Is Pathogen Molecular Testing Reshaping Outpatient Antibiotic Prescribing?

Timely microbiological diagnosis is necessary to optimize treatments, improve antimicrobial stewardship, and improve patient and population health outcomes. For example, in a recent randomized clinical trial of hospitalized patients with community-acquired pneumonia, the use of syndromic polymerase chain reaction (PCR)-based diagnostic test panels led to faster and more targeted microbial treatment.1 Similarly, other hospital-based studies have demonstrated the impact of rapid microbiological diagnosis on faster treatment optimization, improved antimicrobial stewardship, decreased length of hospitalization, and enhanced infectious disease surveillance and infection control practices.2 Yet, in the outpatient setting where 80%–90% of antibiotics are prescribed3,4 and per the United States Centers for Disease Control and Prevention, at least 28% of the 236.4 million antibiotic prescriptions dispensed from community pharmacies in 2022 were considered unnecessary,5 the impact of faster microbiological diagnoses on prescribing practices is unknown. In outpatient settings, patients with suspected infection typically receive broad empiric antimicrobial therapy rather than directed therapy based on the identification of the infecting pathogen.2 Empiric antibiotic prescribing often occurs because of the inability to rapidly identify the pathogen causing an infection.2 Studies indicate that more than 50% of outpatients with acute upper respiratory infections receive antibiotics, even though viruses are typically responsible.6–8 This overprescribing contributes to unnecessary drug exposure for patients and escalates antimicrobial resistance, a top global health threat with significant economic costs.9 While past studies suggest reluctance of clinicians to modify initial empiric antibiotic prescribing patterns despite subsequent availability of antimicrobial susceptibility reports,10 this attitude may be shifting, driven in part by innovations that enable results to be available faster. Improvements in panel and platform design, specimen collection and transport, and result reporting interpretation and turnaround time render diagnostic results that are more rapidly actionable for outpatients. This in turn offers the potential to disrupt empiric antibiotic prescribing. With the aim of understanding, if the availability of such faster testing is being used to guide prescribing decisions, a survey was administered to outpatient health care providers with routine access to syndromic PCR-based tests that provide next-day results in the United States (HealthTrackRx, Denton, Texas). Multiple syndromic panels were available including for respiratory, urinary, gastrointestinal, genitourinary, wound, and other infections. Each panel included unique pathogen targets according to the most common causes for the specific infectious syndrome in the United States. Providers used their discretion on medical necessity, but received information on (1) syndromic panel selection based on relevant patient symptoms, (2) relevance of the organisms on each panel, (3) sample collection, and (4) results interpretation. The providers’ medical specialty, level of training, and familiarity with the laboratory testing platform/service (extremely, very, somewhat, or not familiar) were recorded. Among physicians (Medical Doctors [MD]/Doctors of Osteopathic Medicine [DO]) and nurse practitioners (NP)/physician assistants (PA) very or extremely familiar with the testing and involved with interpreting test reports/providing treatment recommendations, a proportional odds model was used to examine how specialty (primary care [including primary care, pediatric, women’s health, or internal medicine] versus urgent care) and level of training (MD/DO versus NP/PA) influence the impact of the testing on their antibiotic prescribing. Seven hundred and ten providers were surveyed (N = 710). Among 384 (94 MD/DO, 290 NP/PA) respondents, 346 (90.1%) were extremely or very familiar with the testing platform of which 325 (93.9%) were involved with interpreting test reports and providing treatment recommendations. Turnaround time and accuracy were test characteristics ranked by providers as most valuable to their practice. Of the 325 respondents involved with interpreting test reports and providing treatment recommendations, 319 reported data on prescribing decisions. Of these, most (97.5%) use test results to make prescribing decisions with equal numbers prescribing upfront and then adjusting treatment later if needed (ie, empirically prescribing) (48.6%) versus awaiting test results before prescribing antibiotics (48.9%). More NP/PA providers compared with MD/DO providers (52.1% versus 39.0%, respectively, P = 0.0451) and those in primary compare versus urgent care practices (61.2% versus 39.1%; P = 0.0002) reported withholding antibiotic prescriptions pending test results (Table 1). When asked if the provider noticed any change in antibiotic prescribing after incorporating the PCR test in their decision-making, 92.1% confirmed changes, with 9.4% reporting more antibiotic prescribing, 25% noting fewer antibiotic prescriptions, and 57.6% noting improved self-reported accuracy of antibiotic prescriptions. Logistic regression analysis indicated that NP/PAs compared with MD/DOs (odds ratio: 2.27, 95% confidence interval: 1.26–4.09; P = 0.0059) and those in primary care compared with urgent care practices (odds ratio: 2.83, 95% confidence interval 1.71–4.67; P < 0.0001) were more than twice as likely to withhold antibiotic prescriptions pending test results (Table 1). Table 1. - Provider-reported Use of Rapid PCR Test Results to Make Antibiotic Prescribing Decisions. MD/DO (n = 94) NP/PA (n = 290) P value Primary care (n = 153) Urgent care (n = 187) P value Familiar with the test 89 (94.7%) 257 (88.6%) 0.0873 153 (93.5%) 164 (87.7%) 0.0741 Interpret results and provide treatment recommendations 80/89 (89.9%) 245/257 (95.3%) 0.0638 135/153 (88.2%) 160/164 (97.6%) 0.0011 Withhold antibiotic prescriptions pending test results 30/77 (39.0%) 126/242 (52.1%) 0.0451 82/134 (61.2%) 61/156 (39.1%) 0.0002 Adjusted odds ratio of diagnostic versus empiric prescribinga OR = 1.0 OR = 2.2795% CI: 1.26–4.09 0.0059 OR = 2.8395% CI: 1.71–4.67 OR = 1.0 <0.0001 aAdjusted for specialty and care (primary versus urgent) facility.Abbreviations: CI, confidence interval; MD/DO, Medical Doctors/Doctors of Osteopathic Medicine; NP, nurse practitioners; OR, odds ratio; PA, physician assistants. These results suggest that, in comparison with historical practices,10 outpatient providers are now using newly available PCR test panels to guide antibiotic prescribing, with almost half awaiting test results to determine if antibiotics are appropriate. This change in prescribing behavior may lead to more accurate prescribing as well as fewer empiric antibiotic prescriptions. Primary care providers were more likely than urgent care specialists to await test results to prescribe. Empiric prescribing decisions depend on several factors, including how sick the patient is, underlying comorbidities and risk of progression, how quickly the result will return, patient expectations/preferences, and the ability to reach the patient for timely follow-up. Primary care providers may have infrastructure in place and long-term relationships with patients to enable timely follow-up. Further, patients visiting urgent care practices may be sicker, requiring antibiotics upfront. The willingness of NP/PAs to forgo empiric prescribing may also reflect patient acuity. Understanding why some practitioners continued to prescribe empirically despite the availability of rapid and clinically actionable PCR results was beyond the scope of this survey. Additionally, while this survey targeted respondents familiar with the testing platform, limiting the generalizability of the results, the findings suggest that access to next-day test results offers the opportunity to combat overprescribing. To fully appreciate the potential impact of such testing, the cost of multiplex PCR tests must also be considered. Multiplex PCR panels cost between $170 and $643/test, depending on the panel and setting of care,11,12 with patient out of pocket costs typically $150 or less. Prospective, randomized trials in the outpatient setting that take into account the costs of the testing are needed to confirm the impact of rapidly available test results on antibiotic prescribing and, ultimately, to understand how such changes influence patient outcomes and antimicrobial resistance reduction in the long-term. Confirming the cost-effectiveness of the testing would further inform and influence payer reimbursement decisions. In summary, the introduction of rapid, clinically actionable molecular diagnostics in outpatient settings—where most infections are diagnosed and unnecessary antibiotics are prescribed—provides a vital tool for targeted therapy and reducing empiric overprescribing. With the capability to identify the infecting pathogen within 24 hours, outpatient providers are increasingly using these test results to guide antibiotic administration, with many waiting for results before prescribing. Based on users’ perceptions, these tests also appear to be guiding antibiotic selection. Indeed, the availability of rapid molecular diagnostics may be shifting practices away from empiric antibiotic prescribing in outpatient settings. Acknowledgments The authors thank Lindsay Phillips for assistance in data collection and Ben Favret for assistance in preparing the manuscript. Conflicts of Interest Dr Alexander: Consultant: Scynexis, GSK, Astellas, Merck, HealthTrackRx, Basilea; Research Grant to My Institution: Karius; Clinical Trials (site/study Principal Investigator): Scynexis, F2G; Royalties (Chapter Author): UpToDate. Dr Irish: Consultant: HealthTrackRx, Dr Rosato: NCO, Dr Fragala: Consultant: HealthTrackRx; Dr Goldberg: HealthTrackRx; Dr Eisenstein: CARB-XD. The other author has no conflicts of interest to disclose. Funding This research was funded by HealthTrackRx (Denton, Texas). Author Contributions Drs Alexander, Irish, Fragala, and Goldberg: data analysis and manuscript preparation; Drs Rosato, Eisenstein, and Nash: manuscript preparation.