Bacterial Resistance to Antibiotics (page 1)

(This chapter has 4 pages) 2008 Kenneth Todar, PhD Introduction In the past 60 years, antibiotics have been critical in the fight against infectious disease caused by bacteria and other microbes. Antimicrobial chemotherapy has been a leading cause for the dramatic rise of average life expectancy in the Twentieth Century. However, disease-causing microbes that have become resistant to antibiotic drug therapy are an increasing public health problem. Wound infections, gonorrhea, tuberculosis, pneumonia, septicemia and childhood ear infections are just a few of the diseases that have become hard to treat with antibiotics. One part of the problem is that bacteria and other microbes that cause infections are remarkably resilient and have developed several ways to resist antibiotics and other antimicrobial drugs. Another part of the problem is due to increasing use, and misuse, of existing antibiotics in human and veterinary medicine and in agriculture. In 1998, in the United States, 80 million prescriptions of antibiotics for human use were filled. This equals 12,500 tons in one year. Animal and agricultural uses of antibiotics are added to human use. Agricultural practices account for over 60% of antibiotic usage in the U.S., so this adds an additional 18,000 tons per year to the antibiotic burden in the environment. Nowadays, about 70 percent of the bacteria that cause infections in hospitals are resistant to at least one of the drugs most commonly used for treatment. Some organisms are resistant to all approved antibiotics and can only be treated with experimental and potentially toxic drugs. An alarming increase in resistance of bacteria that cause community acquired infections has also been documented, especially in the staphylococci and pneumococci (Streptococcus pneumoniae), which are prevalent causes of disease and mortality. In a recent study, 25% of bacterial pneumonia cases were shown to be resistant to penicillin, and an additional 25% of cases were resistant to more than one antibiotic. Microbial development of resistance, as well as economic incentives, has resulted in research and development in the search for new antibiotics in order to maintain a pool of effective drugs at all times. While the development of resistant strains is inevitable, the slack ways that we administer and use antibiotics has greatly exacerbated the process. Unless antibiotic resistance problems are detected as they emerge, and actions are taken immediately to contain them, society could be faced with previously treatable diseases that have become again untreatable, as in the days before antibiotics were developed. History of antibiotics and emergence of antibiotic resistance The first antibiotic, penicillin, was discovered in 1929 by Sir Alexander Fleming, who observed inhibition of staphylococci on an agar plate contaminated by a Penicillium mold. Fleming was searching for potential antibacterial compounds. He noticed that a patch of the mold Penicillium notatum had grown on a plate containing the bacterium Staphylococcus and that around the mold there was a zone where no Staphylococcus could grow. After more research, he was able to show that culture broth of the mold prevented growth of the Staphylococcus even when diluted up to 800 times. He named the active substance penicillin but was unable to isolate it.

In the center of the plate is a colony of Penicillium notatum, a mold that produces penicillin. After appearance of the mold colony, the plate was overlaid with a bacterial culture of Micrococcus luteus which forms a yellow "lawn" of growth. A zone of inhibition of bacterial growth surrounds the fungal colony where penicillin has diffused into the medium. http://helios.bto.ed.ac.uk/bto/microbes/penicill.htm#Top

Several years later, in 1939, Ernst Chain and Howard Florey developed a way to isolate penicillin and used it to treat bacterial infections during the Second World War. The new drug came into clinical usage in 1946 and made a huge impact on public health. For these discoveries Fleming, Chain and Florey were awarded the Nobel prize in 1945. Their discovery and development revolutionized modern medicine and paved the way for the development of many more natural antibiotics.

While Fleming was working on penicillin, Gerhard Domagk, a German doctor, announced the discovery of a synthetic molecule with antibacterial properties. He named the compound Prontosil, and it became the first of a long series of synthetic antibiotics called sulfonamides or sulfa drugs. Prontosil was introduced to clinical use in the 1930s and was used to combat urinary tract infections, pneumonia and other conditions. While sulfa drugs in many cases are not as effective as natural antibiotics, they are now in widespread use for the treatment of many conditions. Gerhard Domagk was awarded the Nobel prize in 1939 for his discovery of Prontosil. In 1946, penicillin became generally available for treatment of bacterial infections, especially those caused by staphylococci and streptococci. Initially, the antibiotic was effective against all sorts of infections caused by these two Gram-positive bacteria. Penicillin had unbelievable ability to kill these bacterial pathogens without harming the host that harbored them. It is important to note that a significant fraction of all human infections are caused by these two bacteria (i.e., strep throat, pneumonia, scarlet fever, septicemia, skin infections, wound infections, etc.). In the late 1940s and early 1950s, new antibiotics were introduced, including streptomycin, chloramphenicol and tetracycline, and the age of antibiotic chemotherapy came into full being. These antibiotics were effective against the full array of bacterial pathogens including Gram-positive and Gramnegative bacteria, intracellular parasites, and the tuberculosis bacillus. Synthetic antimicrobial agents such as the "sulfa drugs" (sulfonamides) and anti-tuberculosis drugs, such as para aminosalicylic acid (PAS) and isoniazid (INH), were also brought into wider usage. The first signs of antibiotic resistance There has probably been a gene pool in nature for resistance to antibiotic as long as there has been for

antibiotic production, for most microbes that are antibiotic producers are resistant to their own antibiotic. In retrospect, it is not surprising that resistance to penicillin in some strains of staphylococci was recognized almost immediately after introduction of the drug in 1946. Likewise, very soon after their introduction in the late 1940s, resistance to streptomycin, chloramphenicol and tetracycline was noted. By 1953, during a Shigella outbreak in Japan, a strain of the dysentery bacillus (Shigella dysenteriae) was isolated which was multiple drug resistant, exhibiting resistance to chloramphenicol, tetracycline, streptomycin and the sulfonamides. Over the years, and continuing into the present almost every known bacterial pathogen has developed resistance to one or more antibiotics in clinical use. Evidence also began to accumulate that bacteria could pass genes for drug resistance between strains and even between species. For example, antibiotic-resistance genes of staphylococci are carried on plasmids that can be exchanged with Bacillus, Streptococcus and Enterococcus providing the means for acquiring additional genes and gene combinations. Some are carried on transposons, segments of DNA that can exist either in the chromosome or in plasmids. In any case, it is clear that genes for antibiotic resistance can be exchanged between strains and species of bacteria by means of the processes of horizontal gene transmission (HGT). Multiple drug resistant organisms Multiple drug resistant organisms are resistant to treatment with several, often unrelated, antimicrobial agents as described above in Shigella. Some of the most important types of multiple drug resistant organisms that have been encountered include: MRSA - methicillin/oxacillin-resistant Staphylococcus aureus VRE - vancomycin-resistant enterococci ESBLs - extended-spectrum beta-lactamases (which are resistant to cephalosporins and monobactams) PRSP - penicillin-resistant Streptococcus pneumoniae MRSA and VRE are the most commonly encountered multiple drug resistant organisms in patients residing in non-hospital healthcare facilities, such as nursing homes and other long-term care facilities. PRSP are more common in patients seeking care in outpatient settings such as physicians' offices and clinics, especially in pediatric settings. ESBLs are most often encountered in the hospital (intensive care) setting, but MRSA and VRE also have a significant nosocomial ecology. Methicillin-Resistant Staph Aureus. MRSA refers to "methicillin-resistant Staphylococcus aureus", which are strains of the bacterium that are resistant to the action of methicillin, and related beta-lactam antibiotics (e.g. penicillin and cephalosporin). MRSA have evolved resistance not only to beta-lactam antibiotics, but to several classes of antibiotics. Some MRSA are resistant to all but one or two antibiotics, notably vancomycin-resistant. But there have been several reports of VRSA (VancomycinResistant Staph Aureus) that are troublesome in the ongoing battle against staph infections. MRSA are often sub-categorized as Hospital-Associated MRSA (HA-MRSA) or Community-Associated MRSA (CA-MRSA), depending upon the circumstances of acquiring disease. Based on current data, these are distinct strains of the bacterial species. HA-MRSA occurs most frequently among patients who undergo invasive medical procedures or who have weakened immune systems and are being treated in hospitals and healthcare facilities such as nursing homes and dialysis centers. MRSA in healthcare settings commonly causes serious and potentially life threatening infections, such as bloodstream infections, surgical site infections or pneumonia. In the case of HA- MRSA, patients who already have an MRSA infection or who carry the bacteria on their bodies but do not have symptoms (colonized) are the most common sources of transmission. The main mode of transmission to other patients is through human hands, especially healthcare workers' hands. Hands may become contaminated with MRSA bacteria by contact with infected or colonized patients. If appropriate hand hygiene such as washing with soap and water or using an alcohol-based hand sanitizer is not performed, the bacteria can be spread when the healthcare worker touches other patients. MRSA infections that occur in otherwise healthy people who have not been recently (within the past year) hospitalized or had a medical procedure (such as dialysis, surgery, catheters) are categorized as

community-associated (CA-MRSA) infections. These infections are usually skin infections, such as abscesses, boils, and other pus-filled lesions. About 75 percent of CA-MRSA infections are localized to skin and soft tissue and usually can be treated effectively. However, CA-MRSA strains display enhanced virulence, spread more rapidly and cause more severe illness than traditional HA-MRSA infections, and can affect vital organs leading to widespread infection (sepsis), toxic shock syndrome and pneumonia. It is not known why some healthy people develop CA-MRSA skin infections that are treatable whereas others infected with the same strain develop severe, fatal infections. Studies have shown that rates of CA-MRSA infection are growing fast. One study of children in south Texas found that cases of CA-MRSA had a 14-fold increase between 1999 and 2001. CA-MRSA skin infections have been identified among certain populations that share close quarters or experience more skin-to-skin contact. Examples are team athletes, military recruits, and prisoners. However, more and more CA-MRSA infections are being seen in the general community as well, especially in certain geographic regions. Also, CA-MRSA are infecting much younger people. In a study of Minnesotans published in The Journal of the American Medical Association, the average age of people with MRSA in a hospital or healthcare facility was 68. But the average age of a person with CA-MRSA was only 23. More people in the U.S. now die from MRSA infection than from AIDS. Methicillin-resistant Staphylococcus aureus was responsible for an estimated 94,000 life-threatening infections and 18,650 deaths in 2005, as reported by CDC in the Oct. 17, 2007 issue of The Journal of the American Medical Association. The national estimate is more than double the invasive MRSA prevalence reported five years earlier. That same year, roughly 16,000 people in the U.S. died from AIDS, according to CDC. While most invasive MRSA infections could be traced to a hospital stay or some other health care exposure, about 15% of invasive infections occurred in people with no known health care risk. Two-thirds of the 85% of MRSA infections that could be traced to hospital stays or other health care exposures occurred among people who were no longer hospitalized. People over age 65 were four times more likely than the general population to get an MRSA infection. Incidence rates among blacks were twice that of the general population, and rates were lowest among children over the age of 4 and teens. Extended-Spectrum beta-lactamase (ESBL) - producing Gram-negative bacteria Extendedspectrum beta-lactamases (ESBLs) are plasmid-associated beta lactamases that have recently been found in the Enterobacteriaceae. ESBLs are capable of hydrolyzing penicillins, many narrow spectrum cephalosporins, many extended-spectrum cephalosporins, oxyimino-cephalosporins (cefotaxime, ceftazidime), and monobactams (aztreonam). Beta-lactamase inhibitors (e.g. clavulanic acid) generally inhibit ESBL producing strains. ESBL producing isolates are most commonly Klebsiella ssp, predominantly Klebsiella pneumoniae, and E. coli, but they have been found throughout the Enterobacteriaeae. Because ESBL enzymes are plasmid mediated, the genes encoding these enzymes are easily transferable among different bacteria. Most of these plasmids not only contain DNA encoding ESBL enzymes but also carry genes conferring resistance to several non--Lactam antibiotics. Consequently, most ESBL isolates are resistant to many classes of antibiotics. The most frequent coresistances found in ESBL-producing organisms are aminoglycosides, fluoroquinolones, tetracyclines, chloramphenicol, and sulfamethoxazoletrimethoprim. Treatment of these multiple drug-resistant organisms is a therapeutic challenge. ESBL producing strains have been isolated from abscesses, blood, catheter tips, lung, peritoneal fluid, sputum, and throat cultures. They apparently have a world-wide distribution. Rates of isolation vary greatly worldwide and within geographic areas and are rapidly changing over time. In the United States, between 1990 to 1993, a survey of the intensive care units of 400 hospitals recorded an increase from 3.6% to 14.4% in ESBL producing strains of Klebsiella. In 1994, the CDC reported that 8% of Klebsiella spp from a few large centers produced ESBLs. In Europe, as of 1995, ESBLs occurred in 20%-25% of Klebsiella ssp from patients in ICUs, although they were found in patients up to 30%-40% frequency in France. Known risk factors for colonization and/or infection with organisms harboring ESBLs include admission to

an intensive care unit, recent surgery, instrumentation, prolonged hospital stay and antibiotic exposure, especially to extended-spectrum beta-lactam antibiotics. Use of extended-spectrum antibiotics exerts a selective pressure for emergence of ESBL producing strains. The resistance plasmids can then be transferred to other bacteria, not necessarily of the same species, conferring resistance to them. The lower GI tract of colonized patients is the main reservoir of these organisms. Gastrointestinal carriage can persist for months. In some cities in the United States, nursing homes may be an important reservoir of ESBL producing strains. Nursing home patients are more likely to be treated empirically with antibiotics, and thus on admission to a hospital to be more likely to possess an ESBL producing strain. Patient to patient transmission of ESBL producing organisms occurs via the hands of hospital staff. It is known that ESBL producing strains can survive in the hospital environment. Nosocomial infections in patients occur through the administration of extended spectrum beta-lactam antibiotics or via transmission from other patients via health care workers, who become colonized with resistant strains via exposure to patients or other health care workers. Spread of ESBL producing strains can be minimized by good infection control practices, especially by good hand washing technique. Bacterial mechanisms of antibiotic resistance Several mechanisms have evolved in bacteria which confer them with antibiotic resistance. These mechanisms can either chemically modify the antibiotic, render it inactive through physical removal from the cell, or modify target site so that it is not recognized by the antibiotic. The most common mode is enzymatic inactivation of the antibiotic. An existing cellular enzyme is modified to react with the antibiotic in such a way that it no longer affects the microorganism. An alternative strategy utilized by many bacteria is the alteration of the antibiotic target site. These and other mechanisms are shown in the the figure and accompanying table below.

Mechanisms of antibiotic resistance in bacteria

Antibiotic Chloramphenicol Tetracycline

Method of resistance reduced uptake into cell active efflux from the cell

-lactams, Erythromycin, Lincomycin -lactams, Aminoglycosides, Chloramphenicol Sulfonamides, Trimethoprim Sulfonamides, Trimethoprim

eliminates or reduces binding of antibiotic to cell target enzymatic cleavage or modification to inactivate antibiotic molecule metabolic bypass of inhibited reaction overproduction of antibiotic target (titration)

The acquisition and spread of antibiotic resistance in bacteria The development of resistance is inevitable following the introduction of a new antibiotic. Initial rates of resistance to new drugs are normally on the order of 1%. However, modern uses of antibiotics have caused a huge increase in the number of resistant bacteria. In fact, within 8-12 years after wide-spread use, strains resistant to multiple drugs become widespread. Multiple drug resistant strains of some bacteria have reached the proportion that virtually no antibiotics are available for treatment. Antibiotic resistance in bacteria may be an inherent trait of the organism (e.g. a particular type of cell wall structure) that renders it naturally resistant, or it may be acquired by means of mutation in its own DNA or acquisition of resistance-conferring DNA from another source. Inherent (natural) resistance. Bacteria may be inherently resistant to an antibiotic. For example, an organism lacks a transport system for an antibiotic; or an organism lacks the target of the antibiotic molecule; or, as in the case of Gram-negative bacteria, the cell wall is covered with an outer membrane that establishes a permeability barrier against the antibiotic. Acquired resistance. Several mechanisms are developed by bacteria in order to acquire resistance to antibiotics. All require either the modification of existing genetic material or the acquisition of new genetic material from another source. Vertical gene transfer The spontaneous mutation frequency for antibiotic resistance is on the order of about of about 10-8- 10-9. This means that one in every every 108- 109 bacteria in an infection will develop resistance through the process of mutation. In E. coli, it has been estimated that streptomycin resistance is acquired at a rate of approximately 10-9 when exposed to high concentrations of streptomycin. Although mutation is a very rare event, the very fast growth rate of bacteria and the absolute number of cells attained means that it doesn't take long before resistance is developed in a population. Once the resistance genes have developed, they are transferred directly to all the bacteria's progeny during DNA replication. This is known as vertical gene transfer or vertical evolution. The process is strictly a matter of Darwinian evolution driven by principles of natural selection: a spontaneous mutation in the bacterial chromosome imparts resistance to a member of the bacterial population. In the selective environment of the antibiotic, the wild type (non mutants) are killed and the resistant mutant is allowed to grow and flourish Horizontal gene transfer Another mechanism beyond spontaneous mutation is responsible for the acquisition of antibiotic resistance. Lateral or horizontal gene transfer (HGT) is a process whereby genetic material contained in small packets of DNA can be transferred between individual bacteria of the same species or even between different species. There are at least three possible mechanisms of HGT, equivalent to the three processes of genetic exchange in bacteria. These are transduction, transformation or conjugation. Conjugation occurs when there is direct cell-cell contact between two bacteria (which need not be closely related) and transfer of small pieces of DNA called plasmids takes place. This is thought to be the main mechanism of HGT.

Transformation is a process where parts of DNA are taken up by the bacteria from the external environment. This DNA is normally present in the external environment due to the death and lysis of another bacterium. Transduction occurs when bacteria-specific viruses (bacteriophages) transfer DNA between two closely related bacteria.

Mechanisms of horizontal gene transfer (HGT) in bacteria The combined effects of fast growth rates to large densities of cells, genetic processes of mutation and selection, and the ability to exchange genes, account for the extraordinary rates of adaptation and evolution that can be observed in the bacteria. For these reasons bacterial adaptation (resistance) to the antibiotic environment seems to take place very rapidly in evolutionary time. Bacteria evolve fast!

Tests for sensitivity and resistance to antibiotics. (Left) The size of the zones of inhibition of microbial growth surrounding the antibiotic disks on the plate are an indication of microbial susceptibility to the antibiotic. (Right) By the use of these disks it is also possible to detect the occurrence of individual mutants within the culture that have developed antibiotic resistance. This image shows a close-up of the novobiocin disk (marked by an arrow on the whole plate) near which individual mutant cells in the bacterial population that were resistant to the antibiotic and have given rise to small colonies within the zone of inhibition.

Societal, medical and agricultural practices that lead to antibiotic resistance In the face of a microbe's inherent ability to develop antibiotic resistance, many societal. medical and agricultural practices contribute to this process, foremost of which are discussed below. Antibiotics in food and water Prescription drugs are not the only source of antibiotics in the environment. In the United States, antibiotics can be found in beef cattle, pigs and poultry. The same antibiotics then find their way into municipal water systems when the runoff from housing facilities and feedlots contaminates streams and groundwater. So it's a double hit: we get antibiotics in our food and drinking water, and we meanwhile promote bacterial resistance. Routine feeding of antibiotics to animals is banned in the European Union and many other industrialized countries. Maybe they know something we don't. Indiscriminate use of antibiotics in agriculture and veterinary practice The non-therapeutic use of antibiotics in livestock production makes up at least 60 percent of the total antimicrobial production in the United States. Irresponsible use of antibiotics in farm animals can lead to the development of resistance in bacteria associated with the animal or with people who eat the animal. Such resistance can then be passed on to human pathogens by mechanisms of HGT. Of major concern is the use of antibiotics as feed additives given to farm animals to promote animal growth and to prevent infections (rather than cure infections). The use of an antibiotic in this way contributes to the emergence of antibiotic-resistant pathogens and reduces the effectiveness of the antibiotic to combat human infections. Antibiotic resistance in genetically modified crops Antibiotic-resistance genes are used as "markers" in genetically modified crops. The genes are inserted into the plant in early stages of development to in order to detect specific genes of interest . e.g. herbicide-resistant genes or insecticidal toxin genes. The antibiotic-resistance genes have no further role to play, but they are not removed from the final product. This practice has met with criticism because of the potential that the antibiotic-resistance genes could be acquired by microbes in the environment. In some cases these marker genes confer resistance to front-line antibiotics such as the beta-lactams and aminoglycosides. Inappropriate use of antibiotics in the medical environment One problem is the casual use of antibiotics in medical situations where they are of no value. This is the fault of both health care workers and patients. Prescribers sometimes thoughtlessly prescribe 'informed' demanding patients with antibiotics. This leads to use of antibiotics in circumstances where they are of not needed, e.g. viral upper respiratory infections such as cold and flu, except when there is serious threat of secondary bacterial infection. Another problem is patient failure to adhere to regimens for prescribed antibiotics. Patients and doctors need to realize their responsibility when they begin an antibiotic regimen to combat an infectious disease. There are several measures that should be considered. Patients should not take antibiotics for which there is no medical value (corollary: doctors should not prescribe antibiotics for which there is no medical value). Patients should adhere to appropriate prescribing guidelines and take antibiotics until they have finished. Patients should be give combinations of antibiotics, when necessary, to minimize the development of resistance to a single antibiotic (as in the case of TB). Patients need to be given another antibiotic or combination of antibiotics if the first is not working. Combating antibiotic resistance The following are recommendations to combat the development of antibiotic resistance in bacteria and

other microorganisms. Search for new antibiotics. To combat the occurrence of resistant bacteria, biotechnology and pharmaceutical companies must constantly research, develop and test new antimicrobials in order to maintain a pool of effective drugs on the market. Stop the use of antibiotics as growth-promoting substances in farm animals. Of major concern is the use of antibiotics as feed additives given to farm animals to promote animal growth and to prevent infections rather than cure infections. The use of such antibiotics contributes to the emergence of antibioticresistant bacteria that threaten human health and decreases the effectiveness of the same antibiotics used to combat human infections. Use the right antibiotic in an infectious situation as determined by antibiotic sensitivity testing, when possible. Stop unnecessary antibiotic prescriptions. Unnecessary antibiotic prescriptions have been identified as causes for an enhanced rate of resistance development. Unnecessary prescriptions of antibiotics are made when antibiotics are prescribed for viral infections (antibiotics have no effect on viruses). This gives the opportunity for indigenous bacteria (normal flora) to acquire resistance that can be passed on to pathogens. Finish antibiotic prescriptions. Unfinished antibiotic prescriptions may leave some bacteria alive or may expose them to sub-inhibitory concentrations of antibiotics for a prolonged period of time. Mycobacterium tuberculosis is a slow growing bacteria which infects the lung and causes tuberculosis. This disease kills more adults than any other infectious disease. Due to the slow growing nature of the infection, treatment programs last for months or even years. This has led to many cases on unfinished prescriptions and 5% of strains now observed are completely resistant to all known treatments and hence incurable. Several other possible solutions have been proposed or implemented to combat antibiotic resistance. In the pharmaceutical industry, past and current strategies to combat resistance have not been effective. Pharmaceutical companies are seeking new, less costly strategies to develop antibiotics. A decrease in the number of prescriptions for antibiotics, especially in small children, is occurring. Several countries such as the UK have regulations concerning the use of antibiotics in animal feed. Large scale public health education efforts are underway to stress the importance of finishing prescriptions. Indeed, in many places, failure to finish tuberculosis prescriptions can result in jail time. Summary The discovery of antibiotics was a leap in modern medicine. They have been able to stop the growth or kill many different kinds of microorganisms. However, bacteria have proven to be much more innovative and adaptive than we imagined and have developed resistance to antibiotics at an ever increasing pace. Bad practices and mismanagement have only exacerbated the situation. We could soon return to a state of medical health that was as dire as that which occurred prior to antibiotic use. However, with more research, education of the public, and well thought out regulations, the problems can be solved. Several strategies are currently used to find new antibacterial compounds and new strategies are in development and trial. Not only is there a problem in finding new antibiotics to fight old diseases (because resistant strains of bacteria have emerged), there is a parallel problem to find new antibiotics to fight new diseases. In the past three decades, many "new" bacterial diseases have been discovered (E. coli O157:H7 gastric ulcers, Lyme disease, toxic shock syndrome, "skin-eating" streptococci). Already broad patterns of resistance exist in these pathogens, and it seems likely that we will soon need new antibiotics to replace the handful that are effective now against these bacteria, especially as resistance begins to emerge among them in

the selective environment antibiotic chemotherapy. It is said that the discovery and use of antibiotics and immunization procedures against infectious disease are two developments in the field of microbiology that have contributed about twenty years to the average life span of humans in developed countries where these practices are employed. While the greater part of this span in time is probably due to vaccination, most of us are either still alive or have family members or friends who are still alive because an antibiotic conquered an infection that otherwise would have killed them. If we want to retain this medical luxury in our society we must be vigilant and proactive We must fully understand how and why antimicrobial agents work, and why they don't work, and realize that we must maintain a stride ahead of microbial pathogens that can only be contained by antibiotic chemotherapy.

RESISTANCE TO ANTIBIOTICS AND OTHER ANTIMICROBIAL AGENTS CHAPTER 1 INTRODUCTION

1.1 Antibiotics have saved countless lives and transformed the practice of medicine since the first flowering of antimicrobial chemotherapy in the 1930s and '40s.[1] Many people are old enough to remember when these and other antimicrobial drugs were not available. Their memories include patients with pulmonary tuberculosis isolated in sanatoria until either they died or their disease healed itself; frequent postoperative wound infections; bone infections (osteomyelitis) followed by discharging sinuses requiring drainage for year after year; syphilis advancing to its late stages and ending in insanity despite the use of arsenical drugs; cases of tuberculous meningitis, invariably fatal; and simple cuts and scratches giving rise to fatal septicaemia.
Box 1

HISTORY AND TERMINOLOGY

Historically the concept of attacking invading organisms without harming the host was introduced at the turn of the century by the German Paul Ehrlich. This concept he called chemotherapy. The invading organisms he first studied were not bacteria but rather the protozoa that cause malaria and sleeping sickness; but in 1910 he made his great discovery of salvarsan (the 606th synthetic chemical he had tried) which was effective in treating the spirochaete (a type of bacterium) which causes syphilis. He called it a "magic bullet". In the 1930s the sulphonamide drugs were introduced: they were the first effective drugs that attacked the common bacteria such as streptococci and could cure pneumonia and meningitis, although they caused serious problems and side effects. They were not called antibiotics; they were known as "chemotherapeutic agents".

"Antibiotic" was the term originally applied to naturally occurring compounds such as penicillin which attacked infecting bacteria without harming the host. "Antibiotic"

is now regularly used to refer to synthetic compounds as well as natural compounds, and to refer to antiviral as well as antibacterial drugs. In the public mind, however, "antibiotics" are still largely equated with penicillin.

Penicillin, the first antibiotic, was identified in a mould by Alexander Fleming in 1928; but it was not available for use until Florey and Chain and their colleagues purified it in 1940 and showed how effective it could be. Unlike the sulphonamides it seemed completely harmless to the host and very effective against many bacteria. As it was a naturally occurring product, not a synthetic chemical, it was not called a chemotherapeutic drug, although that would have been a perfectly correct description. It would also have been a correct general description to include not only all antibacterial agents but also agents against viruses, protozoa, worms and other parasites, with all of which our report is concerned. However, the word "chemotherapy" is now used and recognised by the public as the term to describe the drug treatment of cancer.

"Cancer chemotherapy" is a legitimate term if we regard cancerous cells as invasive, being therefore, like infecting organisms, foreign to the host. Cancer chemotherapy thus seeks to attack the invader without damaging the host. Chemotherapy against microbial infection is referred to as antimicrobial chemotherapy.

This report is concerned with various forms of antimicrobial chemotherapy, and mainly with antibacterial antibiotics, or antibacterials. It deals also with antiviral antibiotics, or antivirals; and with antimalarial and anthelmintic drugs.

For an historical account of the development of resistance to antibiotics, see the evidence of Professors Phillips and Roberts of the Royal College of Pathologists, p 453.

1.2 L P Garrod wrote in 1968, "No one recently qualified, even with the liveliest imagination, can picture the ravages of bacterial infection which continued until little more than 30 years ago". Since then, many new antibacterial agents have been developed and antiviral chemotherapy, then in its infancy, has become possible for an increasing range of viral diseases. As well as its uses in the direct treatment of infection, antimicrobial chemotherapy has also helped to make possible medical advances such as transplantation and the treatment of many forms of cancer which carry a special risk of infection.

1.3 But the worm in the bud emerged early when, during the development of penicillin, the enzyme which destroys it was isolated and it was presciently predicted (by Abraham and Chain) that penicillin resistance would become a problem. So it has, and so also, at greatly varying intervals following its introduction, has resistance to each new antibiotic. 1.4 "Resistance" means that an organism ceases to be killed or inhibited by a drug. While antibiotics can cause, as can all active therapies, a wide range of adverse effects ranging from trivial to fatal, resistance is a special problem, since the agent loses its former efficacy and the future treatment of other patients is therefore jeopardised.[2] The problem of antibiotic resistance has now become a major concern in medicine throughout the world. 1.5 The fact of antibiotic resistance is widely known, though not so widely understood. In the United Kingdom, the aspect most talked about among the public at large is probably MRSA (methicillin-resistant Staphylococcus aureus), an infection associated principally with hospitals and nursing homes. Other aspects of the problem are familiar to the people affected: for instance, resistant TB (tuberculosis) is a major threat to people with AIDS, while resistant malaria is the scourge of Africa and the Far East. Both Houses of Parliament have debated MRSA and other resistant infections in the past year or two (the Commons on 19 March 1997, the Lords on 4 November 1996). The Government are seized of the issue: it features in the Chief Medical Officer's annual reports for 1995 and 1996, and Ministers are awaiting advice on different aspects of it from the Standing Medical Advisory Committee and the Advisory Committee on the Microbiological Safety of Food (p 373). So we bring the matter before the House confident that it deserves Parliamentary time and attention. This enquiry has been an alarming experience, which leaves us convinced that resistance to antibiotics and other antiinfective agents constitutes a major threat to public health, and ought to be recognised as such more widely than it is at present. 1.6 We begin our report with a brief account of what resistance is and why it matters; for more on these questions, we refer the reader to a recent report by the Parliamentary Office of Science and Technology, Diseases Fighting Back (October 1994). We consider how far resistance can be controlled, and how. We then proceed to consider the evidence we have received on the various means of control: prudent use of antimicrobial agents in human medicine (Chapter 2) and in animals (Chapter 3); infection control (Chapter 4) and disease surveillance (Chapter 5); and development of new drugs (Chapter 6) and vaccines (Chapter 7). Chapter 8 considers the special problems of resistance in viruses, and Chapter 9 considers international issues including malaria. Chapter 10 considers the sources of support for research and data-collection. Our recommendations are set out in Chapter 11, and summarised in Chapter 12. Appendix 7 contains notes on some important antimicrobial agents, Appendix 8 a glossary of other terms, and Appendix 9 a list of acronyms. What is resistance? 1.7 All antibiotic resistance has a genetic basis. Some organisms are inherently resistant to many antibiotics ("innate resistance"). This resistance probably evolved as a response to exposure to antibiotics present in the natural environment. Many such organisms pose no threat to healthy people, but may become important pathogens in vulnerable patients in hospital. Examples include the Pseudomonas species and some Enterococci. 1.8 Acquired resistance can arise by a number of diverse mechanisms:

(i) Mutational resistance. These mutations have occurred randomly in a small proportion of the particular bacterial population. The most familiar example is seen in the bacterium causing tuberculosis, where a few organisms are naturally resistant to, for example, streptomycin. In the presence of streptomycin as a single antibiotic these resistant organisms soon become the dominant population. (ii) By horizontal transfer of genes determining resistance from one organism to another. This can occur by the direct transfer (conjugation) between bacteria of genetic material carried on small pieces of DNA (plasmids) situated within the bacterial cell but outside the bacterial chromosome, or by similar pieces of DNA carried on a bacterial virus, a bacteriophage (transduction), or by direct transfer of naked DNA (transformation). 1.9 While these mechanisms have been known for many years, what has emerged more recently is knowledge of the great frequency and flexibility with which bacteria are able to exchange genetic material. and the crucial importance of these mechanisms in bacterial evolution. It is now known, for example, that genetic interchange can take place between a much more diverse variety of organisms than was formerly thought. and is probably a common event in the natural world. There is a global pool of resistance genes which can spread between different bacterial populations occupying different habitats, e.g. between man, animals and the environment. Genes carrying antibiotic resistance factors are easily able to spread if the host organism gains an evolutionary advantage in acquiring them. The importance of these processes for antibiotic resistance in man and animals is that, by whichever process genes for resistance have been acquired, the presence of an antibiotic in the environment of the bacterium imposes "selection pressure" and encourages resistance to spread. The antibiotic kills all susceptible bacteria, thereby "selecting out" the resistant strain; in this way a previously minor population of antibiotic-resistant organisms rapidly becomes dominant. Although there are enormous variations in the speed with which resistance to any antibiotic emerges, and in its geographical spread once it has emerged, it is indisputable that resistance has developed to many new agents after their introduction, with consequent diminution or actual loss of their former value to medicine. Thus has appeared the vicious circle repeatedly witnessed during the last half century, in which the value of each new antibiotic has been progressively eroded by resistance, leading to the introduction of a new and usually more expensive agent, only for this in its turn to suffer the same fate. Clinical resistance 1.10 Bacterial species differ greatly in their inherent susceptibility or resistance to various antibiotics. There is also a range of susceptibility within any species, so that some organisms are more susceptible than others. Clinical resistance, i.e. whether the antibiotic will or will not work in a patient or animal, is a more complex concept in which many other factors are involved such as the precise location of the infection, the distribution of the drug in body fluids and the state of the patient's immune system. 1.11 Resistance is measured in the clinical microbiology laboratory by qualitative or quantitative methods which attempt to relate the test results to the expected effect in clinical practice, taking into account such factors as the range of serum concentrations achieved when the antibiotic is administered. Most laboratories use an agar disc susceptibility test in which isolates (i.e. samples) are categorised as susceptible, resistant or intermediate.

1.12 There is much continued discussion about the best methods of antibiotic testing, about quality control and about international agreement on methods. In practice, the results of these pragmatic tests often relate well to clinical success or failure, for example in tuberculosis and in gonorrhoea. 1.13 At a more basic level, the biochemical mechanisms responsible for antibiotic resistance have been analysed in great detail. Resistance arises (i) if the bacteria can inactivate the drug before it reaches its target within the bacterial cell; (ii) if the outer layers of the cell are impermeable, and prevent the drug from entering; (iii) if the drug enters but is then pumped back out again ("efflux"); (iv) if the target is altered so that it is no longer recognised by the antibiotic, or (v) if the bacteria acquire an alternative metabolic pathway that renders the antibiotic's target redundant ("by-pass"). Although some hundreds of resistances are known, virtually all can be ascribed to one of these five broad types of mechanism. See Figure 1, which represents the antibiotic as a bullet and the target as a roundel.[3] Cross-infection 1.14 Bacteria inhabit a global environmental pool in which resistant bacteria, and genes transferring antibiotic resistance between bacteria, can and do spread easily between people and animals. A continuous process of exchange of genes takes place within the microbial world. The two variable factors affecting the spread of antibiotic resistance are the selection pressure exerted by antibiotic use, and the ease with which resistant organisms are able to spread between people by "cross-infection". The international dimension 1.15 Because the amount of interaction between human populations, and with them their commensal[4] microbes, varies greatly, the type and frequency of antibiotic resistance in any particular organism differs greatly between geographical locations. In the longer term, however, once antibiotic resistance is established in an organism, it spreads with lesser or greater speed throughout the world. Modern methods of molecular epidemiology have enabled the spread of bacteria to be tracked and it is clear that bacteria, some of them carrying antibiotic resistance factors, can spread between countries and continents with phenomenal speed in this era of mass travel.

1 For terminology, see Box 1, and the glossary in Appendix 8. Back

2 Resistance should be distinguished from tolerance. When a patient develops tolerance to a drug, no other patient is affected; but a resistant organism can infect others. Back 3 We are indebted for this simple account of a complex matter, and for the Figure, to Dr David Livermore, Head of the PHLS Antibiotic Reference Unit. See also Diseases Fighting Back, Parliamentary Office of Science and Technology, October 1994; and the memoranda of the Association of Medical Microbiologists (p 2) and the Society for General Microbiology (p 485). Back 4 Commensal microbes, or "flora", are the numerous and diverse micro-organisms which inhabit the skin, nose, mouth and gut. They do not normally cause disease in healthy people. Back 1.16 Although antibiotic resistance is encountered everywhere, there are special problems in hospitals and other health care institutions. Many organisms which are part of the normal commensal flora of the body pose an important threat to patients whose resistance is lowered by reason of illness, surgery, administration of immunosuppressant drugs, or extremes of age. Many of these organisms, such as VRE (vancomycin-resistant enterococcus[5]), are highly resistant, and even previously susceptible species can easily be replaced by resistant strains. 1.17 The hospital environment, especially in departments such as intensive care units and neonatal units, operates as an epidemiological pressure cooker for the emergence of resistance, combining high infective risks in immunologically compromised patients who are also undergoing invasive procedures, frequent spread of infection, and high usage of antibiotics exerting strong selective pressure on the microbial population. 1.18 Some organisms causing hospital infection, such as MRSA (methicillin-resistant Staphylococcus aureussee below), may also become a significant problem outside hospitals; and, even if they do not, their presence in the normal flora of people in the community is an important factor in the dissemination of these resistant strains. Does antibiotic resistance matter? 1.19 Antibiotic resistance threatens mankind with the prospect of a return to the preantibiotic era. This will not, of course, happen overnight. It is a relatively slow but inexorable process, patchy in its effects but already under way. The options available for the treatment of infections have everywhere become constrained. In some locations, the organisms causing several life-threatening infections are now resistant to all available antibiotics, so that for patients suffering these illnesses the antibiotic era has already ended. Some examples will serve to emphasise the great importance of resistance in bacterial, viral and parasitic disease. 1.20 The organisms causing gut infections such as typhoid and bacillary dysentery are peculiarly liable to become antibiotic resistant. These organisms also acquire resistance to several antibiotics (multiple drug resistance, MDR) with great facility, so that Salmonella species, for example, can become resistant to 8 or 10 antibiotics. The genes conferring these properties are transferred together as a package, and infections caused by these strains are untreatable by any of the antibiotics involved. Thus, when it became possible to treat typhoid

fever with antibiotics, chloramphenicol and later amoxycillin or trimethoprim were widely and successfully used. Resistance to all these agents emerged so that in recent years it has been necessary to employ ciprofloxacin as the only effective agent remaining to us. Now ciprofloxacin-resistant typhoid has emerged in many places and is at present causing an epidemic in Tajikistan. 1.21 Pneumococcus (Streptococcus pneumoniae) is a universal organism often carried in the nose and throat of healthy people. It is the most common cause of pneumonia in adults and children throughout the world, a major cause of otitis media, and one of the three most common causes of bacterial meningitis. Pneumococcal meningitis is an especially serious form, with a high mortality and a high probability of deafness and other neurological consequences in its survivors. For nearly 40 years after penicillin was introduced, the pneumococcus remained fully susceptible. Resistance then emerged and is now in the process of disseminating throughout the world.[6] In some countries resistant strains are already dominant; in the United Kingdom at present they are an increasing problem. A few penicillin-resistant pneumococci are highly resistant and these strains often show multiple resistance to other antibiotics. Few options are available for treating meningitis caused by such strains, and the agents remaining are unavailable in many poorer countries, so that pneumococcal meningitis has become effectively untreatable in these places. 1.22 The lives of patients with meningococcal meningitis were first saved by sulphonamides in the late 1930s. By the 1970s too many strains were sulphonamide-resistant for these drugs to remain in use for a life-threatening infection and penicillin became universally, and effectively, the treatment of choice. Penicillin resistance in meningococci has now emerged in a few places. Its presence anywhere is serious; spread to the meningitis belt of Africa (a wide geographical area subject to large epidemics every few years) would constitute a major disaster in world health. 1.23 Staphylococci are omnipresent on our skins; they are normally benign but capable of causing infections ranging from a boil to life-threatening septicaemia. Methicillin-resistant Staphylococcus aureus (MRSA), often also resistant to many other antibiotics, has become highly prevalent in many hospitals and nursing homes.[7] Only vancomycin and related drugs, toxic, expensive and not always effective agents, remain for their treatment; and the first isolates of vancomycin-resistant S. aureus[8] (VRSA) have already been reported in Japan and the USA. 1.24 Gonorrhoea provides one of the clearest examples of the successive loss of one antibiotic after another because of the inexorable advance of antibiotic resistance. At first sulphonamides were successful but resistance rapidly emerged. A form of penicillin resistance which could be overcome by increasing the dose followed, with a progressive increase in the amount of penicillin needed to effect a cure. Later still, gonococci acquired the ability to make penicillinase, completely vitiating the effect of penicillin in gonococcal infection caused by these strains. Many other formerly effective agents have suffered the same fate, leading to the need for progressive changes in national and WHO recommendations for the treatment of this common worldwide infection. Many of the changes have involved increasingly expensive drugs. 1.25 Tuberculosis (TB) kills around three million people each year, more than any other infectious disease. Resistance to the first antibiotics effective against tuberculosis was detected as soon as they were introduced. How this could be prevented, by using two or three agents in

combination, was then rapidly discovered, and a series of meticulous trials by the United Kingdom Medical Research Council established regimens of treatment which were highly successful both in curing patients and in preventing the emergence of resistance. Resistance has remained rare in countries such as the United Kingdom, where these regimens are generally followed under a well-established system for diagnosis and treatment. Sadly, however, resistance has become important in many countries because, for many reasons, the established regimens have often been neglected by doctors and by patients. 1.26 Multi-drug resistant (MDR) tuberculosis is a newer and different problem, found especially but not exclusively in patients with HIV infection, carrying a high mortality and extremely difficult to treat. Here is another example, in which our antibiotic options are nearly exhausted and when tuberculosis again becomes "the captain of all these men of death".[9] MDR tuberculosis is at present rare in the United Kingdom.[10] 1.27 Malaria is thought to cause two million deaths and several hundred million new infections annually throughout the world. A variety of resistance patterns has emerged, creating serious constraints on the options available for both treatment and prevention. In highly endemic areas it is becoming increasingly difficult to treat life-threatening disease in children. Most important is chloroquine resistance, now widespread in many continents. Chloroquine resistance has necessitated a return to the oldest of all chemotherapeutic agents, quinine, but multiple resistance, including partial resistance to quinine, is now a major problem in SE Asia. The consequences of the current degree of malarial resistance are massive, especially in Africa where Ministries of Health have annual budgets of a few dollars per head. See Chapter 9 below. 1.28 Viruses can also become resistant to the drugs used in their treatment. This takes place inside the cells of the patient, within which the virus multiplies. Viruses such as HIV replicate very rapidly and minor variations of the genome occur with each multiplication, leading to a genetically heterogeneous population of viral particles. It is inevitable that, in the presence of an antiviral drug, variants with increased resistance to its action will show a selective advantage and will soon become the majority species. Much effort is now being undertaken to find how best to use antiviral drugs in ways which make it more difficult for resistance to emerge. See Chapter 8 below. 1.29 The impact of antibiotic resistance on our ability to treat some important infections is summarised in Box 2. 1.30 There are some exceptions to the general onward march of resistance. For example, Streptococcus pyogenes (group A haemolytic streptococcus) has so far remained susceptible to penicillin, although it is often resistant to several other agents. Likewise, resistance has rarely been reported among the chlamydia, an important cause of genital and eye infections including trachoma; and the causative organism of syphilis (Treponema pallidum) has remained susceptible to penicillin. We do not attempt to describe the problem of resistance as it affects all pathogens, but have concentrated on those of particular importance for world health. Can resistance be controlled? 1.31 How to tackle the problems of resistance depends critically on the answer to a question on which science is divided: can the rise in the proportion of resistant strains be reversed, or at least slowed down, or is it inexorable?

1.32 Professor David Reeves, President of the Association of Medical Microbiologists (AMM), is an optimist. "There is plenty of evidence that, if you remove the selection pressure, the organisms will slowly revert [to susceptibility], some types of organisms more quickly than others and to certain antibiotics more quickly than to others" (Q 23, cp McGavock Q 675). Dr Peter Davey, Reader in Clinical Pharmacology at Ninewells Hospital, Dundee, gave us a table of supporting evidence for the proposition that resistance is dependent on usage (p 145), but produced at least one potential confounding variable for each point. Professor A Percival, Professor of Clinical Bacteriology at Liverpool University, put it starkly: "The concept that antibiotic resistance is related somehow to the amount of use is critical, because, if it is not true, then we have no chance of controlling it...Although everybody believes that, the evidence to support it and to demonstrate it in a scientifically acceptable way is largely lacking...worldwide" (Q 73). 1.33 In marked contrast with Professor Reeves' optimism was the approach of Dr Bruce Levin, a population biologist from Emory University, whom we met in Atlanta. He believes that resistance is a one-way street: even if antibiotic use is cut back sharply, the proportion of resistant strains wanes slowly if at all; even moderate use still imposes heavy selective pressure; and, if use is resumed, resistance rises again more rapidly than before. As he put it, "We are committed to an arms race"; disarmament is not an option. See Appendix 6. 1.34 Countries with firmer controls on the supply and use of antibiotics, and more rigorous infection control, have lower rates of resistant strains, and it is generally assumed that these things are connected. "Holland and Denmark have amongst the lowest incidence of MRSA, due to their effective antibiotic and infection control policies" (PHLS p 42). "Spain has the highest consumption of anti-infectives per capita in Europe, and one of the worst records of antibiotic resistance" (ABPI p 176).
Box 2 EXAMPLES OF VALUABLE ANTIBIOTIC THERAPIES NOW LOST OR IMPERILLED BY THE SPREAD OF RESISTANCE Organism Disease Agents lost or threatened[11] Penicillin; many others

Pneumococcus

Pneumonia, otitis, meningitis Meningitis, septicaemia Meningitis Wound infection, sepsis

Meningococcus Haemophilus influenzae Staphylococcus aureus

Sulphonamides; (penicillin) Ampicillin, chloramphenicol Penicillin, penicillinaseresistant penicillins, others Most relevant agents Most relevant agents Sulphonamides, penicillin,

Salmonella typhi Shigella spp. Gonococcus

Typhoid fever Bacillary dysentery Gonorrhoea

tetracycline; (ciprofloxacin) Plasmodium falciparum Severe malaria Chloroquine, pyrimethamine; (mefloquine, quinine) Ampicillin, trimethoprim, others

E. coli (coliforms)

Urinary infection, septicaemia

1.35 Slowing the take-over of resistant strains is one thing; eradicating them once they have arrived is another, and seems to be easier for some organisms than for others. Resistance in the gonococcus to penicillin (PHLS p 68), and in Strep. pyogenes to erythromycin (AMM p 9, Finch p 187), can be reduced; resistance in Staph. aureus to methicillin (PHLS p 43) and in E. coli to streptomycin (PHLS p 52), once prevalent, appears to be stable. 1.36 Throughout our enquiry we have listened for stories of success in reversing the rise in the proportion of resistant strains. Only two such stories have been rigorously researched and written up (Anderson Q 687). In Finland, an increase in resistance of Group A streptococci to erythromycin around 1990 was countered by a policy restricting the use of macrolides in favour of alternative drugs. Consumption was reduced by nearly a half by 1992, and the rate of resistance was nearly halved by 1996. Dr Davey told a similar tale from Iceland (Q 261), where a problem with penicillin-resistant pneumococci was associated with day-care centres for children. "There was an information campaign aimed at the public and doctors, saying that giving antibiotics to children too frequently at day-care centres was not a good idea. They did reduce the antibiotic prescribing, and they have reduced the transmission of these resistant bugs". 1.37 According to the Department of Health, "The role of the use of antimicrobials in the development of antimicrobial resistance is undoubted. Those countries with high usage and uncontrolled availability of `over the counter' antibiotics tend to have higher levels of antimicrobial resistance, whereas Denmark for instance has seen a dramatic reduction in the prevalence of antibiotic resistant micro-organisms since tight controls on antimicrobial usage, together with strict infection control procedures, were introduced" (p 342).[12] On the basis of the Danish experience, the Chief Medical Officer considers that it should be an objective of Government strategy "to reduce the prevalence of micro-organisms resistant to current drugs" (Q 756). 1.38 As for malaria, Dr David Warhurst of the London School of Hygiene and Tropical Medicine told us, "There is evidence that, if drugs are not used, some of these [resistant malarial] organisms will probably get back to their pre-existing state" (Q 493). Dr Deenan Pillay, Director of the PHLS Antiviral Susceptibility Reference Laboratory in Birmingham, told a similar tale from virology (Q 614). 1.39 So what is the policy-maker to conclude? We suggest that the following propositions conform with the present state of knowledge:

(i) Any antimicrobial agent must be expected to encounter resistance sooner or later.

(ii) Resistant strains will take longer to emerge and spread if antimicrobial use is controlled and prudent from the start. (iii) Improving the control of antimicrobial use can be expected to slow down the rise in the proportion of resistant strains. In the case of certain pathogens (e.g. streptococci, pneumococci, gonococci), the proportion may even fall; but this must not be expected to happen in every case. If, following an improvement in control and a fall in resistance, control is once again relaxed, reversion to high levels of resistance may be swift. 1.40 Why reducing the selective pressure of antimicrobials sometimes brings down the level of resistance and sometimes does not, scientists cannot yet say with certainty. As noted above, in cases where it does, it may be because resistance, though in itself an advantage from the microbe's point of view, confers a collateral burden such that, in the absence of selective pressure, the resistant strain is at an evolutionary disadvantage (i.e. in terms of "survival of the fittest", the resistant strain is less fit), and susceptible strains take over again. Where resistance does not evolve away, it may be because the resistant strain has undergone a secondary adaptation and evolved around the burden, so that in the absence of selective pressure it is no longer at a disadvantage. Alternatively, it may be because the plasmid carrying the gene which confers the resistance in question also carries, as a package, genes which code for resistance to other agents (e.g. Salmonella may carry packages of resistance to as many as 10 antibiotics); if just one of these agents remains in use, selective pressure will keep all the resistances in the package at high levels. 1.41 These rules of thumb, though rough and ready, are based on evidence from leaders in the field in both the United Kingdom and the USA; and those with whom we have spoken readily admit that knowledge in this area is incomplete and somewhat anecdotal. The case for continued research is clear: see Chapter 10 below. Acknowledgements 1.42 The enquiry which led to this report was conducted between July 1997 and March 1998 by Sub-Committee I, whose members are listed in Appendix 1. They received evidence from the organisations and individuals listed in Appendix 2, to all of whom we are grateful for their time and trouble. They paid two visits to the Headquarters of the Public Health Laboratory Service (PHLS) in Colindale, where Professor Brian Duerden, Deputy Director, and numerous members of PHLS staff, were extremely helpful: see Appendices 3 and 4. They visited King's College Hospital, on Denmark Hill; we are grateful to Professor Mark Casewell, then Professor of Medical Microbiology, and to members of the medical and nursing staff, for organising a very informative visit: see Appendix 5. In November, four members visited the USA: Appendix 6 describes the visit, and acknowledges the numerous people who gave generously of their time to make it worthwhile. We acknowledge the help of Sub-Committee I's Specialist Advisers, Professor Harold Lambert, Emeritus Professor of Microbial Diseases at St George's Hospital, Tooting, and Professor Richard Wise, Professor of Clinical Microbiology at Birmingham City Hospital. Finally, we are grateful to the Parliamentary Office of Science and Technology (POST), for the report Diseases Fighting Back noted above, and for a report on Vaccines and their Future Role in Public Health.

5 See the evidence of Dr H F Kennedy and Dr J R Michie of the Royal Hospital for Sick Children, Glasgow (p 549). Back

6 See the evidence of Professor Keith Klugman, Director of the South African Institute for Medical Research (p 426). Back 7 See the evidence of Dr R Hill of King's College Hospital, p 417. Back 8 These isolates have also been characterised as "vancomycin-intermediate" (VISA), because, although their susceptibility is reduced, they are not absolutely resistant. We were told in the USA that these isolates, though alarming in their own right, were not what VRSA had been expected to look like. Back 9 John Bunyan, The Life and Death of Mr Badman. Back 10 Dr B Bannister at Coppetts Wood Hospital sees 50 new cases of TB each year; of these, two or three are MDR-TB, usually in people from Turkey or central Africa (p 377). Professor D A Mitchison of St George's Hospital puts the total number of cases of MDR-TB in the United Kingdom at present at fewer than 20 (p 432). Back 11 Because of the widely variable distribution of antibiotic resistance at any time, a clear distinction between `lost' and `threatened' cannot be made. An antibiotic now useless in one place may still be valuable in another. Resistance to the agents in parentheses is so far uncommon. Back 12 Dr Rosamund Williams of WHO (Q 132) told us that the Danish strategy included surveillance of resistance and usage, strict national prescribing guidelines with follow-up of doctors who infringe them, and isolation and screening of infected persons. The Minister for Public Health said that the Department of Health are "very interested" in the Danish approach (Q 755). Back

CHAPTER 2 PRUDENT USE IN HUMAN MEDICINE

2.1 The evidence set out above suggests that one way to tackle drug resistance in the short term is to work towards appropriate and prudent usage of the drugs themselves. In the longer term there may be new drugs and vaccines (see Chapters 6 and 7); but every doctor, dentist and veterinary surgeon can, it seems, affect the situation for better or worse from day to day by more or less appropriate prescribing.[13] 2.2 It must be recognised at once that, even without the issue of resistance, use of medicines is a three-cornered battlefield. Doctors want the freedom to do the best for their patients; those who pay for health care, whether governments or insurers or the patients themselves, want good care at low cost; and pharmaceutical manufacturers want to maximise the return to their shareholders. Discussion of the impact of usage on resistance cannot be divorced from this context. Present use in the United Kingdom 2.3 In the United Kingdom, most antimicrobials used in human medicine are prescribed by GPs (general practitioners, or family doctors). The Association of Medical Microbiologists

(AMM - QQ 45-52, p 9) told us that in England alone, GPs prescribe 270m defined daily doses each year"enough antibiotics to treat every man, woman and child in England for five days a year". This is much more than is administered in hospitals (though data do not permit a precise comparison, and even if they did one would not be comparing like with like). This figure is derived from Prescription Pricing Authority (PPA) data for 1992-94; the same data show an annual increase in prescribing of 5 per cent over the previous three years, with no simultaneous increase in infectious disease to explain it. They also show a tendency to prescribe newer drugs instead of older ones. However the AMM believe around 80 per cent of antimicrobial prescribing in United Kingdom general practice to be "fully justified" (Q 45). 2.4 The Association of the British Pharmaceutical Industry (ABPI) gave us a different set of figures. In 1996, United Kingdom GPs wrote 51m antibiotic prescriptions, which, though a lot, was 2.5m fewer than the previous year. They conclude, "The message about not prescribing antibiotics in diseases that are probably viral appears to be getting through" (p 177). 2.5 The Department of Health produce yet more figures, and analyse them in detail (p 343). Between 1991 and 1996 in England, the number of prescription items for antibacterials increased by only 7 per cent, and the net ingredient cost by only 4 per cent, both much less than the figures for all drugs; between 1995 and 1996 both figures went down, as noted by the ABPI. However there were wide variations from drug to drug. The group which gives "most cause for concern" is the fluoroquinolones: over the five years, use rose by 48 per cent, and cost by 81 per cent. "Ciprofloxacin is the market leader in a group of drugs which is heavily promoted". Use of penicillins rose by 13 per cent; the Department find this "disappointing", though they suggest possible innocent explanations.[14] Use of macrolides rose only marginally, but their cost rose by 58 per cent, probably because erythromycin, the original macrolide, lost market share to newer and more expensive macrolides with additional applications (e.g. azithromycin). "The newer macrolides are heavily promoted". The Department produce further figures to show that antibiotic prescribing varies widely between health authorities. 2.6 From his experience in Oxfordshire, Dr Richard Mayon-White, a Consultant in Communicable Disease Control, considers the antibiotic prescribing of United Kingdom GPs to be "conservative" (p 110). However local monitoring has revealed "wide variations" in prescribing of expensive drugs such as ciprofloxacin. Dr Hugh McGavock, Director of the Drug Utilisation Research Unit at The Queen's University of Belfast, detects a bell-curve (Q 660): "some doctors accepting the guidelines perfectly, the majority moderately well, and some doctors on the far side of the curve hardly paying any attention to them". The Royal College of General Practitioners (RCGP) stand by their profession: "Given the context of diagnostic uncertainty, current prescribing practice of GPs is in general more beneficial than harmful in the care of individual patients" (p 166). However, "There is sufficient evidence of widespread variation in the utilisation of antibiotics to suggest that there is scope for further reduction of their use by some practitioners". Dr Davey is more sceptical about GP prescribing. "We have intense debates about whether children with otitis media should receive antibiotics, or people with sore throats. But our work would suggest that the majority of patients to receive antibiotics...just have runny noses, where there is no evidence that they benefit" (Q 265). 2.7 There is no equivalent to PPA data for hospitals (DH p 342); all one can say for certain is that hospitals dispense a much smaller volume of antimicrobials than are prescribed in general practice. According to the Department of Health, 15-20 per cent of hospital expenditure on drugs goes on antimicrobials; and around one in-patient in four receives at least one course of antibiotics. The AMM produced, by different routes, two different "ballpark" figures: on the

one hand, 1-2m daily doses per year in English hospitals, or on the other the significantly higher figure of something under 5m in the whole United Kingdom (p 9). 20-30 per cent of hospital usage of antimicrobials is for prophylaxis against infection during surgery; and, according to the AMM, "courses given are often longer than necessary". Dr Davey also has doubts about current practice in this area (p 152; QQ 267, 272): "At the moment I do not think we have sufficient debate about what level of benefit [in terms of reduced risk from infection] justifies the use of antibiotics".[15] Towards more prudent use 2.8 The AMM believe that, by a significant and sustained campaign of education, it might be possible to eliminate the "small proportion" of United Kingdom GP prescribing which they believe to be completely unjustified, and to reduce the larger proportion which "is perhaps not founded on the best evidence-based practice but may be justified by medical, cultural or psychological reason". They warned us, "The continuing legitimate use of antibiotics in humans may still sustain and might even increase the amount of resistance"[16]; to say nothing of resistance continuing to be generated in animals, and in humans in other countries where imprudent use persists. However, "the medical profession must put its own house in order before it can expect others to do so" (pp 8-13). On the other hand Dr Davey, who believes that more than half of GP prescribing is justified by nothing more than a runny nose, would happily "let the people with the sore throats get antibiotics but concentrate on the people who do not have any clinical signs which warrant antibiotics" (Q 265). 2.9 These contrasting positions present the question, What constitutes prudence? The RCGP insist that, although doctors have a national, strategic responsibility for public health, this cannot override their primary responsibility to the individual patient (p 167, Q 280). Dr Davey put it more bluntly: "I would much rather...some people received unnecessary treatment than we end up with somebody dying" (Q 268). Dr McGavock was equally blunt on the other side of the dilemma: "[Over-prescription] is a situation that really must be changed if we wish to preserve the antimicrobial era...it may well go on, but if it does our grandchildren will curse us for wasting this limited human resource" (Q 647). Witnesses have pointed to various ways to resolve this dilemma; we consider them below. Formularies, policies and guidelines 2.10 A formulary is a list of available drugs, or of drugs recommended from among those available; a policy gives guidance or instruction on when and how they are to be used. In the United Kingdom, most hospitals and some general practices have at least a formulary of antibacterials recommended for local use, and some hospitals have antibacterial policies. 2.11 The PHLS presented us with the findings of a survey of clinical audit in hospitals in England and Wales, in the context of MRSA (p 41). They found that formularies vary widely in form and content, and "are drawn up with little involvement of the junior staff; comments are often not invited"; that communication of antibiotic policies to new staff is often poor; and that audit of prescribing is often irregular, infrequent and unstructured. The AMM recommend that all hospitals and general practices should have both formularies and policies. These should be produced in a way which gives the doctor "ownership"; and they should be supported by strong encouragement from the health departments, and by audit.

2.12 Policies may operate at any level, from local to global (RCGP p 167; DH p 344; Petrie Q 669). They may reflect the threat of resistance by recommending rotation or combination of drugs (Davey Q 263, Spencer p 519, Tyrell p 528), or by avoiding certain drugs altogether (PHLS p 52). They may be supported by restrictive reporting,[17] by financial incentives (recommended by Dr McGavock, Q 658) and even by compulsion (resisted by Professor Petrie, Q 674). They may be more or less flexible; Glaxo Wellcome advocate flexibility, "to enable nimble local responses to changes in resistance patterns" (p 407). Box 3 gives examples of good practice in this area which have come to our notice.
Box 3

FORMULARIES, POLICIES AND GUIDELINES: EXAMPLES OF GOOD PRACTICE

Grampian Formulary

The Grampian Formulary has broad ownership in both hospital and general practice and vigorous monitoring by ward pharmacists and Health Board Prescribing Advisers. Professor Petrie claims that it achieves GP concurrence or compliance of 90-96 per cent (QQ 648, 672, 684). It is commended by the Scottish Microbiology Association (p 471).

Scottish Intercollegiate Guidelines Network (SIGN)

SIGN (Q 663) is a wide group of professionals and others who are working together to produce evidence-based guidelines for a range of conditions. There are formal procedures for proposing a guideline, prioritising projects, literature reviews, grading of evidence and recommendations, and critical appraisal. 21 guidelines have been published, on paper and on the Internet, and 48 are in preparation; infectious diseases are rising up SIGN's list of priorities. Putting the guidelines into practice has involved groups of GPs, associations of patients, and newspapers; and the Royal Colleges and the Scottish Council for Postgraduate Dental and Medical Education use them in the process of accreditation for training. There are arrangements for audit and feedback. Professor Petrie emphasized the importance of local "ownership", and of implementation; "simply sending out a guideline is a waste of time" (Q 664).

PRODIGY

PRODIGY (Prescribing Rationally with Decision Support in General Practice) is a project to develop an electronic guideline (or "decision support") system, intended for use on a GP's desktop computer. Dr Davey commended PRODIGY (Q 250). So did the RCGP (p 167, Q 290); they told us that PRODIGY is being piloted in 200 practices with funding from the Department of Health, and that other similar systems are under development. The Department observe that guidance delivered through PRODIGY is "adjusted to local conditions to reinforce local policies"; and that computer prescribing systems are in use in some hospitals (p 344).

2.13 A policy is not a panacea. It must be sound in itself, and must be implemented conscientiously and intelligently. It must also be policed: see below, paragraph 2.34. Dr R C Spencer, of the Public Health Laboratory at Bristol Royal Infirmary, points out that policies depend on surveillance, to indicate what the infecting organism is most likely to be, and what resistances it is likely to exhibit (p 513). We consider surveillance in Chapter 5. As for implementation, Professor Percival gave us the example of a policy drawn up for severe community-acquired pneumonia; it was applied to pneumonia of all kinds, and is now blamed for "a tremendous increase in Clostridium difficile side-effects" (Q 104). 2.14 The pharmaceutical industry strongly prefer evidence-based guidelines for appropriate prescribing, rather than crude injunctions to prescribe less. In some situations, they say, certain drugs should be withdrawn, in others they should be used in combination (particularly for TB and HIV) or rotated; in others new and better agents should be used. Guidelines must be based on evidence, which the industry is willing to provide (see Box 11 below). Glaxo Wellcome and SmithKline Beecham make no bones about it: if use of their anti-infective products were restricted beyond a certain point, they would place their investments somewhere else.[18] 2.15 In the USA, one of the pressures towards imprudent prescribing is fear of litigation. In United Kingdom law, a doctor who prescribes in accordance with a local policy is unlikely to be successfully sued for negligence (Davey p 156). Professor James Petrie of the University of Aberdeen described the British legal position in this area as "reassuring" (Q 674). Dr McGavock expressed himself more concerned about being sued for prescribing too much than for prescribing too little (Q 642). Rapid testing 2.16 Susceptibility testing by the standard methods takes 48 hours; the whole process, from the doctor taking the specimen to receiving the result, may take longer. In the meantime, the doctor must prescribe empirically, and may prescribe inappropriately. Many witnesses have put it to us that the cause of prudent use would be much advanced by more rapid testing. According to the ABPI, rapid testing for routine infections is on the way. "People are working on it, and it

will become available in due course"perhaps in 5-10 years (Q 341). SmithKline Beecham call on the Government to support technology development, possibly through the EU (p 485). 2.17 The PHLS agreed that faster testing would be helpful (Q 109). "Genotypic" tests, which use the polymerase chain reaction (PCR) to examine the isolate's DNA, are already in use. However, these "will not replace the existing methods entirely", because they are expensive, and because they answer only one question at a time. Professor Roger Finch of the University of Nottingham believes that the expense of genotypic testing for the wide range of infections encountered in general practice will be prohibitive (Q 385). Dr Davey sounded a further note of caution (QQ 269-271): even if rapid analysis were available, it would say only that a certain organism was present, not that it was necessarily the cause of the disease. 2.18 Even without faster tests, the process could be speeded up if results could be reported electronically. The RCGP told us that systems to do this exist already (Q 294). The recent NHS White Paper The new NHS (Cm 3807) indicates that one immediate application envisaged for the projected "NHSnet", linking all hospitals and GP practices, is to be transmission of test results (paragraph 1.12). Prescription checking and control 2.19 A hospital formulary may incorporate controls, whereby certain agents may be prescribed only with certain levels of authority. According to the AMM, "Recent evidence from the USA shows that it is possible by considerable effort (prior approval from an infection specialist for the use of an antibacterial from a restricted list) to influence prescribing without adversely affecting clinical outcomes and with improvements in the sensitivities of bacteria" (p 10). 2.20 According to the Department of Health (p 344), although routine prescribing of antimicrobials in hospitals is done by junior doctors, "Access to non-routine agents is restricted through the hospital pharmacy", and in most hospitals pharmacists visit the wards every day to check prescriptions and advise.[19] However the PHLS clinical audit project found that practice in this respect, in hospitals in England and Wales, "varied greatly"; and Dr Davey, as a hospital consultant, admitted that prescribing by junior doctors at night is often not reviewed by senior staff in the morning (QQ 245, 268). He suggested that there might be a role here for senior nurses. 2.21 Senior practitioners whom we met in Boston disagreed as to the appropriate level of control. Dr Sherwood Gorbach favoured requiring that every prescription for a drug associated with a resistance problem be accompanied by a "chit" giving the reason for prescribing. Dr Anton Medeiros considered that this would restrict professional freedom to a degree unacceptable in the USA; he believed that, if surveillance was thorough and its findings were properly communicated, doctors would moderate their practice voluntarily. See Appendix 6. 2.22 In general practice, there is a long and strong tradition of clinical freedom and responsibility. Dr McGavock considered (QQ 650, 657-660, and p 300) that liberty has degenerated into licence, and that the seriousness of the threat to the effectiveness of antimicrobials justifies extreme measures. He recommended that GPs should be prohibited from prescribing specific new, expensive, broad-spectrum antibiotics without first receiving microbiological advice. Professor Petrie considered this to be unaffordable and impractical. Dr

McGavock acknowledged that it would be expensive, but insisted that it was a price worth paying to prolong the "antimicrobial era". Pharmaceutical licensing 2.23 It would in principle be possible for a licensing authority, considering a new application, to turn it down on the basis that the new drug, though effective, was no more so than drugs already on the market, and was more likely to induce resistance. This is not however the usual approach. Dr Ross Taylor of the RCGP told us (Q 283), "I think the fundamental problem is that in this country medicines are licensed on the basis of effectiveness, not comparative effectiveness. So, if an antibiotic is licensed, a company can quite legitimately...promote that medicine...whether or not there is another antibiotic already available that might be better". Similarly, it would in principle be possible for a licensing authority to modify or withdraw a licence if resistance induced by the drug rose to a certain level (Finch p 189). Dr McGavock recommended that certain antibiotics should be licensed for hospital use only, so that GPs could not use them at all (p 300). Pharmaceutical salesmanship 2.24 The AMM blame the increase in prescribing of antimicrobials, and the tendency to prescribe more expensive drugs where cheaper ones would do, on advertising and salesmanship by the pharmaceutical industry. The comments of the Department of Health on the increased use of ciprofloxacin and the more expensive macrolides, noted above, suggest that they agree though in oral evidence, the Chief Medical Officer deprecated any suggestion that the industry went beyond reasonable bounds, and said that increased use of macrolides may have been due to a new indication for use (Q 789). Professor Petrie gave a startling example of promotion which he regarded as excessive: provision by a company to a community on-call service of free "starter packs" of the antibiotics co-amoxiclav (Q 654) and clarithromycin (Q 657). The AMM call for "stricter controls" to restrain "over-zealous promotion"; so did Dr Rosamund Williams of WHO (Q 127). 2.25 Dr Mayon-White considers pharmaceutical salesmanship an "important influence" on GP prescribing, but not necessarily a bad one: "it can be a collaborative process in setting policies and getting educational messages across" (Q 158). For this to happen, he says, health authorities and doctors must be "in control of the process". The RCGP likewise declined to condemn pharmaceutical marketing outright (Q 282); they consider it necessary and useful, and believe that GPs are able to retain control. The ABPI observe that their members' marketing is governed by a code of practice administered by an independent body; they consider that the pressure on GPs to prescribe comes more from patients (see below) and from lack of time than from the industry (Q 333).
13 What we have to say about doctors should be taken to apply mutatis mutandis to dentists. Veterinary practice is discussed separately in the next chapter. Back

14 Demographic factors, or an increase in repeat treatments due to rising resistance. Back 15 Cp BMA: "Antibiotic misuse is common and studies have suggested that up to 70 per cent of treatment courses are unnecessary or inappropriate. Therapy is often unnecessarily prolonged

and prophylaxis is often inappropriate or given at the wrong time" (p 381). The National Committee for Microbiology also express concern in this area (p 541). Back 16 Cp Glaxo Wellcome: "The recent lessons learnt from anti-retroviral chemotherapy have demonstrated the very powerful effects that natural selection can have, even when prescribing is entirely appropriate and patients are highly motivated to comply with treatment" (p 407). Back 17 Restrictive reporting: When a doctor sends a specimen to a laboratory for microbiological analysis and susceptibility testing, the microbiologist must decide what to report. Reporting every bacterial isolate would be unhelpful, since it would include normal flora and contaminants; the doctor is interested only in possible pathogens. Reporting the susceptibility of each isolate to every conceivable antimicrobial would be impossible; in practice, only certain susceptibilities are tested or reported. In the United Kingdom, microbiologists turn these facts of life to advantage. According to the AMM, "Reporting is tailored to what is felt to be appropriate for the individual patient and the wider context, and is often made to be concordant with the hospital's formulary and antibacterial policies". Back 18 GW: "The outcome will be a "deprioritisation" of this area in favour of others" (p 407). SKB: "Inappropriate restrictions...would of necessity cause companies to invest their research effort in other, more profitable, therapeutic areas" (p 475). Back 19 The Hammersmith Hospital has appointed a specialist clinical pharmacist working exclusively on antimicrobials. This has led to reduced infection, and annual savings of 77,000 (DH p 345). On the general role of the hospital pharmacist, see the evidence of the Royal Pharmaceutical Society, p 461. Back 2.26 In the United Kingdom at present, generally speaking, systemic antibiotics (i.e. taken internally, as opposed to "topical") are licensed as "POM": i.e. prescription-only medicines, available only on prescription from a doctor, dentist or veterinary surgeon (not from a nurseprescriber). There is however a general trend to deregulate medicines, moving them from POM to "P" (pharmacy); and there is discussion of treating antibiotics in this way. 2.27 The RCGP believe that there is pressure for OTC antibiotics from consumers and from industry. However they have no doubt that OTC availability would mean more use, and they consider it possible that this would mean more resistance. Therefore, "We do not think it is a good idea" (Q 306).[20] Neither does Professor Petrie (Q 659). The ABPI are "generally opposed" to OTC antibiotics, with the possible exception of treatments for uncomplicated lower urinary tract infection (cystitis) (p 177, Q 338). They object because pharmacists do not have the necessary training or suitable premises for the confidential consultations which would sometimes be needed to advise customers on the right choice of medicine, nor access to the microbiology services which would sometimes be needed for diagnosis; and because the imprudent use which contributes to resistance would probably increase. 2.28 Professor Finch (p 186, QQ 364-376), who is a member of the Committee on Safety of Medicines and co-chairman of a BSAC working party on OTC antibiotics, told us that the pressure to deregulate the supply of antibiotics comes not just from industry, but from the regulators themselves, in particular at EU level. He explained that what is under consideration is not a free-for-all, but OTC supply of single doses or short courses for particular indications. In his view, much depends on what drugs are to be deregulated, and for what conditions; he too identified cystitis as one which "seems to be reasonable at first view", along with minor

infections of the skin or eye.[21] It would be necessary to give pharmacists robust guidelines, possibly access to medical records, and perhaps a surveillance role. OTC antibiotics would make commercial sense only if they resulted in increased use; but this would not necessarily mean more resistance, and might even mean less, if P medicines with a lower tendency to induce resistance took market share from POM drugs with a higher tendency. He concludes that more research and consideration is needed before any major change. 2.29 One systemic antimicrobial is already available OTC in the United Kingdom: the antifungal fluconazole, sold in a one-capsule course for Candida vaginitis. In Spain and Greece fluconazole is available OTC, and seems to have given rise to significant resistance. According to Professor Finch, "In the United Kingdom there is no evidence to date that the use in the community of a single capsule for Candida vaginitis is associated with resistance" (Q 375); but this may be because no survey has been undertaken. According to Dr David Denning of the University of Manchester (p 402), fluconazole-resistance in Candida is now common, but the causes are "not known entirely"; any effects of OTC supply are "not being studied". 2.30 The Department of Health drew to our attention the concern of the Medical Devices Agency about wound dressings which incorporate antibiotics (p 345). Under EU law and the Medical Devices Regulations, there is no provision for such dressings to be POM, and they have therefore been available OTC since 1995. The Agency are concerned that uncontrolled use will give rise to resistance. However they have no evidence yet that this is happening; and they have received little support from other Member States "because many antibiotics are already available OTC in a number of EU countries and there seems to be general acceptance of this practice". As to actual OTC antibiotics, however, the Department are confident that they can hold the line. The Chief Medical Officer, Sir Kenneth Calman, has raised the issue with his EU counterparts (Q 757); he considers a ministerial decision to permit OTC availability "very unlikely" (Q 794). Under EU law, any medicine may be confined to the POM category if it presents "danger to health" if used without supervision; and Sir Kenneth assured us that "health" for this purpose included public health (p 371). Medical education: undergraduate, postgraduate and vocational 2.31 The undergraduate medical curriculum is crowded, and several witnesses told us that it tends to devote little time to antimicrobial therapy (e.g. Petrie, McGavock Q 681though Professor Finch disagrees, p 187).[22] This would be understandable, since oral antimicrobials are relatively easy to prescribe, being relatively non-toxic and unlikely to harm the patient directly. As the AMM put it, "antibacterials are victims of their own success". They point out that "antimicrobials are the only class of drugs the prescription of which can have adverse consequences outside individual recipients"i.e. the selection of resistant strains. They recommend, "priority should be given by the medical profession, universities and the General Medical Council to ensuring that a definitive slot on antibacterial use is in all curricula, and that this includes not only technicalities of antibacterials but puts their use into sociological and world contexts" (p 10). Implementing this recommendation would of course involve taking time from other subjects, whose advocates would no doubt make their case with equal vigour. Continuing professional development 2.32 In medicine as in other walks of life, one of the most effective forms of professional development is participation in teaching. In Oxfordshire, Dr Mayon-White reports, "the teaching and training practices tend to prescribe less of the expensive antibioticsand indeed

fewer antibiotics overallthan the non-teaching, non-training practices" (Q 158, cp Davey p 155 and Q 250). In the context of prescribing, another effective form of professional development is participation in the process of creating local formularies and policies (Petrie Q 679). 2.33 Not every doctor can be a trainer or policy-maker; and various ways have been found to deliver professional development to the wider medical community. Dr Jeremy Grimshaw of the Health Services Research Unit at the University of Aberdeen gave us an overview drawn from the findings of the Cochrane Collaboration on Effective Professional Practice, of which he is the co-ordinating editor (Q 672). He indicated that there are no "magic bullets"; which interventions are most effective at changing behaviour depends on the behaviour in question, and on the context (e.g. hospital or general practice), and best results are obtained by intervening in several ways at once. Dr Grimshaw's findings may be crudely summarised as follows:
(i) Passive dissemination, using literature and lectures: in itself, ineffective;

(ii) Interactive workshops: more effective; (iii) Audit and feedback: effective, though the changes achieved may be modest; (iv) Educational outreach, whereby trainers visit professionals individually: "very promising"; (v) Local opinion leaders raising standards by example: "much touted", but requires further research (Q 683). We have received further evidence about (iii) and (iv), as follows. Audit and feedback 2.34 GPs may already opt to receive detailed information ("PACT" data) on their own prescribing from the PPA; and some may have access to additional information from their own IT system. The AMM, who recommend wider use of formularies and policies, call for these to be supported by audit (p 10).[23] They acknowledge that audit is especially difficult in the isolated conditions of general practice; on the other hand, GPs are ahead of hospitals in computerising their records of diagnosis and prescription. They acknowledge that the cost to the NHS of the staff and IT required for a proper system of audit would be high; and they assess the impact, even after 5-10 years, as "Some effect, but not major". Professor Petrie is an advocate of audit (Q 663): "By doing audit of what people are doing, you can get the `outliers' and bring them into the middle group of prescribers voluntarily. If you start going out with guns and statutory controls, people hide." Box 4 gives examples of good practice in this area which have come to our notice.
Box 4

PRESCRIBING AUDIT AND FEEDBACK: EXAMPLES OF GOOD PRACTICE

Northern IrelandCOMPASS

Dr McGavock told us about COMPASSComputerised On-line Monthly Prescribing Analysed for Science and Stewardship (p 281, Q 635). COMPASS is an evidencebased, computerised prescribing interrogation system, run by the Drug Utilisation Research Unit of The Queen's University, Belfast. "COMPASS compares each practice's prescribing every month in Northern Ireland against best practice and it then prints a clear report showing the ways in which the doctors did prescribe and recommends changes to improve the quality of their prescribing". For a cost of 2 per copy, COMPASS typically identifies possible savings of 15 per cent. "COMPASS is taken to every practice annually by the Area Prescribing Advisers, for detailed discussion, but fundholding practices often request this document quarterly, to drive their cost-effectiveness efforts. COMPASS has saved over 11m in Northern Ireland in the past three years, but could save up to 25m annually, if fully utilised, with a striking improvement in the quality of medical treatment". In the first quarter of 1997, 60 non-fund-holding GPs saved 1.23m using COMPASS, backed up with lectures and visits. Dr McGavock finds the results of COMPASS overall "disappointing". He believes that, if it is to change behaviour significantly, such a system requires to be backed up by educational outreach, and by some direct "incentive", e.g. such that a proportion of the money saved on drugs is returned to the practice (QQ 638, 684).

Oxfordshire

Dr Mayon-White told us (p 111) that Oxfordshire Health Authority has begun a programme to improve antibiotic use in general practice. Guidelines have been issued to GPs, recommending first-choice empirical treatments for various infections. The guidelines are supported by prescription monitoring; results are fed back to GPs in general by means of a newsletter, and to individual practices through visits by the Authority's Medical Adviser.

Educational outreach 2.35 Dr Grimshaw defined educational outreach thus: "This is where you have a professional, often a pharmacist, going to visit a general practice or hospital to give a number of very selective messages about good prescribing behaviour, which often use the marketing

techniques of the pharmaceutical industry, to try and identify the specific barriers to the behaviour they want to happen, and modify their message based on these barriers and reinforce that message throughout that contact". He mentioned that the Department of Health is currently funding a large-scale trial of this method, expected to report in 1999 (QQ 672, 683). The USA is ahead of the United Kingdom in this field: see Box 5.
Box 5

EDUCATIONAL OUTREACH IN THE USA

Dr Jerry Avorn of Harvard Medical School is a world expert on educational outreach. His approach is modelled on that of the pharmaceutical industry. He began with focus groups of physicians. These revealed two groups of doctors: some who overprescribe out of ignorance; and others who consciously overprescribe in order to satisfy their patients. For the second group, like Dr Schwartz, he provides "paper placebos". For the first, he sends out "academic detailmen": pharmacists from the medical school who meet physicians one-to-one, on the same basis as salesmen, to talk about prudent prescribing. He has shown that every $1 spent on these actions saves $2 on the drugs bill. His approach has been taken up in various places around the USA; similar approaches have been tried in various parts of the United Kingdom, and adopted nationwide in Australia. He acknowledged that some doctors require to be persuaded that prudent use is not just a euphemism for cutting costs at the expense of patient care.

Dr Ben Schwartz, of the Centers for Communicable Disease Control and Prevention (CDC) in Atlanta, has worked on educating community physicians and their patients, with a view to controlling the rise of penicillin-resistant pneumococcus. In focus groups, physicians acknowledged overusing antibiotics by as much as 50 per cent. They blamed pressure from patients, and shortage of consultation time: it is quicker to prescribe, than to explain why a prescription would be inappropriate. Dr Schwartz has therefore produced the following aids for physicians: professional information sheets; a simple patient information leaflet for the waiting room, explaining that unnecessary antibiotics are bad for the patient; a "non-prescription" form; question-and-answer sheets for parents; and a letter for parents to give to their child-carer. Pilot projects are now under way in five States; CDC is equipping the local health department to train senior doctors to disseminate the concepts and materials to their peers. Evaluation will show whether these approaches reduce inappropriate use, and whether this in turn affects the level of resistance.

Educating the public 2.36 One major factor affecting the prescribing behaviour of GPs is the expectations of their patients[24]or GPs' perception of patients' expectations, which is not necessarily the same thing.[25] Patients' expectations are not uniform; according to Dr Grimshaw, they "vary across different areas, socio-economic groups and cultures" (p 301). 2.37 It may not be the case that people in poorer communities have a higher expectation of receiving antibiotics. Rather, it may be that GPs in such communities are more pressed for time, and therefore inclined to prescribe rather than explain. However a GP under pressure of time may do various things to terminate the consultation besides prescribing. First, he may simply say No: Dr Grimshaw cited evidence from a randomised trial that this reduces the likelihood that the patient will attend again with the same complaint (Q 674). Saying No is not always easy: Dr Grimshaw recommended training for GPs in communication skills (p 301). Dr Davey mentioned two alternative approaches (Q 273): give the patient an information leaflet; or say, "I do not think you need an antibiotic, but if you want to get a prescription you can come back any time, you do not have to make another appointment". Of the latter approach, he commented, "Most of the people who were given that option did not come and get the antibiotics". Dr Taylor of the RCGP mentioned another: a delayed-action prescription, for use if symptoms persist. "The result of that is that many patients do not in fact take the prescription" (Q 303); this strategy is also known to reduce the rate of follow-up appointments. 2.38 Rather than waiting until the patient is in the surgery, the professions may take the message of prudent use out to the public at large. The AMM recommend that the health and education departments, the health professions and the media should all do more to convey the message (p 11). Dr Mayon-White similarly calls for public education; he recommends that health authorities' health education departments should prepare material for schools, while adult education should be delivered through the media (QQ 159-163). 2.39 The RCGP observe that GPs work within, and are to an extent constrained by, a "cultural framework" (Q 281). A GP who unilaterally defies this framework may simply lose his patients. However the culture can be changed, as in the cases of barbiturates and amphetamines. Changing the culture requires public education and consensus-building, based on evidence. In this matter, they feel, the evidence will have to be compelling: "Whether or not, and how far, antibiotic use for relief of symptoms of self-limiting illness should be limited should not be a matter for doctors to decide, but the subject of public policy. In a context in which antibiotics are much more freely used in agriculture and food production, it seems unlikely that there would be much public support for such restriction...Antibiotic use in the United Kingdom is already lower than in many other European countries; in that context it would be difficult to operate a more restrictive policy" (p 167). The RCGP observe that typically the sort of patients who are most insistent on antibiotics belong to the social groups who are hardest to reach with educational material (Q 299, cp Petrie Q 677). Are antibiotics bad for you? 2.40 Persuading the public of the case for reduced use of antibiotics will be much easier if it can be shown that unnecessary use carries risks not only to public health in general, but also to the particular patient under treatment.[26]

2.41 Professor Finch said (in the context of OTC antibiotics), "It is uncommon for resistance to arise in an individual receiving an antibiotic and for this to cause him/her harm" (p 189). (TB is an exception, as are viruses such as HIV.) However he went on, "Agents can affect the susceptibility of the bacteria which make up the normal flora of the skin and gastrointestinal tract. This in turn could give rise to subsequent infection in an individual". Dr Ben Schwartz of the US Centers for Communicable Disease Control and Prevention (CDC) and Dr Michael Bennish of Boston both believe they can show that previous treatment with antibiotics is a risk factor for infection with resistant strains. Dr McGavock cited evidence that, in cases of otitis media, withholding antibiotics for two to three days reduced the rate of recurrence by four fifths (QQ 659, 677). 2.42 Dr Davey also has evidence to support this proposition. "Patients with resistant organisms are more likely to have received prior antibiotic therapy than are controls [i.e. patients with susceptible strains of the same organism]"though, he admits, "Antibiotic use may just be a marker for patients who are more ill" (p 145). Speaking in the context of the successful campaign in Iceland to bear down on penicillin-resistant pneumococci, he added, "The message that we need to get across is that most of the bacteria that live in our bodies do not do us any harm, and if you eliminate them with antibiotics then you allow the bad guys in" (Q 262; cp Q 236, pp 146, 154). He went so far as to say, "Germs are good for you...germs are part of your environment" (Q 274); he admitted that this is "probably not something that people understand". 2.43 The ABPI observe that effects on the gut flora are checked as part of the process of licensing a new antimicrobial. "Whilst it is true that there are occasional changes in gut flora which are limited to the duration of treatment, generally speaking you find that the gut flora returns to normal quickly after the antibiotics are stopped" (Q 331). 2.44 If antibiotics do even slight collateral damage, then it would plainly be better not to use them in situations where they can do no good (e.g. the minor viral infections which Dr Davey believes account for more than half of GPs' prescriptionsQ 265). More difficult are situations where antibiotics do a little good, which must then be weighed against the possibility of harm. The RCGP draw attention to the major grey area in general practice: the use of antibiotics to relieve the symptoms of self-limiting illness such as sore throat, bronchitis and otitis media. They concede that there may be a case for limiting use in such cases on public health grounds, but they insist that such use does have direct benefit (p 166). The ABPI, on the other hand, regard prescribing antibiotics for "the majority of sore throats", or for "a banal, self-limiting, mild condition, e.g. an upper respiratory tract infection", as "irresponsible" (p 176). 2.45 Tessa Jowell MP, the Minister for Public Health, was firmly of the view that the public must not be warned off antibiotics. For her, the issue crystallised around meningitis: nothing must be said to deter parents, in particular, from contacting their doctor at once if they suspected meningitis in their child. She acknowledged that it was difficult to convey the message of appropriate use of antibiotics in a way which was balanced and not confusing; but, she insisted, "We do not want patients to think that antimicrobials are dangerous for the individual...; they are among the safest medicines" (Q 759). The Chief Medical Officer suggested that the problem could be addressed by acknowledging that what was appropriate varied from case to case: for instance, what was appropriate for a young child with a fever was not necessarily appropriate for a middle-aged man with a cold (Q 761). Compliance

2.46 It is usual practice for patients receiving antibiotics to be instructed to "complete the course". Yet the quality of evidence on which the recommended duration of antibiotic treatment is based varies, in fact, greatly from case to case. For some infections, of which tuberculosis is the most notable example, the type and duration of treatment needed to cure the patient (and to prevent resistance in those few patients who are not cured) has been well authenticated by numerous controlled trials. For others, including the many respiratory infections, such as bronchitis, pneumonia, sinusitis and otitis media, which constitute the most frequent occasions of antibiotic use, the optimum duration of courses of treatment is still surprisingly ill-founded. International comparisons reveal these uncertainties vividly, particularly in otitis media, bacterial endocarditis and urinary tract infection (Griffin p 548). There is certainly no virtue in completing the course if the infection was not in fact present in the first place (Davey Q 268), or if the prescribed course was longer than necessary to cure the infection and completing the course only prolongs the selective pressure on the commensal flora. 2.47 Whether or not the particular recommendations for duration of treatment are wellfounded, however, it is clear that compliance with prescribing recommendations is hard to achieve and that non-compliance with treatment can contribute to resistance, especially in the case of tuberculosis and sexually-transmitted infections. The Royal Pharmaceutical Society has addressed this problem in a valuable document entitled From Compliance to Concordance (March 1997). The RCGP observe that compliance can be encouraged by drug and regimen design, and by patient information, on which there is recent EU legislation which they support (p 167). The ABPI are "particularly keen" on improving compliance by means of new formulations (p 176, Q 347); and they support the Patient Pack Initiative on patient information.
20 The RCGP's memorandum (p 167) said, "In principle...we would not object to the direct sale of certain antibiotics". They admitted under oral examination that they would object, up to the point where the Medicines Control Agency licensed antibiotics OTC, at which point "We would have to go along with that" (Q 306). Back

21 Cp Royal Pharmaceutical Society p 463. Back 22 Professor David Greenwood, of Nottingham University Hospital, believes that his department "may be unique" in giving third-year medical students a two-week module on antimicrobial therapy (p 410). Back 23 So do the British Pharmacological Society (p 386). Back 24 For eloquent accounts of this problem from the front line, see the evidence of the Osborne Practice in Southsea (p 440) and Dr John Sterland (p 526). Back 25 "Doctors tend to overestimate the patient's desire for a prescription"-RCGP Q 297, cp Grimshaw Q 674. Back 26 Like all medicines, antibiotics carry risks of direct adverse side-effects in the individual. Although antibiotics are generally safe medicines, unwanted effects do occur, mostly trivial but occasionally life-threatening. Our particular concern, however, is the relationship of antibiotic treatment to the development of resistance. Back