February 28, 2011

Antibiotic Prescribing and Use

CHRT Issue Brief February 2011

The Institute of Medicine and the Centers for Disease Control and Prevention (CDC) have identified antibiotic resistance as a major public health problem in our country.1 One of the biggest contributors to antibiotic resistance is the inappropriate use of antibiotics.

Of particular concern is the prescribing of antibiotics for infections that are usually viral in nature—colds, sore throats, and most upper respiratory infections. These illnesses do not respond to antibiotics and nearly always get better on their own. Since 1995 the CDC has run an outreach campaign to encourage appropriate use of antibiotics, and the overall rate of antibiotic prescribing decreased 23 percent from 1992 to 2000. Nevertheless, a high rate of inappropriate antibiotic use—a rate that continues to put population health at risk—persists. This issue brief examines antibiotic prescribing trends in Michigan in recent years.

1 Mark S. Smolinski, Margaret A. Hamburg, and Joshua Lederberg, Editors, Committee on Emerging Microbial Threats to Health in the 21st Century. (2003). Microbial Threats to Health: Emergence, Detection, and Response. Washington, D.C.: The National Academies Press.

Trends in Antibiotic Use

Two million antibiotic prescriptions were billed to Blue Cross Blue Shield of Michigan (BCBSM) in 2009; more were likely dispensed that cost less than the co-payment. The BCBSM data show that members are prescribed just under one (.9) antibiotic per year. This is comparable to the national average of .88 prescriptions per person.2

Overall antibiotic prescribing for adult BCBSM members decreased 9.3 percent between 2007 and 2009, from .96 prescriptions per member to .87 prescriptions per member.

Many of the CDC-initiated outreach activities aimed at reducing inappropriate antibiotic use have been focused on pediatricians.3 Despite these efforts, antibiotic prescribing rates for BCBSM children increased 4.5 percent between 2007 and 2009. The overall increase in antibiotic use combined with data presented later in this issue brief indicates that there continues to be too high a rate of inappropriate use.

2 Steinman, M.A., Yang, K.Y., Byron, S.C., Maselli, J.H., & Gonzales, R. (2009). Variation in outpatient antibiotic prescribing in the United States. The American Journal of Managed Care, 15(12), 861–868.

3 Grijalva, C.G., Nuorti, J.P., & Griffin, M.R. (2009). Antibiotic prescription rates for acute respiratory tract infections in US ambulatory settings. JAMA: The Journal of the American Medical Association, 302(7), 758–766. doi:10.1001/jama.2009.1163

Types of Antibiotics Prescribed

The National Committee for Quality Assurance (NCQA) has identified a list of “antibiotics of concern.” These antibiotics are called “broad spectrum” antibiotics, meaning they are effective against a wide range of bacteria. While these drugs can be very valuable in treating infections that have not responded to other agents or were caused by multiple types of bacteria, in many cases an infection can be more precisely targeted with a narrow spectrum antibiotic. Since the use of any antibiotic agent can promote resistance to that agent, broad spectrum antibiotics should be reserved for the most serious infections. Overuse of these agents is of concern because it can cause increased resistance to them, leaving even fewer drugs available to treat serious infections.4

In 2009, 49.1 percent of antibiotic prescriptions in the overall BCBSM population were for these broad spectrum antibiotics. This compares to a national rate of 47 percent.2,5 Between 2007 and 2009, the percent of antibiotic prescriptions for antibiotics of concern stayed nearly steady in the adult BCBSM population, decreasing 0.4 percent during that time period. In the same time period, the percentage of antibiotics of concern prescribed to BCBSM children increased 3.4 percent, going from 44.9 percent to 46.4 percent. One possible explanation for this increase in children is the continuing rise in resistant pathogens that are responsible for ear infections in children, necessitating broad spectrum antibiotic treatment.6 Other possible reasons are that children get different infections than adults, and that some drugs that are used in adults are not used for pediatric patients.

4 Tenover, F.C., & Hughes, J.M. (1996). The Challenges of Emerging Infectious Diseases. JAMA: The Journal of the American Medical Association, 275(4), 300&nsahs;304. doi:10.1001/jama.1996.03530280052036

5 National rates are from a study of 229 managed care plans. BCBSM data is from PPO plans.

6 Naseri, I., Jerris, R.C., & Sobol, S.E. (2009). Nationwide Trends in Pediatric Staphylococcus aureus Head and Neck Infections. Arch Otolaryngol Head Neck Surg, 135(1), 14&nsash;16. doi:10.1001/ archoto.2008.511

When antibiotic prescribing is indicated, determining which antibiotic to prescribe is influenced by several factors. The most important factor should be which antibiotic is most effective against the bacteria likely to be responsible for a given infection. However, factors such as physician and patient preference, convenience, and familiarity with specific drugs often play a role.

Figure 4 shows the top antibiotics prescribed for BCBSM adults in 2009, highlighting those categorized by the NCQA as “antibiotics of concern.” Five of the top ten most common antibiotics are classified as “antibiotics of concern.” Azithromycin (also known as Zithromax or Z-Pak) is the most frequently prescribed antibiotic for adults, making up 20 percent of all antibiotic prescriptions.

The most commonly prescribed antibiotic for children in 2009 was Amoxicillin. Only three of the top ten antibiotics are categorized as “antibiotics of concern.”

Geographic Variation in Antibiotic Prescribing7

In general, we know antibiotic prescribing varies widely across regions, health plans, and even individual physicians.2 In Michigan, antibiotic prescribing rates among the BCBSM population vary across hospital service areas (HSA). These differences cannot be explained by different rates of illness alone.8

In the adult population in 2009, the highest overall antibiotic prescribing rate in Michigan was 2.2 times the lowest antibiotic prescription rate. Eleven HSAs had prescribing rates that were at least 30 percent higher than the state average: Cheboygan, Gaylord, Grosse Point, Hillsdale, L’anse, Ludington, Marlette, Pigeon, Sturgis, Watervliet, and West Branch. Five HSAs had prescribing rates 25 percent or less than the state average: Ann Arbor, Chelsea, Greenville, Saline, and Southfield.

Prescribing for the pediatric population showed much higher variation than prescribing for the adult population. In 2009, the rate of antibiotic prescribing for BCBSM children varied by a factor of 4.6, from .54 antibiotics per member in Ann Arbor to 2.2 antibiotics per member in Dowagiac. Thirty-six HSAs had rates more than 30 percent above the state average, while nine had rates at least 25 percent below the state average. Eight of the nine low utilizing HSAs were in Southeast Michigan. High utilizing HSAs were clustered in the Upper Peninsula, in Northern Lower Michigan, and on the Ohio border.

Southeast Michigan has a lower rate of antibiotic use in children and more pediatricians per 1000 population than outstate Michigan. Though there is not a statistically significant association between lower antibiotic use rates and higher numbers of pediatricians per 1000 population at the HSA level, it may account, at least in part, for differences that can be seen between southeast Michigan and the rest of the state.

7 While there is likely an undercounting of antibiotic prescriptions due to prescriptions that cost less than the co-payment and prescriptions filled under $4 prescription programs, the results should not be substantially affected. These results are consistent with health plan results from 2005, before $4 prescription programs had a substantial effect on the market. Prescriptions are also not counted if they are written by a physician but never filled by the patient.

8 Butler, C.C., Hood, K., Verheij, T., Little, P., Melbye, H., Nuttall, J., Kelly, M.J., et al. (2009). Variation in antibiotic prescribing and its impact on recovery in patients with acute cough in primary care: prospective study in 13 countries. BMJ, 338(jun23 2), b2242–b2242. doi:10.1136/bmj.b2242

Prescribing Practices and Variation

Prescribing antibiotics for viral infections, which do not respond to antibiotics and nearly always get better on their own, is a driver of overall antibiotic prescribing rates. One study found that antibiotic prescribing for colds, upper respiratory infections, and bronchitis accounts for nearly one third of antibiotics prescribed in the outpatient setting.9

To quantify the extent to which inappropriate antibiotic prescribing drives variation in overall prescribing rates, we examined the association between measures of appropriate prescribing and overall prescription rates. Our measures of appropriate prescribing are based on measures used in the Healthcare Effectiveness Data and Information Set (HEDIS), a tool used by health plans to measure the effectiveness of healthcare.10

The first measure is the percentage of children with upper respiratory infections (URI) that are given antibiotics. Upper respiratory infections involve the nose, sinuses, throat, and trachea, with the most common example being the common cold. Overall in 2009, 21.9 percent of BCBSM covered children diagnosed with URI were prescribed antibiotics. This percentage excludes children who were prescribed antibiotics for reasons other than the URI, such as ear infections or strep throat. This rate is a decrease from the 44 percent of children who received antibiotics for URIs found in a national study in 1992, but there is still substantial room for improvement, since antibiotics are not found to improve outcomes in URIs in children.

There is a correlation between the percentage of children given antibiotics for URIs and the overall pediatric antibiotic prescribing rate in a geographic region. Figure 9 shows the relationship between those two measures. Regions with a higher overall antibiotic use rate for children also had a higher percentage of antibiotics used for conditions with a viral diagnosis. Of the variation in overall pediatric antibiotic prescribing rates, 31.4 percent is related to variation in prescribing rates for URI. In other words, it appears that high use regions of the state have a higher use, at least in part, due to a higher inappropriate use of antibiotics.

Another measure of appropriateness in antibiotic prescribing for children is whether children who received antibiotics for pharyngitis (sore throat) also received a test to confirm the infection was bacterial. Overall, 56.3 percent of children received the appropriate test when given antibiotics for a sore throat. Appropriate testing for pharangytis is related to overall antibiotic prescribing. That is, places with high rates of testing to confirm bacterial infection also had lower rates of antibiotic prescribing. This relationship is shown in Figure 10. Of the overall variation in pediatric antibiotic prescribing, 12.1 percent is related to testing patterns for pharyngitis. Regions of the state with higher use of antibiotics were also regions with lower rates of testing for the presence of bacterial infection.

For adults, a measure of appropriateness was the percentage of adults diagnosed with bronchitis who were not given antibiotics. Like URI, most cases of bronchitis are viral and improve with appropriate non-antibiotic treatment. Prescribing rates for BCBSM adults with bronchitis in 2009 were very high—77.2 percent received a prescription for antibiotics. Although rates were high throughout the state, there was a statistically significant correlation between the rates of prescribing antibiotics for bronchitis and those regions with high overall use rates. That is, regions with the highest antibiotic dispensing for bronchitis also had the highest antibiotic dispensing overall. Of the variation in overall use rates, 8.8 percent is related to prescribing antibiotics for bronchitis (Figure 11).

9 Gonzales, R., Steiner, J.F., & Sande, M.A. (1997). Antibiotic Prescribing for Adults With Colds, Upper Respiratory Tract Infections, and Bronchitis by Ambulatory Care Physicians. JAMA: The Journal of the American Medical Association, 278 (11), 901 -904. doi:10.1001/ jama.1997.03550110039033

10 HEDIS measures the percentage of children with URI and percentage of adults with bronchitis not given antibiotics. Our measures report the inverse—the percentage of with those conditions who are given antibiotics.

Conclusion

Efforts to educate providers and patients, especially for pediatric providers and patients, have resulted in overall decreased use of antibiotics for viral infections. High rates of inappropriate antibiotic prescribing occur in some parts of Michigan, particularly outside of southeast Michigan. Of particular concern are the lack of testing for presence of a bacterial infection in children prior to prescribing an antibiotic, the continuing high use rates of antibiotics for viral infections in both children and adults, and the relatively high use of broad spectrum antibiotics among the adult population.

A number of efforts in Michigan have focused on reducing the inappropriate use of antibiotics. Most notably, the Michigan Antibiotic Resistance Reduction Coalition has had a number of successes in reducing inappropriate use over time. Overuse of antibiotics continues to be a major health problem in Michigan and nationally. Further efforts to reduce inappropriate use are essential for safeguarding our health and preserving the effectiveness of antibiotics.