence of 500,000 to 1,000,000 cases (CDC, DSTD/HIVP, 1995). Prevalence estimates range from 10 million to 40 million (Fish, 1992); the CDC's estimate is 24 million (CDC, DSTD/HIVP, 1995).
OCR for page 351
--> Treatment. An estimated 20 percent to 30 percent of genital wart infections resolve without treatment (CDC, 1993). The CDC does not recommend a particular treatment for other patients, although some expensive regimens are specifically not recommended. Treatment of HPV often entails multiple office visits, adding to its cost. No estimates of treatment costs or the number of patients treated are available in the literature. Cancer. HPV infection is associated with the risk of genital and anal cancers. Genital warts become precancerous in about 20 percent of affected women, and about 1 percent of these develop into invasive cervical cancer (Fish, 1992). The extent to which HPV infection is responsible for specific cancers, however, is unclear. One recent study found HPV to be responsible for about three-fourths of invasive cervical cancer (Schiffman et al., 1993). Some reviews are more cautious in their assessment of the causative role of this specific agent, noting that epidemiologic evidence does not yet support the theoretical link between HPV and cervical cancer (Reeves et al., 1989; Oriel, 1990; Paavonen et al., 1990). However, there is less reservation about the link between sexually transmitted diseases generally and cervical cancer. The National Institutes of Health estimated the direct cost of cervical cancer to be $610 million in 1990 (1990$) (NIH, 1995) or $791 million in 1994 dollars. Assuming that 50 percent to 90 percent of this cost is attributable to STDs, if not to HPV alone, this would imply an annual cost burden of $396 to $712 million (1994$). Chancroid A total of 773 cases of chancroid were reported to the CDC in 1994 (CDC, DSTDP, 1995). This number represents a decrease from the mid-1980s, when the occurrence of chancroid reached its highest levels since the 1950s (Schmid et al., 1987). Chancroid is difficult to diagnose because of the complexity of laboratory testing procedures. Reliance on clinical diagnosis may result in either underreporting or overreporting (Schmid et al., 1987). The treatment recommended by CDC is azithromycin 1g (CDC, 1993). The current cost for this drug is $24 per treatment regimen. The cost of this treatment, with physician visit, for all cases would be $66,000 (1994$). Hepatitis B Virus Infection Some 200,000 to 300,000 cases of hepatitis B virus infection are estimated to occur annually in the United States, and approximately one million individuals are chronic carriers (Margolis et al., 1991; Hall and Halsey, 1992). Most (90 percent) cases occur among adults. Although drug abuse and occupational exposure are important modes of transmission, sexual intercourse appears to be responsible
OCR for page 352
--> for the more cases than any other mode of transmission-34 percent of cases, according to one source (Margolis et al., 1991), 38 percent according to another (USPSTF, 1996). In at least an additional 30 percent of cases, the mode of transmission is unknown (Margolis et al., 1991). In 1994, the CDC estimated an incidence of 53,000 cases of sexually transmitted hepatitis B virus infection (CDC, DSTD/HIVP, 1995). The cost of hepatitis B virus infection has been a subject of relatively strong interest in the literature; however, no estimates that can reasonably be considered to reflect overall costs are available. Cost estimates for hepatitis B virus infection prepared by the IOM attribute $146.22 million (1984$) to acute infection (IOM, 1985). Using 36 percent (the median of the above estimates) as the proportion attributable to sexual activity, this would imply $111.07 (1994$) in annual costs for acute hepatitis. The sequelae of hepatitis B virus infection (chronic persistent hepatitis, chronic active hepatitis, cirrhosis, and primary hepatocellular carcinoma) are clearly responsible for a large proportion of the costs of this disease. The IOM attributes an additional $138.38 (1984$) to these sequelae. However, this estimate reflects the discounted present value of treatment for these conditions and thus is not an annual cost. More recent studies have focused on the calculation of lifetime costs of hepatitis B virus infection for the purpose of assessing the cost-effectiveness of hepatitis B vaccine (Arevalo and Washington, 1988; Bloom et al., 1993; Margolis et al., 1995). These studies do not present annual costs for sequelae. Indirect Costs Although the estimation of indirect costs was not undertaken for this review, morbidity and mortality are clearly an important part of the burden of STDs. In a recent study, Ebrahim and others (1995) estimated the number of STD-related deaths occurring among women related to STDs to be 9,179 in 1992. This total represents a relatively stable number of STD-related deaths (around 6,500 per year) from non-HIV-related causes and a rapidly increasing annual number of deaths from HIV/AIDS. Cervical cancer is by far the most common cause of STD-related death, representing 5,210 deaths in 1992 (57 percent of the total). (This estimate assumes that sexually transmitted agents are responsible for all deaths from cervical cancer.) The next most common causes of death are HIV/AIDS (2,665 cases, 29 percent of the total), hepatitis B and C virus infection (960 deaths; 10 percent of the total), and PID (220 deaths, 2 percent of the total). An earlier study estimated that STDs were responsible for 20 percent of reproductive mortality among women in 1975. However, this estimate precedes HIV infection and does not include cervical cancer (Grimes, 1986). Indirect costs reflect the quantification of health effects-both morbidity and mortality-in terms of their monetary value. Few studies have examined indirect
OCR for page 353
--> costs of STDs. Those that have done so assumed quantities of work loss due to STD morbidity, including both formal employment and household management (Curran, 1980; Washington et al., 1986, 1987; Washington and Katz, 1991). These assumptions are particularly important in the computation of indirect costs of STDs, because productivity costs related to morbidity appear to dwarf those related to mortality for most STDs (Washington et al., 1986). A preferable option would be to obtain survey data describing the work time lost to STDs, as has been done for other illnesses (Rice et al., 1985); however, this has not yet been attempted. It should also be noted that the value of leisure time is not included in these estimates. The studies assessing indirect costs use gender-specific average wages to place a value on morbidity time. Because many STDs disproportionately affect people earning below average wages-i.e., the poor, certain racial and ethnic groups, and younger people (CDC, DSTDP, 1995)-it can be argued that a lower wage should be used. However, it can also be argued that the use of gender-specific wages for women, because they are lower than those of men, undervalues the economic burden (Curran, 1980). Conclusions The cost estimates reviewed here demonstrate the limited amount of research that has been conducted to quantify the current annual economic burden imposed on society by sexually transmitted diseases. The primary sources of information in this area are the IOM (1985) study, conducted a decade ago, and a series of studies conducted by the Institute for Policy Studies at the University of California at San Francisco (Washington et al., 1986, 1987; Washington and Katz, 1991). Far more research has assessed the cost-effectiveness of specific policies-both proposed and realized-related to sexually transmitted disease. This review also demonstrates the shortage of literature describing the national incidence and prevalence of sexually transmitted diseases and of data documenting the validity of the reporting system for diseases that are reported. Despite the limits of the present review, it is clear that STDs exact a substantial direct cost. Conservatively, the subset of STDs examined here costs the nation at least 4 billion annually. The costs of national STD prevention efforts, including surveillance and education programs, are not included in any studies reviewed in this report. Whether these should be considered part of the burden of disease is not clear. What is clear, however, is that current prevention outlays should not be weighed against current costs of illness. The size of a prevention effort should relate to the benefit of that effort, not to the remaining burden of illness. In fact, the benefit of many preventive measures-those targeting the risky sexual behaviors responsible for the transmission of STDs-includes not only a reduction in the burden
OCR for page 354
--> of STDs, but also reductions in abortions and other consequences of unwanted pregnancy. As described earlier, the estimates compiled in this report must be viewed as only partially describing the economic burden of STDs. Many costs associated with these diseases are not documented in the literature, often because collection of the data necessary to accurately describe these categories of resource use would require large and new research efforts. It should also be noted that this report reviews costs only for a subset of STDs. There are many other STDs, some of which clearly impose a large economic burden. AIDS, for example, was estimated to cost $45,700 (1994$) per patient in one report, a figure that implies an annual cost of $8.4 billion (Hellinger, 1992). Others, such as lymphogranuloma venereum (LGV), are relatively rare and likely to incur low costs. This report thus does not provide a comprehensive estimate of the burden of STDs but summarizes the available information relevant to assessing this burden for a specific subset of sexually transmitted illness. References AMA (American Medical Association). Physician marketplace statistics 1994. Chicago: Center for Health Policy Research, 1994. Arevalo JA, Washington AE. Cost-effectiveness of prenatal screening and immunization for hepatitis B virus. JAMA 1988;259:3:365-9. Begley CE, McGill L, Smith PB. The incremental cost of screening, diagnosis, and treatment of gonorrhea and chlamydial in a family planning clinic. Sex Transm Dis 1989;16:63-7. Bloom BS, Hillman AL, Fendrick M, Schwartz JS. A reappraisal of hepatitis B virus vaccination strategies using cost-effectiveness analysis. Ann Internal Med 1993;118:298-306. Bowie WR. Drug therapies for sexually transmitted diseases: clinical and economic considerations . Drugs 1995;4:496-515. Cates Jr. W. The ''other STDs": do they really matter? JAMA 1988;259:24:3606-8. CDC (Centers for Disease Control and Prevention). 1993 Sexually transmitted diseases treatment guidelines. MMWR 1993;42(No. RR-14):56-66. CDC. Ectopic pregnancy-United States, 1990-1992. MMWR 1995;44:46-8. CDC, DSTD/HIVP (Division of STD/HIV Prevention). Annual report 1994. U.S. Department of Health and Human Services, Public Health Service. Atlanta: Centers for Disease Control and Prevention, 1995. CDC, DSTDP (Division of STD Prevention). Sexually transmitted disease surveillance 1994. U.S. Department of Health and Human Services, Public Health Service. Atlanta: Centers for Disease Control and Prevention, September 1995. Chang TW, O'Keefe P. Cesarean section and genital herpes. N Engl J Med. 1977;296:573. Chuang TY. Neonatal herpes: incidence, prevention, and consequences. Am J Prev Med 1988;4:47-53. Corey L. The current trend in genital herpes: progress in prevention. Sex Transm Dis 1994;21[2 Suppl]:S38-S44. Creinin MD, Washington AE. Cost of ectopic pregnancy management: surgery versus methotrexate. Fertil Steril 1993;60:963-9. Curran JW. Economic consequences of pelvic inflammatory disease in the United States. Am J Obstet Gynecol 1980;138:848-51.
OCR for page 355
--> de Lissovoy G, Zenilman J, Nelson KE, Ahmed F, Celentano DD. The cost of a preventable disease: estimated U.S. national medical expenditures for congenital syphilis, 1990. Public Health Rep 1995;110:403-9. Drug Topics Red Book. Montvale, NJ: Medical Economics Company, Inc., 1995. Ebrahim SH, Peterman TA, Zaidi AA, Kamb ML. Mortality related to sexually transmitted diseases in women, U.S., 1973-1992. Eleventh Meeting of the International Society for STD Research, August 27-30, 1995, New Orleans, LA [abstract no. 343]. Fish RM. Herpes simplex. In: Fish RM, Campbell ET, Trupin SR, eds. Sexually transmitted diseases: problems in primary care. Los Angeles: Practice Management Information Corporation, 1992. Grimes DA. Deaths due to sexually transmitted diseases: the forgotten component of reproductive mortality. JAMA 1986;255:1727-9. Hall CB, Halsey NA. Control of hepatitis B: to be or not to be? Pediatrics 1992;90 [2 Pt. 1] 274-7. Hellinger FJ. Forecasts of the costs of medical care for persons with HIV: 1992-1995. Inquiry 1992;29:356-65. Hibbs JR, Gunn RA. Public health intervention in a cocaine-related syphilis outbreak. Am J Public Health 1991;81:1259-62. IMS America. National disease and therapeutic index (NDTI). Plymouth Meeting, PA: IMS America, 1993. IOM (Institute of Medicine). New vaccine development: establishing priorities; vol. I, Diseases of importance in the United States. Washington, D.C.: National Academy Press, 1985. Johnson RE, Nahmias AJ, Magder LS, Lee FK, Broods CA, Snowden CB. A seroepidemiologic survey of the prevalence of herpes simplex virus type 2 infection in the United States . N Engl J Med 1989;321:7-12. Margolis HS, Alter MJ, Hadler SC. Hepatitis B: evolving epidemiology and implications for control. Semin Liver Dis 1991;11:84-92. Margolis HS, Coleman PJ, Brown RE, Mast EE, Sheingold SH, Arevalo JA. Prevention of hepatitis B virus transmission by immunization. An economic analysis of current recommendations. JAMA 1995;274:1201-8. Moran JS, Kaufman JA, Felsenstein D. Survey of health care providers: who sees patients needing STD services, and what services do they provide? Sex Transm Dis 1995;22:67-9. NIH (National Institutes of Health). Disease-specific estimates of direct and indirect costs of illness and NIH support. Draft document, February 1995. Oriel D. Genital human papillomavirus infection. In: Holmes KK, Mårdh P-A, Sparling PF, Wiesner PJ, Cates W Jr., Lemon SM, Stamm WE, eds. Sexually transmitted diseases. 2nd ed. New York: McGraw-Hill, 1990. Paavonen J, Koutsky LA, Kiviat N. Cervical neoplasia and other STD-related genital and anal neoplasms. In: Holmes KK, Mårdh P-A, Sparling PF, Wiesner PJ, Cates W Jr., Lemon SM, et al., eds. Sexually transmitted diseases. 2nd ed. New York: McGraw-Hill, 1990. Randolph AG, Washington AE, Prober CG. Cesarean delivery for women presenting with genital herpes lesions: efficacy, risks, and costs. JAMA 1993;270:77-82. Reeves WC, Rawls WE, Brinton LA. Epidemiology of genital papillomaviruses and cervical cancer. Rev Infect Dis 1989;11:426-39. Rice DP, Hodgson TA, Kopstein AN. The economic costs of illness: a replication and update. Health Care Financ Rev 1985;7:61-80. Schiffman MH, Bauer HM, Hoover RN, Glass AG, Cadell DM, Rush BB, et al. Epidemiologic evidence showing that human papillomavirus infection causes most cervical intraepithelial neoplasia. J Natl Cancer Inst 1993;85:958-64. Schmid GP, Sanders LL, Blount JH, Alexander ER. Chancroid in the United States: re-establishment of an old disease. JAMA 1987;258:3265-8.
OCR for page 356
--> Singh GK, Mathews MS, Clarke SC, Yannicos T, Smith BL. Annual summary of births, marriages, divorces, and deaths: United States, 1994. Centers for Disease Control and Prevention, National Center for Health Statistics, Monthly Vital Statistics Report, October 23, 1995. USPSTF (U.S. Preventive Services Task Force). Guide to clinical preventive services. 2nd ed. Washington, D.C.: U.S. Department of Health and Human Services, 1996. Washington AE, Arno PS, Brooks MA. The economic cost of pelvic inflammatory disease. JAMA 1986;255:1735-8. Washington AE, Johnson RE, Sanders LL. Chlamydial trachomatis infections in the United States: what are they costing us? JAMA 1987;257:2070-2. Washington AE, Katz P. Cost of and payment source for pelvic inflammatory disease: trends and projections 1983 through 2000 [see comments]. JAMA 1991;266:2565-9. Weström L, Mårdh P-A. Acute pelvic inflammatory disease (PID). In: Holmes KK, Mårdh P-A, Sparling PF, Wiesner PJ, Cates W Jr., Lemon SM, et al., eds. Sexually transmitted diseases. 2nd ed. New York: McGraw-Hill, 1990.
Representative terms from entire chapter: