HISTORICALLY, THE INTERRUPTION OF the transmission of sexually transmitted infections (STIs) through diagnosis and treatment of infected individuals has formed the cornerstone of public health interventions for STI prevention.1 The offering of such services in a confidential environment at no or low cost to the patient has sought to maximize the effectiveness of this strategy. Although the majority of STIs are currently diagnosed in the private sector,2 the provision of STI services through publicly funded STI clinics is still seen as a valuable approach to increase access for socially and economically disadvantaged populations who are at highest risk for STIs.3 However, recent economic developments as well as shifts in public health priorities since the September 11 events have led to changes in public spending. As a result, funding for many public health clinics has decreased, leading to a reassessment of the delivery of such services and finding alternate sources of funding to sustain them, including the institution of patient copayments. Although charging patients a (nominal) fee for services has been widely used in all sectors of health care, not much is known about the effect of patient copayments on the demand for STI services and the ability of the public health system to fulfill its STI control mission.
In the fall of 2002, funding cuts forced the Denver Public Health Department to institute a copayment at its STI clinic. We evaluated the impact of this measure by comparing patient visits and diagnosed STIs during the years immediately before and after the institution of the copayment in December 2002.
The Denver Metro Health Clinic (DMHC) is the largest STI clinic in the Rocky Mountain region, logging 20,000 patient visits in 2002. Starting December 1, 2002, a patient copayment schedule was introduced, including a basic fee of $15 for patients living in the city and county of Denver and the 3 surrounding counties that provide financial support for the clinic. The $15 fee was deemed to be similar to copayments charged elsewhere in the Denver health system, of which DMHC is a part. Patients from outside the 4-county region were charged $65, reflecting the actual cost of a DMHC patient visit. The copay policy was publicized in advance through (e-mail) communications to an extensive list of care providers and health departments in Denver and surrounding counties. The fee structure was posted in the clinic and flyers were distributed to patients when they registered at the clinic. Because of concerns that the copay might have a deterrent effect on patients at highest risk for STIs, a waiver policy was initiated 4 weeks after the initiation of the fee. Patients who, after screening by the clinic clerk, stated that they could not afford the copay were triaged by a clinic nurse and might have their fee waived if they presented with symptoms or if they were referred as contacts to patients with STI diagnosed elsewhere. Follow-up visits remained free of charge. Among patients presenting to DMHC for a new problem during the first full year of fee collection (2003), 8435 patients (74.1%) paid $15,849, 7.5% of patients paid $65.00, and 2099 patients (18.4%) qualified for waiver.
The effects of the patient copayment were evaluated using 3 methods. First, we used the DMHC computerized medical record system to compare patient visits related to a new problem and diagnosed with Neisseria gonorrhoeae (GC) and Chlamydia trachomatis (CT) infections. We compared all visits to the clinic in 2002 and 2003 to examine a potential relationship with the initiation of the copay. To allow for a 1-month “settling” period and to also adjust for seasonal variability, we then conducted comparison analyses for the first 9 months of 2002 and the first 9 months of 2003. We limited our analyses to GC and CT because, first, all men and women presenting with a new problem at DMHC are routinely tested for both CT and GC and, second, these 2 bacterial infections are arguably most significant in terms of prevalence as well as public health benefit through the provision of sexually transmitted disease clinical services. CT and GC testing was conducted using the ProbeTec strand displacement assay (Becton Dickinson, Cockeysville, MD) using first-voided urine (men) and cervical (women) specimens. In addition, GC testing also included routine urethral culture for men and cervical culture for women using Thayer Martin chocolate agar culture plates. Pharyngeal and rectal GC cultures were conducted on indication. GC and CT diagnostic assessments did not change during the study period. For men who have sex with men (MSM), we also examined the effect on the diagnosis of (early) syphilis and HIV infection.
Second, to evaluate whether the copay disproportionately affected persons of lower financial status, we collected socioeconomic information, including household income and number of persons in the household using a brief patient survey that was started 6 weeks before the initiation of the copay and continued thereafter.
Third, to examine the possible compensatory effect of patients accessing care providers outside DMHC, we evaluated all CT and GC case reports from Denver providers to the Colorado Department of Public Health and Environment during 2002 and 2003.
Immediately after the initiation of the copayment, patient visit volume fell from 1639 visits in November 2002 to 1139 a month later, a reduction of 30.5% (Fig. 1). Compared with the first 3 quarters of 2002, DMHC logged 3250 fewer visits in the corresponding period after the copayment took effect, a decline of 28.5%. This decrease was consistent throughout the period, ie, there was no indication that volume increased after the initial decline (Fig. 2). The decline was greater among women (34%) compared with men (25%) and also greater among younger patients (45% among those younger than 20) than among older patients (18.2% among those older than 50). The greatest decline in visits (60.2%) was noted for residents from a neighboring county who were charged the $65 copay.
Diagnosed CT infections declined by 28.1% or 427 cases (from 1519 cases in the first period to 1092 cases in the second) and diagnosed GC infections declined by 38.1% or 332 cases (from 871 to 539 cases). The decline was much stronger among women for CT (38.4% compared with 22.3% for men), but for GC there were no gender differences. Declines also differed by age group. For CT, the strongest decline was among those younger than 20 (43.7%), compared with only 12% and 8.8% for the age groups 30 to 39 and 40 to 49, respectively. By contrast, the decline for GC was higher in older age groups: 47.5% among those aged 40 to 49 compared with 32.8% among those younger than 20 (Fig. 3). There were no clear differences by race/ethnicity. Among persons younger than 25, visits were down by 38%, CT by 38.2% and GC by 33.8%. This age group accounted for 85.6% of fewer diagnosed CT cases and for 39.6% of fewer diagnosed GC cases.
The number of visits for patients who presented with symptoms (presumably less impacted by the copay given the waiver policy in the clinic) declined by 26.3% compared with a decline of 32.1% among those presenting without symptoms. Asymptomatic GC cases declined by 51.0% (from 145 to 71 cases) compared with 35.6% among those who had symptoms (from 726 to 468 cases). Likewise, asymptomatic CT cases showed a 36.7% decline (from 511 to 313 cases) compared with 22.8% among symptomatic cases (from 1008 to 778 cases). Still, of the 332 fewer diagnosed GC cases and of the 418 fewer diagnosed CT cases, symptomatic patients accounted for 77.7% and 53.7%, respectively.
Among MSM, the number of visits declined by 21.1% (from 1085 to 856) and the number of diagnosed GC infections by 40.2% (from 204 to 122 cases) comparing the 2 periods. Early syphilis increased by 8.8% (from 34 to 37 cases) and diagnosed HIV infections declined by 17.8% (from 28 to 23 cases).
The proportion of patients with incomes below 100% of the federal poverty guidelines ($9310 for a 1-person household in 2004) declined from 60.5% during the 6 weeks before the initiation of the copay to 29.4% during the 6-week period thereafter (P <0.0001). During the remainder of 2003, this proportion slowly increased to approximately 41%, still significantly below the level before the start of the copay. However, many patients could not or were unwilling to answer the income question and, as a result, socioeconomic data were available for less than 50% of patients.
During 2003, CT cases were reported from 221 Denver providers compared with 124 providers in 2002 (an increase of 78%). GC cases were reported by121 providers in 2003 compared with 73 in 2002 (an increase of 65.7%). However, all non-DMHC providers combined reported 12.0% fewer CT cases and 8.2% fewer GC cases in 2003 compared with 2002. The ratio of cases reported by DMHC and those reported by non-DMHC providers declined from 0.42 to 0.33 for CT and from 0.94 to 0.61 for GC.
In the period after the institution of a patient copayment, we observed an immediate decline of almost 30%, continuing throughout the observation period, and a concomitant 28% to 38% decline in CT and GC diagnoses. Women and younger patients appeared to be disproportionately affected, especially with regard to CT diagnosis.
As expected, patients of lower socioeconomic status appeared to be disproportionately affected by the copay. The highest copay ($65 charged to residents of certain counties) had the greatest impact on patient visits. Although a differential effect on racial/ethnic groups might have been expected, such an effect was not observed. We speculate that within a setting of an STI clinic, socioeconomic status does not vary much by race/ethnicity such that whites are as severely impacted by a financial barrier as are blacks or Hispanics.
Because symptomatic patients could have their copay waived under certain circumstances, we would have expected symptomatic patients to be less impacted by the copay than patients without symptoms. However, the effect of the copay waiver triage proved difficult to assess. Indeed, the decline of visits by asymptomatic patients was only marginally greater than among patients who presented with symptoms. The absence of a triage effect can be explained in 3 ways. First, the copay was asked of everyone presenting to the clinic regardless of symptoms, and triage only occurred after the patient had indicated that they could not afford the copay. Furthermore, the waiver was not started until a month after the initiation of the copay, and no information was available to assess how well symptomatic patients were reached by the waiver triage. As a result, a number of (symptomatic) patients might have already left the clinic before triage occurred. Second, among asymptomatic patients, there is a significant proportion of “worried well,” which may be more willing or able to pay for services than those who have symptoms. Finally, the publication of the new clinic policy regarding the copay and ensuing word-of-mouth communications may have had a significant deterring effect on patients for whom the nuances of a possible copay waiver were easily lost.
Although for patients with GC and CT the decline among those with symptoms was proportionally smaller than among those without symptoms, symptomatic patients accounted for most of the decline (78% for GC and 54% for CT), indicating that the waiver policy did little to offset the negative impact of the copay.
The major limitation of this article is that it is a simple before–after comparison, which precludes any conclusion about a causative relationship between the institution of the copayment and the subsequent decline in services and diagnosed STIs. However, the immediate and sustained drop of 30% inpatient volume after the institution of the copay and the disproportionate drop in visits among persons of lower socioeconomic status make a causal relationship highly likely. Still, the decline may have been influenced by secular trends unrelated to the effects of the copayment. For example, the decline in CT and GC reports by providers outside DMHC during the study periods could indicate that these infections showed a downward trend outside the institution where the copayment was initiated. However, this decline was much smaller than the one observed at DMHC, and the ratio of cases reported by DMHC and those reported by others also declined. This suggests that even if there was an overall decline in CT and GC incidence during the study period, the stronger decline at DMHC still suggests a role of the copayment. Furthermore, the decline in case reports by other providers could have been a ripple effect caused by fewer contacts of patients diagnosed at DMHC who access services outside DMHC. Alternatively, the lack of a compensatory effect on CT and GC cases reported by other providers may also be explained by underreporting.
Our findings lead us to advocate for the abolishment of copayments in STI clinics or to structure them such that persons would not be denied services on the basis of ability to pay. However, this position may not be tenable given current economic realities that are unlikely to improve anytime soon. Therefore, dependent on budgetary constraints, we must increasingly prioritize STI services that are most likely to yield the largest public health benefit.4 For instance, it could be argued that the diagnosis and treatment of bacterial STIs (gonorrhea, chlamydia, and syphilis) have a greater prevention benefit than the management of viral STIs other than HIV. Hence, at the bare minimum, we should be able to offer diagnosis and treatment for these STIs at no or low cost, whereas other STI services (such as the treatment for genital warts) could be offered using a sliding-fee scale. Along these lines, DMHC has recently initiated a no-fee CT/GC urine screening without examination or other testing. In the meantime, Denver city and county officials are reviewing the copayment data and will determine whether to continue this policy.
In summary, the institution of a copayment in this STI clinic was associated with a sharp decline in services and diagnosed STIs, which may be differentially damaging to vulnerable populations, including the economically disadvantaged, women, and adolescents. Although there are limitations to the generalizability of our results, there is no reason to believe that similar effects are not likely to be observed elsewhere. Because copayments apparently discourage infected individuals from seeking STI services, the time to treatment will lengthen, potentially increasing long-term sequelae. At the same time, the longer period of infectiousness may also lead to more infections, thus reducing the effectiveness of public STI services as a public health intervention.
Postscript: The copayment requirement for Denver county residents was rescinded on Janary 1, 2005.
1. Aral S, Wasserheit J. STD-related health care seeking and health service delivery. In: Holmes K, Mardh P-A, Sparling P, et al., eds. Sexually Transmitted Diseases, 3rd ed. New York: McGraw-Hill, 1999:1295.
2. Brackbill RM, Sternberg MR, Fishbein M. Where do people go for treatment of sexually transmitted diseases? Fam Plann Perspect 1999; 31:10–15.
3. Hook E, Darroch J, Landry D, Mabey D. Sexually transmitted disease and reproductive health clinical services: Categorical clinics and integrated programs. In: Holmes K, Sparling P, Mardh P-A, et al., eds. Sexually Transmitted Diseases, 3rd ed. New York: McGraw-Hill,;1999:1273.
4. Rietmeijer CA. The role of STD outreach testing in times of dwindling STD prevention resources. Sex Transm Dis 2003; 30:659–660.