November and December of 2022 saw growth in rates of flu, RSV, and COVID, that renewed attention to a seasonal trend of increased respiratory illnesses in the winter months. Data from Medicare, Medicaid, and private insurance illustrate the seasonal ups and downs of respiratory infections. Earlier HCCI research showed that hospitalizations from respiratory illness increased in the winter months (November to March) and declined over the summer (June to September). Prescriptions of antibiotics and similar medications also demonstrate a cycle of increased respiratory illness in the winter and decreases in the summer months.
Unsurprisingly, the onset of the COVID-19 pandemic appears to have disrupted this longstanding cyclical trend. HCCI's earlier work showed that COVID hospitalizations dwarfed admissions for other respiratory infections from April through December of 2020. Pneumonia hospitalization rates dropped by more than half, from about 40 per 100,000 in April 2019 to 16 per 100,000 in April 2020, with admission rates remaining in the low teens through the end of the year. This analysis of antibiotic prescriptions illustrates a similar pattern of the seasonal cycle of respiratory illness being disrupted in 2020.
We used HCCI's unique employer sponsored insurance claims dataset, Medicare Part D claims, and Medicaid TMSIS files to examine use of four categories of antibiotic prescription drugs:
Use of Antibiotics was Seasonal with Higher Rates in the Winter
As shown in Figure 1, these four categories of antibiotics showed a distinct seasonal pattern with rates of use that were higher during the winter months and lower in the summer months. The monthly rates shown in Figure 1 for 2018 and 2019 are similar to those we observe in earlier years as well. Penicillin and macrolides, which were the most commonly used antibiotics in our data, had more dramatic increases and decreases than the other antibiotic categories we studied. Among people with employer sponsored insurance (ESI), for example, in 2019, we observed the highest use of penicillin in December with a rate of 2,603 per 100,000 enrollees. That was 70% higher than July - the month with the lowest rate of penicillin prescription fills that year in (1,533 per 100,000 enrollees). ABRT prescriptions followed a similar pattern to penicillin and macrolides.
Although tetracycline had some similar seasonal variation, rates of those prescriptions were more consistent throughout the year, perhaps reflecting the different set of conditions it can be prescribed to treat (e.g., acne). In December 2019, there were 704 per 100,000 tetracycline prescription fills up 20% from 582 per 100,000 in July among those with ESI (Medicaid tetracycline use was 9% higher in December than July from 351 to 381 per 100,000). People with Medicare Part D plans (Medicare Advantage and Traditional Medicare) had higher utilization rates of tetracycline overall and a slightly more distinct seasonal pattern with rates 24 to 28% higher in December 2019 than July of that same year.
Antibiotic Use Dropped Precipitously at the Beginning of the COVID-19 Pandemic
Our data show a significant drop in antibiotic use in the spring of 2020 that coincides with the early months of the COVID-19 pandemic. For example, Figure 2 shows that, among ESI enrollees, penicillin use in April 2020, a rate of 913 per 100,000, was approximately 40 percent lower than 2019's low point (1,533 per 100,000 in July). We saw a similar trend in macrolide, tetracycline and ABRT prescriptions among those with ESI, with rates in April of 2020 that were 47 percent, 19 percent, and 41 percent lower than July 2019, respectively (Figure 2).
We observe a similar pattern among people enrolled in Medicaid. Rates of penicillin use in April 2020 were 44 percent lower than July 2019. Among Medicaid enrollees, rates of macrolide prescriptions were 39 percent lower, rates of tetracycline use were 24 percent, and rates of ABRT use were 40 percent lower in April 2020 when compared to the prior year's July rates. People enrolled in Medicare drug plans had lower utilization rates as well, but the difference between April 2020 and July 2019 was smaller for penicillin, macrolides, and ABRT. Changes in tetracycline rates for those enrolled in Medicare drug plans were aligned with changes observed in ESI and Medicaid populations.
The change that led to the decrease in prescribed antibiotics in early 2020 also seems to have shifted the seasonal cycle of prescribing patterns in that year. Even though rates of antibiotic use slowly increased from July 2020 through the end of the year, the utilization rates in December 2020 were substantially lower than in December 2019. For example, penicillin, macrolides, and ABRT utilization rates in December 2020 were less than half of the December 2019 rates for ESI and Medicaid enrollees. Prescriptions of tetracycline among people with Medicaid and ESI also were lower in December 2020 than in 2019, though the decline was not as dramatic – only 25 to 32 percent lower in December 2020 than in December 2019 across payers (Figure 3).
Antibiotic Use among Children Dropped the Most During the COVID-19 Pandemic
The declines in antibiotic use in 2020 were even more dramatic among children. Compared to December 2019, rates of penicillin, macrolide, and ABRT use in December of 2020 among children aged 0-17 enrolled in ESI and Medicaid were even lower than the general population. While penicillin use among the overall ESI population decreased by 54 percent, it decreased by 79 percent among children. Among Medicaid enrollees, penicillin use among all age groups in December 2020 was 62 percent lower than December 2019. Penicillin use decreased 74 percent among children 0-17 enrolled in Medicaid. Figure 4 shows a similar pattern among macrolides and ABRT. Interestingly, tetracycline showed the opposite pattern, with use among children declining less than among ESI and Medicaid enrollees of all ages.
Annual Antibiotic Use has Declined Slightly in Years Leading up to the Pandemic
This analysis, looking at monthly prescribing rates of antibiotics, also can inform efforts to reduce unnecessary antibiotic prescriptions, which can contribute to increasing antibiotic resistance. The Centers for Disease Control and Prevention (CDC) has reported that at least 28 percent of outpatient antibiotic prescriptions are unnecessary and has implemented surveillance measures and reporting on prescribing patterns. Our data show that rates of antibiotic prescriptions have decreased slightly from 2018 through 2020. Antibiotics are part of a larger group of drugs – antimicrobials – that also include antivirals, anti-infectives, and antifungals. These drugs can treat simple ear infections, sexually transmitted infections, ringworm and even cure Hepatitis C.
The pattern we observe in antibiotic prescribing – a slight decrease over time – is unique among antimicrobials. In contrast, antiviral prescription rates have mostly increased over the past five years, with a slight decrease in recent years. Anti-infective and antifungal rates have remained relatively constant over the same period.
Data Show Drops in Prescribing for Non-COVID Respiratory Illness in 2020, but Seasonal Patterns May Soon Return
The beginning of the COVID-19 pandemic coincided with a significant decrease in antibiotic use, especially among children and teens. Historically low rates of several pathogens in 2020, in particular respiratory syncytial virus (RSV) and seasonal influenza, have been well documented. Other studies suggest that non-pharmaceutical interventions, such as wearing masks and maintaining physical distance, reduced transmission of infections in this time period. Accordingly, the prescribing patterns we observe in 2020 are consistent with lower incidence of respiratory infection over the study period.
At the same time, there is longstanding evidence that suggests a seasonal pattern of respiratory illness with rates of infection highest in the winter and lowest in the summer. In the years since 2020, this cycle may have returned such that a similar analysis of 2021 and 2022 data would reveal a return to pre-pandemic antibiotic prescription rate patterns.
To put these findings into context, we conducted similar analyses on all antimicrobials in our data. Results from all the antimicrobials studied can be found in the Health Care Vitals Dashboard on HCCI's website.
Observing antimicrobial use patterns can help public and private decisionmakers understand treatment and burden of infectious diseases over time among different populations. Tracking utilization can help during periods of cyclical ups and downs or during a global pandemic.
We identified prescription pharmacy drug claims from 2018 to 2020 using the RxNorm category for antimicrobials, as defined by the VA formulary, for the population with pharmacy coverage during each month. Antimicrobials were classified into several broad categories, including antibiotics (antibiotic resistant treatments, cephalosporin, macrolides, lincosamides, nitrofuran, penicillin, quinolone, sulfonamide, tetracycline), antifungals (e.g. fluconazole, terbinafine, nystatin, etc.), antivirals (e.g. valacyclovir, oseltamivir, acyclovir, etc.), and other anti-infectives (e.g. methenamine salts, metronidazole, rifaximin, vancomycin, etc.).
Rates were calculated as total number of unique fills (defined by NDC and fill date) divided by the total population with pharmacy coverage per 100,000 enrollees. All enrollees with known age and sex were included in the denominator. The ESI and Medicaid populations were limited to those under the age of 65, the Traditional Medicare and Medicare Advantage populations were limited to enrollees over the age of 18.