Do Prescriptions for Expedited Partner Therapy for Chlamydia Get Filled? Findings From a Multi-Jurisdictional Evaluation, United States, 2017–2019

Abstract

Background 

Expedited partner therapy (EPT) is commonly provided by prescription; however, the efficacy of this modality is unknown. We examined whether EPT prescriptions are filled when the cost barrier is removed.

Methods 

To track EPT prescription fill rates, we used single-use pharmacy vouchers that covered the cost of azithromycin, 1 g (chlamydia treatment). We recruited clinical sites to distribute vouchers to patients with chlamydia who would receive an EPT prescription under clinic policies. When distributing vouchers, sites recorded and retained: voucher unique identifier, sex and age of index patient, distribution date, and whether partner name was written on the EPT prescription. Pharmacists receiving vouchers entered the identifier, sex and age of presenting person, and redemption date into a standard pharmacy claim transmission system. Data for redeemed vouchers were retrieved from an industry portal and linked with data retained at clinical sites.

Results 

Thirty-two clinical sites distributed 931 vouchers during September 2017 to January 2019; 382 (41%) were redeemed. Vouchers distributed to patients 18 years or younger (49 [30%] of 163) were less likely to be redeemed compared with those distributed to patients older than 18 years (322 [44%] of 736; P = 0.001). Just over half of vouchers were redeemed the same day (196 [56%] of 352) and 1 mile or less from the clinical site (188 [54%] of 349). After excluding an outlier site, vouchers accompanied by EPT prescriptions including a partner name (15 [56%] of 27) were more likely to be redeemed than those lacking a name (83 [34%] of 244; P = 0.03).

Conclusions 

Less than half of EPT prescriptions were filled, even when medication was free. Whenever possible, EPT should be provided as drug-in-hand.

In Brief

Less than half of prescriptions for free expedited partner therapy for chlamydia were filled. Providing medication in-hand may result in more partner treatment.

More than 1.7 million cases of Chlamydia trachomatis (chlamydia) were reported in the United States (US) in 2018, the largest number of reports for any notifiable condition.¹ Repeated chlamydia infection is common,² and sometimes results from resuming sex with an untreated partner.³ Repeat infection increases risk for adverse outcomes, such as ectopic pregnancy and infertility.⁴ Expedited partner therapy (EPT), a strategy for treating the sex partners of an index patient with a sexually transmitted infection (STI), involves a clinician providing or prescribing additional treatment for sex partners without an intervening medical evaluation.⁵ The practice of EPT is legal in most states.⁶

Three randomized controlled trials established the evidence base for EPT preventing repeat chlamydia in the index patient.^7–9 Two trials used medication-EPT (index patients provided with drug-in-hand to deliver to sex partners),^7,9 and in the third trial, medication was distributed through a chain of pharmacies, an STI clinic, or direct mailing.⁸ However, the efficacy of prescription-EPT (whereby index patients are given prescription(s) to deliver to sex partners) has never been studied.

Information on EPT practices at the population-level are limited; however, evaluations in 2 states suggest that approximately half of EPT is provided in the form of prescription.^10,11 There are several reasons prescription-EPT might not be as effective as medication-EPT. First, prescription-EPT requires more steps than medication-EPT,¹² including the need to travel to a pharmacy and render payment for the medication. Paying for medication is a documented barrier to filling EPT prescriptions,¹⁰ and the out-of-pocket cost of EPT treatment for chlamydia (generic azithromycin, 1 gram) has been reported to be as high as US $120.^13,14

Second, neighborhoods with high STI rates may also have a paucity of pharmacies, making it especially difficult for people to fill prescriptions.¹³ Third, many pharmacists lack knowledge about the specifics of EPT laws^13–15 and may refuse to fill EPT prescriptions.¹⁶ In 1 study, pharmacists refused to fill 58% of EPT prescriptions with no patient name in a state where “nameless” EPT prescriptions are legal.¹⁶

The objective of this evaluation was to measure the proportion of EPT prescriptions that get filled, because obtaining medication from a pharmacy might be a rate-limiting step in prescription-EPT. To examine prescription fill rates in a best-case scenario, we utilized a pharmacy voucher to eliminate any barriers posed by the cost of medication and simultaneously monitored the fulfillment of EPT prescriptions.

MATERIALS AND METHODS

Pharmacy Discount Voucher

A pharmaceutical company (RxCrossroads by McKesson) designed a pharmacy voucher in the form of a discount card specifically for this project (Fig. 1). The voucher fully covered the costs for azithromycin, 1 g, tablet form (treatment for chlamydial infection), and was designed to function within a free medication program model typically used by pharmaceutical manufacturers to engage new patients. The program infrastructure enabled automatic reimbursement to pharmacies for every dose of medication dispensed to a person presenting a voucher, and made it possible for the investigators to monitor voucher redemption at pharmacies.¹⁷ Vouchers could be redeemed at any pharmacy in the US, excluding those in states that do not permit EPT for chlamydia (15 states⁶) and states that require sales tax collection on prescription drugs (2 states) (K. Lyman, personal communication).

Figure 1: Pharmacy voucher for free EPT treatment,* with detachable “tear-off tab” to record information on the index patient.† The bottom section shows the pharmacy voucher that fully covered costs for azithromycin, 1 g, tablet form when presented at the pharmacy.* The top section shows the detachable “tear-off tab” that health care providers completed when distributing the card to an index patient with chlamydia. Providers removed this tear-off tab before distribution and placed it in a ballot box for subsequent collection by the investigators.†

The voucher was designed so that a unique identifier appeared in 2 places on each voucher: in the body of the voucher and on a “tear-off tab” attached to the top of the voucher (Fig. 1). Health care providers distributing vouchers recorded limited information on the tear-off tab, and removed and retained the tear-off tab from each voucher before handing to the patient along with an EPT prescription.

When an index patient or sex partner presented a voucher at a pharmacy, the pharmacist transmitted a pharmacy claim to the pharmaceutical company, entering the unique identifier on the voucher. This process follows the National Council for Prescription Drug Programs' (NCPDP) standard for transactions between pharmacies and a third-party payer.^17,18 To further support utilization, instructions for processing were printed on the back of each voucher, including information for a Help Desk for pharmacists to contact with questions.

Each voucher was affixed to a brochure containing information about EPT for sex partners, as required in New York State.¹⁹ Brochures were available in either English or Spanish.

Recruitment and Training of Clinical Sites

To distribute the vouchers, we sought to enroll clinical sites that: (a) diagnosed and treated a substantial number of chlamydia infections among heterosexual patients and (b) had experience with prescription-EPT. We focused on chlamydia diagnosed among heterosexual index patients because the Centers for Disease Control and Prevention do not routinely recommend EPT for men who have sex with men.²⁰ We did not enroll sites already practicing medication-EPT, including the New York City (NYC) Sexual Health Clinics, because we did not want this evaluation of prescription-EPT to undermine a modality with established efficacy.⁵

To identify clinical sites in NYC, we reviewed 2015 to 2016 surveillance data for provider reports of chlamydia that indicated that the index patient had been given prescription-EPT. For clinical sites in the other jurisdictions (Maryland, California, and New York outside NYC), we contacted state and local health departments and relevant faculty at academic institutions for suggestions of sites that met the criteria for participation. New York,²¹ Maryland,²² and California²³ all have legal provisions allowing clinicians to provide EPT for chlamydia.

Clinical sites were enrolled in the project on a rolling basis during September 2017 to November 2018 and received standard training by the project investigators about how providers should distribute the vouchers. For sites located in Maryland, partners at an academic research center who were trained by the investigators provided the project training. To guide the number of vouchers investigators provided, clinical sites were asked to estimate the number of chlamydia infections recently diagnosed in their practice and the percentage of these patients who received prescription-EPT.

Clinical sites were instructed to offer vouchers, along with EPT prescriptions, to each patient diagnosed with chlamydia who would normally receive EPT according to clinic policies and procedures. When distributing a voucher, health care providers recorded the index patient's sex, age, and date of distribution on the tear-off tab (Fig. 1). Midway through the project, materials were reprinted after an extension of the funding timeline. For the reprinted materials, a field was added onto the tear-off tab for providers to indicate whether the EPT prescription accompanying the voucher had a name for the intended sex partner. This field was added in response to concerns reported by clinical sites that “nameless” EPT prescriptions may not be accepted at pharmacies. In both phases of the project, after completing information on the tear-off tab, clinicians then removed the tab and placed it in a ballot box for subsequent collection by the investigators. Finally, providers gave the index patient the voucher(s) and EPT prescription(s) for their sex partners; during training, providers were instructed to advise patients to take the prescription and voucher to the pharmacy.

For 3 clinical sites that expressed concern that index patients might use a voucher for their own, rather than their partner's treatment, we provided spare vouchers to cover medication costs for select index patients (e.g., uninsured). We tracked these vouchers using the unique identifier, and excluded from all analyses. At the conclusion of the project, sites completed a survey to document characteristics of the clinical setting and practices related to distributing the vouchers.

Pharmacy Voucher Redemption

When a voucher was presented at a pharmacy, the pharmacist submitted the claim electronically to the pharmaceutical company, entering the following variables of interest into a NCPDP-standard claim transmission system: voucher unique identifier, date of submission, age, and sex of the person presenting the voucher, and address and National Provider Identifier for the pharmacy. After the claim was adjudicated and approved, a process that happens in real-time,¹⁷ the pharmacist dispensed EPT medication at no cost. Data submitted as part of the claim process were accessible to the investigators within approximately 24 hours through a data portal.

Linking Index Patient Demographics With Data Obtained During the Voucher Claim Process

Midway and at the end of the project, the investigators collected the tear-off tabs from each clinical site. Next, information recorded by providers on the tear-off tabs was entered into an Access database developed by the investigators (Microsoft Corporation, Redmond, WA), and pharmacy data were extracted from the pharmaceutical company portal. These 2 data sets, representing the distribution (tear-off tabs) and redemption (pharmacy portal) of vouchers, were linked based on unique identifier.

Analysis

The primary objective was to determine the percentage of EPT prescriptions filled when the cost barrier is removed, using voucher redemption as a surrogate measure. This evaluation was not a pilot or proof of concept for the practice of distributing a voucher along with an EPT prescription, and as such, we did not collect data to assess the cost of such a program. We calculated the number of vouchers distributed, and the number and proportion of vouchers redeemed, overall and by characteristics of the index patient and site of distribution. Associations between voucher redemption and traits of the index patient and clinical sites were examined using χ² and Fisher exact tests.

For redeemed vouchers, we assessed patterns related to fulfillment. This included quantifying the frequency of redeemed vouchers, by sex and age of the person who presented the voucher, number of days between distribution and redemption, distance between clinical site and redeeming pharmacy, type of pharmacy, and comparison of the sex reported for the index patient who received and person who claimed the voucher.

Pharmacy type was classified as large chain (pharmacy name corresponds with a top 25 pharmacy company²⁴), ambulatory clinic (pharmacy National Provider Identifier indicates a clinical scope of practice and co-location with a clinical site), and all others. The distance between clinical site and pharmacy was ascertained by geocoding the address for both locations and measuring the straight-line (Euclidean) distance using ArcGIS software, version 10 (Environmental Systems Research Institute, Redlands, CA).

Data from the site survey were used to characterize participating clinical sites, including practice type and primary specialty, geography, pharmacy accessibility, and experience with prescription-EPT. We used the 2013 National Center for Health Statistics Urban–Rural Classification Scheme to classify the urbanization of the county where each site was located.²⁵

All analyses were conducted in SAS, version 9.4 (SAS Institute, Inc., Cary, NC). This investigation was considered program evaluation, and therefore was not subject to review by the NYC Department of Health and Mental Hygiene Institutional Review Board.

RESULTS

Participating Clinical Sites

A total of 32 clinical sites were recruited to participate in this project (Table 1), including locations in NYC, New York State (outside NYC), Maryland (Baltimore City and County), and California (Fresno County). A diverse group of clinical settings distributed vouchers, including community health centers (20 [63%] of 32), college student health centers (3 [9%] of 32), hospital-affiliated clinics (6 [19%] of 32), and publicly funded STI clinics (3 [9%] of 32). Nearly all (29 [91%] of 32) sites reported that more than 1 clinician wrote EPT prescriptions as part of this project.

TABLE 1: Number and Characteristics of Participating Clinical Sites (N = 32)

Most clinical sites had longer than 2 years of experience with prescription-EPT (19 [59%] of 32), but 7 (22%) sites were new to prescribing EPT. The majority of clinical sites limited vouchers to index patients with laboratory-confirmed chlamydia (25 [78%] of 32), and placed no restrictions on the number of vouchers that could be given to 1 index patient (29 [91%] of 32).

Allocation of Pharmacy Vouchers

The investigators allocated 3560 vouchers to clinical sites for EPT treatment, including 3020 (85%) with English-language and 540 (15%) with Spanish-language brochures.

Distribution of Pharmacy Vouchers

Of the 3560 vouchers allocated by the investigators, sites reported distributing 931 (26%) to index patients from September 26, 2017, to January 31, 2019. There was wide variation in the number of vouchers distributed by each site, with 1 publicly funded STI clinic distributing 181 vouchers, and 13 sites distributing less than 10 vouchers. Close to 80% of the vouchers (741 of 931) were distributed by 31% of sites (10 of 32).

Among vouchers with complete information, the majority were distributed to patients older than 18 years (736 [82%] of 899) and female patients (651 [71%] of 922). Of note, investigators identified 2 obstetrics and gynecology sites that documented distributing over 55% of vouchers to males. Upon investigation, we learned that these sites were primarily completing tabs with reference to sex partners. To account for this, we reclassified index patient sex as “female” for all tear-off tabs from these 2 sites (n = 43). For vouchers given in the second phase of the project, nearly all (305 [92%] of 333) accompanied an EPT prescription with no name (Table 2).

TABLE 2: Distribution and Redemption of EPT Vouchers, by Characteristics of Index Patients and Clinical Site

Redemption of Pharmacy Vouchers

Among the 931 vouchers distributed for EPT, 382 (41%) were redeemed (Table 2). When examining associations between index patients and voucher redemption, significant relationships emerged for both sex and age. Vouchers were more likely to be redeemed if they were given to females (286 [44%] of 651) compared with males (94 [35%] of 271; P = 0.009), and to patients older than 18 years (322 [44%] of 736) compared with patients 18 years or younger (49 [30%] of 163; P = 0.001) (Table 2). Vouchers were also most likely to be redeemed if they were issued at student health centers (129 [72%] of 180) and sites with an on-site pharmacy (134 [58%] of 232) (P < 0.001). Of note, when we linked tear-off tabs with redeemed vouchers, we identified 33 redeemed vouchers for which we did not have the matching tear-off tab. These missing tabs corresponded with vouchers allocated to 11 sites. We removed these 33 vouchers from analyses.

One college health center distributed a large number of vouchers (n = 148), and also had a redemption percentage in the top quartile (113 [76%] of 148). Clinicians at this site worked in close collaboration with an on-site pharmacy that processed 87% of redeemed vouchers. Given this site's large volume and unique features, we conducted a subanalysis excluding vouchers distributed at this health center.

After excluding the student health center, the redemption percentage decreased to 34% (Table 3). Age remained significantly associated with redemption; a lower percentage of vouchers distributed to patients 18 years or younger (39 [26%] of 151) were redeemed when compared with those older than 18 years (223 [37%] of 604; P = 0.01). A new relationship emerged, such that a greater proportion of vouchers linked to “named” EPT prescriptions (15 [56%] of 27) were redeemed than those with a “nameless” prescriptions (83 [34%] of 244; P = 0.03).

TABLE 3: Distribution and Redemption of EPT Vouchers, by Characteristics of Index Patients, excluding high-volume student health center

The majority of people who redeemed a voucher were female (215 [56%] of 382) and 18 years or older (326 [85%] of 382) (Table 4). Most vouchers were redeemed the same day as distribution (196 [56%] of 352), and at a pharmacy 1 mile or less from the distribution site (188 [54%] of 349). Across the project, 160 pharmacies processed 1 voucher or more. Vouchers were redeemed in many types of pharmacies, including large chain (221 [58%] of 382), ambulatory clinic (103 [27%] of 382), and all others (58 [15%] of 382).

TABLE 4: Redemption of EPT Vouchers, by Characteristics of Patients, Clinical Sites, and Pharmacies (N = 382)

DISCUSSION

Using pharmacy voucher redemption as a proxy measure for EPT prescription fulfillment, we found that less than half of the vouchers were redeemed, suggesting that only a minority of partners get treated when EPT for chlamydia is provided in the form of a prescription. The fulfillment of vouchers was especially low (≤30%) for adolescents. This investigation is among the first to report on the frequency of nameless EPT prescriptions, and we found that nearly all EPT prescriptions lacked a name in the second phase of the project. Moreover, a lower proportion of these nameless EPT prescriptions were filled, though these findings are based on a smaller number of vouchers. These findings provide important insights into how the commonly practiced, yet understudied, method of prescription-EPT works in the real-world.

The voucher redemption percentage we observed aligns closely with findings from a similar real-world investigation in Scotland which found that only 40% of pharmacy vouchers given to index patients with chlamydia for partner treatment were redeemed.²⁶ This would translate into markedly lower rates of partner treatment than those reported by index patients in the randomized clinical trials that established the efficacy of EPT (64% to 86% of sex partners were reported to be “very likely” to have taken the EPT medication).⁵

The proportion of vouchers redeemed varied significantly by index patient age, with only 30% of vouchers given to index patients 18 years or younger being redeemed. This extends findings from an investigation at family planning centers in California where the authors found that patients 18 or younger filled only 47% of prescriptions for their own chlamydia treatments, raising the concern of whether “fill rates may be even lower for receipt of partner treatment.”²⁷ This suboptimal fulfillment suggests that EPT should be provided as medication whenever possible, especially for adolescents.

Although the overall redemption percentage suggests that prescription-EPT may not be an effective partner treatment strategy, in 1 student health center that distributed a high volume of vouchers, 76% of vouchers were redeemed. This site had several unique characteristics, including an on-site pharmacy and collaboration between clinical and pharmacy staff that facilitated consistent fulfillment of EPT prescriptions. This pharmacy filled 87% of redeemed vouchers from this site, indicating that prescription-EPT may be a viable partner treatment option under specific, supportive circumstances.

Nearly all EPT prescriptions (92%) that accompanied vouchers in the second phase of the project had no name documented. After excluding the unique student health center, a lower proportion of vouchers that were accompanied by a nameless prescription were redeemed. It is likely that some pharmacists will not fill nameless EPT prescriptions, as was reported by investigators in Milwaukee who conducted a secret shopper study using nameless EPT prescriptions.¹⁶ Studies in multiple states have documented limited EPT knowledge among pharmacists,^13–15 and because the majority of EPT prescriptions may be written with no name, engaging and educating pharmacists about the legality and process for filling nameless EPT prescriptions is critical.

The findings of this investigation should be interpreted in the context of limitations related to the allocation, distribution, and redemption of vouchers. Only a small proportion of the vouchers that investigators allocated to clinical sites were distributed to index patients (26%). This discrepancy (between anticipated and actual distribution) likely reflects the “presumptive treatment gap,” whereby many patients are presumptively treated before laboratory confirmation of chlamydia, and therefore are not offered or do not return for EPT treatment when their laboratory test results become available.^12,28 However, we do not know how many patients were eligible, or how clinicians decided when to offer vouchers.

To facilitate broad implementation of this project, we directed sites to distribute vouchers based on their standard EPT practices. Consequently, there was wide variation in sites' processes for treating index patients (e.g., e-prescription, on-site treatment) and distributing the vouchers (e.g., limiting EPT to persons with laboratory-confirmed chlamydia, offering EPT only to females). While this assortment of practices reflects the numerous ways that prescription-EPT is implemented in the real-world, it prohibited us from performing subgroup analyses to examine how various factors might impact provision and use of prescription-EPT.

The validity of information on index patient sex is unclear. Sites were instructed to record information about the index patient on tear-off tabs; however, we identified 2 sites that were primarily reporting sex partner sex on the tabs. While we reclassified index patient sex for these 2 sites, we were unable to identify or quantify misclassification of index patient sex at other sites that served both male and female patients.

Furthermore, the collection of tear-off tabs was incomplete, as we found 33 redeemed vouchers that did not have a matching tear-off tab. Although we removed these 33 vouchers from analyses, it is possible that there were other vouchers distributed to index patients that were not redeemed. If so, we may have underestimated the number of distributed vouchers (the denominator), such that we may have overestimated the percentage of vouchers redeemed.

There were 30 vouchers presented at a pharmacy that did not result in provision of free EPT treatment. These claims were rejected based on parameters specified by the investigators when designing the voucher (e.g., redeemable in states where EPT is legal, usable for 1 EPT dose, coverage for generic azithromycin). This unfavorable outcome may have resulted in confusion and inconvenience for the index patient and lack of treatment for the sex partner. While we aimed to avoid voucher rejections through printed instructions and a Help Desk number, this underscores the real-world challenges involved in prescription-EPT.

Most importantly, we used voucher redemption as a proxy for filling an EPT prescription. We could have underestimated the percentage of prescriptions filled if people did not use the voucher when filling an EPT prescription, and we cannot measure the ultimate outcome of interest, that is, whether a partner took the prescribed medication.

The overarching aim of EPT is to treat sex partners, and the effectiveness of prescription-EPT hinges on whether prescriptions get filled. Our findings indicate that less than half of EPT prescriptions get filled. This is in sharp contrast to the high percentages of partner treatment reported by index patients in the randomized trials for EPT⁵ and suggests that prescription-EPT may not be a highly effective mechanism for treating sex partners and may be particularly problematic for adolescents, among whom voucher redemption was even lower. While prescription-EPT may work well in specific settings, such as university health centers equipped with an on-site pharmacy, overall this evaluation indicates that EPT should be provided as medication in-hand whenever possible, and adolescents should be prioritized for medication-EPT if resources are limited.

To support the provision of medication-EPT by health care providers, the California Department of Public Health, in collaboration with Essential Access Health, developed a robust program to distribute free EPT medication to clinics that serve underinsured patient populations at risk for STIs.²⁹ This program distributed over 23,500 free doses of EPT to 188 clinics in 2017 and reported that 38% of chlamydia cases at participating clinics received at least 1 dose of EPT from 2016–2018. This program both expanded reach of medication-EPT across the State and created opportunities for technical assistance.²⁹ Health departments should explore the feasibility of establishing similar programs that purchase and distribute medication-EPT in bulk.

Lastly, the high prevalence of nameless EPT prescriptions documented by this project, and association between nameless prescriptions and lower voucher redemption emphasizes the crucial role of pharmacists in understanding, accepting, and processing EPT prescriptions. As jurisdictions seek to strengthen partner treatment, consideration of medication distribution programs and widespread pharmacist education are critical.