Abstract
Our objective was to determine if oral vancomycin, fidaxomicin, and oral metronidazole use in the United States changed after publication of revised clinical practice guidelines for Clostridium difficile infection (CDI) in February 2018.
We obtained US antibiotic prescription data (IQVIA) from 2006–August 2019 and used guideline-recommended dosing regimens to estimate monthly numbers of 10-day treatment courses of vancomycin, fidaxomicin and metronidazole. Interrupted time-series analyses were performed, adjusted by month. We compared linear trends for monthly numbers of treatment courses in different time periods.
Cumulative treatment courses of oral vancomycin and fidaxomicin increased by 54% (n = 226 166) and 48% (n = 18 518), respectively, in 18 months following guidelines compared with 18 months before; those of oral metronidazole decreased by 3% (n = 238 372). Monthly vancomycin and fidaxomicin use significantly increased throughout the period following revised guidelines (P < .0001 and P = .0002, respectively), whereas that of metronidazole decreased significantly (P < .0001). Monthly vancomycin use increased and metronidazole use decreased to a significantly greater extent after publication of revised guidelines than after publication of clinical trials establishing superiority of vancomycin over metronidazole (P < .0001).
Revised practice guidelines have had a significant impact on CDI treatment in the US. Clinical trial data used for the revised guidelines were available since 2007–2014 and 2011–2012 for oral vancomycin and fidaxomicin, respectively. Guidelines or guidance documents for treating CDI and other infections should be updated in more timely fashion.
(.)
Clostridium (Clostridioides) difficile infection (CDI) is designated as an urgent threat to public health by the US Centers for Disease Control and Prevention (CDC) [1]. The CDC estimates that 225 000–275 000 CDI cases were diagnosed each year between 2012 and 2017 among patients in US hospitals, accounting for 12 800 deaths and $1 billion in attributable healthcare costs annually [1]. Over the same time period, the incidence of community-onset CDI increased in the United States, accounting for almost 50% of CDI cases in 2017 [2, 3]. Metronidazole and oral vancomycin have been primary treatment options against CDI for over 30 years. Oral metronidazole was approved by the US Food and Drug Administration (FDA) in 1963, albeit not with an indication for treatment of CDI. Vancocin (Pfizer, Inc) and generic vancomycin capsules were initially FDA approved against CDI in 1986 and 2012, respectively. A newer oral agent, fidaxomicin (Dificid; Merck & Co), was FDA approved against CDI in 2011. Bezlotoxumab (Zinplava; Merck & Co), a monoclonal antibody directed against C. difficile toxin B, received FDA approval in 2016 as an adjunctive agent to anti-CDI antibiotics in reducing disease recurrences.
The Infectious Diseases Society of America (IDSA) and Society for Healthcare Epidemiology of America (SHEA) first published joint clinical practice guidelines for CDI in 2010, which did not include fidaxomicin (Table 1) [4]. Revised IDSA/SHEA guidelines, published in February 2018, assimilated clinical trial data that became available after 2010 on the performance of oral vancomycin, fidaxomicin, and oral metronidazole in treating CDI in adults [6]. The revised guidelines relied heavily upon pooled data from 2 multinational, randomized clinical trials (RCTs) published in 2014 by Polymer Alternative for CDI Treatment (PACT) investigators that demonstrated superior clinical responses with vancomycin compared with metronidazole [7]. The IDSA/SHEA committee also incorporated newly combined data from 5 CDI RCTs published between 2000 and 2014, demonstrating that vancomycin was superior to metronidazole in achieving initial and sustained resolution of symptoms [6]. Two other pivotal RCTs cited in the revised guidelines were published in 2011 and 2012, comparing vancomycin with fidaxomicin for treatment of nonfulminant CDI. Clinical cure rates were comparable for the drugs, but fidaxomicin was superior in achieving sustained resolution of symptoms, in particular for infections caused by non-NAP1 C. difficile [8, 9]. The revised guidelines endorsed vancomycin or fidaxomicin as preferred treatments for initial and first recurrent nonsevere CDI in adults, rather than metronidazole as previously recommended. Oral vancomycin remained a first-line treatment for initial and first recurrent severe CDI, but fidaxomicin was newly recommended as an alternative first-line agent. Bezlotoxumab was not included in the revised guidelines.
Revised (2018) and 2010 Recommendations for Treatment of Clostridium difficile Infection in Adults
| Clinical Definitiona | Supportive Clinical Dataa | Recommended Treatment | |
|---|---|---|---|
| Revised Guidelines | 2010 Guidelines | ||
| Initial episode, nonsevereb | WBC ≤15 000 cells/mL and serum creatinine <1.5 mg/ dLb | Vancomycin 125 mg po QID or fidaxomicin 200 mg po BID for 10 days | Metronidazole 500 mg po TID for 10–14 days |
| Initial episode, severec | WBC >15 000 cells/mL or serum creatinine >1.5 mg/dLc | Vancomycin 125 mg po QID or fidaxomicin 200 mg po BID for 10 days | Vancomycin 125 mg po QID for 10–14 days |
| Initial episode, fulminantd | Hypotension or shock, ileus, megacolond | Vancomycin 500 mg po QID + metronidazole 500 mg IV Q8 hours for 10 dayse | Vancomycin 500 mg po QID + metronidazole 500 mg IV every 8 hrs for 10–14 dayse |
| First recurrencec | N/A | Vancomycin 125 mg po QID for 10 days or vancomycin po taper/pulse or fidaxomicin 200 mg po BID for 10 days | Same as initial episode |
| Second/subsequent recurrencec | N/A | Vancomycin po taper/pulse or vancomycin 125 mg po QID for 10 days, then rifaximin 400 mg po TID for 20 days or fidaxomicin 200 mg po BID for 10 days or fecal microbiota transplantf | Vancomycin po taper or pulse |
| Clinical Definitiona | Supportive Clinical Dataa | Recommended Treatment | |
|---|---|---|---|
| Revised Guidelines | 2010 Guidelines | ||
| Initial episode, nonsevereb | WBC ≤15 000 cells/mL and serum creatinine <1.5 mg/ dLb | Vancomycin 125 mg po QID or fidaxomicin 200 mg po BID for 10 days | Metronidazole 500 mg po TID for 10–14 days |
| Initial episode, severec | WBC >15 000 cells/mL or serum creatinine >1.5 mg/dLc | Vancomycin 125 mg po QID or fidaxomicin 200 mg po BID for 10 days | Vancomycin 125 mg po QID for 10–14 days |
| Initial episode, fulminantd | Hypotension or shock, ileus, megacolond | Vancomycin 500 mg po QID + metronidazole 500 mg IV Q8 hours for 10 dayse | Vancomycin 500 mg po QID + metronidazole 500 mg IV every 8 hrs for 10–14 dayse |
| First recurrencec | N/A | Vancomycin 125 mg po QID for 10 days or vancomycin po taper/pulse or fidaxomicin 200 mg po BID for 10 days | Same as initial episode |
| Second/subsequent recurrencec | N/A | Vancomycin po taper/pulse or vancomycin 125 mg po QID for 10 days, then rifaximin 400 mg po TID for 20 days or fidaxomicin 200 mg po BID for 10 days or fecal microbiota transplantf | Vancomycin po taper or pulse |
Abbreviations: BID, twice per day; CDI, Clostridium difficile infection; FDA, Food and Drug Administration; IV, intravenous; N/A, not applicable; po, by mouth; QID, 4 times per day; TID, 3 times per day; WBC, white blood cell count.
aClinical definitions used in the updated guidelines.
bThese criteria correspond to the clinical definition of “initial episode, mild or moderate” in the 2010 guidelines.
cDefinition and terminology are the same in the 2010 guidelines.
dThese criteria correspond to the clinical definition of “initial episode, severe, complicated” in the 2010 guidelines.
eUpdated and 2010 guidelines recommend considering rectal installation of vancomycin in the event of ileus.
fIn a small randomized controlled trial, rifamixin given after standard therapy for CDI was associated with a trend toward fewer recurrences than were seen with placebo [5]. Other non–FDA-approved agents that have been used in the past against CDI, including nitazoxanide, fusidic acid, tigecycline, and bacitracin, were not included in the guidelines. Bezlotoxumab, approved by the FDA in 2016 for prevention of CDI recurrences by toxin B–producing strains (as adjunct to anti–CDI antibiotics), does not appear in the guidelines; its use to date has been minimal.
Revised (2018) and 2010 Recommendations for Treatment of Clostridium difficile Infection in Adults
| Clinical Definitiona | Supportive Clinical Dataa | Recommended Treatment | |
|---|---|---|---|
| Revised Guidelines | 2010 Guidelines | ||
| Initial episode, nonsevereb | WBC ≤15 000 cells/mL and serum creatinine <1.5 mg/ dLb | Vancomycin 125 mg po QID or fidaxomicin 200 mg po BID for 10 days | Metronidazole 500 mg po TID for 10–14 days |
| Initial episode, severec | WBC >15 000 cells/mL or serum creatinine >1.5 mg/dLc | Vancomycin 125 mg po QID or fidaxomicin 200 mg po BID for 10 days | Vancomycin 125 mg po QID for 10–14 days |
| Initial episode, fulminantd | Hypotension or shock, ileus, megacolond | Vancomycin 500 mg po QID + metronidazole 500 mg IV Q8 hours for 10 dayse | Vancomycin 500 mg po QID + metronidazole 500 mg IV every 8 hrs for 10–14 dayse |
| First recurrencec | N/A | Vancomycin 125 mg po QID for 10 days or vancomycin po taper/pulse or fidaxomicin 200 mg po BID for 10 days | Same as initial episode |
| Second/subsequent recurrencec | N/A | Vancomycin po taper/pulse or vancomycin 125 mg po QID for 10 days, then rifaximin 400 mg po TID for 20 days or fidaxomicin 200 mg po BID for 10 days or fecal microbiota transplantf | Vancomycin po taper or pulse |
| Clinical Definitiona | Supportive Clinical Dataa | Recommended Treatment | |
|---|---|---|---|
| Revised Guidelines | 2010 Guidelines | ||
| Initial episode, nonsevereb | WBC ≤15 000 cells/mL and serum creatinine <1.5 mg/ dLb | Vancomycin 125 mg po QID or fidaxomicin 200 mg po BID for 10 days | Metronidazole 500 mg po TID for 10–14 days |
| Initial episode, severec | WBC >15 000 cells/mL or serum creatinine >1.5 mg/dLc | Vancomycin 125 mg po QID or fidaxomicin 200 mg po BID for 10 days | Vancomycin 125 mg po QID for 10–14 days |
| Initial episode, fulminantd | Hypotension or shock, ileus, megacolond | Vancomycin 500 mg po QID + metronidazole 500 mg IV Q8 hours for 10 dayse | Vancomycin 500 mg po QID + metronidazole 500 mg IV every 8 hrs for 10–14 dayse |
| First recurrencec | N/A | Vancomycin 125 mg po QID for 10 days or vancomycin po taper/pulse or fidaxomicin 200 mg po BID for 10 days | Same as initial episode |
| Second/subsequent recurrencec | N/A | Vancomycin po taper/pulse or vancomycin 125 mg po QID for 10 days, then rifaximin 400 mg po TID for 20 days or fidaxomicin 200 mg po BID for 10 days or fecal microbiota transplantf | Vancomycin po taper or pulse |
Abbreviations: BID, twice per day; CDI, Clostridium difficile infection; FDA, Food and Drug Administration; IV, intravenous; N/A, not applicable; po, by mouth; QID, 4 times per day; TID, 3 times per day; WBC, white blood cell count.
aClinical definitions used in the updated guidelines.
bThese criteria correspond to the clinical definition of “initial episode, mild or moderate” in the 2010 guidelines.
cDefinition and terminology are the same in the 2010 guidelines.
dThese criteria correspond to the clinical definition of “initial episode, severe, complicated” in the 2010 guidelines.
eUpdated and 2010 guidelines recommend considering rectal installation of vancomycin in the event of ileus.
fIn a small randomized controlled trial, rifamixin given after standard therapy for CDI was associated with a trend toward fewer recurrences than were seen with placebo [5]. Other non–FDA-approved agents that have been used in the past against CDI, including nitazoxanide, fusidic acid, tigecycline, and bacitracin, were not included in the guidelines. Bezlotoxumab, approved by the FDA in 2016 for prevention of CDI recurrences by toxin B–producing strains (as adjunct to anti–CDI antibiotics), does not appear in the guidelines; its use to date has been minimal.
At the time that the revised guidelines were published, comparative effectiveness data for oral vancomycin and metronidazole in treatment of CDI in children were lacking, and fidaxomicin was not approved by the FDA for use in patients younger than 18 years of age. The guidelines endorsed oral vancomycin or oral metronidazole as the best treatment of an initial episode or first recurrence of pediatric CDI. Vancomycin was preferred over metronidazole for initial episodes of severe CDI and for second or greater disease recurrence. Fidaxomicin was not included as a recommended treatment of CDI in children. The 2010 IDSA/SHEA guidelines were specifically for the management of CDI in adults, and they did not include pediatric recommendations.
Our objective in this study was to determine if the use of oral vancomycin, fidaxomicin, or oral metronidazole changed in the United States after publication of the revised guidelines.
METHODS
We obtained US data on dispensed antibiotic prescriptions from IQVIA (Durham, NC) for commercially available oral formulations of vancomycin, fidaxomicin, and metronidazole from March 2006 through August 2019. Bezlotoxumab was not included in our analysis because it did not appear in the revised IDSA/SHEA guidelines and yearly US prescriptions are significantly less than those of the other agents. Prescription data were provided as numbers of units dispensed monthly of the following oral formulations: vancomycin, 125 mg; vancomycin, 250 mg; fidaxomicin, 200 mg; or metronidazole, 500 mg. Data for vancomycin and metronidazole included generic drugs and Vancocin or Flagyl (Pfizer, Inc), respectively; fidaxomicin is available only as Dificid. Oral doses of vancomycin that were compounded from powder or intravenous formulations were not available, nor were data on the use of vancomycin hydrochloride powder for oral suspension (Firvanq; Cutis Pharma). In order to normalize monthly dispensed prescriptions, we converted numbers of units of different agents into numbers of 10-day treatment courses of the following regimens, as recommended in the revised guidelines: vancomycin, 125 mg 4 times per day (QID); vancomycin, 500 mg QID; fidaxomicin, 200 mg twice per day (BID); or metronidazole, 500 mg 3 times per day (TID). For example, 80 units of vancomycin 250 mg were converted to one 10-day treatment course of vancomycin 500 mg QID. Data were expressed as 3-month rolling averages of the numbers of 10-day treatment courses for each drug per month. Vancomycin 125 mg QID and 500 mg QID courses were combined.
Interrupted time-series analyses were performed using numbers of treatment courses, adjusted by month for possible seasonality. Standard errors for coefficients used the Newey-West least-squares estimation (Stata/IC version 16.0; StataCorp). We compared linear trends for numbers of treatment courses with oral vancomycin and oral metronidazole in 3 time periods: (1) baseline (March 2006 through publication of PACT data in May 2014), (2) post-PACT (June 2014 through publication of revised IDSA/SHEA guidelines in February 2018), and (3) post–revised guidelines (March 2018 through August 2019). For vancomycin in the post-PACT period, 2 trend lines were calculated. The first line was for trend through March 2016, which was the month during this period with the peak number of treatment courses. The second line was for trend over the entire time period. The first line, which was steeper, was used for data analysis. We compared linear trends for fidaxomicin in 2 time periods: (1) pre–revised guidelines (January 2013 [when 3-month rolling data were first available] through February 2018) and (2) post–revised guidelines (March 2018 through August 2019).
Drug-pricing data were obtained on 28 January 2020 from Lexicomp, which links to Medi-Span Price Rx [10]. Websites for pricing of specific drugs are listed below:
RESULTS
Cumulative treatment courses of oral vancomycin and fidaxomicin increased by 54% (n = 226 166) and 48% (n = 18 518), respectively, during the 18 months immediately after publication of revised CDI guidelines compared with 18 months immediately before publication (Table 2). During the same periods, cumulative treatment courses of oral metronidazole decreased by 3% (n = 238 372). The combined increase in vancomycin and fidaxomicin use (n = 244 684) was almost perfectly offset by decreased metronidazole use. Cumulative treatment courses of vancomycin increased by 24% (n = 68 725) over the 18 months after and before publication of the PACT studies, while cumulative treatment courses of metronidazole were essentially unchanged (0.1% decrease; n = 7440).
Treatment Courses of Oral Vancomycin, Fidaxomicin, and Oral Metronidazole in the United States
| Time Period and Difference in Treatment Courses | Antibiotic Treatment Coursesa | ||
|---|---|---|---|
| Vancomycin | Fidaxomicin | Metronidazole | |
| March 2018–August 2019 (18 months post–revised guidelines), n | 648 834 | 57 502 | 8 373 688 |
| September 2016–February 2018 (18 months pre–revised guidelines), n | 422 668 | 38 984 | 8 612 060 |
| Difference, n | +226 166 | +18 518 | −238 372 |
| Difference, % | +54 | +48 | −3 |
| Time Period and Difference in Treatment Courses | Antibiotic Treatment Coursesa | ||
|---|---|---|---|
| Vancomycin | Fidaxomicin | Metronidazole | |
| March 2018–August 2019 (18 months post–revised guidelines), n | 648 834 | 57 502 | 8 373 688 |
| September 2016–February 2018 (18 months pre–revised guidelines), n | 422 668 | 38 984 | 8 612 060 |
| Difference, n | +226 166 | +18 518 | −238 372 |
| Difference, % | +54 | +48 | −3 |
aNumbers of 10-day treatment courses during the respective time periods, as calculated from IQVIA prescription data using regimens recommended in revised clinical practice guidelines.
Treatment Courses of Oral Vancomycin, Fidaxomicin, and Oral Metronidazole in the United States
| Time Period and Difference in Treatment Courses | Antibiotic Treatment Coursesa | ||
|---|---|---|---|
| Vancomycin | Fidaxomicin | Metronidazole | |
| March 2018–August 2019 (18 months post–revised guidelines), n | 648 834 | 57 502 | 8 373 688 |
| September 2016–February 2018 (18 months pre–revised guidelines), n | 422 668 | 38 984 | 8 612 060 |
| Difference, n | +226 166 | +18 518 | −238 372 |
| Difference, % | +54 | +48 | −3 |
| Time Period and Difference in Treatment Courses | Antibiotic Treatment Coursesa | ||
|---|---|---|---|
| Vancomycin | Fidaxomicin | Metronidazole | |
| March 2018–August 2019 (18 months post–revised guidelines), n | 648 834 | 57 502 | 8 373 688 |
| September 2016–February 2018 (18 months pre–revised guidelines), n | 422 668 | 38 984 | 8 612 060 |
| Difference, n | +226 166 | +18 518 | −238 372 |
| Difference, % | +54 | +48 | −3 |
aNumbers of 10-day treatment courses during the respective time periods, as calculated from IQVIA prescription data using regimens recommended in revised clinical practice guidelines.
Monthly vancomycin use decreased significantly throughout the baseline period (P = .03) before increasing significantly throughout both the post-PACT (P < .0001) and post–revised guideline (P < .0001) periods (Figure 1). Monthly fidaxomicin use decreased significantly from 2013 (when prescription data were first available) through the post-PACT period (P < .0001), before increasing significantly throughout the post–revised guideline period (P = .0002). Monthly metronidazole use contrasted with that of vancomycin, increasing significantly throughout the baseline period (P < .0001) and decreasing significantly throughout post-PACT (P = .0001) and post–revised guideline (P < .0001) periods.
Treatment courses of oral vancomycin (A), oral fidaxomicin (B), and metronidazole (C) in the United States. Monthly use of oral vancomycin and fidaxomicin increased significantly after publication of revised IDSA/SHEA clinical practice guidelines in February 2018 (second dashed vertical line; P < .0001 and P = .0002, respectively), whereas that of metronidazole decreased significantly (P < .0001). Monthly vancomycin use increased to a significantly greater extent after publication of revised guidelines than it did after publication of PACT data in May 2014 (first dashed vertical line; P < .0001, linear trend). In contrast, monthly metronidazole use decreased to a significantly greater extent after publication of revised guidelines than it did after publication of PACT data (P < .0001, linear trend). The number of treatment courses for each agent, expressed as 3-month rolling averages, appears on the y axis. Note that scales on the y axes differ by antibiotic. Years are shown on the x axis. Each data point (Actual) corresponds to a month within the respective year. Trend lines (Predicted) are shown for vancomycin and metronidazole in 3 time periods: baseline, post-PACT, post–revised guidelines. For vancomycin in the post-PACT period, 2 trend lines are shown: a steeper line for trend through the peak number of treatment courses (dashed line) and a less steep line for trend over the entire time period (solid line). The steeper line was used for data analysis. Trend lines (Predicted) for fidaxomicin are shown in 2 time periods: pre–revised guidelines, post–revised guidelines. P values in the figures are for the change in treatment courses for each drug during the various time periods. Abbreviations: IDSA/SHEA, Infectious Diseases Society of America/Society for Healthcare Epidemiology of America; PACT, Polymer Alternative for CDI Treatment.
Monthly vancomycin use increased to a significantly greater extent after publication of revised guidelines than it did after publication of PACT data (P < .0001, linear trend) (Figure 1). Conversely, monthly metronidazole use decreased to a significantly greater extent after publication of revised guidelines than it did after publication of PACT data (P < .0001, linear trend).
Discussion
Publication of updated IDSA/SHEA clinical practice guidelines for CDI was followed by immediate and highly significant increases in the use of oral vancomycin and fidaxomicin, and a significant decrease in the use of oral metronidazole in the United States. Indeed, changes in vancomycin and metronidazole use after publication of the revised guidelines were greater than those previously observed after publication of the PACT studies that demonstrated superiority of vancomycin over metronidazole. Since oral vancomycin and fidaxomicin are used exclusively against known or suspected CDI, it is reasonable to conclude that the guidelines were successful in changing treatment of the disease. It is notable that 8 years elapsed between iterations of the guidelines, and that data from the PACT studies and pivotal RCTs of vancomycin versus fidaxomicin in treatment of CDI were available 4 and 6–7 years before the revised guidelines, respectively [6–9]. Fidaxomicin was not included in the 2010 guidelines, which were published shortly before RCT data and FDA approval of the agent. Therefore, data used to revise CDI treatment recommendations were available for years before changes in practice were endorsed by IDSA and SHEA. Taken together, our findings support a new IDSA initiative to develop, disseminate, and adopt more timely guidelines and guidance for managing antimicrobial-resistant and other difficult-to-treat infections, as set forth in the society’s 2019 Strategic Plan [11–13].
Two explanations for the changes in use of CDI-active antibiotics are most plausible. The first explanation is behavioral: in response to the revised guidelines, clinicians and antimicrobial stewardship programs (ASPs) may be more inclined to prescribe or approve vancomycin or fidaxomicin in lieu of metronidazole. If correct, this scenario would suggest that clinicians and ASPs were aware of and familiar with the revised guidelines, agreed with their recommendations, expected them to improve patient care and outcomes, and were motivated to change practice (perhaps acknowledging that outcomes were suboptimal with previous approaches) [14]. Moreover, these behavioral motivators were strong enough to overcome significant pharmacy cost disadvantages of vancomycin and fidaxomicin compared with metronidazole. As of January 2020, average wholesale prices (AWPs) of 10-day treatment courses of Vancocin and generic vancomycin (125 mg QID) were $3775 and $1252, respectively; corresponding AWPs for 500-mg QID courses were $13 921 and $4617. Average wholesale prices of 10-day courses of fidaxomicin (200 mg BID) were $4639, whereas those of metronidazole (500 mg TID) were only $10–$21. Average wholesale prices generally overestimate true acquisition costs and do not capture heterogeneity in coverage between healthcare plans [15]. Nevertheless, generic oral vancomycin AWPs have not changed significantly since 2012, and prices of generic vancomycin capsules for most patients typically approach or surpass $1000 per treatment course [15]. If changes in practice were driven by the revised guidelines, it would call into question whether clinicians and ASPs were sufficiently proactive in incorporating clinical data from primary studies into their treatment strategies, or whether they were unduly reliant upon guidelines to define standards of care and justify the use of effective, but expensive antibiotics [11, 12].
A second, non–mutually exclusive explanation for changes in antibiotic use involves drug access: hospitals and health insurance plans may be more likely to include vancomycin and/or fidaxomicin on formulary and to offer the drugs as preferred formulary agents. Indeed, oral vancomycin costs have been a well-recognized barrier to optimal CDI treatment [15, 16]. Antibiotic costs are subsumed within bundled diagnosis-related group payments to hospitals by Medicare, which may create pressures to restrict inpatient use of more expensive agents [11, 12]. Prior to the revised guidelines, many healthcare plans did not cover oral vancomycin or fidaxomicin for outpatients, required prior authorization for these agents, and/or classified them as nonpreferred, specialty tier drugs that carried significant (and often prohibitive) co-pays [15, 16]. The US Department of Veterans Affairs (VA), the largest federal purchaser of prescription drugs, leaves decisions on oral vancomycin use to local facilities [17]. In contrast, fidaxomicin has been restricted to prior authorization since 2011 [17, 18]. Shortly after publication of the revised guidelines, we moved oral vancomycin from prior authorization to unrestricted use at the VA Pittsburgh Healthcare System. We enacted a similar policy for vancomycin 125 mg QID at the University of Pittsburgh Medical Center (UPMC), while retaining restrictions on other vancomycin regimens and fidaxomicin. UPMC also administers several insurance plans, in which oral vancomycin tier status generally has been lowered and fidaxomicin retains higher tier, specialty formulary status. We suspect that the unequivocal endorsement of vancomycin or fidaxomicin over metronidazole in the revised guidelines has led many hospitals and insurers to make similar changes in formulary and tier status, improving access to at least one of the recommended drugs.
In general, adherence to clinical practice guidelines across diseases and medical disciplines is variable [19]. Most studies have shown that implementation of guideline recommendations improves processes of medical care, although the size of improvement varies [19–22]. Effects of guideline adherence on patients’ outcomes are studied less extensively, and results are conflicting [19, 22]. Using data from 10 US geographic areas, CDC reported that 78%, 42%, and 2% of CDI cases in 2013–2015 were treated with metronidazole, vancomycin, and fidaxomicin, respectively [23]. Overall, antibiotic treatment was in accordance with 2010 IDSA/SHEA guidelines in 47% of cases. In other, single-center reports, adherence to 2010 treatment guidelines ranged from 38.5% to 54% [24–27]. In various studies, guideline compliance was associated with significant reductions in overall and attributable CDI mortality, disease recurrence, and/or length of stay compared with noncompliance [24, 26–29]. The extent to which adherence to guidelines reflects clinician and ASP preferences, type of healthcare coverage, and/or formulary status and costs of particular drugs is unclear from previous studies.
There are several limitations to this study. IQVIA data do not provide information on indication for prescribing a drug or on doses or durations of treatment in individual patients. Treatment courses, as determined here, are helpful for comparing use of agents, but do not necessarily correspond to numbers of patients treated for CDI with a given drug. Oral metronidazole, unlike oral vancomycin or fidaxomicin, is used extensively for a wide range of indications, and it is impossible to determine usage against CDI specifically. It is notable, however, that the combined increase in oral vancomycin and fidaxomicin treatment courses after publication of revised CDI guidelines was virtually identical to the decrease in oral metronidazole courses. IQVIA data do not quantitate use of vancomycin hydrochloride powder for oral suspension (Firvanq) or the oral use of compounded or intravenous vancomycin formulations, as administered instead of vancomycin capsules at some hospitals. Compounded or intravenous formulations administered orally are often not included in insurance formularies because they are not FDA approved [15]. Finally, our analyses do not account for changes in CDI incidence, which may confound data interpretation.
In conclusion, revised IDSA/SHEA clinical practice guidelines have had a significant impact on CDI treatment in the United States. During the 8 years between guidelines, outcome data from several RCTs were published and 2 new agents were approved by the FDA. More timely updates following these events may have led to earlier changes in CDI treatment. IDSA has identified real-time development and revision of guidelines or guidance documents in response to evolving data as a top priority and endorsed the possibility of employing electronic or other nontraditional formats [13]. A web-based process for rapid formulation and dissemination of evidence-based, expert-developed recommendations for hepatitis C management by IDSA and the American Association for the Study of Liver Diseases could serve as a model [30]. Turnaround times for guidance documents may be shortened by a narrow focus on treatment rather than comprehensive disease management or by using methods other than a rigorous GRADE (Grading of Recommendations Assessment, Development, and Evaluation) framework, while still relying upon systematic, critical literature review [19, 31]. Even if plans are made to develop more timely guidance for a given disease, clinicians, ASPs, and healthcare plans should not delay in utilizing therapeutics that are shown in clinical trials to be more effective than previous options. Data from several treatment and cost-effective analyses may be needed to define the precise roles of new agents, as has been suggested to be the case for bezlotoxumab [32]. Studies are warranted to determine the impact of revised CDI guidelines on outcomes, cost-effectiveness, and healthcare resource utilization, and to better understand behavioral, drug access, healthcare coverage, and economic issues that limit prescribing of useful antimicrobials in general [12].
Notes
Acknowledgments. The authors thank Alan Carr for his assistance with obtaining and interpreting IQVIA data.
Potential conflicts of interest. C. J. C. has been awarded investigator-initiated research grants from Astellas, Merck, Melinta, and Cidara for studies unrelated to this project; served on advisory boards or consulted for Astellas, Merck, the Medicines Company, Cidara, Scynexis, Shionogi, Qpex, and Needham & Company; and spoken at symposia sponsored by Merck and T2Biosystems. M. H. N. has been awarded investigator-initiated research grants from Astellas, Merck, Melinta, and Cidara for projects unrelated to this study, and served on advisory boards for Astellas, Cidara, Merck, the Medicines Company, Scynexis, and Shionogi. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
