Discussion
Many underlying factors influence physicians' prescribing behavior, including (1) formal medical education and training, which lack courses in rational therapeutic decision-making that emphasize cost-effectiveness comparisons and the necessity for critically analyzing promotional literature, (2) time constraints of a busy practice, which limit thoughtful therapeutic decision-making, (3) fear of malpractice litigation, (4) inability to keep abreast of advances in medicine, (5) lack of knowledge about costs, (6) misinterpretation of test results or even which test to order in a given situation, (7) inexperience, (8) overreliance on clinical experience as a sole guide for practice, (9) patient demand for treatment even when it is unnecessary, (10) influence of opinion leaders or powerful authority figures who set practice patterns for their more junior colleagues sometimes without the benefit of the best or most recent data, (11) financial incentives or insulation from cost considerations because of third-party coverage, and (12) advertising campaigns and detailing programs of sales representatives that effectively communicate, but sometimes inflate, claims that encourage use of their products.[20,24,33,34] Pharmaceutical companies have been very successful at influencing physicians' prescribing patterns by applying principles of communications and behavioral science theories. Although physicians claim they are not heavily influenced by drug advertisements or sales representatives, their beliefs concerning drug effectiveness have been shown to be congruent with messages received through commercial rather than scientific channels.[35] In addition, a temporal relationship between industry's enticements and prescribing patterns has been demonstrated.[36]
Educational intervention programs are tools for improving the quality, safety, effectiveness, and cost-benefit of patient care. The overriding principles are to provide physicians with accurate, unbiased, up-to-date, evidence-based information and encourage appropriate utilization of resources (e.g., drugs, laboratory tests, radiological procedures). Some of the best proven methods of influencing physicians' clinical decision-making include (1) conducting interviews to determine baseline knowledge and motivations for current practice patterns, (2) defining clear educational and behavioral change goals, (3) establishing credibility through a respected organizational identity, referencing authoritative and unbiased sources of information, and presenting both sides of a controversial issue, (4) focusing programs on specific physicians based on prescribing profiles that show heavy use of targeted drugs, tests, or procedures, (5) having face-to-face interactions with prescribing physicians and their opinion leaders, (6) providing educational sessions that stimulate active participation in an informal atmosphere, (7) offering an alternative to the practice being discouraged (e.g., a more rational medication choice or a nonpharmacologic therapy), (8) using visually appealing print materials that emphasize a few key points with graphics and headlines, (9) highlighting and repeating the essential messages, and (10) providing positive reinforcement of improved practices in follow-up visits.[33]
In our case, a preliminary study was not conducted to gain an understanding of why physicians routinely prescribed ribavirin and antimicrobials for infants and young children hospitalized with RSV-related lower-respiratory-tract infection. However, the investigators' conversations with attending physicians and residents revealed that the reasons were multiple and included a lack of knowledge of the 1993 AAP guidelines or interpretation of them as requiring the use of ribavirin, the misconception that concurrent serious bacterial infections were common with RSV-related lower-respiratory-tract infections, the belief that RSV infections could not cause infiltrates to appear on chest x-rays or a mildly elevated white blood cell count with a left shift, a fear of malpractice litigation due to a missed case of sepsis or meningitis, and the belief that administering antimicrobials for a few days could do no harm. In addition, based on the high prescribing rate, it appeared that the pharmaceutical company sales representatives' detailing program had probably influenced physicians' prescribing patterns for ribavirin.
As a result of these informal findings, the major themes of the educational messages were as follows: (1) The physician should prescribe ribavirin only for those infants and children who meet at least one of the criteria for use as recommended in the AAP guidelines, (2) he or she should initiate ribavirin therapy when indicated as early as possible to maximize effectiveness, (3) concurrent serious bacterial infections are very rare in infants and young children hospitalized with RSV lower-respiratory-tract infection, (4) RSV infection can cause fever, infiltrates on chest xrays, and a mildly elevated white blood cell count with a left shift, and (5) treating RSV lower-respiratory-tract infection with broad-spectrum i.v. antimicrobials actually increases the chance of bacterial superinfection. Educational presentations and discussions focused on clinical, laboratory, and radiological findings of RSV lower-respiratory-tract infection, disease pathophysiology, clinical course, and the clinical pharmacology and therapeutic effectiveness of various treatment options. Information presented was based on critical review of the literature, inlcuded pros and cons, and focused on appropriate disease management rather than cost containment. Additionally, the investigators argued that routine full sepsis or meningitis workups at admission for infants and young children with typical signs and symptoms of RSV bronchiolitis are unnecessary and add to the cost, discomfort, and stress of the hospitalization. However, physicians were encouraged to consider laboratory testing for bacterial infections in infants appearing severely ill with an atypical presentation or clinical course because of the small risk of a concurrent serious bacterial infection.
Our multifaceted educational intervention program was somewhat successful at improving physicians' prescribing of ribavirin and broad-spectrum i.v. antimicrobials in infants and young children hospitalized for treatment of RSV lower-respiratory-tract infections. Excessive use of ribavirin and antimicrobials significantly decreased after the educational interventions, resulting in significant cost savings for drug therapy. Although this study did not examine the effects on overall health care costs, the cost of ribavirin therapy ($1100 per day for just the drug) is a major portion of the overall cost of hospitalization for patients admitted to a general pediatrics ward with RSV lower-respiratory-tract infection. The decreased use of ribavirin and antimicrobials did not appear to prolong the hospital length of stay or increase the readmission rate, which remained at 0%. In addition, antimicrobials were still frequently prescribed in patients without an indication for use, although less so than before.
We did not want to restrict ribavirin use or require justification and preapproval for use. We believed education could work and would be better received than placing a clinical pharmacist or infectious diseases physician in the role of "the medication police." Many factors probably contributed to the limited success of our multifaceted educational intervention program and helped achieve changes in both physicians' knowledge and behavior. A number of the strategies discussed previously were employed. Several different channels of communication provided multiple and repeated reinforcements of the desired prescribing practices. Even though our program did not include face-to-face personal educational visits (academic detailing), the combination of printed materials and evidence-based presentations as well as the active involvement of residents helped achieve the goals.
The effect of the multifaceted educational intervention program on ribavirin use was much greater than the effect on antimicrobial use. After the program, only 1.6% of patients without an indication for ribavirin received the drug, while 60.6% of patients were still prescribed an antimicrobial agent. However, the appropriate use of ribavirin also decreased. This was an unexpected and unintended result. It underscores the difficulty of relaying complex messages (e.g., treat only patients with one or more of the following risk factors) versus simple messages (e.g., treat all or treat none). Of potential concern, 97.8% of patients with three or more risk factors received ribavirin before the program, compared with only 39.2% afterward. In addition, 97.8% of the patients admitted to the PICU with two or more risk factors received ribavirin before the program, compared with 36.0% after the program. However, we are unsure if the decreased use of ribavirin in high-risk patients admitted to a general pediatrics ward caused the twofold higher PICU admission rate. A previous study found that only 1.8% of previously healthy full-term infants admitted with RSV lower-respiratory-tract infection were subsequently transferred to the PICU because their condition deteriorated.[37] Other factors, such as admitting only the sickest children to the hospital due to managed care pressures or a more virulent strain of RSV in a given year, may also have been responsible for this finding.
More education needs to be conducted to improve the overall prescribing patterns for ribavirin. This is particularly important because of the trends seen during this seven-year study period. Proportionally, the patients hospitalized appeared to be sicker and have more risk factors over time. This phenomenon is consistent with national health care trends and may be attributable to the fact that more prematurely born babies are being saved and many of them have comorbidities. Therefore, the appropriate use of ribavirin when necessary must be emphasized.
Although the antimicrobial prescribing rate decreased from 85.6% to 60.6%, most of the infants and children treated did not need antimicrobial therapy. Only 1.8% of the patients in the study had a positive bacterial culture. The positive blood cultures were probably caused by contaminants and did not represent bacteremia or sepsis. The positive urine cultures may have represented concurrent urinary tract infections or simply asymptomatic bacteriuria. It is difficult to change deeply held beliefs and fears of undertreating bacterial infections. Data regarding the prevalence of concurrent serious bacterial infections from our own institution have not yet been formally presented to the medical and house staffs and may be helpful in changing such beliefs.[38]
No decay effect was seen during the three-year period following the educational program. The reduction in ribavirin use was maintained. Antimicrobial use varied from year to year, but the decrease was also maintained. This is very unusual, as most educational and academic detailing programs have a short-lived effect.[19,20] Some of the components of the program (morning report discussions, noon conference lectures, and teaching rounds) were continued during the three-year period afterward, but the overall educational intensity was less. Despite the long-term sustained effects, we do believe "booster" programs are needed to reinforce the messages.
Our program had several weaknesses. First, no formal survey was initially conducted to determine baseline knowledge and motivations for physicians' current prescribing patterns. This information is critical to the design of the program and the creation of the key educational messages. We used informal conversations as a surrogate for a formal survey. Second, not all attending physicians were present at every educational meeting or presentation. Thus, some pediatricians may have received only the printed materials. Third, no prescriber profiling and targeting of specific attending physicians were conducted, which could have resulted in better outcomes. Fourth, no face-to-face meetings with individual attending physicians were performed. Finally, attending physicians were not given periodic feedback or praise on changes made in their prescribing patterns for ribavirin and antimicrobials.
This study also had several limitations. First, the retrospective study design relied on chart reviews and had all the inherent weaknesses of such. Second, no random allocation or control group (e.g., another hospital or a group of physicians who did not receive materials and attend meetings) was established to curb the effect of potential confounders and allow for a strong causal association to be made. Third, it is impossible to determine which components of the educational program were effective and what roles other environmental influences played (e.g., managed care, consensus guidelines, physician turnover, and hospitalwide programs). Managed care penetration in Corpus Christi during this time period was low and was probably minimally contributive. However, hospital length of stay for all patients (including those hospitalized with RSV lower-respiratory-tract infection) slowly and steadily decreased during this period, possibly due to pressures from managed care organizations. Publication of the AAP guidelines for ribavirin use in 1993[11] and the revised AAP guidelines in 1996[12] may have been partially responsible for the reduction in ribavirin prescribing over time. The wording change from "should be used" in 1993 to "may be considered" in 1996 was significant in the minds of pediatricians. Pediatricians perceived that the revised guidelines gave them more flexibility in deciding whether or not to initiate ribavirin therapy. However, as stated earlier, consensus statements and practice guidelines have not been shown to be effective on their own.[14,15] In addition, the 1996 AAP guidelines were published after the peak of the 1995-96 RSV season and this could not have substantially contributed to the significant decrease in ribavirin prescribing during that season. Attending-physician turnover was very low during the seven-year study period. Also, no similar hospitalwide programs were conducted for antimicrobial prescribing or management of RSV.
Am J Health Syst Pharm. 2003;60(8) © 2003 American Society of Health-System Pharmacists
Cite this: Effect of an Educational Program on the Treatment of RSV Lower-Respiratory-Tract Infection - Medscape - Apr 15, 2003.
Comments